A macrocyclic chelator with unprecedented Th4+ affinity

Article abstract of DOI:10.1021/ja503456r

A novel macrocyclic octadentate ligand incorporating terephthalamide binding units has been synthesized and evaluated for the chelation of Th ⁴⁺. The thorium complex was structurally characterized by X-ray diffraction and in solution with kinetic studies and spectrophotometric titrations. Dye displacement kinetic studies show that the ligand is a much more rapid chelator of Th⁴⁺ than prevailing ligands (1,4,7,10- tetraazacyclododecane-1,4,7,10-tetraacetic acid and diethylenetriaminepentaacetic acid). Furthermore, the resulting complex was found to have a remarkably high thermodynamic stability, with a formation constant of 10⁵⁴. These data support potential radiotherapeutic applications.

Full text of DOI:10.1021/ja503456r

Article
pubs.acs.org/JACS
A Macrocyclic Chelator with Unprecedented Th4+ Affinity
Tiffany A. Pham,†,‡ Jide Xu,†,‡ and Kenneth N. Raymond*,†,‡
†Department of Chemistry, University of California, Berkeley, Berkeley, California 94720-1460, United States
‡Chemical Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, California 94720-1460, United States
S
* Supporting Information
ABSTRACT: A novel macrocyclic octadentate ligand incor-
porating terephthalamide binding units has been synthesized
and evaluated for the chelation of Th4+. The thorium complex
was structurally characterized by X-ray diffraction and in
solution with kinetic studies and spectrophotometric titrations.
Dye displacement kinetic studies show that the ligand is a
much more rapid chelator of Th4+ than prevailing ligands
(1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid and
diethylenetriaminepentaacetic acid). Furthermore, the result-
ing complex was found to have a remarkably high thermodynamic stability, with a formation constant of 1054. These data support
potential radiotherapeutic applications.
such as DOTA have both slow off and on rates, while acyclic
ligands like DTPA display faster formation kinetics but form
complexes of lower stability. For example, DOTA has a greater
stability with lanthanides than DTPA by about 2 orders of
magnitude, but the latter displays faster kinetics by 4 orders of
magnitude.6
Our siderophore-inspired ligands have been used to complex
actinides for applications in nuclear waste remediation and in
vivo decorporation.7 The terephthalamide (TAM) binding
group has a very high binding affinity for Th4+,7c and its two
amide groups enable its incorporation in a novel ligand
topology intended to address the requirements for a good
radiotherapeutic chelate. The ligand L (Figure 1) was designed
INTRODUCTION
■
α-Particle emitters have been studied for applications in
radiotherapy for over 50 years.1 The shorter range and higher
energy of α-particles relative to β-particles results in greater
cytotoxicity.2 However, clinical investigations of α-therapeutics
are fairly recent. Nuclear therapy was previously limited to β-
emitters and external radiation. To date, the only FDA-
approved α-emitter is 223RaCl2, a calcium mimic effective in the
treatment of bone metastases.3 Coupling the short-range of α-
particles with the precise targeting of a biomolecule, such as an
antibody, could offer radiotherapeutics with both the
advantages of versatility and specificity. Much of the current
research in α-therapy is directed toward using a variety of α-
emitters, e.g., 225Ac, 211At, 227Th, 213Bi,4 in conjunction with a
ligand covalently attached to a biomolecule. This approach has
created a demand for suitable chelators, which must rapidly and
irreversibly form thermodynamically stable complexes with the
α-emitting radioisotope under physiological conditions. Pub-
lished approaches have been limited to aminocarboxylic acids
such as diethylenetriaminepentaacetic acid (DTPA) and
1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid
(DOTA) and their derivatives. While these chelators have
been available for several decades, their effectiveness is in
several respects quite limited.
Figure 1. Octadentate terephthalamide ligand (L) with a topology
incorporating a macrocycle with pendant binding units.
Algeta ASA, the producer of the commercialized 223RaCl2,
has pointed out the potential of 227Th, a β-decay product of
227Ac.5 This α-emitter decays to 223Ra with an 18.7 d half-life, a
convenient time scale for manufacturing and shipping.
However, unlike radium, the chemistry of thorium requires
its sequestration in a stable complex. If then attached to a
targeting agent such as a monoclonal antibody, various 227Th
therapeutic agents could be envisioned. However, current
available chelators do not simultaneously have high thermody-
namic stability and fast association. In general, macrocycles
to combine the favorable thermodynamics of macrocycles with
the favorable kinetics of linear ligands. It features four binding
units (at least eight atoms are required to fill the coordination
sphere of Th4+): two TAMs in a macrocycle and two TAMs on
pendant arms, connected by tris(2-aminoethyl)amine (tren)
backbones. The pendant TAMs provide opportunities for
Received: April 10, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja503456r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
a
Scheme 1. Synthesis of L
a
Complete experimental procedures and abbreviations are detailed in the Experimental Section.
further functionalization, e.g., for linking to a targeting moiety.
In L they were functionalized with methoxyethyl groups to
increase the water solubility of the untargeted ligand. The
suitability of L as a Th4+ chelator for radiotherapeutic
applications was evaluated with thermodynamic, kinetic, and
single-crystal X-ray diffraction (XRD) studies.
thorium complex was in turn crystallized from a methanol and
dimethylformamide solution, into which diethyl ether and
tetrahydrofuran were diffused. Naturally occurring 232Th, which
has a half-life of 1.4 × 1010 years, was used for ease of handling.
XRD experiments gave the structures of L and its thorium
complex (Figure 2, Table 1).
In principle, the macrocyclic TAMs of the ligand could
chelate a metal ion in two possible configurations, either
equatorially, with the pendant units on opposite sides of the
metal−macrocycle plane (pseudo-C2h symmetry) or in a
configuration of lower symmetry (pseudo-C2 symmetry),
where the pendant arms are adjacent to one another (Figure
3). As the protonated ligand, L crystallizes with inversion
symmetry (as in the pseudo-C2h), but was observed to adopt
the C2-symmetric conformation around the metal center when
bound to Th4+. DFT-level calculations [B3LYP/6-31G, SDD-
(f)]10 of the complex in these two possible conformations
suggest that the observed mode is the more thermodynamically
stable one. The calculated C2-symmetric structure with adjacent
pendant units is 22 kcal/mol lower in energy than the
corresponding C2v-symmetric structure, a stabilization that can
be attributed to the macrocyclic strain induced by the large
diameter of Th4+. Increasing the ring size with tris(2-
aminopropyl)amine (trpn) instead of tren in the ligand
backbones reduces this calculated energy difference to 13
kcal/mol. The inner coordination environment of the crystal-
lized complex was analyzed with a quantitative shape measure
(SM),11 which compares the dihedral angles of adjacent faces in
the crystal structure’s coordination polyhedron to those of
idealized polyhedra most common in eight-coordinate
complexes: the bicapped trigonal prism (C2v), trigonal
dodecahedron (D2d), and square antiprism (D4d). The
coordination of ThL about the metal center is intermediate
between the ideal C2v and D2d geometries [SM = 5.3 (C2v), 7.7
(D2d), and 11.1 (D4d)]. Despite considerable distortion, the
TAM binding units can be viewed as spanning the m edges of a
trigonal dodecahedron with angles of 1.0°, 2.7°, 14.7°, and
19.5° from vertical.12 The macrocyclic binding units have the
largest deviation angles, providing further evidence for the
overall chelation mode of the ligand being attributable to ring
strain.
RESULTS AND DISCUSSION
■
Synthesis of L. The synthesis of L (Scheme 1) was adapted
from known preparations of analogous ligands.7a For ease of
purification the singly protected tren 3 was accessed in three
steps; the one-step synthesis8 gave a mixture of singly and
doubly protected amines that could not be separated. Two of
the tren amines were protected with benzyloxycarbonyl (Cbz)
groups (yielding 1), which were removed by hydrogenolysis
following protection of the remaining amine with a tert-
butoxycarbonyl (Boc) group (forming 2). Another key
intermediate was the macrocycle, which was formed with two
high-dilution reactions to prevent undesired polymeric by-
products.9 Compound 5 was formed by the slow addition of
mono-Boc tren (3) into a large excess of the TAM derivative 4,
while 6 was formed by the slow addition of the reactants, in
equimolar amounts, into a large volume of solvent.
