High denticity oxinate-linear-backbone chelating ligand for diagnostic radiometal ions [111In]In3+and [89Zr]Zr4+

Article abstract of DOI:10.1039/d0dt04230g

Advances in nuclear medicine depend on chelating ligands that form highly stable and kinetically inert complexes with relevant radiometal ions for use in diagnosis or therapy. A new potentially decadentate ligand, H₅decaox, was synthesised to i

Full text of DOI:10.1039/d0dt04230g

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
High denticity oxinate-linear-backbone
chelating ligand for diagnostic radiometal ions
Cite this: Dalton Trans., 2021, 50,
3874
[111In]In3+ and [89Zr]Zr4+†‡
b
Lily Southcott, a,b Xiaozhu Wang, a Luke Wharton, a,b Hua Yang,
d
Valery Radchenko,b,c Manja Kubeil, d Holger Stephan,
María de Guadalupe Jaraquemada-Peláez *a and Chris Orvig
*
a
Advances in nuclear medicine depend on chelating ligands that form highly stable and kinetically inert
complexes with relevant radiometal ions for use in diagnosis or therapy. A new potentially decadentate
ligand, H5decaox, was synthesised to incorporate two 8-hydroxyquinoline moieties on either end of a di-
ethylenetriamine backbone decorated with three carboxylic acids, one at each N atom of the backbone.
Metal complexation was assessed using nuclear magnetic resonance (NMR) spectroscopy and high-
resolution mass spectrometry (HR-MS) with In3+, Zr4+ and La3+. Solution thermodynamic studies provided
the stepwise protonation constants and metal formation constants, indicating a high affinity for both In3+
and Zr4+ (pIn = 32.3 and pZr = 34.7), and density functional theory (DFT) calculations provided insight
into the coordination environments with either metal ion. Concentration dependent radiolabeling experi-
ments with [111In]InCl3 and [89Zr]ZrCl4 showed promise as quantitative radiolabeling (>95%) occurred at
micromolar concentrations, under mild, near-physiological conditions of pH 7 and room temperature for
30 minutes. Serum stability of both radiometal complexes was investigated and the [111In]In(decaox)
complex remained 91% intact after 24 hours while the [89Zr]Zr(decaox) complex was 86% intact over the
same time, comparable to other chelating ligands previously assessed with the same methods. The high
radiolabeling yields, limited serum protein transchelation and structural insight of the [89Zr]Zr(decaox)
complex suggest a promising fit between the oxinate-containing ligand and the Zr4+ ion, setting the stage
for further investigations with a functionalised version of the chelator for its potential in PET imaging.
Received 12th December 2020,
Accepted 16th February 2021
DOI: 10.1039/d0dt04230g
rsc.li/dalton
particles, or Auger electrons.1–3 As the accessibility to new iso-
topes is burgeoning with more and better cyclotrons, as well as
Introduction
Nuclear medicine comprises diagnosis and treatment of generators and other production methods being implemented,
disease by exploiting the decay properties of radioactive the type and number of isotopes available is dramatically
nuclides. Radionuclides decay in different modalities, giving increasing. To harness the diagnostic or therapeutic ability of
rise to diagnostic techniques based on γ-detection, positron a radioactive isotope, it must be safely administered and deli-
emission tomography (PET) and single photon emission com- vered to its biological target, otherwise unwanted adverse
puted tomography (SPECT), and therapies based on β− or α effects become prominent. Central to development in metallo-
radiopharmaceuticals is the coordinating ligand that is
designed to chelate a radiometal ion to form a stable complex
aMedicinal Inorganic Chemistry Group, Department of Chemistry, University of
that will not dissociate once administered.2
British Columbia, 2036 Main Mall, Vancouver, British Columbia V6T 1Z1, Canada.
Chelators are often designed specifically for the metal ion
E-mail: orvig@chem.ubc.ca, mdgjara@chem.ubc.ca
of interest – the size, charge and decay properties of the radio-
metal ion itself will influence the stability and inertness of the
overall coordination complex.4 When comparing ligands for
stable metal complexes in radiopharmaceuticals, a balance
between rapid complexation and kinetic inertness is highly
desired. One of the most commonly used in metalloradiophar-
maceuticals is the macrocyclic chelator DOTA (1,4,7,10-tetra-
azacyclododecane-1,4,7,10-tetraacetic acid) (Fig. 1), which forms
bLife Sciences Division, TRIUMF, 4004 Wesbrook Mall, Vancouver, British Columbia,
V6T 2A3, Canada
cDepartment of Chemistry, University of British Columbia, 2036 Main Mall,
Vancouver, British Columbia V6T 1Z1, Canada
dInstitute of Radiopharmaceutical Cancer Research, Helmholtz-Zentrum Dresden
Rossendorf, Bautzner Landstrasse 400, 01328 Dresden, Germany
†Invited for the 50th volume of Dalton Trans.
‡Electronic supplementary information (ESI) available. See DOI: 10.1039/
d0dt04230g
3874 | Dalton Trans., 2021, 50, 3874–3886
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
Fig. 1 Chemical structures of selected chelators and complexes.
highly inert complexes but often requires harsh radiolabeling provide interesting radiometal isotopes for their respective
conditions (high temperatures, long incubation times) that are imaging modalities, SPECT and PET.3,4 111In]In3+ is regularly
[
not suitable for bioconjugates.5,6 DTPA (diethylenetriamine cyclotron-produced, widely commercially available and has
1
pentaacetic acid) is an exemplary acyclic chelator with five car- useful properties for SPECT imaging, including a half-life (t )
2
boxylic acid moieties attached through the linear diethyl- of 2.8 days, and average Eγ = 171 and 245 keV.4 Interest in
enetriamine backbone. It radiolabels metals quickly (under developing PET probes with [89Zr]Zr4+ has increased dramati-
15 minutes) in mild conditions (room temperature (RT)); cally in the last few years as its suitability for antibody target-
1
however, it has inferior in vivo stability compared to the macro- ing radiopharmaceuticals is realised: a mid-range half-life (t =
2
cycle DOTA.5 While both ligands or their targeting derivatives 78 h, long for a positron emitter) and low energy (Eβ+ = 396
are in standard use, combining the rapid complexation of keV) plus availability and ease of production on medical
DTPA with in vivo stability of DOTA would yield a novel chelator cyclotrons.3,28,29
of significant importance to metalloradiopharmaceuticals. To
Combining the idea of non-macrocyclic chelators for rapid
achieve such a balance between fast complexation and in vivo complexation and high kinetic inertness with the oxinate
inertness, non-macrocylic ligands with rigid functional groups moiety found in ionophoric complexes, several multidentate
or backbones have recently been a focus when combined with ligands containing oxine groups have recently been reported
8-hydroxyquinoline as a chelating moiety.7–11
by us – including H2hox,7 H4octox,8 H4bispox,9 H3glyox,10 and
8-Hydroxyquinoline (8-HQ, oxine) is a small, bicyclic mole- H2CHXhox,11 which are six- to eight-coordinating ligands that
cule that has been used extensively in analytical chemistry,12,13 radiolabel at high efficiencies with various metal ions includ-
medicinal chemistry and nuclear medicine.14–17 As an imaging ing [68Ga]Ga3+, [111In]In3+, and [177Lu]Lu3+. To exploit the ver-
agent in nuclear medicine, metal–oxinate complexes typically satility of the ‘ox’ family of ligands and to accommodate larger
have been used as ionophores: lipophilic, low denticity and metal ions with higher coordination numbers, a new, oxine-
meta-stable complexes to allow the transport of a metal ion containing ligand was conceived based upon the diethyl-
into cells. The most widely used example is the tris ligand enetriamine backbone of DTPA, and the utility of our pre-
complex, [111In]In(oxinate)3 (Fig. 1),18–21 which has been used viously reported H5decapa.30 Incorporating the bicycle oxine
for decades to radiolabel leukocytes for SPECT imaging of as arms for the diethylenetriamine backbone could be ben-
inflammation and infection. Similarly reported are M(oxinate) eficial to enhance the rigidity of the ligand and in turn
complexes of various radiometals: [52Mn]Mn(oxinate)2,22 [68Ga] improve the stability of its metal complexes.
