A fluorescent anion sensor that works in neutral aqueous solution for bioanalytical application

Article abstract of DOI:10.1021/ja0175643

Anion recognition and anion sensing are of interest because anions play many important roles in living organisms. Most currently known anion sensors work only in organic solution, but sensors for biological applications are required to function in neutral

Full text of DOI:10.1021/ja0175643

Published on Web 03/21/2002
A Fluorescent Anion Sensor That Works in Neutral Aqueous
Solution for Bioanalytical Application
Shin Mizukami,† Tetsuo Nagano,† Yasuteru Urano,† Akira Odani,‡ and
Kazuya Kikuchi*,§,†
Contribution from the Graduate School of Pharmaceutical Sciences, The UniVersity of Tokyo,
7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan, Research Center for Materials Science,
Graduate School of Sciences, Nagoya UniVersity, Furo-cho, Chikusa-ku,
Nagoya 464-8602, Japan, and PRESTO, JST Corporation, 4-1-8 Honcho,
Kawaguchi 332-0012, Japan
Received November 19, 2001
Abstract: Anion recognition and anion sensing are of interest because anions play many important roles
in living organisms. Most currently known anion sensors work only in organic solution, but sensors for
biological applications are required to function in neutral aqueous solution. We have designed and
synthesized a novel fluorescent sensor for anions. The sensor molecule 1-CdII contains 7-amino-4-
trifluoromethylcoumarin as a fluorescent reporter and CdII-cyclen (1,4,7,10-tetraazacyclododecane) as an
anion host. In neutral aqueous solution, CdII of 1-CdII is coordinated by the four nitrogen atoms of cyclen
and the aromatic amino group of coumarin. When various anions are added to 100 mM HEPES buffer
solution (pH 7.4) containing 1-CdII, the aromatic amino group of coumarin is displaced from CdII, causing
a change of the excitation spectrum. While pyrophosphate and citrate were detected with high sensitivity,
fluoride and perchlorate produced no response. Among organic anions, ATP and ADP gave strong signals,
while cAMP showed little signal. By utilizing the different affinities of the sensor for AMP and cAMP, the
activity of phosphodiesterase, which cleaves cyclic nucleotide, was monitored in real-time. The sensor
should have many biochemical and analytical applications and the sensing principle should be widely
applicable to the sensing of other molecules.
Introduction
and inorganic anions play important roles in living organisms.
Development of anion sensors may also lead to the development
of novel sensors that can detect bioorganic molecules containing
anionic groups intramolecularly. However, most known anion
sensors only function in organic media and it is difficult to use
them in biochemical or physiological experiments.
Fluorescent sensors are useful to analyze and clarify the roles
of biomolecules in living systems; for example, several functions
of intracellular CaII have been elucidated by using fluorescent
CaII indicators.1 Moreover, many other metal sensors and pH
sensors have been developed.2 There is now increasing interest
in anion recognition3 and anion sensing,4 because many organic
Anions are generally larger than cations such as metal ions,
and therefore anions are more subject to solvation than cations.
In organic solvents, it is not so difficult to capture and detect
anions because the solvation energy is relatively small and
electrostatic interactions operate effectively. However, in aque-
ous solvents, which are relevant to biological applications, it is
very difficult to recognize anions because of the strong
hydration. So far, only a few fluorescent anion sensors that work
in aqueous solution have been developed, although many are
known for organic environments.
* To whom correspondence should be addressed. E-mail: kkikuchi@
mol.f.u-tokyo.ac.jp.
† The University of Tokyo.
‡ Nagoya University.
§ JST Corporation.
(1) (a) Grynkiewicz, G.; Poenie, M.; Tsien, R. Y. J. Biol. Chem. 1985, 260,
3440. (b) Minta, A.; Kao, J. P. Y.; Tsien, R. Y. J. Biol. Chem. 1989, 264,
8171.
(2) de Silva, A. P.; Gunaratne, H. Q. N.; Gunnlaugsson, T.; Huxley, A. J. M.;
McCoy, C. P.; Rademacher, J. T.; Rice, T. E. Chem. ReV. 1997, 97, 1515.
(3) (a) Schmidtchen, F. P.; Berger, M. Chem. ReV. 1997, 97, 1609. (b) Kimura,
E.; Koike, T. Chem. Commun. 1998, 1495. (c) Snowden, T. S.; Anslyn, E.
