Lanthanide Complexes of DO3A-(Dibenzylamino)methylphosphinate: Effect of Protonation of the Dibenzylamino Group on the Water-Exchange Rate and the Binding of Human Serum Albumin

Article abstract of DOI:10.1021/acs.inorgchem.9b00267

Protonation of a distant, noncoordinated group of metal-based magnetic resonance imaging contrast agents potentially changes their relaxivity. The effect of a positive charge of the drug on the human serum albumin (HSA)-drug interaction remains poorly understood as well. Accordingly, a (dibenzylamino)methylphosphinate derivative of 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA) was efficiently synthesized using pyridine as the solvent for a Mannich-type reaction of tBu₃DO3A, formaldehyde, and Bn₂NCH₂PO₂H₂ ethyl ester. The ligand protonation and metal ion (Gd³⁺, Cu²⁺, and Zn²⁺) stability constants were similar to those of the parent DOTA, whereas the basicity of the side-chain amino group of the complexes (logK_A = 5.8) was 1 order of magnitude lower than that of the free ligand (log K_A = 6.8). The presence of one bound water molecule in both deprotonated and protonated forms of the gadolinium(III) complex was deduced from the solid-state X-ray diffraction data [gadolinium(III) and dysprosium(III)], from the square antiprism/twisted square antiprism (SA/TSA) isomer ratio along the lanthanide series, from the fluorescence data of the europium(III) complex, and from the ¹⁷O NMR measurements of the dysprosium(III) and gadolinium(III) complexes. In the gadolinium(III) complex with the deprotonated amino group, water exchange is extremely fast (_M = 6 ns at 25 °C), most likely thanks to the high abundance of the TSA isomer and to the presence of a proximate protonable group, which assists the water-exchange process. The interaction between lanthanide(III) complexes and HSA is pH-dependent, and the deprotonated form is bound much more efficaciously (13% and 70% bound complex at pH = 4 and 7, respectively). The relaxivities of the complex and its HSA adduct are also pH-dependent, and the latter is approximately 2-3 times increased at pH = 4-7. The relaxivity for the supramolecular HSA-complex adduct (r₁^b) is as high as 52 mM^-1 s^-1 at neutral pH (at 20 MHz and 25 °C). The findings of this study stand as a proof-of-concept, showing the ability to manipulate an albumin-drug interaction, and thus the blood pool residence time of the drug, by introducing a positive charge in a side-chain amino group.

Full text of DOI:10.1021/acs.inorgchem.9b00267

Article
pubs.acs.org/IC
Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX
Lanthanide Complexes of DO3A−
(Dibenzylamino)methylphosphinate: Effect of Protonation of the
Dibenzylamino Group on the Water-Exchange Rate and the Binding
of Human Serum Albumin
†
†
Peter Urbanovsky, Jan Kotek, Fabio Carniato,‡ Mauro Botta,‡ and Petr Hermann*,†
́
†Department of Inorganic Chemistry, Universita Karlova (Charles University), Hlavova 2030, 12843 Prague 2, Czech Republic
‡
̀
Dipartimento di Scienze e Innovazione Tecnologica, Universita del Piemonte Orientale “A. Avogadro”, Viale T. Michel 11, 15121
Alessandria, Italy
S
* Supporting Information
ABSTRACT: Protonation of a distant, noncoordinated
group of metal-based magnetic resonance imaging contrast
agents potentially changes their relaxivity. The effect of a
positive charge of the drug on the human serum albumin
(HSA)−drug interaction remains poorly understood as well.
Accordingly, a (dibenzylamino)methylphosphinate derivative
of 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid
(DOTA) was efficiently synthesized using pyridine as the
solvent for a Mannich-type reaction of tBu3DO3A, form-
aldehyde, and Bn2NCH2PO2H2 ethyl ester. The ligand
protonation and metal ion (Gd3+, Cu2+, and Zn2+) stability
constants were similar to those of the parent DOTA, whereas
the basicity of the side-chain amino group of the complexes (logKA = 5.8) was 1 order of magnitude lower than that of the free
ligand (log KA = 6.8). The presence of one bound water molecule in both deprotonated and protonated forms of the
gadolinium(III) complex was deduced from the solid-state X-ray diffraction data [gadolinium(III) and dysprosium(III)], from
the square antiprism/twisted square antiprism (SA/TSA) isomer ratio along the lanthanide series, from the fluorescence data of
the europium(III) complex, and from the 17O NMR measurements of the dysprosium(III) and gadolinium(III) complexes. In
the gadolinium(III) complex with the deprotonated amino group, water exchange is extremely fast (τM = 6 ns at 25 °C), most
likely thanks to the high abundance of the TSA isomer and to the presence of a proximate protonable group, which assists the
water-exchange process. The interaction between lanthanide(III) complexes and HSA is pH-dependent, and the deprotonated
form is bound much more efficaciously (∼13% and ∼70% bound complex at pH = 4 and 7, respectively). The relaxivities of the
complex and its HSA adduct are also pH-dependent, and the latter is approximately 2−3 times increased at pH = 4−7. The
relaxivity for the supramolecular HSA−complex adduct (r1 ) is as high as 52 mM−1 s−1 at neutral pH (at 20 MHz and 25 °C).
b
The findings of this study stand as a proof-of-concept, showing the ability to manipulate an albumin−drug interaction, and thus
the blood pool residence time of the drug, by introducing a positive charge in a side-chain amino group.
tunable through a ligand design. In turn, the relaxivity depends
on the number of water molecules directly coordinated in the
complexes (q), on the residence time of the metal-bound water
molecule(s) (τM), and on the overall tumbling of the CAs
(rotation correlation time, τR). Over the years, the gradual
optimization of the properties of these complexes has
substantially improved our understanding of the relationship
between the structure of the CAs and relaxivity.2,3
The first property used to improve the efficiency of MRI CAs
was molecular tumbling because τR represents the main limiting
factor of the relaxivity of small complexes. To increase τR, two
main strategies are applied. In the first strategy, the complex is
INTRODUCTION
■
In medical imaging, magnetic resonance imaging (MRI) stands
out worldwide as one of the most commonly used imaging
methods, especially for soft tissues. In many examinations,
contrast agents (CAs) are used to improve the resolution and
sensitivity of these imaging methods. Clinically approved MRI
CAs are exclusively based on complexes of highly paramagnetic
metal ions, e.g., trivalent gadolinium, with polydentate ligands
such as diethylenetriaminepentaacetic acid (DTPA) or 1,4,7,10-
tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA) or
their derivatives. Most approved CAs have efficiency much
lower than that theoretically predicted1 because their molecular
parameters are far from optimal. The efficiency of MRI CAs
[termed relaxivity, r1, which is the enhancement of the relaxation
time of water protons (T1) in the presence of 1 mM CA] is
Received: January 30, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.9b00267
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
covalently attached to a large molecule (e.g., a polymer), and the
resulting large conjugate has slow molecular motion.4 The other
strategy is based on the noncovalent binding of a CA to a
macromolecule or nanosized scaffold (proteins, liposomes,
micelles, and nanoparticles, among others),5 most commonly to
human serum albumin (HSA),6,7 which is the most abundant
protein in human blood. The interaction between an MRI CA
and HSA improves the efficiency of the CA thanks to its longer
rotational correlation time and pharmacokinetics in blood,
thereby enabling angiographic imaging. HSA has hydrophobic
binding pockets able to interact with various hydrophobic
groups.8 Thus, many drugs and MRI CAs with hydrophobic
substituents capable of interacting with HSA have been
investigated.6 Several of those MRI CAs have been used in
human patients (Eovist, MultiHance, and Ablavar), and they are
gadolinium(III) complexes of ligands shown in Chart 1: EOB-
and serum albumins changes not only its molecular tumbling but
also other molecular parameters that affect the efficiency of the
CA.9−15
To optimize the efficiency of MRI CAs, their molecular
parameters must be properly tuned.16 Thus, relaxivity enhance-
ment by slowing down the molecular tumbling (i.e., lengthening
τR) of the complexes can be achieved only with an appropriate
value (according to the magnetic field used for imaging) of the
residence lifetime of the water molecule bound to the central
metal ion (i.e., for an optimal τM). This is possible when using
complexes with fast exchange rates between coordinated and
bulk water molecules and with a HSA binding moiety. Although
Ablavar almost meets the above criteria, the blood circulation
time of this MRI CA is too long. Hence, Ablavar is more likely to
release toxic gadolinium(III) ions because complexes of open-
chain DTPA derivatives may be insufficiently stable because of
their low kinetic inertness.17 Therefore, the properties of such
CAs should be optimized by (i) changing the HSA binding of
CAs, (ii) tuning the τM and τR parameters, and (iii) using
macrocyclic DOTA derivatives for the high kinetic inertness of
their complexes.7
Chart 1. Structures of Ligands Used in Selected Gadolinium-
Based MRI CAs That Interact with HSA
Introducing a phosphorus acid pendant arm into the structure
of DOTA has improved the relaxivity of Gd(III) complexes3
thanks to their decreased water residence lifetimes (τM = 10−70
ns),18−23 which are significantly shorter than that of [Gd-
(DOTA)(H2O)]−. In addition, these water residence lifetimes
are almost optimal for attaining very high relaxivity when the
complexes are conjugated to a large molecule with slow
tumbling.24,25 Moreover, the hydrophobic binding pocket of
HSA is decorated with positively charged lysine and arginine
residues.26,27 Consequently, the HSA−drug interaction is more
efficient when a hydrophobic substituent and a negative charge
are both present in the molecules, as exemplified in MS-325,
which has both a hydrophobic diphenylcyclohexyl group and a
phosphate anion. Thus, MRI CAs can be optimized using
effective strategies considering HSA−CA interactions, τM and τR
parameters, and the properties of macrocyclic DOTA
derivatives.
DTPA, BOPTA, and ligand in MS-325, respectively. Complex
MS-325 (Ablavar, earlier also called Vasovist) strongly interacts
with HSA, and it has been approved by the Food and Drug
Administration (FDA) in the USA as an angiographic MRI CA.
This compound has been thoroughly studied, which has
gradually improved our understanding of the interactions
between MRI CAs and blood proteins. The results from these
studies have shown that the interaction between the complex
On the basis of these considerations, we aimed to synthesize
the ligand H4L1 (Chart 1) and to investigate its lanthanide(III)
complexes because this macrocyclic ligand contains two
hydrophobic benzyl groups and a protonable amino group.
Scheme 1. H4L1 Synthesis
B
DOI: 10.1021/acs.inorgchem.9b00267
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Hence, its complexes with HSA should interact through the
benzyl groups as a function of the protonation state of the amino
group because only the nonprotonated form can efficiently bind
to HSA. Concomitantly, the acidity of the amino group should
be lower because of the effect of the neighboring electron-
withdrawing phosphinate group,28 and its protonation state
should be switchable near the physiological pH range. Last, the
fast water exchange of the gadolinium(III) complex, induced by
the bulky phosphinate group, is expected to enhance the
relaxivity when the gadolinium(III) complex is bound to HSA.
RESULTS
■
Ligand and Complex Syntheses. The title ligand was
prepared using tBu3DO3A and (dibenzylamino)-
methylphosphinic acid (1, which was obtained using a simplified
and improved procedure based on a published protocol, as
detailed in the Supporting Information, SI)29 as starting reagents
in a straightforward three-step synthesis (Scheme 1) with no
need to isolate or purify the intermediates. Conditions for the
Mannich-like reaction are very mild, and, thus, no phosphinate
oxidation to phosphonate or tBu3DO3A methylation was
observed. In addition, the reaction was quantitative (31P
NMR) with a slight excess of the macrocycle over the
phosphinate. Carboxylate tert-butyl ester groups were removed
with trifluoroacetic acid (overnight heating), and removal of the
phosphinate ethyl group in 10% aqueous pyridine at 75 °C (∼1
day) led to H4L1.
Figure 1. Molecular structure of the [Gd(H2O)(DO3APDBAm)]− anion
found in the crystal structure of (NH4)[Gd(H2O)(DO3APDBAm)]·
3H2O. Carbon-bound hydrogen atoms are omitted for clarity. Only one
position of the disordered benzyl groups is shown.
Lanthanide(III) complexes of H4L1 were prepared in a
slightly acidic solution and purified by column chromatography
on neutral alumina. During chromatography, the complexes (Rf
= 0.3−0.4) were also desalted (NaCl was eluted just before the
complexes, only slightly overlapping). Repeated evaporation of
the pooled pure complex fractions with water led to an almost
complete removal of ammonia. The complexes were generally
1
obtained as thick oils and were directly used for H and 31P
NMR measurements (characterization and pH dependencies).
