Electronic Effects of Ligand Substitution in a Family of CoII2 PARACEST pH Probes

Article abstract of DOI:10.1021/acs.inorgchem.8b01896

We report three new Co₂-based paramagnetic chemical exchange saturation transfer (PARACEST) probes with the ability to ratiometrically quantitate pH. A Co^II₂ complex, [LCo₂(etidronate)]^-, featuring tetra(carboxamide) and OH-substituted etidronate ligands with opposing pH-dependent CEST peak intensities, was previously shown to exhibit a linear correlation between log(CEST_OH/CEST_NH) and pH in the pH range 6.5-7.6 that provided a sensitivity of 0.99(7) pH unit^-1 at 37 °C. Here, we demonstrate through a series of CF₃-functionalized Co^II₂ complexes [(^XL′)Co₂(etidronate)]^- (X = NO₂, F, Me), that modest changes in the electronic structure of Co^II centers through remote ligand substitution can significantly affect the NMR and CEST properties of Co₂-based PARACEST probes. Variable-pH NMR and CEST analyses reveal that the chemical shifts of the ligand protons are highly affected by the nature of the X substituent. The ratios of OH and NH CEST peak intensities at 115 and 88, 93 and 79, and 88 and 76 ppm for X = NO₂, F, and Me, respectively, afford pH calibration curves with remarkably high sensitivities of 1.49(9), 1.48(7), and 2.04(5) pH unit^-1 across the series. The 1.5-2-fold enhancement in pH sensitivity for the CF₃-functionalized Co₂ probes stems from the complete separation of the OH and NH CEST peaks. Furthermore, incorporation of electron-withdrawing CF₃ groups shifts the detection window to a more acidic range of pH 6.2-7.4. Finally, the Co^II₂ complexes are found to be extremely robust toward substitution and oxidation in aqueous solutions. Taken together, these results highlight the unique ability of transition metal-based PARACEST probes to provide a highly sensitive concentration-independent measure of pH and demonstrate that modest ligand modifications can be a powerful tool for optimizing the pH sensing performance of these probes.

Full text of DOI:10.1021/acs.inorgchem.8b01896

Article
Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX
pubs.acs.org/IC
Electronic Effects of Ligand Substitution in a Family of CoII2
PARACEST pH Probes
Agnes E. Thorarinsdottir, Scott M. Tatro, and T. David Harris*
Department of Chemistry, Northwestern University, Evanston, Illinois 60208, United States
*
S Supporting Information
ABSTRACT: We report three new Co -based paramagnetic
2
chemical exchange saturation transfer (PARACEST) probes
II
with the ability to ratiometrically quantitate pH. A Co
2
−
complex, [LCo (etidronate)] , featuring tetra(carboxamide)
2
and OH-substituted etidronate ligands with opposing pH-
dependent CEST peak intensities, was previously shown to
exhibit a linear correlation between log(CESTOH/CESTNH
)
and pH in the pH range 6.5−7.6 that provided a sensitivity of
−
1
0
.99(7) pH unit at 37 °C. Here, we demonstrate through a
II
X
series of CF -functionalized Co
complexes [( L′)-
3
2
−
Co (etidronate)] (X = NO , F, Me), that modest changes
2
2
II
in the electronic structure of Co centers through remote
ligand substitution can significantly affect the NMR and CEST
properties of Co -based PARACEST probes. Variable-pH NMR and CEST analyses reveal that the chemical shifts of the ligand
2
protons are highly affected by the nature of the X substituent. The ratios of OH and NH CEST peak intensities at 115 and 88,
9
3 and 79, and 88 and 76 ppm for X = NO , F, and Me, respectively, afford pH calibration curves with remarkably high
2
−
1
sensitivities of 1.49(9), 1.48(7), and 2.04(5) pH unit across the series. The 1.5−2-fold enhancement in pH sensitivity for the
CF -functionalized Co probes stems from the complete separation of the OH and NH CEST peaks. Furthermore,
3
2
incorporation of electron-withdrawing CF groups shifts the detection window to a more acidic range of pH 6.2−7.4. Finally,
3
II
2
the Co complexes are found to be extremely robust toward substitution and oxidation in aqueous solutions. Taken together,
these results highlight the unique ability of transition metal-based PARACEST probes to provide a highly sensitive
concentration-independent measure of pH and demonstrate that modest ligand modifications can be a powerful tool for
optimizing the pH sensing performance of these probes.
INTRODUCTION
exchange is highly dependent on pH as well as the pK of
a
■
the exchangeable proton and its proximity to the transition
The realization of chemical probes with the ability to
accurately map small pH changes in vivo represents an
6
−8,11
metal ion.
exhibit exchangeable protons with large chemical shifts
and thereby improve sensitivity by minimizing interference
Furthermore, these paramagnetic probes
7c,d,8
important synthetic challenge, as acidic extracellular pH is a1
prominent feature of pathological conditions such as cancer,
12
1
e,2
1g,2c
from biological background signals. Nevertheless, the CEST
signal intensity is also affected by the concentration of the
probe, which significantly complicates in vivo studies.
ischemia,
and inflammation.
Magnetic resonance
imaging (MRI) is an ideal noninvasive imaging modality for
probing pH in vivo due to its high spatiotemporal image
resolution and unlimited tissue penetration depth. Therefore,
3
One strategy to surmount the challenge of concentration
dependence is to employ PARACEST probes featuring two
distinct types of exchangeable protons that exhibit different
pH-dependent changes in CEST signal intensity. The ratio of
the intensities of the two CEST signals provides a
spatial mapping of tissue pH through MRI may aid in the early
detection of pathologies and provide valuable information
about the progression of diseases and the efficacy of
1
,2,4
treatments.
7d,8e,13
concentration-independent measure of pH.
We recently
Toward developing pH-responsive MRI contrast agents,
transition metal complexes that function as paramagnetic
demonstrated that a Co PARACEST probe supported by a
2
5
−8
dinucleating tetra(carboxamide) ligand and an ancillary
etidronate ligand can be employed to ratiometrically map
chemical exchange saturation transfer (PARACEST) agents
are of particular interest owing to the inherent environmental
responsiveness of the exchangeable protons on these
7d
solution pH in a physiologically relevant range. The high pH
−
1
9
sensitivity of 0.99(7) pH unit for this probe is attributed to
compounds. Here, contrast is generated through exchange
of protons on a paramagnetic molecule with bulk H O protons
2
upon frequency-specific irradiation, resulting in a decrease in
Received: July 7, 2018
1
0
the bulk H O signal. Importantly, the rate of proton
2
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.8b01896
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
the base- and acid-catalyzed exchange of the carboxamide
purification. Experimental details on the syntheses of ligands and
organic precursors are provided in the Supporting Information.
(
NH) and hydroxyl (OH) protons on the two ligands,
NO
2
Synthesis of Na[( L′)Co (etidronate)]·1.0NaNO ·1.5MeOH (2-
respectively, which results in CEST intensities that are
proportional and inversely proportional to pH, respectively.
However, considering that the most downfield-shifted CEST
2
3
NO ). A pink solution of Co(NO ) ·6H O (84 mg, 0.29 mmol) in
MeOH (2 mL) was added dropwise to a stirring pale yellow solution
2
3
2
2
NO
2
of H( L′) (110 mg, 0.15 mmol) in MeOH (3 mL) to give a dark
orange solution. To this solution, a colorless solution of etidronic acid
monohydrate (32 mg, 0.14 mmol) in MeOH (2 mL) was added
dropwise, followed by addition of sodium methoxide (39 mg, 0.72
mmol) in MeOH (2 mL) to give a light orange solution. After being
stirred at 25 °C for 3 h, the orange solution was evaporated to
dryness, and the resulting orange residue was stirred in MeCN (10
mL) for 30 min. The orange solid was collected by filtration, washed
peak for this Co complex is comprised of overlapping NH and
2
OH signals, one can envision that PARACEST probes bearing
two types of exchangeable protons that give rise to well-
separated CEST peaks with opposing pH-dependent inten-
sities have the potential to exhibit much higher sensitivities.
Transition metal-based PARACEST probes offer an ideal
platform for the design of highly sensitive ratiometric pH
probes by virtue of their high environmental responsiveness
with MeCN (5 mL) and Et O (5 mL), and dried under reduced
2
6
−8
pressure for 16 h. Diffusion of Et O vapor into a concentrated
2
and excellent tunability through ligand design. In particular,
modest structural changes such as modification of the ligand
backbone or pendent groups of azamacrocyclic ligands can lead
to drastic changes in the chemical shift and intensity of CEST
solution of the orange solid in MeOH (2 mL) afforded a crystalline
light orange solid that was washed with Et O (2 × 3 mL) and dried
2
under reduced pressure for 24 h to give 2-NO (63 mg, 34%) as a
2
light orange solid. Anal. Calcd for C27.5H Co F N Na O18.5P2: C,
36
2
12
8
2
8
e,i,l
peaks.
Nevertheless, there is a dearth of studies that probe
26.89; H, 2.95; N, 9.12%. Found: C, 26.99; H, 3.01; N, 9.20%. ICP-
OES: Co:P = 0.96:1.00. UV−visible absorption spectrum (22 μM; 50
mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES)
electronic effects on the pH dependence of CEST for
paramagnetic compounds. As such, there is a significant
interest in understanding how electronic effects of ligand
substitution influence the pH behavior of transition metal-
based PARACEST probes. Combining such electronic studies
with structural investigations would enhance our under-
standing of the factors that govern the pH dependence of
the CEST effect and facilitate the design of chemical probes
with optimal pH responsiveness for targeted applications.
The observation of highly pH-sensitive CEST peak intensity
−
1
−1
buffered at pH 6.94, 25 °C): 371 nm (ε = 13 800 M cm ). ESI-
NO
2
MS (m/z): Calcd for C26H30Co2F12N O14P2 ([( L′)-
7
−
Co (etidronate)] ): 1071.98, found: 1072.01; calcd for
2
NO
2
+
C H Co F N O P ([( L′)Co (etidronate)+2H] ): 1073.99,
2
6
32
2
12
7
14
2
2
−1
found: 1074.01. FT-IR (ATR, cm ): 3248 (m); 3076 (m); 2969
w); 1738 (w); 1640 (s); 1593 (m); 1507 (w); 1398 (m); 1312 (s);
272 (m); 1156 (s); 1074 (s); 992 (w); 901 (m); 833 (m); 802 (m);
749 (m); 671 (m). Slow diffusion of Et O vapor into a concentrated
(
1
2
solution of 2-NO
in MeOH afforded light orange plate-shaped
crystals of Na[( L′)Co (etidronate)]·1.0NaNO (2′-NO ) suitable
for single-crystal X-ray diffraction analysis.
Synthesis of Na[( L′)Co (etidronate)]·2.2NaNO ·1.0H O (2-F). A
2
2
7d
NO
ratios for the previously reported Co probe, in conjunction
2
3
2
2
with the high modularity of the dinucleating ligand scaffold,
F
2
3
2
prompted us to investigate the pH dependence of CEST
II
pink solution of Co(NO
mL) was added dropwise to a stirring colorless solution of H( L′)
120 mg, 0.17 mmol) in MeOH (3 mL) to give a dark pink solution.
