An overly anionic metal coordination environment eliminates the T1-weighted response of quinol-containing MRI contrast agent sensors to H2O2

Article abstract of DOI:10.1016/j.ica.2019.119045

A previously reported quinol- and Mn(II)-containing MRI contrast agent sensor for H₂O₂ has the drawback of releasing the metal ion upon oxidation of the organic ligand. The release of potentially neurotoxic Mn(II) limits the sensor's

Full text of DOI:10.1016/j.ica.2019.119045

Journal Pre-Proof
Research paper
An Overly Anionic Metal Coordination Environment Eliminates the T -Weigh-
1
ted Response of Quinol-Containing MRI Contrast Agent Sensors to H O
2
2
Tessa E. Hutchinson, Adam Bashir, Meng Yu, Ronald J. Beyers, Christian R.
Goldsmith
PII:
S0020-1693(19)30657-7
https://doi.org/10.1016/j.ica.2019.119045
ICA 119045
DOI:
Reference:
To appear in:
Inorganica Chimica Acta
Received Date:
Revised Date:
Accepted Date:
8 May 2019
29 July 2019
31 July 2019
Please cite this article as: T.E. Hutchinson, A. Bashir, M. Yu, R.J. Beyers, C.R. Goldsmith, An Overly Anionic
Metal Coordination Environment Eliminates the T -Weighted Response of Quinol-Containing MRI Contrast Agent
1
Sensors to H O , Inorganica Chimica Acta (2019), doi: https://doi.org/10.1016/j.ica.2019.119045
2
2
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will
undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing
this version to give early visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
©
2019 Published by Elsevier B.V.

JOURNAL PRE-PROOF
An Overly Anionic Metal Coordination Environment Eliminates the T -Weighted Response
1
of Quinol-Containing MRI Contrast Agent Sensors to H O
2
2
†
†
†
‡
Tessa E. Hutchinson, Adam Bashir, Meng Yu, Ronald J. Beyers, and Christian R.
†,
Goldsmith *
†
Department of Chemistry and Biochemistry, Auburn University, Auburn, AL 36849, USA
Auburn University Magnetic Resonance Imaging Research Center, Auburn, AL 36849, USA
‡
*
Corresponding author: crgoldsmith@auburn.edu
Abstract
A previously reported quinol- and Mn(II)-containing MRI contrast agent sensor for H O
2
2
has the drawback of releasing the metal ion upon oxidation of the organic ligand. The release of
potentially neurotoxic Mn(II) limits the sensor’s in vivo applicability. We prepared N,N’-bis(2,5-
dihydroxybenzyl)ethanediamine-N,N’-diacetic acid (H qc1) as a substitute ligand that could
6
potentially remain bound to the metal ion after oxidation of the quinol portions to more weakly
metal-binding para-quinones. The carboxylic acid groups deprotonate at ambient pH, providing
a more anionic coordination environment that stabilizes its Mn(II) complex in water. Although
the more anionic coordination sphere doesn’t introduce air sensitivity, it does render the metal
center more susceptible to oxidation by hydrogen peroxide, as evidenced by electron
paramagnetic resonance. The oxidation of the metal to less paramagnetic Mn(III) is proposed to
lower the r enough to completely counter any increase in T -weighted relaxivity that would
1
1
result from improved aquation.
Keywords
Ligand design, manganese, magnetic resonance imaging, sensors
1

JOURNAL PRE-PROOF
Introduction
The over-production of reactive oxygen species (ROS) has been associated with a wide
range of health disorders, including several cardiovascular and neurological diseases.[1-5] The
roles of ROS in these pathologies, however, have not been fully clarified. Being able to identify
when and where ROS concentrations spike during disease progression could provide more
effective means to both diagnose and treat these conditions. Spatiotemporal patterns of oxidative
stress in the brain could, for instance, help to differentiate neurological conditions that give rise
to similar clinically observable symptoms. In order to identify these and similar patterns, one
needs a redox-active sensor that can reliably function in vivo. Probes that alter how tissues and
organs appear when visualized by magnetic resonance imaging (MRI) are attractive in that this
spectroscopic technique allows for non-invasive imaging.
Our laboratory has previously reported a series of redox-responsive contrast agents for
MRI.[6-8] Our general strategy has been to complex a redox-active organic ligand to a Mn(II)
ion. The high paramagnetism of high-spin Mn(II) (S = 5/2) results in a high T -weighted
1
relaxivity (r ). The redox activity of manganese allows it to catalyze the oxidation of the organic
1
component by a ROS. Our more recent work has focused on using polydentate ligands with
quinols as the redox-active moieties.[7, 8] Transition metal ion-catalyzed oxidation by H O
2
2
converts these to para-quinones (Scheme 1), which are more readily displaced by water
molecules. The greater aquation increases r , improving MRI contrast, and a ligand with two
1
quinols, H qp2 (Scheme 2), provides a larger response.[8]
4
Scheme 1. Note that this graphic originally appeared in reference [8]
2

JOURNAL PRE-PROOF
Scheme 2
The disadvantage of this strategy is that the oxidized ligand necessarily has a weaker
binding affinity for Mn(II). Indeed, the reaction between H O and [Mn(H qp2)Br ] likely
2
2
4
2
releases a substantial amount of Mn(II), based on the relatively low stability of the pre-oxidized
complex and the 2.8 aquation number (q) measured for the oxidized mixture.[8] The measured q
2
+
is likely elevated by the presence of [Mn(H O) ] (q = 6), but the partial oxidation of the quinols
2
6
prevented us from more definitively assessing the stabilities of the oxidized products.
Approximately 70% of the quinols are oxidized with excess H O ; this results in a mixture of
2
2
Mn(II) complexes with unreacted H qp2, a partially oxidized ligand with only one para-quinone
4
(
H qp2), and the fully oxidized ligand with two para-quinones (qp2). The loss of Mn(II) from
2
the probe is problematic for in vivo applications since free Mn(II) is known to be toxic.[9]
In the current work, we attempt to eliminate oxidation-triggered release of Mn(II) by
substituting carboxylate groups for the pyridines in H qp2. Although such a substitution could be
4
expected to lower the thermodynamic barrier for the oxidation of Mn(II) to Mn(III), Caravan’s
group had recently used phenol- and carboxylate-containing ligands to produce Mn(II)
complexes that are sufficiently stable to air for use as MRI contrast agents.[10-12] Further, the
resulting ligand would have a strong resemblance to N,N’-bis(2-hydroxybenzyl)ethylenediamine-
N,N’-diacetic acid (HBED, Scheme 2), which was synthesized and characterized by Martell and
co-workers and forms a stable complex with Mn(II).[13, 14] Additionally, we have observed that
a sufficiently redox-active ligand can direct redox chemistry away from the metal ion by serving
as a sacrificial oxidant or reductant.[15]
We synthesized N,N’-bis(2,5-dihydroxybenzyl)ethanediamine-N,N’-diacetic acid (H qc1,
6
Scheme 2) as a new hexadentate ligand. This molecule retains the two quinols of H qp2 that are
4
essential to the large H O response of its Mn(II) complex but replaces the pyridines with
2
2
carboxylic acids, which readily deprotonate to carboxylate anions above pH 5.0. The installation
3