Small amounts of the polycrystalline solid samples were
obtained upon preparation of single crystals, and these solids
were used for some physicochemical measurements (lumines-
cence/fluorescence, relaxometry, etc.).
Structures of Complexes in the Solid State. The
molecular structures of the complex anions [Ln(H2O)-
(DO3APDBAm)]− (Ln = Gd, Dy) found in the crystal structures
of (NH4)[Gd(H2O)(DO3APDBAm)]·3H2O and Na[Dy(H2O)-
(DO3APDBAm)]·4H2O are very similar to each other (Figures 1
and 2 and Tables 1 and S4). The central lanthanide(III) ions in
all structurally independent complex molecules are nine-
coordinated, with a metal ion lying between mutually parallel
N4 and O4 planes but closer to the O4 plane (the Gd−QO4 and
Dy−QO4 distances are 0.805 and 0.854/0.832 Å, respectively;
QO4 is the centroid of the O4 plane). An apically bound water
Figure 2. Molecular structures of the [Dy(H2O)(DO3APDBAm)]−
anion found in the crystal structure of Na[Dy(H2O)(DO3APDBAm)]·
4H2O: (A) view along the axial Dy−water bond; (B) side view. Only
one of the two independent complex molecules is shown. Carbon-
bound hydrogen atoms are omitted for clarity. Only one position of the
disordered benzyl groups is shown.
molecule, which caps the O4 plane, closes the coordination
4−
sphere. The (DO3APDBAm
)
ligand anion adopts, in all cases,
the 3,3,3,3-B conformation30 with a twisted-square-antipris-
matic (TSA) geometry in which the pendant arms and the
macrocycle chelate rings have the same signs of rotation,
forming an enantiomeric Λ−λλλλ/Δ−δδδδ pair. The benzyl
groups of the P-substituent are disordered, and the structures are
stabilized by a system of hydrogen bonds involving lattice water
molecules and pendant-arm oxygen atoms.
the ligand distribution diagram is shown in Figure S2. To
estimate the protonation sequence of the ligand, 1H and 31P{1H}
NMR titrations (Figure S3) of the ligand were performed with
CsOH to avoid alkali-metal-ion complexation at high pH.31
More details on the protonation site of the ligand assignment are
given in the SI.
Equilibrium Studies of H4L1 and Its Complexes.
Equilibrium data on the ligand were collected by potentiometry.
Ligand protonation constants are given in Tables 2 and S5, and
C
DOI: 10.1021/acs.inorgchem.9b00267
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 1. Selected Geometric Parameters of the Complex Cage as Found in the Crystal Structures of
(NH4)[Gd(H2O)(DO3APDBAm)]·3H2O and Na[Dy(H2O)(DO3APDBAm)]·4H2O
[Dy(H2O)(DO3APDBAm)]−
parameter
[Gd(H2O)(DO3APDBAm)]−
molecule 1
molecule 2
Distances (Å)
2.583(4)
2.517
1.710(2)
0.805(2)
Angles (deg)
0.9(2)
28.5(2)
24.1(2)
26.6(2)
26.4(2)
Ln−O1 (water O)
2.623(2)
2.542
1.686(1)
0.854(1)
2.748(3)
2.508
1.674(1)
0.832(1)
a
NQ4···QO4
a
Ln···QN4
a
Ln···QO4
N4−O4 plane angle
0.9(1)
1.0(1)
a
a
a
twist N1−QN4−QO4−O11
twist N4−QN4−QO4−O51
twist N7−QN4−QO4−O61
26.0(1)
23.8(1)
23.5(1)
26.7(1)
135.22(8)
138.63(8)
27.3(1)
24.9(1)
26.1(1)
26.9(1)
136.24(9)
139.40(9)
a
twist N10−QN4−QO4−O71
b
O11−Ln−O61
138.5(1)
141.5(1)
b
O51−Ln−O71
a
b
QN4 and QO4 are centroids of the N4 and O4 planes, respectively. “Opening” angle (i.e., transannular O−Ln−O angle).
Table 2. Consecutive Protonation Constants (log KA) of
H4L1 (25 °C and I = 0.1 M (NMe4)Cl] and Their Comparison
with Protonation Constants of Similarly Selected Ligands
(Chart 2)
a
33
species
H4L1
H4DOTA32
H4DO3APABn
HL
12.18
9.16
6.77
4.33
2.39
12.9
9.72
4.62
4.15
2.29
1.34
12.55
9.60
5.11
4.11
2.71
1.54
H2L
H3L
H4L
H5L
H6L
b
b
1.54
a
b
Charges are omitted. Protonation of the side-arm amino group in
boldface.
Figure 3. Distribution diagram of the GdIII-H4L1 system for titration of
the preformed [Gd(H2O)(L1)]− complex. The p[Gd(H2L)]+ label
means an in-cage complex diprotonated on the pendant arms if log KA2
would be fixed to 0.7, which is the lowest limit accessible by our titration
procedure [cGdL1 = 0.003 M, 25 °C, and I = 0.1 M (NMe4)Cl)].
The stability constants of the H4L1 complexes were
determined for the gadolinium(III) ion and for two biologically
important metal ions, copper(II) and zinc(II), by potentiom-
etry. The solutions containing divalent metal ions quickly
reached equilibrium. As expected, the complex of trivalent
gadolinium formed slowly, and, thus, out-of-cell titration was
used. To more accurately determine the protonation constant of
the P-methylamino group, titrations of the preformed GdIII-
H4L1 complex were also performed. The resulting values are
outlined in Tables 3 and S6 and S7, and the distribution
diagrams are shown in Figures 3 and S4 and S5.
Solution Structure of Lanthanide(III) Complexes.
Lanthanide(III) complexes of DOTA-like ligands are commonly
found in solution as square-antiprismatic (SA) and/or TSA +
TSA′ isomers. Knowing their abundance helps with interpreta-
tion of the relaxometric data. Because this study primarily aimed
to assess the effect of the protonation state of the side-arm P-
methylamino group on the properties of whole complexes, the
isomer ratio for the lanthanide(III) complexes of H4L1 was
followed across the lanthanide series and at pHs where the
complexes are either fully protonated or deprotonated. Only two
31P NMR peaks and two sets of 1H NMR signals were observed,
thus confirming that there is one major arrangement around the
phosphorus atom and that only SA and TSA isomers are present
(Figure S6). The signals of the SA/TSA isomers were assigned
based on the lanthanide-induced shift of the “axial” cyclen ring
protons16 and on a comparison of the isomer signal intensity in
1H and 31P NMR spectra. The results showed that the SA/TSA
isomer ratio changes with the size of the lanthanide(III) ion and
a
Table 3. Equilibrium Constants (log KGdL or log KA) of Gadolinium(III) Complexes of H4L1 and Selected Ligands [25 °C and I =
0.1 M (NMe4)Cl]
b
c
34
33
d35
,
,
equilibrium
H4L1
H4DOTA
24.7
H4DO3APABn
24.04
MS-325
22.06
Gd + L ⇄ [Gd(L)]
23.77
e
e f
,
e
[Gd(L)] + H ⇄ [Gd(HL)]
[Gd(HL)] + H ⇄ [Gd(H2L)]
5.63 (5.85)
4.76
f
1.74 (<1)
a
b
c
d
e
Out-of-cell titrations. Charges are omitted. 25 °C and I = 0.1 M NaCl. 25 °C and I = 0.1 M NaClO4. Protonation of the pendant amino group
f
in boldface. Protonation constants of the preformed gadolinium(III) complex in parentheses [25 °C and I = 0.1 M (NMe4)Cl].
D
DOI: 10.1021/acs.inorgchem.9b00267
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
with (de)protonation of the P-methylamino group (Figure 4 and
Table S8). To determine the protonation constants of the SA/
The 1H NMRD profiles of both deprotonated and protonated
forms of the complex at different temperatures (Figure 5) and
Figure 4. Abundance of the TSA/TSA′ isomers of the LnIII-H4L1
complexes in their protonated (blue) and deprotonated (red) forms at
25 °C.
TSA isomers, 1H and 31P NMR titrations were performed for the
EuIII-, TbIII-, and YbIII-H4L1 complexes, and the results are
shown in Table S9 and Figure S7.
To further confirm the presence of the SA/TSA isomers and
to examine their solution structure, other spectral investigations
7
were conducted. High-resolution UV−vis spectra (5D0 ← F0
transition; Figure S8; details are given in the SI) and emission
lifetimes (Figure S9 and Table S10; details are given in SI) of the
EuIII-H4L1 complex in both protonated forms were recorded.
The fluorescence lifetime titrations of the complex in water and
D2O (Figure S9) indicated log KA values of 5.6 and 6.2,
respectively. Dysprosium-induced shift (DIS) of the 17O NMR
signal of bulk water containing a dissolved dysprosium(III)
complex36 showed q = 0.9 and 1.2 in the deprotonated and
protonated forms of the DyIII-H4L1 complex, respectively.
Relaxometric Evaluation of the GdIII-H4L1 Complex.
Variation of the relaxivity as a function of the pH was measured
(20/40 MHz and 37 °C) for relaxometric evaluation of the
GdIII-H4L1 complex. The results closely reproduce the pH-
dependent variation in the relaxivity presented above. The water
proton relaxivity (r1) was constant up to pH ∼ 5 and then
decreased until reaching a new plateau at pH > 7 (Figure S10).
The titration curves indicated a log KA of 5.8 for protonation of
the P-methylamino side arm. However, the pH-dependent
variation of r1 is limited and not clearly associated with a change
in the hydration state of the complex, in contrast to the previous
results.37 For further insight into the physicochemical character-
istics of this novel gadolinium(III) complex and its pH-
1
Figure 5. H NMRD profiles of the GdIII-H4L1 complex at pH = 4.0
(A) and 8.3 (B) at several temperatures. The curves are calculated using
the best-fit parameters (Table 4).
variation of the 17O NMR data with the temperature (Figure
S11) were assessed. The 1H NMRD profiles were recorded over
a proton Larmor frequency range of 0.01−70 MHz. The shape
of the profiles of both protonated and deprotonated complexes
are rather typical of low-molecular-weight gadolinium(III)
chelates. The curves are characterized by a plateau at low
magnetic fields (ca. 0.01−1 MHz), followed by a dispersion
centered around 5 MHz and by a second plateau in the 20−70
MHz frequency range. For both complex forms, the relaxivity
values showed a marked decrease with an increase in the
temperature from 10 to 37 °C, which indicates the occurrence of
fast exchange of the bound water molecule. This implies that r1 is
mainly limited by fast rotation of the complex and not by a long
water-exchange lifetime (τM). In turn, the 1H NMRD data were
analyzed using the standard Solomon−Bloembergen−Morgan
theory for the inner-sphere (IS) contribution to relaxivity38 and
Freed’s model to account for the outer-sphere (OS) contribu-
tion.39 Considering the high number of parameters requiring
fitting, some of them were fixed at known or reasonable values
using a well-established procedure. In this best-fit procedure, the
Gd−Hw distance (rGdH) was set to 3.1 Å, the distance of the
closest approach of the OS water molecules to the gadolinium-
1
dependent relaxivity, detailed H and 17O NMR relaxometric
studies were performed. Variation of the water proton relaxivity
of solutions of gadolinium(III) complexes as a function of the
magnetic field was measured as nuclear magnetic relaxation
dispersion (1H NMRD) profiles. Their analysis based on the
established theory of paramagnetic relaxation1 makes it possible
to calculate the values of several structural, electronic, and
dynamic parameters of these complexes, including the number
of water molecules in the inner and second-coordination spheres
(q and qss, respectively), their distance from the paramagnetic
center (r), and the overall molecular tumbling time of the
complex (τR). In addition, variation of the 17O NMR transverse
relaxation rate, R2, and the shift, Δω, as a function of the
temperature provide accurate information on the kinetics of
coordinated water exchange (kex = 1/τM).