To this solution, a colorless solution of etidronic acid monohydrate
37 mg, 0.17 mmol) in MeOH (2 mL) was added dropwise, followed
) ·6H O (96 mg, 0.33 mmol) in MeOH (2
3 2 2
properties for Co 2 complexes bearing CF -functionalized
3
F
tetra(carboxamide) ligands. Specifically, we sought to examine
how the CEST properties of these compounds are affected by
the nature of the para-substituent on the bridging phenoxo
(
(
II
ligand. Herein, we report a series of new Co 2 complexes as
by addition of sodium methoxide (45 mg, 0.83 mmol) in MeOH (2
mL) to give a pink solution. After being stirred at 25 °C for 3 h, the
pink solution was evaporated to dryness, and the resulting pink
residue was stirred in MeCN (10 mL) for 20 min. The pink solid was
ratiometric PARACEST pH probes and demonstrate that small
II
changes in the electronic structure of Co centers through
remote ligand substitution can lead to notable differences in
CEST profiles. These studies underscore that proper ligand
design is a key attribute toward optimizing the performance of
ratiometric PARACEST pH sensors, particularly in fine-tuning
the pH sensitivity and detection range.
collected by filtration, washed with MeCN (5 mL) and Et O (5 mL),
2
and dried under reduced pressure for 16 h. Diffusion of Et O vapor
2
into a concentrated solution of the pink solid in MeOH (2 mL)
afforded a crystalline pink solid that was washed with Et O (2 × 3
2
mL) and dried under reduced pressure for 24 h to give 2-F (163 mg,
7
5 % ) a s
a
d a r k p i n k s o l i d . A n a l . C a l c d f o r
3.2 19.6
EXPERIMENTAL SECTION
C H Co F N Na O P2: C, 24.52; H, 2.53; N, 9.02%. Found:
■
26 32
2
13 8.2
General Considerations. Unless otherwise specified, the
manipulations described below were carried out at ambient
atmosphere and temperature. Air- and water-free manipulations
were performed under a dinitrogen atmosphere in a Vacuum
Atmospheres Nexus II glovebox or using standard Schlenk line
techniques. Syntheses of metal complexes were carried out in an
MBraun LABstar glovebox, operated under a humid dinitrogen
atmosphere. Glassware was oven-dried at 150 °C for at least 4 h and
allowed to cool in an evacuated antechamber prior to use in the
C, 24.73; H, 2.69; N, 9.25%. ICP-OES: Co:P = 1.01:1.00. UV−visible
absorption spectrum (87 μM; 50 mM HEPES buffered at pH 6.95, 25
−1
−1
°C): 309 nm (ε = 4400 M cm ). ESI-MS (m/z): Calcd for
F
−
C
H
30Co
F
N
O
P
([( L′)Co
(etidronate)] ): 1044.99, found:
26
2
13
6
12
2
2
H
F
1045.08; calcd for
C
3 2 Co
F
N
O
P
([( L′)-
2 6
2
1 3
6
1 2
2
+
Co (etidronate)+2H] ): 1047.00, found: 1047.04. FT-IR (ATR,
2
−
1
cm ): 3249 (m); 3078 (m); 2969 (w); 1738 (m); 1643 (s); 1575
(m); 1470 (m); 1431 (m); 1375 (m); 1352 (m); 1287 (w); 1258
(m); 1154 (s); 1074 (s); 991 (w); 907 (m); 890 (m); 834 (m); 799
gloveboxes. Acetonitrile (MeCN), diethyl ether (Et O), N,N-
(m); 673 (m). Slow diffusion of Et
2
O vapor into a concentrated
2
diisopropylethylamine (DIPEA), and methanol (MeOH) were dried
using a commercial solvent purification system from Pure Process
Technology and stored over 3 or 4 Å molecular sieves prior to use.
solution of 2-F in MeOH afforded pink prism-shaped crystals of
F
Na[( L′)Co
(etidronate)]·1.0NaNO
·1.0MeOH (2′-F) suitable for
(etidronate)]·4.0NaNO ·3.4H O (2-Me).
2
3
single-crystal X-ray diffraction analysis.
Me
H O was obtained from a purification system from EMD Millipore.
Synthesis of Na[( L′)Co
2
3
2
2
Deuterated solvents were purchased from Cambridge Isotope
Laboratories. The syntheses of N,N′-[(2-hydroxy-5-nitro-1,3-
phenylene)bis(methylene)]bis[N-(carboxymethyl)glycineamide]
A pink solution of Co(NO
)
3
2
·6H O (79 mg, 0.27 mmol) in MeOH
2
Me
(2 mL) was added dropwise to a stirring colorless solution of H( L′)
(97 mg, 0.13 mmol) in MeOH (3 mL) to give a dark pink solution.
To this solution, a colorless solution of etidronic acid monohydrate
(33 mg, 0.15 mmol) in MeOH (2 mL) was added dropwise, followed
by addition of sodium methoxide (36 mg, 0.67 mmol) in MeOH (2
(
HL) and Na[LCo (etidronate)]·0.2NaNO ·2.7H O (1) were carried
2 3 2
7d
out as reported previously. All other reagents and solvents were
purchased from commercial vendors and used without further
B
DOI: 10.1021/acs.inorgchem.8b01896
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
mL) to give a pink solution. After being stirred at 25 °C for 3 h, the
pink solution was evaporated to dryness, and the resulting pink
residue was stirred in MeCN (10 mL) for 20 min. The pink solid was
Avance III 500 MHz (11.7 T) system equipped with a DCH
CryoProbe. H and F NMR spectra for compounds 2-X (X = NO2,
1
19
F, Me) in D O and for aqueous solution samples containing 50 mM
2
collected by filtration, washed with MeCN (5 mL) and Et O (5 mL),
HEPES and 100 mM NaCl buffered at various pH values were
collected on an Agilent DD2 500 MHz (11.7 T) spectrometer. For
samples in buffer, spectra were acquired using an inner capillary
2
and dried under reduced pressure for 20 h. Diffusion of Et O vapor
2
into a concentrated solution of the pink solid in MeOH (2 mL)
afforded a crystalline pink solid that was washed with Et O (2 × 3
containing a solution of 3% (v/v) trifluoroacetic acid (TFA) in D O
2
2
mL) and dried under reduced pressure for 24 h to give 2-Me (59 mg,
to lock the samples. All chemical shift values (δ) are reported in ppm
1
3
1%) as a pink solid. Anal. Calcd for C H Co F N Na O P :
and coupling constants (J) are reported in Hertz (Hz). H NMR
2
7
39.8
2
12 10
5
27.4 2
X
C, 22.13; H, 2.74; N, 9.56%. Found: C, 22.03; H, 2.56; N, 9.44%.
ICP-OES: Co:P = 0.92:1.00. UV−visible absorption spectrum (42
μM; 50 mM HEPES buffered at pH 6.95, 25 °C): 307 nm (ε = 3900
spectra for H( L′) and organic precursors are referenced to residual
proton signals from the deuterated solvents (7.26 ppm for CDCl ,
3
4.79 ppm for D
and 1.94 ppm for CD
CD CN are referenced to the CD
solvent (δ = 1.32 ppm). F NMR spectra for ligands and organic
precursors are referenced to an external standard of CFCl (δ = 0
ppm). F NMR spectra for 2-X are referenced to an internal standard
O or H O, the
O, 3.31 ppm for CD
OD, 2.50 ppm for (CDX
CN). C{ H} NMR spectra for H( L′) in
carbon signal from the deuterated
) SO,
2
3
3 2
−
1
−1
13
1
M
(
cm ). ESI-MS (m/z): Calcd for C H Co F N O P
3
2
7
33
2
12
6
12 2
[(MeL′)Co (etidronate)] ): 1041.01, found: 1041.15; calcd for
−
2
3
3
Me
+
19
C H Co F N O P ([( L′)Co (etidronate)+2H] ): 1043.03,
2
7
35
2
12
6
12
2
2
−
1
found: 1043.06. FT-IR (ATR, cm ): 3254 (m); 3094 (m); 2970
w); 1739 (m); 1638 (s); 1568 (m); 1476 (w); 1428 (m); 1372 (m);
350 (m); 1317 (m); 1254 (m); 1157 (s); 1081 (s); 1017 (w); 995
3
1
9
(
1
of TFA at −76.00 ppm. For measurements of 2-X in D
2
2
1
(
(
w); 952 (w); 916 (m); 881 (m); 833 (m); 799 (m); 744 (w); 671
chemical shift of the solvent signal in the H NMR spectra was set to
1
m); 650 (w). Slow diffusion of Et O vapor into a concentrated
0 ppm to simplify comparison between H NMR spectra and the
2
solution of 2-Me in MeOH afforded pink plate-shaped crystals of
corresponding CEST spectra (Z spectra). The MestReNova 10.0
NMR data processing software was used to analyze and process all
recorded NMR spectra.
Me
Na[( L′)Co (etidronate)]·solvent (2′-Me) suitable for single-crystal
2
X-ray diffraction analysis. Note that the term “solvent” in the formula
above denotes a combination of crystallographically disordered
MeOH, Et O, and H O molecules (see below).
Determination of pK
by 1H NMR Analysis. The pH-
a
1
dependent H NMR chemical shift of the Me resonance from
etidronate was used to estimate the pK values for compounds 2-X (X
, F, Me). The change in H NMR chemical shift of the Me
resonance as a function of pH was fitted to a Boltzmann sigmoidal
2
2
X-ray Structure Determination. Single crystals of 2′-X (X =
a
1
NO , F, Me) were directly coated with Paratone-N oil, mounted on a
= NO
2
2
MicroMounts rod, and frozen under a stream of dinitrogen during
data collection. The crystallographic data were collected at 100 K on a
Bruker Kappa Apex II diffractometer equipped with an APEX-II CCD
18
function for each compound to model a single ionization event
according to the following equation:
detector and a Mo Kα IμS microsource with MX Optics (2′-NO , 2′-
2
Me), or on a Bruker Kappa Apex II diffractometer equipped with an
APEX-II CCD detector and a MoKα sealed tube source with a
Triumph monochromator (2′-F). Raw data were integrated and
corrected for Lorentz and polarization effects with Bruker APEX2
A1 − A
2
δ = A +
2
1 + exp pH − pK /dx)
(
(
)
(1)
a
In this equation, δ is the obtained chemical shift, A is the theoretical
chemical shift of the fully deprotonated species, A is the theoretical
chemical shift of the fully protonated species, pK is the inflection
point of the graph, and dx is a parameter describing the steepness of
the curve.
2
14
version 2014.11−0.1 Absorption corrections were applied using the
1
5
program SADABS. Space group assignments were determined by
a
examining systematic absences, E-statistics, and successive refinement
of the structures. Structures were solved using direct methods in
16
SHELXT and refined by SHELXL operated within the OLEX2
CEST Experiments. All CEST experiments were carried out at 37
17
interface. All hydrogen atoms were placed at calculated positions
using suitable riding models and refined using isotropic displacement
parameters derived from their parent atoms. Thermal parameters for
all non-hydrogen atoms were refined anisotropically.
°C on an Agilent DD2 500 MHz (11.7 T) spectrometer. For CEST
experiments, 4 mM or 8−9 mM samples of 2-X (X = NO , F, Me) in
2
aqueous buffer solutions containing 50 mM HEPES and 100 mM
1
NaCl at desired pH values (measured with a pH electrode before H
In the crystal structure of 2′-Me, the solvent molecules were
severely disordered and could not be modeled properly. Therefore,
the solvent masking procedure as implemented in OLEX2 was used.