JOURNAL PRE-PROOF
of the carboxylic acids is anticipated to improve the binding affinity of the ligand by providing a
more anionic coordination environment for cationic metal ions at ambient pH. The ligand is
otherwise similar to H qp2; this enables us to focus on determining how ligand charge impacts
4
the stability of the sensor and its response to H O .
2
2
Experimental Section
Materials
All chemicals were purchased from Sigma-Aldrich and used as received, unless noted
otherwise. Diethyl ether (ether), ethyl acetate (EtOAc), methanol (MeOH), dichloromethane
(
CH Cl ), and ethanol (EtOH) were bought from VWR. 1,2-Ethylenediamine was purchased
2 2
from Fluka. All deuterated solvents were bought from Cambridge Isotopes and used as received.
Instrumentation
1
13
H and C NMR spectra were recorded on either a 400 MHz or 600 MHz AV Bruker
NMR spectrometer. IR spectra were collected with a Nicolet iS-50 spectrometer. Electron
paramagnetic resonance (EPR) spectra were collected on a Bruker EMX-6/1 X-band EPR
spectrometer operated in the perpendicular mode and analyzed with the program EasySpin. All
EPR samples were run as frozen solutions in quartz tubes. High-resolution mass spectrometry
(HR-MS) data were obtained at the Mass Spectrometry Center at Auburn University on a Bruker
microflex LT MALDI-TOF mass spectrometer via direct probe analysis operated in the positive
ion mode.
Magnetic Resonance Imaging (MRI)
All MRI data were collected at the Auburn University MRI Research Center on a
Siemens Verio open-bore 3-T MRI clinical scanner; a 15-channel knee coil was used to
simultaneously image 12-15 samples. The imaging procedure was identical to those used for
similar studies from our laboratory.[6-8, 16] An inversion recovery (IR) sequence was used that
featured a non-selective adiabatic inversion pulse followed by a slice-selective gradient recalled
echo (GRE) readout after a delay period corresponding to the inversion time (TI).[17, 18] The
GRE was a saturation readout, such that only one line of k-space was acquired per repetition
time (TR), in order to maximize both signal strength and the accuracy of the T estimates. The
1
specific imaging parameters were as follows: TR was set to 4 s, TI was varied from 4.8 to 2500
ms over 37 steps, the echo time (TE) was set to 3.6 ms, the flip angle equaled 90°, averages = 1,
4

JOURNAL PRE-PROOF
slice thickness = 5 mm, field of view = 140 × 140 mm, matrix = 128 × 128, resulting in a pixel
size of 1.1 × 1.1 × 5.0 mm. All samples were run in 50 mM solutions of HEPES in water,
buffered to pH 7.0 and kept at 22 °C. The manganese content was systematically varied from
0
.10 to 1.00 mM. The inverses of the T values were plotted versus the concentration of Mn(II)
1
to obtain r values.
1
MRI Data Analysis
Image analysis was performed using custom Matlab programs (Mathworks, Natick, MA).
The initial TI = 4.8 ms image was used as a baseline to determine circular region of interest
ROI) boundaries for each sample; from these, the mean pixel magnitudes for each ROI were
(
calculated. For each of the 36 subsequent TI images, the same ROI boundaries were applied, and
the mean pixel magnitude calculations were repeated. This gave consistent ROI spatial
definitions and a corresponding time course of magnitudes for each of the samples over all the TI
time points. Each sample's complex phase was used to correct the magnitude polarity to produce
a complete exponential T inversion recovery curve. The Nelder-Mead simplex algorithm[19]
1
was applied to each sample's exponential curve to estimate its corresponding T value.
1
Synthesis
2
,5-Bis(tert-butyldimethylsiloxy)benzaldehyde (1).
This compound was prepared through a previously reported procedure.[20] 2,5-
Dihydroxybenzaldehyde (700 mg, 5.07 mmol) and imidazole (1.04 g, 15.2 mmol) were dissolved
in 10 mL of anhydrous N,N-dimethylformamide under a N atmosphere. The resultant solution
2
was then cooled to 0 °C using an ice bath. tert-Butyldimethylsilyl chloride (1.90 g, 12.6 mmol)
was added, and the reaction mixture stirred and warmed to room temperature (RT) over the next
1
6 h. 100 mL of water were then added to quench the reaction, and the crude product was
extracted with three 50 mL portions of EtOAc. The organic layers were combined, washed with
three 100 mL aliquots of water, and dried over sodium sulfate. After the EtOAc was removed by
rotavaporation, the product was purified by column chromatography using 50:1 hexanes/EtOAc
1
as the eluent (R = 0.2) to yield 1.60 g of the product as a yellow oil (85% yield). H NMR (400
f
MHz, CDCl , 293 K): δ 10.39 (s, 1H), 7.24 (m, 1H), 6.97 (m, 1H), 6.76 (m, 1H), 1.02 (s, 9H),
3
1
3
0
1
3
.98 (s, 9H), 0.25 (s, 6H), 0.19 (s, 6H). C NMR (100 MHz, CDCl , 293 K): 189.9, 153.4,
3
49.8, 127.9, 127.4, 121.2, 117.7, 25.7, 25.6, 18.3, 18.1, -4.4, -4.5. MS (ESI): Calcd for MH+,
-1
67.2125; Found, 367.1879. IR (cm ): 2955.36 (m), 2930.04 (m), 2886.00 (w), 2857.88 (m),
5

