Versatile Nickel(II) Scaffolds as Coordination-Induced Spin-State Switches for 19F Magnetic Resonance-Based Detection

Article abstract of DOI:10.1002/anie.202010587

¹⁹F magnetic resonance (MR) based detection coupled with well-designed inorganic systems shows promise in biological investigations. Two proof-of-concept inorganic probes that exploit a novel mechanism for ¹⁹F MR sensing based on con

Full text of DOI:10.1002/anie.202010587

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Versatile Nickel (II) Scaffolds as Coordination-Induced Spin-State
Switches for 19F Magnetic Resonance-Based Detection
Authors: Da Xie, Meng Yu, Zhu-Lin Xie, Rahul T. Kadakia, Chris
Chung, Lauren E Ohman, Kamyab Javanmardi, and Emily
Lauren Que
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202010587
Link to VoR: https://doi.org/10.1002/anie.202010587

10.1002/anie.202010587
Angewandte Chemie International Edition
RESEARCH ARTICLE
Versatile Nickel (II) Scaffolds as Coordination-Induced Spin-State
Switches for 19F Magnetic Resonance-Based Detection
Da Xie[a], Meng Yu[a], Zhu-Lin Xie[a], Rahul T. Kadakia[a], Chris Chung[a], Lauren Ohman[a], Kamyab
Javanmardi[b], Emily L. Que*[a]
Abstract: 19F magnetic resonance-based detection coupled with well-
designed inorganic systems shows great promise in biological
investigations including metabolic tracking, studies of protein
structure and function, molecular imaging, and medical diagnostics,
providing low biological background and high specificity of the
modality. Two proof-of-concept inorganic probes are reported that
exploit a novel mechanism for 19F MR sensing based on converting
from low-spin (S=0) to high-spin (S=1) Ni2+. Activation of diamagnetic
NiL1 and NiL2 via light or β-galactosidase respectively converts them
into paramagnetic NiL0 that displays a single 19F NMR peak shifted by
>35 ppm with accelerated relaxation rates. This large chemical shift
change is facilitated by contact and pseudo-contact interactions
between the 19F reporter and the 5-coordinate Ni2+. This spin-state
switch is effective for sensing light or enzyme expression in live cells
using 19F MR spectroscopy and imaging that differentiate signals
based on chemical shift and relaxation times. This general inorganic
scaffold has potential for developing agents that can sense analytes
These properties have enabled the use of 19F as a unique handle
for investigating at protein structure and function, tracking
metabolism, and developing molecular imaging agents.[6]
Combining 19F magnetic resonance spectroscopy with the rich
magnetic and chemical properties of metal complexes provides a
diverse toolbox for the development of a diverse array of
activatable probes for 19F-based biosensing.
To date, multiple metal complexes have been established as
probes that respond to specific biochemical processes through
manipulating metal-fluorine interactions.[6c] Gd3+ complexes have
been reported for sensing enzymatic activity.[7] In these probes,
the 19F MR signal is quenched due to intramolecular
paramagnetic relaxation enhancement (PRE) from the Gd3+ and
restored following enzyme-specific cleavage of the link between
the Gd3+ and the 19F reporter. Metal redox couples such as
Mn2+/Mn3+,[8] Co2+/Co3+,[9] Cu+/Cu2+,[10] and Eu2+/Eu3+[11] have also
been used; in this case 19F MR signatures (relaxation time and
chemical shift) are modulated by altering the oxidation state and
spin state of metal center in response to redox events. A third
strategy is to exploit changes in donor strength to modulate metal
ranging from ions to enzymes, opening up diverse possibilities for 19
F
MR-based biosensing.
Introduction
spin state and thus 19F chemical shifts including systems based
[12]
on Co2+
and Fe2+ [13]. A fourth design strategy is tuning the
The magnetic properties of metal complexes can be applied in a
number of contexts ranging from data storage[1] to medical
diagnostics[2]. One exciting frontier is in biosensing, where the
magnetism of metal complexes can be exploited and manipulated
to develop smart scaffolds for reporting unique physiochemical
and biochemical signatures in living systems.[3] 19F magnetic
resonance techniques including spectroscopic (19F nuclear
magnetic resonance spectroscopy, NMR/MRS) and imaging
modalities (19F magnetic resonance imaging, MRI) are promising
techniques for biosensing and medical diagnostics with excellent
tissue penetration depth and high sensing specificity.[4] 19F
exhibits excellent MR properties with sensitivity close to 1H and a
large chemical shift range over 350 ppm. Further, there is no
detectable endogenous signal in typical biological samples
including humans; 19F MR signal can be directly correlated to the
concentration of fluorinated probe, enabling quantification.[5]
coordination number and geometry of the metal center to
manipulate the magnetic properties of the complex, affording
distinct 19F MR signals. While lanthanide containing probes using
this strategy have been reported,[14] this remains largely
unexplored with transition metal complexes.
Ni2+ complexes have been investigated for the accessibility of
both high-spin and low-spin states in an aqueous environment,
dependent on the Ni2+ coordination geometry.[15] A square-planar
Ni2+ almost always adopts a diamagnetic configuration (S = 0),
while a five or six-coordinate Ni2+ favors a paramagnetic
configuration (S = 1). Though paramagnetic Ni2+ complexes have
been used as 19F MR tracers for live-cell labeling, examples of
spin-state-switching Ni2+ complexes are rare in the field of MR
biosensors and have not been reported for 19F.[16] Ni2+
azoporphyrin complexes were reported as light-induced spin-
state switches for 1H MRI based on coordination change, but poor
aqueous solubility has limited their application in biological
environments.[17]
[a]
D. Xie, Dr. M. Yu, Dr. Z. Xie, R.T. Kadakia, C. Chung, L. Ohman,
and Prof. Dr. E.L. Que
Department of Chemistry
The University of Texas at Austin
105 E. 24th St Stop A5300, Austin, Texas 78712, United States
E-mail: emilyque@cm.utexas.edu
K. Javanmardi
Department of Molecular Biosciences
The University of Texas at Austin
Herein we present a versatile Ni2+ platform utilizing a side-bridged
cyclam (1,4,8,11-tetraazacyclotetradecane) ligand that can
undergo coordination-induced spin-state switch upon activation
(Figure 1). The modularity of the sensing platform was
demonstrated by two proof-of-concept probes, NiL1 and NiL2, for
[b]
2500 Speedway, Austin, TX 78712, USA
Supporting information for this article is given via a link at the end of
the document.
sensing photoactivation and
respectively, both forming NiL0 as product. Spectroscopic, single-
β -galactosidase activity,
This article is protected by copyright. All rights reserved.