V. Curr. Opin. Chem. Biol. 1999, 3, 740. (d) Bazzicalpi, C.; Bencini, A.;
Bianchi, A.; Cecchi, M.; Escuder, B.; Fusi, V.; Garcia-Espan˜a, E.; Giorgi,
C.; Luis, S. V.; Maccagni, G.; Marcelino, V.; Paoletti, P.; Valtancoli, B. J.
Am. Chem. Soc. 1999, 121, 6807.
(4) (a) De Santis, G.; Fabbrizzi, L.; Licchelli, M.; Poggi, A.; Taglietti, A.
Angew. Chem., Int. Ed. Engl. 1996, 35, 202. (b) Fabbrizzi, L.; Faravelli,
I.; Francese, G.; Licchelli, M.; Perotti, A.; Taglietti, A. Chem. Commun.
1998, 971. (c) Watanabe, S.; Onogawa, O.; Komatsu, Y.; Yoshida, K. J.
Am. Chem. Soc. 1998, 120, 229. (d) Niikura, K.; Metzger, A.; Anslyn, E.
V. J. Am. Chem. Soc. 1998, 120, 8533. (e) Nishizawa, S.; Kato, Y.;
Teramae, N. J. Am. Chem. Soc. 1999, 121, 9463. (f) Miyaji, H.;
Anzenbacher, P., Jr.; Sessler, J. L.; Bleasdale, E. R.; Gale, P. A. Chem.
Commun. 1999, 1723.
A fluorescent anion sensor for use in aqueous solution must
meet two requirements. One is sufficiently strong affinity for
anions in water, and the other is the ability to convert anion
recognition into a fluorescence signal. Most known anion sensors
do not have a sufficiently strong affinity for anions in water,
although they satisfy the latter requirement. Although some
anion hosts can capture anions in aqueous solvent, they are only
host molecules, not sensor molecules. It is difficult to satisfy
both requirements, simultaneously. Czarnik et al. succeeded in
9
3920
J. AM. CHEM. SOC. 2002, 124, 3920-3925
10.1021/ja0175643 CCC: $22.00 © 2002 American Chemical Society

Design and Synthesis of a Novel flUorescent Sensor for Anions
A R T I C L E S
Scheme 1. Synthetic Route to 1 and 1-CdII Complexa
Figure 1. Design concept of sensing anions. Mn+ is a metal ion that can
be chelated stably by cyclen.
the development of fluorescent anion sensors, which work in
aqueous solution.5 The sensing mechanism of these sensors is
based on the anion-binding-induced pKa change of an amino
group adjacent to a fluorophore. However, it has been difficult
to apply this sensing mechanism to biological experiments,
because the sensors have polyamine ligands which chelate metal
ions and change the fluorescence properties by chelating them.
Although a metal ion insensitive sensor is also reported, it is a
sensor for large polyanions such as DNA, not for small
molecular anions.6
In this paper, we present a novel approach to detect anions.
We used a macrocyclic polyamine-metal complex as an anion
host, and a fluorescent aromatic amine as a switch and a
fluorescent reporter. On the basis of this novel anion sensing
mechanism, we developed a practical fluorescent sensor. We
show that this new anion sensor can function in neutral aqueous
solution even containing 100 mM HEPES buffer. Moreover, to
demonstrate the utility of this sensor for bioanalytical applica-
tions, we show that this sensor is applicable for real-time sensing
of phosphodiesterase activity.
a Conditions: (a) TsCl, pyridine. (b) BrCH2CH2Br, Cs2CO3. (c) Con-
centrated H2SO4, 90 °C. (d) Cyclen. (e) Cd(ClO4)2, 60 °C.
(Fuji Silysia Chemical Ltd.) column chromatography. Complex
1-CdII, the CdII complex of 1, was obtained as the perchlorate
salt and recrystallized from 2-propanol/H2O. Details of the
synthetic procedure are described in Supporting Information.
Absorption and Excitation Spectra of 1-Metal Com-
plexes. We measured the absorption and fluorescence spectra
of 1 with ZnII, CdII, CuII, and no metal to find which metal
would fit best with our design concept. These metals are well-
known to be strongly coordinated by cyclen.9 The desired
spectral change was a blue shift of the absorption and the
excitation spectra, because coordination of the 7-amino group
to the metal is required for anion sensing, as shown in Figure
1, and the participation of nitrogen lone pairs in coordinating
the metal would induce such a wavelength shift.
