Novel reversible Zn2+-assisted biological phosphate turn-on probing through stable aryl-hydrazone salicylaldimine conjugation that attenuates ligand hydrolysis

Article abstract of DOI:10.1021/ic4013526

A novel reversible zinc(II) chemosensing ensemble (2·Zn ²⁺) allows for selective turn-on fluorescence sensing of ATP and PPi in aqueous media (detection limits: 2.4 and 1.0 μM, respectively) giving selective binding patterns: ATP ～ PPi > ADP AMP > monophosphates ≈ remaining ions tested. The conjugated hydrazone [C=N - NH - R] resists hydrolysis considerably, compared to the imine [C=N - CH₂ - R, pyridin-2-ylmethanamine] functionality, and generalizes to other chemosensing efforts. Prerequisite Zn²⁺·[O _phenolN_imineN_pyr] binding is selective, as determined by UV-vis and NMR spectroscopy; ATP or PPi extracts Zn²⁺ to regenerate the ligand-fluorophore conjugate (PPi: turn-on, 512 nm; detection limit, 1.0 μM). Crystallography, 2-D NMR spectroscopy, and DFT determinations (B3LYP/631g) support the nature of compound 2. 2-Hydrazinyl-pyridine- salicylaldehyde conjugation is unknown, as such; a paucity of chemosensing-Zn²⁺ binding reports underscores the novelty of this modifiable dual cation/anion detection platform. A combined theoretical and experimental approach reported here allows us to determine both the potential uniqueness as well as drawbacks of this novel conjugation.

Full text of DOI:10.1021/ic4013526

Article
pubs.acs.org/IC
Novel Reversible Zn2+-Assisted Biological Phosphate “Turn-On”
Probing through Stable Aryl-hydrazone Salicylaldimine Conjugation
That Attenuates Ligand Hydrolysis
Olga G. Tsay,† Sudesh T. Manjare,†,‡ Hyungjun Kim,† Kang Mun Lee,† Yoon Sup Lee,†
and David G. Churchill*,†
†Department of Chemistry, Korea Advanced Institute of Science and Technology (KAIST), 373-1 Guseong-dong, Yuseong-gu,
Daejeon, 305-701, Republic of Korea
‡Center for Catalytic Hydrocarbon Functionalizations, Institute for Basic Science (IBS), 373-1 Guseong-dong, Yuseong-gu, Daejeon,
305-701, Republic of Korea
S
* Supporting Information
ABSTRACT: A novel reversible zinc(II) chemosensing
ensemble (2·Zn2+) allows for selective “turn-on” fluorescence
sensing of ATP and PPi in aqueous media (detection limits:
2.4 and 1.0 μM, respectively) giving selective binding patterns:
ATP ∼ PPi > ADP ≫ AMP > monophosphates ≈ remaining
ions tested. The conjugated hydrazone [CNNHR]
resists hydrolysis considerably, compared to the imine [C
NCH2R, pyridin-2-ylmethanamine] functionality, and
generalizes to other chemosensing efforts. Prerequisite Zn2+·[OphenolNimineNpyr] binding is selective, as determined by UV−vis
and NMR spectroscopy; ATP or PPi extracts Zn2+ to regenerate the ligand−fluorophore conjugate (PPi: turn-on, 512 nm;
detection limit, 1.0 μM). Crystallography, 2-D NMR spectroscopy, and DFT determinations (B3LYP/631g*) support the nature
of compound 2. 2-Hydrazinyl-pyridine-salicylaldehyde conjugation is unknown, as such; a paucity of chemosensing-Zn2+ binding
reports underscores the novelty of this modifiable dual cation/anion detection platform. A combined theoretical and
experimental approach reported here allows us to determine both the potential uniqueness as well as drawbacks of this novel
conjugation.
biology. Because of the biological importance of phosphates in
living organisms, they remain the focus of various research. In
recent years various detection methods have been developed;
emergent and improved modalities are in high demand.9
Fluorescence chemosensors can be designed to offer high
sensitivity and rapid response to such analyte concentrations.
They allow for “real-time” detection in genuine biological
samples.10 Among the various detection modes utilized for
phosphate sensing, certain metal ion complexes are considered
as promising molecular analytical platforms for chemical
recognition under aqueous conditions.11,12 In this context,
Zn2+-based sensors have been developed: they mimic metal-
loenzymes in their active site recognition.13−16 Despite the
intensity of this research area, few papers featuring fluorescence
BODIPY−chelator·Zn2+ ensembles for detection of phosphates
have been reported and utilized (BODIPY = boron
difluorodipyrromethene, Figure 1).17 Among the relevant
known reports, a doubly substituted BODIPY with a bipyridyl
unit for zinc chelation was designed. It was found to be
sensitive to inorganic phosphate, but other biological
polyphosphates were not tested.17a Dipicolylamine (DPA)
I. INTRODUCTION
Phosphates are ubiquitous in biology and play significant roles
in a range of essential biological processes. They have
bioenergetic roles and are involved in genome stability, lipid
stability, recognition of organelle membranes, skeletal struc-
tures, protein regulation, cell signaling, and many other
processes and events.1 Nucleoside polyphosphates (e.g.,
adenosine triphosphate, ATP, and adenosine diphosphate,
ADP) constitute an important class of biological phosphates;
adenosine triphosphate (ATP) is a molecule whose concen-
tration signifies “energy currency” of intracellular energy
transfer in cells and is involved in various cellular processes,2
such as the synthesis of macromolecules (including DNA and
RNA),3 protein transport,4 ion channel regulation,5 and both
intra- and extracellular signaling.6 On the other hand,
pyrophosphate (PPi) is the product of ATP hydrolysis and
an essential anion for normal cell function. For example, PPi is
involved in energy transduction mechanisms and participates in
various enzymatic reactions.7 Also, elevated PPi concentration
in the synovial fluid is associated with calcium pyrophosphate
dihydrate (CPPD) crystal deposition disease, and a condition
termed chondrocalcinosis.8 The roles that biological phos-
phates play in tubulin microtubule formation are also extremely
important regarding macromolecular self-assembly essential in
Received: May 29, 2013
Published: August 14, 2013
© 2013 American Chemical Society
10052
dx.doi.org/10.1021/ic4013526 | Inorg. Chem. 2013, 52, 10052−10061

Inorganic Chemistry
Article
Figure 1. Structures from previous reports of BODIPY-based Zn2+ ensembles for phosphate sensing by, e.g., Akkaya, Hamachi, and Hong17 and the
structure pursued in the present work.
