New Caged Coumarin Fluorophores with Extraordinary Uncaging Cross Sections Suitable for Biological Imaging Applications

Article abstract of DOI:10.1021/ja036958m

Photocaged fluorescent molecules are important research tools for tracking molecular dynamics with high spatiotemporal resolution in biological systems. We have designed and synthesized a new class of caged coumarin fluorophores. These coumarin cages disp

Full text of DOI:10.1021/ja036958m

Published on Web 03/20/2004
New Caged Coumarin Fluorophores with Extraordinary
Uncaging Cross Sections Suitable for Biological Imaging
Applications
YuRui Zhao,† Quan Zheng,†,‡ Kenneth Dakin,† Ke Xu,† Manuel L. Martinez,§ and
Wen-Hong Li*,†
Contribution from the Departments of Cell Biology and Biochemistry and Physiology,
UniVersity of Texas Southwestern Medical Center at Dallas, Dallas, Texas 75390, and
Department of Polymer Science and Engineering, UniVersity of Science and
Technology of China, Hefei, Anhui, P. R. China
Received June 28, 2003; E-mail: wen-hong.li@utsouthwestern.edu
Abstract: Photocaged fluorescent molecules are important research tools for tracking molecular dynamics
with high spatiotemporal resolution in biological systems. We have designed and synthesized a new class
of caged coumarin fluorophores. These coumarin cages displayed more than 200-fold fluorescence
enhancement after UV photolysis. Remarkably, the uncaging cross section of a 1-(2-nitrophenyl)ethyl (NPE)-
caged coumarin is 6600 at wavelength of 365 nm, about 2 orders of magnitude higher than previously
described caged fluorophores. Product analysis of the photolytic reaction showed clean conversion of NPE-
caged coumarin to 2-nitrosoacetophenone and the parent coumarin, suggesting that the mechanism of
the photolysis follows the known photochemical reaction pathway of the 2-nitrobenzyl group. We have
also measured the two-photon uncaging cross sections of NPE-caged coumarins 2a and 5 at 740 nm to
be near 1 Goeppert-Mayer (GM). The mechanistic study, together with the two-photon uncaging data,
suggested that the coumarin moiety serves as an antenna to enhance the light harvesting efficiency of the
coumarin cage and that the photonic energy absorbed by coumarin was utilized efficiently to photolyze the
NPE group. Future explorations of this type of “substrate-assisted photolysis” may yield other cages of
high uncaging cross sections. For cellular imaging applications, we prepared a cell permeable and caged
coumarin fluorophore, NPE-HCCC2/AM (10), which can be loaded into fully intact cells to high concentra-
tions. Initial tests of this probe in a number of cultured mammalian cells showed desired properties for the
in vivo imaging applications. The combined advantages of robust fluorescence contrast enhancement,
remarkably high uncaging cross sections, noninvasive cellular delivery, and flexible chemistry for
bioconjugations should generate broad applications of these caged coumarins in biochemical and biological
research.
1. Introduction
Desired properties of caged fluorophores for cellular imaging
applications should include high uncaging cross sections, robust
fluorescence enhancement after uncaging, and flexible chemistry
for bioconjugation and cellular delivery. The uncaging cross
section measures the efficiency of photolysis. It equals the
product of the quantum yield of uncaging (Qu) and the extinction
coefficient (ꢀ) of the molecule at the wavelength of photolysis
(preferably above 350 nm for imaging applications in live cells
using common optics). Thus far the reported caged fluorophores
have uncaging cross sections on the order of a hundred when
photolyzed at 350 nm or above.5 New caged fluorescent
molecules with higher uncaging cross sections are desirable
because we can efficiently turn on their fluorescence while
minimizing side effects of UV illumination on live specimens.
Photocaged fluorescent dyes have wide applications in
tracking the spatiotemporal dynamics of molecular movements
in biological systems.1-4 These caged tracers are weakly or
nonfluorescent when key functional groups of fluorophores are
masked by photolabile protecting groups (cages). Photoactiva-
tion with ultraviolet (UV) light removes the protecting group
(uncage) and abruptly switches on the fluorescence of parent
dyes.
† Departments of Cell Biology and of Biochemistry, University of Texas
Southwestern Medical Center at Dallas.
‡ University of Science and Technology of China.
§ Department of Physiology, University of Texas Southwestern Medical
Center at Dallas.
(1) Adams, S. R.; Tsien, R. Y. Annu. ReV. Physiol. 1993, 55, 755.
(2) Politz, J. C. Trends Cell Biol. 1999, 9, 284.
Another key requirement for caged fluorophores is that parent
fluorophores should be reasonably photostable to resist photo-
(3) Kao, J. P. Y.; Adams, S. R. In Optical Microscopy: Emerging Methods
and Applications; Herman, B., Lemasters, J. J., Eds.; Academic Press: San
Diego, CA, 1993; p 27.
(4) Mitchison, T. J.; Sawin, K. E.; Theriot, J. A.; Gee, K.; Mallavarapu, A.
Methods Enzymol. 1998, 291, 63.
(5) Krafft, G. A.; Sutton, W. R.; Cummings, R. T. J. Am. Chem. Soc. 1988,
110, 301.
9
10.1021/ja036958m CCC: $27.50 © 2004 American Chemical Society
J. AM. CHEM. SOC. 2004, 126, 4653-4663
4653

A R T I C L E S
Zhao et al.
Figure 1. Design and syntheses of caged coumarins. (a) Zn(CN)2, HCl(g). Et2O, 0 °C, and then Et2O, H2O, 100 °C, 22%. (b) Malonic acid, aniline (cat),
pyridine, 44%. (c) (1) SOCl2, cat DMF, CH2Cl2. (2) H-D-Glu(OMe)-OMe (for 1a) or H-D-Glu(OtBu)-OtBu (for 1b), TEA, CH2Cl2, 67%. (d) NPE bromide
(for 2a,d) or NB bromide (for 2b) or DMNB bromide (for 2c), DIEA, CH2Cl2, ∼65-75%. (e) (1) 2d, TFA, Et3SiH, CH2Cl2. (2) AcOCH2Br, DIEA, CH3CN,
60%. (f) Hydrolysis by cell esterases and UV uncaging.
bleaching. A major drawback of existing caged fluorophores
NPE group can be efficiently photolyzed with infrared light
illumination. We have also designed and synthesized a cell
permeable derivative of a coumarin cage, NPE-HCCC2/AM,
that can be loaded into cells noninvasively. Cellular imaging
of this probe demonstrated a combination of advantages.
6
such as caged fluorescein5 or caged resorufin is the rapid
photobleaching of parent dyes after photoactivation.4 Photo-
bleaching introduces uncertainties in the fluorescence quanti-
fication and makes it difficult to image labeled cells repeatedly
over a period of time. In addition, highly reactive oxygen species
that are toxic to cells are generated during photobleaching.
Rhodamines are less prone to photobleaching than fluorescein
or resorufin, and caged rhodamines were subsequently developed
to circumvent the problem.4,7 The uncaging quantum yields of
these caged rhodamines have not been reported yet, but they
were probably comparable to or less than those of caged
fluoresceins.4 Moreover, since two caging groups were used to
protect one rhodamine molecule as bis-carbamates, the ef-
ficiency of photolysis was further reduced because both caging
groups need to be photolyzed in order to generate one
fluorescent molecule. Finally, to apply caged fluorophores to
cell imaging, it would be desirable to devise delivery strategies
to load these probes inside cells to sufficient concentrations.
None of the previously reported caged fluorophores have been
derivatized into cell permeable forms that can be enriched inside
cells to high concentrations noninvasively.