Absorption and excitation spectra of 1 with ZnII, CdII, CuII,
and no metal are shown in Figure 2. We measured these spectra
after sufficient equilibration (more than 3 h), because formation
of these complexes was very slow. The absorption spectral peak
of 1 was at 382 nm, and those of 1 + ZnII, 1 + CdII, and 1 +
CuII were at 388, 342, and 342 nm, respectively. When we added
CdII and CuII to a solution of 1, a blue shift of each absorption
spectrum was observed. This supports the idea that the 7-amino
group coordinates CdII and CuII. The absorption spectrum of 1
+ ZnII was slightly red-shifted. Although this red shift shows
that there is some interaction between ZnII and 7-amino-4-
trifluoromethylcoumarin, the 7-amino group probably does not
coordinate to ZnII, for the reason mentioned above.
The excitation spectra of 1, 1 + ZnII, and 1 + CdII resembled
their absorption spectra, although the fluorescence intensity of
1 + ZnII increased a little. The fluorescence of 1 + CuII
decreased markedly because of the nature of CuII; it is generally
known that CuII ion quenches the fluorescence of various
fluorescent compounds.10
From the principle of anion sensing shown in Figure 1, the
central metal ion needs to be coordinated by five nitrogen atoms
including the 7-amino group. Therefore, if ZnII ion is not
coordinated by the 7-amino group of 1 in neutral aqueous
solution containing 100 mM HEPES, 1-ZnII complex is unable
to act as an anion sensor. Moreover, because the fluorescence
of the 1-CuII complex was strongly quenched, 1-CuII is also
not suitable for a fluorescence sensor. Thus, based on the
Results and Discussion
Design Concept. First, we have to choose an anion host that
can recognize anions in aqueous solution. It is known that the
ZnII-cyclen (1,4,7,10-tetraazacyclododecane) complex is a good
host molecule for anions in neutral aqueous solution, and anions
coordinate to ZnII as the fifth ligand.7 Therefore, we decided to
utilize this complex or an analogous compound. Second, when
we use this complex as the anion host, we have to convert anion
recognition into a fluorescence signal. We thought that if an
intramolecular ligand coordinates to the metal, which is chelated
by cyclen, with a relatively weak affinity, an anion species could
displace the ligand and this event could trigger an optical signal.
So we chose the aromatic amino group of a 7-aminocoumarin
as the fifth ligand. As shown in Figure 1, we expected that the
coordination of an anion to the metal would cause dissociation
of the 7-amino group of 7-amino-4-trifluoromethylcoumarin
from the metal and trigger a change in the fluorescence spectra,
because functional substitution at the 7-position of the coumarin
ring affects the fluorescence spectra of 7-substituted coumarins.8
Preparation of Compounds. The cyclen-conjugated cou-
marin 1 was synthesized in 4 steps from 7-amino-4-trifluoro-
methylcoumarin (Scheme 1), and purified by NH-silica gel
(5) Czarnik, A. W. Acc. Chem. Res. 1994, 27, 302.
(6) Hong, S. Y.; Czarnik, A. W. J. Am. Chem. Soc. 1993, 115, 3330.
(7) (a) Koike, T.; Kajitani, S.; Nakamura, I.; Kimura, E.; Shiro, M. J. Am.
Chem. Soc. 1995, 117, 1210. (b) Kimura, E.; Aoki, S.; Koike, T.; Shiro,
M. J. Am. Chem. Soc. 1997, 119, 3068.
(9) Martell, A. E.; Hancock, R. D. Metal Complexes in Aqueous Solutions;
Plenum Press: New York, 1996; p 98.
(10) Lakowicz, J. R. Principles of Fluorescence Spectrometry, 2nd ed.; Kluwer
Academic/Plenum Publishers: New York, 1999; p 237.
(8) Wheelock, C. E. J. Am. Chem. Soc. 1959, 81, 1348.
9
J. AM. CHEM. SOC. VOL. 124, NO. 15, 2002 3921

A R T I C L E S
Mizukami et al.