10.1002/anie.202010587
Angewandte Chemie International Edition
RESEARCH ARTICLE
Figure 1. Design strategies (left box) and the structures and sensing mechanisms (right) of spin-state switching Ni2+ complexes for 19F magnetic resonance
biosensing.
strong coordinating anions including CN-, SCN- and N3 in
aqueous solution.[19] The bis-ethylene bridge enforces planarity of
the N donors and a small metal binding pocket, both of which
favor a square planar, low-spin Ni2+. On the other hand, although
much less explored, SB-cyclam with pendant coordinating groups
can accommodate high-spin Ni2+ ions.[20] Therefore, we
envisioned to couple coordination-induced spin-state change with
MR-based sensing by appending a “caged” donating ligand to Ni2+
SB-cyclam complexes, where the caging group will be removed
after probe activation by a specific analyte (Figure 1). We chose
a “caged” phenol due to the drastic difference of donating strength
between an ether group and a phenolate. Moreover, there is
literature precedent for high-spin phenolate Ni2+ cyclam
complexes.[21] For a readily tractable proof-of-concept probe, we
protected the phenol with a photo-labile ortho-nitrobenzyl group
to afford NiL1, which will undergo bond cleavage following UV-
irradiation to release phenolate and form NiL0. We also prepared
NiL2 that responds to cleavage by β-galactosidase via a self-
immolative process to also generate NiL0. Both probes were
successfully applied in live cells, opening the possibility of imaging
dynamic biochemical processes in cells and beyond. This probe
platform can be readily adapted to detect any analyte that can
induce a specific cleavage event that transforms the coordination
properties of the ligand scaffold.
-
crystal, and computational studies revealed that NiL1 and NiL2 are
square planar and diamagnetic, while NiL0 is five- coordinate and
paramagnetic. The large difference in relaxation times and
chemical shift between paramagnetic (NiL0) and diamagnetic
species (NiL1 and NiL2) allowed selective 19F MR imaging of
individual complexes in different magnetic states. The spin-
flipping process was successfully assessed in solution and in
cellular environments via 19F MR-based techniques, thus
highlighting Ni2+ side-bridged cyclam scaffolds as reliable and
flexible platforms to develop responsive 19F MR sensors.
Results and Discussion
Probe design and synthesis
To develop Ni2+ complexes as 19F MR probes with switchable spin
states based on coordination geometry, two requirements must
be taken into account: (1) in the low-spin square-planar Ni2+
complex, access of surrounding anions or solvent molecules to
Ni2+ center from the axial position should be minimized; otherwise
it may be easily converted to the high-spin complex in biological
media. (2) After probe activation, axial coordination of an
intramolecular donating ligand should be strong enough to
generate a high-spin Ni2+ complex in order to compensate the
energetic penalty associated with spin-state change. Donating
ligands with less steric bulk and slow ligand exchange kinetics are
favored.
Briefly, the syntheses of NiL1 and NiL2 were carried out starting
from 5-trifluoromethyl-salicylaldehyde. “Activatable” groups
(ortho-nitrobenzyl for NiL1; nitrophenyl-β-galactoside for NiL2)
were attached to the phenol via SN2 reactions with nearly
quantitative yield (Scheme S1-S3). The formed benzaldehyde
was then converted to the corresponding benzyl bromide in two
steps with an overall yield of 40-50%. Following a classical
synthetic procedure, protected cyclam was conjugated with the
We chose a cyclam-based scaffold for our probes given its strong
affinity for 3d transition metals and the broad application of cyclam
metal complexes in biological settings.[18] The specific probe
design was inspired by past studies on Ni2+ side-bridged-cyclam
(SB-cyclam) complexes that are diamagnetic even in presence of
This article is protected by copyright. All rights reserved.

10.1002/anie.202010587
Angewandte Chemie International Edition
RESEARCH ARTICLE
above prepared benzyl bromide to furnish the ligand precursor as
a bromide salt (64-87% yield), followed by reduction via NaBH4 to
produce the ligands L1 and L2 in 74% and 79% yield, respectively.
The Ni2+ complexes were formed by heating a methanolic solution
of the ligand and nickel acetate, with 87% yield for NiL1 and 84%
yield for NiL2. Perchlorate analogues were also synthesized,
however these complexes were less soluble and thus were not
the main focus of this study. To independently investigate the
formed product, we also prepared NiL0 and its cyclam analogue,
NiL3, using synthetic route similar to NiL1 and NiL2. Full synthetic
procedures and characterizations are provided in the supporting
information.
an approximate intensity ratio of 1:1 (Figure S2). Given that
ligand L1 only displayed one singlet 19F peak (-63.04 ppm with
2.30 Hz FWHM, Figure S3), we attribute these peaks to the
presence of diastereomers, with both Ni2+ and donating NBn
serving as stereogenic centers. A variable-temperature 19F NMR
experiment was performed to study the exchange of the two
isomers and revealed only small changes in peak separation from
18.84 Hz at 8 ºC to 14.28 Hz at 80 ºC (Figure S2). This
observation indicates an exchange rate of the two isomers much
smaller than 20 s-1 as indicated by the NMR spectrum across all
temperature. Due to vicinity of the two isomeric peaks, the 19F MR
relaxation times were measured by treating the peaks as one,
giving a T1 of 740 ms and a T2 of 240 ms (Table S1). Both the
longitudinal and transverse relaxation time are consistent with a
diamagnetic Ni2+ complex.