Deprotection of the Boc groups in the presence of the benzyl
protecting groups proved slightly problematic, since the benzyl
groups were also susceptible to cleavage. This problem was
circumvented by using milder reaction conditions (30%
trifluoroacetic acid in CH2Cl2, added dropwise at 0 °C,
followed by reaction at room temperature monitored by TLC)
and prompt quenching with triethylamine as soon as the
deprotection reached completion. Attempts to isolate the
primary amine of 6 were unsuccessful, and coupling to the
pendant TAMs was performed immediately following quench-
ing of the Boc deprotection.
Mass spectral and elemental characterization of L confirmed
1
its identity, but the H NMR in DMSO showed extensive
hydrogen bonding that broke the symmetry, affording a
complicated spectrum. Treatment of L with NaOD in D2O
restored the expected symmetry and yielded well-resolved H
NMR signals.
Structural Characterization of L and ThL. Single crystals
suitable for XRD were grown by the vapor diffusion of
acetonitrile into an aqueous solution of the free ligand. The
1
Kinetics. Radiotherapeutics are inherently time-sensitive, as
the radioisotope must be fully sequestered by the chelate as fast
B
dx.doi.org/10.1021/ja503456r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Furthermore, above pH 3.1, more than 50% of the Th in
solution is hydrolyzed ([ThOH]3−, [Th(OH)2]2−
,
[Th4(OH)12]4+, etc.16), and increasingly more so at higher
pH values. In order to simplify the mechanism of complexation
by L, the metal was maintained as mainly free Th4+ in acidic
solution (pH 1.4, in 79 mM HNO3) until mixing with L, which
was buffered such that mixing with the Th solution resulted in a
reaction mixture at pH 7.4. The reaction studied was then
−8+n
Th4+ + LHn
→ [ThL]4− + nH+
(1)
3−
At this pH, most of the ligand is present as LH44− and LH5
(the species present in greater than 1% are LH44−, LH53−, and
LH62−, at 39.1, 57.8, and 2.6%, respectively). These measure-
ments were conducted using a stopped-flow apparatus, allowing
UV spectra to be taken immediately following the fast mixing of
the metal and ligand solutions. With various excess ligand
concentrations, the complexation was performed under pseudo-
first order kinetic conditions, from which a second-order rate
constant could be calculated. The observed first-order rate
constants [determined using the regression modeling program
Specfit17 to fit the changes in absorbance (Figure 4b) at a range
of wavelengths] were linearly dependent on the ligand
concentration, suggesting a second-order reaction with a rate
constant of k2 = 1.8(1) × 104 M−1 s−1 (the slope of the linear
regression in Figure 4c). The half-life of the complexation of 1
μM Th and 1 μM L is then 57 s, a more than adequately short
time period for radiotherapeutic applications of L. The direct
comparison of the k2 value with those of other ligands is not
possible given the lack of published work on the rate of
complexation of Th4+ by relevant ligands.
Of particular interest is the improved complexation kinetics
of L as compared to DTPA and DOTA, and the stopped-flow
kinetic experiment performed above is not applicable because
the ligands do not absorb in the UV−vis region. The relative
rate of complexation of Th4+ was instead studied indirectly,
using the dye Arsenazo III (D), with which it forms a colored
complex. The addition of ligand (L, DTPA, or DOTA)
displaced the dye, and the formation of a thorium−ligand
complex could be monitored by disappearance of the thorium−
dye complex (Figure 5b). The reaction effectively became the
ligand exchange described by the following general equation:
Figure 2. ORTEP diagrams of (a) the protonated ligand LH8·2HCl
and (b) the [ThL]4− complex. Potassium ions, solvent molecules, and
hydrogens in part b are omitted for clarity. Ellipsoids are shown at 50%
probability (gray, C; red, O; blue, N; green, Cl; lime-green, Th).
as possible to minimize decay prior to administration to the
patient. In addition to entropic effects, the high thermodynamic
stability of macrocycles with metal ions is attributed to the
preorganization of the donor atoms and multiple juxtaposi-
tional fixedness, which describes the greater difficulty with
which the donor atoms dissociate due to the lack of end
groups.13 However, the rigidity of the ligand causing this
stability can also hinder its kinetics of complexation. For
example, the polyaminocarboxylic acid macrocycles DOTA and
HEHA have slow formation rates with Ln(III) ions.6c,14 Despite
this shortcoming, DOTA is among the most commonly studied
macrocycles in radioactive medical applications,15 but an ideal
bifunctional chelator would form the final complex quickly
under mild conditions, to avoid damaging the targeting
biomolecule.
Study of the kinetics of association posed several problems
due to the spectroscopic silence of Th4+, as well as its
propensity to hydrolyze. By UV−vis spectroscopy, complex-
ation could only be monitored by changes in L, as there are no
distinct charge-transfer bands with Th4+. However, at
physiological pH and in excess of ligand, the spectral difference
between the ligand and the ThL complex is small (Figure 4a),
rendering accurate monitoring of the reaction difficult.
(2)
ThD + L → ThL + D
The stoichiometry of the thorium−dye complex, measured
with a Job plot (Figure 5a), was found to be between 1:1 and
1:2 Th:dye, varying with the purity of the dye used. Previous
work18 supports the existence of the 1:2 species, and since the
exact stoichiometry of the complex was not crucial for this
indirect kinetic study, an excess of dye was used to ensure that
all of the thorium was complexed. The Th−dye complex was
formed at acidic pH, to avoid the presence of Th hydroxide
species, and buffered to pH 7.4 prior to mixing with the ligand.
The aim was to extrapolate a second-order rate constant by
performing this displacement reaction under pseudo-first-order
conditions at different concentrations of excess ligand.
However, a rate dependence on the L concentration was not
observed under conditions above 10 equiv of L relative to Th4+.
A second-order rate constant was then calculated only for the
reactions with this ligand concentration, which was sufficient
for qualitative comparisons (Table 2); L was found to complex
Th with an observed rate constant of 147 M−1 s−1, a value 2
orders of magnitude greater than those of DTPA (0.61 M−1
s−1). No dye displacement was observed with DOTA over 2
C
dx.doi.org/10.1021/ja503456r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Table 1. Crystallographic Data and Structure Refinement for LH8·2HCl·4H2O and ThLK4·(DMF)2·(MeOH)2·THF
empirical formula
Mr
C50H72Cl2N10O22
1236.08 g/mol
C62H84K4N12O23Th
1753.85 g/mol
130(2) K
temp
100(2) K
wavelength
crystal system
space group
unit cell dimensions
1.54178 Å
0.71073 Å
monoclinic
triclinic
P21/n
P1
̅
a = 11.0189(6) Å, α = 90°
b = 8.3701(4) Å, β = 94.475(2)°
c = 29.9278(15) Å, γ = 90°
2751.8(2) Å3
a = 11.347(2) Å, α = 79.129(2)°
b = 16.549(2) Å, β = 87.126(2)°
c = 20.044(3)Å, γ = 79.929(2)°
3638.8(8) Å3
volume
Z
2
2
ρcalcd
1.492 g/cm3
1.551 g/cm3
μcalcd
1.849 mm−1
2.357 mm−1
F(000)
1304
1719
crystal size
0.10 × 0.50 × 0.10 mm3
0.21 × 0.09 × 0.06 mm3
2θ range for data collection
index ranges
2.96°−68.34°
−13 ≤ h ≤ 11, −10 ≤ k ≤ 10, −36 ≤ l ≤ 34
48 118
1.03°−50.74°
−13 ≤ h ≤ 13, −19 ≤ k ≤ 16, −24 ≤ l ≤ 24
45 683
reflections collected
independent reflections
completeness to θ = 25.00°
absorption correction
max and min transmission
refinement method
data/restraints/parameters
goodness-of-fit on F2
final R indices [I > 2σ(I)]
R indices (all data)
largest diff peak and hole
5013 [R(int) = 0.0294]
99.6%
13322 [R(int) = 0. 0928]
99.9%
semiempirical from equivalents
0.8367 and 0.2593
full-matrix least-squares on F2
5013/0/432
semiempirical from equivalents
0.8715 and 0.6374
full-matrix least-squares on F2
13322/102/954
1.037
1.038
R1 = 0.0355, wR2 = 0.0919
R1 = 0.0432, wR2 = 0.0965
0.673 and −0.268 e Å−3
R1 = 0.0709, wR2 = 0.1711
R1 = 0.1147, wR2 = 0.1973
1.568 and −1.564 e Å−3
constants of L. Although the pKa values of the hydroxyl groups
of a bidentate TAM unit are 6.0 and 11.0,20 those in L range
from 3.95 to 12.79 (Table 3). Statistical factors alone do not
account for this wide range, indicating further intramolecular
interactions within the various deprotonated states. As is
apparent in the representative titration shown in Figure 6, a
large pH range was required to attain all the protonation states
of L, and the UV−vis spectra of the individual species (Figure
9) display significant overlap.