2. Results and Discussion
2.1. Design and Syntheses of Caged Coumarins. We
developed a new caged fluorophore based on 6-chloro-7-
hydroxy-coumarin 3-carboxamide (1 of Figure 1) because of
its combined advantages including high fluorescence quantum
yield, strong absorption above 400 nm, and insensitivity to
physiological pH fluctuations.8 Starting from 4-chloro-resorcinol,
we prepared 6-chloro-7-hydroxy-coumarin 3-carboxylate in two
steps. Coupling coumarin 3-carboxylate with the methyl ester
of D-glutamate provided coumarin 3-carboxamide (1a, Figure
1). The dye absorbs maximally at 408 nm with an extinction
coefficient of 44 000. The fluorescent quantum yield of the
molecule is 0.93 measured in an aqueous buffer (100 mM KCl,
20 mM Mops, pH 7.35). Three caged derivatives of this
fluorophore were synthesized by masking the 7-hydroxyl group
of coumarin with different cages: 1-(2-nitrophenyl)ethyl (NPE,
2a, Figure 1), 2-nitrobenzyl (NB, 2b), and 4,5-dimethoxy-2-
nitrobenzyl (DMNB, 2c) groups.
We now report a new class of caged coumarin fluorophores
with very high uncaging cross sections at 365 nm. These caged
coumarins exhibit more than 200-fold fluorescence enhancement
after photoconversion. In addition, the two-photon uncaging
cross sections of two NPE-caged coumarins 2a and 5 approach
1 Goeppert-Mayer (GM) at 740 nm. To our knowledge, this
represents the first example that caged molecules based on an
2.2. Photochemical and Fluorescent Properties of Caged
Coumarins. The absorption maxima of all three coumarin cages
center around 360 nm (Figure 2A). The absorption spectra also
contain broad shoulders between 370 and 380 nm. All three
caged coumarins are essentially nonfluorescent, shown by their
(6) Theriot, J. A.; Mitchison, T. J. Nature 1991, 352, 126.
(7) Ottl, J.; Gabriel, D.; Marriott, G. Bioconjugate Chem. 1998, 9, 143.
(8) Zlokarnik, G.; Negulescu, P. A.; Knapp, T. E.; Mere, L.; Burres, N.; Feng,
L.; Whitney, M.; Roemer, K.; Tsien, R. Y. Science 1998, 279, 84.
9
4654 J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004

A New Generation of Caged Fluorophores
A R T I C L E S
Figure 2. (A) Absorption spectra of caged coumarins 2a-c (10 µM). (B) Emission spectra (Ex ) 410 nm) of 2a (1 µM) prior to UV exposure and after
increasing durations of UV illuminations. The inset is the excitation spectrum (Em ) 450 nm) of the exhaustive photolysis product of 2a. All spectra were
taken in 20 mM Mops buffer with 100 mM KCl, pH 7.3.
Table 1. Fluorescent and Photochemical Properties of Caged
that photoconversion of caged coumarins 2a-c produced parent
coumarin 3-carboxamide 1a. Further confirmation on the identity
Coumarinsa
of photolysis products came from the high-pressure liquid
chromatography (HPLC) analysis (vide infra). The fluorescence
enhancements after uncaging ranged from 200-fold (2b) to
nearly 800-fold (2c) (Table 1). This robust change in fluores-
cence quantum yields is important for generating high contrast
optical signals with minimal background.
The extent of photoconversion of caged coumarins to
coumarin 3-carboxamide depended on the length of UV
illumination. NPE-caged coumarin 2a reached 100% conversion
much faster than the other two coumarin cages (Figure 3). The
quantum yields of photolysis (determined by the ferrioxalate
a Qf2: fluorescence quantum yields of caged coumarins. Qf1: fluorescence
quantum yield (0.93) of coumarin 1a. Qf1/Qf2: folds of fluorescence
actinometry)9 and the uncaging cross sections of three caged
fluorophores were measured and are shown in Table 1. There
are remarkable differences in the uncaging cross sections among
three caged coumarins. The uncaging cross section of NPE-
caged coumarin (2a) exceeded 6000, about 2 orders of
magnitude higher than a previously reported NB-caged fluo-
rescein monoether (Quꢀ ) 76 at 350 nm).5 The NB-caged
coumarin (2b) has an uncaging cross section significantly lower
than that of 2a, but is still 6 times higher than that of the DMNB
caged coumarin (2c) or nearly 1 order of magnitude higher than
that of a caged fluorescein.5
enhancement after photoconverting 2a-c to 1a. All measurements were
done in 20 mM Mops buffer containing 100 mM KCl, pH 7.3.
The very high uncaging cross sections of 2a and 2b mainly
resulted from much increased UV absorption. The extinction
coefficients of NB and NPE groups by themselves are less than
400 M-1 cm-1 above 360 nm. The theoretical up limits of their
uncaging cross sections should be no more than 400 M-1 cm-1
above 360 nm even when Qu approaches unity. When these
caging groups are attached to the 7-hydroxyl group of coumarin,
the overall absorbance of the molecule increased by more than
50 times at UV wavelengths. This suggested that the photonic
energy absorbed by the coumarin moiety was utilized to
photolyze caging groups quite efficiently, at least for the NPE
cage 2a with Qu of 33%.
2.3. Mechanistic Studies of Photolysis of NPE-Caged
Coumarins. The exceptional uncaging efficiency of NPE-caged
coumarin 2a prompted us to investigate the mechanism of its
photolysis and to examine the specific roles of NPE group and
coumarin moiety in the photochemical reaction. The mechanism
of photolysis of 2-nitrobenzyl groups such as NB or NPE groups
has been thoroughly studied.10-13 Upon UV excitation, 2-ni-
Figure 3. Time course of the photoconversion of caged coumarins 2a (1),
2b (O), and 2c (4) to 6-chloro-7-hydroxy-coumarin 3-carboxamide. Caged
coumarins (∼1-2 µM in 20 mM Mops buffer with 100 mM KCl, pH 7.3)
illuminated with 365 nm light for various durations. The peak fluorescence
emission at 448 nm was measured after each episode of UV illumination
and normalized against the fluorescence intensity of a standard of 1a.
negligible fluorescence quantum yields (Table 1). UV illumina-
tion (365 nm) removed the caging group and progressively
generated fluorescent products (Figures 2B and 3). The excita-
tion and emission spectra of the exhaustive photolyzed products
of three coumarin cages were identical, and they showed no
difference from those of coumarin 3-carboxamide 1a, suggesting
(9) Hatchard, C. G.; Parker, C. A. Proc. R. Acad. London A 1956, 235, 518.
9
J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004 4655

A R T I C L E S
Zhao et al.
Scheme 1
trobenzyl group or substituted versions such as an NPE group
forms an aci-nitro intermediate either through a singlet or a
triplet excited state (Scheme 1). Once formed, the aci-nitro
intermediate decays to generate the parent substrate and 2-nitroso
benzaldehyde (from NB) or 2-nitrosoacetophenone (from NPE).
If the photolysis of NPE-caged coumarin 2a proceeds
according to the above mechanism, it should generate two
products: 2-nitrosoacetophenone and coumarin 1a. Figure 4
shows HPLC analyses of the product formation after photolyzing
NPE-caged coumarin 2a and 1-(2-nitrophenyl)ethyl acetate
(NPE acetate). We included NPE acetate in the photolysis study
to calibrate the elution time of 2-nitrosoacetophenone on the
HPLC chromatogram. After UV illumination, the peak corre-
sponding to NPE acetate (peak 1, Figure 4A) decreased. This
change was accompanied by the formation of a new peak with
an earlier elution time at 4.3 min (peak 2, Figure 4A). The
absorption maximum of NPE acetate is 262 nm (spectrum 1,
Figure 4B). The absorption spectrum of the exhaustively
photolyzed solution containing NPE acetate (spectrum 2, Figure
4B) showed the absorption maximum at 280 nm with a shoulder
at 315 nm. This spectrum is identical with the absorption
spectrum of peak 2 of trace b in Figure 4A (data not shown),
and it is in excellent agreement with the reported absorption
spectrum of 2-nitrosoacetophenone.14 Thus, we assigned peak
2 of Figure 4A eluting at 4.3 min as 2-nitrosoacetophenone.