Solution behavior of NiL1
All prepared Ni2+ complexes exhibited excellent aqueous solubility
up to 100 mM. The UV-vis absorption spectrum of NiL1 in HEPES
buffer (Figure S1) contained an absorption band centered at 480
nm (ε = 2.7x102 M-1cm-1), which was assigned to a spin-allowed
d-d transition. This absorption feature matched with a previously
reported square planar Ni2+ SB-cyclam complex,[20] supporting the
presence of a square-planar low-spin Ni2+ complex in aqueous
solution. We note that the extinction coefficient of NiL1 is slightly
higher than in perfectly square planar Ni2+ complexes,[22]
potentially indicative of a slight tetrahedral distortion of the
cyclam-N4 donors. Magnetic moment measurement of NiL1
determined via Evan’s method confirmed the largely diamagnetic
nature of the complex in the same buffer (0.51 μB). The slight
paramagnetism is likely due to the unfavored axial access of
ligands (e.g. water and anions) to the Ni2+.
Photo-irradiation of NiL1
To investigate NiL1 as a spin-state switching agent, we irradiated
an aqueous buffered solution of NiL1 at a light power of 3.5
mW/cm2 and tracked the reaction progress via 19F NMR. Prior to
irradiation, only peaks from NiL1 were seen at -61.53 and -61.57
ppm (Figure 2A). Continuous irradiation resulted in generation of
a new species with a characteristic broad peak at -24.76 ppm
(FWHM = 22.7 Hz) and disappearance of the peaks from NiL1.
Both peaks of NiL1 declined simultaneously, consistent with them
being diastereomers with the same chemical reactivity. The newly
generated species was identified as NiL0 according to its 19F NMR
spectrum as well as high-resolution mass spectroscopy (Figure
2A and S4). T1 and T2 were measured for the newly emerged
peak and both fall on the scale of tens of milliseconds (T1 = 33 ms
and T2 = 15 ms), indicating the conversion from the diamagnetic
NiL1 to the
Interestingly, the 19F NMR spectrum of NiL1 showed two adjacent
singlet peaks (-61.53 and -61.57 ppm) separated by 18 Hz with
Figure 2. Characterization of the photochemical conversion of NiL1 to NiL0 in HEPES (50 mM, pH 7.4) at room temperature, assessed by (A) 19F NMR (3.0 mM
NiL1), (B) magnetic moment measurement (5.0 mM NiL1), (C) and (D) UV-vis absorption (1.0 mM NiL1), (E) phantom 19F MRI (4.0 mM NiL1; upper panel, pulse
sequence selected for NiL1; lower panel, pulse sequence selected for NiL0), and (F) phantom 19F MRI signal-to-noise ratio analysis. NiL1 was irradiated by hand-
held UV lamp at a power of 3.5 mW/cm2 for an hour in total.
This article is protected by copyright. All rights reserved.

10.1002/anie.202010587
Angewandte Chemie International Edition
RESEARCH ARTICLE
paramagnetic NiL0. To further demonstrate the spin-state
switching process, the magnetic moment after different durations
of light-irradiation was assessed via Evans’ method (Figure 2B).
A gradual increase in magnetic moment was observed from 0.51
µB to 2.91 µB over an hour, with the final value approaching the
magnetic moment of NiL0 (2.96 µB). The paramagnetism of Ni2+
in NiL0 was further validated by EPR spectroscopy in frozen
aqueous solution; the EPR spectrum displayed a broad peak with
a g factor of 1.99 (Figure S5).
slow evaporation of its solution in non-coordinating solvents
(dichloroethane/pentene). As shown in Figure 3A, the Ni2+ center
is five-coordinate with Ni-N(cyclam) bond lengths ranging from
2.014 Å to 2.085 Å and Ni-O (phenolate) bond distance of 1.999
Å, indicative of a high-spin Ni2+ complex. The strong donating
phenolate caused the cyclam to adopt an unusual distorted trans-
II configuration. While the acetate counterion participates in an
extensive H-bonding network when NiL0 is in the solid state, no
sixth ligand (e.g. acetate, H2O) is observed in the primary
coordination sphere of the metal center, consistent with a less
flexible SB-Cyclam ligand where only intramolecular coordination
to the Ni2+ center is allowed. To evaluate the coordination
geometry of the Ni2+ center, Addison and Reedijk’s τ5 parameter
was determined. An intermediate τ parameter of 0.51 indicated a
structure that is right in-between square pyramidal and trigonal
bipyramidal. This structural result is partly due to the presence of
the chelating phenolate that “lifts” both the donating NBn and the
Ni2+ center up from the average macrocyclic plane as defined by
the other three SB-cyclam nitrogen donors (by 1.116 Å and 0.697
Å, respectively). Furthermore, the average distance between the
fluorine atoms and Ni2+ center was 7.2 Å and assumed to be
similar in solution due to the rigidity of the Ni-phenolate
interactions. This distance falls within the effective range for PRE
to result in shortened relaxation times of fluorine atoms.[24]
The photochemical reaction kinetics were further evaluated
through UV-vis absorption and 19F NMR. During light irradiation,
the characteristic absorbance of NiL1 at 480 nm in the UV-vis
spectrum gradually decreased, with simultaneous increase in a
shoulder at 378 nm and a broad peak at 620 nm (Figure 2C).
Plotting the absorption at 480 nm against irradiation time exhibited
a first-order decay with a half-life of 9.4 minutes (Figure 2D),
consistent with previously reported reaction kinetics for photo-
cleavage of ortho-nitrobenzyl-type groups.[23] Because only one
event is observed, it is likely that upon cleavage of the nitrobenzyl
moiety, the subsequent intramolecular coordination between
phenolate and Ni2+ is too rapid to capture as a separate event by
UV-vis spectroscopy. 19F NMR-based quantification of the decay
process of NiL1 closely mirrored the UV-vis absorption decay of
NiL1, affording the same half-life (9.5 minutes, Figure S6).