Figure 2. Differences in hydrolysis between imine and hydrazone groups.
unit was also incorporated into a BODIPY-based PPi sensor
which operates via chemiluminescence that is generated
electrically.17b Also, a fluorescent BODIPY-based Zn2+ complex
was successfully utilized for the selective detection of
phosphorylated tau protein in histological samples.17c Herein,
we report a new BODIPY-Zn2+ ensemble for the detection of
phosphates under biologically relevant conditions; a novel
chelation pocket helps balance (i) cation/anion selectivity and
(ii) imine hydrolysis tendencies (Figures 1 and 2). The probe
was found to be sensitive to ATP and PPi when screened with
other competing anion and phosphate analytes; the chelator
pocket itself shows reasonable selectivity to Zn2+. BODIPY−
Zn2+ ensemble formation is possible in the presence of other
interfering metal ions. Due to the strong affinity of phosphates
to zinc metal ion, extraction may be effected, and fluorescence
recovery, as well as probe reversibility, can be studied.9g There
are a limited number of systems for the reversible detection of
biophosphates.12a,15a Full characterization of the ligand has
been made, and the examination of the MO patterns of this
substrate via DFT geometry optimization calculations reveals
important patterns pertinent to future receptor design.
Previous metal binding motifs have been explored but with
limited expressed emphasis on chemosensing,18 Zn2+ binding,19
or fluorescence. The choice of hydrazinopyridine over the
ubiquitous picoline-based chelations allows for the study of
possible differences in solubility and hydrolysis rates.21 The
design involves a single Zn2+ center and a [OphenolNimineNpyr]
chelation pocket along with a hydrazone linkage allowing for
“greater intrinsic hydrolytic stability” via resonance stabilization
(see Figure 2).20
II. EXPERIMENTAL SECTION
Materials and Physical Measurements. All chemicals used
herein [e.g., 2,4-dimethylpyrrole, 2,3-dichloro-5,6-dicyano-1,4-benzo-
quinone (DDQ), BF3·OEt2, MnO2, 2-hydrazinopyridine) were of
analytical grade and used as received from commercial suppliers
(Aldrich, TCI, and Junsei chemical companies]. All solvents used for
spectral measurements were of analytical or HPLC grade. BODIPY
compound 1 was prepared as previously reported.21 Analytical thin
layer chromatography (TLC) was performed on cuts of original plates
10053
dx.doi.org/10.1021/ic4013526 | Inorg. Chem. 2013, 52, 10052−10061

Inorganic Chemistry
Article
Scheme 1. Synthesis for Compound 2
Figure 3. Front view of compound 2 as a thermal ellipsoid drawing for the crystallographic determination (for atomic numbering, see SI). “Washout”
graphics of side- and top-views of the same molecular structure. Thermal ellipsoids are drawn at 30%. All hydrogen atoms are omitted for clarity.
Monoclinic, P21/n, a = 10.814(2) Å, b = 14.744(3) Å, c = 16.812(3) Å, β = 96.75(3)°, V = 2662.0(9) Å3, Z = 4.
(20 × 20 cm2, 60 F silica gel, Merck). A Vilber Lourmat-4LC UV lamp
was used to probe the fluorescence of reaction “spots” assayed by TLC
(lamp: 4 W, 365 nm, 50/60 Hz). Silica gel used in column
chromatography was of diameter 0.040−0.063 mm (Merck). Solvents
used in NMR spectral analyses were purchased commercially and were
5.01 Hz, 1H, H22), 7.93 (s, 1H, H17), 7.66 (dd, 3JH−H = 8.4 Hz, 4JH−H
=
7.2 Hz, 1H, H20), 7.15 (dd, 3JH−H = 8.1 Hz, 4JH−H = 2.24 Hz, 1H, H16),
3
7.09−7.12 (m, 3H, H12,15), 7.04 (d, JH−H = 8.4 Hz, 1H, H19), 6.85
3
4
(dd, JH−H = 7.2 Hz, JH−H = 4.9 Hz, 1H, H21), 6.00 (s, 2H, H2, H6),
2.49 (s, 6H, H10), 1.48 (s, 6H, H9). 13C NMR spectral data (100 MHz,
CD2Cl2: 54 ppm): δ 158.1 (C14), 155.8 (C18), 155.7 (C3,5), 148.4
(C22), 143.7 (C7a), 143.7 (C8a), 142.2 (C17), 141.6 (C13), 138.9 (C20),
132.2 (C1,7), 130.4 (C16), 129.7 (C12), 126.3 (C11), 121.5 (C2,6), 119.6
(C8), 117.9 (C15), 117.0 (C21), 106.8 (C19), 14.9 (C10), 14.7 (C9).
HRMS (ESI) Calcd for C25H24BF2N5O: 459.2042. Found: m/z
460.2115 (M + H)+.
1
of spectroscopy grade. H and 13C NMR spectra were recorded on a
Bruker Avance 400 MHz spectrometer with TMS used as an internal
standard. UV spectra were recorded with JASCO V-503 UV−vis
spectrometer (1000 nm min−1). Emission spectra were obtained using
a SHIMADZU RF-5310 pc spectrofluorophotometer. HRMS data
were obtained with a micro-TOF-QII (Bruker, Daltonik). Single-
crystal X-ray data were collected on a Bruker P4 diffractometer
equipped with a SMART CCD detector. The molecular structure of
compound 2 was elucidated using direct methods; standard difference
map techniques were used to determine the crystal structure. Full-
matrix least-squares refinement procedures were performed on F2
values with the SHELXTL program (version 5.10).
Theoretical Calculations. A computational study was undertaken
to aid the understanding of the HOMO−LUMO arrangement in
compound 2 and its derivatives. Density functional theory (DFT)
calculations were carried out with the Gaussian 09 program. The
molecular structure of 2 was optimized using the B3LYP method
together with the 631g* basis set; where applicable, the zinc atom was
treated separately with the sdd basis set. The input coordinates for gas
phase optimizations were obtained from the crystal structure
coordinates. HOMO and LUMO energy levels of the molecule
suggested that electron density is present over the core BODIPY
molecule; for HOMO − 1 and LUMO + 1, electron density existed
over the remainder of the molecule including the PET (photoinduced
electron transfer) donor 2-hydrazinyl-pyridine-salicylaldehyde moiety.