E
DOI: 10.1021/acs.inorgchem.9b00267
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 4. Selected Relaxometric Parameters Assessed by the Simultaneous Analysis of 1H NMRD Profiles and 17O NMR Data
(11.7 T) on Gadolinium(III) Complexes of H4L1 and Related Ligands (Chart 2)
a
H4L1
H5DO3AP
b
b
c
d
e
2
parameter
pH = 8.3
pH = 4.0
H4DOTA
H4DO3APOEt
pH = 7.0
pH = 2.5
CS-(DO3APNBn
)
f
f
f
r1298 (20 MHz)/mM−1 s−1
Δ2/1019 s−2
5.5
2.9 0.2
13
6.3
2.0 0.1
4.7
1.6
11
0.41
77
3.15
49.8
−3.7
1
4.5
4.25
8.7
2.0
84
3.1
61.4
−3.3
1
4.57
20.7
3.9
4.32
6.1
9.1
4.6
2.70
70
3.1
55.8
−3.28
1
2.3
15.9
1.9
183
3.1
39
τV298/ps
1
15
6.7 0.3
102
1
kex298/107 s−1
τR298/ps
20.0 0.5
7.14
83
92
3.1
4
3
g
g
r
Gd−H/Å
3.1
3.1
ΔH#M/kJ mol−1
34.1 0.6
40.3 1.9
76.0
−3.28
AO/ℏ/106 rad s−1
−3.5 0.1
−3.3 0.2
−2.89
1
g
g
q
qSS
298τRSS/ps
1
1
1
1
62
g
g
1
1
1
1
36
1
23
30
3
59
4
a
b
c
From ref 18. Values of 4.0 Å, 2.24 × 10−5 cm2 s−1, and 17 and 1 kJ mol−1 were used for the parameters a, 298D, ER, and EV, respectively. From ref
d
e
f
g
44. From ref 20. From ref 45. Relaxivity at 310 K. Fixed in the fitting procedure.
Figure 6. Reduced 17O NMR transverse relaxation rates (T2r, top) and 17O NMR chemical shifts (Δω, bottom) for GdIII-H4L1 at pH = 4.0 (left) and
8.3 (right) measured at 67.8 MHz (11.74 T). The solid lines were calculated using the parameters listed in Table 4.
(III) ion (aGdH) was fixed to 4.0 Å, and the values of 1.3, 2.24,
and 3.1 (×10−5 cm2 s−1) were used for the water-solute relative
diffusion coefficient (D) at 283, 298, and 310 K, respectively.
The number of coordinated water molecules (q) was fixed to 1,
similar to complexes of related DOTA-like ligands.40 The fit was
performed using the following adjustable parameters: the overall
molecular tumbling time of the complex (τR) and the electronic
relaxation parameters Δ2 (trace of the squared zero-field-
splitting, ZFS, tensor) and τV (correlation time for modulation
of the transient ZFS). The best-fit parameters are listed in Table
4 and compared with those of related gadolinium(III)
macrocyclic complexes of similar sizes with q = 1.
reorientation (i.e., long τR). Conversely, the profiles were well
reproduced, accounting for a sizable second-sphere (SS)
contribution. This contribution was defined by the presence of
water molecules, which closely diffused to the complex at a
sufficiently short distance from the paramagnetic ion (ca. <4 Å)
and at a residence time (τMSS) long enough to be affected by the
rotation.41,42 These short distances and long times are
commonly found in systems with phosphonic/inic acid donor
groups.18−21,41,42 Therefore, in this model, two additional
parameters, the number of SS water molecules (qSS) and their
rotational correlation time (τRSS), were included in the analysis
(Table 4). The average distance of the water protons from the
paramagnetic center was arbitrarily fixed at 3.6 Å, an
intermediate value between those of water molecules in the
inner (3.1 Å) and outer (4.0 Å) solvation shells. Thus, the
resulting best-fit parameters are completely in line with those
Analysis of the experimental 1H NMRD curves using a model
that considers only IS and OS contributions to the relaxivity
failed to provide satisfactory and plausible results. The relaxivity
values were too high to be explained only by slow molecular
F
DOI: 10.1021/acs.inorgchem.9b00267
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
assessed in similar complexes of similar dimensions and
molecular geometries (see below).
The variation of 17O NMR R2 and Δω as a function of the
temperature was measured (14.1 T, 20 mM gadolinium(III)
solution, and pH = 8.3 and 4.0) and then analyzed using the well-
established set of Swift−Connick equations.43 The depend-
encies of the R2 (1/T2r) and Δω values are shown in Figure 6.
The increase of 1/T2r with a decrease in the temperature over a
wide range of values confirms the fast water-exchange rate, as
expected. 17O NMR R2 data primarily depend on the electronic
relaxation times (T1,2e), the hyperfine Gd−17Owater coupling
constant (AO/ℏ), τM, and q. Information on q and AO/ℏ are
derived from the variation of Δω with the temperature.
Additional parameters, relevant to the exchange process, are
associated with the exchange lifetime of the coordinated water
Figure 7. Relaxometric titration of the GdIII-H4L1 complex with HSA at
different pH values (25 °C, 20 MHz, and cGd = 0.2 mM).
#
molecule(s), τM, and with its enthalpy of activation, ΔHM
.
Typical values of 1.0 and 17 kJ mol−1 were assigned to the
activation energy for modulation of the ZFS interaction (Ev) and
for rotational motion of the complex (ER), respectively. The
best-fit parameters are reported in Table 4.
binding isotherm even though the presence of multiple (low)
affinity sites on HSA cannot be ruled out. At low HSA
concentrations, the deprotonated complex showed a marked
enhancement in R1 (pH = 7.0 and 8.3), which indicates that this
form of the chelate has a good affinity for the protein pockets. log
Interaction between the GdIII-H4L1 Complex and HSA.
HSA has long been known to bind noncovalently to various
hydrophobic molecules, thereby increasing their blood circu-
lation time.46 For this reason, MRI CAs with hydrophobic side
groups have already been developed as blood pool agents, and
some are used in clinical practice.6 However, their pharmaco-
logical properties are not entirely satisfactory, and their design
can be improved by enhancing specific properties. Because the
GdIII-H4L1 complex contains a hydrophobic side arm, we
decided to investigate the binding interaction of GdIII-H4L1
with HSA. The experiments with the protonated form of GdIII-
H4L1 were performed at pH = 4.0 because the abundance of this
form is higher than 95% at this pH. Data on the deprotonated
form of GdIII-H4L1 were collected at pH ∼ 7 (85−90%
abundance) to avoid major structural changes of HSA, which
occur at pH > 7.8.47
K
aff of the sample at neutral pH is 2.9 with an r1b value of 52.0
mM−1 s−1, a rather high value, which suggests a low degree of
rotational freedom of the complex in the binding site. At 0.2 mM
GdIII-H4L1, 4.5% HSA, and 25 °C, the fraction of the complex
bound to HSA is approximately 55%.
The relaxometric titration at pH = 4.0 clearly indicates a weak
interaction between the protonated complex and protein (log
K
aff = 2). Under conditions identical with those at neutral pH,
b
the fraction of the bound complex is only ca. 7%, with an r1
value of 47 mM−1 s−1. These results confirm the initial
hypothesis according to which protonation of the side arm
amino group would significantly weaken the interaction between
the complex and HSA. Unsurprisingly, as shown in Figure 7, the
binding affinity of GdIII-H4L1 at pH = 8.3 is rather similar to that
at pH = 7, with log Kaff = 3.3. However, the relaxivity of the
The interaction between GdIII-H4L1 and HSA was inves-
tigated by 1H NMR relaxometric titrations in the presence and
absence of substrates known to tightly bind to the protein, i.e.,
warfarin and ibuprofen. These two drugs selectively target
binding sites BS-I (warfarin) or BS-II (ibuprofen) with log Kaff >
6.48 The titration experiment consists of measuring the increase
in the water proton longitudinal relaxation rate (R1) with the
concentration of HSA at a given proton Larmor frequency and
temperature (20 MHz and 298 K). R1 is enhanced with an
increase in the fraction of the bound complex due to a decrease
in its tumbling motion. Three different relaxometric titrations
(at pH = 7) were performed in (i) a dilute solution of the
complex, (ii) a mixture of the complex and ibuprofen in a 1:5
molar ratio, and (iii) a mixture of the complex and warfarin in a
1:5 molar ratio (Figure S12). The binding curves of the solutions
containing the Gd(III) complex and the complex−warfarin
mixture were identical. Conversely, the relaxation rates were
lower than those in the absence of ibuprofen, which indicates
preferential complex binding to the BS-II binding site. The
relaxometric titration experiments were also used to estimate the
affinity constant, Kaff, of the protein and relaxivity of the fully
bound complex is lower (r1 = 30 mM−1 s−1), an unexpected
b
result likely associated with changes in the protein conformation
at basic pH and with a loss of the quaternary structure.
Competitive assays using selective high-affinity ligands for BS-
I and BS-II followed by fluorescence spectroscopy were also
used to determine the binding sites and affinity constant of the
complex for HSA (details are given in the SI and Figure S13).
The experiments confirmed the above relaxometric data:
interaction with the BS-II site and log Kaff ∼ 3.0 with n = 0.9
(n is the number of binding sites on HSA) for the deprotonated
form as well as much lower affinity for the protonated form.
1
The H NMRD profiles of GdIII-H4L1 (0.2 mM) in the
presence of a large excess of HSA (1.3 mM) at different pH
values were also recorded in the frequency range of 0.01−70
MHz and at 25 °C (Figures 8 and S14). Under these conditions,
the paramagnetic complex is predominantly found in the
protein-bound form. In all cases, the field dependence of the
relaxivity is quite typical of slowly tumbling macromolecular
systems, with a peak centered around 20 MHz associated with a
pronounced decrease of the tumbling motion of the metal
chelate (i.e., with a long τR).
49
b
bound complex, r1 .
Dilute (0.2 mM) aqueous solutions of the complex were
titrated with HSA at pH = 4.0 (fully protonated complex), 7.0
(large population of the deprotonated complex), and 8.3 (fully
deprotonated complex but nonnative conformation of HSA), at
298 K and 20 MHz (Figure 7). All data were fitted to a 1:1
DISCUSSION
■
The ligand was synthesized according to the most commonly
applied approaches using tBu3DO3A and 1 as starting materials.
G
DOI: 10.1021/acs.inorgchem.9b00267
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
H4L1 were 60−70%, although the conversions (in situ 31P
NMR) of all reactions were quantitative. These decreases in
yield were caused by some losses of the ligand on the cation-
exchange resin and by the slight solubility of H4L1 in
tetrahydrofuran (THF) during the trituration procedure. The
lanthanide(III) complexes were obtained following a procedure
commonly used to prepare LnIII-H4DOTA complexes, and they
were desalted on alumina.
Single crystals of the free ligand in zwitterionic form (for a
discussion of the structure, see the SI) and of its GdIII and DyIII
complexes in fully deprotonated forms were obtained in this
study. In the solid state, [Ln(H2O)(L1)]− (Ln = Gd, Dy) anions
are found as TSA isomers. This TSA arrangement is common
among structurally described lanthanide(III) complexes of other
HnDO3APR derivatives,18,50−52 and the present structures
follow the norm. Water can surprisingly coordinate up to a
dysprosium(III) ion, although lanthanide(III)−water distances
are rather long (2.623 and 2.748 Å for each complex unit; Table
1). Nevertheless, this distance is similar to that observed in
hydrated lanthanide(III) complexes of other HnDO3APR
derivatives51 because the smaller lanthanide(III) ion sinks
deeper into the ligand cavity and increasingly so with the
narrowing of the “opening angles” (i.e., O−Ln−O angles
involving the mutually trans oxygen atoms in the O4 plane).
Consequently, this sinking leads to steric crowding at the water
binding site above the O4 plane. An opening angle value of
∼135° has been suggested as the water coordination thresh-
old.3,51 The corresponding values of the dysprosium(III)
complex units (Table 1) lie around this threshold. This
observation is in line with the solution data, where the IS
hydration starts to decrease after gadolinium(III) (Figure 4) but
still is significant for the DyIII-H4L1 complex, as confirmed by
the DIS results.
The lanthanide(III) complexes were prepared to investigate
variation of the HSA binding with the pH. Thus, the
protonation/stability constants and data on the protonation
sites of the ligand and complexes are crucial information.
Accordingly, equilibrium data were collected through potentio-
metric and NMR titrations. The log KA values determined by
potentiometric titration match those assessed by analysis of the
NMR signals of the nuclei of the pendant arms (Table 2). The
ligand protonation sites were also estimated by analysis of the
NMR titration curves (for details, see the SI).31,33,53 Thus,
similar to other cyclen derivatives, the first two protons should
be bound to the ring amine groups, which are, as expected, less
Figure 8. 1H NMRD profiles of the GdIII-H4L1 complex in the presence
of HSA (A) and calculated relaxivities of the complex fully bound to
HSA (B) at several pH values (cHSA = 1.3 mM, ccomplex = 0.2 mM, and 25
°C).
