Balance between fluorescence enhancement and association affinity in fluorescent heteroditopic indicators for imaging zinc ion in living cells

Article abstract of DOI:10.1021/ic201728f

A fluorescent heteroditopic indicator for the zinc(II) ion possesses two different zinc(II) binding sites. The sequential coordination of zinc(II) at the two sites can be transmitted into distinct fluorescence changes. In the heteroditopic ligand system that our group developed, the formations of mono- and dizinc(II) complexes along an increasing gradient of zinc(II) concentration lead to fluorescence enhancement and an emission bathochromic shift, respectively. The extents of these two changes determine the sensitivity and, ultimately, the effectiveness of the heteroditopic indicator in quantifying zinc(II) ion over a large concentration range. In this work, a strategy to increase the degree of fluorescence enhancement upon the formation of the monozinc(II) complex of a heteroditopic ligand under simulated physiological conditions is demonstrated. Fluorination of the pyridyl groups in the pentadentate N,N,N′-tris(pyridylmethyl)ethyleneamino group reduces the apparent pK_a value of the high-affinity site, which increases the degree of fluorescence enhancement as the monozinc(II) complex is forming. However, fluorination impairs the coordination strength of the high-affinity zinc(II) binding site, which in the triply fluorinated ligand reduces the binding strength to the level of the low-affinity 2,2′-bipyridyl. The potential of the reported ligands in imaging zinc(II) ion in living cells was evaluated. The subcellular localization properties of two ligands in five organelles were characterized. Both benefits and deficiencies of these ligands were revealed, which provides directions for the near future in this line of research.

Full text of DOI:10.1021/ic201728f

ARTICLE
pubs.acs.org/IC
Balance between Fluorescence Enhancement and Association
Affinity in Fluorescent Heteroditopic Indicators for Imaging Zinc Ion
in Living Cells
Gui-Chao Kuang,† John R. Allen,‡ Michelle A. Baird,‡ Brian T. Nguyen,† Lu Zhang,† Thomas J. Morgan, Jr.,§
Cathy W. Levenson,§ Michael W. Davidson,*,‡ and Lei Zhu*,†
†Department of Chemistry and Biochemistry, Florida State University, Tallahassee, Florida 32306-4390, United States
‡National High Magnetic Field Laboratory and Department of Biological Science, Florida State University, 1800 East Paul Dirac Drive,
Tallahassee, Florida 32310, United States
§College of Medicine, Florida State University, 1115 West Call Street, Tallahassee, Florida 32306-4300, United States
S Supporting Information
b
ABSTRACT: A fluorescent heteroditopic indicator for the zinc(II) ion
possesses two different zinc(II) binding sites. The sequential coordination of
zinc(II) at the two sites can be transmitted into distinct fluorescence changes. In
the heteroditopic ligand system that our group developed, the formations of
mono- and dizinc(II) complexes along an increasing gradient of zinc(II)
concentration lead to fluorescence enhancement and an emission bathochromic
shift, respectively. The extents of these two changes determine the sensitivity
and, ultimately, the effectiveness of the heteroditopic indicator in quantifying
zinc(II) ion over a large concentration range. In this work, a strategy to increase
the degree of fluorescence enhancement upon the formation of the monozinc-
(II) complex of a heteroditopic ligand under simulated physiological conditions
is demonstrated. Fluorination of the pyridyl groups in the pentadentate N,N,N0-
tris(pyridylmethyl)ethyleneamino group reduces the apparent pKa value of the
high-affinity site, which increases the degree of fluorescence enhancement as the monozinc(II) complex is forming. However,
fluorination impairs the coordination strength of the high-affinity zinc(II) binding site, which in the triply fluorinated ligand
reduces the binding strength to the level of the low-affinity 2,20-bipyridyl. The potential of the reported ligands in imaging
zinc(II) ion in living cells was evaluated. The subcellular localization properties of two ligands in five organelles were
characterized. Both benefits and deficiencies of these ligands were revealed, which provides directions for the near future in
this line of research.
’ INTRODUCTION
properties of the ligand, as well as the impact of the environment
(e.g., solvent, counterion, etc.) on the coordination chemistry
and photophysical profiles of the ligand.
A fluorescent heteroditopic ligand contains two metal coordi-
nation sites; one has a higher affinity, or a different selectivity,
than the other. Over the concentration gradient of a metal ion (M
in Figure 1), one or two of the three coordination states—non-,
mono-, and dicoordinated—predominates, which upon excita-
tion expresses distinct steady-state fluorescence as a function of
As early as 2000, modulation of the fluorescence of a hetero-
ditopic ligand via sequential metal coordination was observed by
Rurack et al.2,5ꢀ7 It was suggested that such a platform could be
applied in sensing technologies that have large concentration
coverage or multiplexed capabilities. The potential ability of a
fluorescent heteroditopic ligand to cover a large sensing or
imaging concentration range is particularly relevant to the studies
of biological zinc(II) ion,8ꢀ18 whose concentration may vary from
a subnanomolar regime to approaching millimolar on certain
occasions.19 In the resting states of most mammalian cells, the
free or mobile zinc(II)20 concentrations in cytosol are at nano-
molar or lower levels.20ꢀ22 In certain zinc(II)-storing subcellular
metal ion concentration within a broad range (Figure 1).1,2
A
fluorescent heteroditopic ligand by design undergoes a number
of relaxation processes upon photoexcitation, including electron
transfer, energy transfer, or fluorescence.1,3 The relative rates of
these processes are acutely dependent on the coordination status
of the ligand.4 The construction of a fluorescent heteroditopic
ligand that is capable of generating a large fluorescence contrast
between its three coordination states is fundamentally interesting
and challenging, because such an exercise requires a precise, and
oftentimes improved understanding of the relevant stepwise
coordination chemistry and its effect on the photophysical
Received: August 8, 2011
Published: September 12, 2011
r
2011 American Chemical Society
10493
dx.doi.org/10.1021/ic201728f Inorg. Chem. 2011, 50, 10493–10504
|