Ga(oxine)3,23 and [89Zr]Zr(oxinate)4.24,25 [89Zr]Zr(oxinate)4 has
The potentially decadentate ligand, H5decaox (Fig. 1), was
been used increasingly in mouse models for tracking of cell- synthesised and its utility was tested by selecting metal ions by
based therapies,24 liposomes,26 and bone marrow27 – this charge (3+ vs. 4+) and size to identify any coordination prefer-
leads to applications in PET imaging of cancer, arthritis and ences, consistent with the availability of relevant radioisotopes.
acute bone injury response. Both [111In]In3+ and [89Zr]Zr4+ In3+ is a metal ion with borderline hardness, while the smaller
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 3874–3886 | 3875

View Article Online
Paper
Dalton Transactions
Zr4+ prefers hard, anionic donors, allowing for a comparison fied synthetic procedures and optimised yield. This resulted in
to be made upon complexation with H5decaox to understand a ligand similar to the acyclic DTPA, exchanging the two flex-
its coordination preferences. Further, with increasing interest ible carboxylic acid moieties on either end for more rigid,
31
in radiometals for therapy, [225Ac]Ac3+ (t = 9.9 days, α-decay)
8-hydroxyquinoline arms. Compared to previous ‘ox’ ligands,
1
2
and nonradioactive surrogate La3+ were also chosen to investi- this is the highest denticity example, and the most compli-
gate the potential of the large backbone of H5decaox to accom- cated to synthesise.
modate larger metal ions.
Metal complexation
Metal complexation of H5decaox with In(ClO4)3, La(NO3)3, and
ZrCl4 was achieved and characterised with 1H NMR, 1H–1H
correlation spectroscopy (COSY) and high resolution mass
Results and discussion
Synthesis and characterisation
spectrometry (HR-MS). All complexes were confirmed by
Starting from diethylenetriamine, H5decaox was synthesised in HR-MS, but the inherent insolubility of the complexes limited
five steps with an initial nosyl protection step and alkylation complete structural characterisation. The 1H NMR of In
with tert-butyl bromoacetate (Scheme 1). After the nosyl de- (decaox) indicates a metal complex species with one oxinate
protection via thiophenol, the product was alkylated with the arm metal-bound and one oxine unbound (Fig. 2). This is
synthesised building block, 2-(bromomethyl)quinoline-8-ol. evinced by five sharp, resolved peaks in the aromatic region
Final ligand deprotection proceeded with 6 M HCl and the (7.58–9.12 ppm) representing the metal bound 8-hydroxy-
desired product was purified via reverse-phase C18 automated quinolinate, while the remaining small, broad aromatic peaks
column chromatography. Although the overall yield is only (6.66–7.09 ppm) correspond to the more fluxional unbound
8.4%, efforts to improve yields and solubility with a hydroxyl arm. The proton correlations of the five bound aromatic
protecting group for 8-hydroxyquinoline derivatives were protons are clear in the COSY spectrum (Fig. 2C), while the
attempted (acetate, tert-butyldimethylsilyl, benzene sulfonyl- remaining peaks of the unbound arm are found overlapping at
oxy). While the protecting groups improved ligand solubility, 7.58 ppm, indicating the two distinct hydroxyquinoline arm
they ultimately impeded the succeeding reactions or were environments and accounting for all expected protons in the
cleaved during alkylation steps, leading to a mixture of par- aromatic region. Further, the diastereotopic splitting of the
tially protected products and more complicated purification methylene protons of the acetate arms and the additional
steps. Proceeding without protecting groups and continuing to splitting of the diethylene backbone protons provides evidence
the final H5decaox deprotection without purification simpli- of one isomer with minimised fluxional interconversion as
Scheme 1 H5decaox synthesis, reagents and conditions (i) Na2CO3, THF, 0 °C → RT, 18 h, 52%, (ii) K2CO3, CH3CN, reflux, 24 h, 96%, (iii) K2CO3,
CH3CN, reflux, 24 h, 56%, (iv) NaBH4, MeOH, 0 °C → RT, 18 h, 75%, (v) PBr3, CH3CN, 0 °C, 5 h, 74%, (vi) K2CO3, CH3CN, reflux, 24 h and (vii) 6 M HCl,
THF, 50 °C, 18 h (30% over 2 steps).
3876 | Dalton Trans., 2021, 50, 3874–3886
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
Fig. 2 (A) 1H NMR spectrum (D2O, RT, 300 MHz) of H5decaox and (B) 1H NMR spectrum (D2O, RT, 400 MHz) In(decaox) and (C) partial 1H–1H COSY
NMR spectrum (D2O, RT, 400 MHz) of In(decaox).
opposed to multiple species (Fig. 2B). The complexity of the coordinated, this structure presents a decreased rigidity and is
NMR spectra of the Zr4+ and La3+ complexes (Fig. S16–S19†) lower in energy; however, lower kinetic inertness is predicted
suggests species present with much greater fluxionality and due to the exceedingly accessible capped carboxylate oxygen in
spectral complexity. NMR cannot fully characterise the Zr4+ the bipyramidal structure.
and La3+ complexes due to precipitation, as noted with other
For the Zr4+ complexes, two minimum energy confor-
ligands of high coordination number and DTPA-like back- mations were found: seven-coordinated (N2O5) and eight-co-
bones, limiting the structural characterization of said com- ordinated (N3O5) Zr4+ environments (Fig. 3A and S32†), with
plexes in solution.30,32
only a 12.5 kJ mol−1 difference in free energy. The more stable
conformation, [Zr(decaox)]− A, is the seven coordinated Zr4+
ion with a capped octahedral coordination environment.
Density functional theory calculations
In the absence of X-ray crystal structures, DFT calculations Although only seven-coordinate, isomer A is lower in energy,
were undertaken to rationalise the ligand–metal connectivity likely due to the shorter bond distances of the bound oxygen
(Fig. 3). The calculations provide optimised structures showing to the Zr4+ ion (Table S3†).
seven and eight-coordinated Zr4+ ions, plus seven-coordinated
Two possible conformations were found for the La3+ metal
In3+ ions and ten-coordinated La3+ ions. Of particular interest complexes, both with N5O5 coordination environments saturat-
are the seven-coordinated In3+ ions: the more stable isomer ing the coordination sphere with the decadentate denticity of
[In(decaox)]2− B presents a pentagonal bipyramidal coordination the ligand (Fig. S33†). Calculations including a coordinated
environment (N3O4) and has one oxine pendant uncoordinat- water molecule as seen in the crystal structure of the octaden-
ing, consistent with the observations in the H NMR solution tate chelator H4octox,8 showed it was impossible to coordinate
1
study (Fig. 3B and S31†). With only one hydroxyquinoline arm any water molecules even with the inclusion of a second
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 3874–3886 | 3877

View Article Online
Paper
Dalton Transactions
Fig. 3 Optimised structures of the (A) [Zr(decaox)]− A and (B) [In(decaox)]2− B complex anions obtained with DFT calculations. Bond distances and
free energies in Tables S2 and S3.†
sphere of explicit water. It appears that H5decaox, having H5decapa. The second and third protonation equilibria (log K2
shorter overall bond distances of either the three acetate O–M = 10.65 (1) and log K3 = 9.92 (1), species H2L3− and H3L2−
)
and the two oxine O–M than [La(octox)]−, encapsulates La3+ involve the protonation of the oxinate oxygens (OHox) and are
better and prevents the coordination of a water molecule.
consistent with the same functional group protonations in
H4octox. Protonation of the two remaining tertiary amines
(log K4 = 9.12 (1) and log K5 = 7.32 (1); species H4L− and H5L)
are in the same order of magnitude as in H5decapa. The last
three protonation equilibria are attributed to the carboxylic
acid substituents (log K6 = 3.64 (1), log K7 = 2.63 (1) and log K8
= 1.56 (1)), species H6L+, H7L2+ and H8L3+. Protonation of the
quinolinium nitrogens was not observed in the experimental
conditions but might occur below pH 0.2 (Fig. S23A†).
Solution thermodynamics
To assess the stability of complexes with various metal ions,
solution thermodynamic studies can be used to quantify the
chelator’s ionizable protons and complex formation equilibria.