NMR and CEST data collection) were measured. Z-spectra (CEST
spectra) were obtained according to the following protocol: H NMR
1
spectra were acquired from −50 to 150 ppm with a step increase of 1
3
Two void volumes of 955.4 Å , each with 242 electrons, were
ppm using a presaturation pulse applied for 4 s at a power level (B )
1
estimated per unit cell and ascribed to a combination of MeOH, Et O,
of 21 μT. An inner capillary containing a solution of 3% (v/v) TFA in
2
and H O solvent molecules. Due to this disorder, the nomenclature of
D O was placed within the NMR sample tube to lock the sample. The
2
2
Me
the single crystals of 2′-Me is noted as Na[( L′)Co (etidronate)]·
normalized integrations of the H O signal from the obtained spectra
2
2
+
solvent. The Na ions in all three crystal structures were disordered,
were plotted against frequency offset to generate a Z-spectrum, where
resulting in large atomic displacement parameters. Specifically, in the
direct saturation of the H O signal was set to 0 ppm.
2
+
crystal structure of 2′-Me, the occupancy of Na was freely refined
Exchange rate constants (k ) were calculated following a
ex
1
9
2
ex
over two positions using the EADP constraint, and in the crystal
previously reported method, where the x-intercept (−1/k ) was
+
structure of 2′-F, three out of four Na ions in the asymmetric unit
obtained from a plot of M /(M − M ) (M and M are the
z
0
z
z
0
were positionally disordered over two positions and their occupancy
was freely refined over the two positions. In the crystal structure of 2′-
NO , the CF groups and NaNO cocrystallite were severely
magnetization of the on- and off-resonance, respectively) against 1/
2
−1
1
ω1 (ω1 in rad s ). H NMR spectra were acquired at various
presaturation power levels ranging from 13 to 21 μT applied for 4 s at
2
3
3
disordered. This disorder was modeled by using a combination of
the restraints SADI, DFIX, and ISOR. Crystallographic data for
37 °C. The B values were calculated based on the calibrated 90°
1
pulse on a linear amplifier. To correct for baseline variations, a linear
compounds 2′-X (X = NO , F, Me) and the details of data collection
baseline was applied for two CEST regimes. For 2-NO , the data
2
2
are listed in Table S1.
points at 130 and 72 ppm and at 72 and 30 ppm were employed for
the two regimes. For 2-F, the data points at 130 and 68 ppm and at 68
and 37 ppm were used for the two regimes. For 2-Me, the data points
at 130 and 68 ppm and at 68 and 37 ppm were used for the two
regimes in the pH range 6.01−7.20, whereas the data points at 102
and 68 ppm and at 68 and 37 ppm were employed in the pH range
7.38−7.80 owing to a poor baseline in the downfield region of the
spectra. Reported values of %CEST [(1 − M /M ) × 100%] are the
NMR Spectroscopy. H and 19F NMR spectra for the ligands
1
X
H( L′) (X = NO , F, Me) and organic precursors were collected at 25
2
°
C at 500 and 470 MHz frequencies, respectively, on Agilent DD2
00 MHz (11.7 T) or Varian Inova 500 MHz (11.7 T) spectrometers,
5
or on an automated Agilent DD MR 400 MHz (9.4 T) spectrometer
1
3
1
at 400 and 376 MHz frequencies, respectively. C{ H} NMR spectra
X
for H( L′) were collected at 126 MHz frequency using a Bruker
z
0
C
DOI: 10.1021/acs.inorgchem.8b01896
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
−
2+
−
X
−
2+
Figure 1. Reaction of L , Co , and etidronate to form [LCo (etidronate)] , as observed in 1 (left). Reaction of ( L′) , Co , and etidronate to
2
X
−
form [( L′)Co (etidronate)] , as observed in 2-X (X = NO , F, Me) (right).
2
2
differences in %H O signal reduction between applied on-resonance
presaturations (raw data) and the values obtained by inserting the
corresponding frequencies into the linear baseline equations. To
reflectance measurements. Solution spectra were collected for 97−370
2
X
μM samples of ligands H( L′) (X = NO , F, Me) in MeOH, and for
2
22−87 μM samples of compounds 2-X (X = NO , F, Me) in aqueous
2
calculate k , the CEST intensities at the frequency offsets
buffer solutions containing 50 mM HEPES and 100 mM NaCl at
three different pH values, covering the range used for CEST
experiments. Diffuse reflectance spectra were collected on crystalline
ex
corresponding to maximum H O signal reductions at 21 μT power
2
level were monitored for each pH value. The pH calibration curves
were generated by taking the logarithm with base 10 (log ) of the
1
0
samples of 2′-X (X = NO , F, Me). Samples for measurements were
2
ratios of two CEST signal intensities after a baseline correction was
applied.
Solution Magnetic Measurements. The solution magnetic
prepared by mixing polycrystalline samples of the compounds with
BaSO powder for a 2-fold dilution to give smooth, homogeneous
4
powders. The data were treated with a background correction of
moments of compounds 2-X (X = NO , F, Me) were determined
2
BaSO and the spectra are reported as normalized Kubelka−Munk
4
20
1
using the Evans method by collecting variable-pH H NMR spectra
at 37 °C (310 K) on an Agilent DD2 500 MHz (11.7 T)
spectrometer. In a typical experiment, the compound (3−5 mM)
was dissolved in a mixture of 2% (v/v) tert-butanol in an aqueous
solution containing 50 mM HEPES and 100 mM NaCl buffered at a
specific pH value. The resulting solution was placed in an NMR tube
containing a sealed capillary with the same solvent mixture but
without the to-be-characterized paramagnetic compound as a
reference solution. Diamagnetic corrections were carried out based
on the empirical formula of each compound (as determined by
transformation F(R) of the raw diffuse reflectance spectra, where F(R)
for each compound was normalized with the strongest absorbance set
to F(R) = 1.
Electrochemical Measurements. Cyclic voltammetry measure-
ments were carried out in a standard one-compartment cell under a
dinitrogen atmosphere using CH Instruments 760c potentiostat. The
cell consisted of a platinum electrode as a working electrode, a
platinum wire as a counter electrode, and a saturated calomel
electrode (SCE) as a reference electrode. Analytes were measured in
aqueous solutions with 100 mM NaCl and 50 mM HEPES buffered at
pH 7.3. All potentials were converted and referenced to the normal
21
elemental analysis) using Pascal’s constants. The paramagnetic
para
M
3
−1
molar susceptibility χ
following equation:
(cm mol ) was calculated using the
22
hydrogen electrode (NHE) using a literature conversion factor.
20
Other Physical Measurements. Electrode-based pH measure-
ments were carried out using a Thermo Scientific Orion 9110DJWP
double junction pH electrode connected to a VWR sympHony B10P
pH meter. The pH meter was calibrated using standardized pH buffer
solutions at 4.01, 7.00, and 10.00 purchased from LaMotte Company.
Elemental analyses of all compounds were conducted by Midwest
3
4
ΔvMw
para
− χ Md ia
χ
=
M
πv m
(2)
0
In this equation, Δν is the frequency difference (Hz) between the tert-
butyl resonance of tert-butanol in the sample and reference solutions,
−
1
M is the molecular mass of the paramagnetic compound (g mol ),
Microlab Inc. Infrared spectra were recorded for solid samples of
w
X
ν is the operating frequency of the NMR spectrometer (Hz), m is the
ligands H( L′) (X = NO
2
, F, Me) and Co
2
complexes 2-X (X = NO
,
2
0
−
3
dia
concentration of the paramagnetic compound (g cm ), and χ
is
F, Me) on a Bruker Alpha FTIR spectrometer equipped with an
attenuated total reflectance accessory. These data are provided in
Figures S1 and S2. Electrospray ionization mass spectrometry (ESI-
MS) measurements were performed on a Bruker AmaZon SL
quadrupole ion trap instrument. All measurements were carried out in
MeOH carrier solvent using positive and/or negative ionization
M
3
−1
the diamagnetic contribution to the molar susceptibility (cm mol ).
UV−Visible Absorption Spectroscopy. Solution and solid-state
UV−visible spectra were collected at ambient temperature in the
2
00−800 nm range on an Agilent Cary 5000 UV−visible−NIR
spectrophotometer equipped with an integrating sphere for diffuse
D
DOI: 10.1021/acs.inorgchem.8b01896
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
X
−
Figure 2. Crystal structures of the anionic complexes [( L′)Co (etidronate)] , as observed in 2′-X (X = NO , F, Me). Purple, magenta, green, red,
2
2
blue, and gray spheres represent Co, P, F, O, N, and C atoms, respectively; H atoms are omitted for clarity.
Table 1. Selected Mean Interatomic Distances (Å) and Angles (deg) in 1′ and 2′-X (X = NO , F, Me) at 100 K
2
d
1
′
2′-NO2
2′-F
2′-Me
Co−Ophenoxo
Co−Oamide
Co−Ophosphonate
Co−N
Co···Co
Co−Ophenoxo−Co
O−P−O
2.0920(2)
2.1127(1)
2.0618(1)
2.1558(2)
3.6740(3)
122.837(4)
113.882(3)
111.486(6)
170.324(1)
62.27(2)
2.0655(2)
2.1180(1)
2.0500(1)
2.1674(1)
3.6664(3)
125.133(3)
117.080(3)
108.147(5)
170.153(1)
54.09(2)
2.0373(3)
2.1316(2)
2.0835(2)
2.1543(2)
3.6265(5)
125.754(7)
116.823(6)
109.86(1)
169.479(1)
65.68(3)
2.0240(2)
2.1379(1)
2.0735(1)
2.1538(2)
3.5806(3)
124.394(3)
116.531(3)
109.144(2)
169.031(1)
60.41(2)
P−C−P
a
trans-O−Co−E
b
∑sum
∑
ω
5.19(1)
49.120(4)
4.51(1)
54.641(4)
5.47(1)
53.491(4)
5.03(1)
51.177(3)
mean
c
a
b
E denotes either a N or an O atom from the [CoNO ] coordination sphere. Octahedral distortion parameter (∑) = absolute deviation from 90°
5
c
−
X
−
d
of each 12 cis angle in [CoNO ]. Dihedral angle between the Co−O
obtained from reference 7d.
−Co plane and the plane of the phenolate ring of L or ( L′) . Values
5
phenoxo
mode. Inductively coupled plasma optical emission spectroscopy
ICP-OES) was performed on a Thermo iCAP 7600 dual view ICP-
OES instrument equipped with a CETAC ASX520 240-position
autosampler. Samples were dissolved in a 3% aqueous nitric acid
solution and the emissions for Co and P compared to standard
solutions.
substituent X on the bridging phenoxo moiety is varied from a
strongly electron-withdrawing NO2 group to an electron-
donating Me group. These dinucleating ligands were accessed
(
through S 2 reactions between 2,2′-(azanediyl)bis(N-(2,2,2-
N
trifluoroethyl)acetamide) and para-substituted 2,6-bis-
(
bromomethyl)phenol derivatives similarly to HL (see
experimental details and Schemes S1 and S2 in the Supporting
RESULTS AND DISCUSSION
■
Information).