Determination of Binding Constant between 2 and 2·Zn2+.
Binding constant was estimated by using the following equations:30
Details for UV−Vis and Fluorescence Measurements. Stock
solutions of 2 (1.0 mM) were prepared in absolute methanol. The 0.01
M stock solutions of metal ions (Na+, K+, Cs+, Mg2+, Ca2+, Mn2+, Fe2+,
Co2+, Ni2+, Cu2+, Ag+, Zn2+, Cd2+, Hg2+, Pb2+, and Al3+) were prepared
in distilled water as their perchlorate salts. Stock solutions of anions as
−
sodium salts (F−, Cl−, Br−, I−, CH3COO−, CN−, SO42−, HCO3 ,
−
−
−
NO3 , CO32−, H2PO4 , HPO4 , P2O74−) (0.01 M) and ATP, ADP,
and AMP (adenosine monophosphate) (0.01 M) were prepared in
distilled water. Typically, for absorption and emission trials, test
solutions were prepared by placing 15.0 μL of stock solution of 2 into
a cuvette and diluting the solution to 3.0 mL with methanol/HEPES
buffer solution (2:1, 0.1 M, pH = 7.2). To a solution of 2, 10.0 equiv of
metal ions were added, and emission spectra were then recorded
immediately in which the excitation wavelength was set to that for the
absorption maxima. Excitation and emission slit widths were 3.0 nm.
Synthesis of Compound 2. To a solution of 1 (0.04 g, 0.11
mmol) in CH2Cl2 (10 mL) was added 2-hydrazinopyridine (0.024 g,
0.22 mmol). The reaction mixture was refluxed for 2 h and monitored
via TLC. The solvent was then removed under reduced pressure; the
solid was then washed with a small amount of CH2Cl2 and methanol.
Recrystallization from CH2Cl2/CH3OH gave 2 (0.04 g) as a red
I
I0
= 1 + Ka[Zn]
(1)
⎡
⎢
⎢
⎣
⎛
⎜
⎝
⎞
⎟
⎠
1
2
1
Ka
[Zn] =
Zn − H −
0
0
⎤
⎥
⎥
⎦
2
⎛
⎜
⎝
⎞
⎟
⎠
4Zn0
Ka
1
Ka
+
Zn − H −
+
0
0
(2)
powder in 80% yield. 1H NMR spectral data (400 MHz, CD2Cl2: 5.30
Here, H0 is the total concentration of 2 which is considered to be
constant throughout the duration of the titration; Zn0, total
3
ppm): δ 10.9 (br s, 1H, OH), 8.86 (br s, 1H, NH), 8.17 (d, JH−H
=
10054
dx.doi.org/10.1021/ic4013526 | Inorg. Chem. 2013, 52, 10052−10061

Inorganic Chemistry
Article
Figure 4. (a) Emission spectra of 2 (5.0 μM) upon addition of various concentrations of Zn2+ (0−25 μM). Inset: plot of emission intensity at 512
nm versus [Zn2+]. (b) Competitive experiment of 2 (5.0 μM) with different metal ions in aqueous solutions (CH3OH/HEPES buffer, 2:1, pH 7.2);
λemis = 512 nm. [Mn+] = 50 μM; [Zn2+] = 5 μM. λex = 500 nm.
Figure 5. (a) Emission spectra of 2·Zn2+ (5.0 μM) with various anions, including PPi, AMP, ADP, and ATP (50.0 μM). (b) Fluorescence titrations
(in triplicate, STD) of 2·Zn2+ (5.0 μM) with PPi, AMP, ADP, and ATP (0−100.0 μM). (c) Fluorescence intensity enhancement (I/I0) of 2·Zn2+
(5.0 μM) at 512 nm in the presence of different anions and AMP, ADP, and ATP (50.0 μM) in CH3OH/HEPES buffer (2:1, pH 7.2). I0 is the
emission intensity of 2·Zn2+. (d) Emission intensity at 512 nm of 2·Zn2+·ATP in the presence of different anions: [2·Zn] = 5.0 μM, [ATP] =
[anions] = 50.0 μM. Conditions: CH3OH/:HEPES buffer (2:1, pH 7.2); λem = 512 nm, λex = 500 nm.
concentration of added Zn2+; Ka, binding constant; [Zn], concen-
III. RESULTS AND DISCUSSION
tration of the free Zn2+ in the equilibrium.
Synthesis. BODIPY 2 was synthesized from the previously
Determination of Stoichiometry and Binding Constant for
ATP and PPi Binding with 2·Zn2+. The interaction between 2·Zn2+
reported compound 121 through the addition of 2-hydrazino-
pyridine in methylene chloride under reflux (Scheme 1).
Detailed synthetic procedures and characterization data are
provided in the Experimental Section. Also, a single-crystal X-
and polyphosphates was analyzed according to the Benesi−Hildebrand
equation for spectrofluorometric titration:28
ray structure was obtained (Figure 3). Compound 2 was
I0
I0 − I
B
1
= A +
characterized by H and 13C NMR spectroscopy and ESI-
Ka[analyte]2
HRMS.
Compound 2 was found in the P21/n space group in a crystal
grown from a CH2Cl2/CH3OH solvent mixture. There is no
solvent of crystallization. Structurally, the expected 2-hydroxy-
2-(2-pyridinyl)hydrazone benzaldehyde-based meso group is
present: the meso group−BODIPY rotor angle was found to be
∼89°. The hydrazone group is effectively coplanar (non-H
Here, Ka, I0, and I are, respectively, the binding constant, the observed
fluorescence intensity (in the absence of analyte), and the observed
fluorescence intensity (in the presence of analyte). Binding constants
were estimated from titration experiments as the ratio between the y-
intercept and the slope of the linear “best fit.”