Figure 3. Diagrams illustrating possible chelation modes of Th4+ by L.
(a) Pseudo-C2h symmetric structure, where the pendant TAMs lie on
opposite sides of the equatorially bound macrocyclic TAMs. (b)
Pseudo-C2 symmetric structure as observed in the crystal structure,
where the pendant TAMs are on the same side of the macrocycle.
Measurement of the Th4+ stability constant required the use
of a competing ligand in excess; otherwise, ThL would already
be formed at the start of the titration at pH 2. In the presence
of DTPA, the Th−DTPA complex instead forms at this low
pH, and at approximately pH 4, the [ThLH]3− complex begins
to form, which can be monitored spectrophotometrically. The
largest change in absorbance occurs at pH 5−6, when the
[ThLH]3− becomes deprotonated to form [ThL]4− (Figure 7).
The log β110 (where βmlh is the cumulative stability constant
for the equilibrium mM + lL +hH ⇌ MmLlHh) was found to be
53.7(5), 24 orders of magnitude greater than that of DTPA
(28.78,21 Table 4). Its greater log β110 value relative to the log
β140 = 45.54 for ethyl-TAM7b suggests that the higher denticity
and/or macrocyclic topology confers substantial thermody-
namic stability to the ThL complex. A measure of the complex
stability is also provided by the pTh value,22 or the −log of the
free thorium concentration at a specific pH and [Th]tot = 1 μM
and [L]tot = 10 μM (the speciation diagram in Figure 8 shows
the dominant species in the pH 1−13 range). This value
accounts for the effect of protonation on metal−ligand
equilibria, providing a direct comparison of the complexing
abilities of various ligands, which have different protonation
constants under physiologically relevant conditions. While
months at room temperature; it only proceeded within a
practicable time upon heating (100 °C). Despite the uncertain
mechanism in these dye displacement experimentsthe rate-
limiting step is most likely dye dissociation rather than ligand
association in the presence of a fast ligandL is clearly a much
faster thorium chelator than the macrocyclic DOTA and the
linear DTPA. These faster kinetics can be attributed to the
greater affinity of TAM relative to the O and N donors in
DTPA and DOTA but likely also to the partially macrocyclic
topology of L. Much like a linear ligand, its pendant arms afford
greater flexibility that enable faster association to the metal ion.
Thermodynamics. The thermodynamic stability of the
ThL complex was also evaluated. This property is of high
importance, as the release of the metal ion would be
catastrophic. Thorium forms hydroxides and complexes with
many proteins, amino acids, and other biological molecules,
which can aggregate in internal organs, such as the liver and
spleen.19 Due to the high number of pKa values and moderate
water solubility, UV−vis spectrophotometric titrations were
performed to determine the protonation and Th formation
D
dx.doi.org/10.1021/ja503456r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 5. Indirect kinetics. (a) UV−vis spectra of free Arsenazo III dye
(blue, 50 μM) and of Th−dye complex (red, 25 μM dye and 25 μM
Th) at pH 7.4 (100 mM HEPES, 0.1 M KCl). Spectra in black are of
Th and dye at intermediate stoichiometries (arrows denote spectral
changes with increasing concentrations of Th). (b) A vs t plot at 669
nm, showing the displacement of dye from the Th−dye complex upon
addition of L at t = 0 (5 μM Th−dye, 50 μM L, pH 7.4, 100 mM
HEPES, 0.1 M KCl, 25 °C). Black circles are data points; the red line
is the first-order exponential decay fit (R2 = 0.9974).
a
Table 2. Indirect Kinetic Data for L, DTPA, and DOTA
ligand
kobs (M−1 s−1
)
L
147
0.61
0.93
DTPA
DOTA
b
a
Reaction conditions: 5 μM Th−Arsenazo III complex, 50 μM ligand,
pH 7.4 (100 mM HEPES, 100 mM KCl), 25 °C. Rates calculated from
the decrease in absorbance at 669 nm (Th−Arsenazo III complex).
b
100 °C.
Table 3. pKa Values of L
Figure 4. Direct kinetics. Stopped-flow experimental conditions: pH
7.4 upon 1:1 mixing of Th4+ (pH 1.4, 79 mM HNO3) and L (pH 8.2,
200 mM HEPES, 10% DMSO), 25 °C. (a) UV−vis spectra of L
(before mixing with Th) and ThL (10 s after mixing, 5 μM Th, 10
equiv of L). (b) A vs t plot of complexation of Th by L upon mixing at
t = 0 (5 μM Th, 15 equiv of L) at 380 nm (black circles represent
experimental values; the red line is the first-order fit). (c) Observed
first-order rate constants (error bars are standard deviations)
calculated at varying concentrations of L relative to thorium. The
second-order rate constant was extrapolated from the slope of the
linear regression.
pKal
12.79(2)
11.9(4)
11.1(2)
9.3(4)
pKa2
pKa3
pKa4
pKa5
pKa6
pKa7
pKa8
pKa9
pKa10
7.57(9)
6.1(6)
5.73(6)
3.95(4)
3.30(7)
1.9(2)
DTPA has a higher pTh at low pH, due to its greater acidity, L
has a remarkably higher pTh at neutral and higher pH, implying
negligible Th dissociation from L in vivo.
dye kinetics enabled the direct comparison of the metal
complexation rates of L with other ligands, in particular the
prevailing chelators DTPA and DOTA. In the metal complex
the terephthalamide binding units are arranged in a novel
topology that may explain the fast complexation and high
stability.
CONCLUSION
■
Solution thermodynamic studies have shown an unprecedented
Th4+ binding affinity for the ligand L, and while kinetic studies
did not provide a similarly quantitative measure, the indirect
E
dx.doi.org/10.1021/ja503456r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 7. Spectrophotometric competition titration of ThL. Starting
conditions: 50 μM Th, 50 μM L, 1 mM DTPA, and 0.1 M KCl (25
°C). (a) A vs wavelength plot at varying pH (data abridged for clarity;
spectra normalized for dilution). (b) A vs pH at 380 nm (experimental
data points are blue circles, red crosses are calculated absorbances)
overlaid onto speciation.
Figure 6. Spectrophotometric titration of L. Starting conditions: 70
μM L, 0.1 M KCl, and 1 mM each of MES, HEPES, and CHES (25
°C). (a) A vs wavelength plot at varying pH (data abridged for clarity;
spectra normalized for dilution). (b) A vs pH at 380 nm (experimental
data points are blue circles, red crosses are calculated absorbances)
overlaid onto speciation.
EXPERIMENTAL SECTION
■
Synthesis. All chemicals were used as supplied without further
purification, unless otherwise noted. Characterization data were
obtained at facilities at the University of California, Berkeley, except
for the elemental analysis of L, which was also performed by Columbia
Analytical Services (Tucson, AZ). NMR spectra were obtained at
room temperature on Bruker AV-500 or AV-600 spectrometers.
Chemical shifts are reported in parts per million (ppm) relative to
solvent residual signals. Mass spectra were obtained at the QB3/
Chemistry Mass Spectrometry Facility, on a Finnigan LTQ FT high-
resolution electrospray ionization (ESI) mass spectrometer. Yields
indicate the amount of isolated material, and reactions were not
optimized. Silica gel (230−400 mesh) was used for column
chromatography purifications. Terephthalamide derivative 4 was
prepared as described previously.23
Figure 8. Speciation diagram at [L] = 10 μM and [Th] = 1 μM.