Combined UV-potentiometric titrations were performed to
identify the protonation constants and species present over a
pH range of 0.2–11.4 of H5decaox. Analysis of the potentio-
metric and spectrophotometric data with the HyperQuad,33
HypSpec2014,34 and Hyss35 computer programmes yielded the
protonation constants in Table 1, the molar absorptivities of
Complex formation equilibria of H5decaox with In3+, Zr4+ and
La3+
eight absorbing species, and the speciation diagram of Upon determining the protonation constants of the free
H5decaox (Fig. S24A and S24B†). ligand, similar approaches (titrations and in-batch studies) can
As H5decaox was designed with the oxinate moiety of be used to assess thermodynamically the stability of the
H4octox8 and the diethylenetriamine backbone of H5decapa,30 metal–ligand complexes and determine if such complexes will
useful comparisons among protonation constants of the three be suitable in physiological applications. This stability is con-
ligands can be made (Table 1). The first protonation constant veyed via the formation constant, log K, which represents the
(log K1 = 11.0 (1), species HL4−) can be assigned to the central thermodynamic driving force for complex formation through
tertiary amine in the backbone and is similar to the log K1 in the sum of equilibrium equations of each step, expressed as:
Table 1 Protonation constants (log Kq) of discussed ligands
Equilibrium reaction
H5decaox
H5decapac
H4octoxd
L + H+ ⇆ HL
11.00 (1)a (Nbackbone
)
11.03 (3) (Nbackbone
)
10.65 (1) (OHox
10.02 (1) (OHox
9.03 (1) (Nbackbone
5.18 (1) (Nbackbone
3.05 (1) (COOH)
2.03 (2) (COOH)
)
)
HL + H+ ⇆ H2L
H2L + H+ ⇆ H3L
H3L + H+ ⇆ H4L
H4L + H+ ⇆ H5L
H5L + H+ ⇆ H6L
H6L + H+ ⇆ H7L
H7L + H+ ⇆ H8L
∑log K {[HqL]/[Hq-1L][H+]}
10.65 (1)a (OHox
)
9.20 (3) (Nbackbone
6.86 (4) (Nbackbone
)
)
9.92 (1)a (OHox
)
)
)
9.12 (1)a (Nbackbone
7.32 (1)a (Nbackbone
3.64 (1)a(COOH)
2.63 (1)a (COOH)
1.56 (1)b (COOH)
55.80 (1)
)
)
4.43 (4) (py-COOH)
3.46 (5) (py-COOH)
2.84 (6) (COOH)
2.52 (4) (COOH)
ND
−0.31 (8) (Nox
−0.67 (7) (Nox
38.98 (7)
)
)
40.34 (4)
a Using UV-potentiometric titrations. b Using UV batch titration, T = 25 °C, I = 0.16 M NaCl. c From ref. 30. d From ref. 8. Charges omitted for
simplicity.
3878 | Dalton Trans., 2021, 50, 3874–3886
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
pM + qH + rL ↔ MpHqLr (charges omitted for clarity), repre- species (Table 2). The metal complex speciation (Mn+ = Zr4+,
senting metal ion (M) and ligand (L) association; protons (H+) In3+ and La3+) was determined with the calculated stability
here are also important as they compete with metals for occu- constants (Fig. 4). It is interesting to note that despite the
pation of a chelate atom electron pair. Further, the pM value is similar log KML for the Zr4+ and In3+ complexes, pZr is 2.4
another way to demonstrate the metal sequestering ability of a units higher than is pIn because of the pKa of the neutral
ligand in terms of free metal ion concentration under standard species, Zr(Hdecaox) (log KMHL = 9.39 (3)) is higher than in
conditions (pM = −log([M]) when [L] = 10 µM, [M] = 1 µM, pH [In(Hdecaox)]− (log KMHL = 7.53 (2)), translating into less proton
7.4, T = 25 °C).36 Complex formation equilibria studies of competition at pH 7.4. At physiological pH the single species
H5decaox were undertaken with natural stable isotopes of In3+, Zr(Hdecaox) predominates, which will be highly advantageous
La3+ and Zr4+.
for in vitro or in vivo assays and medicinal applications.
All three metal complexes are mononuclear and present Finally, a comparison of the state-of-art of Zr4+ and In3+ chela-
different protonation degrees (species MH3L, MH2L, MHL, tors can be made in terms of pM values (Table 3). While
ML, M(OH)L and additional MH4L species for La3+ – charges H5decaox compares amongst the highest pM values for In3+
omitted for simplicity – Table 2). From the spectroscopic fea- and Zr4+ chelating ligands, the kinetic inertness of these metal
tures of the oxine chromophores in these experiments and chelates remains reliable evidence for potential in vivo appli-
compared with other well-studied analogous metal complexes cations and these promising thermodynamic stability results
with either La3+ or In3+ and Ga3+,7,8,10,11,37 we can assume that are essential for further in vitro/in vivo studies.
the higher deprotonations of either MH2L and MHL species
correspond to the phenol–OH in the oxine moieties. The most
Radiolabeling and serum stability
acidic protonations are attributed either to tertiary amines in
Based on the nonradioactive metal complexation, solution
the backbone or the carboxylic substituents; however, insolubi-
studies and available isotopes, radiolabeling with both [111In]
lity limited structural elucidation with either NMR or X-ray
InCl3 and [89Zr]ZrCl4 proceeded in a concentration dependent
manner. Since the attempted radio-TLC plates and conditions
crystallography.
H5decaox has a higher affinity for the smaller In3+ and Zr4+
gave inconsistent results for [111In]InCl3 radiolabeling, radio-
HPLC was used to determine the radiochemical yield during
concentration dependent studies with both [111In]InCl3 and
[89Zr]ZrCl4 (Fig. 5 and Table S1†). With [111In]InCl3 (3 MBq per
reaction), radiolabeling was quantitative with ligand concen-
trations 100, 10 and 1 µM and 89% at 0.1 µM in NaOAc buffer
(0.1 M, pH 7) at room temperature for 30 minutes.
Radiolabeling with [89Zr]ZrCl4 (2 MBq per reaction) under the
same conditions (NaOAc buffer, 30 min, RT) showed similar
results for 100 and 10 µM, but RCY dropped off to 89% at
1 µM. Additionally, the distribution coefficient of the [89Zr]Zr
(decaox) complex at pH 7.4 was investigated using a two-phase
extraction system from 1-octanol/aqueous buffer to understand
the hydrophilic nature of the complex. The log D7.4 was deter-
mined to be −0.822 0.004, a reasonably hydrophilic complex
comparable to the recently reported DFO2 (−0.71) although
more lipophilic than DFO itself (−2.70).42 Finally, radiolabel-
ions, indicated by the La3+ metal complex stability log KML
being almost half the log value for the corresponding [Zr(L)]−
Table 2 Stability constants (log KML) and corresponding stepwise pro-
tonation constants log K1q1(MHqL)a of H5decaox with In3+, La3+ and Zr4+
(T = 25 °C, I = 0.16 M NaCl)
Zr4+
In3+
La3+
log K101 (ML)
log K111 (MHL)
log K121 (MH2L)
log K131 (MH3L)
log K141 (MH4L)
log K1–11 (M(OH)L)
pM
43.06 (3)
9.39 (3)
5.13 (2)
3.53 (2)
—
9.44 (3)
34.7
42.32 (2)
7.53 (2)
5.02 (3)
2.40 (1)
—
10.51 (3)
32.3
24.6 (2)
9.17 (1)
7.84 (1)
5.12 (2)
3.64 (2)
10.66 (2)
16.5
a K1q1 = [MHqL]/[MHq−1L][H]q; (q − 1) = −1 denotes OH.
Fig. 4 Speciation plots of H5decaox metal complexes: (A) Zr4+, (B) In3+, and (C) La3+ ; dashed line indicates pH 7.4.
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 3874–3886 | 3879

View Article Online
Paper
Dalton Transactions
Table 3 pMa values of selected chelating ligands (Mn+ = In3+, Zr4+
)
Table 4 Serum stability of selected chelators with [111In]In3+
pIn
pZr
1 h stability
(%)
24 h stability
(%)
5 day stability
(%)
Complex
H5decaox
8-Hydroxyquinoline
DTPA
H4octox
H3glyox
H4pypa
H4octapa
DOTA
32.3b
34.130
25.730
25.08
3410
H5decaox
DFO
DTPA
3,4,3-LI(1,2-HOPO)
THPN
AHA
34.7b
32.238
33.938
44.039
42.840
28.938
[
[
[
[
[
[
111In][In(decaox)]2−
111In][In(octox)]− a
111In][In(decapa)]2− b
111In][In(octapa)]− b
111In][In(DOTA)]− b
111In][In(DTPA)]2− b
97.3 0.4
96.2 0.4
89.7 1.6
93.8 3.6
89.6 2.1
86.5 2.2
91
1
63
6
91.4 0.6
89.1 1.7
92.3 0.04
89.4 2.2
88.43 2.2
83.6 1.4
ND
ND
ND
ND
30.541
26.530
18.830
a From ref. 8. b From ref. 30.
a pM = −log[Metal]free, [L] = 10 μM, [M] = 1 μM, pH = 7.4.36 b This work.
methodology, including
[
111In][In(decapa)]2−
,
[
111In][In
ing with the larger, therapeutic radiometal ion, [225Ac]Ac3+ was
unsuccessful in varying conditions: 0.1 M or 1 M NaOAc, pH 7,
RT and 80 °C. This corroborates with the solution studies with
La3+ (cold metal surrogate of 225Ac3+) that suggested weak
thermodynamic stability of the [La(decaox)]2− complex,
especially in comparison to the In3+ and Zr4+ complexes.