Figure 3. Excitation spectra of 5 µM 1-CdII upon addition of sodium
pyrophosphate (0, 0.003, 0.01, 0.03, 0.1, 0.3, 3, 10 mM) in 100 mM HEPES
buffer (pH 7.4) at 25 °C. The emission wavelength was 500 nm.
14.3, respectively.11 If the coordination forms of 1-metals are
the same as those of cyclen, log K(1-ZnII) might be expected
to be larger than log K(1-CdII), but this is not the case; namely,
the order of relative stability of the 1-metal complexes was
reversed as compared to the cyclen-metal complexes. This
result suggests that the coordination form of the 1-CdII complex
is different from that of the 1-ZnII complex. We consider that
ZnII is not coordinated by the aromatic amino group: water
molecules may prevent the aromatic amino group from coor-
dinating to ZnII even in the absence of anions that bind ZnII
strongly. On the other hand, the results are consistent with the
idea that CdII is chelated by five nitrogen atoms, i.e., the four
amino groups of cyclen and the aromatic amino group. Thus,
the pH titration results also suggest that the 1-CdII complex meet
the design principle in Figure 1.
Figure 2. (a) Absorption spectra and (b) excitation spectra (emission
wavelength is 500 nm) of 1 with various metals. The concentration of 1
was 5 µM. All metals were added at 1.2 equiv with respect to 1. Solutions
were buffered with 100 mM HEPES (pH 7.4).
Fluorescence Detection of Anions with 1-CdII. We added
various anions to a solution of synthesized 1-CdII (5 µM) to
investigate what effects various anions have on the excitation
spectra. All spectra were measured in the presence of 100 mM
HEPES buffer (pH 7.4) at 25 °C.
Table 1. Stability Constants log âpqr for MpLqHr Systems
(M ) ZnII, CdII, L ) 1)a
M
p
q
r
log âpqr
CdII
1
1
1
1
1
1
0
0
1
1
1
1
1
1
1
1
0
11.38 ( 0.05
4.46 ( 0.06
-3.95 ( 0.08
10.80 ( 0.08
3.87 ( 0.09
4.61 ( 0.12
9.67 ( 0.01
18.06 ( 0.01
-1
-2
0
-1
-2
1
When sodium pyrophosphate (PPi) solution was added to a
solution of 1-CdII, the excitation spectrum shifted dose-
dependently toward longer wavelength. As shown in Figure 3,
the λmax shifted from 342 to 383 nm after addition of 10 mM
PPi. The excitation spectrum of 1-CdII + 10 mM PPi resembled
that of the free base 1 (Figure 2); for instance, both peak
wavelengths were about 380 nm. As described in the Design
Concept section, in the case of 7-aminocoumarins, a batho-
chromic shift of the excitation spectrum indicates that the
electron density of the 7-amino group has increased.8 These
results indicate that the labile fifth ligand (the aromatic amino
group) was dose-dependently replaced by PPi anion, as we
expected (Figure 1). In short, in neutral aqueous solution, PPi
has a larger affinity for the CdII-cyclen complex than the
aromatic amino group has, even though the amino group is
located intramolecularly.
ZnII
H+
2
a âpqr ) [MpLqHr]/[M]p[L]q[H]r. Conditions: 0.1 M Et4NClO4 (I ) 0.1),
25 °C. [L] ) 0.5 mM, 0.17 mM.
absorption and excitation spectra, we concluded that 1-CdII
would be the best fluorescent anion sensor based on our design
principles, as shown in Figure 1.
Potentiometric pH Titration. Next, we carried out poten-
tiometric pH titration experiments. We measured the pH titration
curves of 1‚4HClO4, 1‚4HClO4 + Zn(ClO4)2, and 1‚4HClO4
+ Cd(ClO4)2, and calculated the ligand protonation constants
and the metal complexation constants from these data by using
a computer program. The results are shown in Table 1. The
protonation equilibrium constants of 1 were 9.67 (log K1 ) log
Next, we examined the effects of other anions. Some other
anions caused spectral changes of 1-CdII and the apparent
dissociation constant (Kd) values are shown in Table 2. These
Kd values were calculated by fitting the fluorescence intensity
changes as described in the Experimental Section. As we
expected, citrate showed strong affinity. Phosphate had a much
â011) and 8.39 (log K2 ) log â012 - log â011). The other lower
pKa values could not be calculated.