10055
dx.doi.org/10.1021/ic4013526 | Inorg. Chem. 2013, 52, 10052−10061

Inorganic Chemistry
Article
Figure 6. 1H NMR spectrum for (a) 2 in DMSO; (b) 2 in the presence of 1.2 equiv of Zn2+, to give 2·Zn2+·nH2O; and (c) 2 in the presence of 1.2
equiv Zn2+ and 2.0 equiv of PPi (relative to Zn2+) to give 2·Zn2+·PPi. [2] = 0.02 M.
mean-plan deviation: 0.16). The imine C17−N bond length is
1.28 Å; the N−C18 bond length is 1.38 Å. The N−N bond
length is 1.37 Å; the C13−C17 bond length is 1.46 Å, consistent
with element−element single bonds. The ortho-Npy is rotated
distal from the hydroxyl group. Also, the entire 8-substituent
has a bent countour (Figure 3). The deviation in the BODIPY
mean plane is minimal (0.024 Å); alternate CC and CC
bond lengths are clearly evident.
signifying the formation of the 2·Zn2+ complex (Figure 4). For
1:1 binding stoichiometry, a binding constant of Ka = 1.1 × 107
M−1 was obtained (Figure S12). The formation of the zinc
complex was also tracked by mass spectrometry (HRMS)
(Figure S6). The peak at 522.1260 corresponds to a 1:1
complex formation [C25H24BF2N5O + Zn]+, consistent with the
calculated m/z mass value of 522.1255.
Sensing of Polyphosphates with a BODIPY−Zn2+
Ensemble. As the 2·Zn2+ ensemble is weakly fluorescent, the
sensing ability of this system toward various anions, including
biological phosphates, was evaluated by further fluorescence
spectroscopic measurements. The zinc complex 2·Zn2+ (5.0
μM) was prepared in situ and treated with various anions
Formation of BODIPY−Zn2+ Ensemble. Compound 2
has moderate fluorescence under aqueous conditions (ΦF =
0.27 0.01;22 CH3OH/HEPES buffer, 2:1, pH 7.2). However,
in the presence of Zn2+ (5.0 μM), the fluorescence of 2 drops
dramatically (ΦF = 0.038 0.002) (Figure 4). Other metal ions
(Na+, K+, Cs+, Mg2+, Ca2+, Mn2+, Fe2+, Co2+, Ni2+, Cu2+, Ag+,
Zn2+, Cd2+, Hg2+, Pb2+, Al3+) were also examined. A decrease of
emission intensity was observed in the presence of Co2+, Cu2+,
Cd2+, and to a lesser extent Hg2+, Pb2+, Fe2+, Ni2+ (Figure S10).
While these reductions in fluorescence were not as dramatic as
those found for Zn2+ they cannot be ignored as potential
competitors for the Zn2+. After equilibration was established in
these standard competntiion-type reactions, when Zn2+ and a
competitor both show a decrease in fluorescence intensity, it
requires careful work to determine the extent of mixing. In the
worst case, such interference may be inconclusive. Interestingly,
Cd2+, a common interferent in Zn2+ sensing, gives only a
moderate change in fluorescence intensity for 2. When 10 equiv
of various competitive metal ions were assayed, in the case of
Zn2+, only 1 equiv was enough to “turn off” the emission.
Additionally, competitive experiments revealed that Zn2+
complexation is predominant in the presence of other metal
ions (Figure 4b), even with Cu2+.
−
including F−, Cl−, Br−, I−, CH3COO−, CN−, SO42−, HCO3 ,
−
−
NO3 , CO32−, H2PO4 , HPO42−, P2O74− (PPi), as well as ATP,
ADP, and AMP. Figure 5 clearly shows that 2·Zn2+ is highly
sensitive to polyphosphates PPi and ATP. The emission
intensity of the 2·Zn2+ ensemble increases upon addition of
ATP and PPi; the other anions, including inorganic phosphates
and AMP, give no observable changes. Interestingly, ADP gives
moderate emission changes. Fluorescence titration profiles of 2·
Zn2+ (5.0 μM) with different concentrations of polyphosphates
(0−100.0 μM) indicate that fluorescence intensity increases
gradually with increasing concentration of polyphosphates and
reaches saturation at ∼20 μM for ATP and PPi, constituting
∼95% and ∼70%, of the fluorescence signal for 2, respectively
(Figure 5b). Since the addition of ATP and PPi gives a recovery
of fluorescence identical to that for 2, we propose that the
mechanism of sensing involves initial bidentate binding of one
phosphate to 2·Zn2+ (see Figure 9) followed by the full
displacement of zinc ion by a second chelating polyphosphate.
Binary-type compounds [Zn2+·2PPi] or [Zn2+·2ATP] would
thus form, liberating compound 2. The decomplexation of zinc
ion by polyphosphates has been previously reported, but
Fluorescence titration experiments for 2 (5.0 μM) with
different concentrations of Zn2+ show gradual decreases in
emission intensity; saturation is reached with 1.0 equiv of Zn2+
10056
dx.doi.org/10.1021/ic4013526 | Inorg. Chem. 2013, 52, 10052−10061

Inorganic Chemistry
Article
Figure 7. (a) Fluorescence experiments showing significant and reversible “on−off” responses for the detection of PPi by 2·Zn2+. [2] = [Zn2+] = 5.0
μM, [PPi] = 10 μM; A = 2·Zn2+, B = A + [PPi], C = B + [ Zn2+], D = C + [PPi], E = D + [Zn2+], F = E + [PPi], G = F + [Zn2+], H = G + [PPi], I =
H + [Zn2+], J = I + [PPi], K = J + [Zn2+], L = K + [PPi]. (b) Photographs of 2 (5.0 μM) and 2·Zn2+ (5.0 μM) under UV excitation (365 nm),
[ATP] = 10 μM.
Figure 8. Summary of hydrolytic pathways for probe 2·Zn2+ and related systems.21
stepwise mechanistic studies are absent,15a,23 thus spurring an
entry into theoretical calculations shown below (Figure 9) as
well as mass spectrometric studies.
presence of other anions, including phosphates, such as
−
HPO42−, H2PO4 , AMP, and ADP (Figure 5d).