Bis(Cbz)tren (1). Benzyl phenyl carbamate (Sigma-Aldrich, 20.5 mL,
103.8 mmol) was added dropwise to a stirring CH2Cl2 (50 mL,
degassed) solution of tren (TCI, 7.0 mL, 46.8 mmol) in an ice bath.
Following melting of the ice bath, the reaction was stirred at room
temperature for 12 h. The clear, light yellow reaction mixture was
reduced in vacuo to a yellow oil and purified by silica column
chromatography (5−15% MeOH in DCM). The product fractions
were rotary evaporated and further dried on the vacuum line for 8 h
158.4. HRMS-ESI (m/z): [M + H]+ calcd for C22H31O4N4+ 415.2340,
found 415.2333.
Mono(Boc)-bis(Cbz)tren (2). A solution of NaCl (4.789 g, 81.95
mmol) and NaHCO3 (2.997 g, 35.67 mmol) in 60 mL water was
added to a solution of 1 (10.2754 g, 24.05 mmol) in 50 mL of CH2Cl2.
Di-tert-butyl dicarbonate (Sigma-Aldrich, 5.7373 g, 26.29 mmol)
dissolved in 10 mL of CH2Cl2 was added to the biphasic mixture,
resulting in immediate bubbling. The reaction was let stir at room
temperature overnight, filtered, and extracted with CH2Cl2 (3 × 40
mL). The combined organic extracts were reduced to a viscous, clear,
light yellow oil that was purified by silica column chromatography (0−
10% MeOH in CH2Cl2). The product fractions were rotary evaporated
1
with light (50 °C) heating to a light yellow, translucent oil (72%). H
NMR (600 MHz, CDCl2, δ): 2.45−2.65 (m, NCH2CH2NH2 and
NCH2CH2NHCbz, 8H), 3.24 (s, NCH2CH2NHCbz, 4H), 5.11 (s, Bn
CH2, 4H), 5.16 (br s, amine H, 1H), 6.57 (s, amide H, 2H), 7.33−7.36
(m, Ph H, 10H). 13C NMR (151 MHz, CDCl2, δ): 39.7, 40.0, 49.5,
54.5, 57.3, 66.6, 116.2, 119.2, 128.2, 128.7, 128.8, 129.8, 137.7, 157.3,
1
to a clear, yellow, viscous oil (84%). H NMR (500 MHz, CDCl2, δ):
F
dx.doi.org/10.1021/ja503456r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
CH2), 7.19 (d, J = 10.5 Hz, 1H, ArH), 7.35−7.43 (m, 10H, Ph), 7.84
(d, J = 10.5 Hz, 1H, ArH), 8.01 (br t, 1H, amide H). 13C NMR (100
MHz, CD2Cl2): δ 28.8, 39.5, 55.7, 58.4, 70.8, 76.1, 76.7, 124.0, 126.4,
128.1, 128.4, 128.6, 128.6, 128.6, 128.9, 130.3, 133.6, 135.9, 137.1,
149.6, 150.2, 163.9, 166.7, 201.7. HRMS-ESI (m/z): [M + H]+ calcd
+
for C28H29O5N2S2 537.1512, found 537.1518.
Mono(Boc)-bis(BnTAMthiaz)tren (5). A solution of 3 (0.2792 g,
1.13 mmol) in 200 mL of CHCl3 (with 3 drops of triethylamine) was
added dropwise to a stirring solution of 4 (27.7 g, 47.7 mmol) in 500
mL of CH2Cl2 (with 5 drops triethylamine). The high-dilution slow
addition was performed at room temperature over 18 h. The reaction
mixture was reduced in volume by rotary evaporation and separated by
silica column chromatography (2−5% MeOH in CH2Cl2). 5 was
eluted with 4% methanol and was isolated as a yellow foamy solid after
1
removal of the solvent (66%). H NMR (500 MHz, CD2Cl2): δ 1.36
(s, 9H, CH3), 2.39 (t, J = 6.5 Hz, 4H, tren CH2), 2.46 (t, J = 6.0 Hz,
2H, tren CH2), 2.99 (d, J = 6.0 Hz, 2H, tren CH2), 3.03 (t, J = 7.3, 4H,
tren CH2), 3.23 (q, J = 6.2 Hz,4H, thiaz CH2), 4.40 (t, J = 7.5 Hz, 4H,
thiaz CH2), 5.031 (br t, 1H, amide H), 5.10 (s, 4H, Bn CH2), 5.12 (s,
4H, Bn CH2), 7.17 (d, J = 8.0 Hz, 2H, ArH), 7.35−7.39 (m, 20H, Ph),
7.70 (d, J = 8.0 Hz, 2H, ArH), 7.76 (br t, 2H, amide H). 13C NMR
(100 Hz, CD2Cl2): δ 28.2, 28.9, 37.6, 55.7, 76.2, 76.8, 124.1, 126.2,
128.1, 128.4, 128.6, 128.7, 128.8, 128.9, 130.7, 133.4, 136.0, 137.1,
149.5, 150.1, 164.2, 166.7, 201.7. HRMS-ESI (m/z): [M + H]+ calcd
Figure 9. Molar extinction coefficients of species of L and ThL.
Table 4. log βmlh and pTh Values for Various Ligands
pH
L
DTPA
28.78
ETAM
45.54
log β110
53.7(5)
log β111; pKa
58.9(6); 5.2(2)
30.94; 2.16
15.4
+
a
for C61H65O10N6S4 1169.3640, found 1169.3617.
pTh
3.0
7.4
9.0
12.0
39.1
45.4
7.4
(TrenBoc)2(BnTAM)2 (6). Two 450 mL CHCl3 solutions, one of 5
(0.8503 g, 0.727 mmol) with 5 drops of triethylamine and one of 3
(0.19981 g, 0.811 mmol) with 15 drops of triethylamine, were
combined by dropwise addition into a flask containing 1000 mL of
CH2Cl2. The high-dilution addition took place over 72 h at room
temperature. The reaction mixture became progressively less yellow, as
the starting material 5 is yellow, but thiazolidine and TAM are
colorless in solution. Since the reaction was still yellow once the
addition was complete, diethylenetriamine (1 mL, 9 mmol) was added
to the reaction mixture to facilitate separation from remaining 5, as the
primary amine renders its Rf on silica much lower. After 1.5 h, the
reaction mixture had become colorless and was directly applied on a
flash gradient silica column (0−3% MeOH in CH2Cl2). The product
eluted with 3% methanol and was dried in vacuo to afford a white solid
(87%). 1H NMR (500 MHz, CD2Cl2): δ 1.33 (s, 18H, CH3), 2.45 (br
t, 4H, tren CH2), 2.54 (t, J = 5.3 Hz, 8H, tren CH2), 2.99 (br q, 4H,
tren CH2), 3.38 (q, J = 5.2 Hz, 8H, tren CH2), 4.92 (br s, 2H, amide
H), 5.03 (s, 8H, Bn CH2), 7.10 (s, 4H, ArH), 7.36 (m, 20H, Ph), 7.67
(br s, 4H, amide H). 13C NMR (100 MHz, CD2Cl2): δ 28.1, 37.1,
37.7, 52.6, 76.7, 78.7, 125.1, 128.4, 128.5, 128.6, 131.8, 136.6, 150.4,
25.6
19.4
24.7
28.2 6
a
pTh = −log[Th4+]free; [Th4+]tot = 10−6 M, [L]tot = 10−5 M.
1.36 (s, t-Bu, 9H), 2.51 (br t, NCH2CH2NHBoc, 2H), 2.54 (br t,
NCH2CH2NHCbz, 4H), 3.10 (br t, NCH2CH2NHBoc, 2H), 3.19 (br
t, NCH2CH2NHCbz, 4H), 5.05 (s, Bn H, 4H), 5.10 (br s, RNHBoc,
1H), 5.55 (br s, RNHCbz, 2H), 7.28−7.32 (m, Ph H, 10H). 13C NMR
(151 MHz, CDCl2, δ): 28.7, 39.2, 39.6, 54.6, 54.9, 66.9, 79.5, 128.4,
128.5, 128.9, 139.7, 157.2. HRMS-ESI (m/z): [M + H]+ calcd for
+
C27H39O6N4 515.2864, found 515.2847.