Following successful radiolabeling with both [111In]InCl3
and [89Zr]ZrCl4, the kinetic inertness of the resulting com-
plexes towards human serum proteins was investigated in a
protein challenge experiment using either PD-10 size exclusion
(DOTA)]− and [111In][In(DTPA)]2− (Table 4). However, the five-
day stability is only compared to that of [In(octox)]−, which
indicates the smaller oxinate ligand H4octox may be better
suited for [111In]In3+ (Table 4). The DFT calculated structure
indicating decreased rigidity with only one hydroxyquinoline
bound and labile points from the lack of encapsulation near
the carboxylic acid provides an explanation of the susceptible
[
111In]In3+ ion towards serum proteins.
After 24 hours [89Zr]Zr(decaox) remained 86% intact, indi-
desalting columns43 with
[
111In]In(decaox) complex, and
cating the initial decomplexation is common in both com-
plexes. Over 7 days, [89Zr]Zr(decaox) remained 75% intact, indi-
cating a higher stability than for [111In]In(decaox), and slightly
less stable when compared to DFO and DFO-type ligands using
the same method (Table 5).42 Interestingly, the [89Zr]Zr(decaox)
complex is significantly more stable than is [89Zr]Zr(decapa),44
which incorporates picolinic acid moieties in lieu of the oxi-
nates found in H5decaox. This indicates that the rigidity of the
bicyclic oxine groups has had the desired improved stability in
this case compared to picolinic acid ligands. The observed
decrease in stability in both [111In][In(decaox)]2− and [89Zr][Zr
protein precipitation with cold acetonitrile42 for the [89Zr]Zr
(decaox) complex. Many endogenous proteins found in human
serum, such as apo-transferrin and albumin, are capable of
competing for metal ion binding, and a successful radio-
metal–ligand complex must withstand transchelation towards
such proteins. For the
[
111In]In(decaox) complex, within
24 hours a slight decrease to 91% intact was observed, which
continued to slowly decomplex to only 63% intact complex at
the final five-day time point (Fig. 5B). This 24-hour decrease is
similar to previously reported ligands following the same
Fig. 5 (A) Concentration dependent radiolabeling of H5decaox with [111In]In3+ (black) and [89Zr]Zr4+ (red) in NaOAc (0.1 M, pH 7), 30 min, RT, n = 2,
monitored by radio-HPLC. (B) Serum stability of [111In]In(decaox) and [89Zr]Zr(decaox), n = 3.
3880 | Dalton Trans., 2021, 50, 3874–3886
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
Table 5 Serum stability of selected chelators with [89Zr]Zr4+
received unless otherwise indicated. Aluminum-backed ultra-
pure silica gel 60 Å, 250 µm thickness analytical TLC plates
were used for reaction monitoring. Flash column chromato-
graphy was performed using silica gel sourced from Silicycle
(Siliaflash irregular silica gels F60, 60 Å pore size, 40–63 mm
particle size). Automated purification was performed using a
Teledyne Isco CombiFlash Rf automated system, using RediSep
Rf Gold normal-phase silica, neutral alumina, and reverse-
phase C18 re-useable column cartridges. 1H, 13C{1H} and
COSY NMR spectra were measured on Bruker AV300 and AV400
spectrometers at ambient temperature unless otherwise stated
and interpreted using Mestrenova spectral visualization soft-
ware. Low-resolution mass spectrometry was carried out using
a Waters liquid-chromatography – mass-spectrometer system
consisting of a Waters ZQ quadrupole spectrometer coupled to
an ESCI ion source and a Waters 2695 HPLC system. HRMS
was performed on a Waters Micromass LCT – TOF instrument
at the University of British Columbia. Microanalyses for C, H,
and N were performed on a Carlo Erba elemental analyzer EA
1108. [111In]InCl3 was produced and purchased from BWX
Technologies as a 0.05 M HCl solution. [225Ac]Ac3+ was pro-
vided by Canadian Nuclear Laboratories (CNL) from decay of
229Th and separation from 233U. Upon arrival, [225Ac]Ac3+ was
purified by branched DGA resin column to remove chemical
impurities and re-constituted in 0.01 M HNO3. [89Zr]Zr4+ was
produced from the [89Y(p,n)89Zr] reaction at either HZDR or
University of Alabama Birmingham as [89Zr]ZrCl4 or [89Zr]Zr
(oxalate), respectively. Radiolabeling progression was moni-
tored using radioHPLC with a Knauer Smartline System con-
sisting of Smartline 1000 pump, K2501 UV detector, Raytest
Gabi Star activity detector, Chromgate 2.8 software and a
Smartline 5000 manager with a Phenomenex Synergi 4-micron
Hydro/HP 80 A 250 × 4.6 mm or an Agilent 1260 Infinity II
Quaternary Pump, Agilent 1260 Autosampler, Raytest Gabi Star
NaI(Tl) radiation detector, Agilent 1260 variable wavelength
detector, with a Phenomenex Luna C18 analytical column (4.6
× 250 mm, 5 µm).
1 day stability 3 day stability 7 day stability
Complex
(%)
(%)
(%)
[
[
[
[
89Zr][Zr(decaox)]−
86.3 0.1
86
1
74.6 0.1
ND
87
81
89Zr][Zr(decapa)]− a ∼26
∼17
∼96
∼83
89Zr][Zr(DFO2)]− b
89Zr][Zr(DFO)]− b
∼93
∼80
1
4
a From ref. 44. b From ref. 42.
(decaox)]− may also be due to non-specific protein interactions
rather than decomplexation, considering the hydrophobic
nature of the complexes; this has been reported with several,
similar lipophilic M–oxine complexes.45–47 Further, in vitro
serum stability assessments are not always predictive the
in vivo behaviour of metal–ligand complexes, and overall, the
mild and rapid radiolabeling conditions while remaining com-
parably stable over 24 hours suggest the promise of the
H5decaox new chelating ligand radiopharmaceutical
development.
Conclusions and outlook
A new potentially decadentate, oxine-containing ligand was
synthesised from the diethylenetriamine backbone. This
ligand has been assessed by both radioactive and non-radio-
active metal complexation studies with various metal ions
([nat/111In]In3+; [nat/89Zr]Zr4+; natLa3+/[225Ac]Ac3+). The increased
denticity and cavity size of the ligand compared to previously
reported ‘ox’ ligands can accommodate larger metal ions or
different charged ions, as indicated in the high thermo-
dynamic stability constants with In3+ and Zr4+. Mild radiolabel-
ing conditions led to high radiochemical yields at micromolar
concentrations with both [111In]InCl3 and [89Zr]ZrCl4 and the
intact complex after 24 hours in the presence of serum pro-
teins indicates H5decaox as a promising chelate for radiometal
isotopes. DFT calculations complement the solution thermo-
dynamics and serum stability, evinced in the seven-coordinate
complexes with Zr4+ and In3+, having N2O5 and N3O4 atom
environments, respectively. H5decaox appears to be extremely
well-suited for the harder metal ion Zr4+, with shorter Zr–O
bond lengths, and a better match for the hard donor hydroxy-
quinoline units. Further, the central carboxylic acid moiety
presents an interesting accessible position for functionali-
zation to imbue a targeted, bifunctional chelator with similar
capabilities; synthesis and characterisation of a functionalised
version of this highly promising ligand is underway and will
be assessed as a bioconjugate.
Synthesis and characterisation
8-Hydroxy-2-carbaldehyde (1). Compound 1 was prepared
according to the literature with appropriate characterisation
and spectra.7,16
2-(Hydroxymethyl)quinolin-8-ol (2). Compound 1 (499 mg,
2.88 mmol) was dissolved in MeOH (20 mL) and cooled to 0 °C
in an ice-water bath. NaBH4 (220 mg, 5.76 mmol) was added to
the stirring solution, which was subsequently warmed to RT
and stirred overnight. The reaction mixture was quenched with
H2O (20 mL), acidified with 1 M HCl (5 mL) and extracted with
CH2Cl2 (3 × 25 mL). The organic phases were combined and
dried over Na2SO4 and evaporated to give an off white solid
1
(381 mg, 75%). H NMR (CDCl3, 400 MHz) δ 8.19 (d, 1H, J =
Experimental section
Materials and methods
8.5 Hz), 7.48 (t, 1H, J = 7.9 Hz), 7.40 (m, 2H, J = 8.5 Hz), 7.25
(d, 1H, J = 7.6 Hz), 5.00 (s, 2H). 13C{1H} NMR (CDCl3, 75 MHz)
All reagents and solvents were purchased from commercial δ 137.3, 127.6, 119.4, 118.2, 111.1, 64.9. LR-ESI-MS: calcd for
suppliers (Sigma-Aldrich, AKScientific, TCI) and used as [C10H9NO2]: 175.1; found: [M + H]+: 176.3
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 3874–3886 | 3881

View Article Online
Paper
Dalton Transactions
2-(Bromomethyl)quinoline-8-ol (3). Compound 2 (217 mg, solved in water (50 mL) and extracted with EtOAc (3 × 50 mL).