Figure 4C shows HPLC chromatograms of NPE-caged
coumarin 2a and its photolysis products. The chromatograms
were displayed by taking the absorbance at 310 nm to show
the caged coumarin 2a (peak 3, Figure 4C) and two photolyzed
products (peaks 2 and 4, Figure 4C) on the same chromatogram
with comparable peak heights. The HPLC chromatogram of the
synthesized coumarin 1a (trace d, Figure 4C) was also listed as
a reference. Prior to UV illumination, NPE-caged coumarin 2a
was eluted off the column after 9.2 min (trace a, peak 3 of Figure
4C). After UV photolysis (trace b, Figure 4C), peak 3 decreased,
and two new peaks appeared at 2.2 min (peak 2) and 4.3 min
(peak 4), respectively. Peak 2 of Figure 4C was identified as
2-nitrosoacetophenone because its elution time (4.3 min) and
absorption spectrum are the same as those of peak 2 in Figure
4A. Peak 4 corresponded to the synthesized coumarin 1a
because it showed the same elution time (2.2 min), absorption,
excitation, and emission spectra (data not shown). The identity
Figure 4. HPLC analysis of the photolysis products. (A) Chromatograms
of NPE acetate before (trace a) and after (trace b) UV illumination (365
nm, 1000 s). The absorbance of trace b was scaled down 3 times relative
to trace a. (B) Absorption spectra of NPE acetate (90 µM) before (1) and
after (2) UV illumination. (C) Chromatograms of NPE-caged coumarin 2a
before (trace a) and after photolysis at 365 nm (trace b, 50 s UV; and trace
c, 200 s UV). Trace d is the chromatogram of the synthesized coumarin
1a. The minor left shoulder of peak 3 is believed to be a stereoisomer of
compound 2a. All samples were eluted with an isocratic mixture of 70%
acetonitrile and 30% water at a flow rate of 1 mL/min.
(10) McCray, J. A.; Trentham, D. R. Annu. ReV. Biophys. Biophys. Chem. 1989,
18, 239.
(11) Corrie, J. E.; Barth, A.; Munasinghe, V. R.; Trentham, D. R.; Hutter, M.
C. J. Am. Chem. Soc. 2003, 125, 8546.
(12) Schupp, H.; Wong, W. K.; Schnabel, W. J. Photochem. 1987, 36, 85.
(13) Zhu, Q. Q.; Schnabel, W.; Schupp, H. J. Photochem. 1987, 39, 317.
(14) Walker, J. W.; McCray, J. A.; Hess, G. P. Biochemistry 1986, 25, 1799.
of the product in peak 4 was further confirmed by mass
spectroscopy analysis, which showed the expected molecular
weight peak at 397.86 ([M + H]+, 397.06 calcd for coumarin
9
4656 J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004

A New Generation of Caged Fluorophores
A R T I C L E S
Scheme 2
Scheme 3
1a of formula C17H16ClNO8). Longer UV illumination com-
pleted the photoconversion of caged coumarin 2a to 2-nitrosoac-
etophenone and free coumarin 3-carboxamide 1a (trace c, Figure
4C).
The above product studies confirmed photolytic decomposi-
tion of NPE-caged coumarin 2a to the expected products based
on the known mechanism of photochemical reaction of 2-ni-
trobenzyl groups. Recently, 6-bromo-7-hydroxycoumarin-4-
ylmethyl (Bhc) and analogous groups have been developed as
photolabile protecting groups mainly for two-photon uncaging
applications.15 In addition, Dore and co-workers have reported
the synthesis and photochemistry of 8-bromo-7-hydroxyquino-
line-2-ylmethyl (BHQ) group and its applications in photocag-
ing.16 The one-photon uncaging cross sections of these cages
range from several hundreds to slightly over a thousand at 365
nm.
energy to the NPE group after absorbing UV light. This greatly
enhances the overall uncaging efficiency of the coumarin cage.
Since 6-chloro substituent is next to the photocleavage site, we
examined whether chlorine substitution at the 6-position of
coumarin is necessary for the “coumarin-assisted photolysis”.
To address the question, we prepared an NPE-caged 7-hydroxy-
coumarin 3-carboxamide (NPE-HCC1/Me, compound 5, Scheme
3). The compound has an even more impressive uncaging
quantum yield of 0.53 at 365 nm (ꢀ365nm ) 18 400, Quꢀ ) 9752,
measured in pH 7.3 Mops buffer), about 60% higher than the
NPE-caged coumarin 2a. Product analysis by HPLC after
photolyzing 5 also confirmed expected photolytic reaction
generating 2-nitrosoacetophenone and parent 7-hydroxy cou-
marin 3-carboxamide 6 based on the known mechanism of the
NPE group (Figure S2, Supporting Information). The increase
in Qu after removing 6-chlorine substitution probably reflected
a decrease in steric hindrance between the NPE group and
coumarin moiety so that NPE group may adopt more favorable
conformations for receiving photonic energy from coumarin.
In addition, chlorine substitution should have effects on
intersystem crossing or on the relative rates of excited-state
decay, etc. These variations may also affect the efficiency of
photolysis. Further studies would be required to address the
issue.
These examples showed that caged substrates can dramatically
improve the efficiency of photolysis of caging groups by
enhancing the absorption of UV light. Future explorations of
this type of “substrate-assisted photolysis” may yield other cages
of high uncaging cross sections. Since the NPE group by itself
has fairly weak absorption above 350 nm, it is unlikely that the
photonic energy absorbed by the coumarin moiety was trans-
ferred to the NPE group mainly through the Fo¨rster’s type of
dipole coupling. An alternative pathway is that the NPE group
and the coumarin moiety are strongly coupled to form a
“semiconjugated chromophore” in which excitation is delocal-
ized over caged coumarin 2a. Support for this hypothesis came
from the comparison of the absorption spectra of 6-chloro-7-
methoxy-coumarin 3-carboxamide and NPE-caged coumarin 2a.
Both the Bhc group and our NPE-caged coumarin 2a share
a coumarin structural core. However, NPE-caged coumarin 2a
reported here is distinguished from Bhc or analogous groups in
at least two major aspects. First, in our system, 6-chloro-7-
hydroxy coumarin 3-carboxamide was the fluorescent substrate
that was caged by an NPE group, whereas Bhc, BHQ, or related
groups were designed as photolabile protecting groups for caging
other molecules. Unlike Bhc or BHQ group, 6-chloro-7-hydroxy
coumarin 3-carboxamide such as 1a is fairly resistant to UV
illumination. In addition, continuous illumination of 6-chloro-
7-acetoxy-coumarin 3-carboxamide (4 of Scheme 2) with 365
nm UV light over 5 min had little effect on the integrity of the
molecule as shown by the HPLC analysis (Figure S1, Supporting
Information). This further suggested that 6-chloro-7-hydroxy-
coumarin 3-carboxamide by itself is inert to UV light, at least
under the same condition where NPE-caged coumarin 2a was
completely photolyzed.
Second, the mechanism of photolytic cleavage of coumarin-
4-ylmethyl group has been postulated to be a photo SN1 reaction,
and a coumarinylmethyl cation intermediate has been suggested
based on 18O labeling experiments.17 In contrast, the mechanism
of photochemical reaction in our system involves the classical
2-nitrobenzyl photolysis as inferred from the HPLC and
spectroscopic analyses.
During the photolysis of caged coumarin 2a, the coumarin
moiety most likely plays an “antenna” role by transferring
(15) Furuta, T.; Wang, S. S.; Dantzker, J. L.; Dore, T. M.; Bybee, W. J.;
Callaway, E. M.; Denk, W.; Tsien, R. Y. Proc. Natl. Acad. Sci. U.S.A.