Mono(Boc)tren (3). 2 (11.220 g, 20.20 mmol) was dissolved in 120
mL of MeOH in a glass vessel containing a magnetic stir bar. A 5 mL
MeOH suspension of Pd/C (Sigma-Aldrich, 10 wt %, type E101 NE/
W; 1.1318 g) was added to the solution and the glass vessel was placed
in a Parr bomb. The bomb was purged three times with 750 psi of H2
before being filled to 1250 psi of H2. The reaction was stirred at room
temperature for 24 h with a sustained pressure of 1000 psi. The
reaction mixture was then filtered through a fine glass frit and rotary
evaporated to a light yellow, viscous oil. The oil was coevaporated with
MeOH on the vacuum line and dried for 24 h. NMR showed the
presence of carbonate adduct in the product, which was dissolved in
water and purified by elution through an ion-exchange column
(Dowex 21K XLT resin). Evaporation afforded a clear, yellow oil
(90%), which was stored at 4 °C as a CH2Cl2 solution containing 1
mol equiv of triethylamine. 1H NMR (500 MHz, CDCl2, δ): 1.38 (s, t-
Bu, 9H), 2.44−2.49 (m, NCH2CH2NH2 and NCH2CH2NHBoc, 6H),
2.66 (t, J = 6.0 Hz, 4H), 3.10 (br q, NCH2CH2NHBoc, 2H), 5.78 (br
t, amine H, 1H). 13C NMR (151 MHz, CDCl2, δ): 28.7, 39.3, 40.0,
51.6, 57.8, 79.1, 156.8. HRMS-ESI (m/z): [M + H]+ calcd for
+
155.8, 165.7. HRMS-ESI (m/z): [M + H]+ calcd for C66H81O12N18
1177.5968, found 1177.5962.
Tren2(BnTAM)2[BnTAM(OMeEtN)]2 (8). A solution of 6 (0.26435 g,
0.225 mmol) in 7 mL of CH2Cl2 was placed in an ice−water bath and
stirred. Three milliliters of trifluoroacetic acid (TFA) was added
dropwise over 5 min, and the reaction progress was monitored by
TLC. The reaction mixture was stirred at 0 °C for 10 min and at room
temperature for 20 min. The reaction mixture was dried in vacuo,
affording a pink-orange oil that was redissolved in 2 mL of
dichloromethane and subsequently dried in vacuo (this was performed
twice). One milliliter of triethylamine and 1 mL of CH2Cl2 were added
to the foamy oil at 0 °C, and the mixture was stirred at room
temperature for 45 min before the addition of a solution of 7 (0.56 g,
1.05 mmol) in 8 mL of CH2Cl2. The reaction mixture was stirred at
room temperature for 34 h. Separation of the product from
triethylamine was achieved by extraction with 10 mL of water. The
aqueous layer was acidified to pH 8 (from pH 10) with HCl and
extracted with CH2Cl2 (3 × 5 mL). The combined organic layers were
rotary evaporated and applied onto a gradient silica column (1−5%
MeOH in CH2Cl2). The product was eluted with 5% methanol and
further purified on a silica chromatron plate. The byproducts were
eluted with 2.5% methanol in CH2Cl2, and a slow gradient to 6% was
used to collect the desired product. Rotary evaporation resulted in a
light yellow foamy solid (80%). 1H NMR (500 MHz, CD2Cl2): δ 2.45
(t, J = 8.0 Hz, 4H, tren CH2), 2.58 (t, J = 6.0 Hz, 8H, tren CH2), 3.21
+
C11H27O2N4 247.2129, found 247.2127.
BnTAM(thiaz)(OMeEtN) (7). A solution of 2-methoxyethanamine
(Aldrich, 0.35 mL, 4.03 mmol) in 200 mL of CHCl3 (with four drops
of triethylamine) was added dropwise to a stirring solution of
BnTAM(thiaz)2 (4) (30.010 g, 51.7 mmol), in 1000 mL of CH2Cl2.
The high-dilution slow addition was performed at room temperature
over 16 h. The reaction mixture was then directly applied onto a
gradient silica column (0−10% methanol in dichloromethane), and the
desired product was eluted with 8% methanol. Rotary evaporation
gave a yellow, foamy solid (91%) that became an oil upon storage at 4
°C. 1H NMR (500 MHz, CD2Cl2): δ 3.03 (t, J = 9.0 Hz, 1H,
methoxyethanamide CH2), 3.24 (s, 3H, CH3), 3.39−3.42 (m, 2H,
thiaz CH2), 3.47−3.51 (m, 2H, methoxyethanamide CH2), 4.40 (t, J =
9.3 Hz, 2H, thiazolidine CH2), 5.11 (s, 2H, Bn CH2), 5.12 (s, 2H, Bn
G
dx.doi.org/10.1021/ja503456r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(s, 6H, CH3), 3.23−3.28 (m, 4H, methoxyethanamide CH2), 3.37−
3.41 (m, 12H, methoxyethanamide CH2, tren CH2), 3.49 (q, J = 6.7
Hz, tren CH2), 4.89 (s, 8H, Bn CH2), 5.06 (s, 4H, Bn CH2), 5.07 (s,
4H, Bn CH2), 6.90 (s, 4H, ArH), 7.31−7.42 (m, 40H, Ph), 7.43 (d, J =
10.5 Hz, 2H, ArH), 7.49 (d, J = 10.5 Hz, 2H, ArH), 7.62 (t, J = 6.5 Hz,
2H, amide H), 7.69 (t, J = 6.5 Hz, 4H, amide H), 8.07 (t, J = 5.5 Hz,
2H, amide H). 13C NMR (100 MHz, CD2Cl2): δ 36.7, 37.0, 39.6, 51.8,
52.4, 58.3, 70.6, 76.3, 76.8, 76.9, 124.7, 125.4, 125.8, 128.3, 128.3,
128.5, 128.5, 128.5, 128.6, 130.2, 132.0, 136.1, 136.3, 136.7, 150.3,
150.5, 150.5, 164.3, 164.7, 164.7. HRMS-ESI (m/z): [M + Na]+ calcd
for C106H110O18N10Na+ 1833.7892, found 1833.7866.
aromatic rings). The free ligand crystallized with two chlorine atoms
(balancing the charge on the protonated tertiary amines) and four
water molecules per ligand molecule. The thorium complex crystal-
lized with four potassium atoms, as well as two dimethylformamide,
one tetrahydrofuran, and two methanol molecules per asymmetric
unit. Disordered solvent molecules in this crystal structure (all but one
methanol) were modeled with occupancies of less than 1. One of the
methoxyethyl substituents (atoms C37, O18, C38 and the attached
hydrogens) on the pendant terephthalamide units showed some
disorder, which could be satisfactorily modeled over two positions.
The structure diagrams in Figure 2 were created using ORTEP-32.27
Computational Studies. DFT calculations were performed at the
UC Berkeley Molecular Graphics and Computation Facility with
Gaussian 09 software and the GaussView graphical user interface.28
The geometries and energies of the complexes were optimized at the
B3LYP level with the 6-31G basis set for all atoms except the thorium
atoms, for which the Stuttgart/Dresden ECP60MWB_SEG basis set
was used to model a 60-electron small core pseudopotential,
incorporating quasi-relativistic effects.10
LH8·3HCl. A 5 mL portion of 12.1 N HCl was added to a solution
of 8 (0.32466 g, 0.179 mmol) dissolved in 5 mL of glacial acetic acid.