Design and Syntheses. The NO -substituted tetra-
carboxamide) dinucleating ligand HL and the corresponding
X
2
Reaction of H( L′) with two equivalents of Co(NO ) ·
3
2
(
6
H O, one equivalent of etidronic acid, and five equivalents of
II
2
Co complex Na[LCo (etidronate)]·0.2NaNO ·2.7H O (1)
2
2
3
2
Na(OMe) in MeOH afforded Co complexes analogous to that
2
featuring ancillary etidronate ligand were prepared as
NO
2
7d
in 1. Namely, Na[( L′)Co
2
(etidronate)]·1.0NaNO
3
·
previously described (see Figure 1, left). With the goal of
II
2
1.5MeOH (2-NO
2
) was isolated as a light orange solid,
achieving Co complexes that exhibit more intense CEST
F
Na[( L′)Co (etidronate)]·2.2NaNO ·1.0H O (2-F) as a dark
2
3
2
effects for carboxamide protons in the pH range 6.0−6.5, we
Me
pink solid, and Na[( L′)Co (etidronate)]·4.0NaNO ·3.4H O
2
3
2
synthesized CF -functionalized analogues of HL. Here,
3
(
2-Me) as a pink solid (see Figure 1, right). Note that all four
replacement of the primary carboxamide groups (−CONH )
with secondary amides bearing CF3 substituents
2
Co complexes feature the ancillary bisphosphonate etidronate.
2
The central OH group on this ligand has been shown to give
(
−CONHCH CF ) serves to lower the pK of the amide
2
3
a
7
d
rise to highly shifted and pH-sensitive CEST peaks,
protons by virtue of the electron-withdrawing character of the
rendering it suitable for incorporation into PARACEST pH
sensors.
CF groups. Such an increase in acidity of the labile protons
engenders faster proton exchange at more acidic pH values,
and thus provides greater CEST peak intensities.
3
8
a−e,h,13,23
Crystal Structures. Single crystals of 2′-X (X = NO , F,
2
Me) suitable for X-ray diffraction analysis were obtained by
Furthermore, to more thoroughly investigate the electronic
effects of ligand substitution on the CEST properties of
transition metal-based PARACEST probes, three derivatives of
slow diffusion of Et O vapor into a concentrated solution of 2-
2
X (X = NO , F, Me) in MeOH. Specifically, light orange plate-
2
X
NO
2
the CF -functionalized ligand H( L′) (X = NO , F, Me) were
shaped crystals of Na[( L′)Co (etidronate)]·1.0NaNO (2′-
3
2
2
3
F
targeted. For this ligand series, the identity of the para-
NO ), pink prism-shaped crystals of Na[( L′)-
2
E
DOI: 10.1021/acs.inorgchem.8b01896
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Co (etidronate)]·1.0NaNO ·1.0MeOH (2′-F), and pink plate-
spectra for 2-F and 2-Me in pH 6.95 buffer solution exhibit
2
3
Me
−1
−1
shaped crystals of Na[( L′)Co (etidronate)]·solvent (2′-Me)
weaker absorptions centered at 309 nm (ε = 4400 M cm )
2
−
1
−1
were obtained for the three compounds. Analysis at 100 K
revealed that 2′-NO2 and 2′-Me crystallized in the
orthorhombic space groups P2 2 2 and Pccn, respectively,
and 307 nm (ε = 3900 M cm ), respectively (see Figures S4
and S5). We assign these absorption bands to ligand−metal
charge transfer (LMCT) transitions from the bridging
1
1 1
II
whereas 2′-F crystallized in the triclinic space group P1
̅
(see
phenolate to Co , in accord with similar phenoxo-bridged
25c,26
Table S1). The different space groups across the 2′-X series
most likely result from different cocrystallization of salts and
solvent molecules in these structures.
Co complexes.
The red shift of the LMCT band and
2
dramatic increase in molar absorptivity in moving from 2-Me
to 2-NO (see Figure S6) are in good agreement with the
2
Similar to that in Na[LCo (etidronate)]·6.8H O (1′), the
influence of the para-substituent on the extent of π-
2
2
general structure of the anionic complexes in 2′-X consists of
two nearly identical Co centers in distorted octahedral
coordination environments. Each coordination sphere is
made up of two amide O atoms, a μ-phenoxo O atom, and a
conjugation in the phenolate ligands. Note that the same
X
trend is observed for MeOH solutions of the ligands H( L′)
(see Figure S7). The intensities of the absorption bands for 2-
X increase slightly when the pH is raised from 5.95 to 7.96 (see
Figures S3−S5), suggesting a minimal increase in ligand donor
X
−
N atom from ( L′) , together with two O atoms from the
2
4
28
μ −κ etidronate ligand (see Figure 2). The mean dihedral
angle between the Co−O −Co trigonal plane and the
strength with increased pH. Finally, the close similarities
between the diffuse reflectance spectra collected for crystalline
samples of 2′-X (see Figures S8−S10) and the spectra
obtained for aqueous solutions of 2-X indicate that the
structures of the Co2 complexes determined from X-ray
diffraction are preserved in HEPES buffer solutions in the pH
range 6.0−8.0.
phenoxo
X
−
hexagonal plane of the phenolate ring of ( L′) ranges from
5
1.177(3)° for 2′-Me to 54.641(4)° for 2′-NO . These angles
2
are slightly larger than observed for 1′, in accord with the
bulkier amide substituents in 2′-X (see Table 1).
The mean Co−O distances of 2.0655(2), 2.1180(1), and
2
.0500(1) Å for the Co−O
, Co−O
, and Co−
Solution Magnetic Properties. To further probe the
phenoxo
amide
O
bonds in 2′-NO , respectively, are similar to those
solution electronic structures of 2-X (X = NO , F, Me) and
phosphonate
2
2
observed in 1′. For 2′-F and 2′-Me, the mean Co−O
assess their magnetic properties, variable-pH dc magnetic
phenoxo
, and Co−
distance is slightly shorter, and the Co−O
susceptibility data were collected at 37 °C for aqueous buffer
amide
2
0
O
distances are slightly longer, as compared to the
solutions using the Evans method (see Experimental
phosphonate
NO -substituted analogue. The decrease in Co−O
bond
Section). The plots of χ T versus pH, as depicted in Figures
2
phenoxo
M
length across the 2′-X series is consistent with higher negative
charge density on the μ-phenoxo O atom for ligands bearing
electron-donating para-substituents than electron-withdrawing
S11−S13, reveal that χ T varies insignificantly with pH in the
M
pH range 5.8−8.1. Over this pH range, average values of χ T =
M
3
−1
5.75(8), 5.6(1), and 5.5(1) cm K mol were obtained for 2-
2
4
substituents. The mean Co−N distance is similar for all four
NO , 2-F, and 2-Me, respectively (see Table S2). Assuming
2
II
Co2 complexes, ranging from 2.1538(2) Å in 2′-Me to
two magnetically noninteracting Co centers, these values
2
.1674(1) Å in 2′-NO . In contrast, the average intramolecular
correspond to g = 2.48(2), 2.44(6), and 2.42(3) for 2-NO , 2-
2
2
Co···Co distance decreases from 3.6664(3) Å in 2′-NO to
F, and 2-Me, respectively. These data are consistent with
2
3
.6265(5) Å in 2′-F and 3.5806(3) Å in 2′-Me. Note that this
decrease in intramolecular Co···Co distance across the 2′-X
distance. This
values reported for structurally similar high-spin Co
2
3
II
7d,9a,25
complexes featuring octahedral S = / Co centers.
2
series follows the same trend as the Co−O
Taken together, the solution magnetic measurements of 2-X
phenoxo
II
observation, in conjunction with the similar intramolecular
corroborate the high-spin assignment of the Co ions evident
Co···Co distance for 1′ and 2′-NO , implies that the electronic
from UV−visible spectroscopy and illustrate that the magnetic
2
properties of the bridging phenoxo ligand significantly affect
properties of this family of Co complexes do not change
2
the structural features of these Co complexes.
significantly in the physiological pH range.
2
1
9
The mean intramolecular Co−O
−Co angle does not
F NMR Spectroscopy. To further investigate and
phenoxo
vary significantly across the 2′-X series, but was found to be
compare the solution electronic structures and properties of
1
9
slightly larger for 2′-X than 1′. This difference in Co−
2-X (X = NO , F, Me), F NMR spectra were collected for
2
O
−Co angle most likely stems from the increased steric
aqueous solutions buffered at selected pH values between 6.0
phenoxo
X
−
−
bulk of the pendent amides in ( L′) compared to those in L .
Accordingly, the O−P−O and P−C−P bond angles increase
and decrease slightly, respectively, for 2′-X, compared to the
corresponding angles for 1′. Taken together, the interatomic
distances and angles for 2′-X, in conjunction with an average
octahedral distortion parameter (∑) ranging from 4.51(1) to
and 7.8. The spectrum for 2-NO at pH 5.97 features two sets
2
of two closely separated peaks positioned at −67.14 and
−68.42 ppm and at −73.13 and −74.05 ppm, respectively (see
Figure S14). These resonances have equal integration and are
assigned to the four inequivalent CF groups in 2-NO , in
3
2
accord with the pseudo-C symmetry of the complex. Upon
2
5
=
.47(1)° across the series, are consistent with two high-spin S
raising the pH to 7.77, the less upfield-shifted peaks shift
slightly upfield, while the frequencies of the more upfield-
shifted set reveal no change. The spectra for 2-F and 2-Me
exhibit analogous features and pH behavior as those for 2-NO2
(see Figures S15 and S16), but significant differences in
chemical shifts are observed (see Figure 3, right). Interestingly,
the chemical shifts of the more upfield-shifted sets of peaks are
nearly identical for all three compounds, while the resonance
frequencies of the less upfield-shifted peaks are drastically
different. These peaks shift upfield with increasing electron-
withdrawing character of X. Consequently, the separation
3
II
/ Co ions bridged by a deprotonated OH group from
2
X
7d,24−27
H( L′).
Notably, the solid-state structures of 2′-X are
influenced by a combination of electronic and steric factors.
UV−Visible Spectroscopy. To probe the electronic
structures of compounds 2-X (X = NO , F, Me) in aqueous
2
solution, UV−visible absorption spectra were collected for
samples in 50 mM HEPES buffers containing 100 mM NaCl.
The spectrum for 2-NO in pH 6.94 buffer shows an intense
2
−
1
−1
absorption band at 371 nm (ε = 13 800 M cm ) (see Figure
S3) in analogy to the spectral features of 1. In contrast, the
7d
F
DOI: 10.1021/acs.inorgchem.8b01896
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure S17, bottom). The etidronate Me and OH peaks for 2-
7d
NO are shifted downfield compared to those obtained for 1,
2
whereas the opposite trend is observed for the carboxamide
peaks. Interestingly, the four amide proton resonances for 2-
NO are well shifted from the H O resonance, indicating that
2
2
these protons adopt a trans configuration with respect to the
29
carbonyl O atom. The line widths of the NH peaks increase
markedly when the pH is raised from 5.97 to 7.77, whereas the
OH resonance shifts downfield by 5 ppm and becomes sharper
(
see Figure S18). These pH-dependent properties of the
exchangeable proton resonances for 2-NO are consistent with
2
those observed for 1 and indicate base- and acid-catalyzed
exchange of the NH and OH protons, respectively.