To correlate the solution structure of NiL0 with its solid-state
structure, we measured the phenol pKa of NiL0 to be 5.33 ± 0.01
in aqueous buffer using UV-vis (Figure S8). The acidic pKa value
indicates that in biologically relevant media at neutral pH, the
phenol moiety will be largely deprotonated and will coordinate to
Ni2+. Further, the proton relaxivity r1 measured in aqueous solution
of NiL0 was 0.020 ± 0.002 mM-1 s-1, similar to octahedral Ni2+
complexes with no vacant site for water coordination.[25] In
comparison, NiL3, the cyclam analogue of NiL0 in which there is
no side-bridge, can coordinate water to form a 6-coordinate
species[26] and displayed a proton relaxivity of 0.13 ± 0.01 mM-1 s-
1, 6.5x larger than NiL0. These results support a five-coordinate
structure of NiL0 in aqueous solution with no bound H2O,
facilitated by the presence of the cyclam side-bridge. The defined
solution structure of NiL0, as suggested by its single 19F NMR
peak, brings a significant advantage over chemical-shift based
lanthanide probes whose MR signals are diluted by the presence
of multiple isomers in aqueous solution.[27]
To evidence the application of NiL1 as a probe for 19F MRI, we
performed phantom imaging of an irradiated sample of NiL1
(Figure 2E). Given the distinction of both chemical shift and
relaxation time between NiL1 and its irradiation product, NiL0, we
employed two different pulse sequences to selectively image NiL1
and NiL0 individually. A Rapid Acquisition with Relaxation
Enhancement (RARE) pulse sequence with long echo time (TE =
40 ms) and long repetition time (TR = 1000 ms) was optimized for
NiL1, while NiL0 was best imaged with a RARE sequence with
short echo time (TE = 2.0 ms) and short repetition time (TR = 55
ms). Consequently, there was zero cross-talk between the 19F MR
images of the two complexes. The initial signal for NiL1 (signal-to-
noise ratio, SNR = 49.5) diminished over the course of irradiation
(Figure 2F). Conversely, the signal intensity for NiL0 steadily
increased, plateauing at SNR ~150 after irradiation for 60 minutes
(Figure 2F). The higher SNR after irradiation points to the higher
sensitivity of NiL0 compared to NiL1 as the shorter T1 of NiL0
allows for increased acquisition replicates within a fixed imaging
time. To quantify the sensitivity difference, we measured the limit-
of-detection (LOD) of NiL1 and NiL0 (Figure S7), yielding values
of 0.93 mM and 0.20 mM respectively in a ten-minute scan. This
is consistent with reported LODs for diamagnetic and
paramagnetic 19F MR probes.[9a, 22a]
We performed DFT calculations on a level of B3LYP/6-31G(d,p)
to understand the electronic structure of NiL0. The optimized
structure of NiL0 well mimics its X-ray structure; superimposed
structures are shown in Figure S9, with τ5 = 0.53 for the optimized
structure. The calculated d-orbital splitting (Figure 3B) indicates
that the strong coordination between the phenolate and Ni2+
!
raises the energy levels of the d!" and d# orbitals close to the
Structural insights for NiL0 and NiL1
!
!
d!
resulting in a high-spin Ni2+ center (S = 1). In this
$"
To gain a deeper understanding of the spin-state difference
calculation, the difference in the energy of the filled d!" and a
between NiL0 and NiL1, single crystals of NiL0 were obtained by
This article is protected by copyright. All rights reserved.

10.1002/anie.202010587
Angewandte Chemie International Edition
RESEARCH ARTICLE
Figure 3. (A) ORTEP representation of NiL0 from a square pyramidal view. Thermal ellipsoid plot at 50% probability level. Solvent molecules, counter anion (acetate),
and hydrogen atoms are omitted for clarification. (B) Calculated d-orbital diagram and d8-electron fillings of NiL0. (C) Optimized structure of NiL1 by DFT calculations.
Bond lengths of Ni-Ncyclam are 1.97, 1.94, 1.95, and 1.95 Å. (D) Calculated d-orbital diagram and d8-electron fillings NiL1.
!
!
half-filled d!
orbital is small (0.13 eV), indicating there is a
unchanged (λmax = 452 nm for NiL1 and λmax = 381, 472, and 616
nm for NiL0), indicating little to no trans-metalation occurred.
Further, this indicated that even in the presence of excess Cl- or
$"
low-lying excited state that could couple with the ground state to
generate large zero-field splitting (D) and introduce high magnetic
anisotropy.[28] High-spin Ni2+ coordinated in trigonal-bipyramid
geometry has been shown to induce large negative D values in
design of single-molecular magnets.[29] Compared to the chemical
shifts of L0 (-60.69 ppm) and NiL1 (-61.01 ppm), NiL0 displayed a
peak at -24.76 ppm, a downfield shift by more than 35 ppm
(Figure S10). Moreover, the diamagnetic Zn2+ analogue of NiL0
(ZnL0, -60.45 ppm) only showed a 0.24 ppm downfield shift
comparing to the ligand L0 (Figure S10), excluding the effect of
the metal Lewis acidity on the 19F chemical shift of the -CF3 group.
These observations emphasize the pivotal role of Ni2+-phenolate
coordination in inducing magnetic anisotropy of the paramagnetic
Ni2+ center and strengthening through-bond contact between Ni2+
and 19F. Both the contact shift and pseudocontact shift contribute
to the large downfield shift of the 19F NMR resonance in NiL0.
NO3 , the primary coordination sphere of the Ni2+ did not change
-
in either complex. While [Ni(cyclam)]2+ is well known for its
extreme kinetic stability,[30] the double ethylene bridge present in
the ligand framework likely reduces the flexibility of macrocycle
and further enhances the inertness of the complexes.