Inorganic Chemistry
ARTICLE
vesicular structures (e.g., those found in hippocampal mossy fiber
neuron terminals), the concentrations of free zinc(II) are much
higher.23,24 When cells are subjected to environmental changes,
free zinc(II) concentrations may increase drastically due to the
stimulated release of zinc(II) from zinc(II)-storing vesicles. For
example, when insulin is released from pancreatic β cells, the
corelease of zinc(II) leads to a rapid increase of extracellular
concentration of free zinc(II).25 When hippocampal neurons are
under stimulation, the flow of neural transmitter glutamate to
the synaptic cleft prior to entering the postsynaptic neuron is
accompanied by zinc(II) ion.26ꢀ28 These events involve large
swings of concentrations of zinc(II), for which indicators with
single zinc(II) binding sites are unable to cover.29
The use of two indicators that are effective in two different
concentration regimes may collectively record the full profiles of
zinc(II) dynamics during these events. This approach has been
demonstrated on several occasions.30ꢀ32 However, the use of
two or multiple indicators may suffer uneven cellular uptake
efficiencies and/or varied subcellular localization properties.
Therefore, a single indicator with broad concentration coverage
is advantageous in quantitative profiling of zinc(II) distribution
and dynamics when large swings of zinc(II) concentration occur.
The fluorescent heteroditopic ligand platform promises to
produce indicators with large zinc(II) concentration coverage
for quantitative biological sensing/imaging purposes.
Our group reported fluorescent heteroditopic ligand 1 (Figure 2),
whose fluorescence is dependent on free zinc(II) concentration
over a range of 6 orders of magnitude under zinc(II)-buffered, pH-
neutral aqueous conditions.33 Compound 1 contains a phenylvinylꢀ
bipyridyl fluorophore whose fluorescence is quenched via intramo-
lecular photoinduced electron transfer (PET)34 from the amino
groups in the high-affinity pentadentate metal coordination site
(boxed in blue). Zinc(II) coordination at the high-affinity site
lowers the rate of PET, hence allowing the fluorescence to occur.
The formation of the dizinc(II) complex at high zinc(II) concen-
trations leads to a bathochromic shift of emission due to the
stabilization of a charge-transferred excited state by the presence of
the zinc(II) ion.35
The fluorescence enhancement factor, defined as the ratio of
the fluorescence quantum yields of the monozinc(II) complex
and the free ligand (ϕ1/ϕ0), and the emission band shift (Δλ) are
the two metrics used in evaluating the heteroditopic indicator in a
sensing/imaging context.4 A large Δλ may allow the use of two
separate fluorescence microscopic filter sets to record the
intensity changes in both fluorescence channels. A large ϕ1/ϕ0
is imperative for creating an indicator that is highly sensitive to
the variation of free zinc(II) within low concentration regimes
(i.e., nanomolar or below). For compound 1 under physiological
conditions, the values of ϕ1/ϕ0 and Δλ are 1.9 and 30 nm,
respectively. Developing strategies to increase Δλ of a hetero-
ditopic indicator is the topic of a separate project.3 Herein, we
introduce a solution to increase the fluorescence enhancement
factor ϕ1/ϕ0 of 1. We will also evaluate the live-cell imaging
potentials of the family of fluorescent heteroditopic indicators
derived from the structure of compound 1.
Figure 1. Metal coordination-dependent fluorescence of a fluorescent
heteroditopic ligand. Square, high-affinity binding site; circle, low-
affinity binding site; M, the metal ion of interest, in this study zinc(II);
λ, emission wavelength; ϕ, fluorescence quantum yield. Subscripts
denote the coordination status. In this work, the monozinc(II) complex
formation results in a fluorescence enhancement, whereas the dizinc(II)
complex formation leads to an emission band shift. Ka and Kd are the
acidity constant and the metal complex dissociation constant, respec-
tively, of a metal coordination site. “1” and “2” denote the high- and low-
affinity site, respectively.
’ RESULTS AND DISCUSSION
A. Design. The fluorescence quantum yield of compound 1
under zinc(II)-free physiological conditions is 0.30.33 We
Figure 2. The protonation of the high-affinity binding site of ligand 1 and the formation of the monozinc(II) complex. Blue and red frames box the high-
and low-affinity binding sites, respectively.
10494
dx.doi.org/10.1021/ic201728f |Inorg. Chem. 2011, 50, 10493–10504

Inorganic Chemistry
ARTICLE
Chart 1. Structures
obtained following the reductive amination reactions of 13
and 14 with aldehyde 15. The preparations of nonfluorinated
compounds 5 and 6 are described in the Supporting Infor-
mation.
C. pH Profiles of the Ligands. Due to the structural similarity
of heteroditopic ligands 1ꢀ6, monofluorinated ligand 2 is
chosen as an example in the following description of the
investigation on the pH dependence of fluorescence. There are
five different (six overall, as marked by arrows in Figure 3)
protonatable sites within the pH range of 2ꢀ10 in ligand 2, which
challenges the determination of the pH-fluorescence profile. In
the six protonatable sites, two are tertiary amino groups and four
are pyridyl/6-fluoropyridyl groups. The acid/base equilibria of
four sites are expected to impact the fluorescence property of
ligand 2: the protonation of the tertiary amino groups (marked
by blue arrows) would inhibit the fluorescence quenching PET
processes, whereas the protonation of the bipy (2,20-bipyridyl)
nitrogen atoms (marked by red arrows) would enhance the
charge-transfer character of the fluorophore. We do not expect
the protonation of the other three unconjugated pyridyl/6-
fluoropyridyl groups (marked by green arrows) to significantly
affect the fluorescence of 2. Since the pKa values of a tertiary
amino and a pyridyl group are well-separated, we should be able
to use two apparent pKa values for the tertiary amino containing
high-affinity binding site (pKa1 in Figure 1) and the bipy-
containing low-affinity binding site (pKa2 in Figure 1), respec-
tively, to characterize the pH-fluorescence profile of ligand 2.
The fluorescence of ligand 2 rises as pH decreases from 9.9.
After the intensity maximizes at a pH ∼ 5.3, the emission
undergoes a bathochromic shift as the solution turns further
acidic (Figure 4A). Similar to the fluorescence response of the
heteroditopic ligand to a zinc(II) gradient,1 the high-affinity site
is protonated during the first phase (starting from pH 9.9
downward), which results in a fluorescence enhancement,
whereas the protonation of the low-affinity bipy site in the
second phase leads to the bathochromic shift of the emission.
The two-phase transition that the fluorescence of ligand 2
experiences as pH decreases from 9.9 to 2.0 can be monitored
at two different wavelength channels (Figure 4B). By fitting the
portions of largest changes of the three isotherms40 with a
modified HendersonꢀHasselbalch equation (see details in the
Supporting Information), the apparent pKa values of the high-
(note that in the high-affinity site there are two amino groups
which are difficult to distinguish) and low-affinity binding sites of
ligand 2 were determined (Table 1). The pKa values of other
ligands (Table 1) were obtained using the same approach.
As shown in Table 1, the substitution of fluorine on the pyridyl
groups in the high-affinity binding site of 1 to afford ligands 2ꢀ4
reduces the pKa value of the high-affinity site (pKa1) without
significantly altering that of the low-affinity site (pKa2). Single
fluorine substitution (compound 2) on the pyridyl group
attached to the amino group two bonds away from the fluoro-
phore (N(α) in Figure 2) reduces the pKa1 value by 1.7 unit
(from 8.0 to 6.3). Double-fluorination (compound 3) on the two
pyridyl groups attached to the amino group further away from
the fluorophore (N(β) in Figure 2) has a similar effect on pKa1.
Triple-fluorination further reduces the pKa1 to 6.0 in compound
4, thus resulting in a pentadentate ligand that is largely unpro-
tonated at neutral pH. The pKa2 value drops slightly upon
fluorine substitution at the high-affinity site, up to 0.2 unit in
the triply fluorinated ligand 4. The replacement of N(α)-picolyl
in 1 with a N(α)-methyl group, which results in compound 5,
hypothesized that the protonation of the electron-donating
tertiary amino groups reduces the efficiency of PET, which
results in high background fluorescence (Figure 2). Therefore,
decreasing the pKa of the amino groups by strategically installing
electron-withdrawing components in proximity shall increase the
efficiency of PET, hence lowering the background fluorescence.
Two strategies were devised for raising the electron-with-
drawing capability of the substituents on the amino groups: (1)
The double-bond in 1 is replaced with a triple bond to afford
compound 6. Not only the electron-withdrawing ability of the
fluorophore, which is covalently attached to the amino quencher
via a methylene group, is enhanced, but the possibly interfering
alkene photoisomerization process is eliminated.36 (2) The
electron-withdrawing fluorine is placed on the pyridyl groups
within the high-affinity binding site to result in compounds 2ꢀ4
(see Chart 1). This design shall selectively lower the pKa of the
amino groups without altering the photophysical properties of
the phenylvinylꢀbipy fluorophore in 1.37
B. Synthesis. The syntheses of fluorinated ligands 2ꢀ4 are
shown in Scheme 1. 2-Fluoro-6-bromomethylpyridine (7)38
undergoes an SN2 substitution with N,N-dipicolyl-1,2-ethylene-
diamine (8)39 to afford the monofluorinated pentadentate ligand
9. Reductive amination between compounds 9 and 15 results in
ligand 2. The difluorinated N,N-di(2-fluoro-6-pyridylmethyl)-
ethylenediamine (12) was prepared using the two-step sequence
developed for compound 8.39 Compound 12 subsequently
undergoes reductive amination with 2-pyridinecarboxaldehyde
or a substitution reaction with compound 7 to afford pentaden-
tate ligands 13 and 14, respectively. Ligands 3 and 4 were
10495
dx.doi.org/10.1021/ic201728f |Inorg. Chem. 2011, 50, 10493–10504