The compounds 2-F and 2-Me exhibit similar spectral
1
7d
Figure 3. Portions of H NMR spectra for aqueous solutions of 2-
NO (blue), 2-F (green), and 2-Me (red) buffered at pH 7.38, 7.40,
2
and 7.38, respectively, highlighting the chemical shift of the etidronate
profiles and pH dependences as does 2-NO in the pH range
2
Me group (left). 19F NMR spectra for aqueous solutions of 2-NO2
6
.0−7.8, albeit with some key differences in chemical shifts and
(
blue), 2-F (green), and 2-Me (red) buffered at pH 7.38, 7.40, and
.38, respectively (right). All spectra were collected for solutions in 50
mM HEPES buffers with 100 mM NaCl at 37 °C. The H NMR
chemical shifts are reported as frequency offsets with the H O
chemical shift set to 0 ppm. The chemical shift of TFA internal
standard was set to −76.00 ppm for F NMR data, as denoted by the
line widths (see Figures S19−S23). First, the resonances for 2-
F and 2-Me span a smaller chemical shift range than those for
7
1
2
-NO , from −75 to 175 ppm versus H O. Moreover, the
X −
2
2
2
protons on the phenolate ring of ( L′) are shifted downfield
by 7 and 9 ppm for 2-F and 2-Me, respectively. The opposite
trend in chemical shift changes across the series is observed for
the resonances from the ancillary etidronate. For solutions
buffered at pH 7.4, the etidronate Me peak shifts from 54 to 73
1
9
asterisk.
between the two sets of CF peaks is largest for 2-Me and
smallest for 2-NO . These experiments demonstrate that
NMR is a useful tool for probing the electronic structures of 2-
X in solutions. Finally, note that the additional peak at ca. −
9.3 ppm in the spectra for 2-F corresponds to the para-F
substituent and is not significantly affected by pH in the range
studied (see Figure S15).
H NMR Spectroscopy. To probe the effects of ligand
substitution on NMR properties in this family of Co
3
1
9
ppm in moving from X = Me to X = NO (see Figure 3, left).
F
2
2
Furthermore, the OH resonance shifts from 92 to 117 ppm
across the series (see Figure S23). The OH peak becomes
sharper with increasing pH for all complexes; however, the
4
peak for 2-F is significantly broader than for 2-NO and 2-Me.
2
This observation suggests a faster OH proton exchange for 2-F
1
than for the other two Co complexes between pH 6.0 and 7.8.
2
Lastly, the two most downfield-shifted NH resonances are
shifted upfield by ca. 10 ppm for 2-F and 2-Me, as compared
2
1
complexes, H NMR spectra were collected for buffer solutions
to those for 2-NO , whereas the less shifted sets of NH peaks
of 2-X (X = NO , F, Me) in the pH range 6.0−7.8. The
2
2
1
spectrum for 2-NO at pH 6.48 exhibits paramagnetically
are shifted slightly downfield. Overall, the H NMR chemical
2
shifted resonances with chemical shifts ranging from −105 to
shifts of the 2-X series are extremely sensitive to the electronic
properties of the X substituent. Remarkably, the resonance
frequencies of the protons on the distant etidronate are highly
affected, demonstrating that even modest changes in ligand
structure can lead to drastic changes in NMR properties.
CEST Properties. To investigate and compare the potential
1
85 ppm versus H O (see Figure S17, top). The sharp
2
resonances at 32 and 35 ppm are assigned to the two protons
NO
2
−
on the phenolate ring of ( L′) and the intense peak at 71
ppm corresponds to the etidronate Me group. Furthermore,
the resonances at 44, 48, 88, and 92 ppm correspond to two
sets of slightly inequivalent carboxamide NH protons, whereas
the etidronate OH group resonates at 115 ppm. The
assignment of the exchangeable proton resonances was verified
of 2-X (X = NO , F, Me) as pH-responsive PARACEST
2
probes, CEST spectra were collected for 8−9 mM solutions of
2-X in 50 mM HEPES buffers with 100 mM NaCl at pH 6.0−
by their disappearance in the spectrum recorded in D O (see
7.8 (see Figure 4). The spectrum for 2-NO at pH 5.97
2
2
Figure 4. Variable-pH CEST spectra collected at 37 °C for 8 mM of 2-NO (left), 9 mM of 2-F (center), and 8 mM of 2-Me (right) in aqueous
2
solutions containing 50 mM HEPES and 100 mM NaCl buffered at pH 6.0−7.8 (red to blue). Values of pH are given in the legend. Insets:
Expanded views of the CEST peaks of interest.
G
DOI: 10.1021/acs.inorgchem.8b01896
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
displays three main features, namely a peak at 112 ppm with
1
1% H O signal reduction, two overlapping peaks at 87 and 91
2
ppm, and a broad signal centered at 45 ppm. As the pH is
raised to 7.77, the peak at 112 ppm shifts to 117 ppm, and its
intensity decreases to 3.0%. In stark contrast, the positions of
the other CEST peaks are relatively insensitive to pH changes
and the intensities of these peaks greatly increase to 15−17%
(
see Figure 4, left). These data are consistent with variable-pH
1
H NMR experiments, and the three CEST features
correspond to the etidronate OH group and two sets of
overlapping NH protons, where the further downfield-shifted
set can be resolved into two peaks. Note that the OH and NH
CEST peaks for 2-NO are well separated from one another, in
2
contrast to the overlapping peaks observed at ca. 104 ppm in
the spectra for 1. Furthermore, the NH CEST effects for 2-
NO are significant at pH 5.97, whereas they are barely
2
7d
detectable below pH 6.50 for 1. Therefore, incorporation of
electron-withdrawing CF groups serves to (1) shift the OH
3
peak downfield and (2) increase the NH CEST intensities at
lower pH values, and as such has dramatic effects on the CEST
properties of Co complexes.
2
The CEST spectra for 2-F and 2-Me reveal analogous
features and pH-dependent changes in signal intensities as
observed for 2-NO (see Figure 4, center and right). The NH
2
peaks for 2-F are shifted relative to those for 2-NO , to 49, 79,
2
and 81 ppm, and the peaks for 2-Me are further shifted to 51,
1
7
6, and 79 ppm. These data are in accord with H NMR
analysis. Notably, the separation between the OH and NH
CEST peaks for 2-X decreases substantially when X is varied
from NO to Me. Additionally, while the OH peak for 2-NO2
Figure 5. Scheme depicting the mechanism for pH-induced changes
2
in NMR frequencies of etidronate resonances for 2-X (X = NO , F,
shifts 5 ppm downfield between pH 6.0 and 7.8, this peak shifts
downfield by 7 ppm for the other two complexes in this pH
range, from 90 to 97 ppm and 85 to 92 ppm for 2-F and 2-Me,
respectively. The higher susceptibility of the OH CEST
frequencies for 2-F and 2-Me to pH variations than observed
for 2-NO likely arises from the higher pK values of one of the
2
Me). The ionization process taking place on etidronate is highlighted
with the light red circles.
pH, in particular one of the cobalt-coordinated O
atom. This hypothesis was supported by (1) dramatic
chemical shift changes for the bisphosphonate resonances that
phosphonate
7d
2
a
coordinated Ophosphonate atom in these compounds (see below).
The pH dependences of the CEST effects for 2-X are
summarized in Figures S24−S29. The CEST intensities of the
OH peak in the middle of the frequency range and of one peak
for each NH CEST feature were monitored for each Co2
followed sigmoidal pH profiles, (2) pK values of the free
a
30
bisphosphonic acids, and (3) previous reports of protonated
31
Ophosphonate atoms in transition metal complexes. Indeed, the
chemical shifts of the etidronate Me group for 2-X exhibit
pronounced pH dependences. The Me resonance for 2-NO2
complex. The OH CEST effects for 2-NO , 2-F, and 2-Me at
2
1
15, 93, and 88 ppm, respectively, show very similar pH
shifts from 52.87 to 72.62 ppm versus H O between pH 2.30
2
dependences in the pH range 6.0−7.8 (see Figure S30).
Likewise, the intensities of the NH CEST peaks at 88, 79, and
and 7.77, and exhibits a sigmoidal pH profile. Similarly, the Me
resonances for 2-F and 2-Me shift by 17.56 and 18.76 ppm,
respectively, in the pH ranges 3.01−7.80 and 2.72−7.80,
respectively. Fits of the Me chemical shifts as a function of pH
7
6 ppm for 2-NO , 2-F, and 2-Me, respectively, increase
2
analogously with increasing pH (see Figure S31). Furthermore,
the maximum OH CEST effects for the three compounds vary
similarly when the pH is raised from 6.0 to 7.4, revealing a near
linear intensity decrease. However, slight discrepancies are
observed above pH 7.4. While the intensity of the OH CEST
to eq 1 afforded pK = 4.76(7), 5.41(6), and 5.38(6) for 2-
a
NO , 2-F, and 2-Me, respectively (see Figures S33−S35). The
2
pK value obtained for 2-NO is considerably lower than the
a
2
7d
value of 5.01(3) obtained for 1, as expected given the
peak for 2-Me keeps decreasing, the intensities for 2-NO and
electron-withdrawing nature of the CF substituents. Fur-
2
3
2
-F reach a plateau (see Figure S32). In sum, the frequencies
thermore, the pK for 2-NO is lower than the values for 2-F
and 2-Me by ca. 0.6. This observation agrees well with the
a
2
of the OH and NH CEST peaks for 2-X are greatly affected by
the nature of the para-substituent X, whereas the CEST
smaller change in OH CEST frequency for 2-NO in the pH
2
intensities and their pH profiles are significantly less affected.
range 6.0−7.8. To illustrate, the etidronate ligands in 2-F and
2-Me have greater contributions from both protonation states
in this pH range, which renders the etidronate resonances
more sensitive to small variations in pH. Finally, note that
1
pK Determination by H NMR Spectroscopy. The
a
variations in the frequencies giving rise to maximum OH
CEST effects for 2-X (X = NO , F, Me) in the pH range 6.0−
2
1
7
.8 suggest that these Co complexes undergo modest pH-
changes in the H NMR frequency of the etidronate OH peak
2
induced structural changes in solution (see Figure 5). We
previously hypothesized that these CEST peak shifts result
from protonation of the ancillary bisphosphonate ligand at low
cannot be employed directly to estimate the pK values for 2-X
owing to the broadness of this peak below pH 6.0 for all
compounds.
a
H
DOI: 10.1021/acs.inorgchem.8b01896
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 6. Plots of the ratios of OH and NH CEST intensities from presaturation at 115 and 88 ppm for 8 mM aqueous buffer solutions of 2-NO2
left), at 93 and 79 ppm for 9 mM solutions of 2-F (center), and at 88 and 76 ppm for 8 mM solutions of 2-Me (right) versus pH. Insets: Semilog
(
forms of the plots. Colored circles denote experimental data and the black lines correspond to linear fits of the data. The black numbers in the insets
represent the slopes of the linear fits of the data.