The solution was stirred at room temperature for 64 h and dried in
vacuo. The resulting light yellow cake was suspended in 10 mL of
MeOH and dried 3 times to give a light yellow solid. Further drying of
this solid under vacuum overnight yielded a light gray solid (89%). 1H
NMR (500 MHz, D2O + NaOD): δ 2.883 (t, J = 7.3 Hz, 4H, tren
CH2), 2.93 (t, J = 6.0 Hz, 8H, tren CH2), 3.34 (s, 6H, CH3), 3.41 (t,
6.3 Hz, 8H), 3.49−3.53 (m, 8H, tren CH2, methoxyethanamide CH2),
3.60 (t, J = 5.5 Hz, 4H, methoxyethanamide CH2), 6.54 (s, 4H, ArH),
6.87 (s, 2H, ArH), 6.87 (s, 2H, ArH). 13C NMR (125 Hz, DMSO-d6):
δ 34.3, 34.4, 39.5, 49.0, 50.4, 51.6, 51.9, 58.4, 70.5, 116.6, 116.7, 116.9,
117.0, 117.3, 118.4, 119.6, 149.5, 150.5, 150.6, 150.9, 169.0, 169.1,
170.8. HRMS-ESI (m/z): [M + Na]+ calcd for C50H62O18N10Na+
Kinetic Studies. Indirect Kinetics. Arsenazo III (Sigma-Aldrich)
was twice purified by HPLC prior to use. The dye (100 mg) was
dissolved in Millipore water, filtered through a 0.22 uM nylon syringe
filter, and separated on a preparatory Varian Dynamax 250 × 41.1 mm
C18 column. A solvent gradient of 0−18% acetonitrile/H2O (0.1%
TFA) was used to collect three fractions with intense UV absorbances.
The fraction with the most intense absorbance (tR 15−16 min) was
collected as the main product, without the front or back tailing
material. This fraction was rotary evaporated, and the dark purple-
green solid was resuspended in methanol before drying under vacuum
−
1113.4136, found 1113.4126; [M − H]− calcd for C50H61O18N10
1089.4171, found 1089.4170. Anal. Calcd (found) for C50H62O18N10·
3HCl: C, 50.03 (50.38); H, 5.46 (5.58); N, 11.67 (11.51). 11.51. Anal.
Calcd (found) for C50H62O18N10·3HCl (Columbia Analytical Serv-
ices): C, 50.03 (50.01); Cl, 8.9 (7.0); H, 5.46 (5.64); N, 11.67 (11.30).
[ThL]4K. LH8·3HCl (13.65 mg, 0.0114 mmol) was suspended in 7
mL of methanol at 45 °C. A solution of Th(NO3)4·4H2O (Alfa, 6.12
mg, 0.0111 mmol) in 1 mL of MeOH was added dropwise to the
ligand solution while stirring, causing an immediate color change to
yellow. A stoichiometric amount of 1 M KOH in methanol (0.09 mL,
0.0910 mmol) was added to the ligand suspension dropwise, to a pH
of 8, solubilizing the reaction mixture. (The thorium solution can be
added after the base.) The reaction mixture was refluxed under
nitrogen flow for 3 h. Once cooled to room temperature, the product
was precipitated by addition of diethyl ether and concentration of the
reaction mixture. The tan precipitate was filtered and dried overnight
overnight. HRMS-ESI (m/z): [M
−
2H]2− calcd for
2−
C22H16As2O14N4S2 386.9274, found 386.9279. Anal. Calcd (found)
for C22H18As2O14N4S2·2H2O: C, 32.53 (32.34); H, 2.73 (2.73); N,
6.90 (6.71).
UV−vis spectra were acquired on a Hewlett-Packard 8453 diode
array spectrometer at 25 °C. The temperature was controlled by a
recirculating water bath connected to the jacketed cell holder. Samples
were measured in small volume (50 μL) quartz cuvettes with 1 cm
path lengths. Solutions were buffered with 100 mM HEPES at pH 7.4
(at 25 °C, with 0.1 M KCl). Complexation of the thorium and dye was
performed by combining aqueous stock solutions of standardized Th4+
in nitric acid (pH 1.4, 1 mM, diluted from a 10 mM solution, the
standardization is described in the next section) and an excess of
purified dye (1 mM). The solution was let sit on the bench at room
temperature for 30 min, diluted in buffer, and let sit at room
temperature for 1 h. The absorbance was monitored starting from the
addition of ligand, as a solution of DMSO ligand stock solution (5
mM) and buffer (equilibrated for 15 min beforehand). DOTA was
purchased from Sigma-Aldrich. The A vs t plots were fit with a first-
order decay equation in Origin 6.1 at wavelengths corresponding to
the appearance of free dye (550, 560 nm) and the disappearance of
thorium−dye complex (614, 669, 700 nm).
Stopped-Flow Kinetics. The stopped-flow kinetic experiments were
performed using a stopped-flow apparatus equipped with an OLis
rapid-scanning monochromator 1000 and a 75 W Xe lamp. Upon
electronic activation, the apparatus (powered by Ar flow) mixes 100
μL of each solution in the two syringes into a mixing chamber. This
mixture is injected into a separate chamber, where the flow is stopped
and UV spectra are taken. One syringe contained 10 μM Th4+ in 79
mM HNO3 (pH 1.4), while the other held a solution of 50−300 μM
ligand, in 200 mM HEPES (pH 8.3) and 10% DMSO. The 1:1 mixture
of these solutions was 5 μM Th, 25−150 μM ligand, 100 mM HEPES
(pH 7.4). UV spectra in the 305−454 nm range were collected for
1.5−10 s, depending on the concentrations of the starting materials, at
rates of 31−1000 scans/s. The data were analyzed using the program
SpecFit to simultaneously use the absorbance at all of the wavelengths
to obtain a second-order rate constant.
1
under vacuum, resulting in a light brown solid (15.11 mg, 93%). H
NMR (500 MHz, D2O + NaOD + DMSO-d6): 2.264 (br t, 2H, tren
CH2), 2.365 (br t, 4H, tren CH2), 2.647 (br t, 4H, tren CH2), 3.069 (s,
6H, CH3), 3.133−3.259 (m, 14H, tren CH2, methoxyethanamide
CH2), 3.510 (d, J = 10 Hz), 4H, methoxyethanamide CH2), 3.743 (d, J
= 11.5 Hz, tren CH2), 6.662−6.681 (d, 2H, ArH; s, 4H, ArH), 6.785
(d, J = 8.5 Hz, ArH). 13C NMR (125 Hz, DMSO-d6): δ 36.2, 38.2,
38.8, 40.3, 56.1, 58.3, 58.4, 58.5, 70.8, 72.2, 109.6, 109.8, 114.5, 114.6,
114.7, 168.3, 168.5, 168.9, 169.0, 171.0, 171.2, 171.4. HRMS-ESI (m/
z): [M + H]3− calcd for C50H54O18N10Th3− 438.1338, found 438.1355.
X-ray Crystallography. Single crystals of LH8 suitable for XRD
were grown by the vapor diffusion of acetonitrile into an aqueous
solution of the protonated ligand. Single crystals of ThLK4 were grown
by the vapor diffusion of 1:1 diethyl ether:tetrahydrofuran into a
solution of the complex in 1:20 dimethylformamide:methanol.