For the phospha-Mannich reaction with H−P ester 2, dry N,N-
dimethylformamide was also tested as a solvent, but the
necessary reaction time was 7−8 days. Other solvents
commonly used in phospha-Mannich reactions, such as ethanol
(EtOH), acetonitrile, or toluene, led to inefficient reactions
requiring an excess of the phosphinate ester and higher
temperatures. The reactions generally led to difficult-to-purify
mixtures. Acetate tert-butyl esters were cleaved using a standard
procedure with CF3CO2H. For the final phosphinic ester
deprotection, the reaction in aqueous pyridine required about 1
day at 75 °C. Deprotection in 1 M aqueous potassium hydroxide
was also tested.20 At room temperature, the reaction was very
slow (requiring up to 2 weeks for complete conversion); at 75
°C, the ligand slowly decomposed. The overall isolated yields of
Chart 2. Structures of Other Ligands Discussed in the Text
H
DOI: 10.1021/acs.inorgchem.9b00267
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
basic than those of H4DOTA32 and similar to those of
monophosphinic acid analogues,33 e.g., H4DO3APABn (Chart
2). The following proton is attached to the dibenzylamino group
in the phosphorus pendant arm (log KA = 6.77). The value is
lower than that expected for Me-NBn2 (predicted log KA ∼ 7.8)
due to the presence of the electron-withdrawing phosphinate
group. The next three protonations occur on the carboxylate
groups. Thus, protonation of the pendant-arm amino group
occurs within the physiological pH range.
The divalent metal and gadolinium(III) ions are fully
complexed at pH > 4−5 and 3, respectively. The thermody-
namic stability of the complexes is similar to that of H4DOTA
complexes or of their monophosphinate derivatives.33 The first
protonation constants of the complexes in the slightly acidic
region can be explained by protonation of the side-arm P-
methylamino group. The value of the protonation constant
remains unchanged between the out-of-cell titration and the
titration of the preformed gadolinium(III) complex (the latter is
more accurate than the former). Although the amino group is far
away from the central metal ion, the group in the gadolinium-
(III) complex becomes more acidic (log KA1 = 5.85 for the
preformed gadolinium(III) complex) than expected and falls
slightly below the desired physiological pH range. The titration
of the preformed gadolinium(III) complex qualitatively proved
that the complex is kinetically inert because the complex does
not decompose upon simple acidification of the reaction mixture
to pH = 1.5, and its protonation may occur only in very acidic
solutions (log KA2 < 1). Thus, the data show that both
deprotonated and monoprotonated GdIII-H4L1 species are fully
thermodynamically stable.
The abundances of the SA and TSA isomers of Ln(III)
complexes of DOTA-like ligands in solution depend on the
ligand structure and the size of the lanthanide(III) ion.40 The
isomerism originates from the mutual arrangement of cyclen
chelate rings and from the direction of rotation of the pendant
arms. Among complexes of DOTA-like ligands, the abundance
of the TSA isomer decreases along the lanthanide series;
however, its abundance increases again when the coordinated
axial water molecule is expelled in complexes with the small
lanthanide(III) ion (such an “anhydrous” isomer is commonly
abbreviated as TSA′). In addition, the phosphorus atom
becomes chiral after coordination of the phosphinate group. A
combination of these structural elements leads to four possible
diasteroisomers in lanthanide(III) complexes of monophos-
phinic acid analogues of DOTA.19−21,53 However, only one
preferred arrangement on the phosphorus atom (>90%) is
commonly observed in complexes of the ligand with P-alkyl
substituent(s).19,21,42,54 Here, only one set of 1H NMR signals of
the TSA and SA isomers and two 31P NMR signals were also
observed (Figure S5). The SA/TSA isomers have different MRI-
related properties.3 The SA/TSA isomer ratio of complexes of
different lanthanide(III) ions varies as expected.40,55 The
hydrated TSA arrangement prevails among large ions, its
abundance is minimal in the middle of the lanthanide series, and
it increases again among smaller ions, albeit now as TSA′ species
(Figure 4 and Table S8). The data clearly show that the TSA
isomers are more abundant in deprotonated than in protonated
complexes. Furthermore, in solution, the IS water molecule
should be present up to the terbium(III) complex because
population of the anhydrous TSA′ isomer becomes noticeable
only for the complexes of the heavier lanthanide(III) ions
(Figure 4). In the solid state, the IS water molecule was observed
up to the dysprosium(III) complex (see above). The SA/TSA
isomer ratio of LnIII-H5DO3AP complexes varies similarly as a
function of the pH (and, thus, of the protonation state of the
complexes; the structure of H5DO3AP is shown in Chart 2).18
The abundances of TSA isomers in the deprotonated [Gd-
(H2O)(DO3APDBAm)]− and protonated [Gd(H2O)-
(HDO3APDBAm)] complexes are 60−70% and 40−50%,
respectively. Thus, these abundances of TSA isomers are higher
than those of the [Gd(H2O)(DOTA)]− complex (∼15%)55 and
similar to those of gadolinium(III) complexes of other
monophosphorus acid analogues of DOTA.18−21 The proto-
nation constants of the side-arm P-methylamino group of EuIII-,
1
TbIII-, and YbIII-H4L1 complexes were determined by H/31P
NMR titration, and similar to the gadolinium(III) complex, they
are approximately 1 order of magnitude lower than that of the
free ligand (Table 3) because coordination of the phosphinate
group to the metal ion further increases the electron-
withdrawing effect of the group. For all complexes, the SA
isomer is more basic than the TSA isomers (log KA values are
∼6.3 and ∼5.8, respectively; Table S9). The order of the
protonation constants of the isomers, SA > TSA, is opposite to
that of the phosphonate group in the SA/TSA isomers of LnIII-
H5DO3AP complexes.18
The UV−vis and emission spectra of the europium(III)
complex (if the expected extensive SS hydration due to the
presence of phosphinate42 and closely located amino groups is
taken into account) confirm the presence of the hydrated
complex for both protonated and deprotonated species (for a
more detailed discussion, see the SI). Fluorescence lifetime
titration (Figure S9) confirmed the value of the protonation
constant of the side-arm methylamino group. The abundances
of the SA and TSA isomers calculated from the absorption
spectra match those determined from 31P NMR spectra (40%
SA and 60% TSA and 30% SA and 70% TSA for protonated and
deprotonated complexes, respectively, at 25 °C). The results of
DIS measurement also support a hydration state corresponding
to q = 1 even for the dysprosium(III) complex.
Overall, the above data show that LnIII-H4L1 complexes are
present in aqueous solution as a mixture of SA and TSA isomers
and that the gadolinium(III) complex should be monohydrated
in both deprotonated and protonated forms. However, the
coordinated water molecule should be highly labilized because
of the easily accessible transition state at q = 0 (see the
relaxometry part below).
The relaxivity of the GdIII-H4L1 complex is pH-dependent
with pKA ∼ 5.8, and the value is similar to those assessed using
other methods. Therefore, relaxivity-related parameters were
determined for both forms of the GdIII-H4L1 complex,
[Gd(H2O)(HL1)] and [Gd(H2O)(L1)]− (Table 4). Most of
them are similar to those of gadolinium(III) complexes of
H4DOTA and/or of complexes of their previously investigated
monophosphorus acid analogues. The fitted values of the 17O
NMR hyperfine coupling constant, AO/ℏ, of ca. −3.5 ( 0.1) ×
10−6 rad s−1, independently support the presence of one IS water
molecule in both forms of the gadolinium(III) complex.
Variation of the relaxivity with the pH (5.5 vs 6.3 mM−1 s−1 at
pH = 7 and 4, respectively, 20 MHz, and 25 °C) is not explained
by changes in the common determinants of the IS mechanism.
Protonation of the complex is not associated with a change in the
molecular size and, therefore, cannot imply a significant
variation in τR. More specifically, the rotational correlation
time, τR, is on the order of 100 ps, i.e., rather close to the values
reported for the GdIII−H4DOTA complex and related chelates
with similar molecular masses. The tumbling motion of the
I
DOI: 10.1021/acs.inorgchem.9b00267
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
protonated complex is somewhat slower, most likely because the
SS water molecules have a different arrangement caused by P-
methylamine protonation and, thus, a different shape of the
entire molecule. It might be caused by a different hydrogen-
bonding network involving protonated amine, oxygen atoms of
pendant arms, and/or water molecules in the first/second
coordination sphere. Conversely, the increase in r1 for the
protonated complex is too small to be attributed to a change in
the hydration number q [although partial hydration for the
deprotonated complex cannot be fully ruled out because the
hydration break takes place just behind the gadolinium(III)
complex]. The values of the parameters describing the electronic
relaxation, Δ2 and τV, are in agreement with those found in other
gadolinium(III) complexes of DOTA-like ligands. This implies a
similar molecular geometry characterized by high symmetry and
stereochemical rigidity. However, the observed change in r1 with
protonation can be accounted for by differences in the
contribution of SS hydration (see also below). An effective
hydrogen-bonding interaction of the tetrahedral phosphinate
group with proximate water molecules is at the origin of this
large relaxivity value could be further enhanced after GdIII-H4L1
conjugation with a slowly tumbling substrate because r1 would
not be limited by the low water-exchange rate.
Hydrophobic molecules show an affinity for the two main
hydrophobic cavities of the protein, commonly known as
Sudlow’s binding sites I (BS-I) and II (BS-II). The hydrophobic
pockets are located deep inside the protein structure, and the
connecting channels have positively charged lysine and histidine
residues close to its opening. Therefore, the ligand−HSA
interaction is enhanced by the presence of a negative charge in
the interacting molecules or could be destabilized by the
presence of a positive charge close to their hydrophobic sites. To
the best of our knowledge, no report has been published thus far
showing the effect of the presence of a positive charge near the
binding group on the binding interaction to HSA. The title
ligand was also designed to enable triggering of the HSA binding
by varying the pH due to the presence of a protonable amino
group. The nonprotonated complex should form a more stable
complex−HSA conjugate than its protonated form, thereby
slowing the molecular tumbling and increasing the relaxivity.
Protonation of the amino group brings a positive charge close to
the hydrophobic benzyl groups, which should weaken the
interaction between the complex and HSA. Although proto-
nation of the side-arm dibenzylamino group occurs at
nonoptimal pH (see above), our hypothesis was tested in this
proof-of-concept study based on well-established approaches
using fluorescence and relaxometric measurements.
effect.18−23,56 In particular, this contribution is more significant
SS
for the protonated complex because of the higher value of τR
,
which suggests a stronger interaction between the water
molecules and protonated complex. Accordingly, protonation
of the complex does not modify its hydration state but disturbs
the structure of the second coordination sphere, which is
reflected in variation of the relaxivity.
Similar to other complexes of monophosphorus acid DOTA
analogues,18−21 the coordinated water residence lifetimes, τM, of
both forms of GdIII-H4L1 are rather short, in the range of 5−15
ns at 298 K. The water residence lifetime of the deprotonated
complex (τM ∼ 5 ns) is among the shortest ever reported for
monohydrated gadolinium(III) complexes with ligands of the
DOTA family.23 It is comparable to the residence lifetime values
of complexes of the DOTA-like ligands with a noninteger value
of the hydration number.22,23 The lability of the bound water
molecule likely originates from its easy access to the eight-
coordinated, nonhydrated transition state of the deprotonated
complex following a dissociative water-exchange mechanism. In
fact, gadolinium(III) is located just before the onset of the
appearance of the TSA′ isomer in the lanthanide series, which
starts with the terbium(III) complex (Figure 4). Such a
significant acceleration of the water-exchange rate, induced by
the presence of a phosphonate group, has also been observed in
complexes of monophosphonic acid cyclen derivatives with
various structures22,23 and in complexes of octadentate ligands
based on, e.g., ethylenediamine.57 Thus, we can also speculate
that hydrogen bonding to the proximate amino group can assist
the exchange of the bound water molecule. This effect is more
efficient in the deprotonated complexes in which an O−H···N
interaction between the bound water and amino group can
occur. A similar water-exchange acceleration in the deproto-
nated complex was observed in the GdIII-H5DO3AP complex at
neutral pH.18 Moreover, a recent study has shown that, in
lanthanide(III) complexes of DOTA-tetraamides derived from
linear diamines, the interaction between the protons of
coordinated water and adjacent free amino groups of the
pendant arm significantly accelerates the water-exchange
process.58 Notwithstanding, this water-exchange process may
also be affected by the presence of a bulky dibenzylamino group
close to the axial water binding site.