Inorganic Chemistry
ARTICLE
Scheme 1. Reagents and Conditions: (a) dry ethanol, K2CO3, reflux, 1 h, 25%; (b) 15, NaBH(OAc)3, rt, 6 h; (c) dry ethanol,
Na2CO3, reflux, 5 h, 74%; (d) TFA, CH2Cl2, rt, 1.5 h, 95%; (e) 2-pyridinecarboxaldehyde, NaBH(OAc)3, rt, 6 h, 36%; (f) 7, dry
ethanol, K2CO3, reflux, 3 h, 22%
raises the pKa1 value to 8.8, suggesting that the 2-picolyl group is
more electron-withdrawing than the methyl group.
The pKa effect on the background fluorescence (in the absence
of zinc(II) addition) of a ligand at a physiological pH of 7.4 can be
assessed using the ratio of integrated fluorescence intensity at a
pH of 7.4 and the maximum intensity achievable over a pH
gradient (e.g., at pH 5.3 for ligand 2, see Figure 4A). The values of
I7.4/Imax of ligands 1ꢀ6 are listed in Table 1. Fluorination at the
high-affinity site suppresses the background fluorescence (I7.4/
I
max e 33% of 2ꢀ4), while I7.4/Imax of nonfluorinated ligand 1 is
Figure 3. The protonatable sites in ligand 2 (marked by arrows). The
blue arrows represent the tertiary amino groups. The red arrows
represent the bipy nitrogens. The green arrows represent the unconju-
gated pyridyl groups.
69%. In the case of the nonfluorinated ligand 5, the I7.4/Imax is
96%, indicating that there is little room for further enhancement
upon binding zinc(II). It was later shown from the zinc(II)
titration experiment that the addition of zinc(II) only reduces the
fluorescence of the solution containing ligand 5.
The formal oxidation of the carbonꢀcarbon double bond in 1
to a triple bond in 6 also decreases the basicity of the amino group
N(α), which has a pKa1 value of 7.3. This strategy on raising the
electron-withdrawing ability of the fluorophore is similar to that
developed by Lippard et al. where fluorination of the xanthene
fluorophore moiety lowers the pKa of the di(2-picolyl)amino-
based zinc(II) binding site in the zinpyr series of indicators.41
Structural modification of the fluorophore in 1 to afford 6 also
decreases the basicity of the bipy low-affinity site so that the pKa2
value drops to 4.0. The formal oxidation of the fluorophore
reduces the charge-transfer character of the fluorophore,36 which
creates shorter excitation (320 nm) and emission (380 nm)
wavelengths and a smaller emission band shift (∼20 nm) upon
zinc(II) coordination (see section E). The excitation and emis-
sion profiles fall into the UV region, which challenges the glass-
based conventional fluorescence microscopes that we apply in
cell imaging experiments. Therefore, compound 6 was not
considered an indicator for zinc(II) in imaging applications.
D. Comments on the Metal Buffering System. Before we
present the zinc(II)-coordination data of compounds 1ꢀ6 in
the next section, some comments on the solvent system that we
used in the study are warranted. The dissociation constants of the
ligand/zinc(II) complexes were determined under metal-buffered
10496
dx.doi.org/10.1021/ic201728f |Inorg. Chem. 2011, 50, 10493–10504

Inorganic Chemistry
ARTICLE
Figure 4. (A) Fluorescence spectra of 2 (5.5 μM, λex = 350 nm) collected at various pH’s (conditions: 10% DMSO, [mixed pH buffer] = 90 mM (see
Supporting Information), [EGTA] = 9 mM). Blue and red spectra were acquired at the highest pH and the lowest pH values, respectively. (B) The
fluorescence intensity dependence on pH at 400 nm (diamonds) and 470 nm (triangles). Blue and red lines are fitting curves based on a modified
HendersonꢀHasselbalch equation. The fitted apparent pKa values are listed in green.
Table 1. pKa Values of High- and Low-Affinity Binding Sites
of 1ꢀ6a
neutral aqueous conditions, where the total zinc(II) concentration
([Zn]t) remains high (close to millimolar level). However, since
the majority of zinc(II) ions are bound by the metal cheletor (or
metal buffer) in the solution, the free zinc(II) (unbound from
organic ligands, hydrated zinc(II) ion)20 is kept at much lower
levels. Therefore, this situation mimics that in the intracellular
milieu, where total zinc(II) abundance is high, while the majority
of it is associated with zinc(II)-binding proteins.20,42 The fluctua-
tion in the free zinc(II) concentration ([Zn]f) may stimulate the
release or uptake of zinc(II) by zinc(II)-binding proteins, such as
metallothionein,20,43,44 to offset the change, therefore creating a
bufferingsystemthatispivotalinmaintainingzinc(II) homeostasis
required for proper physiological functions.
pKa
1
2
3
4
5
6
high-affinity (pKa1)
low-affinity (pKa2)
8.0
4.4
69%
6.3
4.4
27%
6.6
4.3
33%
6.0
4.2
15%
8.8
4.6
96%
7.3
4.0
60%
b
I7.4/Imax
a The curve-fitting plots based on fluorescence titration data of ligands 1,
3ꢀ6 are included in the Supporting Information. b The ratio of
integrated fluorescence intensity at pH 7.4 over the maximal intensity
achievable in the blue channel.
where [Zn]f is in the nanomolar range, [Zn*ind] is in the
micromolar range, and [Zn*buffer] is in the millimolar range.
Therefore, [Zn]t ≈ [Zn*buffer], yet [Zn]f is controlled in the
nanomolar range via equilibrium (eq 1). What the indicator is
reporting is [Zn]f under zinc(II)-buffered conditions in the
presence of a large zinc(II) reservoir, where the presence of a
minuscule amount of the indicator does not affect the [Zn]f that
is controlled according to eq 1. Under the described thermo-
dynamically equilibrating zinc(II)-buffered conditions, assuming
a 1:1 binding stoichiometry between the indicator and zinc(II),
the fluorescence intensity measured at various [Zn]f’s is inde-
pendent of the indicator concentration as shown by Tsien et al. in
analogous studies of calcium(II)-buffered systems (eq 3), when
Imin and Imax are known.45 The zinc(II)-buffering system has
been applied to characterize a large number of zinc(II)
The interplay between the zinc(II) ion, the metal buffer, and
the indicator can be understood by analyzing the following two
0
equilibria. Kd and Kd represent the dissociation constants of
zinc(II) with the metal buffer and the indicator, respectively.
2þ
ꢁ
Zn buffer S Zn þ buffer
½Znꢂf ½bufferꢂ
ð1Þ
0
Kd
¼
ꢁ
½Zn bufferꢂ
2þ
ꢁ
Zn ind S Zn þ ind
½Znꢂf ½indꢂ
Kd ¼
ð2Þ
ꢁ
½Zn indꢂ
indicators20,31,42,46ꢀ71
Assuming that the total indicator concentration ([ind]t) is
2 μM,
that operate under physiological condi-
tions. However, we feel that a clear clarification and a justification
to a broader audience for applying the metal-buffered conditions
in characterizing zinc(II) indicators for biological uses are
needed on the basis of the questions that we have received over
the past few years regarding the appropriateness of this particular
experimental approach. We consider that the application of
zinc(II)-buffered conditions generates more reproducible affinity
data than a direct titration between zinc(II) and an indicator,
because the zinc(II)-buffered system significantly reduces the
effect of interfering metal ions and inaccuracy in indicator
ꢁ
½indꢂt ¼ ½indꢂ þ ½Zn indꢂ ¼ 2 μM
When [ind] = [Zn*ind] = 1 μM, where [Zn]f falls at the center of
the effective coverage of the indicator, [Zn]f = Kd, which, in this
demonstration, is within the nanomolar range or below. Under
such conditions, the total Zn2+ concentration
ꢁ
½Znꢂt ¼ ½Znꢂf ðnMÞ þ ½Zn indꢂ ðμMÞ
ꢁ
þ ½Zn bufferꢂ ðmMÞ
10497
dx.doi.org/10.1021/ic201728f |Inorg. Chem. 2011, 50, 10493–10504