The stability constants of 1 for ZnII and CdII, which are shown
as log â110 in Table 1, were calculated; the log K(1-ZnII) and
log K(1-CdII) values are 10.80 and 11.38, respectively. The
reported log K(cyclen-ZnII) and log K(cyclen-CdII) are 16.2 and
(11) Kodama, M.; Kimura, E. J. Chem. Soc., Dalton Trans. 1977, 2269.
9
3922 J. AM. CHEM. SOC. VOL. 124, NO. 15, 2002

Design and Synthesis of a Novel flUorescent Sensor for Anions
A R T I C L E S
Table 2. Apparent Dissociation Constants (Kd) of Sensor 1-CdII
for Anions in 100 mM HEPES Buffer (pH 7.4)a
anion
K (M)
d
anion
K (M)
d
pyrophosphate
7.5 × 10-5
9.0 × 10-5
9.2 × 10-3
1.5 × 10-2
3.2 × 10-2
9.0 × 10-2
b
ATP
1.4 × 10-5
2.6 × 10-5
4.8 × 10-5 c
4.4 × 10-4
1.7 × 10-3
1.9 × 10-3
b
citrate
ADP
GMP
AMP
UMP
CMP
cAMP
I-
phosphate
Br-
Cl-
F-
ClO4
-
b
a All anions were added as sodium or potassium salts. All Kd values
were calculated under the condition with 100 mM HEPES buffer (pH 7.4).
b Kd was too large to be calculated. c This value is smaller than the real Kd,
because dynamic quenching was observed besides static quenching.
weaker affinity than PPi, probably because the protonated forms,
HPO42- and H2PO4-, are dominant species in aqueous solution
of pH 7.4 and they have smaller numbers of negative charges
than PPi. Perchlorate did not cause any change even at 100 mM.
With regard to oxoacids, multivalent anions bound to 1-CdII
more strongly than mono- or divalent anions. These data suggest
that anions interact with 1-CdII through multipoint recognition,
probably with both the metal and NH protons of the cyclen
ring.
For halogen ions, the spectral changes occurred in the order
I- > Br- > Cl- . F-, reflecting the degree of soft basicity.
This is consistent with the fact that CdII is a soft acid and has
a high affinity for soft bases.
We next examined some nucleotides, because they are
negatively charged organic compounds with key roles in living
organism, especially cyclic nucleotides such as cAMP and
cGMP work as intracellular second messengers. Therefore,
sensing of various nucleotides is very important. First, four
nucleoside monophosphates, AMP, GMP, CMP, and UMP, were
added to a solution of 5 µM 1-CdII. Interestingly, only GMP
quenched the fluorescence and the other three nucleotides caused
a red shift of the excitation spectrum, like inorganic anions
(Figure 4a). The quenching mechanism of the guanine nucleotide
is probably based on photoinduced electron transfer. The
apparent dissociation constants are shown in Table 2. These
results suggest that both the shape of the excitation spectrum
and the dissociation constants are dependent on the type of
nucleobase. One possible explanation is that coumarin interacts
directly with the base of nucleotides.
Although ADP and ATP had higher affinities for 1-CdII than
AMP, there was no large difference about binding strength
between ADP and ATP. Two cyclic nucleotides, cAMP and
cGMP, had lower affinities for 1-CdII than their hydrolysis
products, AMP and GMP, respectively. The spectral data for
cAMP and AMP are shown in Figure 4b. When even 10 mM
cAMP was added, the excitation spectral change of 1-CdII was
small. This spectrum closely resembled the spectrum upon
addition of 100 µM AMP. Addition of less than 1 mM cAMP
scarcely changed the excitation spectrum of 1-CdII. Considering
these results, the difference in binding strength between AMP
and cAMP was probably based on the number of anionic sites.
Reversibility of Anion Sensing. We examined the revers-
ibility of anion sensing. If the sensing system is reversible,
depletion of the anion that coordinates CdII must produce a blue
shift of the excitation spectrum, causing it to revert to the
original spectrum. An excess amount of MgII was added to this
Figure 4. Comparison of the excitation spectra of 5 µM 1-CdII upon
addition of (a) AMP, GMP, CMP, and UMP and (b) AMP and cAMP. The
concentration of each nucleotide was 10 mM.