1999, 96, 1193.
(16) Fedoryak, O. D.; Dore, T. M. Org. Lett. 2002, 4, 3419.
(17) Schade, B.; Hagen, V.; Schmidt, R.; Herbrich, R.; Krause, E.; Eckardt, T.;
Bendig, J. J. Org. Chem. 1999, 64, 9109.
Figure 5 shows that the “composite spectrum” (trace d, Figure
5) of equimolar 1-(2-nitrophenyl)ethanol and 6-chloro-7-meth-
9
J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004 4657

A R T I C L E S
Zhao et al.
sensitivity for the two-photon photolysis.16 These newly devel-
oped caging groups have been used to protect carboxylates,15,16
phosphates,17,19,21 diols,22 aldehydes, and ketones.23 Molecules
caged by these groups display two-photon uncaging cross
sections about 1 GM at 740 nm. In addition, 4-methoxy-7-
nitroindoline (MNI) group has also been applied as a two-photon
caging group, though the δu (0.06 GM at 730 nm) of a MNI-
caged glutamate is significantly lower than those of molecules
based on Bhc or BHQ groups.20,24
In contrast to these recent developments, caged compounds
employing 2-nitrobenzyl or related protecting groups have fairly
low δu. DMNB-caged compounds such as DM-nitrophen and
DMNB-acetate have δu of ∼0.01-0.03 GM between 720 and
740 nm. Other caged molecules based on NB or NPE groups
have negligible δu.15,18 Since our mechanistic studies on the
photolytic reaction of NPE-caged coumarin indicated energy
coupling between coumarin 3-carboxamide and the nearby NPE
caging group, and because 7-hydroxycoumarin and analogous
chromophores have relatively high two-photon absorption cross
sections,15,25 we proceeded to investigate the sensitivity of NPE-
caged coumarins 2a and 5 (NPE-HCC1/Me, Scheme 3) toward
two-photon photolysis.
To calculate δu, we used 6-bromo-7-hydroxycoumarin-4-
ylmethyl acetate (Bhc-OAc) as a reference compound.15,16 Bhc-
OAc and caged coumarins 2a or 5 in microcuvettes were
illuminated using focused infrared light (740 nm) from a
femtosecond-pulsed and mode-locked Ti:sapphire laser. The
two-photon uncaging action cross section can be determined
using the equation:15
Figure 5. Absorption spectra of NPE-caged coumarin 2a (trace a), 6-chloro-
7-methoxy-coumarin 3-carboxamide (trace b), and 1-(2-nitrophenyl)ethanol
(trace c). All three spectra were taken in 20 mM Mops buffer (pH 7.3)
containing 10 µM substrate. The composite spectrum of equimolar 1-(2-
nitrophenyl)ethanol and 6-chloro-7-methoxy-coumarin 3-carboxamide is
represented by trace d.
oxy-coumarin 3-carboxamide nearly overlapped with the mea-
sured spectrum of 6-chloro-7-methoxy-coumarin 3-carboxamide
(trace b). The spectra (traces b and d) showed fairly low
absorbance above 400 nm. In contrast, the UV-vis spectrum
of caged coumarin 2a (trace a) showed much stronger absorption
above 400 nm, and the absorbance of caged coumarin 2a is
higher than the “composite spectrum” (trace d) throughout. This
probably indicates that there is a sizable coupling between the
NPE group and the coumarin moiety in 2a. The coupling
probably also exists in the NB- and DMNB-caged coumarins
(compounds 2b and 2c, Figure 1) as indicated by the similarity
of the absorption spectra of three caged fluorophores (Figure
2A). However, the photonic energy absorbed by the coupled
chromophore was utilized to photolyze three caging groups with
quite different efficiencies (Table 1). The order of uncaging
quantum yields among caged coumarins 2a-c is consistent with
the general observation made in many other caged molecules:
for the same caged substrate, the NPE cage usually has the
highest uncaging quantum yield, whereas the DMNB cage is
photolyzed least efficiently.3,5
2.4. Two-Photon Uncaging of NPE-Caged Coumarins. A
recent development in the photouncaging field was the introduc-
tion of the two-photon uncaging technique.15,18-20 This method
involves applying high fluxes of infrared (IR) laser light to excite
samples in a highly restricted focal volume, ∼1 µm3. Simul-
taneous absorptions of two or more IR photons with a combined
energy equivalent to one UV photon pump the molecule to the
excited states. By comparison with the usual one-photon UV
photolysis, two-photon uncaging offers the advantages of
releasing active molecules with excellent three-dimensional
resolution and minimizing photo toxicity to cells. The efficiency
of two-photon photolysis is measured as the two-photon
uncaging action cross section, δu. This parameter is the product
of the two-photon absorbance cross section, δa, and the uncaging
quantum yield, Qu2, and is measured in the unit of Goeppert-
Mayer (1 GM ) 10-50cm4‚s/photon).15,18 Ideally, δu of caged
molecules should exceed 0.1 GM for biological applications in
live specimens.15
2Np
δu )
(1)
Cs I0 (t) S2(r)dV
2
∫
where Np is the number of product molecules formed per unit
2
time, I0 (t) is the mean squared light intensity, S(r) is a unitless
spatial distribution function, the integral is over the volume of
the microcuvette, and Cs is the substrate concentration. If caged
compounds are photolyzed under the identical setting with the
2
same two-photon laser power, then the term I0 (t) ∫S2(r)dV
should be about the same for different cages. Thus, δau of cage
a can be determined by comparing its rate of product formation
(Nap) with that of reference cage b with known δbu according to
the equation:
δau NaPCbs
δbu NbPCas
)
(2)
The formation of coumarins after two-photon photolysis of
2a and 5 was quantified similarly as in the one-photon uncaging
experiments by measuring the fluorescence enhancement. The
photolysis of Bhc-OAc was quantified by HPLC analysis as
described.15,16 Figure 6 shows time courses of the photolysis of
Bhc-OAc and NPE-caged coumarins 2a and 5. Both caged
Tsien and co-workers first developed Bhc as a two-photon
cage of general usages.15 More recently, Dore and co-workers
described using BHQ as another protecting group of sufficient
(20) Matsuzaki, M.; Ellis-Davies, G. C.; Nemoto, T.; Miyashita, Y.; Iino, M.;
Kasai, H. Nat. Neurosci. 2001, 4, 1086.
(21) Furuta, T.; Iwamura, M. Methods Enzymol. 1998, 291, 50.
(22) Lin, W.; Lawrence, D. S. J. Org. Chem. 2002, 67, 2723.
(23) Lu, M.; Fedoryak, O. D.; Moister, B. R.; Dore, T. M. Org. Lett. 2003, 5,
2119.
(24) Canepari, M.; Nelson, L.; Papageorgiou, G.; Corrie, J. E.; Ogden, D. J.
Neurosci. Methods 2001, 112, 29.
(25) Parma, L.; Omenetto, N. Chem. Phys. Lett. 1978, 54, 541.
(18) Brown, E. B.; Shear, J. B.; Adams, S. R.; Tsien, R. Y.; Webb, W. W.
Biophys. J. 1999, 76, 489.
(19) Ando, H.; Furuta, T.; Tsien, R. Y.; Okamoto, H. Nat. Genet. 2001, 28,
317.
9
4658 J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004

A New Generation of Caged Fluorophores
A R T I C L E S
Figure 6. Time course of two-photon uncaging of NPE-caged coumarin
2a (1), 5 (9), and Bhc-OAc (4) at 740 nm (average power 590 mW exiting
the cuvette). The formation of coumarin 1a from 2a (initial concentration
10 µM) was measured by fluorescence spectroscopy. The photolysis of Bhc-
OAc (initial concentration 100 µM) was followed by HPLC. Solid lines
are least-squares fits of simple decaying exponentials.