JOURNAL PRE-PROOF
1
1
8
5
686.29 (m), 1608.23 (w), 1485.10 (s), 1419.96 (m), 1387.18 (m), 1362.25 (w), 1300.30 (w),
253.61 (s), 1209.72 (s), 1150.68 (s), 1006.29 (w), 983.85 (m), 909.32 (s), 836.98 (s), 821.89 (s),
05.21 (s), 777.97 (s), 732.67 (w), 684.67 (m), 668.07 (w), 640.88 (w), 611.26 (w), 581.31 (w),
37.82 (w), 449.69 (w).
N,N’-Bis[2,5-bis(tert-butyldimethylsiloxy)benzyl]-1,2-ethanediamine (2).
,5-Bis(tert-butyldimethylsiloxy)benzaldehyde (1.10 g, 3.00 mmol) and ethylenediamine
90 mg, 1.50 mmol) were dissolved in 20 mL of dry MeOH and stirred for 16 h under N at RT.
2
(
2
1H NMR analysis of an aliquot taken at this time indicated that the aldehyde was fully consumed.
The mixture was then cooled to 0 °C with an ice bath. Sodium borohydride (153 mg, 4.05 mmol)
was added to the cooled solution, and the resultant mixture was stirred for 5 h. The MeOH was
then stripped from the solution, and the crude product was dissolved in CH Cl . The crude
2
2
product was washed with three portions of saturated NaHCO . Additional product was extracted
3
from the washes with CH Cl , after which all the organic layers were combined. After the
2
2
solution was dried over sodium sulfate, the CH Cl was removed to yield the product as a
2
2
1
yellow/brown oil (1.10 g, 75% yield). H NMR (400 MHz, CDCl , 293 K): δ 6.70 (d, J = 3.2 Hz,
3
2
1
4
H), 6.58 (m, 4H), 3.69 (s, 4H), 2.71 (s, 4H), 0.98 (s, 18H), 0.97 (s, 18H), 0.19 (s, 12H), 0.16 (s,
1
3
2H). C NMR (100 MHz, CDCl , 293 K): δ 149.37, 147.89, 131.51, 121.33, 118.85, 118.5,
3
+
9.34, 48.76, 25.84, 25.74, 18.17, 18.15, -4.16, -4.43. MS (ESI): Calcd for MH , 761.4960;
-1
Found, 761.4390. IR (cm ): 2954.75 (m), 2928.86 (m), 2885.76 (w), 2857.11 (m), 1490.24 (s),
1
1
5
471.75 (m), 1418.47 (w), 1399.57 (w), 1361.42 (w), 1251.29 (s), 1222.56 (s), 1156.34 (w),
094.93 (w), 982.76 (w), 910.63 (s), 835.60 (s), 799.82 (s), 776.11 (s), 684.52 (m), 666.23 (w),
85.37 (w), 449.33 (w).
N,N’-Bis[2,5-bis(tert-butyldimethylsiloxy)benzyl]-N,N’-bis(tert-butyl-acetate)-1,2–
ethanediamine (3).
N,N’-Bis[2,5-bis(tert-butyldimethylsiloxy)-benzyl]-1,2-ethanediamine (1.10 g, 1.44
mmol), tert-butylbromoacetate (676 mg, 3.47 mmol) and N,N’-diisopropylethylamine (467 mg,
3
.61 mmol) were dissolved in 35 mL of CH Cl and stirred for 24 h. The solution was diluted
2 2
with 75 mL of CH Cl and washed with three 100 mL portions of saturated NaHCO . The
2
2
3
organic layer was then washed with brine, dried over Na SO for 30 min, and rotavapped to yield
2
4
a brown oil. The dark brown oil was then dissolved in minimal MeOH. Over the next 16 h, 0.398
1
g of the product precipitated as a white solid (28% yield). H NMR (400 MHz, CDCl , 293 K): δ
3
6

JOURNAL PRE-PROOF
6
4
.95 (d, J = 2.8 Hz, 2H), 6.60 (d, J = 8.2, 2H), 6.54 (m, 2H), 3.71 (s, 4H), 3.27 (s, 4H), 2.82 (s,
1
3
H), 1.43 (s, 18H), 0.98 (s, 18H), 0.96 (s, 18H), 0.16 (s, 12H), 0.15 (s, 12H). C NMR (100
MHz, CDCl , 293 K): δ 171.02, 149.54, 147.82, 130.63, 121.06, 118.93, 118.35, 80.42, 55.77,
3
5
9
1
1
2.88, 52.78, 28.19, 25.92, 25.77, 18.28, 18.16, -4.13, -4.42. MS (ESI): Calcd for MH+,
-1
89.6322; Found, 989.6359. IR (cm ): 2950.62 (m), 2928.32 (m), 2894.88 (w), 2857.30 (m),
719.16 (s), 1487.17 (s), 1470.49 (m), 1416.31 (w), 1390.24 (w), 1364.91 (m), 1288.47 (m),
246.74 (s), 1200.42 (s), 1150.96 (m), 1125.18 (w), 1083.85 (w), 999.09 (w), 973.86 (m), 951.58
(
w), 911.60 (s), 888.53 (m), 840.27 (s), 823.08 (s), 693.85 (m), 625.48 (w), 587.08 (m), 542.50
w), 492.11 (w), 455.49 (w).
(
N,N’-Bis(2,5-dihydroxybenzyl)-N,N’-bis(tert-butyl-acetate)-1,2–ethanediamine (4).
N,N’-Bis[2,5-bis(tert-butyldimethylsiloxy)benzyl]-N,N’-bis(tert-butyl-acetate)-1,2–
ethanediamine (536 mg, 0.542 mmol) was dissolved in 40 mL of tetrahydrofuran (THF) and
cooled to 0 °C. Acetic acid (0.50 mL, 8.7 mmol) and 2.71 mL of a 1.0 M solution of tetra-N-
butylammonium fluoride (TBAF) in THF (2.71 mmol) were added. After the reaction mixture
stirred for 3 h, 0.60 additional mL of the 1.0 M TBAF solution were added. The resultant
solution stirred for another 2 h to ensure full removal of the tert-butyldimethylsilyl groups. 100
mL of water were added to quench the reaction, and the THF was removed by rotavaporation.
The product was extracted from the aqueous solution with three 50 mL aliquots of CH Cl . The
2
2
combined organic layers were washed with 150 mL of brine before being dried over Na SO for
2
4
3
0 min. The crude was purified by column chromatography using 1:1 hexanes/EtOAc as the
eluent (R = 0.25) to yield 222 mg of the product as a yellow oil (77% yield). We were unable to
f
1
remove impurities that retained either one or two of the silane protecting groups. H NMR (400
MHz, CD CN, 293 K): δ 9.06 (bs, 2H), 6.62 (m, 4H), 6.43 (d, J = 2.4, 2H), 6.40 (bs, 2H), 3.61
3
1
3
(
s, 4H), 3.15 (s, 4H), 2.62 (s, 4H), 1.45 (s, 18H). C NMR (100 MHz, CD CN, 293 K): δ
3
1
70.87, 150.24, 149.61, 122.83, 116.40, 116.23, 115.49, 81.41, 56.78, 55.50, 49.29, 27.32. Calcd
+
-1
for MH , 533.2863; Found, 533.2871. IR (cm ): 3211.89 (w), 2964.06 (m), 1726.78 (s), 1496.02
(
s), 1366.96 (s), 1226.52 (s), 1148.54 (s), 983.99 (w), 893.64 (m), 816.58 (m), 776.75 (s), 735.93
w), 597.05 (w).
(
N,N’-Bis(2,5-dihydroxybenzyl)ethanediamine-N,N’-diacetic acid (H qc1).
6
N,N’-Bis(2,5-dihydroxybenzyl)-N,N’-bis(tert-butyl-acetate)-1,2-ethanediamine (222 mg,
0
.417 mmol) was dissolved in 10 mL of 90:5:5 trifluoroacetic acid/water/triisopropylsilane (TIS)
7