As a step towards bio-application of this coordination-based
strategy, we studied the photochemical conversion of NiL1 to NiL0
in the context of live cells. HeLa cells were incubated with NiL1,
cell pellets were collected and 19F NMR and MRI were taken
before and after light irradiation (λ = 365 nm). Interestingly, before
irradiation, the 19F NMR spectrum was silent (Figure 4). The
disappearance of the NiL1 signal is likely caused by the
interaction between the hydrophobic ortho-nitrophenyl group in
the complex and cellular proteins. Indeed, the 19F NMR spectrum
of NiL1 in the presence of fetal bovine serum displayed a much
broader and featureless peak (FWHM = 123 Hz, Figure S12)
compared to the peaks seen in aqueous buffer (FWHM = 5.70,
5.38 Hz). Signal quenching (significant T2 decrease) because of
hydrophobic interactions has been reported for analogous metal-
cyclen complexes.[31] After light irradiation, a strong signal was
observed at –24.89 ppm (FWHM = 25.38 Hz), corresponding to
the expected formation of NiL0, which was sharper and better
mimicked the peak observed in aqueous HEPES buffer in this
case (FWHM = 32.2 Hz in FBS versus FWHM = 21.2 Hz in
HEPES, Figure S12). 19F MR phantom imaging was performed
by selectively observing the signal from NiL0 due to its lower
detection limit and less hydrophobic-interaction-induced signal
loss. As shown in Figure 4, only background signal was detected
for the cell pellets before irradiation (SNR 2.6, 30-minute scan).
However, after one-hour light exposure, a moderate signal could
be observed with an SNR of 6.5. These imaging results
The DFT-optimized structure of NiL1 clearly showed shorter bond
lengths of Ni2+-N(cyclam) (c.a. 1.94 Å to 1.97 Å), characteristic of
a diamagnetic Ni2+ center (Figure 3C). The phenol-ether oxygen
atom lies 3.75 Å away from Ni2+ that precludes its coordination
and therefore, Ni2+ is best described as a distorted square-planar
coordination geometry. Correspondingly, the d-orbital splitting
!
!
diagram of NiL1 exhibits a high-lying unfilled d!
orbital that is
$"
!
well above the closest d# orbital (separated by 4.56 eV) and,
consequently, all the d-electrons are paired, resulting in a
diamagnetic NiL1 (Figure 3D).
Photo-conversion of NiL1 to NiL0 in cell culture
Prior to cellular testing, we monitored the stability of NiL1 and NiL0
in presence of bio-available metal ions (100 mM KNO3, 100 mM
NaCl, 10 mM MgCl2, 10 mM CaCl2, and 10 mM ZnCl2) using UV-
vis absorption (Figure S11). After 24-hour incubation, the
characteristic d-d bands of both complexes remained nearly
This article is protected by copyright. All rights reserved.

10.1002/anie.202010587
Angewandte Chemie International Edition
RESEARCH ARTICLE
demonstrate the potential application of Ni2+ SB-Cyclam-based
complexes for 19F MR sensing in complex cellular environments.
800-second period (Figure S14). At different NiL2 concentrations,
all the absorption increases displayed a first-order exponential
increase. The enzyme kinetic parameters were further calculated
according to Lineweaver-Burk reciprocal plot, which returned a
Michaelis constant Km of (9.7 ± 1.4) x102 µM and a catalytic rate
kcat of 15 ± 2 s-1 (Figure 5B). These results fall on the same order
but suggest slower kinetics compared to published results[33] on
In vitro enzymatic cleavage of NiL2
The promising results of implementing a spin-state switching
imaging agent in cells using highly controlled chemistry alluded to
the potential of adapting this platform to detect biological activity,
such as the activity of enzymes. As an example, we prepared NiL2
that incorporates a galactose moiety for sensing the activity of β-
galactosidase (β-Gal), an exoglycosidase that hydrolyzes the β-
glycosidic bond. β-Gal was chosen because it has been widely
engineered as a co-expressed reporter protein for human-made
constructs in mammalian cells and in organisms and has been
used in cellular and in vivo imaging.[32] In our study, we designed
NiL2 to undergo a self-immolative process in response to β-Gal to
generate NiL0 for 19F MR-based sensing.
o-nitrophenyl β-D-galactoside,
a
standard substrate for
spectroscopic detection of β-Gal activity, due partly to the larger
size of the NiL2 complex. Regardless, NiL2 appeared to be an
efficient substrate for β-Gal.
To demonstrate the use of NiL2 for probing β-Gal activity by 19F
MRI, we cultured NiL2 with regular and heat-inactivated β-Gal
under the same conditions used for NMR experiments (Figure
5C). As expected, NiL2 in absence of β-Gal or in presence of heat-
inactivated β-Gal did not show any conversion into NiL0. When
NiL2 is mixed with active β-Gal, a time-course phantom 19F MRI
with 5-minute interval well mimicked the progress of β-Gal-
catalyzed cleavage as seen by 19F NMR (Figure S15), with signal
enhancement for NiL0 over a period of 40 minutes. While the
signal for NiL2 almost faded away after 20 minutes of incubation,
NiL0 signal increases were still observed, consistent with the
higher sensitivity of the paramagnetic NiL0 compared to the
diamagnetic NiL2.
Detection of β-Gal enzymatic activity at a cellular level
We further studied the capability of NiL2 in detecting cellular β-
Gal activity via genetically encoded expression of β-Gal in HEK
293T cells using the immediate-early promoter from
cytomegalovirus (pcDNA3.1/His/LacZ).[32b] NiL2 was then co-
cultured with HEK 293T cells for two hours, and the media was
collected and subjected to 19F NMR/MRI analysis. As shown in
Figure 5D, 19F NMR clearly indicated the conversion of NiL2, with
a characteristic singlet peak at -24.78 ppm assigned to NiL0. At a
cell population of ~1.0 million and a NiL2 dosage of 1.0 µmol, cell
uptake of NiL2 was determined to be 1.5 ± 0.1 fmol/cell and the
conversion of NiL2 was 70 ± 10 %, a sign of efficient cell uptake,
enzymatic transformation, and trans-membrane diffusion of the
cleaved product (NiL0). On the other hand, when NiL2 was
incubated with cells not expressing β-Gal, only slight conversion
(c.a. 8.0 ± 0.5 %) was noticed. Similar result was observed when
NiL2 itself was incubated in cell culture media under the same
conditions. To explain this “background” reactivity, we further
tested the stability of NiL2 in presence of glutathione (reduced
type). The moderate reactivity of NiL2 as observed (9.9 %
conversion of NiL2, Figure S16) indicates that free thiol-groups
could cleave the galactose side chain and compromise the
stability of NiL2 in culture media. We note, however, due to the
stochiometric nature and slow kinetics of the reaction between
NiL2 and thiol groups, our probe is still favorable for probing the
catalytic activity of β-Gal in biological systems, as seen by around
Figure 4. Detection of photoactivation of NiL1 in HeLa cells through 19F NMR
and 19F MRI. (Left) 19F NMR spectra at 9.4 T of HeLa cell pellets labeled with 1
mM NiL1 before (bottom spectrum) and after (top spectrum) photo-irradiation.