Fluorescence measurements were used to probe the affinity of
GdIII-H4L1 for the protein binding sites and to assess whether
this interaction depends on the pH. These measurements
suggest a preferred interaction between deprotonated GdIII-
H4L1 and BS-II, and the GdIII-H4L1 conjugate with HSA is
formed mainly in a 1:1 molar ratio, with log Kaff ∼ 3.0 and with a
number n of binding molecules of ∼0.9. The corresponding
relaxometric competitive experiments confirmed BS-II as a
binding site because GdIII-H4L1 specifically competes with
ibuprofen, a high-affinity ligand for this site. Data analysis also
provided conditional affinity constants, log Kaff, with values of
2.0 and 3.3 at pH = 4 and 7 (25 °C), respectively. These values
imply that ca. 7% and 55% of the complex is bound to HSA at pH
= 4 and 7 (ccomplex = 0.2 mM and 25 °C), respectively. At pH =
8.3, the fraction of the bound GdIII-H4L1 complex decreases
because of the loss of the quaternary structure of HSA,
accompanied by a decrease in the relaxivity of the system.
Thus, the main GdIII-H4L1 binding site of HSA is identical with
that of MS-325 (Ablavar). However, the fraction of the bound
complex at pH = 7 is significantly smaller than that of MS-325
(88% at 0.1 mM of complex; 4.5% HSA; 37 °C).12 The
protonation constant log KA of 5.7 for GdIII-H4L1, determined
by relaxometric pH titration in the presence of HSA, is similar to
the value obtained using other methods (see above). The
relaxivity changes with the pH (Figure 8), and this variation
results from differences in the affinity for the protein between
the two forms of the complex.
The 1H NMRD profiles of the HSA adducts have the typical
shape of systems with long τR featuring the characteristic peak
centered at 20 MHz. The height of the peak and large r1b value
calculated in this analysis, 52 mM−1 s−1 (pH = 7.0, 25 °C, and 20
MHz), strongly suggest a marked restriction of the internal
mobility of the chelates in the protein binding pockets. This
value is very similar to that found for gadolinium(III) complexes
of DOTA derivatives with (benzyloxy)methylene groups under
similar experimental conditions.49 Furthermore, a relaxivity of
On the basis of that above, both forms of GdIII-H4L1 exhibit
significantly higher relaxivity than [Gd(H2O)(DOTA)]−. The
J
DOI: 10.1021/acs.inorgchem.9b00267
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
50.8 mM−1 s−1 (37 °C and 20 MHz) was reported for MS-325
the pH does not lead to any change in its first coordination
sphere and, in parallel, in the hydration state of the complex, thus
maintaining its high stability.
fully bound to HSA.12 The r1 values decrease with the
b
temperature over the entire frequency range, thus indicating the
occurrence of fast exchange conditions. Accordingly, the
relaxivity is not limited by τM, because it is typically observed
in most protein-bound gadolinium(III) complexes, but instead
by the extent of motional coupling between the global rotation
of HSA and local reorientation of the complex at the binding site.
CONCLUSIONS
■
A (N,N-dibenzylamino)methylphosphinic acid analogue of
DOTA was prepared to assess the effect of introducing a
protonable group near the water binding site of lanthanide(III)
complexes of macrocyclic ligands based on the DOTA family.
New synthetic conditions using pyridine as a solvent in a
Mannich-type reaction between a secondary amine and a
hydrogen phosphinate ester led to an improved yield and to a
significantly clearer reaction mixture, thereby enabling a simple,
one-pot, gram-scale synthesis of the ligand. The thermodynamic
properties of the ligand and of its complexes are similar to those
of DOTA and of its monophosphorus acid analogues. The
phosphorus side-arm amino group of lanthanide(III) complexes
is protonated below pH = 6. The hydration break in complexes
of lanthanide(III) ions appears just after gadolinium and, thus,
the gadolinium(III) complex is fully hydrated in both
protonation states. A small energy difference between species
with coordination numbers 9 and 8 and the presence of the
amino group near the IS water molecule account for the
extremely fast water-exchange rate, unprecedented for
gadolinium(III) complexes of DOTA-like ligands with q = 1.
The amino group likely assists the transfer of the bound water
molecule to the bulk through hydrogen-bonding interactions,
thus highlighting a new approach to modulating the water-
exchange rate of MRI CAs. The hydrophobic dibenzylamino
moiety interacts with HSA in a protonation-dependent manner,
and binding with its positively charged protonated form is much
weaker. The gadolinium(III) complex exhibits pH-dependent
relaxivity in both free and HSA-bound forms. Hence, the
gadolinium(III) complex can be considered a pH-sensitive MRI
probe, and the sensitivity to the pH might be enhanced in the
blood pool because HSA binding further changes the relaxivity
value. This pH-dependent HSA interaction has never been used
to modulate the pharmacokinetics of drugs in the blood pool.
Protonation at a lower pH (e.g., in tumor or kidney tissues)
should lead to decomposition of the supramolecular HSA−drug
complex and to drug clearance. This study can be considered the
proof-of-concept of this approach. Here, the binding ability of
the deprotonated dibenzylamino group is suboptimal and the
basicity of the group is too low. By an increase in the relative
strength of the interaction between the deprotonated form of a
complex (a drug, in general) and HSA, compared with that of
the protonated form, and adjustment of the corresponding pKA
value, the blood residence time of such a complex may be tightly
modulated. In general, the findings of this paper provide new
options for designing fast water-exchanging or pH-responsive
MRI CAs and for varying the pharmacokinetics of drugs in blood
by tissue pH.
b
Interestingly, variation of the frequency as a function of r1
measured at pH = 4, at which the complex has the lowest affinity
for the protein, is very similar to that assessed at neutral pH. The
presence of the positive charge on the complex and the different
conformations of HSA at pH = 4 may have only a negligible
effect on the overall rotational dynamics of the complex at the
binding site. However, this limited effect is highly unlikely
because a number of exchangeable protons bound to the protein
surface and the formation, upon hydrophobic interaction of the
complexes with HSA, of a clathrate-like, second-coordination-
sphere arrangement of water molecules also contribute to the
relaxivity enhancement of the complex−protein adduct.49,59
Hence, we may further speculate that the less favorable local
rotational dynamics is offset by the greater contribution of these
mechanisms.
Conversely, at pH = 8.3, at which the affinity of the
nonprotonated complex for HSA is high, the relaxivity
enhancement is markedly limited. Even in this case, it is difficult
to associate this effect only with differences in the rotational
dynamics because the modest relaxivity gain is rather uniform
throughout the frequency range investigated. A series of factors
could be involved, including (i) a partial decrease in the number
of hydration q due to displacement of the water molecule by
donor atoms (likely carboxylate) of the protein,60 (ii)
lengthening of the exchange lifetime τM at the binding site due
to a more ordered arrangement of water molecules in the
complex−protein contact region,49 and (iii) a negligible
b
contribution to r1 due to the second-coordination-sphere
mechanism and to exchange of the mobile protons of the
protein.
obs
In summary, the pronounced variation of R1 with the pH
from 7 to 4 in the presence of HSA is mainly due to variation in
the affinity of the complex for the protein (Kaff), which is
reflected in a marked difference in the populations of the free
and bound complex. The remarkable value of r1pb at neutral pH
appears as a result of the lack of a limiting effect of a relatively
relatively long τM and of a good motional coupling between
global rotation and local motion at the binding site, favored by
the presence of two interacting adjacent benzyl groups and by
the short length of the linker.
Caravan and co-workers have reported an interesting example
of the gadolinium(III) complex of a DOTA derivative,
containing a biphenylsulfonamide arm, known for its ability to
coordinate the gadolinium(III) ion in a pH-dependent
manner.61 At acidic pH solutions, the complex exhibits q = 2
and high relaxivity, while deprotonation occurs with an increase
in the pH, which decreases the hydration state to q = 0, thereby
decreasing the relaxivity. The variation in r1p with the pH is
greater than that observed in the present study because it is
caused by an increase in q from 0 to 2. However, the
corresponding variation of the relaxivity with the pH is
attenuated in the presence of HSA because of displacement of
the coordinated water molecules when the complex binds to the
protein.61 The remarkable positive aspect of the GdIII-H4L1
complex is that an increase in the relaxivity with an increase in
EXPERIMENTAL SECTION
■
General Information. The chemicals and solvents were used as
purchased from commercial sources (Sigma-Aldrich, Fluka, Strem,
CheMatech, or Penta). Dry pyridine was bought from Acros. The
compound 1,4,7-tris[(tert-butoxycarbonyl)methyl]cyclen hydrobro-
mide (tBu3DO3A·HBr) was prepared as previously reported.62
Deionized water (Millipore) was used throughout the work. NMR
spectra were recorded on VNMRS300 (7.0 T), Varian UNITY INOVA
400 (9.4 T), Bruker Avance III 400 (9.4 T), Bruker Avance III 500
(11.7 T), Bruker Avance III 600 (14.3 T), and Magritech Spinsolve 43
K
DOI: 10.1021/acs.inorgchem.9b00267
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
MHz (0.94 T) spectrometers using 5 mm sample tubes. NMR chemical
shifts are expressed as parts per million, and coupling constants are
reported in hertz. Unless stated otherwise, all NMR spectra were
collected at 25 °C. For 1H and 13C{1H} NMR measurements in D2O,
tBuOH was used as an internal standard [δ(1H) = 1.25 ppm; δ(13C) =
30.3 ppm]. For 31P NMR spectra, 85% aqueous H3PO4 (0.0 ppm) was
used in an insert NMR tube. MS spectra were acquired on a Bruker
ESQUIRE 3000 spectrometer with electrospray ionization and ion-trap
detection. Thin-layer chromatography (TLC) was performed on silica
60 F254 or aluminum oxide 60 F254 aluminum-backed TLC sheets
(Merck), with detection by UV (254 nm), by dipping into a 10%
aqueous CuSO4 solution coupled with mild heating, or by iodine
vapors. Elemental analyses were performed at the Institute of
Macromolecular Chemistry (Prague, Czech Republic). pD values
were determined using an electrode system calibrated with standard
buffers, and the pD was read after the following correction: pD = pH +
(bm, 4H, H2C3), 3.49 (s, 2H, N−CH2−CO2H), 3.59−3.70 (bm, 2H,
H2C6), 3.81−3.99 (m, 4H, 2 × N−CH2−CO2H), 4.45 (bs, 4H, N−
CH2−Ph), 7.47−7.60 (m, 10H, Ph). 13C{1H} NMR (D2O, pD = 3.9): δ
48.8 (C6), 50.4 (d, 3JCP = 3.5, C2), 51.1 (C3), 52.2 (C5), 52.4 (d, 1JCP
=
82.3, P−CH2−NBn2), 53.9 (d, 1JCP = 105.1, Nring−CH2−P), 54.0 (1 ×
N−CH2−CO2H), 57.8 (2 × N−CH2−CO2H), 58.6 (N−CH2−Ph),
129.7 (Ph), 129.9 (Ph), 130.7 (Ph), 132.0 (Ph), 171.1 (2 × N−CH2−
CO2H), 175.0 (1 × N−CH2−CO2H). 31P{1H} NMR (D2O, pD = 3.9):
δ 25.0. MS(+): 672.2 (calcd 672.3, [M + K]+), 694.2 (calcd 694.2, [M +
Na + K − H]+), 710.2 (calcd 710.2, [M + 2K − H]+). MS(−): 670.0
(calcd 670.2, [M − 2H + K]−), 708.0 (calcd 708.2, [M − 3H + 2K]−),
730.0 (calcd 730.2, [M − 4H + 2Na + K]−). Elem anal. Found (calcd
for C30H50N5O11P, H4L1·3H2O): C, 52.09 (52.39); H, 7.15 (7.33); N,
10.20 (10.18). TLC (SiO2, 1:5 concentrated aqueous NH3/EtOH): Rf
= 0.3.
Crude Ester Intermediate 3. 31P{1H} NMR (pyridine): 48.4.
MS(+): 830.5 (calcd 830.5, [M + H]+), 852.5 (calcd 852.5, [M + Na −
H]+). MS(−): 800.2 (calcd 800.5, [M − Et]−). TLC (SiO2, EtOH): Rf
= 0.8.