PPi-containing solution to bind PPi and prevent PPi from
coordinating CdII. When 10 mM Mg(ClO4)2 was added to a
solution of 1-CdII containing 1 mM PPi, the excitation spectrum
was indeed blue-shifted and the λmax was at 345 nm (Figure
5a). This spectrum almost exactly coincided with the spectrum
of 1-CdII containing no PPi.
Next, we examined the effect of hydrolyzing pyrophosphate.
Inorganic pyrophosphatase is known to convert one pyrophos-
phate ion to two phosphate ions.12 Because phosphate coordi-
nates 1-CdII more weakly than does PPi, cleavage of PPi should
induce a spectral change similar to that seen with a large amount
of MgII ion. When 0.33 unit/mL of inorganic pyrophosphatase
and 0.5 mM Mg(ClO4)2 were added to 1-CdII solution containing
1 mM pyrophosphate at pH 7.4, the peak of the excitation
spectrum quickly shifted toward short wavelength (Figure 5b).
These results indicate that this sensing mechanism is reversible.
Phosphodiesterase Assay. To examine the potential of this
system for biochemical applications, we tried to monitor the
activity of phosphodiesterase (PDE). We used phosphodiesterase
3′:5′-cyclic mononucleotide,13 which catalyzes the conversion
of cyclic nucleotide to nucleoside monophosphate by cleavage
of the phosphodiester of the cyclic nucleotide, e.g., converting
cAMP to AMP (Scheme 2). Such enzymes participate in
(12) (a) Moe, O. A.; Butler, L. G. J. Biol. Chem. 1972, 247, 7308. (b)
Cooperman, B. S.; Chiu, N. Y.; Bruckmann, R. H.; Bunick, G. J.; McKenna,
G. P. Biochemistry 1973, 12, 1665.
(13) Ho, H. C.; Teo, T. S.; Desai, R.; Wang, J. H. Biochim. Biophys. Acta 1976,
429, 461.
9
J. AM. CHEM. SOC. VOL. 124, NO. 15, 2002 3923

A R T I C L E S
Mizukami et al.
Figure 6. Phosphodiesterase activity assay. (a) Change of the excitation
spectra of 1-CdII. The emission wavelength was 500 nm. (b) Fluorescence
intensity vs time plot. The excitation/emission wavelengths were 380 nm/
500 nm.
Figure 5. (a) Excitation spectra of 1-CdII with 1 mM Na4P2O7 (no symbol)
and 1 mM Na4P2O7 + 10 mM Mg(ClO4)2 (open circle). (b) Excitation
spectra of 1-CdII with 1 mM Na4P2O7 before (no symbol) and 20 min after
(closed circle) addition of inorganic pyrophosphatase. The concentration
of 1-CdII was 5 µM. The emission wavelength was 500 nm.
3′:5′-cyclic nucleotide activity. This approach is expected to be
applicable to the assay of other enzyme activities.
Scheme 2
Summary and Conclusion
In conclusion, we have developed a novel fluorescent sensor
for anions, which works in neutral aqueous solution. The sensing
principle is novel and based on coordination chemistry. The
sensor 1-CdII has a CdII ion as the central metal of the anion
host region, and CdII is critical to convert anion recognition to
a fluorescence signal. When we used ZnII and CuII instead of
CdII, we could not obtain desirable characteristics for anion
sensing. On the basis of spectral and pH potentiometric data,
we concluded that CdII was effective because it is weakly
coordinated by the 7-amino group of coumarin, so that some
anions could displace it and coordinate to the CdII ion in its
place. We confirmed that this sensing mechanism was reversible.
We proved that the activity of phosphodiesterase, which
hydrolyzes the phosphodiester bond of cyclic nucleotides, could
be monitored by using 1-CdII. No other known sensor has this
characteristic, so 1-CdII could become the preferred anion sensor
in many biological and analytical applications. The same sensing
principle should also be widely applicable to the sensing of other
molecules.
intracellular signal transduction, so the development of sensitive
detection systems for their activity14 is very important.
The developed anion sensor 1-CdII recognized AMP much
more strongly than cAMP. Therefore, we tried real-time
detection of PDE’s activity by monitoring the increase of AMP.