(Right) 19F MRI imaging on a 7T scanner of NiL0 in HeLa cell pellets before and
after photo-irradiation.
Similar to NiL1, NiL2 is diamagnetic in aqueous buffer, displaying
a sharp singlet peak with long 19F relaxation times (T1 = 790 ms,
T2 = 320 ms), and a d-d transition band at 480 nm with ε = 2.8x102
M-1cm-1 in its UV-vis absorption spectrum (Figure 5A and S1). In
the presence of β-Gal, the NMR peak for NiL2 at
-61.42 ppm
declined while the peak for NiL0 at -24.76 ppm gradually
increased. The efficient conversion of NiL2 to NiL0 via β-Gal was
confirmed by a test trial between 1.5 μmol NiL2 and 3 U enzyme,
where >95% conversion was observed after 40 minutes (Figure
5A). No cleavage was found when NiL2 was incubated with heat-
inactivated β-Gal (Figure S13).
Along with the formation of NiL0, enzymatic cleavage of NiL2
produced an intense yellow color, corresponding to the absorption
of 4-(hydroxymethyl)-2-nitrophenate motif. Its high molar
absorptivity (~103) allows analysis of β-Gal enzyme kinetics
through monitoring the absorption increase at 403 nm. A fixed
amount of β-Gal (1.5 U) was incubated with a buffered solution of
NiL2 at 37 ºC and absorption at 403 nm was recorded over an
8-fold
enhancement
of
NiL2
This article is protected by copyright. All rights reserved.

10.1002/anie.202010587
Angewandte Chemie International Edition
RESEARCH ARTICLE
Figure 5. (A) Detection of in vitro β-galactosidase activity by NiL2 through 19F NMR. (B) Lineweaver-Burk reciprocal plot for determination of enzyme kinetics.
Michaelis constant Km and kcat were found to be (9.7 ± 1.4) x102 μM and 15 ± 2 s-1, respectively. (C) Phantom 19F MRI of in vitro β-Gal activity at 7 T. Pulse sequences
were selected for NiL2 and NiL0, respectively. (D) Detection of β-Gal activity in HEK 293T cells through 19F NMR and 19F MRI using NiL2. A RARE pulse sequence
selective for NiL0 was applied for 19F MRI.
conversion in presence versus in absence of cellular β-Gal. 19F
MRI was further conducted to demonstrate the use NiL2 in
sensing activity of cellular β-Gal (Figure 5D). NiL0 was selectively
imaged using a RARE pulse sequence due to its lower limit-of-
detection. Without encoded expression of β-Gal, the low
conversion of NiL2 could not be visualized and the signal of NiL0
was on the level of noise (SNR ~ 3). When β-Gal was present, a
strong signal was tracked, giving an SNR of 26.6. The sharp
contrast between images well correlated with the observation in
19F NMR and emphasized the viability of using 19F MR imaging to
study cellular activities in a “turn-on” manner via tuning the spin-
state of a Ni2+ center.
efforts include developing probes that have higher SNR, require
shorter scan times, and exhibit improved pharmacokinetics in
order to better facilitate in vivo applications. This represents a new
platform for monitoring biochemical events that should be readily
adapted for the detection of a range of analytes in biological
contexts including small molecules and enzymes that are of
interest to medical diagnostics.
Experimental Section
Photo-irradiation assay of NiL1
NiL1 was dissolved in HEPES buffer (pH 7.4, 50 mM, added 100 mM
KNO3) and irradiated from the side by hand-held UV lamp at 365 nm at a
power of 3.5 mW/cm2. Spectroscopic measurement (including UV-vis
absorption, 1H NMR, and 19F NMR) was performed at following indicated
time points within an overall photo-irradiation time of an hour: UV-vis
absorption spectrum: 0, 1, 2, 3, 5, 7, 10, 13, 16, 20, 25, 30, 35, 40, 50, and
60 min; 1H NMR spectrum: 0, 1, 3, 6, 10, 20, 30, 60 min; 19F NMR
spectrum: 0, 1, 3, 6, 10, 15, 20, 30, 40, 60 min.
Conclusion
Taken together, we have established Ni2+ SB-Cyclam as a
versatile spin-state switching platform for designing aqueous 19
F
MR probes for both in vitro and cellular applications. Our
mechanism relies on distinct changes in the Ni2+ coordination
environment that trigger changes in the electronic structure of the
metal center, converting it from a low-spin S=0 species to a high
spin S=1 species with low-lying excited electronic states. This
change in coordination also modulates the through bond
interactions between the metal and the 19F reporter. Importantly,
the product paramagnetic Ni-phenolate species exhibits large
changes in 19F chemical shift relative to the initial probe, resulting
from modulation of both contact and pseudocontact shift effects.