1.24 mmol) was dissolved in dry CH2Cl2 (12 mL) and cooled to The organic phases were combined, washed with sat. NaHCO3
0 °C in an ice-water bath. PBr3 (128 µL, 1.36 mmol) dissolved (30 mL), dried over MgSO4 and the solvents evaporated in
in dry CH2Cl2 (1 mL) was added dropwise to the solution with vacuo. The crude material was purified via flash column
stirring. The reaction mixture was stirred at 0 °C for 5 h before (Al2O3) chromatography (Combiflash automated purification
sat. NaHCO3 (20 mL) was added to quench. The reaction system, A: Hexanes, B: EtOAc; 100% A to 20% B) to give a
1
mixture was extracted with CH2Cl2 (3 × 25 mL), and the com- yellow oil (432 mg, 56%). H NMR (CDCl3, 100 MHz) δ 3.33 (s,
bined organic phases were dried over Na2SO4 and evaporated 2H), 3.31 (s, 4H), 2.80 (t, 4H, J = 6.0 Hz), 2.66 (t, 4H, J = 6.0
to give the title compound as a light-yellow residue (218 mg, Hz), 1.45 (s, 18 H), 1.44 (s, 9H). 13C{1H} NMR (CDCl3,
74%). 1H NMR (CDCl3, 400 MHz) δ 8.19 (d, 1H, J = 8.5 Hz), 100 MHz) δ 171.6, 171.0, 81.1, 80.9, 55.8, 54.1, 51.6, 47.3, 28.3,
7.61 (d, 1H, J = 8.5 Hz), 7.49 (t, 1H, J = 7.8 Hz), 7.35 (d, 1H, J = 28.2. LR-ESI-MS: calcd for [C22H43N3O6]: 445.32; found [M +
8.2 Hz), 7.22 (d, 1H, J = 8.6 Hz), 4.73 (s, 2H). 13C{1H} NMR H]+: 446.3.
(CDCl3, 75 MHz) δ 137.5, 128.4, 122.0, 117.8, 110.8, 33.9.
LR-ESI-MS: calcd for [C10H8BrNO]: 236.98; found [M(79Br) + solved in dry CH3CN and the flask purged with N2. K2CO3
H]+: 238.1, and [M(81Br) + H]+: 240.2.
(130 mg, 0.94 mmol) was added to the solution. Compound 3
H5decaox (7). Compound 6 (101 mg, 0.23 mmol) was dis-
N,N′-(2-Nitrobenzensulfonamide)-1,2-triaminodiethane (4). (130 mg, 0.53 mmol) was dissolved in dry CH3CN and added
Diethylenetriamine (1.5 mL, 15.2 mmol) was dissolved in dry to the stirring reaction mixture. The reaction mixture was
THF (90 mL) and cooled to 0 °C. Na2CO3 (3.21 g, 30.4 mmol) heated to 60 °C and stirred overnight, after which time it
was added to the stirring solution, followed by the slow turned a deep brown colour. The reaction mixture was
addition of 2-nitrobenzene sulfonyl chloride (6.75 g, removed from heat and cooled to room temperature before
30.4 mmol). Upon complete addition of the reagents, the solu- salts were filtered out. The collected filtrate was evaporated,
tion was warmed to RT and stirred overnight. The salts were dissolved in H2O and the pH was adjusted to 5 before extract-
removed by filtration and the solvent evaporated in vacuo to ing with EtOAc (3 × 15 mL). The combined organic layers were
give a yellow solid, which was recrystallized in DCM to give an dried over Na2SO4, filtered and evaporated to give a brown
1
off-white solid (3.2 g, 52%). H NMR (CD3CN, 400 MHz) δ 8.03 residue that was used without purification in the next de-
(m, 2H), 7.85 (m, 2H), 7.80 (m, 4H), 2.99 (t, 4H, J = 5.7, 6.2 Hz), protection step. The brown residue was dissolved in THF
2.54 (t, 4H, J = 5.6, 6.2 Hz). 13C{1H} NMR (CD3CN, 100 MHz) δ (2 mL) and 6 M HCl (2 mL) was added. The reaction mixture
134.1, 132.9, 132.7, 130.6, 125.1, 47.2, 42.9. LR-ESI-MS: calcd was heated to 40 °C and stirred for 18 h before evaporating the
for [C16H19N5O8S2]: 473.07; found [M + H]+: 473.9.
solvents for purification. The crude product was purified by
Di-tert-butyl2,2′-((((2-(tert-butoxy)-2-oxo ethyl) aza nediyl) reverse phase C18 column chromatography (Combiflash auto-
bis (ethane-2,1-diyl)) bis ((phenylsulfonyl) aza nediyl))diace- mated purification system, A: H2O + 0.01% TFA, B: CH3CN;
tate (5). K2CO3 (2.15 g, 15.5 mmol) was added to a stirring 95% A to 45% A). Combined fractions yielded the product
1
solution of compound 4 (1.05 g, 2.22 mmol) in dry CH3CN H5decaox as a yellow residue (76 mg, 30% over two steps). H
(20 mL). To this solution was added tert-butyl bromoacetate NMR (D2O, 300 MHz) δ 8.85 (d, 2H, J = 8.5 Hz), 7.89 (d, 2H, J =
(1.08 mL, 7.32 mmol) and the mixture was heated at 60 °C for 8.6 Hz), 7.56 (d, 2H, J = 7.9), 7.46 (t, 2H, J = 16.0, 7.9), 6.83 (d,
24 h. The salts were filtered from the reaction mixture and the 2H, J = 7.7), 4.47 (s, 4H), 4.15 (s, 2H), 3.58 (s, 8H), 3.37 (t, 4H,
solvents were evaporated in vacuo. The resulting residue was J = 11.6, 5.8). 13C{1H} NMR (D2O, 75 MHz) δ 175.8, 169.7,
re-dissolved in H2O (30 mL) and extracted with EtOAc (3 × 154.3, 147.4, 146.5, 130.7, 129.3, 129.2, 122.5, 119.6, 116.8,
50 mL). The combined organic layers were washed once with 56.3, 54.9, 54.6, 51.1. LR-ESI-MS calcd for [C30H33N5O8] 591.23;
sat. NaHCO3 (25 mL) and water (25 mL) and dried over MgSO4. [M + H]+: 592.3. HR-ESI-MS calcd for [C30H33N5O8 + H]:
The solvents were evaporated in vacuo to give a yellow oil 592.2407; found [M + H]+: 591.2401. Elemental Analysis calc’d
1
(1.74 g, 96%). H NMR (CDCl3, 400 MHz) δ 8.05 (m, 2H), 7.66 % for H5decaox·5HCl·2.8H2O (C30H33N5O8·5HCl·2.8H2O): C
(m, 4H), 7.55 (m, 2H), 4.14 (s, 4H), 3.44 (t, 4H, J = 6.9 Hz), 3.22 43.71, H 5.33, N 8.50. Found: C 43.93, H 5.52, N 8.26.
(s, 2H), 2.85 (t, 4H, J = 6.9 Hz), 1.42 (s, 9H), 1.32 (s, 18H). 13C
Metal complexation
{1H} NMR (CDCl3, 100 MHz) δ 167.8, 147.9, 133.5, 131.8, 130.9,
123.9, 82.3, 60.4, 53.2, 49.5, 46.7, 28.1, 27.9, 21.1, 14.2. A stock solution of H5decaox (10 mM) was prepared in D2O for
LR-ESI-MS: calcd for [C34H49N5O14S2]: 815.27; found [M + H]+: metal complex NMR characterisation. Metal ions of interest
816.5.
Di-tert-butyl 2,2′-((((2-(tert-butoxy)-2-oxo ethyl) aza nediyl) in D2O (10 mM) and complexes were prepared by combining
bis (ethane-2,1-diyl)) bis aza nediyl))-diacetate (6). 1 : 1.1/L : M (Vt ∼400 µL) and adjusting the pH with freshly pre-
(M = In3+, La3+, Zr4+) were similarly prepared as stock solutions
(
Compound 5 (1.40 g, 1.72 mmol) was dissolved in dry CH3CN pared 0.1 M NaOD. Solutions stood at RT for at least
(20 mL) and the flask purged with nitrogen. Thiophenol 15 minutes prior to collecting NMR spectra.