To further confirm the binding mode and to support the
Fluorescence spectra confirm that for the total release of 2
from 2·Zn2+, 2.0 equiv of ATP or PPi is required, leading to a
proposed 1:2 interaction between 2·Zn2+ and polyphosphate,
supported by Benesi−Hildebrand analysis (Figure S13). At 2.0
equiv of analyte loading, it is proposed that phosphate binding
out-competes core [OphenolNimineNpyr] binding leading to the
liberation of 2 and proposed formation of a Zn2+·2PPi type
species. Estimated binding constants for ATP and PPi were
determined on the basis of this ratio (Figure S13). Detection
limits for ATP and PPi were evaluated as 2.4 and 1.0 μM,
respectively (Figure S14), much lower than that concentrations
found in cells.29 Also, ATP sensing by 2·Zn2+ is possible in the
proposed mechanism, NMR spectroscopic titrations and MS
studies were carried out. An H NMR tube experiment was
1
performed and supports the occurrence of a clean complexation
event of 2 with Zn2+ giving the downfield spectral shift for (or
complete removal of) signals assigned to protons of
neighboring groups: hydrazone (8.23 → 8.57 ppm), pyridyl
(Δδ = 0.22 and 0.12 ppm), and phenol (see Figure 6). Small
upfield shifts for the NH and H22 protons were also detected.
This experiment further supports phenolic-, pyridyl-, and
hydrazonyl-Zn2+ binding, in which the [Naniline] group is
directed away (Figure 1). After the addition of 2 equiv of PPi to
1
2·Zn2+, the H NMR spectra resembles that formed for 2,
10057
dx.doi.org/10.1021/ic4013526 | Inorg. Chem. 2013, 52, 10052−10061

Inorganic Chemistry
Article
Figure 9. HOMO − 1, HOMO, LUMO, and LUMO + 1 (from left to right) of geometry-optimized molecules 2, 2·Zn2+, 2·Zn2+·PPi, and 2a (from
top to bottom) in gas phase calculations (B3LYP method and 631g* basis set; sdd basis set for the zinc atom).
indicating a complete dissociation of Zn2+ from 2, enabled by
the strong competitive chelating action of multiple equivalents
of PPi.
fluorescence intensity is increased. Further addition of Zn2+
resulted in the recovery of emission intensity suggesting the in
situ reformation of 2·Zn2+. The cycle can be repeated with
alternate fluorescence “turn-on”, “turn-off” behavior clearly
observed beyond 5 cycles of repetition (Figure 7).
Mass spectrometry provides additional support for the
liberation of 2. These spectra (Figure S7) reveal values
consistent with zinc-bound PPi species along with a peak
corresponding to 2 in independent form.24 Also, a mass peak
consistent with the formation of complex [2 + Zn + PPi] was
observed; here, a m/z peak of 803.2391 was assigned as [2 + Zn
+ PPi + 2H2O + 3Na + H]+. Thus, a mechanism involving the
binding of polyphosphate to 2·Zn2+, with subsequent removal
of Zn2+ from 2·Zn2+ and release of 2, is further supported.
Reversibility, Hydrolysis, and Context of Chemo-
sensing Action. Reversibility and hydrolytic action are
important criteria in chemosensing, especially when the
hydration sphere of the metal ion is considered. Since the
detection mode is based on the removal of the Zn2+ from 2·
Zn2+ by phosphates, probe reversibility was tested via PPi
titrations. When PPi (2 equiv) was added to 2·Zn2+, the
The chelator design in 2 is minimal, compared to that for the
conjugate reported by Hong and co-workers. In fact, our 4-
phenoxy and neighboring 3-substituent position are borrowed
motifs. Importantly, potential Schiff base hydrolysis is
surmounted by the presence of the hydrazone group; sensor
2 does not produce the bare salicylaldehyde derivative (1)
(Figure 8). 2,2′-Dipicolylamine (DPA) is a well-known binding
moiety for zinc ion, and many zinc sensors based on 2,2′-
dipicolylamine (DPA)13 and Zn2+−DPA complexes for
phosphate sensing have already been developed.13,16,25
Importantly, the 2-picolylamine Schiff base version of
compound 2 was also prepared for comparison (Supporting
Information); however, the ligand is highly unstable toward
hydrolysis under aqueous conditions thus precluding sub-
10058
dx.doi.org/10.1021/ic4013526 | Inorg. Chem. 2013, 52, 10052−10061

Inorganic Chemistry
Article
sequent Zn2+ complex formation studies (Figure S8).21 For 2·
Zn2+, this hydrolytic pathway is suppressed for hydrazone 2
(based on resonance stabilization, vide infra). The hydrazone
bond in 2 under aqueous conditions and Zn2+-complexes26 is
stable20 and can be easily prepared and used further for sensing
of biologically relevant analytes. This motif can become a
general design parameter in metal-ion-based chemosensing.
What makes 2 important and largely different in utility from
those shown in Figure 1 is its demonstrated ability to operate
with cations and anions that are both biorelevant in nature. The
hydrolytic differences of ligand 2, compound 1, and related
systems and their fluorescence capacities are summarized above
(Figure 8). There is also a relationship to salicylaldehyde itself,
determined to be of low neurotoxicity, and relevant to biology
and proteins.27 Additionally, the general lack of toxicity of
BODIPY is encouraging when pursuing conjugates such as
compound 1. The ensemble is different enough from closely
related systems (Figure 1): 2 itself leads to fluorescence
enhancement and resists hydrolysis. In a related case, 1 does
form.21 But, with 2-(2-pyridinyl)hydrazone benzaldehyde
conjugation, generation of probe 2 is retained cleanly, enabling
cycling as shown (Figure 7).
Computational HOMO−LUMO Analysis. Geometry
optimizations of 2, 2·Zn2+, and 2·Zn2+·PPi were undertaken.
Structural parameters and coordinates were used, and the
generation of frontier molecular orbital depictions and their
relative electronic energies were obtained. HOMO/LUMO
comparisons were made with ligand 2 (Figure 9). Subsequent
to this, and on the basis of geometries of closely related
structural fragments, structures of 2·Zn2+(H2O)2 and 2·Zn2+PPi
were successfully optimized and assessed electronically (see
Table 1 and SI). An assessment of very different electronic
calculations in which the [NH] unit was formally switched to
a [CH2] unit (2a), despite the experimentally observed lack of
hydrolytic stability. Aside from the HOMO − 1, the HOMO−
LUMO differences are very similar (Table 1). Next, the metal
center environment of both the 2·Zn2+·(H2O)2 and 2·Zn2+·PPi
is square pyramidal; in 2·Zn2+·(H2O)2, bound water hydrogens
were reached by proximal heteroatoms in a small hydrogen-
bonding network.