The distinction in magnetic properties between high-spin and low-
spin fluorinated Ni2+ complexes allows NMR/MRI tracking of both
species separately, making it possible to monitor reactions with
this complex in biological contexts. The probes reported herein
are limited, however, due to their low 19F content that will limit the
ability to detect their signal in an efficient manner in vivo. Ongoing
Enzymatic assay of NiL2
NiL2 was dissolved in HEPES buffer (50 mM, pH 7.4, added 10 mM MgCl2
and 100 mM KCl) and was cultured with b-galactosidase at an activity
dosage of 2U/µmol NiL2. 19F NMR spectrum was taken at 0, 5, 7.5, 10, 15,
20, 30 and 40-min time point to evaluate the enzymatic conversion of NiL2
to NiL0.
To assess the enzyme kinetics, a series of parallel enzymatic reactions
were performed at 310 K with NiL2 concentration from 25 μM to 250 μM
and a fixed β-gal activity of 1.5 U. The reaction progress was tracked by
UV-vis absorption at 403 nm for a duration of 800 seconds, and this time-
dependent absorption enhancement was fitted to first-order kinetics for
estimation of the initial reaction rate v0. Michaelis constant Km and catalytic
This article is protected by copyright. All rights reserved.

10.1002/anie.202010587
Angewandte Chemie International Edition
RESEARCH ARTICLE
rate kcat were further calculated according to Lineweaver-Burk reciprocal
plot3 (1/v0 versus 1/[NiL2]).
[14] a) K. H. Chalmers, E. De Luca, N. H. M. Hogg, A. M. Kenwright, I. Kuprov,
D. Parker, M. Botta, J. I. Wilson, A. M. Blamire, Chem. Eur. J. 2010, 16,
134-148; b) A. M. Kenwright, I. Kuprov, E. De Luca, D. Parker, S. U.
Pandya, P. K. Senanayake, D. G. Smith, Chem. Commun. 2008, 2514-
2516; c) M. Yu, D. Xie, R. T. Kadakia, W. Wang, E. L. Que, Chem
Commun (Camb) 2020, 56, 6257-6260; d) G. Gambino, T. Gambino, R.
Pohmann, G. Angelovski, Chem Commun (Camb) 2020, 56, 3492-3495.
[15] F. Meyer, H. Kozlowski, in Comprehensive Coordination Chemistry II
(Eds.: J. A. McCleverty, T. J. Meyer), Pergamon, Oxford, 2003, pp. 247-
554.
Acknowledgements
This research was supported by funding from the Welch
Foundation (F-1883) and the National Science Foundation (CHE-
1945401). We gratefully acknowledge Prof. Andrew Ellington and
Shaunak Kar for helpful discussions regarding cellular enzyme
expression experiments. MR phantoms were acquired at the
Biomedical Imaging Center at the Dell Medical School at UT
Austin.
[16] a) J. Blahut, K. Bernášek, A. Gálisová, V. Herynek, I. Císařová, J. Kotek,
J. Lang, S. Matějková, P. Hermann, Inorg. Chem. 2017, 56, 13337-
13348; b) J. Blahut, P. Hermann, A. Gálisová, V. Herynek, I. Císařová,
Z. Tošner, J. Kotek, Dalton Trans. 2016, 45, 474-478.
[17] a) M. Dommaschk, M. Peters, F. Gutzeit, C. Schütt, C. Näther, F. D.
Sönnichsen, S. Tiwari, C. Riedel, S. Boretius, R. Herges, J. Am. Chem.
Soc. 2015, 137, 7552-7555; b) S. Thies, H. Sell, C. Schütt, C. Bornholdt,
C. Näther, F. Tuczek, R. Herges, J. Am. Chem. Soc. 2011, 133, 16243-
16250.
Keywords: nickel •fluorine • magnetic resonance spectroscopy
and imaging • coordination chemistry • biosensing
[18] X. Liang, P. J. Sadler, Chem Soc Rev 2004, 33, 246-266.
[19] K. P. Wainwright, Inorg. Chem. 1980, 19, 1396-1398.
[20] L. Fabbrizzi, F. Foti, M. Licchelli, A. Poggi, Inorg. Chem. 2002, 41, 4612-
4614.
[1]
[2]
a) D. N. Woodruff, R. E. P. Winpenny, R. A. Layfield, Chem. Rev. 2013,
113, 5110-5148; b) G. A. Craig, M. Murrie, Chem. Soc. Rev. 2015, 44,
2135-2147.
a) R. B. Lauffer, Chem. Rev. 1987, 87, 901-927; b) J. Wahsner, E. M.
Gale, A. Rodríguez-Rodríguez, P. Caravan, Chem. Rev. 2019, 119, 957-
1057; c) A. Gupta, P. Caravan, W. S. Price, C. Platas-Iglesias, E. M. Gale,
Inorg Chem 2020, 59, 6648-6678.
[21] E. Kimura, T. Koike, K. Uenishi, M. Hediger, M. Kuramoto, S. Joko, Y.
Arai, M. Kodama, Y. Iitaka, Inorg. Chem. 1987, 26, 2975-2983.
[22] a) D. Xie, L. E. Ohman, E. L. Que, Magn. Reson. Mater. Phys., Biol. Med.
2019, 32, 89-96; b) E. Kimura, T. Koike, H. Nada, Y. Iitaka, Inorg. Chem.
1988, 27, 1036-1040.
[3]
a) G. Angelovski, Angew. Chem. Int. Ed. Engl. 2016, 55, 7038-7046; b)
M. C. Heffern, L. M. Matosziuk, T. J. Meade, Chem. Rev. 2014, 114,
4496-4539; c) E. L. Que, C. J. Chang, Chem. Soc. Rev. 2010, 39, 51-60;
d) L. M. De Leon-Rodriguez, A. J. M. Lubag, C. R. Malloy, G. V. Martinez,
R. J. Gillies, A. D. Sherry, Acc. Chem. Res. 2009, 42, 948-957.
J. Ruiz-Cabello, B. P. Barnett, P. A. Bottomley, J. W. M. Bulte, NMR
Biomed. 2011, 24, 114-129.
[23] P. Klán, T. Šolomek, C. G. Bochet, A. Blanc, R. Givens, M. Rubina, V.
Popik, A. Kostikov, J. Wirz, Chem. Rev. 2013, 113, 119-191.