(440 µL, 4.31 mmol) was added to the stirring solution fol-
Na[In(Hdecaox)]. HR-ESI-MS calcd for C30H30InNaN5O8
lowed by K2CO3 (816 mg, 5.70 mmol). The reaction mixture 726.1031; measured 726.1027.
was heated at 60 °C and stirred for 24 h, before cooling to RT
[Zr(Hdecaox)]. HR-ESI-MS calcd for C30H30N5O8Zr 678.1141;
and removing salts via centrifugation. The residue was re-dis- measured 678.1144.
3882 | Dalton Trans., 2021, 50, 3874–3886
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
[La(H2decaox)]. HR-ESI-MS calcd for C30H31139LaN5O8 chelation was analyzed on day 1, 3, and 7 by adding 300 µL ali-
728.1236; measured 728.1234.
quots to ice cold CH3CN (700 µL) to precipitate the serum pro-
teins. This sample was centrifuged for 10 min at 10 000 rpm,
and the supernatant was decanted and counted in a dose cali-
brator. A second wash of H2O/CH3CN (30 : 70) was added to
the pellet, and re-centrifuged for another 10 min at 10 000
rpm. The supernatant was decanted and counted in a dose
calibrator. The amount of [89Zr]Zr4+ that remained within the
centrifuged pellet was assumed to be transchelated, while the
Radiolabeling and serum stability assays
[
111In]In3+ radiolabeling. A stock solution of H5decaox (1 ×
10−3 M) was prepared in H2O and used in serial dilutions for
concentration dependent radiolabeling experiments.
Radiolabeling reactions were prepared to a total volume of
500 µL in NaOAc buffer (0.1 M, pH 7), with ligand concen-
trations of 1 × 10−4 to 10−8 M, upon which 1 µL (∼3 MBq) of
amount of
[
89Zr]Zr4+ in the combined supernatant was
assumed to be intact complex, and this ratio was used to deter-
mine the percent stability.
[
111In]InCl3 was added and incubated at RT for 30 minutes.
Reaction progress was monitored by radio-HPLC using a linear
gradient of 5%–60% A (A = CH3CN, B = H2O + 0.1% TFA) over
20 minutes, where free [111In]InCl3 elutes at Rt = 4.8 min and
radiolabeled complex elutes at Rt = 11.6 min. The area of each
peak was integrated and used to determine RCY%.
The distribution coefficient (log D) was determined at pH
7.4, performed in duplicate and each duplicate measured
twice. A stock solution of ZrCl4 (1 × 10−4) and [89Zr]ZrCl4
(4 MBq) was prepared and 50 µL was added to a solution of
H5decaox (1 × 10−4 M) to a total reaction volume of 500 µL in
NaOAc buffer (0.1 M, pH 7). Quantitative radiolabeling was
confirmed by HPLC, and 50 µL of the reaction mixture was
added to a mixture of octanol (500 µL) and HEPES buffer/
NaOH (450 µL, pH 7.4). These samples were mixed for
30 minutes at RT and then allowed to separate. 400 µL from
each octanol and water phase was removed and centrifuged.
Duplicates of each sample (100 µL) from each sample were
then counted in a gamma counter (NaI(Tl) scintillation
counter automatic gamma counter 1480, Wizard 3″,
PerkinElmer) and used to determine the octanol/water distri-
bution ratio of [89Zr]Zr(decaox) following eqn (2).
Serum stability of the [111In]In(decaox) complex was evalu-
ated using GE Healthcare Life Sciences PD-10 desalting
column (size exclusion <5000 Da) method as previously
described.43 In triplicate, a quantitatively labeled reaction
([L] = 1 × 10−5 M) with ∼37 MBq (1 mCi) of activity was combined
with an equal volume (500 µL) of human serum and incubated
at 37 °C. At time points of 1 h, 1 day, 3 days and 5 days, ali-
quots of 100, 200, 200 and 400 µL respectively, were diluted to
2.5 mL and the activity was measured with a CRC55tR dose
calibrator. The dilution was then loaded onto an equilibrated
PD-10 desalting column, and the activity of the empty vial was
measured to determine the “residual activity”. Upon column
adsorption, the proteins (MW >5000 Da) were eluted with
3.5 mL of PBS. The activity of the elution was then measured,
and the stability of the complex in % was calculated as in
eqn (1).
ꢀ
ꢁ
Counts in octanol phase
Counts in aqueous phase
log D7:4 ¼ log10
ð2Þ
Solution thermodynamics
ꢀ
ꢁ
ð1Þ
ðActivity of elutionÞ
% stability ¼ 1 ꢀ
Protonation constants and metal stability constants were deter-
mined from a combined UV-potentiometric titration system
using a Metrohm Titrando 809 equipped with a Ross com-
ðActivity of loadÞ ꢀ ðActivity residualÞ
[89Zr]Zr4+ radiolabeling. A stock solution of H5decaox (1 × bined electrode, a Metrohm Dosino 800, and a Varian Cary 60
10−3 M) was prepared in H2O and diluted further for concen- UV/vis spectrophotometer (200–500 nm spectral range) con-
tration dependent experiments. Radiolabeling reactions took nected to a 0.2 cm path length optic dip probe submerged in
place in a total reaction volume of 200 µL, at ligand concen- the titration cell. The titration cell consisted of a 20 mL ther-
trations of 1 × 10−4 to 10−7 M in NaOAc buffer (0.1 M, pH 7) mostatted glass cell at 25 °C with an inlet–outlet tube for nitro-
and ∼2 MBq of [89Zr]ZrCl4. The reaction progress was moni- gen gas to exclude any CO2 before and throughout the titra-
tored at 30 minutes using radio-HPLC on a C18 column with tion. The electrode was calibrated daily for hydrogen ion con-
solvents A (H2O + 0.1% TFA) and B (CH3CN + 0.1% TFA) in a centration using a standard HCl and calibration data were ana-
gradient of 100% A to 100% B over 20 minutes, where free lysed with the Gran procedure48 to obtain calibration para-
[89Zr]ZrCl4 elutes at Rt = 3.30 min and complexed [89Zr]Zr meters E0 and pKw. Solutions were titrated with carbonate-free
(decaox) elutes at Rt = 8.8 min. The area of each peak was inte- NaOH (0.16 M) that was standardised against freshly recrystal-
grated and used to determine RCY%.
lised potassium hydrogen phthalate. Protonation equilibria of
The serum stability of a quantitatively radiolabeled sample the ligand were studied via combined UV-potentiometric titra-
of [89Zr]Zr(decaox) was determined using a protein precipi- tions of a solution containing H5decaox ([L] = 9.62 × 10−4 M,
tation method with cold CH3CN following literature pre- 25 °C, l = 0.2 cm and 0.16 M NaCl). Electromotive force values
cedents.42 In triplicate, 400 µL of a quantitatively radiolabeled and spectra were recorded after each NaOH addition, and the
sample of [89Zr]Zr(decaox) (∼4 MBq, [L] = 1 × 10−4 M) was apparatuses were synchronized to have constant delays
added to a vial with 600 µL of human serum and incubated at between each titrant addition and sufficient time to reach
37 °C over a 7 day period. The stability of the complex to trans- equilibrium.33 The dissociation of the most acidic protons
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 3874–3886 | 3883

View Article Online
Paper
Dalton Transactions
were evaluated via in batch UV-spectrophotometric samples accuracy for the large system sizes under study.70,71 The inte-
([L] = 2.25 × 10−5 M) measured and recorded in a 1 cm path gral equation formalism of the polarizable continuum model
length cuvette, as the pH was below the electrode threshold.
(IEFPCM) was used as an implicit water model in all calcu-
Complex formation equilibria with In3+ or La3+ and lations to simulate the average dielectric effects of the solvent.