Exchange Rate Analysis. The proton exchange rates for
Me reveal substantial and very similar pH dependences (see
Figures 6 and S55). The ratiometric values decrease markedly
between pH 6.2 and 7.4, whereas near no change is observed at
higher pH. Importantly, linear pH calibration curves could be
2
-X (X = NO , F, Me) were estimated at 37 °C using the
2
19
Omega plot method to gain further insight into the pH-
dependent changes in NMR and CEST signal intensities. For
33
2
-NO , the rate constant (k ) for OH proton exchange
generated for 2-X by plotting the logarithm of the intensity
2
ex
3
2
−1
7d
decreases from 1.0(1) × 10 to 3.1(1) × 10 s in the pH
ratios versus pH, in analogy to those reported for 1. Linear
range 5.97−7.77 (see Figure S36 and Table S3). In contrast,
fits of the log(CESTOH/CESTNH) versus pH data for 2-X
afforded the following equations (see Figure 6, insets and
Figure S56):
2
k
for the NH protons increase from 3.3(1) × 10 (43−45
ex
2
2
ppm), 2.5(4) × 10 (88−89 ppm), and 2.2(3) × 10 (91−92
−
1
3
3
ppm) s to 1.7(2) × 10 (43−45 ppm), 1.5(2) × 10 (88−89
3
−1
NO : log(CEST115ppm/CEST88ppm) = −1.49 × pH + 10.0
ppm), and 1.6(2) × 10 (91−92 ppm) s over the pH range
2
6
.39−7.77 (see Figures S37−S39 and Table S3). The opposite
(3)
pH behavior for OH and NH proton exchange rates (see
1
Figure S40) is consistent with H NMR and CEST data and
F: log(CEST93ppm/CEST79ppm) = −1.48 × pH + 9.9
illustrates the acid- and base-catalyzed exchange of the OH and
(
4)
5)
7d,8a−f,h,13,23,32
NH protons, respectively.
The exchange rate constants for OH and NH protons in 2-F
and 2-Me vary similarly with pH between pH 6.0 and 7.8 as
Me: log(CEST88ppm/CEST76ppm) = −2.04 × pH + 13.7
(
those for 2-NO (see Figures S41−S50 and Tables S4 and S5).
2
Nevertheless, close comparison of the pH dependences of k
Remarkably, the slope of the calibration curve for 2-NO2
ex
across the 2-X series reveals some important differences.
represents 1.5-fold enhancement in pH sensitivity compared to
7d
Specifically, k for OH proton exchange in 2-F is slightly larger
that obtained for 1, owing to the complete separation of OH
ex
than the corresponding rate constants for the other two Co2
complexes (see Figure S51). These data support the hypothesis
and NH CEST peaks and thus more contrasting pH-
dependent intensity changes. Furthermore, the CF -function-
3
1
that the broader OH resonances in the H NMR spectra for 2-
alized Co complex enables detection of pH down to 6.2,
2
F, compared to those for 2-NO and 2-Me, originate from a
compared to pH 6.5 for 1. However, the trade-off of separating
the two CEST peaks is a loss of signal intensity. Therefore, an
ideal ratiometric PARACEST pH probe should demonstrate a
compromise between pH sensitivity and CEST peak
intensities.
2
faster proton exchange. Furthermore, the NH proton exchange
rates for 2-X become considerably different at the basic end of
the pH range investigated, where increasing electron-with-
drawing character of X leads to faster exchange (see Figures
S52−S54). In addition, note that k for the OH protons reveal
Interestingly, the pH calibration curves obtained for 2-NO2
and 2-F in the pH range 6.2−7.4 are almost identical, while the
curve for 2-Me has a significantly steeper slope of −2.04(5) pH
ex
a slight increase above pH 7.4 for all compounds (see Figure
S51). This observation suggests that even though OH proton
exchange is primarily acid-catalyzed in the pH range 6.0−7.8,
catalysis by base might become important at more alkaline pH
values.
−
1
unit (see Figure S56). Note, however, that the linear range of
measuring pH using 2-Me is between pH 6.4 and 7.4. The
higher pH sensitivity observed for 2-Me than 2-NO may be
2
Ratiometric CEST Properties. To assess and compare the
attributed to the larger change in the chemical shift of the OH
peak for 2-Me in this pH range, which leads to more dramatic
changes in OH CEST intensity (see Figure S30). Along these
lines, the reason for the close similarities between the pH
ability of 2-X (X = NO , F, Me) to quantify pH in a ratiometric
2
manner, the pH dependences of the ratios of OH and NH
CEST intensities (CESTOH/CESTNH) were investigated. In
particular, the intensities of the OH peak midway between its
position at pH 6.0 and pH 7.8, and the second most downfield-
responsiveness of 2-NO and 2-F, despite the greater change in
2
OH chemical shift for the latter, remains unclear. However, it
is important to note that the slopes of the pH calibration
curves for 2-X are highly affected by the choice of OH CEST
frequencies because of the variations in the frequencies giving
rise to maximum OH CEST effects with pH.
shifted NH peak were employed for each Co complex to
2
provide significant CEST intensities across the whole pH
range. The ratios of CEST intensities at 115 and 88 ppm for 2-
NO , at 93 and 79 ppm for 2-F, and at 88 and 76 ppm for 2-
2
I
DOI: 10.1021/acs.inorgchem.8b01896
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Concentration Effects. To investigate the effects of probe
concentration on the pH calibration curves, we collected
variable-pH CEST spectra for 4 mM aqueous buffer solutions
acidic range, from pH 6.5−7.6 to pH 6.2−7.4. Furthermore,
CEST frequencies were found to be highly affected by the
ligand electronic properties in a series of CF -functionalized
3
II
2
of 2-X (X = NO , F, Me) in analogy to the 8−9 mM samples.
Co complexes 2-X (X = NO , F, Me). Across the series, the
2
2
All compounds exhibit similar pH-dependent changes in CEST
frequencies and intensities for the OH and NH peaks as
observed for the more concentrated samples (see Figures S57−
OH peak shifts downfield, and the separation between the two
sets of NH peaks increases with increasing electron-with-
drawing character of X. In contrast, the pH-dependent changes
in CEST intensities were found to be significantly less affected
by the identity of X.
S80 and Tables S6−S8). Accordingly, the plots of CEST
/
OH
CEST
versus pH and the corresponding pH calibration
NH
II
2
curves for each Co complex are nearly identical for the
While the CF -functionalized Co PARACEST probes are
2
3
different concentrations (see Figures S81−S91). Specifically,
attractive candidates for ratiometric pH quantitation due to
their high pH sensitivities and stabilities in aqueous solutions,
the variations in the frequencies providing maximum OH
CEST intensities with pH are not ideal for intensity-based
CEST probes. To address this issue, we will seek to design
related probes with a compromise between pH sensitivity and
CEST intensities by using the lessons learned from this study
and taking advantage of the chemical tunability of the
phenoxo-bridged dinuclear platform. In particular, work is
underway to modify the bisphosphonate ligand through
incorporation of different substituents and CEST-active
functional groups in an effort to minimize the pH dependence
of CEST peak frequencies.
linear fits of the log(CEST /CEST ) versus pH data for 4
OH
NH
mM of 2-X provided the following pH calibration equations
(
see Figures S81, S84, S87, and S91):
NO : log(CEST115ppm/CEST88ppm) = −1.51 × pH + 10.2
2
(
(
(
6)
F: log(CEST93ppm/CEST79ppm) = −1.48 × pH + 9.9
Me: log(CEST88ppm/CEST76ppm) = −2.0 × pH + 13.7
7)
8)
These experiments reveal that the pH calibration curves for
-X are not significantly affected by the concentration of the
2
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.8b01896.
■
Co complex in this concentration range, which therefore
S
2
demonstrates that these PARACEST probes provide a
concentration-independent measure of pH in the pH range
.2−7.4 for 2-NO2 and 2-F, and 6.4−7.4 for 2-Me.
Additionally, note that 2-F affords a linear pH calibration
6
Additional experimental details and characterization data
curve over the pH range 6.2−7.6 (see Figure S92) and thus has
for 2-X (X = NO , F, Me), including crystallographic
2
the largest pH detection window of the three Co probes.
2
data for 2′-X (X = NO , F, Me) (PDF)
Stability Studies. Finally, we sought to examine the
2
stability of the Co complexes in aqueous solutions. Cyclic
voltammetry measurements were carried out for solutions of 2-
2
Accession Codes
CCDC 1849034−1849036 contain the supplementary crys-
tallographic data for this paper. These data can be obtained
free of charge via www.ccdc.cam.ac.uk/data_request/cif, by
emailing data_request@ccdc.cam.ac.uk, or by contacting The
Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK; fax: +44 1223 336033.
X (X = NO , F, Me) in HEPES buffer at pH 7.3. The cyclic
voltammograms of 2-Me and 2-F each exhibit an irreversible
2
oxidation process centered at ca. 1150 mV versus NHE (see
II
III
Figure S93), which we assign to the metal-based Co /Co
oxidation. As a comparison, the voltammogram of 2-NO2
reveals a less obvious oxidation process at the onset of the
potential window of the solvent (see Figure S93). These
AUTHOR INFORMATION
Corresponding Author
E-mail: dharris@northwestern.edu.
■
observations indicate that the electron-withdrawing CF
3
II
III
substituents in 2-X serve to anodically shift the Co /Co
oxidation wave by ca. 600 mV compared to that observed for
*
7
d
ORCID
1.
Most importantly, these studies demonstrate that 2-X are
34
T. David Harris: 0000-0003-4144-900X
inert toward reaction with O in aqueous solutions. Indeed,
2
these Co2 complexes are extremely robust in aqueous
solutions, as evident from their identical NMR and CEST
properties after weeks in HEPES buffers at pH 6.0−7.8.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
SUMMARY AND OUTLOOK
■
This research was funded by the Air Force Research
Laboratory under Agreement FA8650-15-2-5518 and North-
western University. T.D.H. thanks the Alfred P. Sloan
Foundation, and A.E.T. gratefully acknowledges the Leifur
Eiriksson Foundation for funding support. X-ray crystallog-
raphy and NMR spectroscopy were performed at the
Integrated Molecular Structure Education and Research
Center (IMSERC) at Northwestern University, which has
received support from the Soft and Hybrid Nanotechnology
Experimental (SHyNE) Resource (NSF NNCI-1542205), the
State of Illinois, and International Institute for Nano-
technology (IIN). Quantitative analysis of Co and P was
The foregoing results demonstrate that modest variations in
II
the electronic structure of Co centers through remote ligand
substitution can lead to significant changes in the CEST
II
properties of Co 2 complexes and can be employed for
optimizing their pH sensing performance. Specifically,
incorporation of CF -substituted amides into a ratiometric
3
II
2
Co PARACEST pH probe afforded a 1.5-fold enhancement
in pH sensitivity owing to the complete separation of OH and
NH CEST peaks that exhibit opposing pH-dependent intensity
changes. In addition, the introduction of electron-withdrawing
CF groups led to a shift of the detection window to a more
3
J
DOI: 10.1021/acs.inorgchem.8b01896
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
performed at the Northwestern University Quantitative Bio-
element Imaging Center (QBIC). We thank Prof. T. J. Meade
for use of his HPLC system and lyophilizer, Prof. E. A. Weiss
for use of her UV−visible−NIR spectrophotometer, Ms. N. S.
Pappas for experimental assistance, and Mr. K. Du and Dr. D.
Z. Zee for helpful discussions.
Chemical Exchange Saturation Transfer Magnetic Resonance Imaging
Contrast Agents for Molecular Imaging Studies. Acc. Chem. Res. 2009,
4
2, 915−924. (f) Viswanathan, S.; Kovacs, Z.; Green, K. N.; Ratnakar,
S. J.; Sherry, A. D. Alternatives to Gadolinium-Based Metal Chelates
for Magnetic Resonance Imaging. Chem. Rev. 2010, 110, 2960−3018.