JOURNAL PRE-PROOF
and stirred for 24 h. The solvents were removed by rotavaporation. The product was washed with
ether multiple times to remove the residual TIS and dried over air to yield 150 mg of the product
1
as a white powder (86% yield). H NMR (400 MHz, DMSO-d , 293 K): δ 8.83 (s, 2H), 6.62 (m,
6
1
3
6
1
4
H), 3.93 (s, 4H), 3.60 (s, 4H), 3.14 (s, 4H). C NMR (100 MHz, DMSO-d , 293 K): δ 170.87,
6
50.22, 149.20, 120.27, 118.56, 116.94, 116.51, 53.22, 52.84, 50.11. MS (ESI): Calcd for MH+,
-1
21.1611; Found, 421.1596. IR (cm ): 3043.9 (s), 1669.7 (s), 1513.11 (w), 1459.2 (m), 1435.4
(
w), 1385.3 (m), 1365.6 (m), 1319.7 (w), 1261 (m), 1193.2 (s), 1131.2 (s), 1010.6 (w), 991.5 (w),
68.3 (w), 907.4 (w), 840.6 (m), 799.9 (m), 760.1 (m), 719.1 (m). UV/vis (50 HEPES buffered to
9
-1
-1
pH 7.4): 294 nm (3500 M cm ).
Scheme 3. Synthesis of H qc1
6
Results
Synthesis of H qc1
6
The ligand can be prepared in five steps from commercially available starting materials
(Scheme 3); the first intermediate, 2,5-bis(tert-butyldimethylsiloxy)benzaldehyde (1), was
previously synthesized by another research group.[20] The synthesis of H qc1 is more difficult
6
than those used to prepare H qp2 and the related H qp1 (Scheme 2) for several reasons. First,
4
2
two additional steps are required due to the need to protect both the quinols and the carboxylic
acids. With H qp1 and H qp2, conversely, the quinols are added to the ligand framework late
2
4
8

JOURNAL PRE-PROOF
enough in the synthesis to obliviate the need for their protection.[7, 8] Second, two of the
intermediates need to be purified by column chromatography. H qp1 can be obtained in high
2
purity without any chromatography;[7] whereas, H qp2 requires just a single column at the end
4
of the synthesis.[8] Third, the addition of the protected carboxylic acids to yield 3 proved to be
challenging. We eventually found that highly pure material would precipitate from MeOH, albeit
in low yield (28%). Alternative means of purifying 3, including column chromatography, did not
effectively remove impurities. The immediate precursor to H qc1, 4, could not be obtained in
6
high purity, but the impurities were removed in the final step of the synthesis, as confirmed by
NMR.
We attempted to isolate a Mn(II) complex with H qc1 but were unable to precipitate
6
clean material from a variety of solvents, including water, MeCN, CH Cl , and MeOH. Given
2
2
this, studies on the Mn(II) were done on samples prepared in situ; this has been done with other
manganese-containing complexes, including some MRI contrast agents.[11-13]
Aqueous Characterization of the H qc1 Ligand
6
We analyzed the behavior of the H qc1 ligand by itself in an aqueous solution containing
6
1
00 mM KCl via potentiometric pH titrations. Our best fitting model for the titration data
displays five ionization events as the pH increased from 2 to 10.5 (Table 1, Figure S19). We did
not collect or model data past pH 10.5, as was done with HBED,[13, 14] since the ligand
displays irreversible degradation under those conditions. Using the speciation of HBED as a
2
+
model, the H qc1 ligand likely exists as [H qc1] under extremely acidic conditions, with the
6
8
extra two protons localized on each of the amines.[14] The first three ionization events
correspond to pK values of 2.45 (±0.3), 2.89 (±0.05), and 4.31 (±0.05). These likely correspond
a
to the deprotonation of the first ammonium and the two carboxylic acids and the formation of
+
-
H qc1 , H qc1, and H qc1 . The amine and the carboxylate conjugate bases are anticipated to
7
6
5
accept intramolecular hydrogen bonds from the remaining ammonium and the two quinols, as
was seen in the crystal structure of the free H qp1 ligand.[21] The other two pK values resulting
2
a
from the model are 8.80 (±0.05) and 11.67 (±0.3); the associated deprotonation reactions yield
2
-
3-
H qc1 and H qc1 . The error in the 11.67 pK is larger since we could not fully generate the
4
3
a
species during the titrations. The related HBED ligand displays similar acid/base chemistry, with
pK values of 1.7, 2.53, 4.72, 8.44, and 11.00, plus another at 12.60.[14] The UV/vis spectrum of
a
H qc1 in water changes substantially as the pH increases from 7.4 to 9.0 (Figure S21). At low pH
6
9