1
0.41. H NMRD profiles 1/T1 were measured on a fast-field-cycling
Stelar SmartTracer relaxometer (Mede, Pv, Italy) over a continuum of
magnetic field strengths from 0.00024 to 0.25 T (corresponding to
0.01−10 MHz proton Larmor frequencies). The temperature was
controlled with a Stelar VTC-91 airflow heater equipped with a
calibrated copper−constantan thermocouple (uncertainty of 0.1 K).
Additional data points in the range of 20−70 MHz were obtained on a
Stelar relaxometer equipped with a Bruker WP80 NMR electromagnet.
Noncovalent interactions between the complex and HSA were
evaluated using the proton relaxation enhancement method: R1 was
monitored as a function of the amount of HSA at 298 K and 20 MHz.
Details on the synthesis of compounds 1 and 2 and other
experimental information (more details on X-ray diffraction studies,
experimental for equilibrium studies, NMR titrations, high-resolution
UV−vis absorption spectra, excitation lifetime measurements, fluo-
rescence measurements, and DIS 17O NMR measurements) are given
in the SI.
Crude Ester Intermediate 4. 31P{1H} NMR (H2O): 53.4 (s).
MS(+): 684.3 (calcd 684.3, [M + Na]+), 700.9 (calcd 700.3, [M + K]+).
TLC (SiO2, 1:5 concentrated aqueous NH3/EtOH): Rf = 0.6.
Synthesis of Lanthanide(III) Complexes of H4L1. The ligand
H4L1·3H2O (75 mg, 0.11 mmol), in slight excess (1.05 equiv), and
LnCl3·6H2O (Ln = Ce−Lu, excluding Pm) or LaCl3·7H2O were
dissolved in water (2 mL). The solution pH was maintained at ∼5.5 by
the periodic addition of diluted aqueous NaOH. After pH stabilization
(∼30 min), the mixture was stirred and heated to 60 °C for 3 days.
Then, the volatiles were removed in vacuo, and the oil was
chromatographed on neutral Al2O3 (20 mL) with 10:1:2 iPrOH/
concentrated aqueous NH3/water. Then, 7.5 mL fractions were
collected; those with pure complex (TLC on alumina, Rf = 0.3, and
the same eluent) were combined, and volatiles were removed in vacuo,
producing ammonium salts of the complexes as yellowish oils (60−80
mg). For NMR measurements, the samples of complexes were
dissolved in water (ccomplex ∼ 60 mM), and the solution pH was
adjusted by diluted aqueous HCl or aqueous NaOH to pH ∼ 3
(protonated complex) or 8 (deprotonated complex). Characterization
data are given in Table S1. To prepare solid samples, solutions
containing pure complexes were concentrated in vacuo, and the
residues were dissolved in a small amount of MeOH. Off-white powder
precipitated upon the addition of excess Et2O, which was then filtered
off (S4), washed with Et2O, and dried on air.
Synthesis of H4L1. tBu3DO3A·HBr (3.00 g, 5.04 mmol, 1.1 equiv)
and paraformaldehyde (2.04 g, 68.1 mmol, 15 equiv) were added to the
pyridine solution (20 mL) of compound 2 [for the synthesis of 2, see
the SI; prepared from 1.25 g of 129 (4.54 mmol, 1 equiv)] in a 25 mL
round-bottom flask, which was then tightly closed with a stopper. The
mixture was stirred at 30 °C for 1 day and thereafter at 40 °C for 2 days
for full conversion (31P NMR). The excess of paraformaldehyde was
filtered off, and the volatiles were evaporated in vacuo. The oily residue
was dissolved in diethyl ether (Et2O; 50 mL), hexane was added until
cloudy (∼5 mL), and the mixture was left to stand overnight at room
temperature. Then, the nonreacted excess of tBu3DO3A and some
N,N′-dicyclohexylurea were filtered off. The filtrate was evaporated to
dryness in vacuo, and the oily residue of full ester 3 was used directly in
another reaction. Crude 3 was dissolved in trifluoroacetic acid/CHCl3
(1:1, v/v, 50 mL), and the solution was refluxed (75 °C) overnight.
Then, the solution was cooled to ambient temperature, and the volatiles
were removed in vacuo to give P-ethyl ester 4 as an oil, which was used
directly in the next step. Crude ester 4 was dissolved in aqueous
pyridine (∼75% py, v/v, 100 mL), and the mixture was heated at 75 °C
overnight (until the reaction was complete; 31P NMR). Then, the
mixture was concentrated in vacuo, and the residue was adsorbed on a
cation exchanger (Dowex 50, 300 mL, H+ form). The column was
washed with water (2 column volumes), EtOH (∼3 column volumes),
and water (1 column volume), and ligand H4L1 was eluted off with 10%
aqueous pyridine. The solvents were removed in vacuo, and the crude
oily product was triturated in boiling THF (150 mL). The reddish
powder was filtered off, dried in air, and then triturated in hot azeotropic
EtOH (150 mL) by ultrasonication (5 min). Pure H4L1 was filtered off,
washed with Et2O, and dried in air, resulting in a pure white powder of
H4L1·3H2O (2.17 g, 68% based on tBu3DO3A·HBr). A single crystal of
the ligand, suitable for X-ray diffraction, was obtained by the slow
diffusion of THF vapor into an aqueous solution of the zwitterionic
H4L1 (pH ∼ 4) at room temperature (Figure S1).
X-ray Diffraction. Single crystals of H4L1·7.5H2O were grown by
THF vapor diffusion into an aqueous solution of the ligand. Single
crystals of (NH4)[Gd(H2O)(DO3APDBAm)]·3H2O were grown by
acetone vapor diffusion into a solution of the complex just after
chromatographic purification. The solution was prepared by mixing
equimolar amounts of the ligand and GdCl3, followed by neutralization
with aqueous ammonia (pH = 7 and 60 °C, overnight) and purification
by chromatography on neutral alumina with 10:1:2 iPrOH/
concentrated aqueous NH3/water. Single crystals of Na[Dy(H2O)-
(DO3APDBAm)]·4H2O were grown by acetone vapor diffusion into a
solution of the complex prepared by mixing equimolar amounts of the
ligand and DyCl3 neutralized with aqueous NaOH (1 M) to pH = 7. In
addition, single crystals of tBu3DO3A·HCl·CHCl3 (CCDC 1888344)
were also obtained during the recovery of excess tBu3DO3A in the
H4L1 synthesis (above), and its solid-state structure was determined.
More details are given in the SI.
Selected crystals were mounted on a glass fiber in a random
orientation, and diffraction data were collected at 150 K (Cryostream
Cooler, Oxford Cryosystem) on a Nonius Kappa CCD diffractometer
equipped with a Bruker APEX-II CCD detector using monochromat-
ized Mo Kα radiation (λ = 0.71073 Å). Data were analyzed using the
SAINT V8.27B (Bruker AXS Inc., 2015) software package. Data were
corrected for absorption effects using the multiscan method (SADABS).
All structures were solved by direct methods (SHELXS97)63 and
refined using full-matrix least-squares techniques (SHELXL2014).64
For further information, refer to the SI. Table S2 contains selected
crystallographic parameters of the structures reported in this paper. The
1H NMR (D2O, pD = 3.9; for atom labeling, see the crystal structure
of H4L1 in the SI): δ 3.02 (d, 2JHP = 9.3, 2H, Nring−CH2−P), 2.96−3.06
(bm, 2H, H2C6), 3.06−3.19 (bm, 4H, H2C2), 3.31−3.41 (bm, 8H,
2
H2C3 + H2C5), 3.46 (d, JHP = 7.3, 2H, P−CH2−NBn2), 3.42−3.55
L
DOI: 10.1021/acs.inorgchem.9b00267
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
structures H4L1·7.5H2O, (NH4)[Gd(H2O)(DO3APDBAm)]·3H2O,
and Na[Dy(H2O)(DO3APDBAm)]·4H2O were deposited at the Cam-
bridge Crystallographic Data Centre as CCDC 1888345−1888347,
respectively.
ACKNOWLEDGMENTS
■
This study was funded by the Grant Agency of Charles
University (Grant 1170317 to P.U.) and by the Ministry of
Education of the Czech Republic (Grant LTC 17607 to P.U.,
J.K., and P.H.). M.B. and F.C. thank Universita del Piemonte
Orientale for financial support (Ricerca locale 2016). The
17O NMR. The stock solution of the GdIII-H4L1 complex after
purification by chromatography on alumina was diluted with water to a
concentration of ∼25 mM of the complex. The solution pH (4.0 and
8.3) was adjusted with diluted aqueous HCl/aqueous NaOH, and the
exact concentration of gadolinium(III) was determined using the Evans
method.65 Then, this solution (188 μL) was mixed with H217O (10%
17O, 10 μL) and 0.1% tBuOH in D2O (22 μL). The reference was
prepared by mixing distilled water (188 μL) and H217O (10% 17O, 10
μL) with 0.1% tBuOH in D2O (22 μL). The measurements were
performed on a 500 MHz Bruker Avance III spectrometer in the 2−77
°C range (5 °C intervals). Transverse relaxation rates were calculated
from line widths at half-height.
̀
́
̌
́
authors thank Dr. I. Cısarova for X-ray data acquisition and Z.
Bohmova for titration experiments. The work was performed in
the framework of the EU COST CA15209 Action.
́
̈
REFERENCES
■
́
́
(1) Toth, E.; Helm, L.; Merbach, A. In The Chemistry of Contrast
Agents in Medical Magnetic Resonance Imaging, 2nd ed.; Merbach, A.,
́
́
Relaxometric Measurements. The stock solution of the GdIII-
H4L1 complex was diluted with water to ∼4 mM (the exact
concentration was determined using the Evans method). Then, the
solution pH was adjusted with diluted aqueous NaOH or aqueous HCl
by adding a few microliters into 1.00 mL of the sample solution. The
change in concentration was thus negligible. 1H NMRD measurements
were performed in a magnetic field corresponding to 0.01−70 MHz 1H
resonance frequencies at 10, 25, and 37 °C. GdIII-H4L1 complex
measurements with HSA were performed at pH = 4.0, 7.0, and 8.3 (20
MHz and 25 or 37 °C), and the solutions were also used to acquire the
1H NMRD data. Titrations of the GdIII-H4L1 complex with solid HSA
and in the presence of an inhibitor (ibuprofen or warfarin;
approximately 5 equiv over GdIII-H4L1) were performed at 20 MHz
(25 °C).
Helm, L., Toth, E., Eds.; Wiley: Chichester, U.K., 2013; pp 25−82.
(2) The Chemistry of Contrast Agents in Medical Magnetic Resonance
́
́
Imaging, 2nd ed.; Merbach, A., Helm, L., Toth, E., Eds.; Wiley:
Chichester, U.K., 2013.
̌
̌
(3) Hermann, P.; Kotek, J.; Kubícek, J.; Lukes, I. Gadolinium(III)
complexes as MRI contrast agents: ligand design and properties of the
complexes. Dalton Trans. 2008, 3027−3047.
(4) (a) Zhou, Z.; Lu, Z.-R. Gadolinium-based contrast agents for
magnetic resonance cancer imaging. WIREs Nanomed. Nanobiotechnol.
2013, 5, 1−18. (b) Li, Y.; Beija, M.; Laurent, S.; Elst, L. V.; Muller, R.
N.; Duong, H. T. T.; Lowe, A. B.; Davis, T. P.; Boyer, C.
Macromolecular Ligands for Gadolinium MRI Contrast Agents.
Macromolecules 2012, 45, 4196−4204.
(5) (a) Liu, Y. J.; Zhang, N. Gadolinium loaded nanoparticles in
theranostic magnetic resonance imaging. Biomaterials 2012, 33, 5363−
5375. (b) Zhang, L.; Liu, R.; Peng, H.; Li, P.; Xu, Z.; Whittaker, A. K.
The evolution of gadolinium based contrast agents: from single-
modality to multi-modality. Nanoscale 2016, 8, 10491−10510.
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.9b00267.
́
(c) Yang, C.-T.; Padmanabhan, P.; Gulyas, B. Z. Gadolinium(III)
based nanoparticles for T1-weighted magnetic resonance imaging
probes. RSC Adv. 2016, 6, 60945−60966. (d) Cao, Y.; Xu, L.; Kuang,
Y.; Xiong, D.; Pei, R. Gadolinium-based nanoscale MRI contrast agents
for tumor imaging. J. Mater. Chem. B 2017, 5, 3431−3461.