Selected crystals were mounted in Paratone N oil at the end of a
captan loop and frozen in place under a low-temperature nitrogen
stream. The data were collected on Bruker MicroSTAR-H X8 APEX-II
CCD with Cu Kα radiation (LH8) or SMART APEX-I CCD with Mo
Kα radiation (ThLK4) X-ray diffractometers. Intensity data were
extracted from the frames with the program APEX2. The data were
corrected for Lorentz and polarization effects, and an empirical
absorption correction was applied using the SADABS program.24 The
structures were solved by direct methods and refined using full-matrix
least-squares refinements based on F2 in SHELXL-97.25 Crystallo-
graphic analyses were performed using the WinGX system of
programs.26 All non-hydrogen atoms were refined anisotropically,
while hydrogen atoms were assigned to idealized positions (with the
exception of the hydrogens in LH8 bound to heteroatoms and
Solution Thermodynamics. All spectrophotometric titrations
were carried out with constant stirring and a blanket of Ar flow in a
jacketed cell connected to a recirculating water bath to maintain the
temperature at 25 °C. The ionic strength of all solutions was
H
dx.doi.org/10.1021/ja503456r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
maintained at 0.1 M with 0.1 M KCl in titrand solutions and 0.1 M
acid and base titrants. A standardized stock solution of 0.0010(1) M
Th4+ was prepared by dissolving Th(NO3)4·4H2O (Alfa Products) in
Millipore water with concentrated HNO3 (79 mM, pH 1.4). This
solution was complexometrically titrated with Na2H2EDTA (volu-
metric standard, Sigma-Aldrich) to the yellow end point, using
pyrocatechol violet as the indicator.29 L was added as a 50 mM DMSO
solution, prepared by dissolution of the solid ligand, weighed on an
analytical balance accurate to 0.01 mg. The HCl and KOH solutions
were prepared by dilution of Dilut-It (J.T. Baker, ampules)
concentrated solutions with degassed Millipore water. The 0.1 M
HCl solution was standardized by the potentiometric or colorimetric
(using bromocresol green as indicator) titration of tris-
(hydroxymethyl)aminomethane, and the 0.1 M KOH solution was
standardized by potentiometric titration of potassium hydrogen
phthalate or the standardized HCl solution. The KOH solution was
stored under a blanket of Ar flow and standardized before every
titration. The glass electrode (Metrohm Microtrode) used for the pH
measurements was calibrated by the titration of 1.000 mL of
standardized 0.1 M HCl in 25.0 mL of 0.1 M KCl with standardized
0.1 M KOH to pH 11.6. The titration was analyzed using the program
GLEE30 to refine for the E° and slope. This calibration was also
performed prior to each titration. The automated titration system was
controlled by a Metrohm Titrando 907 and the program Tiamo light.
Two-milliliter Dosino 800 burets dosed the titrant into the titration
vessel (5−90 mL). UV−vis spectra were acquired with an Ocean
Optics USB4000-UV−vis spectrometer equipped with a dip probe (set
to a 10 mm path length) and a DH-2000 light source (deuterium and
tungsten lamps), using the program Spectra Suite. Titrations were
performed at least in triplicate.
ASSOCIATED CONTENT
■
S
* Supporting Information
CIF files for LH8 and ThLK4. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
raymond@socrates.berkeley.edu
Notes
The authors declare the following competing financial
interest(s): Kenneth N. Raymond has a financial interest in
Lumiphore Inc., which has licensed University of California
technology in this general field.
ACKNOWLEDGMENTS
■
This research is supported by the Director, Office of Science,
Office of Basic Energy Sciences, and the Division of Chemical
Sciences, Geosciences, and Biosciences of the U.S. Department
of Energy at LBNL under Contract No. DE-AC02-
05CH11231. The authors would like to thank Dr. Manuel
Sturzbecher-Hoehne and Dr. Peter Gans for their help with
titrations, Dr. Antonio DiPasquale for his help with the XRD
data, and Dr. Casey J. Brown for helpful discussions. The
computational work was made possible by NSF grant CHE-
0840505, which funds the UC Berkeley Molecular Graphics
Facility.
Ligand Titration. Ten-milliliter solutions of 50 μM ligand and 20
equiv (1 mM) each of MES, HEPES, and CHES were titrated forward
and backward between pH 1.8 and 12.7 with standardized 0.1 M KOH
and 0.1 M HCl. Data points (pH readings and UV−vis spectra) were
collected following 0.040−1.000 mL titrant additions, with an
equilibration time of 180 s. The 1 mL titrant additions were
performed above pH 11 due to the buffering capacity of water at high
pH and to minimize the exposure of the electrode to basic solution;
0.1 M HCl was added to the solution prior to the forward titration to
start at a pH below 2. All absorbance measurements used in the
refinement were no more than 1.0 absorbance units. Spectra at 250−
450 nm (at intervals of approximately 0.1 nm) were analyzed
(simultaneously) with the program Hypspec.31
REFERENCES
■
(1) (a) Lawrence, J. H.; Tobias, C. A.; Born, J. L. Trans. Am. Clin.
Climatol. Assoc. 1961, 73, 176−185. (b) Lawrence, J. H.; Tobias, C. A.;
Born, J. L.; Gottschalk, A.; Linfoot, J. A.; Kling, R. P. J. Am. Med. Assoc.
1963, 186, 236−245. (c) Lawrence, J. H.; Tobias, C. A.; Born, J. L.;
Linfoot, J. A.; Kling, R. P.; Gottschalk, A. Trans. Am. Clin. Climatol.
Assoc. 1964, 75, 111−116. (d) Kim, Y.-S.; Brechbiel, M. W. Tumor
Biol. 2012, 33, 573−590.
(2) (a) Barendsen, G. W.; Beusker, T. L. J.; Vergroesen, A. J.; Budke,
L. Radiat. Res. 1960, 13, 841−849. (b) Ritter, M. A.; Cleaver, J. E.;
Tobias, C. A. Nature 1977, 266, 653−655.
(3) Kluetz, P. G.; Pierce, W.; Maher, V. E.; Zhang, H.; Tang, S.; Song,
P.; Liu, Q.; Haber, M. T.; Leutzinger, E. E.; Al-Hakim, A.; Chen, W.;
Palmby, T.; Alebachew, E.; Sridhara, R.; Ibrahim, A.; Justice, R.;
Pazdur, R. Clin. Cancer Res. 2014, 20, 9−14.
(4) (a) Dahle, J.; Krogh, C.; Melhus, K. B.; Borrebaek, J.; Larsen, R.
H.; Kvinnsland, Y. Int. J. Radiat. Oncol. Biol. Phys. 2009, 75, 886−895.
(b) Miederer, M.; Scheinberg, D. A.; McDevitt, M. Adv. Drug Delivery
Rev. 2008, 60, 1371−1382. (c) Boswell, C. A.; Brechbiel, M. W. Nucl.
Med. Biol. 2007, 34 (7), 757−778. (d) Zalutsky, M. R.; Pozzi, O. R. Q.
J. Med. Mol. Imaging 2004, 48, 289−296.
(5) Larsen, R.; Bruland, Oeyvind, B. Thorium-227 for Use in
Radiotherapy of Soft Tissue Disease. WO2004091668A1, October 28,
2004.
(6) (a) Harrison, A.; Walker, C. A.; Parker, D.; Jankowski, K. J.; Cox,
J. P. L.; Craig, A. S.; Sansom, J. M.; Beeley, N. R. A.; Boyce, R. A.;
Chaplin, L.; Eaton, M. A. W.; Farnsworth, A. P. H.; Millar, K.; Millican,
A. T.; Randall, A. M.; Rhind, S. K.; Secher, D. S.; Turner, A. Nucl. Med.
Biol. 1991, 18 (5), 469−476. (b) Ruegg, C. L.; Anderson-Berg, W. T.;
Brechbiel, M. W.; Mirzadeh, S.; Gansow, O. A.; Strand, M. Cancer Res.
1990, 50 (14), 4221−4226. (c) Kodama, M.; Koike, T.; Mahatma, A.
B.; Kimura, E. Inorg. Chem. 1991, 30, 1270−1273.
(7) (a) Gorden, A. E. V.; Raymond, K. N. Chem. Rev. 2003, 103,
4207−4282. (b) Gramer, C. J.; Raymond, K. N. Inorg. Chem. 2004, 43,
6397−6402. (c) Durbin, P. W. Health Phys. 2008, 95 (5), 465−492.
(d) Abergel, R. A.; Raymond, K. N. Hemoglobin 2011, 35, 276−290.