Inorganic Chemistry
ARTICLE
Figure 5. (AꢀE) Fluorescence spectra of ligand 3 (4.1 μM) in the presence of increasing [Zn]f (from blue to red traces) in the buffering ranges of
EDTA, HEDTA, EGTA, NTA, and citrate, respectively. The instrumental parameters were unchanged in the titration experiments. The ranges of y axes
are intentionally kept identical so that the data collected in the five buffered conditions can be compared. (F) Best fits of fluorescence changes using eq 3
at 400 nm (blue) and 470 nm (red) in solutions buffered by EGTA and citrate, respectively.
concentration measurement in affinity determinations, in addi-
tion to providing an environment that is close to physiological
conditions for studying the coordination chemistry of zinc(II).
(eq 3). The breakdown of the zinc(II)-buffered conditions would
take place when (1) the solution is not sufficiently buffered for
metal ions (e.g., when the indicator concentration is high enough
to affect the buffering system),72 (2) the chemical or biochemical
event under investigation is not in a thermodynamic equilibrium
(e.g., when zinc(II) concentration spikes up and down on a time
scale that is faster than the association or dissociation rate of the
zinc(II)ꢀindicator complex), or (3) a 1:1 binding stoichiometry
cannot be used to approximate the association between the
indicator and zinc(II).
ðIminKd þ Imax½ZnꢂÞ
I ¼
ð3Þ
ðKd þ ½ZnꢂÞ
When the conditions to establish the zinc(II)-buffered envi-
ronment do not hold, the correlation between fluorescence
intensity and [Zn]f could not be described by the Tsien equation
10498
dx.doi.org/10.1021/ic201728f |Inorg. Chem. 2011, 50, 10493–10504

Inorganic Chemistry
ARTICLE
E. Zinc(II) Titration Experiments in Metal-Buffered Aqu-
eous Solutions. The zinc(II) titration experiments were carried
out in aqueous solutions (pH 7.4) where the [Zn]f was buffered
in different ranges applying the following five metal chelators:
EDTA (Kd = 7.6 ꢃ 10ꢀ14 M), HEDTA (Kd = 2.3 ꢃ 10ꢀ12 M),
EGTA (Kd = 7.8 ꢃ 10ꢀ9 M), NTA (Kd = 1.4 ꢃ 10ꢀ8 M), and
citrate (Kd = 1.2 ꢃ 10ꢀ5 M).73 Five fluorescence titration
experiments were carried out for each of the ligands 1ꢀ6 in
aqueous solutions that contained individual metal chelators within
their respective buffering ranges. The data were used to estimate
the zinc(II) affinities, assuming a 1:1 coordination stoichiometry,
of the high- and low-affinity sites of 1ꢀ6.
Table 2. Fluorescence Quantum Yields (O0ꢀO2) and Emi-
ssion Band Maxima (λ1, λ2) of 1ꢀ6, and the Dissociation
Constants (pKd1, pKd2) of Their Mono- and Di-Zinc(II)
Complexes under Metal-Buffered Aqueous Conditions
a
a
a
ligand pKd1 pKd2 ϕ0
ϕ1
ϕ2 ϕ1/ϕ0 λ1/nmb λ2/nmb Δλ/nmc
1
2
3
4
5
6
12.8
9.5
5.2 0.30 0.57 0.65 1.9
5.2 0.23 0.64 0.75 2.8
4.9 0.11 0.59 0.53 5.4
5.4 0.14 NDd 0.52 NDd
6.2 0.44 0.43 0.64 0.98
405
405
402
400
398
380
433
433
434
430
427
401
28
28
32
30
29
21
8.1
NDd
NDd
12.8 ∼5.2 0.17 0.48 0.35 2.8
The analysis of the zinc(II) binding profile of difluorinated
indicator 3 (Figure 5) is described as an example. The increases
of [Zn]f within the buffering ranges of EDTA (Figure 5A) and
HEDTA (Figure 5B) lead to only modest fluorescence enhance-
ment, suggesting that the high-affinity site of ligand 3 has a
a ϕ0, ϕ1, and ϕ2 are fluorescence quantum yields of free ligand, monozinc-
(II) complex, and dizinc(II) complex, respectively. b λ1 and λ2 are the
short and long emission band maxima, respectively. c Δλ = λ2 ꢀ λ1.
d ND: not determined.
smaller affinity constant than those of EDTA and HEDTA (Kd1
>
Kd(EDTA or HEDTA)). The fluorescence of 3 responds to [Zn]f
variation sensitively in the solution buffered by EGTA, indicating
that Kd1 is on par with that of the EGTA/zinc(II) complex. When
NTA is used to control [Zn]f, fluorescence enhancement and
bathochromic shift occur simultaneously, implying that both
high- and low-affinity sites of 3 can be populated in this particular
range of [Zn]f. In the citrate-buffered solution, the fluorescence
spectrum immediately shifts to a longer wavelength with the
addition of zinc(II), indicating the rapid saturation of the high-
affinity binding site and the coordination at the bipy secondary
site. On the basis of the data in Figure 5, the apparent 1:1 binding
constant between zinc(II) and the high-affinity binding site of
ligand 3 (Kd1) was satisfactorily fitted using the titration trace at
400 nm collected in EGTA-buffered solution (blue in Figure 5F).
The zinc(II) affinity assuming a 1:1 binding stoichiometry of the
bipy low-affinity site (Kd2) was estimated74 by fitting the titration
trace at 470 nm collected in the citrate-buffered solution.
Other ligands (1, 2, 4ꢀ6) were analyzed using the same
approach. The curve fittings of ligand 1 in both EDTA- and
HEDTA-buffered solutions were satisfactory (Figure S6, Sup-
porting Information). Therefore, the pKd1 (ꢀlog Kd1) of ligand 1
was calculated as the average of the values determined in the two
titration experiments. The pKd1 of ligand 2 was best fitted in a
titration experiment using a solution containing both HEDTA
and EGTA (Figure S7, Supporting Information). The fluore-
scence of ligand 4 does not change much in buffers containing
chelators EDTA, HEDTA, EGTA, and NTA (Figure S8, Sup-
porting Information), suggesting that triple-fluorination signifi-
cantly decreases the affinity of the high-affinity site to zinc(II).
Therefore, only one dissociation constant was reported for ligand
4, which was determined in the citrate-buffered solution. Unlike
ligands 1ꢀ4, no fluorescence enhancement was observed in the
titration experiments involving ligand 5 (Figure S9, Supporting
Information), where a methyl group replaces the 2-picolyl group
at N(α) close to the fluorophore. As shown in Table 1, the high-
affinity site of ligand 5 is almost fully protonated at pH 7.4, which
restores the fluorescence of 1 that is quenched via the PET
pathway in the unprotonated form. Therefore, little fluorescence
alteration occurs when the high-affinity site of 5 takes up zinc(II).
The fluorescence quantum yields and emission maxima of non-,
mono-, and dicoordinated forms of ligands 1ꢀ6 as well as their
dissociation constants are listed in Table 2.
those of ligands 2ꢀ4 in Table 2). In the triply fluorinated ligand
4, the affinity of the high-affinity, pentadentate site drops to the
level that is indistinguishable from the bidentate bipy low-affinity
site. On the other hand, fluorination on the high-affinity site of 1
does drop the apparent pKa1 value, thus increasing the fluores-
cence enhancement factor (ϕ1/ϕ0 in Table 2) from 1.9 of the
nonfluorinated ligand 1 to 5.4 of the difluorinated ligand 3. The
fluorination at the high-affinity site does not appear to affect the
photophysical properties of the fluorophore as the emission
maxima of the free and zinc(II)-bound forms are similar in
ligands 1ꢀ5. In ligand 6, the triple-bond-containing fluorophore
possesses a lesser degree of internal charge transfer (ICT)
character than the phenylvinylꢀbipy fluorophore in the other
ligands.36 Therefore, the zinc(II)-coordination-enabled emission
bathochromic shift of 6 is smaller than those of 1ꢀ5.
F. Potentials of Ligands 1 and 3 in Imaging Zinc(II) in
Living Cells. Compounds 1ꢀ5 show little toxicity as determined
using an In Vitro Toxicology Assay Kit (sulforhodamine B based;
Sigma Tox-6). HeLa (S3) cells were incubated in either DMEM
(Dulbecco’s Modified Eagle Medium) or HBSS (Hank’s Buffered
Salt Solution) in the presence of various ligand molecules up to
10 μM, which is beyond the upper limit of the concentration
range of the indicator used in the fluorescence imaging experi-
ments. There is no apparent cell toxicity in DMEM during the
first 6 h of incubation as measured by SRB (Figures S11ꢀ20,
Supporting Information). After 24 h, various degrees of cell
decline that are proportional to the dosage of the indicators were
observed. The ligands show little toxicity in HBSS, in which the
HeLa (S3) cells were treated with the ligands for up to 6 h. We
concluded that the toxicity of ligands 1ꢀ5 at low micromolar
loading concentrations is insignificant over the time scale of the
imaging experiments, which last up to 1 h.
The abilities of ligands 1ꢀ5 to fluorometrically follow mam-
malian intracellular zinc(II) concentration as affected by the
medium zinc(II) content were evaluated using HeLa (S3) cells as
a model system. Invariably, the supplementation of zinc(II) in
the media drives up the intracellular fluorescence. The concen-
tration dependence of the fluorescence intensity varies in each
case. The imaging data of compounds 1 (nonfluorinated) and 3
(difluorinated) are discussed in the following text. The data
collected using other ligands are included in the Supporting
Information (Figures S21ꢀ23).
Fluorination of the high-affinity binding site of ligand 1 lowers
its affinity to zinc(II) (see the values of pKd1 of ligands 1 and 6 vs
HeLa (S3) cells were incubated in HBSS buffer containing
2 μM of compound 1 for 30 min at 37 °C. After the indicator
10499
dx.doi.org/10.1021/ic201728f |Inorg. Chem. 2011, 50, 10493–10504