Figure 7. Fluorescence imaging of Hela cells loaded with a caged and
cell permeable coumarin 3. Hela cells were loaded with 2 µM 3 for 1 h in
the Hanks balanced salt solution (HBSS, pH 7.35) containing 10 mM Hepes
buffer and 5.5 mM glucose. After washing, cells on glass coverslips were
imaged on an inverted microscope (Axiovert 200, Carl Zeiss). During the
imaging, cells were briefly illuminated with UV light (330-380 nm),
indicated by the arrows in part C. The first two arrows denote 0.1 s UV
exposures, and the last two arrows represent 0.2 s UV. (A and B) Fluorescent
images of cells before and after UV uncaging, respectively. (C) Average
fluorescence intensity of cells in the field of view over the course of four
episodes of UV photolysis.
coumarins 2a, 5, and Bhc-OAc were photolyzed by 740-nm
laser light quite efficiently, yet Bhc-OAc and 5 were about two
times more sensitive to two-photon uncaging than 2a. Since
the reported δu of Bhc-OAc has been measured at 740 nm at
the laser power of 345 mW exiting the cuvette,16 we also
compared the rates of photolysis of 2a, 5, and Bhc-OAc at this
power level. After 20 min of continuous illumination, 5.5, 10,
and 10.6% of photolysis occurred for compounds 2a, 5, and
Bhc-OAc, respectively. Using the reported δu of Bhc-OAc (0.72
GM at 740 nm)15,16 and eq 2, we estimated δu of 2a and 5 to be
0.37 and 0.68 GM, respectively. In addition, product analysis
by HPLC after two-photon photolysis of caged coumarins 2a
and 5 showed the same chromatograms as those obtained after
one-photon uncaging (data not shown), suggesting that the NPE
group was converted to 2-nitrosoacetophone efficiently after
two-photon illuminations.
Thus far caged compounds employing NPE or NB groups
have shown unmeasurably low two-photon uncaging action cross
sections. To our knowledge, our newly developed NPE-caged
coumarins such as 2a and 5 represented the first example that
δu of caged compounds based on 2-nitrobenzyl groups can
approach 1 GM. This result further supported our proposed
mechanism that the coumarin moiety serves as an antenna to
enhance the light harvesting capability of the molecule and to
boost the photolytic efficiency of NPE group.
2.5. Caged and Cell Permeable Coumarins for in Vivo
Fluorescence Imaging Applications. To apply these caged
coumarins for imaging applications in live cells, we attempted
to derivatize them into cell membrane permeant forms. We
focused our initial efforts on NPE-caged coumarin 3-carbox-
amide such as 2a because of its combined advantages, including
high uncaging cross sections and robust fluorescence enhance-
ment after uncaging. In addition, its parent fluorophore 7-hy-
droxy-6-chloro-coumarin-3-carboxamide (HCCC1, Figure 1) is
relatively insensitive to physiological pH fluctuations. We first
attempted to protect two carboxylates of D-glutamate of 2a with
acetoxymethyl (AM) esters (compound 3, Figure 1). AM esters
are stable in physiological solutions but are susceptible to
hydrolysis by cellular esterases. The enzymatic hydrolysis
regenerates negatively charged carboxylates which are trapped
inside cells.26 Initial attempts of hydrolyzing two methyl esters
of 2a under a number of conditions were accompanied by the
concomitant hydrolysis of the amide bond of coumarin 3-car-
boxamide. Instead, we found that protecting two carboxylates
of glutamate with tert-butyl groups provided a more feasible
synthetic route. Thus, acid treatments of the intermediate 2d
(prepared from 1b, Figure 1) followed by esterification with
AM bromide afforded the AM ester of NPE-caged coumarin 3
(NPE-HCCC1/AM, Figure 1).
Figure 7 shows fluorescence images of Hela cells loaded with
coumarin cage 3. Fluorescent imaging was taken by using
excitation bandpath filters with cut-on wavelength longer than
420 nm. Prior to uncaging, continuous monitoring of cells loaded
with 3 showed that there was no difference in fluorescence
intensity of loaded cells from control cells, indicating that the
background fluorescence of coumarin cage is negligible (Figure
7A). Subsequent UV illumination generated a sudden jump of
cellular blue fluorescence (Figure 7B), and repetitive uncaging
caused stepwise increases of coumarin fluorescence (Figure 7C).
To test if NPE-HCCC1/AM performs the same as a fluores-
cent marker in other types of cells, we further evaluated its
fluorescence imaging properties in different cell lines. After
loading and uncaging of NPE-HCCC1/AM in cultured primary
human fibroblasts, we noticed that the coumarin fluorescence
in these cells was not as stable as that in Hela cells. Figure 8A
shows that the coumarin fluorescence intensity of HCCC1grad-
ually decayed in primary fibroblasts after whole field uncaging.
(26) Li, W. H.; Llopis, J.; Whitney, M.; Zlokarnik, G.; Tsien, R. Y. Nature
1998, 392, 936.
9
J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004 4659

A R T I C L E S
Zhao et al.
Figure 8. Fluorescence signal of HCCC1 (A) is less stable than that of HCCC2 (B) in primary human fibroblasts. Human primary fibroblasts were loaded
with caged coumarin NPE-HCCC1/AM (A) or NPE-HCCC2/AM (B), imaged, and uncaged similarly as in Figure 7. Arrows indicate UV flashes of various
durations (seconds): 0.2, 0.4, 0.8, 1.2, 1.6, and 2 (from left to right) (A) and 0.2, 0.2, 0.4, 0.4, and 0.8 (B).
t
Figure 9. Synthesis of a caged and cell permeable coumarin, NPE-HCCC2/AM. (a) (1) Trifluoroacetic anhydride, CH2Cl2, r.t.. (2) BrCH2CO2 Bu, KHCO3,
DMF, 36% for two steps. (b) (1) 1 N NaOH, THF. (2) 6-Chloro-7-hydroxy-coumarin 3-carboxylate, SOCl2, cat DMF, CH2Cl2, Et3N, 24% for two steps. (c)
NPE bromide, DIEA, CH2Cl2, 57%. (d) (1) TFA, Et3SiH, CH2Cl2. (2) BrCH2OAc, DIEA, CH2Cl2, 33% for two steps. (e) Hydrolysis by cell esterases and
UV uncaging.
The declination of fluorescence signal of HCCC1 was also
observed in a few other cell lines tested, such as COS cells and
NIH3T3 fibroblasts, yet the rates of its intensity decay were
slower than what was observed in primary fibroblasts. At the
moment we are not yet clear about the causes for the gradual
loss in coumarin fluorescence intensity, though we speculated
that it might be due to a slow metabolism of 7-hydroxy-6-chloro-
coumarin 3-carboxamide (HCCC1, Figure 1) in these cells.
To improve the fluorescence signal stability of coumarin
3-carboxamide in cells, we designed a new coumarin fluoro-
phore HCCC2 (11, Figure 9) in which 1,4-cis-diaminocyclo-
hexane N,N-diacetate was conjugated with coumarin 3-carbox-
ylate. We reasoned that increasing steric hindrance of the amide
bond of coumarin 3-carboxamide should enhance its enzymatic
stability. In addition, incorporating N,N-diiminoacetates moiety
in the molecule should improve its cellular retention. The
synthesis of its caged and cell permeable precursor NPE-
HCCC2/AM (10) is shown in Figure 9.
including Hela cells, NIH3T3 fibroblasts, COS cells, CHO cells,
and 1321N1 astrocytoma cells. In all these cells, the compound
showed essentially the same desired imaging properties, includ-
ing ease of loading, high sensitivity toward UV photolysis, and
stability of coumarin fluorescence signal after uncaging. Thus,
NPE-HCCC2/AM is of more general utility as a fluorescent
marker in mammalian cells than NPE-HCCC1/AM.
In summary, we have designed and synthesized a new class
of caged coumarins suitable for in vivo imaging applications.