JOURNAL PRE-PROOF
values, there is a strong feature at 299 nm, which is consistent with a phenol or quinol group.[8,
1
0] As the solution becomes more basic, however, a lower-energy band at 330 nm develops; the
energy of this feature is more consistent with a phenolate or quinolate group. The proton that is
lost from the ligand as the pH increases from 7.4 to 9.0 appears to be at least partially localized
on the quinol. Between pH 7.0 and 7.4, the H qc1 ligand therefore predominantly exists as
6
-
2-
H qc1 , with a trace amount of H qc1 (Figure 1).
5
4
Figure 1. Predicted speciation as a function of pH for 1.0 mM H qc1 in 100 mM KCl solution.
6
Table 1. pMn and pK Values for the Ligands and Mn(II) Complexes with H qc1 Determined by
a
6
Potentiometric Titration at 25 °C.
a
pKa1b
pKa2b
pKL1
11.670 (±0.3)
8.02 (±0.06)
6.19 (±0.06)
pKL2a
8.80 (±0.05)
pKL3a
4.31 (±0.05)
pKa3b
4.24 (±0.06)
pKL4a
2.89 (±0.05)
2.45 (±0.3)
log K(MnH qc1)
c
c
c
15.59
10.11
5.55
3
pKL5a
log K(MnH qc1)
4
log K(MnH qc1)
5
pMn (pH 7.4)d
6.67
a
-
Ligand pK values correspond to the following equilibrium constants: K = [(H qc1)
a
L1
3
+
2-
2-
+
-
]
[H ]/[(H qc1) ], pK = logβ – logβ ; K = [(H qc1) ][H ]/[(H qc1) ], pK = logβ210
–
=
=
4
L1
110
-
010
L2
4
5
L2
+
logβ ;
[
K
=
[(H qc1) ][H ]/[(H qc1)], pKL3
=
logβ310
–
logβ ; KL4
1
10
L3
5
6
210
+
+
+
+
2+
(H qc1)][H ]/[(H qc1) ], pK = logβ410 – logβ310; KL5 = [(H qc1) ][H ]/[(H qc1) ], pK
6 7 L4 7 8 L5
logβ – logβ .
5
10
410
b
Metal complex pK values correspond to the following equilibrium constants: K
=
a
a1
1
0

JOURNAL PRE-PROOF
2
-
+
-
-
[
]
[Mn(H qc1)] ][H ]/[[Mn(H qc1)] ], pK = logβ011 – logβ-111; K
= [[Mn(H qc1)]
2
3
a1
a2
3
+
+
+
[H ]/[[Mn(H qc1)]], pK = logβ111 – logβ .; K = [[Mn(H qc1)]][H ]/[[Mn(H qc1) ]], pK
4
a12
011
a3
4
5
a13
=
logβ211 – logβ111.
c
Metal complex stability constants correspond to the following equilibrium constants:
K(MnH qc1) = [[Mn(H qc1) ]]/[Mn ][H qc1 ]; K(MnH qc1) = [[Mn(H qc1)]]/[Mn ][H qc1
-
2+
3-
2+
2-
3
3
3
4
4
4
+
2+
-
]
; K(MnH qc1) = [[Mn(H qc1) ]]/[Mn ][H qc1 ]
log(free Mn(II)) at pH 7.4 with 1.0 mM Mn(II) and 1.0 mM H qc1.
5 5 5
d
6
1
1

JOURNAL PRE-PROOF
Aqueous Characterization of the Mn(II) Complex with H qc1
6
The speciation of Mn(II) complexes with H qc1 in water was analyzed using a 100 mM
6
solution of KCl containing a 1:1 mixture of MnCl and ligand. The curves resulting from the
2
potentiometric pH titrations (Figure S22) look distinct from those for the metal-free H qc1, and
6
only four clear ionization events are observed as the pH is increased from 2.6 to 10.2. As with
our analysis of the titration data for free H qc1, the inclusion of additional species into the model
6
worsened the fit (Figure S23). The lowest pH ionization event corresponds to the release of
Mn(II) from the ligand. The next three ionization events are assigned to (de)protonation events
for the H qc1-Mn(II) complex and are correspond to pK values of 4.24 (±0.06), 6.19 (±0.06),
6
a
and 8.02 (±0.06) (Table 1). The UV/vis bands of an aqueous 1:1 mixture of MnCl and H qc1
2
6
change substantially from pH 5 to 9 (Figure S24), leading us to assign the 6.19 and 8.02 pKa
values to the deprotonation of the Mn(II)-bound quinols.[8] The acid/base behavior is similar to
that of the Mn(II) complex with HBED, which has pK values of 6.58 and 7.66 for the
a
deprotonation of Mn(II)-bound phenols.[13]
The speciation of the Mn(II) as a function of pH is shown in Figure 2. Our model of the
data suggests that the Mn(II)-H qc1 complex is indeed more stable than Mn(II)-H qp2 but is less
6
4
stable than Mn(II)-H qp1 The pMn value at pH 7.4, 1.0 mM total Mn(II), and 1.0 mM total
2
ligand is 6.67; the pMn values for the H qp1 and H qp2 systems under identical conditions,
2
4
conversely, were found to be 7.25 and 5.36, respectively.[8, 22] At pH 7, the Mn(II) exists
-
2-
predominantly as [Mn(H qc1)] , with a substantial amount of [Mn(H qc1)] . Since we could not
3
2
4
-
assess the stability of the fully deprotonated ligand, H qc1 , we were unable to measure a K
2
eq
value for the complexation of the tetraanionic ligand with Mn(II), but we were able to measure
-
2-
3-
values for the Mn(II) complexes with H qc1 , H qc1 , and H qc1 (Table 1). The Mn(II)
5
4
3
complexes with doubly and triply deprotonated HBED appear to be less stable than their H qc1
6
analogs, with log K values of 5.56 and 9.98 respectively.[13]
1
2

JOURNAL PRE-PROOF
Figure 2. Predicted speciation of Mn(II) as a function of pH for an aqueous solution containing
1
.0 mM MnCl , 1.0 mM H qc1, and 100 mM KCl.
2 6
Oxidative Stability of the H qc1 Ligand and its Mn(II) Complex
6
Neither the ligand nor its Mn(II) complex appears to react with O in aqueous solution
2
over several hours as assessed by UV/vis analysis of solutions of these compounds exposed to air
(Figures S25 and S26). The UV/vis spectrum of each compound remains approximately the
same. The stability of the Mn(II) complex to air was also confirmed by EPR (Figure S27).
Although the H qc1 ligand by itself does not react with H O over the course of 1 h (Figure S28),
6
2
2
its Mn(II) complex reacts quickly with this oxidant as assessed by both UV/vis and EPR. In 50
mM HEPES buffered to pH 7.0, the Mn(II) complex initially displays three bands at 298 nm, 368
nm, and 481 nm (Figure 3). Upon the addition of H O , all three peaks increase in intensity over
2
2
the course of 5 min but then weaken. The band at 298 nm eventually disappears completely,
suggesting that the quinols have been oxidized. Over the course of 1 h, a broad band centered at
4
82 nm steadily develops. UV/vis peaks with similar energies have been previously observed for
Mn(III) species.[23, 24] Parallel reactions analyzed by EPR confirm that the metal is being
oxidized to an EPR-silent species, with the Mn(II) signal steadily vanishing over 30 min (Figure
4
). We are currently unable to separate these oxidized manganese products and have therefore
been unable to obtain stability measurements for any of the individual species.
We were unable to obtain clean samples of the demetallated and oxidized form(s) of the
ligand. We instead analyzed the ligand oxidation by adding Zn(OTf) to reactions containing 1:1
2
mixtures of H qc1 and MnCl with or without H O . The Zn(II) displaces the manganese from
6
2
2
2
1
the ligand, yielding species that can be readily visualized by H NMR. The addition of H O
2
2
1
3