Inorganic Chemistry
ARTICLE
Figure 6. Fluorescent images of HeLa (S3 line) cells loaded with nonfluorinated indicator 1 (2.0 μM) in the presence of supplemental ZnCl2 (0ꢀ
50 μM) in HBSS. Scale bar in A = 10 μm. Excitation 355ꢀ405 nm; emission 420ꢀ480 nm.
molecules were removed by replacing fresh media (HBSS) with
graded ZnCl2 concentrations, the samples were imaged after
10 min of incubation at 37 °C. Imaged using the emission filter
set QMax Blue, whose emission spectral window covers 420ꢀ
460 nm, the collected signals primarily come from the long
emission band centered at 433 nm (λ2 in Table 2). The short-
wavelength emission band of 1 centers at 405 nm, which is largely
outside the detection range of the QMax Blue filter set. Upon
increasing the supplemental zinc(II) concentration ([Zn]s) in
the media, a monotonous fluorescence enhancement was ob-
served (Figure 6). The mean fluorescence intensity per cell was
plotted against [Zn]s (Figure 7). The zinc(II)-mediated fluores-
cence enhancement (I/I0) vs [Zn]s is shown in Figure S24
(Supporting Information).
During the initial phase of increasing [Zn]s, the enhancement
of intensity from indicator 1 is moderate. The intensity starts to
increase sharply at 25 μM onward to 50 μM. We interpreted that
when [Zn]s is below 20 μM, the fluorescence enhancement
occurs at the shorter wavelength band which centers at 405 nm.
Figure 7. Mean fluorescence intensity per cell (on a scale of 0ꢀ255) of
This emission band is largely outside the coverage of the QMax
Blue filter set. Therefore, little intensity change is discernible in
indicators 1 (blue) and 3 (red) vs supplemental [ZnCl2] ([Zn]s) in the
media. A minimum of 30 cells were analyzed in image J for each
data point.
the first three (AꢀC) frames. Due to the large affinity gap
between the high- and low-affinity sites in 1 (pKd1 = 12.8, pKd2
=
5.2), the fluorescence intensity does not vary significantly during
the intermediate range of the zinc(II) gradient (DꢀF). When the
[Zn]s is high enough for zinc(II) to occupy the low-affinity site,
the longer emission band centered at 433 nm starts to rise, which
is captured by the QMax Blue filter set as seen in the last three
frames (GꢀI) and in Figure 7.
regime is not long enough to be captured fully using the
commercial QMax Blue filter set. Therefore, the fluorescence
information in the low concentration regime remains largely
uncovered. These limitations will be addressed in the next phase
of this project.
The difluorinated ligand 3 also displays a fluorescence en-
hancement over an increasing zinc(II) gradient (Figure 8). The
dynamic range (Figure 7, red trace) is smaller compared to that of
the nonfluorinated 1. The intracellular distribution is also visibly
different. Ligand 3 tends to form bright particulates intracellularly.
As [Zn]s increases, the density of the particulates grows, which
gives rise to an overall higher mean intensity. However, because
the fluorescence intensity of most of the particulates is at the
This experiment demonstrated that the intracellular fluore-
scence of 1 is dependent on the concentration of supplemental
zinc(II). The overall zinc(II)-mediated fluorescence enhance-
ment is up to 17-fold, which constitutes a large signal dynamic
range. The limitation of compound 1 is its inability to be analyzed
by two separate emission filter sets because the emission band
separation is not wide enough. Furthermore, the wavelength of
the emission that responds to zinc(II) in the low concentration
10500
dx.doi.org/10.1021/ic201728f |Inorg. Chem. 2011, 50, 10493–10504