By comparison with previous caged fluorophores, our newly
developed caged coumarins show a number of advantages,
including photostability of parent dyes, high uncaging efficiency,
and much improved cell loading properties. Since the coumarins
described here emit blue light, they spectrally complement many
existing fluorescent sensors emitting at green or red regions.
The combinatory uses of these coumarin cages with other
fluorescent dyes or indicators which emit at longer wavelengths
may allow us to expand the fluorescent imaging window and
to carry out photouncaging and multicolor imaging simulta-
neously in live cells. Since NPE-caged coumarins such as 2a
and 5 possess high two-photon uncaging action cross sections,
this class of caged coumarins may also have applications in
biological systems where the two photon uncaging technique
To our satisfaction, caged coumarin NPE-HCCC2/AM loaded
quite well into cells. Repetitive photouncaging generated
stepwise increases in cell coumarin fluorescence, which re-
mained stable in primary fibroblasts (Figure 8B). We also tested
the compound in other types of cultured mammalian cells,
9
4660 J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004

A New Generation of Caged Fluorophores
A R T I C L E S
calcd for C17H16ClNO8; obsd: 398.59 (M + H)+, 420.56 (M + Na)+,
436.54 (M + K)+.
is preferred to UV photolysis. Finally, the synthetic strategy
outlined in Figure 1 would allow assembling other derivatives
of these caged coumarins by conjugating coumarin 3-carboxylate
with various amines. This provides a facile route of introducing
different reactive groups into the molecule for the bioconjuga-
tion. The syntheses of these derivatives and their applications
in biological imaging are in progress and will be reported in
due course.
6-Chloro-7-[1-(2-nitrophenyl)ethoxy]-coumarin 3-Carboxamide
(2d). A solution of 1b (45.3 mg, 94 µmol), NPE bromide (25.9 mg,
0.113 mmol), and DIEA (32.7 µL, 0.187 mmol) in acetonitrile (300
µL) was stirred at 45 °C for 10 h. After cooling, the reaction mixture
was directly purified by silica gel chromatography with CH2Cl2/MeOH
(98:2) as eluant. The desired product was obtained as a solid after drying
1
(42.8 mg, 72.5%). H NMR (CD3Cl, δ ppm): 9.05 (d, J ) 8.0 Hz,
1H), 8.66 (s, 1H), 8.05 (d, J ) 8.0 Hz, 1H), 7.58-7.72 (m, 3H), 7.45
(m, 1H), 6.67 (d, J ) 1.6 Hz, 1H), 6.22 (q, J ) 6.4 Hz, 1H), 4.62 (m,
1H), 2.13-2.34 (m, 4H), 1.76 (d, J ) 6.0 Hz, 3H), 1.43 (s, 9H), 1.36
(s, 9H). MS: 630.20 calcd for C31H35ClN2O10; obsd: 653.99 (M +
Na)+, 669.97 (M + K)+.
3. Experimental Section
3.1. Materials and General Methods. All reagents were purchased
from Aldrich or Fluka. Anhydrous solvents for organic syntheses were
purchased from Aldrich and stored over activated molecular sieves (4
Å). Thin-layer chromatography (TLC) was performed on precoated
silica gel 60F-254 glass plates. Reaction products were purified by low-
pressure flash chromatography (FC) using silica gel 60 (∼63-200 µm,
EM Science). Syntheses of caged compounds sensitive to near-UV light
were carried out in an isolated room equipped with red safety lamps.
1H NMR spectra were obtained using Varian 300 or 400 MHz
spectrometers. Chemical shifts (δ, ppm) were reported against TMS
(0 ppm). MALDI-TOF mass spectroscopy was performed on a Voyager-
DE PRO biospectrometry workstation (Applied Biosystems) using 2,5-
dihydroxy benzoic acid as the matrix. UV-vis spectra were recorded
in a 1-cm path quartz cell on a Shimadzu 2401 PC spectrometer.
Fluorescence excitation and emission spectra were recorded on a
Fluorolog 3 spectrometer (Jobin-Yvon Horiba, Edison, NJ). Bhc-OAc
was synthesized as described.15
Compounds 2a-c were synthesized similarly as 2d from 1a.
1
2a: H NMR (CD3Cl, δ ppm): 9.08 (d, J ) 8.0 Hz, 1H), 8.69 (s,
1H), 8.10 (d, J ) 8.0 Hz, 1H), 7.6-7.75 (m, 3H), 7.4-7.5 (m, 1H),
6.71 (d, J ) 2.0 Hz, 1H), 6.27 (q, J ) 6.4 Hz, 1H), 4.78 (m, 1H), 3.76
(d, 3H), 3.64 (d, 3H), 2.23-2.5 (m, 3H), 2.03-2.18 (m, 1H), 1.80 (d,
J ) 6.1 Hz, 3H). MS: 546.10 calcd for C25H23ClN2O10; obsd: 569.8
(M + Na)+, 585.82 (M + K)+.
1
2b: H NMR (CD3Cl, δ ppm): 9.14 (d, J ) 7.6 Hz, 1H), 8.76 (s,
1H), 8.25 (dd, J ) 8.4, 1H), 8.00 (d, J ) 8.2 Hz, 1H), 7.76 (t, J ) 7.6,
1H), 7.72 (s, 1H), 7.57 (t, J ) 7.6 Hz, 1H), 7.05 (s, 1H), 5.62 (s, 2H),
4.80 (m, 1H), 3.77 (s, 3H), 3.67 (s, 3H), 2.3-2.5 (m, 3H), 2.08-2.18
(m, 1H). MS: 532.09 calcd for C24H21ClN2O10; obsd; 533.94 (M +
H)+, 555.89 (M + Na)+, 571.88 (M + K)+.
1
2c: H NMR (CD3Cl, δ ppm): 9.14 (d, J ) 7.6 Hz, 1H), 8.76 (s,
1H), 7.81 (s, 1H), 7.75 (s, 1H), 7.6 (s, 1H), 7.12 (s, 1H), 5.62 (s, 2H),
4.80 (m, 1H), 4.05 (s, 3H), 3.98 (s, 3H), 3.77 (s, 3H), 3.65 (s, 3H),
2.3-2.52 (m, 3H), 2.08-2.18 (m, 1H). MS: 592.11 calcd for C26H25-
ClN2O12; obsd; 593.69 (M + H)+, 615.69 (M + Na)+, 631.67 (M +
K)+.
3.2. Syntheses. 2.4-Dihydroxy-5-chlorobenzaldehyde. The com-
pound was synthesized from 4-chlororesorcinol using a previously
described procedure.27 1H NMR (CDCl3, δ ppm): 11.25 (s, 1H), 9.69
(s, 1H), 7.52 (s, 1H), 6.62 (s, 1H), 6.18 (br s, 1H).
6-Chloro-7-hydroxy-coumarin 3-Carboxylate. A suspension of the
above benzaldehyde (0.58 g, 3.36 mmol), malonic acid (0.72 g, 6.9
mmol), and a catalytic amount of aniline in pyridine (3.0 mL) was
stirred at room temperature (rt) for 72 h. EtOH (5.0 mL) was then
added. The mixture was stirred at this temperature for 1 h and filtered.
The filtrate was washed sequentially with 0.1 N HCl, H2O, and Et2O
to afford a yellow solid. The solid was dried under high vacuum
overnight to give the desired product as a yellow powder (0.35 g,
43.7%). 1H NMR (DMSO-d6, δ ppm): 8.62 (s, 1H), 7.96 (s, 1H), 6.87
(s, 1H). MS: 239.98 calcd for C10H5ClO5; obsd: 241.47 (M + H)+,
263.47 (M + Na)+, 279.45 (M + K)+.