JOURNAL PRE-PROOF
decreases the intensity of the aromatic peaks relative to the benzylic ones; this is consistent with
the loss of the OH protons and the oxidation of the quinols to para-quinones (Figure S29).[7, 8]
Figure 3. Spectrophotometric response of an aqueous solution containing 0.1 mM MnCl , 0.1
2
mM H qc1, and 50 mM HEPES buffered to pH 7.0 to 43 mM H O .
6
2
2
Figure 4. EPR spectra showing the oxidation of an aqueous 1:1 mixture of MnCl and H qc1 by
2
6
H O over 30 min. All solutions contained 1.0 mM MnCl , 1.0 mM H qc1, and 50 mM HEPES
2
2
2
6
buffered to pH 7.0. 10 mM of H O was added, and the reaction was allowed to proceed at RT.
2
2
Aliquots were removed and frozen for EPR analysis at 15 min and 30 min.
Measurement of T -Weighted Relaxivity for the Mn(II) and its Oxidized Products
1
The Mn(II) complex with H qc1 displays distinct spectroscopic changes upon reaction
6
with H O , but not O , prompting us to investigate whether the compound could selectively
2
2
2
1
4

JOURNAL PRE-PROOF
detect H O via changes in its T -weighted relaxivity (r ). We characterized the reaction between
2
2
1
1
H O and the Mn(II)-H qc1 complex using methodology that we had developed for prior H O
2
2
2
6
2
sensors.[6-8] We measured the T values of 1:1 mixtures of MnCl and H qc1 in aqueous
1
2
6
solutions buffered to pH 7.0 using a 3T MRI scanner. The concentration of Mn(II) was varied
from 0.10 mM to 1.0 mM, and the slope of (1/T ) versus the concentration provides r . A second
1
1
set of solutions contained 10 mM H O in order to oxidize the Mn(II)-H qc1 complex; these
2
2
6
were allowed to react for 30 min in order to ensure that the reactions went to completion.
Unexpectedly, the presence of H O neither increased nor decreased the r1 of the
2
2
manganese/H qc1 mixtures. In phosphate buffer, the r values before and after adding H O , 3.48
6
1
2
2
-1
-1
and 3.46 mM s , are within error of each other (Figure 5). A significant response to H O was
2
2
likewise absent when the complex was studied in 50 mM HEPES buffered to pH 7.0. In one
instance, we observed a 10% increase in r (Figure S30), but this was not reproducible.
1
Figure 5. T -weighted relaxivity (r ) data for 1:1 mixtures of MnCl and H qc1 in the absence
1
1
2
6
and presence of excess (10 mM) H O in an aqueous solution containing 50 mM Na PO
2
2
3
4
buffered to pH 7.00 All samples were analyzed at 298 K with a 3 T clinical MRI scanner. All
.
samples were prepared under air and run within 30 min of preparation.
Discussion
Redox-active ligands have found increasing use within many subfields of inorganic
chemistry.[25-31] Our own interest in this area has centered on using quinol-containing
molecules to prepare complexes that display oxidation-triggered increases in their T -weighted
1
relaxivity (r ) and/or catalyze the degradation of superoxide.[7, 8, 21] Both applications rely
1
upon the quinols within the organic ligands reversibly oxidizing to para-quinones (Scheme 1).
1
5

JOURNAL PRE-PROOF
The first two polydentate ligands, H qp1 and H qp2 (Scheme 2), use amines and pyridines as the
2
4
other chelating groups. Although their complexes with Zn(II) thus far appear to be highly stable
in water,[21] these neutral N-donors bind much more weakly to Mn(II).[8] As a consequence,
the H qp2 complex with Mn(II) is just barely stable enough to function at pH 7, and its oxidation
4
to di-para-quinone (qp2) and mono-para-quinone forms (H qp2) appear to release substantial
2
amounts of Mn(II).
In order to prepare a ligand that can more tightly coordinate Mn(II), we have replaced the
pyridines of H qp2 with carboxylic acids to yield H qc1 (Scheme 2). Although the ligand is
4
6
much more challenging to synthesize than H qp2, the changes to the molecule do improve its
4
affinity for Mn(II). The pMn value for H qc1 (pH 7.4, 1.0 mM total Mn(II), 1.0 mM total ligand)
6
is 6.67, which is over an order of magnitude improvement over the 5.36 value reported for the
H qp2 system but is worse than the 7.25 value measured for H qp1.[8, 22] Substantial metal
4
2
dissociation is still observed under highly acidic conditions (Figure 2). The superior binding
affinity of H qp1 for Mn(II) is initially difficult to rationalize, since the latter ligand cannot attain
2
as negative a charge. The quinol portions of the ligands, however, appear to bind to metal ions
poorly in their neutral forms, and they are often detached from the metal ions in crystal
structures.[8, 21] Under neutral to acidic conditions, H qp1 consistently provides five strong
2
donor atoms as opposed to the four from H qc1. We believe that the extra effective donor atom
6
of H qp1 more than compensates for its less negative charge, particularly under acidic
2
conditions.
Unexpectedly, the quinols appear to bind to the Mn(II) much more tightly than phenols.
-
The stability constants of the [Mn(H qc1)] and [Mn(H qc1)] complexes (Table 1) are higher
3
4
than those for their analogs with the related HBED ligand, which contains phenols instead of the
quinols found in H qc1 (Scheme 2).[13] This appears to be a reproducible effect since the Mn(II)
6
complex with H qp1 is substantially more stable than the Mn(II) complex with Hptp1 (Scheme
2
4
), which has a pMn of 5.40 at pH 7.4.[15] The OH group para to the coordinating O-donor of
the quinol is a more strongly electron-donating substituent than the H group that occupies its
place in a phenol. We speculate that the OH for H substitution renders quinolates more electron-
rich and thereby increases their affinities for metal ions.
1
6