Inorganic Chemistry
ARTICLE
Figure 8. Fluorescent images of HeLa (S3 line) cells loaded with difluorinated indicator 3 (2.0 μM) in the presence of supplemental ZnCl2 (0ꢀ50 μM)
in HBSS. Scale bar in A = 10 μm. Excitation 355ꢀ405 nm; emission 420ꢀ480 nm.
Table 3. Subcellular Localization Profiles, As Represented by Pearson’s Correlation Coefficients (PCC),75 of Indicators 1 and 3
under Zinc(II)-Deficient and Zinc(II)-Enriched Conditions
nucleus
mitochondrion
Golgi complex
endoplasmic reticulum
endosome
1
17 ( 5%
44 ( 12%
7 ( 7%
90 ( 4%
68 ( 5%
58 ( 10%
64 ( 5%
74 ( 7%
66 ( 15%
66 ( 10%
67 ( 8%
86 ( 5%
84 ( 7%
71 ( 8%
77 ( 4%
82 ( 3%
79 ( 5%
58 ( 5%
75 ( 7%
1 + Zn2+
3
3 + Zn2+
42 ( 7%
saturation level, the mean intensity is underestimated, which leads
to an overall smaller dynamic range than that of ligand 1. We
attribute the particulate formation to the tendency of fluorinated
ligands (2ꢀ4) to aggregate. This hypothesis is consistent with the
observation of a long wavelength tail of the absorption spectrum of
the difluorinated ligand 3 at 10 μM in a neutral aqueous solution
(Figure S25A, Supporting Information), which is an indication of
inadequate solubility under the applied conditions. On the basis of
the concentration dependence of absorption at 480 nm (Figure
S25B, Supporting Information), presumably solely arising from
aggregation, we estimated that the solubilities of 1 and 3 under the
working conditions are 10 μM and 2 μM, respectively. The actual
origin of the fluorescence of the particulates (or aggregates)
requires detailed photophysical characterization which is beyond
the scope of this report.
In summary, the following beneficial features of zinc(II)
indicators 1ꢀ5 in fluorescence imaging experiments of living
cells are demonstrated: (1) Very low loading concentrations
(e.g., 2 μM) are required, which provides little disturbance to the
native state of the cellular sample under analysis.72 Most reported
indicators for zinc(II) including the commercial products are
often recommended for use at 5 μM or higher loading levels. (2)
The uptake of the indicator molecules is rapid; a 30-min
incubation is sufficient. (3) The toxicities of 1ꢀ5 are low. On
the other hand, the following deficiencies of 1ꢀ5 might hamper a
fluorescence imaging experiment: (a) Short wavelength, high-
energy excitation is needed, which leads to photobleaching over
time. (b) The particulate formation of fluorinated ligands and the
lack thereof of nonfluorinated 1 in living cells are difficult to
predict and to characterize chemically. (c) The uncertainty in
subcellular localization needs to be addressed.
High spatial resolution is one of the most distinctive advan-
tages of fluorescence microscopy over other minimally invasive
imaging methods for analyzing living biological samples.76,77
The subcellular localization properties of ligands 1 and 3 in five
organelles, nucleus, mitochondrion, Golgi complex, endoplasmic
reticulum, and endosome, were characterized (Table 3). In each
case, fusions of the red fluorescent protein, mCherry (610 nm
emission maximum),78 were used to determine the degree of
colocalization with the indicators.
Ligand 1 shows high degrees of localization (Pearson’s cor-
relation coefficient (PCC) over 70%) in four organelles—
mitochondrion (Figure 10), endoplasmic reticulum (Figures S27
and S28, Supporting Information), Golgi complex (Figures S29
and S30, Supporting Information), and endosome (Figures S31
and S32, Supporting Information). Little localization was found
in the nucleus (Figure 9AꢀD, PCC = 17%). Upon zinc(II)
supplementation at 50 μM in the media, the degree of colocaliza-
tion in nuclei increases sharply (PCC = 44%; Figure 9EꢀH),
which correlates with the slight reduction of the PCC values in
the other four organelles. This correlation might indicate a
translocation of the zinc(II) complex of 1 from cytoplasmic
organelles to the nucleus, due to the enhanced intercalating
ability of 1 upon binding zinc(II). For compound 3 (Figures
S33ꢀS42, Supporting Information), the degrees of localization
are lower than those of 1 across the board. However, the PCC
10501
dx.doi.org/10.1021/ic201728f |Inorg. Chem. 2011, 50, 10493–10504

Inorganic Chemistry
ARTICLE
Figure 9. Images of HeLa (S3 line) cells. From left to right: DIC images; green channels representing the emission from indicator 1 (2.0 μM) captured
by QMax Blue filter set (excitation 355ꢀ405 nm; emission 420ꢀ480 nm); red channels showing the emission from mCherryꢀH2BꢀN-6 that localizes
in nuclei, captured by QMax Red filter set (excitation 530ꢀ570 nm, emission 600ꢀ650 nm); images after merging green and red channels. Scale bar =
10 μm. (AꢀD): [Zn]s = 0. (EꢀH): [Zn]s = 50 μM.
Figure 10. Images of HeLa (S3 line) cells. From left to right: DIC images; green channels representing the emission from indicator 1 (2.0 μM) captured
by QMax Blue filter set (excitation 355ꢀ405 nm; emission 420ꢀ480 nm); red channels showing the emission from mCherryꢀPDHA1ꢀN-10 that
localize in mitochondria, captured by QMax Red filter set (excitation 530ꢀ570 nm, emission 600ꢀ650 nm); images after merging green and red
channels. Scale bar = 10 μm. (AꢀD): [Zn]s = 0. (EꢀH): [Zn]s = 50 μM.
increases uniformly upon zinc(II) addition, an indication of the
organelles picking up the zinc(II) complex of 3 from cytosol. The
aggregates of ligand 3 in cytosol may be broken up to some
degrees upon dizinc(II) complex formation, which is then
distributed to various organelles, giving rise to the increase of
localizations in these organelles.
formation. The fluorescence enhancement factor of 1 is low (ϕ1/
ϕ0 = 1.9). In this work, we demonstrated a strategy to increase
ϕ1/ϕ0, which consequently leads to a high sensitivity of zinc(II)
quantification at a low concentration regime. We found that
fluorination of the pyridyl groups in the pentadentate N,N,N0-
tris(pyridylmethyl)ethyleneamino group in 1, a well-known ligand
for zinc(II) and an electron donor in designs of photoinduced
electron transfer (PET)-switching based metal ion indicators,
leads to reduced pKa values of the ligand. Consequently, quench-
ing via PET of the attached fluorophore, in the current case
phenylvinylꢀbipyridyl, becomes more efficient under pH neutral
conditions, which leads to a larger fluorescence enhancement
’ SUMMARY
Heteroditopic indicators 1ꢀ6 are evaluated by two para-
meters: the fluorescence enhancement factor, which is the
fluorescence quantum yield increase upon monozinc(II) com-
plex formation, and emission band shift upon dizinc(II) complex
10502
dx.doi.org/10.1021/ic201728f |Inorg. Chem. 2011, 50, 10493–10504