6-Chloro-7-hydroxy-coumarin 3-Carboxamide (1b). A droplet of
DMF was added to a suspension of 6-chloro-7-hydroxy-coumarin
3-carboxylate (85 mg, 0.35 mmol) and SOCl2 (0.253 mL, 3.477 mmol)
in CH2Cl2 (5.0 mL). The mixture was stirred at 45 °C for 4 h and then
evaporated to dryness. The residue was resuspended in 2 mL dry CH2-
Cl2. To this suspension a solution of H-D-di-tert-butyl glutamate
hydrogen chloride (157 mg, 0.53 mmol) and Et3N (196 µL, 1.4 mmol)
in CH2Cl2 (3.0 mL) was added. The resulting mixture was stirred at rt
overnight. The solution was then poured into EtOAc, extracted with
0.1 N HCl, washed with brine, dried over Na2SO4, and concentrated to
a residue. The residue was purified by FC (CH2Cl2/MeOH, 95:5) to
afford 1b as a pale yellow powder (89 mg, 52.6%). 1H NMR (CDCl3,
δ ppm): 9.24 (d, J ) 8.0 Hz, 1H), 8.65 (s, 1H), 7.56 (s, 1H), 6.98 (s,
1H), 4.64 (m, 1H), 2.31-2.42 (m, 2H), 2.18-2.28 (m, 2H), 1.47 (s,
9H), 1.41 (s, 9H). MS: 481.15 calcd for C23H28ClNO8; obsd: 504.70
(M + Na)+, 520.67 (M + K)+.
NPE-HCCC1/AM (3): A solution of 2d (12 mg, 0.0191 mmol),
CF3CO2H (38.4 µL, 0.498 mmol), triethylsilane (7.5 µL, 0.048 mmol),
and CH2Cl2 (100 µL) was stirred at rt. The reaction was continued
until two tert-butyl esters were completely hydrolyzed, as monitored
by TLC. The reaction mixture was concentrated under high vacuum
and used directly for the esterification. After being dried, the resulting
residue was resuspended in acetonitrile (200 µL). Bromomethyl acetate
(3.5 µL, 0.0466 mmol) and DIEA (11.1 µL, 0.636 mmol) were then
added. The mixture was left at rt overnight. After the solvent was
removed, water (200 µL) was added. The mixture was extracted with
CH2Cl2 (3 × 10 mL). The combined extracts were dried (Na2SO4) and
concentrated. The residue was purified by flash chromatography eluting
with CH2Cl2/MeOH (98/2) to give 6.3 mg of desired product (60%).
1H NMR (CD3Cl, δ ppm): 9.09 (d, J ) 7.6 Hz, 1H), 8.68 (s, 1H),
8.11 (d, J ) 8.0 Hz, 1H), 7.6-7.8 (m, 3H), 7.49 (m, 1H), 6.71 (s, 1H),
6.27 (m, 1H), 5.69-5.81 (m, 4H), 4.79 (m, 1H), 2.55-2.52 (m, 2H),
2.31-2.40 (m, 2H), 2.11 (s, 3H), 2.09 (s, 3H), 1.81 (d, J ) 6.4 Hz,
3H). MS: 662.12 calcd for C29H27ClN2O14; obsd: 663.86 (M + H)+,
685.86 (M + Na)+.
N,N-di-tert-butylacetate-N′-trifluoroacetamide 1,4-cis-Diaminocy-
clohexane (7). Trifluoroacetic anhydride (0.91 mL, 6.550 mmol) was
added to a solution of 1,4-cis-diaminocyclohexane (0.68 g, 5.955 mmol)
in CH2Cl2 at 0 °C. The resulting mixture was stirred at rt overnight. It
was then diluted with CH2Cl2, washed with saturated NaHCO3 and
brine, and concentrated to give a white solid. This was used directly
for the next step without further purification.
The crude monotrifluoroacetamide (0.30 g, 1.427 mmol) obtained
above was mixed with KHCO3 (0.36 g, 3.596 mmol) in DMF (2.0 mL).
tert-Butyl bromoacetate (0.47 mL, 3.139 mmol) was added dropwise
at rt. The resulting mixture was stirred overnight and then diluted with
EtOAc (300 mL), washed with saturated NH4Cl solution, dried over
Na2SO4, and concentrated under vacuum. The residue was purified by
FC eluted with hexane/EtOAc (9:1 to 4:1) to afford 7 as a colorless oil
Compound 1a was synthesized similarly from H-D-dimethyl glutamate
1
hydrogen chloride. 1a: H NMR (CDCl3, δ ppm): 9.17 (d, J ) 7.6
Hz, 1H), 8.71 (s, 1H), 7.64 (s, 1H), 7.04 (s, 1H), 4.79 (m, 1H), 3.78 (s,
3H), 3.67 (s, 3H), 2.4-2.5 (m, 2H), 2.15-2.40 (m, 2H). MS: 397.06
(27) Tsien, R. Y.; Zlokarnik, G. U.S. Patent 6,472,205, 2002.
9
J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004 4661

A R T I C L E S
Zhao et al.
1
(223 mg, 35.6% for two steps). H NMR (400 MHz, δ ppm, CDCl3):
6.14 (d, J ) 7.3 Hz, 1H), 3.71 (m, 1H), 3.43 (s, 4H), 2.68 (m, 1H),
2.04 (d, J ) 12.0 Hz, 2H), 1.94 (d, J ) 12.0 Hz, 2H), 1.42 (s, 18H),
1.19-1.38 (m, 4H). MS: 438.23 calcd for C20H33F3N2O5; obsd: 439.14
(M + H+), 461.12 (M + Na+).
420 nm or above to minimize photolysis during fluorescence measure-
ments. The percentage of photolysis was calculated from the fluores-
cence enhancement normalized against the fluorescence intensity of
pure 6-chloro-7-hydroxy-coumarin 3-carboxamide (compound 1a;
Figure 1). The progress of photochemical reaction was plotted as simple
decaying exponentials. Quantum yields of uncaging (Qu) were calculated
from the equation Qu ) -log(C/C0)/(Iσt), where I is the irradiation
intensity in einsteins‚cm-2‚s-1, σ is the decadic extinction coefficient
in cm2‚mol-1 (103 times of the usual extinction coefficient ꢀ in M-1
cm-1), t is the irradiation time in seconds, and C and C0 are
concentrations of caged coumarins at time t and at the beginning of
photolysis, respectively.26 The total UV intensity I was measured with
the ferrioxalate actinometry,9 and it was typically in the range of ∼0.5-
1.02 × 10-8 E/(cm2‚s).
3.4. Product Analysis and Mechanistic Studies of Photolysis of
Caged Coumarins. The photolysis products were analyzed by HPLC.
A Shimadzu HPLC system (SCL-10A vp controller) equipped with an
auto injector, a photodiode array detector, and a fluorescence detector
was used. Separations were carried out on an analytical reverse-phase
column (250 mm × 4.6 mm, Luna C-18/5 µm, Phenomenex, Torrance,
CA) using an isocratic flow of CH3CN/H2O (7:3). Caged coumarins
(compounds 2a or 5, 70 µM) in Mops buffer (20 mM Mops, 100 mM
KCl, pH 7.3) were photolyzed with 365 nm of light. The sample was
then diluted with acetonitrile and analyzed by HPLC. To calibrate the
retention time of 2-nitrosoacetophenone, 1-(2-nitrophenyl)ethyl acetate
was photolyzed and analyzed by HPLC under the same condition.
6-Chloro-7-hydroxy-coumarin 3-Carboxamide (8). Monotrifluo-
roacetamide 7 (75 mg, 0.172 mmol) was saponified with NaOH (171
µL, 1 N) in MeOH (2 mL) overnight. It was then diluted with EtOAc
(100 mL), washed with saturated NaHCO3, dried over Na2SO4, and
concentrated to a residue which was used immediately for the next
step.