IV. CONCLUSION
In conclusion, we report a new hydrolysis-resistant zinc-ion-
based detection modality for biological polyphosphate binding/
detection. Herein, a specific BODIPY-conjugate for selective
“turn-on” fluorescence sensing of pyrophosphate and ATP at a
single Zn2+ center was demonstrated in aqueous media. The
sensing mode involves demetalation of the weakly fluorescent
2·Zn2+ by ATP and PPi, leading to fluorescence enhancement
due to production of the fluorescent compound 2. The probe
ensemble shows competitive selectivity for ATP and PPi over
other anions, including inorganic phosphates. The ensemble is
different from previous closely related systems: the production
of 2, and not 1, dictates adequate fluorescence enhancement,
and importantly demonstrates clean optical signal cycling via
Zn2+/PPi addition. Detection limits were estimated on the
order of micromolar concentration: 2.4 μM for ATP and 1.0
μM for PPi, lower than levels found biologically. Probe 2 was
not explicitly considered in the detection of Zn2+ in its own
right (turn-off), but is selective as demonstrated when Zn2+ is
held in competition with various other metal ions. The picoline
version (pyridin-2-ylmethanamine) hydrolyzes readily upon
ligand synthesis and is thus not practical for tandem metal ion−
anion screening. Substitution at 2 by aryl amino groups is
subtle: (i) imine bonds are vulnerable upon Schiff base
formation; (ii) heteroatom placements steer metal ion/anion
binding/recognition; (iii) metal·n(anion) fragment dissociation
leads to probe 2 liberation. The probe is reversible: after
binding with PPi or ATP, 2·Zn2+ can be easily regenerated by
the addition of Zn2+ in situ. Triggered drug delivery and
concepts in molecular neurodegenerative disease research will
be pursued in future work.
Table 1. Relative Energy Levels (eV) for Compounds 2, 2·
Zn2+, and 2·Zn2+·PPi
2 (2a)
2·Zn2+·2H2O
2·Zn2+·PPi
LUMO + 1
LUMO
3.66 (3.64)
2.99 (3.00)
0.00
2.69
2.98
2.17
0.39
HOMO
0.00
0.00
HOMO − 1
−0.33 (−0.91)
−1.14
−0.42
distributions in LUMO and LUMO + 1 levels suggest the
origin of the donor−acceptor dyad. In the LUMO + 1 diagram,
the donor group is clearly blanketed by electron density.
Furthermore, geometric features, such as those found for the
crystal structure, are retained: the calculated imine bond length
is 1.29 Å (versus 1.28 Å determined from X-ray structural
data); the N−C18 bond length is 1.39 Å; the N−N bond length
is 1.34 Å; finally, the C13−C17 bond length is 1.45 Å. Structural
values, respectively, of 1.38, 1.37, and 1.46 Å show accord
between theory and experiment. The HOMO levels in both 2
and 2·Zn2+ are localized mainly on the BODIPY fluorophore.
However, differences exist at the LUMO level. For compound
2, the LUMO is localized on the fluorophore core, whereas the
LUMO of the zinc complex is located on the hydrazone moiety.
A complete HOMO → LUMO photoexcited electron transfer
mechanism is therefore proposed. Further, addition of PPi leads
to a delocalization of electron density from the zinc binding site
to the BODIPY fluorophore; here, the nonfluorescent nature of
the zinc complex has been incapacitated allowing for a
redistribution of electron density which causes an increase in
fluorescence. The imine version was also studied via
ASSOCIATED CONTENT
■
S
* Supporting Information
Crystallographic data in CIF format. Synthesis, characterization,
and spectroscopic and binding studies of 2 and 2·Zn2+ with
phosphates. This material is available free of charge via the
Internet at http://pubs.acs.org. The CIF data of crystallo-
graphic determinations have also been deposited with the
Cambridge Crystallographic Data center. These data are
available at www.ccdc.cam.ac.uk/conts/retrieving.html or from
the CCDC, 12 Union Road, Cambridge CB2 IEZ, United
Kingdom. E-mail: deposit@ccdc.cam.ac.kr.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: dchurchill@kaist.ac.kr. Phone: +82-42-350-2845. Fax:
+82-42-350-2810.
Notes
The authors declare no competing financial interest.
10059
dx.doi.org/10.1021/ic4013526 | Inorg. Chem. 2013, 52, 10052−10061

Inorganic Chemistry
Article
Pan, J.; Kim, H. N.; Park, S.; Kim, K. S.; Yoon, J. J. Am. Chem. Soc.
2009, 131, 15528−15533.
ACKNOWLEDGMENTS
■
The Molecular Logic Gate Laboratory (D.G.C.) acknowledges
research support from the National Research Foundation
(NRF) (Grants 2011-0017280, 2010-0013660, 2009-0070330).
S.T.M. acknowledges support from Institute for Basic Science
(IBS) for his postdoctoral fellowship. Mr. Hack Soo Shin and
Prof. Youngkyu Do are acknowledged, respectively, for
facilitating the acquisition of NMR spectroscopic and crystallo-
graphic data. The research support staff at KAIST facilitated the
acquisition of MS data.
(15) Selective for PPi: (a) Pathak, R. K.; Rai, A.; Panda, D.; Rao, C.
P. Anal. Chem. 2012, 84, 5117−5123. (b) Ravikumar, I.; Ghosh, P.
Inorg. Chem. 2011, 50, 4229−4231. (c) Shao, N.; Wang, H.; Gao, X.;
Yang, R. H.; Chan, W. H. Anal. Chem. 2010, 82, 4628−4636. (d) Lee,
D. H.; Kim, S. Y.; Hong, J.-I. Angew. Chem., Int. Ed. 2004, 43, 4777−
4780. (e) Cho, H. K.; Lee, D. H.; Hong, J.−I. Chem Commun 2005,
1690−1692. (f) Kim, S. K.; Lee, D. H.; Hong, J.−I.; Yoon, J. Acc.