Inorganic Chemistry
ARTICLE
when zinc(II) binds. On the other hand, overfluorination, such as
in the triply fluorinated ligand 4, leads to a much reduced zinc(II)
ion affinity of the pentadentate ligand. The fluorinated pentaden-
tate ligands introduced in this work could be readily grafted to
other metal ion receptor scaffolds. The pKa of the ligand and the
metal ion affinity (Kd) have to be balanced in choosing the strategy
of fluorinating the pyridyl group on the basis of the possible
concentration range of zinc(II) in a particular zinc(II) sensing
application.
(4) Younes, A. H.; Zhang, L.; Clark, R. J.; Davidson, M. W.; Zhu, L.
Org. Biomol. Chem. 2010, 8, 5431–5441.
(5) Rurack, K.; Bricks, J. L.; Schulz, B.; Maus, M.; Reck, G.;
Resch-Genger, U. J. Phys. Chem. A 2000, 104, 6171–6188.
(6) Rurack, K.; Koval’chuck, A.; Bricks, J. L.; Slominskii, J. L. J. Am.
Chem. Soc. 2001, 123, 6205–6206.
(7) For more published examples, see ref 2.
(8) For review articles in the past decade on the topic of zinc(II)
sensing and imaging, see refs 8ꢀ18. Kimura, E.; Aoki, S. Biometals 2001,
14, 191–204.
(9) Kikuchi, K.; Komatsu, H.; Nagano, T. Curr. Opin. Chem. Biol.
The potential of ligands 1ꢀ5 in imaging zinc(II) in living cells
was assessed. All ligands permeate through the cell membrane
readily. Little toxicity was observed over the time frame of the
imaging experiments. The ligands show the ability to report
intracellular zinc(II) variations via fluorescence change. The loading
level is as low as 2 μM, lower than the recommended dosage of most
currently known indicators for zinc(II). The major deficiencies
of these indicators are also revealed. The tendency of fluorinated
ligands to aggregate complicates the interpretation of the fluore-
scence observation. More importantly, the uncertainty in subcellular
localization of the reported indicators, as well as most products
reported in literature, compromises the unique capacity of modern
fluorescence microscopy for high resolution imaging. The need for
developing organelle-specific indicator molecules to complement
the growing spatial resolution achievable by the state-of-art fluore-
scence microscopy has been noted in recent reviews.17,79 Several
groups have recently reported various strategies in developing
organelle-specific indicators for zinc(II),80ꢀ84 which will undoubt-
edly elevate the utilities of synthetic indicators in studying intra-
cellular zinc(II) distribution and dynamics at organellar levels.
Addressing these deficiencies is the focus of the ongoing work
in our laboratory.85
2004, 8, 182–191.
(10) Jiang, P.; Guo, Z. Coord. Chem. Rev. 2004, 248, 205–229.
(11) Lim, N. C.; Freake, H. C.; Bruckner, C. Chem.—Eur. J. 2005,
11, 38–49.
(12) Thompson, R. B. Curr. Opin. Chem. Biol. 2005, 9, 526–532.
(13) Chang, C. J.; Lippard, S. J. Met. Ions Life Sci. 2006, 1, 321–370.
(14) Dai, Z.; Canary, J. W. New J. Chem. 2007, 31, 1708–1718.
(15) Carol, P.; Sreejith, S.; Ajayaghosh, A. Chem. Asian J. 2007,
2, 338–348.
(16) Nolan, E. M.; Lippard, S. J. Acc. Chem. Res. 2009, 42, 193–203.
(17) Tomat, E.; Lippard, S. J. Curr. Opin. Chem. Biol. 2010, 14,
225–230.
(18) Xu, Z.; Yoon, J.; Spring, D. R. Chem. Soc. Rev. 2010, 39, 1996–
2006.
(19) Frederickson, C. J. Int. Rev. Neurobiol. 1989, 31, 145–238.
(20) See an interpretation of “free” or “mobile” zinc(II): Krez_el, A.;
)
Maret, W. J. Biol. Inorg. Chem. 2006, 11, 1049–1062.
(21) Outten, C. E.; O’Halloran, T. V. Science 2001, 292, 2488–2492.
(22) Finney, L. A.; O’Halloran, T. V. Science 2003, 300, 931–936.
(23) Frederickson, C. J.; Klitenick, M. A.; Manton, W. I.; Kirkpatrick,
J. B. Brain Res. 1983, 273, 335–339.
(24) Assaf, S. Y.; Chung, S.-H. Nature 1984, 308, 734–736.
(25) Qian, W.-J.; Aspinwall, C. A.; Battiste, M. A.; Kennedy, R. T.
Anal. Chem. 2000, 72, 711–717.
(26) Li, Y.; Hough, C. J.; Suh, S. W.; Sarvey, J. M.; Frederickson, C. J.
J. Neurophysiol. 2001, 86, 2597–2604.
’ ASSOCIATED CONTENT
(27) Sindreu, C. B.; Varoqui, H.; Erickson, J. D.; Pꢀerez-Clausell, J.
Cereb. Cortex 2003, 13, 823–829.
(28) Frederickson, C. J.; Koh, J.-Y.; Bush, A. I. Nat. Rev. Neurosci.
2005, 6, 449–462.
(29) Que, E. L.; Domaille, D. W.; Chang, C. J. Chem. Rev. 2008,
108, 1517–1549.
S
Supporting Information. Procedures for syntheses,
b
characterizations of new compounds, and additional spectra
and images. This material is available free of charge via the
Internet at http://pubs.acs.org.
(30) Shults, M. D.; Pearce, D. A.; Imperiali, B. J. Am. Chem. Soc.
2003, 125, 10591–10597.
’ AUTHOR INFORMATION
(31) Komatsu, K.; Kikuchi, K.; Kojima, H.; Urano, Y.; Nagano, T.
J. Am. Chem. Soc. 2005, 127, 10197–10204.
Corresponding Author
*E-mail: davidson@magnet.fsu.edu, lzhu@chem.fsu.edu.
(32) Nolan, E. M.; Jaworski, J.; Okamoto, K.-I.; Hayashi, Y.; Sheng,
M.; Lippard, S. J. J. Am. Chem. Soc. 2005, 127, 16812–16823.
(33) Zhang, L.; Murphy, C. S.; Kuang, G.-C.; Hazelwood, K. L.;
Constantino, M. H.; Davidson, M. W.; Zhu, L. Chem. Commun
2009, 7408–7410.
(34) de Silva, A. P.; Moody, T. S.; Wright, G. D. Analyst 2009, 134,
2385–2393.
(35) Younes, A. H.; Zhang, L.; Clark, R. J.; Zhu, L. J. Org. Chem.
2009, 74, 8761–8772.
(36) Zhang, L.; Zhu, L. J. Org. Chem. 2008, 73, 8321–8330.
(37) See an example of fluorine substitution in altering the coordina-
tion chemistry of a pyridyl-containing multidentate ligand: Sumaleksh-
my, S.; Henary, M. M.; Siegel, N.; Lawson, P. V.; Wu, Y.; Schmidt, K.;
Bredas, J.-L.; Perry, J. W.; Fahrni, C. J. J. Am. Chem. Soc. 2007, 129,
11888–11889.
’ ACKNOWLEDGMENT
This work was supported, in part, by a New Investigator
Research grant from the James and Esther King Biomedical
Research Program administered by the Florida Department of
Health (08KN-16), National Science Foundation (CHE0809201),
and National Institute of General Medical Sciences (R01GM08-
1382). We thank J. Tyler Simmons for determining the aqueous
solubility of ligands 1 and 3.
’ REFERENCES
(1) Zhang, L.; Clark, R. J.; Zhu, L. Chem.—Eur. J. 2008, 14, 2894–
2903.
(38) Kuang, G.-C.; Michaels, H. A.; Simmons, J. T.; Clark, R. J.; Zhu,
L. J. Org. Chem. 2010, 75, 6540–6548.
(2) Zhu, L.; Zhang, L.; Younes, A. H. Supramol. Chem. 2009, 21,
268–283.
(3) Wandell, R. J.; Younes, A. H.; Zhu, L. New J. Chem. 2010, 34,
2176–2182.
(39) Kawabata, E.; Kikuchi, K.; Urano, Y.; Kojima, H.; Odani, A.;
Nagano, T. J. Am. Chem. Soc. 2005, 127, 818–819.
(40) The fluorescence at the high and low ends of the pH range
deviates from the fitting. At the low end, the two pyridyl moieties in bipy
10503
dx.doi.org/10.1021/ic201728f |Inorg. Chem. 2011, 50, 10493–10504