The above amine was dissolved in 2 mL of CH2Cl2. 6-Chloro-7-
hydroxycoumarin 3-carbonyl chloride (0.0894 mmol, prepared as in
making 1b) suspended in CH2Cl2 (2.0 mL) was then added. Et3N (37.4
µL, 0.274 mmol) was added shortly afterward. The resulting mixture
was stirred at rt overnight and then diluted with EtOAc (150 mL),
washed with brine, and dried over Na2SO4. The concentrated residue
was purified by FC (hexane/EtOAc, 3:2 to 1:1) to afford 8 as a yellow
solid (12.2 mg, 24% for two steps). 1H NMR (400 MHz, δ ppm,
CDCl3): 8.74 (s, 1H), 8.62 (d, J ) 8.0 Hz, 1H), 7.66 (s, 1H), 7.06 (s,
1H), 3.86 (m, 1H), 3.44 (s, 4H), 2.70 (m, 1H), 2.09 (d, J ) 10.6 Hz,
2H), 1.94 (d, J ) 10.6 Hz, 2H), 1.45 (s, 18H), 1.24-1.39 (m, 4H).
MS: 564.22 calcd for C28H37ClN2O8; obsd: 565.87 (M + H+), 587.86
(M + Na+).
7-[1-(2-Nitrophenyl)ethoxy]-6-chloro-coumarin 3-Carboxamide
(9). 6-Chloro-7-hydroxy-coumarin 3-carboxamide (8, 7.4 mg, 13 µmol),
NPE-Br (4.5 mg, 20 µmol), and DIEA (6.8 µL, 39 µmol) were mixed
in CH3CN-CH2Cl2 (0.5 mL, 1:1). The solution was stirred at 40 °C
for 8 h and then cooled. The reaction mixture was diluted with EtOAc
(10 mL), washed with saturated NH4Cl and dried over Na2SO4. The
concentrated residue was purified by FC eluted with hexane/EtOAc
(4:1 to 7:3) to afford NPE-caged coumarin 9 as a slightly yellow solid
3.5. Measurements of Two-Photon Uncaging Cross Sections.
Two-photon photolysis was carried out using a procedure similar to
the one first described by Furuta et al.15 Samples in 100 mM Mops
buffer (pH 7.3) were transferred to a microcuvette (Hellma 105.251-
QS). The filling volume of the cuvette was 45 µL. The laser beam
from a femtosecond-pulsed and mode-locked Ti:sapphire laser (Mira
900-F pumped by a Verdi, Coherent, Santa Clara, CA) was focused
into the center of the cuvette with a focusing lens (01 LPX 029/077,
focal length 25 mm, Melles-Griot, Carlsbad, CA). After irradiation with
740 nm of light, samples were collected, and the formation of the
products was quantified by measuring fluorescence enhancements or
by HPLC analysis. Two-photon uncaging cross sections (δu) were
calculated from eq 2 using the reported δu of 0.72 GM for Bhc-OAc at
345 mW laser power exiting the cuvette.16 Since the percent of
photolysis of caged coumarins 2a and 5 changed little at concentrations
10 µM or above (data not shown), we used 10 µM of 2a and 5 in the
microcuvette during the two-photon photolysis.
1
(5.4 mg, 57.4%). H NMR (400 MHz, δ ppm, CDCl3): 8.71 (s, 1H),
8.48 (d, J ) 8.0 Hz, 1H), 8.10 (dd, J ) 8.0, 1.0 Hz, 1H), 7.40-7.78
(m, 4H), 6.69 (s, 1H), 6.25 (q, J ) 6.4 Hz, 1H), 3.83 (m, 1H), 3.45 (s,
4H), 2.69 (m, 1H), 2.07 (d, J ) 10.9 Hz, 2H), 1.93 (d, J ) 10.6 Hz,
2H), 1.80 (d, J ) 6.4 Hz, 3H), 1.42 (s, 18H), 1.2-1.43 (m, 4H). MS:
713.27 calcd for C36H44ClN3O10; obsd: 714.17 (M + H+), 736.16 (M
+ Na+).
NPE-HCCC2/AM (10): NPE-caged coumarin 9 (5.4 mg, 7.56 µmol)
and triethylsilane (5.9 µL, 38 µmol) were dissolved in CH2Cl2 (200
µL). CF3CO2H (15.7 µL, 0.204 mmol) was added. The resulting solution
was stirred at rt in the dark for 4.5 h and then evaporated to a residue.
The dried residue was dissolved in acetonitrile (150 µL). To this
solution, bromomethyl acetate (3.7 µL, 38 µmol) and DIEA (13.2 µL,
76 µmol) were added. The solution was stirred at rt in the dark
overnight, diluted with CH2Cl2 (10 mL), washed with saturated NH4-
Cl, dried over Na2SO4, and concentrated. The residue was purified by
FC eluting with hexane/EtOAc (3:2) to afford the final product 7 (1.8
mg, 32.5% for two steps) as a yellow solid. MS: 745.19 calcd for
C34H36ClN3O14; obsd: 746.25 (M + H+), 768.23 (M + Na+). 1 H NMR
(400 MHz, δ ppm, CDCl3): 8.71 (s, 1H), 8.49 (d, J ) 8.0 Hz, 1H),
8.10 (dd, J ) 8.0, 1.2 Hz, 1H), 7.73 (dd, J ) 8.0, 1.2 Hz, 1H), 7.67 (s,
1H), 7.64 (m, 1H), 7.49 (m, 1H), 6.69 (s, 1H), 6.25 (q, J ) 6.4 Hz,
1H), 5.75 (s, 4H), 3.82 (m, 1H), 3.62 (s, 4H), 2.71 (m, 1H), 2.04-2.18
(m, 8H), 1.92 (d, J ) 6.4 Hz, 2H), 1.80 (d, J ) 6.4 Hz, 3H), 1.20-
1.42 (m, 4H).
3.6. Uncaging and Fluorescence Imaging of Caged Coumarins
in Cultured Cells. Hela cells were cultured in high glucose DMEM
medium (Gibco) containing 10% fetal bovine serum and 1% penicillin/
streptomycin. Human primary fibroblasts isolated from foreskin (gener-
ously supplied by Dr. Frederick Grinnell) were cultured in the same
medium and were used until 20 doublings were reached. Cells to be
imaged were cultured on 35-mm Petri dishes with glass bottoms
(MatTek, Ashland, MA). Fluorescence microscopy was carried out on
an inverted fluorescence microscope (Axiovert 200, Carl Zeiss) with a
40× objective. Epifluorescence was detected with a cooled CCD camera
(ORCA-ER, Hamamatsu). Cells were excited with light from a 175 W
xenon lamp after passing through appropriate bandpath filters. Switching
of excitation wavelengths between UV uncaging and image acquisition
was controlled by a high-speed wavelength switcher (Lambda DG-4,
Sutter Instrument). Image acquisition and analysis were accomplished
using an integrated imaging software (Openlab, Improvision).
3.3. Quantum Yields of One-Photon Photolysis. The one-photon
photolysis quantum yields were determined by irradiating ∼1-2 µM
of caged coumarins in a buffered solution (10 mM Kmops, 100 mM
KCl, pH 7.3) with 365-nm UV light from a mercury lamp (B-100 AP;
UVP, Upland, CA). Durations of UV irradiation were controlled by
an electronic shutter (Uniblitz, Vincient Associates, Rochester, NY).
After each episode of UV illumination, fluorescence emission spectra
of the samples were measured. The excitation wavelength was set at
Acknowledgment. This research was supported by a research
grant (I-1510) from the Welch Foundation and a Career
Development Award from the American Diabetes Associations
to W.-h.L. We thank Dr. Helen Yin for generously allowing us
9
4662 J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004

A New Generation of Caged Fluorophores
A R T I C L E S
to use her two-photon laser systems. We also thank Dr.
Frederick Grinnell for providing primary human fibroblasts.
5 (PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
Supporting Information Available: Synthesis of coumarin
derivatives and HPLC analysis of photolysis products of 4 and
JA036958M
9
J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004 4663