H5decaox were directly studied via UV-potentiometric titra- Default IEFPCM parameters were used, as implemented in
tions. Ligand–metal solutions were prepared by adding the Gaussian (ε = 78.36, van der Waals surface without “added
atomic absorption (AA) standard metal ions solutions to a spheres”). Each calculation was repeated with 0 and 4 explicit
known concentration of H5decaox (([L] = [In3+] = 7.67 × 10−4 M) water molecules which were placed randomly by the PACKMOL
and ([L] = [La3+] = 8.31 × 10−4 M)) at 25 °C and ionic strength code72 in a sphere up to 4.5 Å from each system’s centre of
of 0.16 M NaCl. For highly acidic dissociation constants with mass. Adjustments were then made before optimisation to
In3+ ([L] = [In3+] = 2.25 × 10−5 M) and all samples with Zr4+ move the randomly placed water molecules closer to hydrogen
([L] = [Zr4+] = 2.0 × 10−5 M), in batch samples were prepared bond centers, such as hydroxyquinolinate and carboxyl
and UV-spectrophotometry was measured individually in a groups. Calculation results were visualized using GaussView
1 cm path length cuvette. All potentiometric measurements version 6.073 and Avogadro version 1.2.0.74
were processed using Hyperquad2013 software,33 while the
obtained spectrophotometric data were processed with the
HypSpec2014 program.34 Molar absorptivities of protonated
species of H5decaox were included in metal stability calcu-
Conflicts of interest
lations, and proton dissociation constants corresponding to The authors declare no competing financial interest.
the hydrolysis of In3+ and La3+ aqueous ions were taken from
Baes and Mesmer.49,50 By convention, a complex containing a
metal ion, M, proton, H, and ligand, L, has the general
formula MpHqLr and the overall equilibrium formation con-
Acknowledgements
stant is designated log β and is characterized by the general We gratefully acknowledge the Natural Sciences and
equilibrium: pM + qH + rL = MpHqLr (charges omitted). The Engineering Research Council (NSERC) of Canada for CGS-M/
stoichiometric indices p might also be 0 in the case of protona- PGS-D scholarships (L. S.), UBC for a Four Year Fellowship
tion equilibria, and negative values of q refer to proton (L. S.) and the NSERC CREATE IsoSiM program at TRIUMF for
removal from metal-ion-coordinated water, equivalent to research stipends (L. S., L. W.). We thank both NSERC and
hydroxide ion addition to the complex. Stepwise equilibrium CIHR for financial support via a Collaborative Health Research
constants, log K, correspond to the difference in log units Project (CHRP – CO) and NSERC Discovery (CO). We thank
between the overall constants of sequentially protonated (or WestGrid and Compute Canada for access to their compu-
hydroxide) species. pM is defined as (−log[Mn+
]free) and was tational resources. We gratefully acknowledge Dr Maria Ezhova
calculated using the Hyss software35 from the stability con- for her expert advice on NMR experiments, Karin Landrock for
stants for each system using the standards: [Mn+] = 1 μM, [Lx−
= 10 μM, pH 7.4, and 25 °C.36
]
technical assistance at HZDR, Dr Martin Walther for providing
89Zr at HZDR and Hayden Scheiber for helpful DFT discus-
sions. We heartily thank Randy Perron and Dr Patrick Causey
at CNL for providing the 225Ac. HZDR is thanked for support-
Density functional theory calculations
All DFT calculations were performed using Gaussian 16 revi- ing the IsoSiM International Research Experience program;
sion c01.51 Self-consistent field (SCF) convergence criteria were TRIUMF receives federal funding via a contribution agreement
set to their default values (SCF = Tight in Gaussian). Structure with the National Research Council of Canada.
optimisations were performed without symmetry constraints
using the Berny algorithm52 with default settings, starting
from initial structures built manually. Each structure was opti-
mised and free energies were calculated using DFT with the
References
PBE0 hybrid exchange–correlation functional,53 added D3
(BJ)54–56 dispersion corrections, and the def2-TZVP basis set57
for all non-metal atoms. Effective core potentials (ECPs) were
used to account for scalar relativistic effects in metal core elec-
trons. Specifically, def2-TZVP basis sets were used with ECPs
for zirconium58 and for indium57,59 while the Stuttgart RSC
segmented valence basis set and ECP60,61 was used for lantha-
num. Metal basis sets were downloaded from the Basis Set
Exchange website.62–64 The PBE0 functional was chosen due to
its previous success in structure prediction and thermochem-
istry of transition metal complexes,65–69 while the def2-TZVP
basis was chosen as a reasonable trade-off between speed and
1 M. D. Bartholomä, Inorg. Chim. Acta, 2012, 389, 36–51.
2 D. Brasse and A. Nonat, Dalton Trans., 2015, 44, 4845–4858.
3 E. Boros and A. B. Packard, Chem. Rev., 2019, 119, 870–901.
4 T. I. Kostelnik and C. Orvig, Chem. Rev., 2019, 119, 902–
956.
5 C. F. Ramogida and C. Orvig, Chem. Commun., 2013, 49,
4720–4739.
6 E. W. Price and C. Orvig, Chem. Soc. Rev., 2014, 43, 260–
290.
7 X. Wang, M. de G. Jaraquemada-Peláez, Y. Cao, J. Pan,
K.-S. Lin, B. O. Patrick and C. Orvig, Inorg. Chem., 2019, 58,
2275–2285.
3884 | Dalton Trans., 2021, 50, 3874–3886
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
8 X. Wang, M. de G. Jaraquemada-Peláez, C. Rodríguez- 29 M. A. Deri, B. M. Zeglis, L. C. Francesconi and J. S. Lewis,
Rodríguez, Y. Cao, C. Buchwalder, N. Choudhary, Nucl. Med. Biol., 2013, 40, 3–14.
U. Jermilova, C. F. Ramogida, K. Saatchi, U. O. Häfeli, 30 E. W. Price, J. F. Cawthray, G. A. Bailey, C. L. Ferreira,
B. O. Patrick and C. Orvig, J. Am. Chem. Soc., 2018, 140,
15487–15500.
E. Boros, M. J. Adam and C. Orvig, J. Am. Chem. Soc., 2012,
134, 8670–8683.
9 N. Choudhary, A. Dimmling, X. Wang, L. Southcott, 31 S. Poty, L. C. Francesconi, M. R. McDevitt, M. J. Morris and
V. Radchenko, B. O. Patrick, P. Comba and C. Orvig, Inorg.
Chem., 2019, 58, 8685–8693.
J. S. Lewis, J. Nucl. Med., 2018, 59, 878–884.
32 W. Y. Hsieh and S. Liu, Inorg. Chem., 2004, 43, 6006–6014.
10 N. Choudhary, M. de G. Jaraquemada-Peláez, K. Zarschler, 33 P. Gans, A. Sabatini and A. Vacca, Talanta, 1996, 43, 1739–
X. Wang, V. Radchenko, M. Kubeil, H. Stephan and
C. Orvig, Inorg. Chem., 2020, 59, 5728–5741.
1753.
34 P. Gans, A. Sabatini and A. Vacca, Ann. Chim., 1999, 89, 45–
11 X. Wang, M. de G. Jaraquemada-Peláez, C. Rodriguez-
49.
Rodriguez, J. Pan, Y. Wang, K. Saatchi, U. O. Ha, K. Lin and 35 L. Alderighi, P. Gans, A. Ienco, D. Peters, A. Sabatini and
C. Orvig, Inorg. Chem., 2020, 59, 4895–4908. A. Vacca, Coord. Chem. Rev., 1999, 184, 311–318.
12 K. Soroka, R. S. Vithanage, D. A. Phillips, B. Walker and 36 W. R. Harris, C. J. Carrano and K. N. Raymond, J. Am.
P. K. Dasgupta, Anal. Chem., 1987, 59, 629–636. Chem. Soc., 1979, 101, 2213–2214.
13 K. F. Sugawara, H. H. Weetall and G. D. Schucker, Anal. 37 É. A. Enyedy, O. Dömötör, E. Varga, T. Kiss, R. Trondl,
Chem., 1974, 46, 489–492.
14 V. Prachayasittikul, S. Prachayasittikul, S. Ruchirawat and
C. G. Hartinger and B. K. Keppler, J. Inorg. Biochem., 2012,
117, 189–197.
V. Prachayasittikul, Drug Des., Dev. Ther., 2013, 7, 1157– 38 Y. Toporivska and E. Gumienna-Kontecka, J. Inorg.
1178. Biochem., 2019, 198, 110753.
15 Y. Song, H. Xu, W. Chen, P. Zhan and X. Liu, 39 M. Sturzbecher-Hoehne, T. A. Choi and R. J. Abergel, Inorg.
MedChemComm, 2015, 6, 61–74. Chem., 2015, 54, 3462–3468.
16 S. H. Chan, C. H. Chui, S. W. Chan, S. H. L. Kok, D. Chan, 40 C. Buchwalder, M. de G. Jaraquemada-Peláez, J. Rousseau,
M. Y. T. Tsoi, P. H. M. Leung, A. K. Y. Lam, A. S. C. Chan,
K. H. Lam and J. C. O. Tang, ACS Med. Chem. Lett., 2013, 4,
170–174.
H. Merkens, C. Rodríguez-Rodríguez, C. Orvig, F. Bénard,
P. Schaffer, K. Saatchi and U. O. Häfeli, Inorg. Chem., 2019,
58, 14667–14681.
17 G. Karpińska, A. P. Mazurek and J. C. Dobrowolski, J. Mol. 41 L. Li, M. de G. Jaraquemada-Peláez, H. T. Kuo, H. Merkens,
Struct.: THEOCHEM, 2010, 961, 101–106.