JOURNAL PRE-PROOF
Scheme 4
The more anionic charge of the H qc1 ligand has a weak impact on the interaction
6
between the quinols and the metal center. The Mn(II) complexes with H qc1 and H qp2 have
6
4
similar pK values for their Mn(II)-quinol groups: 6.19 and 8.02 for H qc1 versus 5.82 and 7.14
a
6
for H qp2.[8] The acid/base behavior of the Mn(II) complex with H qc1 strongly resembles that
4
6
observed for its analog with HBED, which has pK values of 6.58 and 7.66.[13]
a
4
-
Highly anionic ligands, such as ethylenediaminetetraacetate (EDTA ), often stabilize
Mn(III) species,[32] and the more negatively charged coordination sphere provided by H qc1
6
around pH 7 does render the metal center more susceptible to oxidation. Neither the H qp2 nor
4
the H qc1 complex with Mn(II) displays any substantial spectroscopic change upon a 4 h
6
reaction with air. [Mn(H qp2)Br ] is slowly oxidized a Mn(III) species upon reaction with H O ,
4
2
2
2
but this is not noticeable by EPR at 30 min.[8] The Mn(II) complex with H qc1, conversely,
6
quickly reacts with H O to yield what we believe are one or more Mn(III)-containing products,
2
2
as assessed by both UV/vis and EPR (Figure 3, Figure 4). Oxidation to Mn(IV) is unlikely. This
oxidation state is readily observable by EPR, but new signals do not appear in the EPR spectrum
of the H qc1 complex as the reaction with H O proceeds. The ligand appears to be oxidized as
6
2
2
1
well, as evidenced by both UV/vis and H NMR (Figure 3, Figure S29). We have thus far been
unable to isolate any individual manganese-containing products from these reaction mixtures.
It is challenging to predict the overall impact of H O on the T -weighted relaxivity of the
2
2
1
manganese complex with H qc1 since both the ligand and metal are being oxidized during the
6
reaction. The Mn(II) sensors with H qp1 and H qp2 react with H O to yield Mn(II) species with
2
4
2
2
higher r values, which we attributed to water molecules displacing the para-quinone groups of
1
the oxidized ligands (Scheme 1).[7, 8] More highly aquated metal centers tend to have higher r1
1
7

JOURNAL PRE-PROOF
values. The oxidation of the metal center by H O , conversely, would be anticipated to worsen
2
2
the r since the metal ion is rendered less paramagnetic by this process.[10, 33] Another factor
1
that impacts r is the rate of water exchange. Oxidation to Mn(III) would be anticipated to slow
1
the rate of exchange,[34] but the value that would optimize r has not yet been established for
1
either Mn(II) or Mn(III) complexes. Unexpectedly, the oxidation of 1:1 Mn(II)/H qc1 mixtures
6
by H O results in no observable change to the r at pH 7.0 (Figure 5). The relaxivity is impacted
2
2
1
by the choice of buffer; the Mn(II)-H qc1 complex is more effective in 50 mM HEPES solution
6
-1
-1
-1 -1
(
3.98 mM s ) than in 50 mM phosphate (3.48 mM s ). Phosphate is known to compete with
superoxide in manganese-containing superoxide dismutase mimics,[15, 35, 36] and this may
decrease r by competing with water for vacant coordination sites on the metal ion. In one set of
1
experiments, we observed a 10% increase in r in HEPES solution upon adding H O (Figure
1
2
2
S30), which is much smaller than the 30% increase we observed for the H qp2 system,[8] but we
4
could not replicate this in other experiments. Determining whether the metal center is indeed
become more highly aquated during the reaction with H O is complicated by the lack of a
2
2
protocol for measuring solution state aquation numbers for Mn(III) species.
Conclusion
The substitution of two carboxylate groups for the pyridines in the H qp2 ligand allows
4
the resultant H qc1 ligand to bind more tightly to Mn(II). Unfortunately, the more anionic
6
coordination sphere facilitates oxidation of the metal center, which in turn, eliminates the H O -
2
2
triggered increase in T -weighted relaxivity that was observed for the related H qp2 probe.
1
4
Although the new ligand is unsuitable for manganese-containing MRI contrast agent sensors for
H O , the presence of redox-active functionalities and its strongly anionic charge may make this
2
2
ligand useful for other applications.
Acknowledgements
This work was supported by Auburn University and a grant from the National Science
Foundation (NSF-CHE-1662875). We thank Prof. Evert Duin for his assistance with acquiring
the EPR data and Mr. Owen Chandler for helpful discussions regarding the synthesis of the
ligand.
1
8