(6) Chan, K. W.-Y.; Wong, W.-T. Small molecular gadolinium(III)
complexes as MRI contrast agents for diagnostic imaging. Coord. Chem.
Rev. 2007, 251, 2428−2451.
(7) Caravan, P. Protein-targeted gadolinium-based magnetic reso-
nance imaging (MRI) contrast agents: Design and mechanism of
action. Acc. Chem. Res. 2009, 42, 851−862.
(8) (a) Fanali, G.; di Masi, A.; Trezza, V.; Marino, M.; Fasano, M.;
Ascenzi, P. Human serum albumin: From bench to bedside. Mol.
Aspects Med. 2012, 33, 209−290. (b) Yamasaki, K.; Chuang, V. T. G.;
Maruyama, T.; Otagiri, M. Albumin-drug interaction and its clinical
implication. Biochim. Biophys. Acta, Gen. Subj. 2013, 1830, 5435−5443.
(c) Liu, Z.; Chen, X. Simple bioconjugate chemistry serves great clinical
advances: albumin as a versatile platform for diagnosis and precision
therapy. Chem. Soc. Rev. 2016, 45, 1432−1456.
Additional synthetic details, experimental data for X-ray
diffraction studies, additional crystal structures, NMR
titration data and discussion, potentiometric experimental
data and distribution diagrams, along with the results and
discussion, characterization data of lanthanide(III)
complexes and their NMR data and abundance at
different pH values, UV−vis and fluorescence data of
europium(III) complexes and discussion, 17O NMR
relaxation data, competitive fluorescence, and NMR
experiments in the presence of HSA (PDF)
Accession Codes
CCDC 1888344−1888347 contain the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge via www.ccdc.cam.ac.uk/data_request/cif, or by email-
ing data_request@ccdc.cam.ac.uk, or by contacting The Cam-
bridge Crystallographic Data Centre, 12 Union Road, Cam-
bridge CB2 1EZ, UK; fax: +44 1223 336033.
(9) (a) Parmelee, D. J.; Walovitch, R. C.; Ouellet, H. S.; Lauffer, R. B.
Preclinical evaluation of the pharmacokinetics, biodistribution, and
elimination of MS-325, a blood pool agent for magnetic resonance
imaging. Invest. Radiol. 1997, 32, 741−747. (b) Lauffer, R. B.; Parmelee,
D. J.; Dunham, S. U.; Ouellet, H. S.; Dolan, R. P.; Witte, S.; McMurry,
T. J.; Walovitch, R. C. MS-325: albumin-targeted contrast agent for MR
angiography. Radiology 1998, 207, 529−538.
AUTHOR INFORMATION
■
Corresponding Author
́
(10) Muller, R. N.; Raduchel, B.; Laurent, S.; Platzek, J.; Pierart, C.;
̈
*E-mail: petrh@natur.cuni.cz. Tel.: +420-221951263. Fax:
+420-221951253.
Mareski, P.; Vander Elst, L. Physicochemical characterization of MS-
325, a new gadolinium complex, by multinuclear relaxometry. Eur. J.
Inorg. Chem. 1999, 1999, 1949−1955.
ORCID
(11) Aime, S.; Chiaussa, M.; Digilio, G.; Gianolio, E.; Terreno, E.
Contrast agents for magnetic resonance angiographic applications: 1H
and 17O NMR relaxometric investigations on two gadolinium(III)
DTPA-like chelates endowed with high binding affinity to human
serum albumin. JBIC, J. Biol. Inorg. Chem. 1999, 4, 766−774.
Jan Kotek: 0000-0003-1777-729X
Petr Hermann: 0000-0001-6250-5125
Notes
The authors declare no competing financial interest.
M
DOI: 10.1021/acs.inorgchem.9b00267
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
́
(12) Caravan, P.; Cloutier, N. J.; Greenfield, M. T.; McDermid, S. A.;
Dunham, S. U.; Bulte, J. W. M.; Amedio, J. C., Jr.; Looby, R. J.;
Supkowski, R. M.; Horrocks, W. D., Jr.; McMurry, T. J.; Lauffer, R. B.
The interaction of MS-325 with human serum albumin and its effect on
proton relaxation rates. J. Am. Chem. Soc. 2002, 124, 3152−3162.
(13) Eldredge, H. B.; Spiller, M.; Chasse, J. M.; Greenwood, M. T.;
Caravan, P. Species dependence on plasma protein binding and
relaxivity of the gadolinium-based MRI contrast agent MS-325. Invest.
Radiol. 2006, 41, 229−243.
(14) Zech, S. G.; Eldredge, H. B.; Lowe, M. P.; Caravan, P. Protein
binding to lanthanide(III) complexes can reduce the water exchange
rate at the lanthanide. Inorg. Chem. 2007, 46, 3576−3584.
(15) Caravan, P.; Parigi, G.; Chasse, J. M.; Cloutier, N. J.; Ellison, J. J.;
Lauffer, R. B.; Luchinat, C.; McDermid, S. A.; Spiller, M.; McMurry, T.
J. Albumin binding, relaxivity, and water exchange kinetics of the
diastereoisomers of MS-325, a gadolinium(III)-based magnetic
resonance angiography contrast agent. Inorg. Chem. 2007, 46, 6632−
6639.
Bioconjugate Chem. 2006, 17, 975−987. (b) Rudovsky, J.; Hermann, P.;
̌
Botta, M.; Aime, S.; Lukes, I. Dendrimeric Gd(III) complex of a
monophosphinated DOTA analogue: optimizing relaxivity by reducing
internal motion. Chem. Commun. 2005, 2390−2392.
́
̌
(25) Kotkova, Z.; Helm, L.; Kotek, J.; Hermann, P.; Lukes, I.
Gadolinium complexes of monophosphinic acid DOTA derivatives
conjugated to cyclodextrin scaffolds: efficient MRI contrast agents for
higher magnetic fields. Dalton Trans. 2012, 41, 13509−13519.
(26) Wanwimolruk, S.; Birkett, D. J.; Brooks, P. M. Structural
requirements for drug binding to site II on human serum albumin. Mol.
Pharmacol. 1983, 24, 458−463.
(27) Yamasaki, K.; Chuang, V. T. G.; Maruyama, T.; Otagiri, M.
Albumin-drug interaction and its clinical implication. Biochim. Biophys.
Acta, Gen. Subj. 2013, 1830, 5435−5443.
(28) Lukes, I.; Kotek, J.; Vojtísek, P.; Hermann, P. Complexes of
tetraazacycles bearing methylphosphinic/phosphonic acid pendant
arms with copper(II), zinc(II) and lanthanides(III). A comparison with
their acetic acid analogues. Coord. Chem. Rev. 2001, 216−217, 287−
312.
̌
̌
(16) Helm, L.; Morrow, J. R.; Bond, C. J.; Carniato, F.; Botta, M.;
Braun, M.; Baranyai, Z.; Pujales-Paradela, R.; Regueiro-Figueroa, M.;
̌
́
(29) Kotek, J.; Lebduskova, P.; Hermann, P.; Vander Elst, L.; Muller,
R. N.; Geraldes, C. F. G. C; Maschmeyer, T.; Lukes, I.; Peters, J. A.
́
̌
Esteban-Gomez, D.; Platas-Iglesias, C.; Scholl, T. J. In Contrast Agents
for MRI. Experimental Methods; Pierre, V. C., Allen, M. J., Eds.; Royal
Society of Chemistry: Croydon, U.K., 2018; Chapter 2, pp 121−242.
Lanthanide(III) complexes of novel mixed carboxylic-phosphorus acid
derivatives of diethylenetriamine: A step towards more efficient MRI
contrast agents. Chem. - Eur. J. 2003, 9, 5899−5915.
(30) Meyer, M.; Dahaoui-Gindrey, V.; Lecomte, C.; Guilard, R.
Conformations and coordination schemes of carboxylate and
carbamoyl derivatives of the tetraazamacrocycles cyclen and cyclam,
and the relation to their protonation states. Coord. Chem. Rev. 1998,
178−180, 1313−1405.
(31) Taborsky, P.; Lubal, P.; Havel, J.; Kotek, J.; Hermann, P.; Lukes,
I. Thermodynamic and kinetic studies of lanthanide(III) complexes
with H5do3ap (1,4,7,10-tetraazacyclododecane-1,4,7-triacetic-10-
(methylphosphonic acid)), a monophosphonate analogue of H4dota.
Collect. Czech. Chem. Commun. 2005, 70, 1909−1942.
(32) Pniok, M.; Kubícek, V.; Havlíckova, J.; Kotek, J.; Sabatie-Gogova,
A.; Plutnar, J.; Huclier-Markai, S.; Hermann, P. Thermodynamic and
kinetic study of scandium(III) complexes of DTPA and DOTA: A step
toward scandium radiopharmaceuticals. Chem. - Eur. J. 2014, 20, 7944−
7955.
́ ́
(17) Brucher, E.; Tircso, G.; Baranyai, Z.; Kovacs, Z.; Sherry, A. D. In
̈
The Chemistry of Contrast Agents in Medical Magnetic Resonance Imaging,
́
́
2nd ed.; Merbach, A., Helm, L., Toth, E., Eds.; Wiley: Chichester, U.K.,
2013; pp 157−208.
́
̌
(18) Rudovsky, J.; Cígler, P.; Kotek, J.; Hermann, P.; Vojtísek, P.;
Lukes, I.; Peters, J. A.; Vander Elst, L.; Muller, R. N. Lanthanide(III)
̌
́
́
̌
complexes of a mono(methylphosphonate) analogue of H4dota: The
influence of protonation of the phosphonate moiety on the TSAP/SAP
isomer ratio and the water exchange rate. Chem. - Eur. J. 2005, 11,
2373−2384.
́
̌
(19) Rudovsky, J.; Kotek, J.; Hermann, P.; Lukes, I.; Mainero, V.;
Aime, S. Synthesis of a bifunctional monophosphinic acid DOTA
analogue ligand and its lanthanide(III) complexes. A gadolinium(III)
complex endowed with an optimal water exchange rate for MRI
applications. Org. Biomol. Chem. 2005, 3, 112−117.
̌
̌
́
́
́
̌
́
́
(20) Lebduskova, P.; Hermann, P.; Helm, L.; Toth, E.; Kotek, J.;
́
̌
̈
́
́
́
́
Binnemans, K.; Rudovsky, J.; Lukes, I.; Merbach, A. E. Gadolinium(III)
complexes of mono- and diethyl esters of mono-phosphonic acid
analogue of DOTA as potential MRI contrast agents: solution
structures and relaxometric studies. Dalton Trans. 2007, 493−501.
(21) Vitha, T.; Kubícek, V.; Kotek, J.; Hermann, P.; Vander Elst, L.;
Muller, R. N.; Lukes, I.; Peters, J. A. Gd(III) complex of a
(33) Forsterova, M.; Svobodova, I.; Lubal, P.; Taborsky, P.; Kotek, J.;
Hermann, P.; Lukes, I. Thermodynamic study of lanthanide(III)
̌
complexes with bifunctional monophosphinic acid analogues of H4dota
and comparative kinetic study of yttrium(III) complexes. Dalton Trans
2007, 535−549.
(34) Cacheris, W. P.; Nickle, S. K.; Sherry, A. D. Thermodynamic
study of lanthanide complexes of 1,4,7-triazacyclononane-N,N’,N“-
triacetic acid and 1,4,7,10-tetraazacyclododecane-N,N’,N”,N’’’-tetra-
acetic acid. Inorg. Chem. 1987, 26, 958−962.
(35) Caravan, P.; Comuzzi, C.; Crooks, W.; McMurry, T. J.; Choppin,
G. R.; Woulfe, S. R. Thermodynamic stability and kinetic inertness of
MS-325, a new blood pool agent for magnetic resonance imaging. Inorg.
Chem. 2001, 40, 2170−2176.
(36) Djanashvili, K.; Peters, J. A. How to determine the number of
inner-sphere water molecules in lanthanide(III) complexes by 17O
NMR spectroscopy. A technical note. Contrast Media Mol. Imaging
2007, 2, 67−71.
̌
̌
monophosphinate-bis(phosphonate) DOTA analogue with a high
relaxivity: Lanthanide(III) complexes for imaging and radiotherapy of
calcified tissues. Dalton Trans. 2009, 3204−3214.