[24] I. Bertini, C. Luchinat, G. Parigi, E. Ravera, in NMR of Paramagnetic
Molecules (Second Edition) (Eds.: I. Bertini, C. Luchinat, G. Parigi, E.
Ravera), Elsevier, Boston, 2017, pp. 175-253.
[4]
[5]
[6]
[25] A. O. Olatunde, S. J. Dorazio, J. A. Spernyak, J. R. Morrow, J. Am. Chem.
Soc. 2012, 134, 18503-18505.
I. Tirotta, V. Dichiarante, C. Pigliacelli, G. Cavallo, G. Terraneo, F. B.
Bombelli, P. Metrangolo, G. Resnati, Chem. Rev. 2015, 115, 1106-1129.
a) H. Chen, S. Viel, F. Ziarelli, L. Peng, Chem. Soc. Rev. 2013, 42, 7971-
7982; b) S. L. Cobb, C. D. Murphy, J. Fluorine Chem. 2009, 130, 132-
143; c) K. L. Peterson, K. Srivastava, V. C. Pierre, Front. Chem. 2018, 6;
d) E. N. G. Marsh, Y. Suzuki, ACS Chem. Biol. 2014, 9, 1242-1250; e) D.
Xie, M. Yu, R. T. Kadakia, E. L. Que, Acc Chem Res 2020, 53, 2-10.
a) K. Akazawa, F. Sugihara, T. Nakamura, S. Mizukami, K. Kikuchi,
Bioconjugate Chemistry 2018, 29, 1720-1728; b) S. Mizukami, R.
Takikawa, F. Sugihara, M. Shirakawa, K. Kikuchi, Angew. Chem. Int. Ed.
Engl. 2009, 48, 3641-3643; c) S. Mizukami, R. Takikawa, F. Sugihara, Y.
Hori, H. Tochio, M. Wälchli, M. Shirakawa, K. Kikuchi, J. Am. Chem. Soc.
2008, 130, 794-795.
[26] N. Sengottuvelan, D. Saravanakumar, M. Kandaswamy, Inorg. Chem.
Comm. 2005, 8, 297-300.
[27] A. C. Harnden, D. Parker, N. J. Rogers, Coord. Chem. Rev. 2019, 383,
30-42.
[28] S. Gomez-Coca, E. Cremades, N. Aliaga-Alcalde, E. Ruiz, J. Am. Chem.
Soc. 2013, 135, 7010-7018.
[7]
[29] K. E. R. Marriott, L. Bhaskaran, C. Wilson, M. Medarde, S. T. Ochsenbein,
S. Hill, M. Murrie, Chem. Sci. 2015, 6, 6823-6828.
[30] E. J. Billo, Inorg. Chem. 1984, 23, 236-238.
[31] K. Srivastava, E. A. Weitz, K. L. Peterson, M. Marjańska, V. C. Pierre,
Inorg. Chem. 2017, 56, 1546-1557.
[32] a) J.-X. Yu, V. D. Kodibagkar, R. R. Hallac, L. Liu, R. P. Mason, Bioconj.
Chem. 2012, 23, 596-603; b) S. Mizukami, H. Matsushita, R. Takikawa,
F. Sugihara, M. Shirakawa, K. Kikuchi, Chem. Sci. 2011, 2, 1151-1155;
c) J. L. Major, T. J. Meade, Acc. Chem. Res. 2009, 42, 893-903.
[33] S. Jambovane, E. C. Duin, S.-K. Kim, J. W. Hong, Anal. Chem. 2009, 81,
3239-3245.
[8]
[9]
H. Chen, X. Tang, X. Gong, D. Chen, A. Li, C. Sun, H. Lin, J. Gao, Chem
Commun (Camb) 2020, 56, 4106-4109.
a) M. Yu, B. S. Bouley, D. Xie, J. S. Enriquez, E. L. Que, J. Am. Chem.
Soc. 2018, 140, 10546-10552; b) M. Yu, D. Xie, K. P. Phan, J. S.
Enriquez, J. J. Luci, E. L. Que, Chem. Commun. 2016, 52, 13885-13888.
[10] a) R. T. Kadakia, D. Xie, D. Martinez, M. Yu, E. L. Que, Chem. Commun.
2019, 55, 8860-8863; b) J. S. Enriquez, M. Yu, B. S. Bouley, D. Xie, E.
L. Que, Dalton Trans. 2018, 47, 15024-15030; c) D. Xie, S. Kim, V. Kohli,
A. Banerjee, M. Yu, J. S. Enriquez, J. J. Luci, E. L. Que, Inorg. Chem.
2017, 56, 6429-6437; d) D. Xie, T. L. King, A. Banerjee, V. Kohli, E. L.
Que, J. Am. Chem. Soc. 2016, 138, 2937-2940; e) R. T. Kadakia, D. Xie,
H. Guo, B. Bouley, M. Yu, E. L. Que, Dalton Trans 2020.
[11] L. A. Basal, M. D. Bailey, J. Romero, Meser M. Ali, L. Kurenbekova, J.
Yustein, R. G. Pautler, M. J. Allen, Chem. Sci. 2017, 8, 8345-8350.
[12] V. V. Smirnov, E. K. Woller, S. G. DiMagno, Inorg. Chem. 1998, 37, 4971-
4978.
[13] A. I. Gaudette, A. E. Thorarinsdottir, T. D. Harris, Chem. Commun. 2017,
53, 12962-12965.
This article is protected by copyright. All rights reserved.

10.1002/anie.202010587
Angewandte Chemie International Edition
RESEARCH ARTICLE
Entry for the Table of Contents
RESEARCH ARTICLE
Text for Table of Contents
Da Xie, Meng Yu, Zhu-Lin Xie, Rahul T.
Kadakia, Chris Chung, Lauren Ohman,
Kamyab Javanmardi, Emily L. Que*
Page No. – Page No.
Versatile Nickel (II) Scaffolds as
Coordination-Induced Spin-State
Switches for 19F Magnetic
Resonance-Based Detection
This article is protected by copyright. All rights reserved.