Chem. Res. 2009, 42, 23−31. (g) Wen, J.; Geng, Z.; Yin, Y.; Zhang, Z.;
Wang, Z. Dalton Trans. 2011, 40, 1984−1989. (h) Roy, B.; Rao, A. S.;
Ahn, K. H. Org. Biomol. Chem. 2011, 9, 7774−7779. (i) Rao, A. S.;
Singh, S.; Choi, W.; Ahn, K. H. Org. Biomol. Chem. 2012, 10, 8410−
8417. (j) Chen, W.−H.; Xing, Y.; Pang, Y. Org. Lett. 2011, 13, 1362−
1365. (k) Zhang, J. F.; Kim, S.; Han, J. H.; Lee, S.-J.; Pradhan, T.; Cao,
Q. Y.; Lee, S. J.; Kang, C.; Kim, J. S. Org. Lett. 2011, 13, 5294−5297.
(16) Sakamoto, T.; Ojida, A.; Hamachi, I. Chem. Commun. 2009,
141−152 and references therein.
REFERENCES
■
(1) (a) Hirsch, A. K. H.; Fischer, F. R.; Diederich, F. Angew. Chem.,
Int. Ed. 2007, 46, 338−352. (b) Bowler, M. W.; Cliff, M. J.; Waltho, J.
P.; Blackburn, G. M. New J. Chem. 2010, 34, 784−794.
(2) Berg, J. M.; Tymoczko, J. L.; Stryer, L. Biochemistry, 5th ed.; W.
H. Freeman and Company: New York, 2002.
(3) (a) Stubbe, J. J. Biomol. Chem. 1990, 265, 5329−5332. (b) Shen,
X.; Mizuguchi, G.; Hamiche, A.; Wu, C. Nature 2000, 406, 541−544.
(c) Kornberg, A. J. Biol. Chem. 1988, 263, 1−4.
(4) Eckford, P. D. W.; Sharom, F. J. Chem. Rev. 2009, 109, 2989−
3011.
(17) (a) Coskun, A.; Baytekin, B. T.; Akkaya, E. U. Tetrahedron Lett.
2003, 44, 5649−5651. (b) Shin, I.-S.; Bae, S. W.; Kim, H.; Hong, J.-I.
Anal. Chem. 2010, 82, 8259−8265. (c) Ojida, A.; Sakamoto, T.; Inoue,
M.; Fujishima, S.-h.; Lippens, G.; Hamachi, I. J. Am. Chem. Soc. 2009,
131, 6543−6548.
(18) (a) Meng, Z.-S.; Guo, F.-S.; Liu, J.-L.; Leng, J.-D.; Tong, M.-L.
Dalton Trans. 2012, 418, 2320−2329. (b) Larionov, S. V.; Savel’eva, Z.
A.; Klevtsova, R. F.; Uskov, E. M.; Glinskaya, L. A.; Popov, S. A.;
Tkachev, A. V. Russ. J. Coord. Chem. 2011, 37, 1−7. (c) Ghosh, K.;
Kumar, P.; Tyagi, N.; Singh, U. P. Inorg. Chem. 2010, 49, 7614−7616.
(d) Hosny, N. M.; Mahmoud, M. A.l-M.; Aly, M. R. E. Synth. React.
Inorg., Met.-Org., Nano-Met. Chem. 2010, 40, 439−446. (e) Enomoto,
K. Jpn. Kokai Tokkyo Koho JP 2004134289 A 20040430, 2004.
(f) Swamy, S. J.; Reddy, S. R. Indian Chem., Sect. A 2001, 40A, 1093−
1096. (g) Chia, L. M. L.; Wheatley, A. E. H.; Feeder, N.; Davies, J. E.;
Halcrow, M. A. Polyhedron 2000, 19, 109−114. (h) Vecchio-Sadus, A.
M. Trans. Met. Chem. 1995, 20, 256−61. (i) Mohan, M.; Gupta, N. S.;
Chandra, L.; Jha, N. K. J. Inorg. Biochem. 1987, 31 (1), 7−27.
(j) Anderson, R. G.; Nickless, G. Talanta 1967, 14 (11), 1221−8.
(19) For reports about 2-(2-pyridinyl)hydrazone benzaldehyde and
sensors see: (a) Gong, Z.-L.; Ge, F.; Zhao, B.-X. Sens. Actuators, B
2011, 159, 148−153. (b) Panchal, R. G.; Kota, K. P.; Spurgers, K. B.;
Ruthel, G.; Tran, J. P.; Boltz, R. C.; Bavari, S. J. Biomol. Screening 2010,
15 (7), 755−765.
(5) (a) Blanco, G.; Mercer, R. W. Am. J. Physiol. 1998, 275, F633−
F650. (b) Ashcroft, F. M.; Gribble, F. M. Diabetologia 1999, 42, 903−
919.
(6) Burnstock, G. Pharmacol. Rev. 2006, 58, 58−86.
(7) (a) Wilcox, D. E. Chem. Rev. 1996, 96, 2435−2458. (b) Cleland,
W. W.; Hengge, A. C. Chem. Rev. 2006, 106, 3252−3278.
(8) Doherty, M.; Belcher, C.; Regan, M.; Jones, A.; Ledingham, J.
Ann. Rheum. Dis. 1996, 55, 432−436.
(9) (a) Liu, X.; Ngo, H. T.; Ge, Z.; Butler, S. J.; Jolliffe, K. A. Chem.
Sci. 2013, 4, 1680−1686. (b) Butler, S. J.; Jolliffe, K. A. Chem.Asian
J. 2012, 7, 2621−2628. (c) Ambrosi, G.; Formica, M.; Fusi, V.; Giorgi,
L.; Guerri, A.; Macedi, E.; Micheloni, M.; Paoli, P.; Pontellini, R.;
Rossi, P. Inorg. Chem. 2009, 48, 5901−5912. (d) Nonaka, A.; Horie, S.;
James, T. D.; Kubo, Y. Org. Biomol. Chem. 2008, 6, 3621−3625.
(e) Chen, Z.-h.; Lu, Y.; He, Y.-b.; Huang, X.-h. Sens. Actuators, B 2010,
149, 407−412. (f) Lee, D.-N.; Jo, A.; Park, S. B.; Hong, J.-I.
Tetrahedron Lett. 2012, 53, 5528−5530. (g) Kim, K.; Ha, Y.; Kaufman,
L.; Churchill, D. G. Inorg. Chem. 2012, 51, 928−938.