́
̌
(22) (a) Polasek, M.; Caravan, P. Is macrocycle a synonym for kinetic
inertness in Gd(III) complexes? Effect of coordinating and non-
coordinating substituents on inertness and relaxivity of Gd(III) chelates
with DO3A-like ligands. Inorg. Chem. 2013, 52, 4084−4096. (b) Boros,
E.; Karimi, S.; Kenton, N.; Helm, L.; Caravan, P. Gd(DOTAlaP):
Exploring the boundaries of fast water exchange in gadolinium-based
magnetic resonance imaging contrast agents. Inorg. Chem. 2014, 53,
6985−6994.
(23) (a) Dumas, S.; Jacques, V.; Sun, W.-C.; Troughton, J. S.; Welch, J.
T.; Chasse, J. M.; Schmitt-Willich, H.; Caravan, P. High relaxivity
magnetic resonance imaging contrast agents. Part 1: Impact of single
donor atom substitution on relaxivity of serum albumin-bound
gadolinium complexes. Invest. Radiol. 2010, 45, 600−612. (b) Jacques,
V.; Dumas, S.; Sun, W.-C.; Troughton, J. S.; Greenfield, M. T.; Caravan,
P. High-relaxivity magnetic resonance imaging contrast agents. Part 2:
Optimization of inner- and second-sphere relaxivity. Invest. Radiol.
2010, 45, 613−624.
(37) Lowe, M. P.; Parker, D.; Reany, O.; Aime, S.; Botta, M.;
Castellano, G.; Gianolio, E.; Pagliarin, R. pH-Dependent modulation of
relaxivity and luminescence in macrocyclic gadolinium and europium
complexes based on reversible intramolecular sulfonamide ligation. J.
Am. Chem. Soc. 2001, 123, 7601−7609.
(38) (a) Banci, L.; Bertini, I.; Luchinat, C. Nuclear and Electron
Relaxation. The Magnetic Nucleus-Unpaired Electron Coupling in
Solution; VCH: Weinheim, Germany, 1991. (b) Bloembergen, N.;
Morgan, L. O. Proton relaxation times in paramagnetic solutions.
Effects of electron spin relaxation. J. Chem. Phys. 1961, 34, 842−850.
(39) Freed, J. H. Dynamic effects of pair correlation functions on spin
relaxation by translational diffusion in liquids. II. Finite jumps and
independent T1 processes. J. Chem. Phys. 1978, 68, 4034−4037.
́
(24) (a) Rudovsky, J.; Botta, M.; Hermann, P.; Hardcastle, K. I.;
̌
Lukes, I.; Aime, S. PAMAM dendrimeric conjugates with a Gd-DOTA
phosphinate derivative and their adducts with polyaminoacids: The
interplay of global motion, internal rotation, and fast water exchange.
N
DOI: 10.1021/acs.inorgchem.9b00267
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(40) Peters, J. A.; Djanashvili, K.; Geraldes, C. F. G. C.; Platas-Iglesias,
C. In The Chemistry of Contrast Agents in Medical Magnetic Resonance
benzylphosphinate group. J. Chem. Soc., Dalton Trans. 1997, 3623−
3636.
́
(55) Aime, S.; Botta, M.; Fasano, M.; Marques, M. P. M.; Geraldes, C.
F. G. C.; Pubanz, D.; Merbach, A. E. Conformational and coordination
equilibria on DOTA complexes of lanthanide metal ions in aqueous
́
Imaging, 2nd ed.; Merbach, A., Helm, L., Toth, E., Eds.; Wiley:
Chichester, U.K., 2013; pp 209−276.
(41) Botta, M. Second coordination sphere water molecules and
relaxivity of gadolinium(III) complexes: Implications for MRI contrast
agents. Eur. J. Inorg. Chem. 2000, 2000, 399−407.
(42) Kotkova, Z.; Pereira, G. A.; Djanashvili, K.; Kotek, J.; Rudovsky,
J.; Hermann, P.; Vander Elst, L.; Muller, R. N.; Geraldes, C. F. G. C.;
1
solution studied by H-NMR spectroscopy. Inorg. Chem. 1997, 36,
2059−2068.
̌
(56) Blahut, J.; Hermann, P.; Tosner, Z.; Platas-Iglesias, C. A
́
́
combined NMR and DFT study of conformational dynamics in
lanthanide complexes of macrocyclic DOTA-like ligands. Phys. Chem.
Chem. Phys. 2017, 19, 26662−26671.
̌
Lukes, I.; Peters, J. A. Lanthanide(III) complexes of phosphorus acid
analogues of H4DOTA as model compounds for the evaluation of the
second-sphere hydration. Eur. J. Inorg. Chem. 2009, 2009, 119−136.
(43) Swift, T. J.; Connick, R. E. J. NMR (nuclear magnetic
resonance)-relaxation mechanisms of 17O in aqueous solutions of
paramagnetic cations and the lifetime of water molecules in the first
coordination sphere. J. Chem. Phys. 1962, 37, 307−312.
(57) (a) Mato-Iglesias, M.; Platas-Iglesias, C.; Djanashvili, K.; Peters,
́
́
J. A.; Toth, E.; Balogh, E.; Muller, R. N.; Vander Elst, L.; de Blas, A.;
Rodriguez-Blas, T. The highest water exchange rate ever measured for a
Gd(III) chelate. Chem. Commun. 2005, 4729−4731. (b) Balogh, E.;
́
́
Mato-Iglesias, M.; Platas-Iglesias, C.; Toth, E.; Djanashvili, K.; Peters, J.
A.; de Blas, A.; Rodríguez-Blas, T. Pyridine- and phosphonate-
containing ligands for stable Ln complexation. Extremely fast water
exchange on the GdIII chelates. Inorg. Chem. 2006, 45, 8719−8728.
(44) Powell, D. H.; Dhubhghaill, O. M. N.; Pubanz, D.; Helm, L.;
Lebedev, Ya. S.; Schlaepfer, W.; Merbach, A. E. Structural and dynamic
parameters obtained from 17O NMR, EPR, and NMRD studies of
monomeric and dimeric Gd3+ complexes of interest in magnetic
resonance imaging: An integrated and theoretically self-consistent
approach. J. Am. Chem. Soc. 1996, 118, 9333−9346.
́
(58) Zhang, L.; Martins, A. F.; Zhao, P.; Wu, Y.; Tircso, G.; Sherry, A.
D. Lanthanide-based T2ex and CEST complexes provide insights into
the design of pH sensitive MRI agents. Angew. Chem., Int. Ed. 2017, 56,
16626−16630.
́
(45) Rudovsky, J.; Botta, M.; Hermann, P.; Koridze, A.; Aime, S.
(59) Liepinsh, E.; Otting, G. Proton exchange rates from amino acid
side chains - implications for image contrast. Magn. Reson. Med. 1996,
35, 30−42.
Relaxometric and solution NMR structural studies on dimeric
lanthanide(III) complexes of a phosphinate analogue of DOTA with
a fast rate of water exchange. Dalton Trans. 2006, 2323−2333.
(46) Ghuman, J.; Zunszain, P. A.; Petitpas, I.; Bhattacharya, A. A.;
Otagiri, M.; Curry, S. Structural basis of the drug-binding specificity of
human serum albumin. J. Mol. Biol. 2005, 353, 38−52.
(60) Aime, S.; Gianolio, E.; Terreno, E.; Giovenzana, G. B.; Pagliarin,
R.; Sisti, M.; Palmisano, G.; Botta, M.; Lowe, M. P.; Parker, D. Ternary
Gd(III)L-HSA adducts: evidence for the replacement of inner-sphere
water molecules by coordinating groups of the protein. Implications for
the design of contrast agents for MRI. JBIC, J. Biol. Inorg. Chem. 2000, 5,
488−497.
(47) (a) Fanali, G.; Ascenzi, P.; Fasano, M. Effect of prototypic drugs
ibuprofen and warfarin on global chaotropic unfolding of human serum
heme-albumin: A fast-field-cycling 1H-NMR relaxometric study.
Biophys. Chem. 2007, 129, 29−35. (b) Peters, T. All about Albumin:
Biochemistry, Genetics, and Medical Applications; Academic Press:
Orlando, FL, 1995; Chapter 2, pp 9−75.
(48) Er, J. C.; Vendrell, M.; Tang, M. K.; Zhai, D.; Chang, Y.-T.
Fluorescent dye cocktail for multiplex drug-site mapping on human
serum albumin. ACS Comb. Sci. 2013, 15, 452−457.
(49) Aime, S.; Botta, M.; Fasano, M.; Crich, S. G.; Terreno, E. Gd(III)
complexes as contrast agents for magnetic resonance imaging: a proton
relaxation enhancement study of the interaction with human serum
albumin. JBIC, J. Biol. Inorg. Chem. 1996, 1, 312−319.
(61) Moriggi, L.; Yaseen, M. A.; Helm, L.; Caravan, P. Serum albumin
targeted, pH-dependent magnetic resonance relaxation agents. Chem. -
Eur. J. 2012, 18, 3675−3686.
(62) Moore, D. A. Selective trialkylation of cyclen with tert-butyl
bromoacetate [1,4,7,10-tetraazacyclododecane-1,4,7-triacetic acid, tri-
tert-butyl ester hydrobromide]. Org. Synth. 2008, 85, 10−14.
(63) (a) Sheldrick, G. M. SHELX97. Program for Crystal Structure
̈
̈
Solution from Diffraction Data; University of Gottingen: Gottingen,
Germany, 1997. (b) Sheldrick, G. M. A short history of SHELX. Acta
Crystallogr., Sect. A: Found. Crystallogr. 2008, 64, 112−122.
(64) Sheldrick, G. M. SHELXL-2014/7. Program for Crystal Structure
́
́
̌
́
(50) Prochazkova, S.; Kubícek, V.; Kotek, J.; Vagner, A.; Notni, J.;
Hermann, P. Lanthanide(III) complexes of monophosphinate/mono-
phosphonate DOTA-analogues: effects of the substituents on the
formation rate and radiolabelling yield. Dalton Trans. 2018, 47, 13006−
13015.
̈
̈
Refinement from Diffraction Data; University of Gottingen: Gottingen,
Germany, 2014.
(65) Corsi, D. M.; Platas-Iglesias, C.; van Bekkum, H.; Peters, J. A.
Determination of paramagnetic lanthanide(III) concentrations from
bulk magnetic susceptibility shifts in NMR spectra. Magn. Reson. Chem.
2001, 39, 723−726.
̌
́
(51) Vojtísek, P.; Cígler, P.; Kotek, J.; Rudovsky, J.; Hermann, P.;
̌
Lukes, I. Crystal structures of lanthanide(III) complexes with cyclen
derivative bearing three acetate and one methylphosphonate pendants.
Inorg. Chem. 2005, 44, 5591−5599.
́
̌
(52) Kotek, J.; Rudovsky, J.; Hermann, P.; Lukes, I. Three in one:
TSA, TSA’, and SA units in one crystal structure of a yttrium(III)
complex with a monophosphinated H4dota analogue. Inorg. Chem.
2006, 45, 3097−3102.
(53) Lima, L. M. P.; Esteves, C. V.; Delgado, R.; Hermann, P.; Kotek,
̌
̌
́
J.; Sevcíkova, R.; Lubal, P. Tris(phosphonomethyl) cyclen derivatives:
Synthesis, acid-base properties and complexation studies with Cu2+ and
Zn2+ ions. Eur. J. Inorg. Chem. 2012, 2012, 2533−2547.
(54) (a) Aime, S.; Batsanov, A. S.; Botta, M.; Howard, J. A. K.; Parker,
D.; Senanayake, K.; Williams, J. A. G. Solution and solid-state
characterization of highly rigid, eight-coordinate lanthanide(III)
complexes of a macrocyclic tetrabenzylphosphinate. Inorg. Chem.
1994, 33, 4696−4706. (b) Aime, S.; Batsanov, A. S.; Botta, M.; Dickins,
R. S.; Faulkner, S.; Foster, C. E.; Harrison, A.; Howard, J. A. K.;
Moloney, J. M.; Norman, T. J.; Parker, D.; Royle, L.; Williams, J. A. G.
Nuclear magnetic resonance, luminescence and structural studies of
lanthanide complexes with octadentate macrocyclic ligands bearing
O
DOI: 10.1021/acs.inorgchem.9b00267
Inorg. Chem. XXXX, XXX, XXX−XXX