N. Choudhary, K. Gitschtaler, U. Jermilova, N. Colpo,
C. Uribe-Munoz, V. Radchenko, P. Schaffer, K. S. Lin,
F. Bénard and C. Orvig, Bioconjugate Chem., 2019, 30, 1539–
1553.
18 M. A. Green and J. C. Huffman, J. Nucl. Med., 1988, 29, 417.
19 M. Roca, E. F. J. de Vries, F. Jamar, O. Israel and A. Signore,
Eur. J. Nucl. Med. Mol. Imaging, 2010, 37, 835–841.
20 E. Schell-Frederick, J. Frühling, P. Van Der Auwera, Y. Van 42 E. K. Sarbisheh, A. K. Salih, S. J. Raheem, J. S. Lewis and
Laethem and J. Klastersky, Cancer, 1984, 54, 817–824. E. W. Price, Inorg. Chem., 2020, 59, 11715–11727.
21 C. M. Intenzo, A. G. Desai, M. L. Thakur and C. H. Park, 43 P. Comba, U. Jermilova, C. Orvig, B. O. Patrick,
J. Nucl. Med., 1987, 28, 438–441.
C. F. Ramogida, K. Rück, C. Schneider and M. Starke,
22 P. Gawne, F. Man, J. Fonslet, R. Radia, J. Bordoloi,
Chem. – Eur. J., 2017, 23, 15945–15956.
M. Cleveland, P. Jimenez-Royo, A. Gabizon, P. J. Blower, 44 K. Arane, MSc Thesis, University of British Columbia, 2015.
N. Long and R. T. M. De Rosales, Dalton Trans., 2018, 47, 45 O. K. Abou-Zied, N. Al-Lawatia, M. Elstner and
9283–9293.
T. B. Steinbrecher, J. Phys. Chem. B, 2013, 117, 1062–1074.
23 S. M. Valiahdi, P. Heffeter, M. A. Jakupec, R. Marculescu, 46 É. A. Enyedy, O. Dömötör, K. Bali, A. Hetényi, T. Tuccinardi
W. Berger, K. Rappersberger and B. K. Keppler, Melanoma
Res., 2009, 19, 283–293.
24 N. Sato, H. Wu, K. O. Asiedu, L. P. Szajek, G. L. Griffiths
and P. L. Choyke, Radiology, 2015, 275, 490–500.
25 P. Charoenphun, L. K. Meszaros, K. Chuamsaamarkkee,
and B. K. Keppler, J. Biol. Inorg. Chem., 2015, 20, 77–88.
47 A. A. Hummer, C. Bartel, V. B. Arion, M. A. Jakupec,
W. Meyer-Klaucke, T. Geraki, P. D. Quinn, A. Mijovilovich,
B. K. Keppler and A. Rompel, J. Med. Chem., 2012, 55,
5601–5613.
E. Sharif-Paghaleh, J. R. Ballinger, T. J. Ferris, M. J. Went, 48 G. Gran, Analyst, 1952, 77, 661.
G. E. D. Mullen and P. J. Blower, Eur. J. Nucl. Med. Mol. 49 C. F. Baes Jr. and R. E. Mesmer, The Hydrolysis of Cations,
Imaging, 2015, 42, 278–287.
John Wiley & Sons, New York, 1976.
26 P. J. Gawne, F. Clarke, K. Turjeman, A. P. Cope, N. J. Long, 50 C. F. Baes Jr. and R. E. Mesmer, Am. J. Sci., 1981, 281, 935–
Y. Barenholz, S. Y. A. Terry and R. T. M. de Rosales,
Theranostics, 2020, 10, 3867–3879.
27 K. O. Asiedu, M. Ferdousi, P. T. Ton, S. S. Adler,
P. L. Choyke and N. Sato, EJNMMI Res., 2018, 8, 109.
28 J. R. Dilworth and S. I. Pascu, Chem. Soc. Rev., 2018, 47,
2554–2571.
962.
51 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,
G. A. Petersson, H. Nakatsuji, X. Li, M. Caricato,
A. V. Marenich, J. Bloino, B. G. Janesko, R. Gomperts,
B. Mennucci, H. P. Hratchian, J. V. Ortiz, A. F. Izmaylov,
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 3874–3886 | 3885

View Article Online
Paper
Dalton Transactions
J. L. Sonnenberg, D. Williams-Young, F. Ding, F. Lipparini, 61 X. Cao and M. Dolg, J. Mol. Struct.: THEOCHEM, 2002, 581,
F. Egidi, J. Goings, B. Peng, A. Petrone, T. Henderson, 139–147.
D. Ranasinghe, V. G. Zakrzewski, J. Gao, N. Rega, G. Zheng, 62 K. L. Schuchardt, B. T. Didier, T. Elsethagen, L. Sun,
W. Liang, M. Hada, M. Ehara, K. Toyota, R. Fukuda,
J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
V. Gurumoorthi, J. Chase, J. Li and T. L. Windus, J. Chem.
Inf. Model., 2007, 47, 1045–1052.
H. Nakai, T. Vreven, K. Throssell, J. A. J. Montgomery, 63 B. P. Pritchard, D. Altarawy, B. Didier, T. D. Gibson and
J. E. Peralta, F. Ogliaro, M. J. Bearpark, J. J. Heyd, T. L. Windus, J. Chem. Inf. Model., 2019, 59, 4814–4820.
E. N. Brothers, K. N. Kudin, V. N. Staroverov, T. A. Keith, 64 D. Feller, J. Comput. Chem., 1996, 17, 1571–1586.
R. Kobayashi, J. Normand, K. Raghavachari, A. P. Rendell, 65 J. Vícha, M. Patzschke and R. Marek, Phys. Chem. Chem.
J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi,
J. M. Millam, M. Klene, C. Adamo, R. Cammi, 66 M. Bühl, C. Reimann, D. A. Pantazis, T. Bredow and
J. W. Ochterski, R. L. Martin, K. Morokuma, O. Farkas, F. Neese, J. Chem. Theory Comput., 2008, 4, 1449–1459.
J. B. Foresman and D. J. G. Fox, Gaussian 16, Revision C.01, 67 J. Vícha, J. Novotný, M. Straka, M. Repisky, K. Ruud,
Phys., 2013, 15, 7740–7754.
Gaussian, Inc., Wallingford, CT, 2016.
52 X. Li and M. J. Frisch, J. Chem. Theory Comput., 2006, 2,
835–839.
53 C. Adamo and V. Barone, J. Chem. Phys., 1999, 110, 6158–
6170.
S. Komorovsky and R. Marek, Phys. Chem. Chem. Phys.,
2015, 17, 24944–24955.
68 M. M. Quintal, A. Karton, M. A. Iron, A. D. Boese and
J. M. L. Martin, J. Phys. Chem. A, 2006, 110, 709–716.
69 S. Dohm, A. Hansen, M. Steinmetz, S. Grimme and
M. P. Checinski, J. Chem. Theory Comput., 2018, 14, 2596–
2608.
54 S. Grimme, S. Ehrlich and L. Goerigk, J. Comput. Chem.,
2011, 32, 1456–1465.
55 S. Grimme, J. Antony, S. Ehrlich and H. Krieg, J. Chem. 70 K. N. Kirschner, D. Reith and W. Heiden, Soft Mater., 2020,
Phys., 2010, 132, 154104. 18, 200–214.
56 D. G. A. Smith, L. A. Burns, K. Patkowski and C. D. Sherrill, 71 A. Bauzá, D. Quiñonero, P. M. Deyaì and A. Frontera,
J. Phys. Chem. Lett., 2016, 7, 2197–2203. J. Phys. Chem. A, 2013, 117, 2651–2655.
57 F. Weigend and R. Ahlrichs, Phys. Chem. Chem. Phys., 2005, 72 L. Martinez, R. Andrade, E. G. Birgin and J. M. Martinez,
7, 3297–3305. J. Comput. Chem., 2009, 30, 2157–2164.
58 D. Andrae, U. Haeussermann, M. Dolg, H. Stoll and 73 R. Dennington, T. A. Keith and J. M. Millam, GaussView,
H. Preuss, Theor. Chim. Acta, 1990, 77, 123–141. Version 6.0.16, Semichem Inc. Shawnee Mission KS, 2016.
59 B. Metz, H. Stoll and M. Dolg, J. Chem. Phys., 2000, 113, 74 M. D. Hanwell, D. E. Curtis, D. C. Lonie,
2563–2569.
T. Vandermeersch, E. Zurek and G. R. Hutchison,
60 X. Cao and M. Dolg, J. Chem. Phys., 2001, 115, 7348–7355.
J. Cheminf., 2012, 4, 17.
3886 | Dalton Trans., 2021, 50, 3874–3886
This journal is © The Royal Society of Chemistry 2021