(g) Soesbe, T. C.; Wu, Y.; Sherry, A. D. Advantages of Paramagnetic
CEST Complexes Having Slow-to-Intermediate Water Exchange
Properties as Responsive MRI Agents. NMR Biomed. 2013, 26, 829−
838.
REFERENCES
■
(
6) Srivastava, K.; Ferrauto, G.; Young, V. G., Jr.; Aime, S.; Pierre, V.
(
1) (a) Tannock, I. F.; Rotin, D. Acid pH in Tumors and Its
Potential for Therapeutic Exploitation. Cancer Res. 1989, 49, 4373−
384. (b) Vaupel, P.; Kallinowski, F.; Okunieff, P. Blood Flow,
1
9
C. Eight-Coordinate, Stable Fe(II) Complex as a Dual F and CEST
Contrast Agent for Ratiometric pH Imaging. Inorg. Chem. 2017, 56,
4
1
(
2206−12213.
7) (a) Jeon, I.-R.; Park, J. G.; Haney, C. R.; Harris, T. D. Spin
Crossover Iron(II) Complexes as PARACEST MRI Thermometers.
Oxygen and Nutrient Supply, and Metabolic Microenvironment of
Human Tumors: A Review. Cancer Res. 1989, 49, 6449−6456.
(
c) Gillies, R. J.; Raghunand, N.; Karczmar, G. S.; Bhujwalla, Z. M.
II
Chem. Sci. 2014, 5, 2461−2465. (b) Du, K.; Harris, T. D. A Cu
Paramagnetic Chemical Exchange Saturation Transfer Contrast Agent
MRI of the Tumor Microenvironment. J. Magn. Reson. Imaging 2002,
6, 430−450. (d) Gatenby, R. A.; Gillies, R. J. Why do Cancers Have
High Aerobic Glycolysis? Nat. Rev. Cancer 2004, 4, 891−899.
e) Gillies, R. J.; Raghunand, N.; Garcia-Martin, M. L.; Gatenby, R. A.
2
1
Enabled by Magnetic Exchange Coupling. J. Am. Chem. Soc. 2016,
1
38, 7804−7807. (c) Du, K.; Waters, E. A.; Harris, T. D. Ratiometric
(
pH Imaging: A Review of pH Measurement Methods and
Quantitation of Redox Status with a Molecular Fe Magnetic
2
Applications in Cancers. IEEE Eng. Med. Biol. Mag. 2004, 23, 57−
Resonance Probe. Chem. Sci. 2017, 8, 4424−4430. (d) Thorarinsdot-
6
4. (f) Hashim, A. I.; Zhang, X.; Wojtkowiak, J. W.; Martinez, G. V.;
tir, A. E.; Du, K.; Collins, J. H. P.; Harris, T. D. Ratiometric pH
II
Gillies, R. J. Imaging pH and Metastasis. NMR Biomed. 2011, 24,
Imaging with a Co
2
MRI Probe via CEST Effects of Opposing pH
5
82−591. (g) Kato, Y.; Ozawa, S.; Miyamoto, C.; Maehata, Y.;
Dependences. J. Am. Chem. Soc. 2017, 139, 15836−15847.
Suzuki, A.; Maeda, T.; Baba, Y. Acidic Extracellular Microenviron-
ment and Cancer. Cancer Cell Int. 2013, 13, 89−96.
(
Extracellular pH in the Brain During Ischemia: Relationship to the
Severity of Lactic Acidosis. J. Cereb. Blood Flow Metab. 1991, 11, 597−
(8) (a) Dorazio, S. J.; Tsitovich, P. B.; Siters, K. E.; Spernyak, J. A.;
Morrow, J. R. Iron(II) PARACEST MRI Contrast Agents. J. Am.
Chem. Soc. 2011, 133, 14154−14156. (b) Tsitovich, P. B.; Morrow, J.
R. Macrocyclic Ligands for Fe(II) ParaCEST and Chemical Shift MRI
Contrast Agents. Inorg. Chim. Acta 2012, 393, 3−11. (c) Olatunde, A.
O.; Dorazio, S. J.; Spernyak, J. A.; Morrow, J. R. The NiCEST
Approach: Nickel(II) ParaCEST MRI Contrast Agents. J. Am. Chem.
Soc. 2012, 134, 18503−18505. (d) Dorazio, S. J.; Morrow, J. R.
Iron(II) Complexes Containing Octadentate Tetraazamacrocycles as
ParaCEST Magnetic Resonance Imaging Contrast Agents. Inorg.
Chem. 2012, 51, 7448−7450. (e) Dorazio, S. J.; Olatunde, A. O.;
Spernyak, J. A.; Morrow, J. R. CoCEST: Cobalt(II) Amide-Appended
ParaCEST MRI Contrast Agents. Chem. Commun. 2013, 49, 10025−
2) (a) Katsura, K.; Ekholm, A.; Asplund, B.; Siesjo, B. K.
̈
5
99. (b) Sun, P. Z.; Wang, E.; Cheung, J. S. Imaging Acute Ischemic
Tissue Acidosis with pH-Sensitive Endogenous Amide Proton
Transfer (APT) MRI−Correction of Tissue Relaxation and
Concomitant RF Irradiation Effects Toward Mapping Quantitative
Cerebral Tissue pH. NeuroImage 2012, 60, 1−6. (c) Rajamaki, K.;
̈
̈
Nordstrom, T.; Nurmi, K.; Åkerman, K. E. O.; Kovanen, P. T.; Oorni,
̈ ̈
K.; Eklund, K. K. Extracellular Acidosis Is a Novel Danger Signal
Alerting Innate Immunity via the NLRP3 Inflammasome. J. Biol.
Chem. 2013, 288, 13410−13419.
10027. (f) Tsitovich, P. B.; Spernyak, J. A.; Morrow, J. R. A Redox-
(
3) (a) Lauffer, R. B. Paramagnetic Metal Complexes as Water
Activated MRI Contrast Agent that Switches Between Paramagnetic
and Diamagnetic States. Angew. Chem., Int. Ed. 2013, 52, 13997−
Proton Relaxation Agents for NMR Imaging: Theory and Design.
Chem. Rev. 1987, 87, 901−927. (b) Caravan, P.; Ellison, J. J.;
McMurry, T. J.; Lauffer, R. B. Gadolinium(III) Chelates as MRI
Contrast Agents: Structure, Dynamics, and Applications. Chem. Rev.
1
4000. (g) Olatunde, A. O.; Cox, J. M.; Daddario, M. D.; Spernyak, J.
II
II
A.; Benedict, J. B.; Morrow, J. R. Seven-Coordinate Co , Fe and Six-
Coordinate Ni Amide-Appended Macrocyclic Complexes as
II
1
999, 99, 2293−2352. (c) Rieke, V.; Pauly, K. B. MR Thermometry.
J. Magn. Reson. Imaging 2008, 27, 376−390.
4) (a) Wike-Hooley, J. L.; Haveman, J.; Reinhold, H. S. The
ParaCEST Agents in Biological Media. Inorg. Chem. 2014, 53,
8
311−8321. (h) Dorazio, S. J.; Olatunde, A. O.; Tsitovich, P. B.;
(
Morrow, J. R. Comparison of Divalent Transition Metal Ion
Relevance of Tumour pH to the Treatment of Malignant Disease.
Radiother. Oncol. 1984, 2, 343−366. (b) Gerweck, L. E.; Seetharaman,
K. Cellular pH Gradient in Tumor versus Normal Tissue: Potential
Exploitation for the Treatment of Cancer. Cancer Res. 1996, 56,
ParaCEST MRI Contrast Agents. JBIC, J. Biol. Inorg. Chem. 2014,
1
9, 191−205. (i) Olatunde, A. O.; Bond, C. J.; Dorazio, S. J.; Cox, J.
M.; Benedict, J. B.; Daddario, M. D.; Spernyak, J. A.; Morrow, J. R.
II
II
II
Six, Seven or Eight Coordinate Fe , Co or Ni Complexes of Amide-
Appended Tetraazamacrocycles for ParaCEST Thermometry. Chem. -
Eur. J. 2015, 21, 18290−18300. (j) Tsitovich, P. B.; Cox, J. M.;
Spernyak, J. A.; Morrow, J. R. Gear Up for a pH Shift: A Responsive
Iron(II) 2-Amino-6-Picolyl-Appended Macrocyclic ParaCEST Agent
That Protonates at a Pendent Group. Inorg. Chem. 2016, 55, 12001−
1
194−1198. (c) Wojtkowiak, J. W.; Verduzco, D.; Schramm, K. J.;
Gillies, R. J. Drug Resistance and Cellular Adaptation to Tumor
Acidic pH Microenvironment. Mol. Pharmaceutics 2011, 8, 2032−
2
038. (d) Pilon-Thomas, S.; Kodumudi, K. N.; El-Kenawi, A. E.;
Russell, S.; Weber, A. M.; Luddy, K.; Damaghi, M.; Wojtkowiak, J.
W.; Mule, J. J.; Ibrahim-Hashim, A.; Gillies, R. J. Neutralization of
́
12010. (k) Burns, P. J.; Cox, J. M.; Morrow, J. R. Imidazole-Appended
Tumor Acidity Improves Antitumor Responses to Immunotherapy.
Macrocyclic Complexes of Fe(II), Co(II), and Ni(II) as ParaCEST
Agents. Inorg. Chem. 2017, 56, 4545−4554. (l) Abozeid, S. M.;
Snyder, E. M.; Tittiris, T. Y.; Steuerwald, C. M.; Nazarenko, A. Y.;
Morrow, J. R. Inner-Sphere and Outer-Sphere Water Interactions in
Co(II) ParaCEST Agents. Inorg. Chem. 2018, 57, 2085−2095.
(9) (a) James, T. L. Nuclear Magnetic Resonance in Biochemistry:
Principles and Applications; Academic Press: New York, 1975.
(b) Bertini, I.; Luchinat, C. NMR of Paramagnetic Molecules in
Biological Systems; The Benjamin/Cummings Publishing Company,
Inc.: Menlo Park, 1986. (c) Bertini, I.; Luchinat, C.; Parigi, G. Solution
NMR of Paramagnetic Molecules: Applications to Metallobiomolecules
and Models; Elsevier Science B.V.: Amsterdam, 2001.
Cancer Res. 2016, 76, 1381−1390.
(
5) Selected references: (a) Zhang, S.; Winter, P.; Wu, K.; Sherry, A.
D. A Novel Europium(III)-Based MRI Contrast Agent. J. Am. Chem.
Soc. 2001, 123, 1517−1518. (b) Zhang, S.; Merritt, M.; Woessner, D.
E.; Lenkinski, R. E.; Sherry, A. D. PARACEST Agents: Modulating
MRI Contrast via Water Proton Exchange. Acc. Chem. Res. 2003, 36,
7
83−790. (c) Zhou, J.; van Zijl, P. C. M. Chemical Exchange
Saturation Transfer Imaging and Spectroscopy. Prog. Nucl. Magn.
Reson. Spectrosc. 2006, 48, 109−136. (d) Woods, M.; Woessner, D. E.;
Sherry, A. D. Paramagnetic Lanthanide Complexes as PARACEST
Agents for Medical Imaging. Chem. Soc. Rev. 2006, 35, 500−511.