We measured the excitation spectrum (emission wavelength/
500 nm) several times after addition of 2 U/mL (final concen-
tration) of PDE to 5 µM 1-CdII solution containing 100 mM
HEPES, 10 mM cAMP, 10 µM CaCl2, and 0.02 U/mL of PDE
activator at pH 7.4. As shown in Figure 6a,b, the fluorescence
intensity (excitation wavelength, 380 nm/emission wavelength,
500 nm) gradually increased. This spectral change shows the
conversion from cAMP to AMP. These results show 1-CdII
is useful for real-time fluorescence assay of phosphodiesterase
Experimental Section
Measurement of Absorption and Excitation Spectra. All spectra
of 1 or 1-CdII were measured at 5 µM concentration and 25 °C, using
a UV-1600 UV-visible spectrometer (Shimadzu, Kyoto, Japan). The
fluorescence spectrometer was an F-4500 (Hitachi, Tokyo, Japan).
(14) (a) Rutten, W. J.; Schoot, B. M.; De Pont, J. J. H. H. M. Biochim. Biophys.
Acta 1973, 315, 378. (b) Thompson, W. J.; Appleman, M. M. Biochemistry
1971, 10, 311. (c) Hiratsuka, T. J. Biol. Chem. 1982, 257, 13354.
9
3924 J. AM. CHEM. SOC. VOL. 124, NO. 15, 2002

Design and Synthesis of a Novel flUorescent Sensor for Anions
A R T I C L E S
Determination of Apparent Dissociation Constants (Kd) of Sensor
1-CdII for Several Anions. The final concentration of compound 1-CdII
was 5 µM. All the solutions contained 100 mM HEPES buffer (pH
7.4). Fluorescence values at the peak of the excitation spectra were
plotted against the concentration of anions and fitted to the following
equation:
(cAMP), 10 µM CaCl2, 5 µM 1-CdII, 2 U/mL of phosphodiesterase
3′:5′-cyclic nucleotide activator, and 0.02 U/mL of phosphodiesterase
3′:5′-cyclic nucleotide (PDE). The total volume was 3 mL. Excitation
spectra were measured with emission at 500 nm. The reaction
temperature was 30 °C.
Acknowledgment. The authors thank Professor Donald
Hilvert for helpful discussions. We thank Prof. K. Yamaguchi
and S. Sakamoto of Chiba University for HRMS measurement.
This work was supported in part by the Ministry of Education,
Science, Sports and Culture of Japan (Grants 11794026,
12470475, 12557217, 11771467, and 12045218), by the
Mitsubishi Foundation, and by the Research Foundation for
Opt-Science and Technology. S.M. thanks the Japan Society
for the Promotion of Science for a JSPS Research Fellowship
for Young Scientists. K.K. is grateful for financial support
from Nissan Science Foundation, Shorai Foundation of Science
and Technology, and Uehara Memorial Foundation.
F ) (F0 + Fmax[A]/Kd)/(1 + [A]/Kd)
where F0 is the initial F value without anions, Fmax is the maximum F
value, and [A] is the final concentration of the anion added to the
solution.
Potentiometric pH Titration. The potentiometric pH titration
experiments were carried out with an ionic strength, I, of 0.1 M
(Et4NClO4) at 25 °C. We conducted titration experiments at 0.17 mM
and 0.5 mM 1:1 metal-1 systems. The detailed procedure has been
described elsewhere.15 A ligand protonation constant, metal complex-
ation constants, and distribution diagrams were calculated with use of
the computer programs SUPERQUAD16 and MINIQUAD.17
Phosphodiesterase Assay. The reaction solution contained 100 mM
HEPES (pH 7.4), 10 mM adenosine 3′:5′-cyclic monophosphate
Supporting Information Available: Supplemental procedures
for materials and the synthetic procedure of 1 and 1-CdII and
characterization data for the synthetic compounds (PDF). This
material is available free of charge via the Internet at
http://pubs.acs.org.
(15) Odani, A. Physical Methods in Supramolecular Chemistry; Comprehensive
Supramolecular Chemistry, Vol. 8; Davies, J. E. D., Ripmeester, J. A.,
Eds.; Pergamon: New York, 1996; p 445.
(16) Gans, P.; Sabatini, A.; Vacca, A. J. Chem. Soc., Dalton Trans. 1985, 1195.
(17) Sabatini, A.; Vacca, A.; Gans P. Talanta 1974, 21, 53.
JA0175643
9
J. AM. CHEM. SOC. VOL. 124, NO. 15, 2002 3925