JOURNAL PRE-PROOF
1
9

JOURNAL PRE-PROOF
References
[
[
[
1] L. Tretter, I. Sipos, V. Adam-Vizi, Neurochem. Res., 29 (2004) 569-577.
2] C.K. Roberts, K.K. Sindhu, Life Sci., 84 (2009) 705-712.
3] R.L. Mosley, E.J. Benner, I. Kadiu, M. Thomas, M.D. Boska, K. Hasan, C. Laurie, H.E.
Gendelman, Clin. Neurosci. Res., 6 (2006) 261-281.
[
[
[
4] I.M. Fearon, S.P. Faux, J. Mol. Cell. Cardiol., 47 (2009) 372-381.
5] G. Eskici, P.H. Axelsen, Biochemistry, 51 (2012) 6289-6311.
6] M. Yu, R.J. Beyers, J.D. Gorden, J.N. Cross, C.R. Goldsmith, Inorg. Chem., 51 (2012) 9153-
9
[
155.
7] M. Yu, S.L. Ambrose, Z.L. Whaley, S. Fan, J.D. Gorden, R.J. Beyers, D.D. Schwartz, C.R.
Goldsmith, J. Am. Chem. Soc., 136 (2014) 12836-12839.
8] M. Yu, M.B. Ward, A. Franke, S.L. Ambrose, Z.L. Whaley, T.M. Bradford, J.D. Gorden, R.J.
Beyers, R.C. Cattley, I. Ivanović-Burmazović, D.D. Schwartz, C.R. Goldsmith, Inorg. Chem., 56
[
(
[
[
2017) 2812-2826.
9] S.L. O’Neal, W. Zheng, Curr. Environ. Health Rep., 2 (2015) 315-328.
10] E.M. Gale, S. Mukherjee, C. Liu, G.S. Loving, P. Caravan, Inorg. Chem., 53 (2014) 10748-
1
[
1
0761.
11] E.M. Gale, I.P. Atanasova, F. Blasi, I. Ay, P. Caravan, J. Am. Chem. Soc., 137 (2015)
5548-15557.
[12] G.S. Loving, S. Mukherjee, P. Caravan, J. Am. Chem. Soc., 135 (2013) 4620-4623.
[13] F. L'Eplattenier, I. Murase, A.E. Martell, J. Am. Chem. Soc., 89 (1967) 837-843.
[14] R.J. Motekaitis, A.E. Martell, M.J. Welch, Inorg. Chem., 29 (1990) 1463-1467.
[15] I. Kenkel, A. Franke, M. Dürr, A. Zahl, C. Dücker-Benfer, J. Langer, M. Filipovic, R., M.
Yu, R. Puchta, S.R. Fiedler, M.P. Shores, C.R. Goldsmith, I. Ivanović-Burmazović, J. Am.
Chem. Soc., 139 (2017) 1472-1484.
[
[
16] Q. Zhang, J.D. Gorden, R.J. Beyers, C.R. Goldsmith, Inorg. Chem., 50 (2011) 9365-9373.
17] M.A. Bernstein, K.F. King, X.J. Zhou, Handbook of MRI Pulse Sequences, Elsevier
Academic Press, Amsterdam, 2004.
18] E.M. Haacke, R.W. Brown, M.R. Thompson, R. Venkatesan, Magnetic Resonance Imaging:
Physical Principles and Sequence Design, John Wiley & Sons, New York, NY, 1999.
[
[
[
[
19] J.A. Nelder, R. Mead, Comput. J., 7 (1965) 308-313.
20] J. Yoon, J.-S. Ryu, Bioorg. Med. Chem. Lett., 20 (2010) 3930-3935.
21] M.B. Ward, A. Scheitler, M. Yu, L. Senft, A.S. Zillmann, J.D. Gorden, D.D. Schwartz, I.
Ivanović-Burmazović, C.R. Goldsmith, Nature Chem., 10 (2018) 1207-1212.
22] A. Franke, A. Scheitler, J.L. Moore, S. Sader, S. Ison, I. Ivanović-Burmazović, C.R.
Goldsmith, in, 2019.
[
[
[
[
[
23] C.R. Goldsmith, A.P. Cole, T.D.P. Stack, J. Am. Chem. Soc., 127 (2005) 9904-9912.
24] T.J. Hubin, J.M. McCormick, N.W. Alcock, D.H. Busch, Inorg. Chem., 40 (2001) 435-444.
25] J.T. Henthorn, S. Lin, T. Agapie, J. Am. Chem. Soc., 137 (2015) 1458-1464.
26] I.R. Corn, P.D. Astudillo-Sánchez, M.J. Zdilla, P.E. Fanwick, M.J. Shaw, J.T. Miller, D.H.
Evans, M.M. Abu-Omar, Inorg. Chem., 52 (2013) 5457-5463.
[
[
[
[
27] M.R. Haneline, A.F. Heyduk, J. Am. Chem. Soc., 128 (2006) 8410-8411.
28] T. Hirao, Coord. Chem. Rev., 226 (2002) 81-91.
29] T.W. Myers, L.A. Berben, Inorg. Chem., 51 (2012) 1480-1488.
30] D.L.J. Broere, R. Plessius, J.I. van der Vlugt, Chem. Soc. Rev., 44 (2015) 6886-6915.
2
0

JOURNAL PRE-PROOF
[31] O.R. Luca, R.H. Crabtree, Chem. Soc. Rev., 42 (2013) 1440-1459.
[32] F.A. Cotton, G. Wilkinson, Advanced Inorganic Chemistry, 5 ed., John Wiley & Sons, New
York, 1988.
[
[
33] P. Caravan, J.J. Ellison, T.J. McMurry, R.B. Lauffer, Chem. Rev., 99 (1999) 2293-2352.
34] A. Cusanelli, U. Frey, D.T. Richens, A.E. Merbach, J. Am. Chem. Soc., 118 (1996) 5265-
5
[
271.
35] D. Lieb, F.C. Friedel, M. Yawer, A. Zahl, M.M. Khusniyarov, F.W. Heinemann, I.
Ivanović-Burmazović, Inorg. Chem., 52 (2013) 222-236.
36] F.C. Friedel, D. Lieb, I. Ivanović-Burmazović, J. Inorg. Biochem., 109 (2012) 26-32.
[
An Overly Anionic Metal Coordination Environment Eliminates the T -Weighted Response
1
of Quinol-Containing MRI Contrast Agent Sensors to H O
2
2
†
†
†
‡
Tessa E. Hutchinson, Adam Bashir, Meng Yu, Ronald J. Beyers, and Christian R.
†,
Goldsmith *
†
Department of Chemistry and Biochemistry, Auburn University, Auburn, AL 36849, USA
Auburn University Magnetic Resonance Imaging Research Center, Auburn, AL 36849, USA
‡
*Corresponding author: crgoldsmith@auburn.edu
Graphical Abstract
Graphical Abstract Synopsis
A hexadentate anionic ligand with two redox-active quinol groups was synthesized and tested
as a component for a MRI contrast agent. Although the ligand binds more strongly to Mn(II)
than a previously reported ligand and can be used to prepare a Mn(II) complex that does not
2
1

JOURNAL PRE-PROOF
react with O , its Mn(II) complex does not display a T -weighted relaxivity response to H O .
2
1
2 2
An Overly Anionic Metal Coordination Environment Eliminates the T -Weighted Response
1
of Quinol-Containing MRI Contrast Agent Sensors to H O
2
2
†
†
†
‡
Tessa E. Hutchinson, Adam Bashir, Meng Yu, Ronald J. Beyers, and Christian R.
†,
Goldsmith *
†
Department of Chemistry and Biochemistry, Auburn University, Auburn, AL 36849, USA
Auburn University Magnetic Resonance Imaging Research Center, Auburn, AL 36849, USA
‡
*Corresponding author: crgoldsmith@auburn.edu
Highlights


A more anionic version of a diquinol ligand used for redox-responsive MRI contrast
agents was prepared
The ligand binds to Mn(II) approximately 50 times more strongly than a related ligand
with a lesser anionic charge


The Mn(II) complex reacts with hydrogen peroxide, but not air
Hydrogen peroxide oxidizes the metal ion in the complex, preventing a T -weighted
1
relaxivity response
2
2