(
e) Ali, M. M.; Liu, G.; Shah, T.; Flask, C. A.; Pagel, M. D. Using Two
K
DOI: 10.1021/acs.inorgchem.8b01896
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(
10) Ward, K. M.; Aletras, A. H.; Balaban, R. S. A New Class of
Contrast Agents for MRI Applications. Magn. Reson. Med. 2002, 47,
639−648.
(24) Daumann, L. J.; Comba, P.; Larrabee, J. A.; Schenk, G.;
Stranger, R.; Cavigliasso, G.; Gahan, L. R. Synthesis, Magnetic
Properties, and Phosphoesterase Activity of Dinuclear Cobalt(II)
Complexes. Inorg. Chem. 2013, 52, 2029−2043.
(25) Selected references: (a) Sakiyama, H.; Ito, R.; Kumagai, H.;
Inoue, K.; Sakamoto, M.; Nishida, Y.; Yamasaki, M. Dinuclear
Cobalt(II) Complexes of an Acyclic Phenol-Based Dinucleating
Ligand with Four Methoxyethyl Chelating Arms − First Magnetic
Analyses in an Axially Distorted Octahedral Field. Eur. J. Inorg. Chem.
Contrast Agents for MRI Based on Proton Chemical Exchange
Dependent Saturation Transfer (CEST). J. Magn. Reson. 2000, 143,
7
(
9−87.
11) Charisiadis, P.; Kontogianni, V. G.; Tsiafoulis, C. G.; Tzakos, A.
1
G.; Siskos, M.; Gerothanassis, I. P. H-NMR as a Structural and
Analytical Tool of Intra- and Intermolecular Hydrogen Bonds of
Phenol-Containing Natural Products and Model Compounds.
Molecules 2014, 19, 13643−13682.
(
12) (a) Wolff, S. D.; Balaban, R. S. Magnetization Transfer
Contrast (MTC) and Tissue Water Proton Relaxation In Vivo. Magn.
Reson. Med. 1989, 10, 135−144. (b) Zaiss, M.; Bachert, P. Chemical
Exchange Saturation Transfer (CEST) and MR Z-Spectroscopy In
Vivo: A Review of Theoretical Approaches and Methods. Phys. Med.
Biol. 2013, 58, R221−R269.
2001, 2001, 2027−2032. (b) Hossain, M. J.; Yamasaki, M.; Mikuriya,
M.; Kuribayashi, A.; Sakiyama, H. Synthesis, Structure, and Magnetic
Properties of Dinuclear Cobalt(II) Complexes with a New Phenol-
Based Dinucleating Ligand with Four Hydroxyethyl Chelating Arms.
Inorg. Chem. 2002, 41, 4058−4062. (c) Tian, J.-L.; Gu, W.; Yan, S.-P.;
Liao, D.-Z.; Jiang, Z.-H. Synthesis, Structures and Magnetic Properties
of Two Binuclear Cobalt(II) Complexes Bridged by Bis(p-nitro-
phenyl) Phosphate or Diphenylphosphinate Ion. Z. Anorg. Allg. Chem.
(
13) (a) Aime, S.; Castelli, D. D.; Terreno, E. Novel pH-Reporter
MRI Contrast Agents. Angew. Chem., Int. Ed. 2002, 41, 4334−4336.
III
(
b) Castelli, D. D.; Terreno, E.; Aime, S. Yb -HPDO3A: A Dual pH-
and Temperature-Responsive CEST Agent. Angew. Chem., Int. Ed.
011, 50, 1798−1800. (c) Liu, G.; Li, Y.; Sheth, V. R.; Pagel, M. D.
2
(
008, 634, 1775−1779.
2
26) Johansson, F. B.; Bond, A. D.; Nielsen, U. G.; Moubaraki, B.;
Imaging In Vivo Extracellular pH with a Single Paramagnetic
Chemical Exchange Saturation Transfer Magnetic Resonance Imaging
Contrast Agent. Mol. Imaging 2012, 11, 47−57. (d) Sheth, V. R.; Liu,
G.; Li, Y.; Pagel, M. D. Improved pH Measurements with a Single
PARACEST MRI Contrast Agent. Contrast Media Mol. Imaging 2012,
Murray, K. S.; Berry, K. J.; Larrabee, J. A.; McKenzie, C. J. Dicobalt
II−II, II−III, and III−III Complexes as Spectroscopic Models for
Dicobalt Enzyme Active Sites. Inorg. Chem. 2008, 47, 5079−5092.
(
27) Seidler-Egdal, R. K.; Johansson, F. B.; Veltze, S.; Skou, E. M.;
́
II III
II
Bond, A. D.; McKenzie, C. J. Tunability of the M M /M and
M
2
7, 26−34. (e) Sheth, V. R.; Li, Y.; Chen, L. Q.; Howison, C. M.; Flask,
III
II III
/M M (M = Mn, Co) Couples in Bis-μ-O,O′-carboxylato-μ-
2
C. A.; Pagel, M. D. Measuring In Vivo Tumor pHe with CEST-FISP
MRI. Magn. Reson. Med. 2012, 67, 760−768. (f) Castelli, D. D.;
Ferrauto, G.; Cutrin, J. C.; Terreno, E.; Aime, S. In Vivo Maps of
Extracellular pH in Murine Melanoma by CEST−MRI. Magn. Reson.
Med. 2014, 71, 326−332. (g) Rancan, G.; Castelli, D. D.; Aime, S.
OR Bridged Complexes. Dalton Trans. 2011, 40, 3336−3345.
(
28) Nazeeruddin, M. K.; Zakeeruddin, S. M.; Kalyanasundaram, K.
Enhanced Intensities of the Ligand-to-Metal Charge-Transfer
Transitions in Ru(III) and Os(III) Complexes of Substituted
Bipyridines. J. Phys. Chem. 1993, 97, 9607−9612.
3
+
3+
MRI CEST at 1T with Large μ Ln Complexes Tm -HPDO3A:
eff
(29) (a) Wayland, B. B.; Drago, R. S.; Henneike, H. F. Amide
Contact Shift Studies and the Assignment of the Methyl Peaks in
N,N-Dimethylamides. J. Am. Chem. Soc. 1966, 88, 2455−2458.
(b) Stewart, W. E.; Siddall, T. H. Nuclear Magnetic Resonance
Studies of Amides. Chem. Rev. 1970, 70, 517−551. (c) Ming, L.-J.;
Lauffer, R. B.; Que, L., Jr. Proton Nuclear Magnetic Resonance
Studies of Iron(II/III)−Amide Complexes. Spectroscopic Models for
Non-Heme Iron Proteins. Inorg. Chem. 1990, 29, 3060−3064.
(30) (a) Raymond, G. G.; Born, J. L. An Updated pKa Listing of
Medicinal Compounds. Drug Intell. Clin. Pharm. 1986, 20, 683−686.
An Efficient MRI pH Reporter. Magn. Reson. Med. 2016, 75, 329−
3
36. (h) Krchova, T.; Galisova, A.; Jirak, D.; Hermann, P.; Kotek, J.
́ ́ ́ ́
Ln(III)-Complexes of a DOTA Analogue with an Ethylenediamine
Pendant Arm as pH-Responsive PARACEST Contrast Agents. Dalton
Trans. 2016, 45, 3486−3496.
(
14) APEX2, version 2014.11-0; Bruker Analytical X-ray Systems,
Inc.: Madison, WI, 2014.
15) Sheldrick, G. M. SADABS, version 2.03; Bruker Analytical X-ray
Systems, Inc.: Madison, WI, 2000.
16) (a) Sheldrick, G. M. SHELXTL, version 6.12; Bruker Analytical
(
(
(b) Vepsal
2002, 9, 1201−1208. (c) Alanne, A.-L.; Hyvo
Ylisirnio, M.; Turhanen, P.; Kolehmainen, E.; Pera
Vepsalainen, J. J. Systematic Study of the Physicochemical Properties
of a Homologous Series of Aminobisphosphonates. Molecules 2012,
7, 10928−10945.
31) (a) Hu, J.; Zhao, J.; Hou, H.; Fan, Y. Syntheses, Structures and
̈
ai
̈
nen, J. J. Bisphosphonate Prodrugs. Curr. Med. Chem.
nen, H.; Lahtinen, M.;
niemi, S.;
X-ray Systems, Inc.: Madison, WI, 2000. (b) Sheldrick, G. M.
SHELXT − Integrated Space-Group and Crystal-Structure Determi-
nation. Acta Crystallogr., Sect. A: Found. Adv. 2015, 71, 3−8.
̈
̈
̈
̈
̈
(
17) Dolomanov, O. V.; Bourhis, L. J.; Gildea, R. J.; Howard, J. A.
1
(
K.; Puschmann, H. OLEX2: A Complete Structure Solution,
Refinement and Analysis Program. J. Appl. Crystallogr. 2009, 42,
Fluorescence Studies of Two New Cadmium(II) Pyridyl-Diphosph-
onates. Inorg. Chem. Commun. 2008, 11, 1110−1112. (b) Guo, Z.-F.;
Li, B.; Guo, J.-Z.; Yang, P.; Shi, L.-F.; Liu, L. Syntheses, Structures,
Thermal Stabilities and Magnetic Properties of Two-Layered Metal
Diphosphonates. Transition Met. Chem. 2014, 39, 353−360.
3
(
2
(
39−341.
18) OriginPro, version 9.0; OriginLab, Corp.: Northampton, MA,
003.
19) Dixon, W. T.; Ren, J.; Lubag, A. J. M.; Ratnakar, J.; Vinogradov,
E.; Hancu, I.; Lenkinski, R. E.; Sherry, A. D. A Concentration-
Independent Method to Measure Exchange Rates in PARACEST
Agents. Magn. Reson. Med. 2010, 63, 625−632.
(
32) Huang, C.-H.; Morrow, J. R. Cerium(III), Europium(III), and
Ytterbium(III) Complexes with Alcohol Donor Groups as Chemical
Exchange Saturation Transfer Agents for MRI. Inorg. Chem. 2009, 48,
(
20) (a) Evans, D. F. The Determination of the Paramagnetic
7
(
237−7243.
Susceptibility of Substances in Solution by Nuclear Magnetic
Resonance. J. Chem. Soc. 1959, 2003−2005. (b) Schubert, E. M.
Utilizing the Evans Method with a Superconducting NMR
Spectrometer in the Undergraduate Laboratory. J. Chem. Educ.
33) Note that the common logarithm with base 10 was used
throughout this study. For brevity, the notation “log” is employed in
the pH calibration equations and in the insets of Figure 6 rather than
the precise notation “log ”.
1
0
1
(
992, 69, 62.
21) Bain, G. A.; Berry, J. F. Diamagnetic Corrections and Pascal’s
Constants. J. Chem. Educ. 2008, 85, 532−536.
22) Sawyer, D. T.; Sobkowiak, A.; Roberts, J. L., Jr. Electrochemistry
for Chemists, 2nd ed.; John Wiley & Sons: New York, 1995.
23) Aime, S.; Barge, A.; Castelli, D. D.; Fedeli, F.; Mortillaro, A.;
(34) Wood, P. M. The Potential Diagram For Oxygen at pH 7.
Biochem. J. 1988, 253, 287−289.
(
(
Nielsen, F. U.; Terreno, E. Paramagnetic Lanthanide(III) Complexes
as pH-Sensitive Chemical Exchange Saturation Transfer (CEST)
L
DOI: 10.1021/acs.inorgchem.8b01896
Inorg. Chem. XXXX, XXX, XXX−XXX