(20) Kalia, J.; Raines, R. T. Angew. Chem., Int. Ed. 2008, 47, 7523−
7526.
(21) Tsay, O. G.; Lee, K. M.; Churchill, D. G. New J. Chem. 2012, 36,
(10) Valeur, B. Molecular Fluorescence: Principles and Applications;
Wiley−VCH: Weinheim, 2002.
(11) (a) O’Neil, E. J.; Smith, B. D. Coord. Chem. Rev. 2006, 250,
3068−3080. (b) Beer, P. D.; Gale, P. A. Angew. Chem., Int. Ed. 2001,
1949−1952.
(22) Average molecular fluorescence quantum yields are provided
from triplicate measurements determined in CH3OH. See, e.g. :
Demas, , J. N. In Optical Radiation Measurements; Mielenz, , K. D.,
Ed.; Academic Press: New York, 1982,; Vol. 3, (Measurement of
Photoluminescence), pp 195−248.
́
́
40, 486−516. (c) Martinez-Manez, R.; Sancenοn, F. Chem. Rev. 2003,
̃
103, 4419−4476.
(12) (a) Deng, J.; Yu, P.; Yang, L.; Mao, L. Anal. Chem. 2013, 85,
2516−2522. (b) Shao, N.; Jin, J.; Wang, G.; Zhang, Y.; Yang, R.; Yuan,
J. Chem. Commun. 2008, 9, 1127−1129. (c) Ganjali, M. R.; Hosseini,
M.; Aboufazeli, F.; Faridbod, F.; Goldooz, H.; Badiei, A. R.
Luminescence 2012, 27, 20−23.
(23) (a) Wezenberg, S. J.; Anselmo, D.; Escudero-Adan, E. C.;
Buchholz, J. B.; Kleij, A. J. Eur. J. Inorg. Chem. 2010, 29, 4611−4616.
(b) Wezenberg, S. J.; Anselmo, D.; Escudero-Adan, E. C.; Benet-
Buchholz, J.; Kleij, A. J. Org. Lett. 2008, 10, 3311−3314. (c) Wild, A.;
Winter, A.; Hager, M. D.; Schubert, U. S. Analyst 2012, 137, 2333−
2337.
(24) The m/z value of 499.1907 constitutes [2·PPi + Zn + 2H2O + 2
Na+ + 4H]+. The m/z value of 543.1539 is assigned to a product
including Zn·PPi2: [2·PPi + Zn + 2H2O + 4Na+ + 2H]+. The peak at
515.1551 corresponds to free 2: [2 + Na + CH3OH + H]+.
(25) (a) Oh, J.; Hong, J.-I. Org. Lett. 2013, 15, 1210−1213. (b) Rhee,
H.-W.; Choi, S. J.; Yoo, S. H.; Jang, Y. O.; Park, H. H.; Pinto, R. M.;
Cameselle, J. C.; Sandoval, F. J.; Roje, S.; Han, K.; Chung, D. S.; Suh,
J.; Hong, J.-I. J. Am. Chem. Soc. 2009, 131, 10107−10112. (c) Ngo, H.
T.; Liu, X.; Jolliffe, K. A. Chem. Soc. Rev. 2012, 41, 4928−4965.
(26) Cao, J.; Zhao, C.; Wang, X.; Zhang, Y.; Zhu, W. Chem. Commun.
2012, 48, 9897−9899.
(13) (a) Xu, Z.; Yoon, J.; Spring, D. R. Chem. Soc. Rev. 2010, 39,
1996−2006. (b) Lee, H. N.; Xu, Z.; Kim, S. K.; Swamy, K. M. K.; Kim,
Y.; Kim, S.-J.; Yoon, J. J. Am. Chem. Soc. 2007, 129, 3828−3829.
(14) Reported selective systems for ATP: (a) Kaur, J.; Singh, P.
Chem. Commun. 2011, 47, 4472−4474. (b) Rao, A. S.; Kim, D.; Nam,
H.; Jo, H.; Kim, K. H.; Ban, C.; Ahn, K. H. Chem. Commun. 2012, 48,
3206−3208. (c) Strianese, M.; Milione, S.; Maranzana, A.; Grassi, A.;
Pellecchia, C. Chem. Commun. 2012, 48, 11419−11421. (d) Ojida, A.;
Miyahara, Y.; Wongkongkatep, J.; Tamaru, S.-i.; Sada, K.; Hamachi, I.
Chem.Asian J. 2006, 1, 555−563. (e) Kurishita, Y.; Kohira, T.; Ojida,
A.; Hamachi, I. J. Am. Chem. Soc. 2012, 134, 18779−18789. (f) Kaur,
J.; Singh, P. Chem. Commun. 2011, 47, 4472−4474. (g) Moro, A. J.;
Cywinski, P. J.; Korsten, S.; Mohr, G. J. Chem. Commun. 2010, 46,
̈
1085−1087. (h) Kurishita, Y.; Kohira, T.; Ojida, A.; Hamachi, I. J. Am.
Chem. Soc. 2010, 132, 13290−13299. (i) Xu, Z.; Singh, N. J.; Lim, J.;
10060
dx.doi.org/10.1021/ic4013526 | Inorg. Chem. 2013, 52, 10052−10061

Inorganic Chemistry
Article
(27) Ha, Y.; Liew, H.; Park, H. Y.; Kim, K.; Suh, Y.-H.; Churchill, D.
G. Neurosci. Res. 2011, 71, 168−177.
(28) Benesi, H. A.; Hildebrand, J. H. J. Am. Chem. Soc. 1949, 71,
2703−2707.
(29) (a) Traut, T. W. Mol. Cell. Biochem. 1994, 140, 1−22.
(b) Barshop, B. A.; Adamson, D. T.; Vellom, D. C.; Rosen, F.; Epstein,
B. L.; Seegmiller, J. E. Anal. Biochem. 1991, 197, 266−272. (c) Russell,
R. G.; Bisaz, S.; Donath, A.; Morgan, D. B.; Fleisch, H. J. Clin. Invest.
1971, 50, 961−969.
(30) Thordarson, P. Chem. Soc. Rev. 2011, 40, 1305−1323.
10061
dx.doi.org/10.1021/ic4013526 | Inorg. Chem. 2013, 52, 10052−10061