Inorganic Chemistry
ARTICLE
are individually protonated, thus breaking the planarity of the fluor-
ophore to affect the fluorescence. At the high end of the pH range, the
aggregation of the ligand may have become an issue which affects its
fluorescence. Therefore, the fitting of the data in the range of 7.5ꢀ3.5,
which is the pH range of interest in this study and related fluorescence
microscopy experiments, affords the best estimates of the pKa values of
the two metal coordination sites.
(41) Chang, C. J.; Nolan, E. M.; Jaworski, J.; Burdette, S. C.; Sheng,
M.; Lippard, S. J. Chem. Biol. 2004, 11, 203–210.
(42) Bozym, R. A.; Thompson, R. B.; Stoddard, A. K.; Fierke, C. A.
ACS Chem. Biol. 2006, 1, 103–111.
(73) [Zn]f’s were calculated using the MaxChelator program avail-
able at http://www.stanford.edu/∼cpatton/webmaxc/webmaxcS.htm.
Patton, C.; Thompson, S.; Epel, D. Cell Calcium 2004, 35, 427–431.
(74) Bipy may bind zinc(II) in a stoichiometry from 1:1 to 3:1
(ligand/zinc(II)) depending on the counteranion and solvent, see: Eorn,
G. H.; Park, H. M.; Hyun, M. Y.; Jang, S. P.; Kim, C.; Lee, J. H.; Lee, S. J.;
Kim, S.-J.; Kim, Y. Polyhedron 2011, 30, 1555–1564. In CH3CN, a 1:1
binding stoichiometry between ZnCl2 and bipy was reported. See:
Younes, A. H.; Zhang, L.; Clark, R. J.; Zhu, L. J. Org. Chem. 2009, 74,
8761–8772.
(75) Dunn, K. W.; Kamocka, M. M.; McDonald, J. H. Am. J. Physiol.
(43) Krez_el, A.; Maret, W. J. Am. Chem. Soc. 2007, 129, 10911–10921.
Cell Physiol. 2010, 300, C723–C742.
)
(44) Krez_el, A.; Maret, W. J. Biol. Inorg. Chem. 2008, 13, 401–409.
(45) Grynkiewicz, G.; Poenie, M.; Tsien, R. Y. J. Biol. Chem. 1985,
260, 3440–3450.
(76) Fernꢀandez-Suꢀarez, M.; Ting, A. Y. Nat. Rev. Mol. Cell Biol. 2008,
9, 929–943.
(77) McRae, R.; Bagchi, P.; Sumalekshmy, S.; Fahrni, C. J. Chem.
)
(46) Thompson, R. B.; Maliwal, B. P.; Fierke, C. A. Anal. Chem.
Rev. 2009, 109, 4780–4827.
1998, 70, 1749–1754.
(47) Thompson, R. B.; Maliwal, B. P.; Fierke, C. A. Anal. Biochem.
1999, 267, 185–195.
(48) Fahrni, C. J.; O’Halloran, T. V. J. Am. Chem. Soc. 1999, 121,
11448–11458.
(49) Walkup, G. K.; Burdette, S. C.; Lippard, S. J.; Tsien, R. Y. J. Am.
(78) The emission of ligand 1 or 3 falls largely within 400ꢀ450 nm
where mCherry barely absorbs (Figure S26, Supporting Information).
Therefore, the probability of intermolecular FRET between the indi-
cator at a micromolar concentration and mCherry is low.
(79) Domaille, D. W.; Que, E. L.; Chang, C. J. Nat. Chem. Biol. 2008,
4, 168–175.
Chem. Soc. 2000, 122, 5644–5645.
(50) Hirano, T.; Kikuchi, K.; Urano, Y.; Higuchi, T.; Nagano, T.
J. Am. Chem. Soc. 2000, 122, 12399–12400.
(51) Burdette, S. C.; Walkup, G. K.; Spingler, B.; Tsien, R. Y.;
Lippard, S. J. J. Am. Chem. Soc. 2001, 123, 7831–7841.
(52) Gee, K. R.; Zhou, Z.-L.; Qian, W.-J.; Kennedy, R. T. J. Am.
Chem. Soc. 2002, 124, 776–778.
(80) Tomat, E.; Nolan, E. M.; Jaworski, J.; Lippard, S. J. J. Am. Chem.
Soc. 2008, 130, 15776–15777.
(81) Dittmer, P. J.; Miranda, J. G.; Gorski, J. A.; Palmer, A. E. J. Biol.
Chem. 2009, 284, 16289–16297.
(82) Vinkenborg, J. L.; Nicolson, T. J.; Bellomo, E. A.; Koay, M. S.;
Rutter, G. A.; Merkx, M. Nat. Methods 2009, 6, 737–U10.
(83) Qin, Y.; Dittmer, P. J.; Park, J. G.; Jansen, K. B.; Palmer, A. E.
Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 7351–7356.
(53) Gee, K. R.; Zhou, Z.-L.; Ton-That, D.; Sensi, S.; Weiss, J. H. Cell
Calcium 2002, 31, 245–251.
(84) Masanta, G.; Lim, C. S.; Kim, H. J.; Han, J. H.; Kim, H. M.; Cho,
(54) Dai, Z.; Xu, X.; Canary, J. W. Chem. Commun. 2002, 1414–1415.
(55) Hirano, T.; Kikuchi, K.; Urano, Y.; Nagano, T. J. Am. Chem. Soc.
2002, 124, 6555–6562.
B. R. J. Am. Chem. Soc. 2011, 133, 5698–5700.
(85) Sreenath, K.; Allen, J. R.; Davidson, M. W.; Zhu, L. Chem.
Commun., DOI: 10.1039/c1cc14580k.
(56) Maruyama, S.; Kikuchi, K.; Hirano, T.; Urano, Y.; Nagano, T.
J. Am. Chem. Soc. 2002, 124, 10650–10651.
(57) Burdette, S. C.; Frederickson, C. J.; Bu, W.; Lippard, S. J. J. Am.
Chem. Soc. 2003, 125, 1778–1787.
(58) Chang, C. J.; Jaworski, J.; Nolan, E. M.; Sheng, M.; Lippard, S. J.
Proc. Natl. Acad. Sci. U. S. A. 2004, 101, 1129–1134.
(59) Henary, M. M.; Wu, Y.; Fahrni, C. J. Chem.—Eur. J. 2004,
10, 3015–3025.
(60) Hanaoka, K.; Kikuchi, K.; Kojima, H.; Urano, Y.; Nagano, T.
J. Am. Chem. Soc. 2004, 126, 12470–12476.
(61) Taki, M.; Wolfold, J. L.; O’Halloran, T. V. J. Am. Chem. Soc.
2004, 126, 712–713.
(62) Royzen, M.; Durandin, A.; Young, V. G. J.; Geacintov, N. E.;
Canary, J. W. J. Am. Chem. Soc. 2006, 128, 3854–3855.
(63) Goldsmith, C. R.; Lippard, S. J. Inorg. Chem. 2006, 45, 555–561.
(64) Parkesh, R.; Lee, T. C.; Gunnlaugsson, T. Org. Biomol. Chem.
2007, 5, 310–317.
(65) Komatsu, K.; Urano, Y.; Kojima, H.; Nagano, T. J. Am. Chem.
Soc. 2007, 129, 13447–13454.
(66) Wang, H.-H.; Gan, Q.; Wang, X.-J.; Xue, L.; Liu, S.-H.; Jiang, H.
Org. Lett. 2007, 9, 4995–4998.
(67) Xue, L.; Wang, H.-H.; Wang, X.-J.; Jiang, H. Inorg. Chem. 2008,
47, 4310–4318.
(68) Kim, H. M.; Seo, M. S.; An, M. J.; Hong, J. H.; Tian, Y. S.; Choi,
J. H.; Kwon, O.; Lee, K. J.; Cho, B. R. Angew. Chem., Int. Ed. 2008,
47, 5167–5170.
(69) Mizukami, S.; Okada, S.; Kimura, S.; Kikuchi, K. Inorg. Chem.
2009, 48, 7630–7638.
(70) Taki, M.; Watanabe, Y.; Yamamoto, Y. Tetrahedron Lett. 2009,
50, 1345–1347.
(71) Xu, Z.; Baek, K.-H.; Kim, H. N.; Cui, J.; Qian, X.; Spring, D. R.;
Shin, I.; Yoon, J. J. Am. Chem. Soc. 2010, 132, 601–610.
(72) Dineley, K. E.; Malaiyandi, L. M.; Reynolds, I. J. Mol. Pharmacol.
2002, 62, 618–627.
10504
dx.doi.org/10.1021/ic201728f |Inorg. Chem. 2011, 50, 10493–10504