Th4+ Competition Titration. Ten-milliliter solutions of 50 μM L, 50
μM Th4+, and 500 μM DTPA (Pharm-Eco) were titrated forward and
backward with between pH 1.8 and 10.0 with 0.1 M KOH and 0.1 M
HCl. The Th4+ was added to a solution of DTPA at pH 7.4, and this
solution was allowed to equilibrate for at least 24 h prior to the
addition of L. The data points (pH readings and UV−vis spectra) were
collected following 0.040−0.100 mL of titrant, with an equilibration
time of 10 min following KOH additions (forward titrations) or 20
min following HCl additions (backward titrations); 0.1 M HCl was
added to the solution prior to the forward titration to start at a pH
below 2. All absorbance measurements used in the refinement were no
more than 1.1 absorbance units. Spectra of 250−450 nm (at intervals
of approximately 0.1 nm) were analyzed (simultaneously) with the
program Hypspec. The following values for the log β for the formation
of Th hydroxides were included in the refinement: [ThOH]3+, −2.5;
[Th(OH)2]2+, −6.2; Th(OH)4, −17.4; [Th2(OH)2]6+, −5.9;
[Th2(OH)3]5+, −6.8; [Th4(OH)8]8+, −20.4; [Th4(OH)12]4+, −26.6;
[Th6(OH)14]10+, −36.8; [Th6(OH)15]9+, −36.8.16 The following
protonation and stability constants for DTPA and the Th−DTPA
complex were also included: logβ011, 10.4; log β012, 18.95; log β013
,
23.23; log β014, 25.93; log β015, 27.93; log β016, 29.53; log β017, 30.23;
log β110, 28.78; log β111, 30.94; log β11−1, −8.88.21,32 Spectra and
protonation constants of the free ligand, refined from the ligand
titrations, were set constant in the refinement.
I
dx.doi.org/10.1021/ja503456r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(8) Benito, J. M.; Gomez-Garcia, M.; Mellet, C. O.; Baussane, I.;
Defaye, J.; Garcia Fernandez, J. M. J. Am. Chem. Soc. 2004, 126,
10355−10363.
(24) SADABS: Bruker Nonius Area Detector Scaling and Absorption
V. 2.05; Bruker Analytical X-ray Systems, Inc.: Madison, WI, 2003.
(25) Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112−122.
(26) Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 837.
(27) ORTEP-3 for Windows: Farrugia, L. J. J. Appl. Crystallogr. 1997,
30, 565.
(9) A recent application of the high-dilution technique is featured
here: Xu, J.; Corneillie, T. M.; Moore, E. G.; Law, G.-L.; Butlin, N. G.;
Raymond, K. N. J. Am. Chem. Soc. 2011, 133 (49), 19900−19910.
(10) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652.
(b) Miehlich, B.; Savin, A.; Stoll, H.; Preuss, H. Chem. Phys. Lett. 1989,
157, 200−206. (c) Lee, C.; Yang, W.; Parr, G. Phys. Rev. B 1988, 37,
785−789. (d) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys.
1972, 56, 2257−2261. (e) http://www.theochem.uni-stuttgart.de/
pseudopotentials/clickpse.en.html. (f) Haeusermann, U.; Dolg, M.;
Stoll, H.; Preuss, H. Mol. Phys. 1993, 78, 1211−1224. (g) Leininger,
T.; Nicklass, A.; Stoll, H.; Dolg, M.; Schwerdtfeger, P. J. Chem. Phys.
1996, 105, 1052−1059. (h) Kuechle, W.; Dolg, M.; Stoll, H.; Preuss,
H. J. Chem. Phys. 1994, 100, 7535−7542. (i) Cao, X.; Dolg, M.; Stoll,
H. J. Chem. Phys. 2003, 118, 487−496. (j) Cao, X.; Dolg, M. J. Mol.
Struct. (THEOCHEM) 2004, 673, 203−209.
(28) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone,V.; Mennucci, B.;
Petersson, G. A., Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.;
Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota,K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Keith,T.; Kobayashi, R.; Normand, J.;
Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.;
Cossi, M.; Rega, N.; Millam, J. M.; Klene,M.; Knox, J. E.; Cross, J. B.;
Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.;
Yazyev,O.; Austin, A. J.; Cammi, R.; Pomelli,C.; Ochterski, J. W.;
Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador,
P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.;
Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09,
Revision B.01; Gaussian, Inc.: Wallingford, CT, 2010.
(29) Schwarzenbach, G. Complexometric Titrations; Interscience
Publishers: New York, 1957; p 75−76.
(30) Gans, P.; O’Sullivan, B. Talanta 2000, 51, 33−37.
(31) (a) Gans, P.; Sabatini, A.; Vacca, A. Talanta 1996, 43, 1739−
1753. (b) Gans, P.; Sabatini, A.; Vacca, A. Ann. Chim. 1999, 89, 45−49.
(32) NIST Standard Reference Database 46: NIST Critically Selected
Stability Constants of Metal Complexes Database, Version 8.0, U.S.
Department of Commerce, 2004.
(11) An algorithm for calculating the dihedral angles in a
coordination polyhedron (only taking into account the metal and
donor atoms) and minimizing their variance from those of idealized
polyhedra was written some years ago ( Xu, J.; Radkov, E.; Ziegler, M.;
Raymond, K. N. Inorg. Chem. 2000, 39, 4156−4164 ) and recently
updated (Tatum, D.; Raymond, K.N. Automated δ dihedral angle
shape analysis for coordination numbers four through nine. Manu-
script in preparation.). The values stated here were calculated using the
updated algorithm..
(12) The m edges of the trigonal dodecahedron are those lying on
the mirror planes: Hoard, J. L.; Silverton, J. V. Inorg. Chem. 1963, 2,
235−250 . An ideal trigonal dodecahedron for the ThL complex was
generated by rotation of the binding unit with the smallest deviation
about the pseudo-S4 axis of the inner coordination environment. The
angles of deviation of the TAMs from ideal m edges were then
calculated from the superposition of the crystal structure coordinates
onto this ideal dodecahedron..
(13) Busch, D. H.; Farmery, K.; Goedken, V.; Katovic, V.; Melnyk, A.
C.; Sperati, C. R.; Tokel, N. Adv. Chem. Ser. 1971, 100, 44.
(14) (a) Wang, X. W.; Jin, T.; Comblin, V.; Lopez-Mut, A.; Merciny,
E.; Desreux, J. F. Inorg. Chem. 1992, 31, 1095−1099. (b) Choi, K.;
Hong, C. P. Bull. Korean Chem. Soc. 1994, 15 (4), 293−297.
(15) (a) Leon-Rodriguez, L. M.; Kovacs, Z. Bioconjugate Chem. 2008,
19 (2), 391−402. (b) Viola-Villegas, N.; Doyle, R. P. Coord. Chem. Rev.
2009, 253 (13−14), 1906−1925.
(16) Rand, M.; Fuger, J.; Grenthe, I.; Neck, V.; Rai, D. In Chemical
Thermodynamics of Thorium; Monpean, F. J., Perrone, J., Illemassene,
M., Eds.; OECD Nuclear Energy Agency (NEA) “Chemical
Thermodynamics” series; OECD Publishing: Paris, France, 2007;
Vol. 11, p 170.
(17) Binstead, R. A.; Zuberbuhler, A. D.; Jung, B. Specfit, version
3.0.30, 1993−2002.
(18) Basargin, N. N.; Ivanov, V. M.; Kuznetsov, V. V.l; Mikhailova. J.
Anal. Chem. 2000, 55, 204−210.
(19) (a) Stradling, G. N.; Gray, S. A.; Pearce, M. J.; Wilson, I.;
Moody, J. C.; Burgada, R.; Durbin, P. W.; Raymond, K. N. Human
Exp. Toxicol. 1995, 14, 165−169. (b) Stradling, G. N.; Hodgson, S. A.;
Pearce, M. J. Radiat. Prot. Dosim. 1998, 79, 445.
(20) Garrett, T. M.; Miller, P. W.; Raymond, K. N. Inorg. Chem.
1989, 28, 128−133.
(21) Brown, M. A.; Paulenova, A.; Gelis, A. V. Inorg. Chem. 2012, 51
(14), 7741−7748.
(22) Harris, W. R.; Carrano, C. J.; Cooper, S. R.; Sofen, S. R.; Avdeef,
A.; McArdle, J. V.; Raymond, K. N. J. Am. Chem. Soc. 1979, 101,
6097−6104.
(23) (a) Doble, D. M. J.; Melchior, M.; O’Sullivan, B.; Siering, C.;
Xu, J.; Pierre, V. C.; Raymond, K. N. Inorg. Chem. 2003, 42, 4930−
4937. An alternative synthesis also yields desired product: (b) Nagao,
Y.; Miyasaka, T.; Hagiwara, Y.; Fujita, E. J. Chem. Soc. Perkin Trans. 1
1984, 183−187.
J
dx.doi.org/10.1021/ja503456r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX