Synthesis and stability studies of Ga-67 labeled phosphonium salts

Article abstract of DOI:10.1002/jlcr.3448

Delocalized lipophilic cations such as tri- and tetra-arylphosphonium are able to diffuse across the mitochondrial membrane, which allows them to selectively accumulate in cells with a high transmembrane potential (ΔΨ_m). The mitochondrial membr

Full text of DOI:10.1002/jlcr.3448

Received: 17 August 2016
DOI 10.1002/jlcr.3448
Accepted: 30 August 2016
R E S E A R C H A R T I C L E
Synthesis and stability studies of Ga‐67 labeled phosphonium salts
1
2
1
Mingyue Kardashinsky | Nigel Lengkeek | Louis M. Rendina
1
School of Chemistry, The University of Sydney,
Delocalized lipophilic cations such as tri‐ and tetra‐arylphosphonium are able to
diffuse across the mitochondrial membrane, which allows them to selectively accu-
Sydney, NSW 2006, Australia
2
ANSTO Life Sciences, Australian Nuclear
Science and Technology Organisation, NSW 2232,
Australia
mulate in cells with a high transmembrane potential (ΔΨ ). The mitochondrial
m
membrane potential of cancer cells and cardiomyocytes has been reported to be
significantly higher than that of normal epithelial cells. This feature can be exploited
for the selective accumulation of phosphonium derivatives for the purposes of molec-
ular imaging using radionuclides. Four structurally related Ga(III)‐phosphonium salts
were synthesized and fully characterized and found to be modest in toxicity toward
T98G human glioblastoma cells (IC50 > 4 mM). High‐activity (100 MBq) analogs
containing Ga‐67 were also synthesized and their stabilities in phosphate‐buffered
saline and human serum were determined.
Correspondence
Louis M. Rendina, School of Chemistry, The
University of Sydney, Sydney, NSW 2006,
Australia.
Email: lou.rendina@sydney.edu.au.
KEYWORDS
cancer, mitochondria, PET chemistry, phosphonium
1
5
1
| INTRODUCTION
myocardial blood flow agent for PET imaging. More
125
recently,
[
[
[
I]‐p‐iodobenzyl triphenyl phosphonium,
I]‐p‐iodobenzyl dipropylphenyl phosphonium, and
I]‐p‐iodobenzylmethyldiphenylphosphonium were also
1
1
25
25
Delocalized lipophilic cations (DLCs) such as tri‐ and tetra‐
arylphosphonium are able to diffuse across the mitochondrial
membrane as its membrane potential is negative and the pro-
cess occurs without the aid of endogenous transporters.
DLCs have been shown to selectively accumulate in cells
with high transmembrane potential (ΔΨm).
the difference in ΔΨ between the colon carcinoma cell line
investigated as potential myocardial SPECT imaging agents,
and all these agents showed good heart tissue uptake.
1
–6
16
Most phosphonium compounds have used several syn-
thetic steps following radiolabeling, an inefficient strategy
when the short half‐lives of many radioisotopes are consid-
ered. This reduces the overall radiochemical yield and
limits potential radionuclides to those with longer half‐lives.
Previously, our research group has investigated similar
Gd(III) phosphonium salts as potential tumor mitochondrial
targeting agents for neutron capture and photon activation
7–9
For example,
m
17
CX‐1 and the control green monkey kidney epithelial cell line
(CV‐1) is reported to be approximately 60 mV (163 mV in
9
tumor cells vs 104 mV in normal cells). By exploiting the
difference in ΔΨm between the 2 mammalian cell lines,
several researchers have reported the synthesis of several
Cu‐64 labeled phosphonium complexes for positron emission
tomography (PET) imaging, some of which have shown a
18,19
therapies.
In vitro studies of these compounds confirmed
their modest toxicity (IC50 > 1 mM), high uptake by T98G
human glioblastoma cells (>1500 ng Gd/mg protein), and
high selectivity of tumor to normal cells compared with the
control noncancerous SVGp12 fetal glial cell line. All these
factors make the phosphonium compounds good potential
imaging agents with radiometals such as Ga‐67. The primary
aim of this study was to develop a rapid method for the
synthesis of structurally related Ga‐67 phosphonium‐based
agents in high radiochemical yield and purity for PET
imaging. Herein, we report the synthesis of 4 structurally
related Ga(III)‐phosphonium salts and radiolabeled analogs
1
0–12
high selectivity for tumors.
The incorporation of radiolabels such as H‐3, C‐11, F‐18,
and I‐125 into phosphonium salts have also proved useful for
medical (PET and SPECT) imaging. For example, tritiated
3
phosphonium cations such as [ H]‐tetraphenylphosphonium
13
have been widely used as in vitro tumor probes, and
1
1
[
C]‐triphenylmethylphosphonium has been used to
determine the kinetics and tumor selectivity in canine brain
glioma by PET imaging. 4‐([ F]‐fluorophenyl)triphenyl‐
phosphonium is another probe developed as a potential
1
4
18
4
Copyright © 2016 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/jlcr
J Label Compd Radiopharm 2017; 60: 4–11

KARDASHINSKY ET AL.
5
containing Ga‐67. Their stabilities in phosphate‐buffered
saline (PBS) and human serum were also determined.
2
2
| RESULTS AND DISCUSSION
.1 | Syntheses
Four structurally related phosphonium ligands L1 to L4
containing a phosphonium center, a xylyl linker and the
macrocycle DO3A (1,4,7,10‐tetraazacyclododecane,‐N,N′,N
1
8,19
″
‐triacetic acid) were synthesized
according to Scheme 1
and purified using reversed‐phase high‐pressure liquid
chromatography (HPLC) on a C18 column. The “cold” Ga
complexes were synthesized in high purity by addition of
GaCl to ligands L1 to L4 in aqueous solution.
3
2.2 | Radiolabeling studies
67
3+
Ligands L1 to L4 were efficiently complexed with Ga
under typical conditions (80°C, 10 minutes), giving
radiochemical yields (based on radio‐HPLC) of >95% for
all complexes (Table 1).
The Ga‐67 radiolabeled complexes containing L1 to L4
were purified using a C18 SPE and optimally eluted (>90%
recovery of activity) in a 1:1 mixture of EtOH/0.9% saline
solution. Ethanol content of <40% gave poor recoveries of
6
7
.
67
.
the more lipophilic complexes
Ga L2 and
Ga L4.
Removal of the EtOH gave a suitably formulated product
for stability and serum protein binding studies. In all cases,
except for 67Ga L3, the final product purity was >95%, as
determined by radio‐HPLC. An uncharacterized minor impu-
.
6
7
rity (<2%) in L3 appears to complex Ga preferentially. The
modification of reaction parameters (time, temperature, and
ligand stoichiometry) had little effect on the product distribu-
6
7
.
tion. Attempts to purify Ga L3 by the above‐mentioned
method also proved problematic. Although adequate separa-
tion of the 2 products was achieved, the overall recovery of
injected activity was <30%. Recoveries were significantly
nat
.
improved by the addition of Ga L3 to the injections, but
this product was unsuitable for the stability studies.
2.3 | Stability studies
The stabilities of the Ga‐67 radiolabeled complexes
containing L1 to L4 were examined in Dulbecco's PBS
SCHEME 1 Synthesis of Complexes 1–4.
(dPBS). Data for all the time points examined in this study
are presented in Table 2. All of the radiolabeled complexes
demonstrated good stability in dPBS. HPLC chromatograms
and 67Ga L4 all showed the rapid (>1 hour) association of
5% to 6% of their activity with serum proteins (Table 3). This
association increased slightly over time, but it appears to be a
system of limited capacity. At the 24‐hour time point, 2% to
3% of the activity was associated with human serum albumin
proteins.
.
6
7
3+
indicated minor loss of Ga from the complexes as well as
the formation of polar fragments, presumably formed by the
67
.
cleavage of the Ga DO3A‐xylyl moiety from the phospho-
nium salt.
Studies involving human serum allowed for the collection
of 2 sets of data: serum protein binding (Table 3) and stability
67
The stability of the Ga complexes in human serum was
examined by removing the majority of proteins (>99%) by
in human serum (Table 4). Complexes 6 Ga L1, Ga L2,
7
.
67
.

6
KARDASHINSKY ET AL.
TABLE 1 Radiochemical yield and purity and SPE purification recovery of
Ga‐67 complexes containing L1 to L4
TABLE 4 Stability of Ga‐67 complexes with L1 to L4 in human serum at
all time points
Radiochemical
yield (HPLC)
Radiochemical
SPE purification
recovery (%)
Complex
1 h
2 h
4 h
24 h
a
a,b
Complex
purity
6
7
.
Ga L1
84.45
90.70
90.78
92.08
76.30
49.77
92.21
54.67
73.19
44.56
91.06
43.04
56.26
28.16
72.49
14.40
6
6
6
6
7
7
7
7
.
Ga L1
98.3 ± 0.5 (n = 6)
97.2 ± 0.5 (n = 6)
97.6 ± 0.4 (n = 3)
96.9 ± 0.6 (n = 3)
95.6 ± 0.9 (n = 3)
97.4 ± 1.2 (n = 3)
91.7 ± 1.8 (n = 3)
98.5 ± 0.4 (n = 3)
98.2 ± 0.9 (n = 3)
94.4 ± 2.4 (n = 3)
95.6 ± 1.4 (n = 3)
91.5 ± 1.0 (n = 3)
67
67
67
.
Ga L2
.
Ga L2
.
Ga L3
.
Ga L3
.
Ga L4
.
Ga L4
a
Based on HPLC peak integrations.
67
.
67
.
b
Ga L3, and Ga L4). All complexes were stable in PBS
After SPE purification.
but showed signs of metabolic degradation after 24 hours in
human serum. The limited solution stability of the complexes
after extended periods allows for possible future application
TABLE 2 Stability of Ga‐67 complexes with L1 to L4 in dPBS at all time
points
17
to the shorter‐lived Ga‐68 isotope. PET has several
technical merits over SPECT, such as higher spatial resolu-
tion and more accurate attenuation correction. Ga‐68 is an
excellent positron emitter (511‐keV annihilation radiation;
t1/2 = 68 minutes) suitable for PET imaging and is available
from a Ge‐68/Ga‐68 generator, which renders it independent
of an on‐site cyclotron. The short overall time required for
labeling and purification using the method reported here
allows for future imaging applications involving selected
Ga‐68 complexes.
Complex
1 h
2 h
4 h
24 h
6
6
6
6
7
7
7
7
.
Ga L1
93.5
94.7
94.4
96.1
94.0
95.8
99.8
96.2
88.2
92.5
99.7
93.9
.
Ga L2
.
Ga L3
100
96.9
100
97.1
.
Ga L4
ultracentrifugation using a 3000 MWCO filter, providing a
filtrate that was suitable for injection on a standard C18
HPLC column. Table 4 shows the stability of these com-
67
.
plexes at all time points. Two of the complexes ( Ga L1
6
7
.
and Ga L3) showed good stability for up to 2 hours. All
complexes showed some degradation after 24 hours, but
6
7
.
Ga L3 remained mostly intact.
4 | EXPERIMENTAL
The metabolic profile of the complexes Ga L1, 67Ga.
6
7
.
6
7
.
67
.
L2, Ga L3, and Ga L4 were investigated using radio‐
HPLC after 24 hours (Figure 1). Because the complexes are
metabolizing in serum, it is unclear if the binding is associ-
ated with the breakdown products or the original complexes.
The main metabolite of these complexes mainly shows up in
4
.1 | Materials and methods
Distilled water was used for all experiments requiring water.
THF and MeCN were dried before use by following the
methods of Armarego and Chai. THF was dried over
sodium wire and freshly distilled from benzophenone ketyl.
Anhydrous MeCN was freshly distilled from CaH . All other
solvents were used without further purification.
20
2
regions (1.5–2.3 and 2.4–3.7 minutes) of the HPLC chro-
2
matograms. The low retention time points to the metabolites
being polar in nature likely formed from the cleavage of the
phosphonium group, presumably at the xylyl‐bridging group.
The first region in the radio‐HPLC chromatogram comprises
unbound Ga and Ga(DO3A), and the second region
comprises Ga(DO3A)‐xylyl metabolites.
All precursor chemicals were commercially available.
1
,4,7,10‐Tetraazacyclododecane (cyclen) was purchased
67
67
from Nowapharm (China). Gallium(III) chloride was pur-
chased from Strem. All other chemicals were purchased from
Sigma‐Aldrich Co.
67
Column chromatography was conducted on Grace
Davison LC60A 40–63 μm silica column. Thin‐layer
chromatography was conducted on Merck Kieselgel 60 F254
aluminum back plates. Visualization of plates was achieved
by using a 254 nm light.
3
| CONCLUSION
We have synthesized and evaluated the stability of 4 new
Ga‐67‐labeled phosphonium complexes ( Ga L1, Ga L2,
6
7
.
67
.
TABLE 3 Binding of Ga‐67 complexes with L1 to L4 to human serum proteins at all time points
Complex 1 h 2 h
4 h
24 h
HSA
Other
4.73
5.97
0
HSA
Other
4.72
6.59
0
HSA
Other
6.90
6.93
0
HSA
1.95
3.08
2.20
2.98
Other
5.44
7.38
0
6
6
6
6
7
7
7
7
.
Ga L1
0
0
0
0
0
0
0
0
0
0
0
0
.
Ga L2
.
Ga L3
.
Ga L4
5.77
5.75
6.08
7.78
HSA indicates human serum albumin.

KARDASHINSKY ET AL.
7
FIGURE
1
Radio‐HPLC chromatograms of
serum ultrafiltrates after 24 hours
4.2 | Instrumentation
4.3.2 | Method 2
1
13
1
31
1
Semipreparative: An Atlantis T3 semipreparative
(10 mm × 250 mm, 5 μm pore size) column was used. The
flow rate was 3 mL/min. The mobile phase was isocratic
30% MeCN, 60% water and with 10% 0.1 M Na HPO buff-
2 4
ered at pH 7.
Analytical: An Atlantis T3 analytical (4.6 mm × 150 mm,
3 μm pore size) column was used. The flow rate was
All H, C{ H}, and P{ H} NMR spectra were recorded
1
13
on a Bruker Avance200 ( H at 200 MHz and C at
0 MHz), a Bruker Avance300 ( H at 300 MHz, C at
5 MHz, and P at 121 MHz), or a Bruker Avance 500
H at 500 MHz, C at 125 MHz, and P at 202 MHz)
NMR spectrometer at 300 K. All NMR signals (δ) are
reported in parts per million. H NMR spectra recorded in
CDCl were referenced to TMS and to their residual solvent
peaks in all other solvents. C{ H} NMR spectra were
1
13
5
7
(
31
III
1
13
31
1
0
.8 mL/min. The mobile phase was isocratic with 0.01 M
3
1
3
1
Na HPO in 30% MeCN and 70% water buffered at pH 7.
2
4
1
3
1
referenced to their solvent peaks (except for D O). C{ H}
2
NMR spectra in D O were referenced according to an inter-
2
4.3.3 | Method 3
3
1
1
nal standard of TMS (25.145020 MHz). P{ H} NMR spec-
tra were referenced according to internal standards of H PO
Semipreparative: An Atlantis T3 semipreparative
410 mm × 250 mm, 5 μm pore size) column was used.
The flow rate was 3 mL/min. The mobile phase was isocratic
3
4
(
n
(
40.480742 MHz). Coupling constants ( J ) are reported in
ij
Hz. Peak multiplicities have been abbreviated as s (singlet),
d (doublet), t (triplet), br (broad), and m (multiplet—
unassignable multiplicity).
Low‐resolution ESI‐MS were recorded on a Finnigan
LCQ mass spectrometer. High‐resolution ESI‐FT‐ICR‐MS
data were recorded on a Bruker 7.0T mass spectrometer.
2
5% MeCN, 65% water and with 10% 0.1 M Na HPO buff-
2 4
ered at pH 7.
Analytical: An Atlantis T3 analytical (4.6 mm × 150 mm,
3
0
μm pore size) column was used. The flow rate was
.8 mL/min. The mobile phase was isocratic with 0.01 M
Na HPO in 20% MeCN and 80% water buffered at pH 7.
2
4
4.3 | HPLC methods
4
.4 | In vitro cytotoxicity assays
All HPLC methods used a Waters HPLC system equipped
with a UV/vis detector (λ = 254, 230 nm).
In vitro cytotoxicity assays were performed using human
glioblastoma multiforme (T98G) cells. The cytotoxicity of
complexes was assessed using 3‐(4,5‐dimethylthiazol‐2‐yl)‐
4
.3.1 | Method 1
2
,5‐diphenyltetrazolium bromide (MTT) assay. Cells were
4
Preparative: A Waters C18 preparative (19 mm × 150 mm,
μm pore size) column was used. The flow rate was 7 mL/
min. The mobile phase had a gradient 0 to 45 minutes, from
00% solvent A (0.1% TFA in water) and 0% solvent B (0.1%
seeded (density 1 × 10 cells per well) in growth medium
(minimum essential medium Eagle supplemented with 10%
fetal bovine serum, penicillin (100 U), streptomycin
(100 μg/mL), and L‐glutamine (2.5 mM, 100 μL) using 96‐
well plates and were allowed to adhere overnight at 37°C.
Cells were incubated at 37°C in a humidified atmosphere of
5
1
TFA in MeCN) to 0% solvent A and 100% solvent B.
Analytical: A Waters C18 analytical (4.6 mm × 150 mm,
5
0
μm pore size) column was used. The flow rate was
.7 mL/min. The mobile phase had a gradient 0 to 45 minutes,
5% CO in the presence of 1 or the vehicle (control). Serial
2
dilutions of the complex were added to triplicate wells. Max-
imum concentration (MaxC) for the experiments was 4 mM.
After 72 hours, the MTT solution in PBS (30 μL, 0.17% w/v)
from 100% solvent A (0.1% TFA in water) and 0% solvent B
0.1% TFA in MeCN) to 0% solvent A and 100% solvent B.
(

8
KARDASHINSKY ET AL.
t
was added, and the incubation was continued. After a further
5.4 | General synthetic method 2: butyl‐protected
phosphonium‐DO3A ligands
4
hours, the culture medium and excess MTT solution were
removed and the resulting MTT‐formazan crystals dissolved
by addition of 150 μL DMSO. Cell viability was determined
by measuring the absorbance at 600 nm using a Victor3V
microplate reader (PerkinElmer). All readings were corrected
for absorbance from wells containing the vehicle alone, and
the level of MTT was expressed relative to the corresponding
vehicle‐treated controls as % viability.
5
5
| SYNTHESES
.1 | General synthetic method 1: bromoxylyl
.
phosphonium salt precursors
t
.
3
A solution of DO3A‐ Bu HBr, phosphonium salt, and
A solution of the phosphine in PhMe was added dropwise to
a solution of a,a ‐dibromo‐m‐xylene in the same solvent. For
the synthesis of P3, the resulting solution was heated at reflux
for 4 hours, whereas for the synthesis of P4, the resulting
solution was stirred at room temperature for 20 hours. The
precipitate was filtered off and washed with PhMe and
diethyl ether.
Na CO in MeCN was stirred at reflux for 20 hours. The
2
3
mixture was filtered, the solvent was removed in vacuo, and
the residue was recrystallized from acetone/diethyl ether to
yield a colorless solid.
5
.5 | Diphenyl(pyridin‐2‐yl)(4‐((4,7,10‐tris(2‐(tert‐
butoxy)‐2‐oxoethyl)‐1,4,7,10‐tetraazacyclododecan‐1‐yl)
methyl)benzyl)phosphonium bromide (C3)
5
.2 | (4‐(Bromomethyl)benzyl)diphenyl(pyridin‐2‐yl)
phosphonium bromide (P3)
Yield: 1.46 g (72.4%). ESI‐MS: m/z 447.93 ([M‐Br ] ). H
−
+
Yield: 1.24 g (84.6%). ESI‐MS: m/z 880.52 ([M‐Br ] ).
−
+
1
1
H NMR (CDCl ) δ 8.17–8.07 (m, 2H, Ph), 7.78–7.64
3
2
NMR (CDCl ) δ 8.93 (d, 1H, Ph, J = 4.5 Hz), 8.39–
8
3
HP
(m, 2H, Ph), 7.53–7.42 (m, 8H, Ph), 7.22–7.19 (m, 2H,
.34 (m, 1H, Ph), 8.09–8.07 (m, 1H, Ph), 7.80–7.74 (m,
Ph), 7.10–7.07 (m, 2H, Ph), 5.35 (br, 2H, CH ), 4.12 (s,
2
1
1H, Ph), 7.15–7.07 (m, 4H, Ph), 5.48 (d, 2H, Ph,
9
H, CH ), 3.59 (s, 2H, CH ), 3.03–2.23 (br m, 25H,
3
3
4
13
13
JHP = 15 Hz), 4.38 (s, 2H, CH ). C NMR (CDCl ) δ
2
3
CH ), 1.46 (s, 18H, CH ), 1.45 (s, 9H, CH ). C NMR
2 3 3
1
1
1
1
52.0 (d, Ph, J = 105 Hz), 144.8 (s, Ph), 143.3 (s, Ph),
CP
(CDCl ) δ 173.5 (s, C = O), 172.6 (s, C = O), 134.7–
3
3
3
38.4 (d, Ph, JCP = 21 Hz), 138.1 (d, Ph, J = 15 Hz),
CP
128.5 (m, Ph), 83.0 (s, C), 82.8 (s, C), 56.1 (m, CH ),
2
2
3
1
35.2 (s, Ph), 134.6 (d, Ph, J = 39 Hz), 131.1 (d, Ph,
CP
49.7 (m, CH ), 28.1 (s, CH ), 28.0 (s, CH ). P NMR
2
3
3
1
2
JCP = 93 Hz), 131.6 (d, Ph, J = 39 Hz),130.1 (d, Ph,
CP
(CDCl ) δ 19.7 (s).
3
1
3
JCP = 48 Hz), 129.5 (s, Ph, J = 12 Hz), 128.3 (s, Ph),
CP
2
1
3
27.5 (d, Ph, JCP = 33 Hz), 117.1 (s, Ph), 115.9 (s, Ph),
1
31
2.8 (s, CH ), 29.5 (d, CH , J = 183 Hz). P NMR
2
2
CP
5
.6 | Tri‐p‐tolyl(4‐((4,7,10‐tris(2‐(tert‐butoxy)‐2‐
(CDCl ) δ 19.5 (s).
3
oxoethyl)‐1,4,7,10‐tetraazacyclododecan‐1‐yl)methyl)
benzyl)phosphonium bromide (C4)
5
.3 | (4‐(Bromomethyl)benzyl)tri‐p‐tolylphosphonium
−
+
1
Yield: 0.89 g (85.0%). ESI‐MS: m/z 922.00 ([M‐Br ] ). H
bromide (P4)
NMR (CDCl ) δ 7.58–7.52 (m, 6H, Ph), 7.46–7.42 (m, 6H,
3
−
+
1
Yield: 0.89 g (94.9%). ESI‐MS: m/z 535.00 ([M‐Br ] ). H
Ph), 7.07–7.05 (m, 4H, Ph), 5.18 (br, 2H, CH ), 3.69 (s,
2
NMR (CDCl ) δ 7.62 (m, 9H, Ph), 7.60 (m, 6H, Ph), 7.58
9H, CH ), 3.72 (s, 2H, CH ), 3.05–2.28 (br m, 25H, CH ),
3
3
3
2
13
(
m, 1H, Ph), 7.55 (m, 1H, Ph), 7.37 (s, 4H, Ph), 7.05 (d,
1.49 (s, 18H, CH ), 1.47 (s, 9H, CH ). C NMR (CDCl3)
3 3
4
2
1
3
1
1
1
1
H, Ph, J = 1.16 Hz), 5.15 (d, 2H, CH , J = 14.52 Hz),
δ 173.4 (s, C = O), 172.5 (s, C = O), 146.2 (s, Ph), 135.8
HP
2
HP
1
3
3
.90 (s, 2H, CH ), 3.90 (s, 9H, CH ). C NMR (CDCl ) δ
46.3 (s, Ph), 137.9 (d, Ph, J = 15 Hz), 134.8 (s, Ph),
34.1 (d, Ph, J = 39 Hz), 131.7 (d, Ph, J = 21 Hz),
30.8 (d, Ph, J = 54 Hz), 129.2 (d, Ph, J = 9 Hz),
(s, Ph), 134.1 (d, Ph, J = 39 Hz), 131.3 (s, Ph), 130.9
2
3
3
CP
3
3
(d, Ph, JCP = 54 Hz), 130.6 (s, Ph), 127.0 (d, Ph,
JCP = 33 Hz), 114.9 (s, Ph), 113.7 (s, Ph), 82.8 (s, C),
CP
2
3
2
CP
CP
1
3
82.4 (s, C), 58.0 (s, CH ), 56.6 (s, CH ), 55.8 (s, CH ),
CP
CP
2
2
2
2
1
27.5 (d, Ph, JCP = 36 Hz), 118.2 (s, Ph), 117.0 (s, Ph),
50.8 (m, CH ), 49.1 (m, CH ), 30.9 (d, CH , J = 195 Hz),
2 2 2 CP
28.0 (s, CH ), 27.8 (s, CH ), 21.8 (s, CH ). P NMR
3
1
1
14.0 (s, Ph), 112.8 (s, Ph), 32.7 (s, CH ), 30.3 (d, CH ,
2
2
3
3
3
1
31
JCP = 189 Hz), 21.7 (s, CH3). P NMR (CDCl ) δ 21.2 (s).
(CDCl ) δ 22.9 (s).
3
3

KARDASHINSKY ET AL.
9
2
5
.7 | General synthetic method 3: phosphonium‐DO3A
(d, Ph, J = 54 Hz), 113.0 (s, Ph), 110.5 (s, Ph), 57.2
CP
ligands
(s, CH
4
2
2
), 54.7 (s, CH
9.5 (m, CH ), 48.4 (m, CH ), 29.7 (d, CH , J = 198 Hz),
0.8 (s, CH3). P NMR (D O) δ 21.3 (s).
2
), 53.0 (s, CH ), 51.1 (m, CH ),
2 2
1
2
2
2
CP
31
2
5
.10 | General synthetic method 4: “cold” Ga(III)
complexes
.
.
The coupled ligand C3 or C4 (0.5 g) was dissolved in
trifluoroacetic acid (5 mL) and stirred at room temperature
for 24 hours. The solvent was removed in vacuo, and the
crude residue dissolved in H O (100 mL) and extracted with
2
CHCl (3 × 20 mL). The aqueous layer was reduced in vacuo
3
Equimolar amounts of the deprotected ligands L1 to L4
to afford an off‐white solid. This solid was then purified by
reversed‐phase HPLC (method 1), and the product fractions
were lyophilized to afford a fluffy white solid.
(
50 mg) and GaCl was stirred inwater (5 mL) at room temper-
3
ature for 72 hours. The reaction mixture was then purified by
reversed‐phase HPLC (method 1), and the product fractions
were lyophilized to yield the product as a fluffy white powder.
Ligands L1 and L2 were prepared as reported
18,19
previously.
5
.11 | 4‐Iodophenyl)diphenyl(4‐((4,7,10‐tris(carboxy‐
5
.8 | Diphenyl(pyridin‐2‐yl)(4‐((4,7,10‐tris
methyl)‐1,4,7,10‐tetraazacyclododecan‐1‐yl)methyl)
benzyl) phosphoniumgallium(III) trifluoroacetate (1)
(carboxymethyl)‐1,4,7,10‐tetraazacyclododecan‐1‐yl)
methyl)benzyl)phosphonium trifluoroacetate (L3)
Yield: 11.8 mg (24.7%). HPLC retention time (method 1)
Yield: 0.37 g (86.5%). HPLC retention time (method
+
−
+
1
=
19.38 minutes. ESI‐FT‐ICR‐MS for [M‐CF CO ] calcu-
1) = 16.32 minutes. ESI‐MS: m/z 836.93 ([M‐Br ] ). H
3
2
1
lated m/z 903.12935; found 903.12859. H NMR (D O) 8.08–
NMR (D O) δ 7.90–7.87 (m, 2H, Ph), 7.80–7.75 (m, 2H,
2
2
8
.06 (m, 2H, Ph), 7.91–7.90 (m, 2H, Ph), 7.71–7.69 (m, 8H,
Ph), 7.61–7.50 (m, 8H, Ph,), 7.38–7.36 (m, 2H, Ph), 7.27–
Ph), 7.35–7.31 (m, 4H, Ph), 7.07–7.06 (m, 2H, Ph), 4.84 (m,
7
.21 (m, 2H, Ph), 7.02–7.00 (m, 2H, Ph), 4.72 (d, 2H, CH2,
2
2
H, CH ), 4.06–3.94 (m, 8H, CH ), 3.57–3.54 (m, 6H, CH ),
2
JHP = 14.7 Hz), 4.35 (br s, 2H, CH ), 4.03 (br s, 2H,
CH ), 3.4–3.1 (br m, 20H, CH ). C NMR (CDCl ) δ
2
2
2
2
1
3
3
.42–3.32 (m, 8H, CH ), 2.98 (d, 2H, CH , J = 11.5 Hz).
2
2
HH
2
2
3
13
C NMR (D O) δ 173.8 (s, C = O), 173.6 (s, C = O), 163.0
1
62.1 (s, C = O), 162.6 (s, C = O), 152.0 (d, Ph,
2
1
2
1
(d, Ph, J = 145 Hz), 139.1 (s, Ph, J = 50 Hz), 135.0 (d,
JCP = 75 Hz), 143.7 (s, Ph), 142.2 (s, Ph), 138.6 (d, Ph,
CP CP
3
3
2
1
2
Ph, J = 40 Hz), 134.0 (d, Ph, J = 40 Hz), 132.0 (d, Ph,
JCP = 42 Hz), 134.2 (d, Ph, J = 39 Hz), 130.0 (s, Ph,
CP
CP
CP
5
4
3
4
JCP = 15 Hz), 131.3 (d, Ph, J = 25 Hz), 130.8 (d, Ph,
JCP = 15 Hz), 130.0 (d, Ph, J = 50 Hz), 128.8 (d, Ph,
JCP = 35 Hz), 119.8 (s, Ph), 117.5 (s, Ph), 117.1 (d, Ph,
JCP = 51 Hz), 118.2 (s, Ph), 116.4 (s, Ph), 115.2 (s, Ph),
CP
2
1
14.4 (s, Ph), 110.5 (s, Ph), 57.1 (s, CH ), 55.0 (s, CH ),
CP
2
2
5
3.2 (s, CH ), 50.8 (m, CH ), 48.5 (m, CH ), 28.7 (d, CH ,
2 2 2 2
1JCP = 55 Hz), 116.4 (d, Ph, J = 50 Hz), 115.2 (s, Ph),
2
1
31
JCP = 192 Hz). P NMR (D O) δ 22.8 (s).
CP
2
4
1
12.9 (s, Ph), 103.4 (d, Ph, J = 15 Hz), 64.2 (s, CH ), 63.3
CP
2
(
s, CH ), 59.4 (s, CH ), 57.3 (s, CH ), 57.0 (s, CH ), 54.5 (s,
2 2 2 2
5
.9 | Tri‐p‐tolyl(4‐((4,7,10‐tris(carboxymethyl)‐1,4,7,10‐
1
31
CH ), 54.1 (s, CH ), 29.1 (d, CH , J = 195 Hz). P NMR
2
2
2
CP
tetraazacyclododecan‐1‐yl)methyl)benzyl)phosphonium
trifluoroacetate (L4)
(D O) δ 23.2 (s).
2
Yield: 0.32 g (79.9%). HPLC retention time (method 1)
5
.12 | Tris(4‐methoxyphenyl)(4‐((4,7,10‐tris(carboxy‐
−
+
1
=
21.17 minutes. ESI‐MS: m/z 753.73 ([M‐Br ] ). H NMR
methyl)‐1,4,7,10‐tetraazacyclododecan‐1‐yl)methyl)
benzyl)phosphoniumgallium(III) trifluoroacetate (2)
(
(
D O) δ 7.36–7.24 (m, 14H, Ph), 6.97–6.95 (m, 2H, Ph), 4.56
2
2
d, 2H, CH , J = 15 Hz), 4.25 (br s, 2, CH ), 4.02 (br s, 2,
2
HP
2
1
3
CH ), 3.4–3.1 (br m, 20H, CH ), 2.27 (s, 9 H, CH ). C NMR
Yield: 28.3 mg (58.5%). HPLC retention time (method 1)
= 20.22 minutes. ESI‐FT‐ICR‐MS for [M‐CF CO ] calcu-
lated m/z 867.26440; found 867.26308. H NMR (D O) δ
7.49–7.42 (m, 6H, Ph), 7.29–7.26 (m, 2H, Ph), 7.13–7.12
2
2
3
+
(
1
(
D O) δ 162.7 (s, C = O), 162.2 (s, C = O), 146.9 (s, Ph),
2
3
2
2
1
33.5 (d, Ph, J = 39 Hz), 132.0 (s, Ph), 131.3 (s, Ph), 130.6
CP
2
1
d, Ph, J = 51 Hz), 122.1 (s, Ph), 118.2 (s, Ph), 114.3
CP

1
0
KARDASHINSKY ET AL.
6
7
(
m, 6H, Ph), 7.03–7.00 (m, 2H, Ph), 4.62 (d, 2H, CH2,
GaCl in 0.1 M HCl and HEPES buffer (1.0 M) were
3
2
JHP = 14.7 Hz), 4.07–4.03 (m, 2H, CH ), 3.94 (s, 6H,
mixed in a ratio of 2:1 in an acid‐washed 1.5 mL Eppendorf
tube. The pH was adjusted with 0.5 M KOH to 4 to 4.5, and
5 nmol of deprotected ligand (L1–L4) was added. The mix-
ture was heated on a heating block at 80°C for 10 minutes.
The crude product was purified by HPLC (method 2 for
2
CH ), 3.89 (s, 9H, CH ), 3.5–3.3 (m, 14H, CH ), 2.99 (d,
2
C = O), 164.3 (s, Ph), 163.1 (s, Ph), 162.6 (s, Ph), 135.8
3
3
2
2
13
H, CH , J
= 10.2 Hz). C NMR (D O) δ 173.7 (s,
HH 2
2
2
(d, Ph, J = 45 Hz), 131.8 (s, Ph), 131.2 (s, Ph), 130.5
CP
67
.
67
.
67
.
67
.
(
s, Ph), 129.4 (s, Ph), 118.3 (s, Ph), 115.6 (d, Ph,
Ga L1, Ga L2, and Ga L4; method 3 for Ga L3).
1
JCP = 54 Hz), 114.4 (s, Ph), 108.8 (s, Ph), 107.6 (s, Ph),
The purified fractions were diluted to 20 mL and loaded onto
Oasis HLB (hydrophilic‐lipophilic–balanced reversed‐phase
sorbent) sample extraction product and washed with water
5
5
9.4 (s, CH ), 57.3 (s, CH ), 57.0 (s, CH ), 55.8 (s, CH ),
2 2 2 2
3
1
4.5 (s, CH ), 54.0 (s, CH ). P NMR (D O) δ 20.5 (s).
2
2
2
(3 × 3 mL). The compound was eluted off with a 1:1 mixture
of EtOH/0.9% saline solution, which was then removed using
a centrifuge evaporator.
5.13 | Diphenyl(pyridin‐2‐yl)(4‐((4,7,10‐tris(carbo‐
xymethyl)‐1,4,7,10‐tetraazacyclododecan‐1‐yl)methyl)
benzyl) phosphoniumgallium(III) trifluoroacetate (3)
ACKNOWLEDGMENTS
The authors thank the ARC for funding and ANSTO for
access to their radiochemistry facilities.
Yield: 4.4 mg (9.2%). HPLC retention time (method 1)
+
=
14.51 minutes. ESI‐FT‐ICR‐MS for [M‐CF CO ] calcu-
3
2
1
lated m/z 778.22795; found 778.22777. H NMR (D O) δ
2
REFERENCES
8
4
7
.93–8.92 (m, 1H, Ph), 8.11–8.09 (m, 1H, Ph), 7.91–7.86 (m,
H, Ph), 7.69–7.29 (m, 8H, Ph), 7.16–7.14 (m, 2H, Ph), 7.08–
1
. Adams DM, Ji W, Barth RF, Tjarks W. Comparative in vitro evaluation of
dequalinium B, a new boron carrier for neutron capture therapy (NCT). Anti-
cancer Res. 2000;20:3395–3402.
.05 (m, 2H, Ph), 4.90–4.85 (m, 2H, CH ), 4.05–3.93 (m, 8H,
2
CH ), 3.76–3.73 (m, 2H, CH ), 3.55–3.29 (m, 12H, CH ),
2
2
2
2. Modica‐Napolitano JS, Aprille JR. Delocalized lipophilic cations selectively
target the mitochondria of carcinoma cells. Adv Drug Deliv Rev.
3
1
3
.01–2.98 (m, 2H, CH2). P NMR (D O) δ 19.5 (s).
2
2001;49:63–70.
3. Rowe TC, Weissig V, Lawrence JW. Mitochondrial DNA metabolism
targeting drugs. Adv Drug Deliv Rev. 2001;49:175–187.
5
1
.14 | Tri‐p‐tolyl(4‐((4,7,10‐tris(carboxymethyl)‐
,4,7,10‐tetraazacyclododecan‐1‐yl)methyl)benzyl)
4
. Takasu K, Shimogama T, Saiin C, et al. Synthesis and evaluation of β‐
carbolinium cations as new antimalarial agents based on π‐delocalized
lipophilic cation (DLC) hypothesis. Chem Pharm Bull. 2005;53:653–661.
phosphoniumgallium(III) trifluoroacetate (4)
5
. Modica‐Napolitano JS, Nalbandian R, Kidd ME, Nalbandian A, Nguyen CC.
The selective in vitro cytotoxicity of carcinoma cells by AZT is enhanced by
concurrent treatment with delocalized lipophilic cations. Cancer Lett.
Yield: 26.0 mg (53.7%). HPLC retention time (method 1)
+
=
20.87 minutes. ESI‐FT‐ICR‐MS for [M‐CF CO ] calcu-
3
2
1
2003;198:59–68.
lated m/z 819.27965; found 819.27947. H NMR (D O) δ
2
6
7
. Galeano E, Nieto E, Garcia‐Perez AI, et al. Effects of the antitumoural
dequalinium on NB4 and K562 human leukemia cell lines. Mitochondrial
implication in cell death. Leuk Res. 2005;29:1201–1211.
7
.76–7.60 (m, 12H, Ph), 7.39 (br s, 2H, Ph), 7.21 (br s, 2H,
Ph), 4.41 (br s, 2H, CH ), 4.04– 3.99 (m, 4H, CH ), 3.5–
3
2
2
1
3
.1 (m, 24H, CH2). C NMR (D O) δ 163.1 (s, C = O),
2
. Johnson LV, Walsh ML, Chen LB. Localization of mitochondria in living
cells with rhodamine 123. Proc Natl Acad Sci U S A. 1980;77:990–994.
1
62.7 (s, C = O), 133.0 (s, Ph), 131.3 (s, Ph), 131.0 (d, Ph,
2
1
JCP = 36 Hz), 129.0 (d, Ph, J = 48 Hz), 128.7 (s, Ph),
CP
8. Lampidis TJ, Bernal SD, Summerhayes IC, Chen LB. Selective toxicity of
1
18.1 (s, Ph), 114.2 (s, Ph), 57.3 (s, CH ), 55.1 (s, CH ),
rhodamine 123 in carcinoma cells in vitro. Cancer Res. 1983;43:716–720.
2
2
5
2
3.0 (s, CH ), 51.1 (s, CH ), 49.4 (s, CH ), 48.3 (s, CH ),
9. Modica‐Napolitano JS, Aprille JR. Basis for the selective cytotoxicity of rho-
damine 123. Cancer Res. 1987;47:4361–4365.
2
2
2
2
1
31
9.6 (d, CH , J = 261 Hz). P NMR (D O) δ 38.2 (s).
2
CP
2
1
0. Wang J, Yang CT, Kim YS, et al. 64Cu‐Labeled triphenylphosphonium and
triphenylarsonium cations as highly tumor-selective imaging agents. J Med
Chem. 2007;50:5057‐5069.
6
| RADIOLABELING STUDIES
1
1. Yang CT, Kim YS, Wang J, et al. 64Cu‐labeled 2‐(diphenylphosphoryl)ethyl
diphenylphosphonium cations as highly selective tumor imaging agents:
effects of linkers and chelates on biodistribution characteristics. Bioconjug
Chem. 2008;19:2008‐2022.
6.1 | General synthetic method 5: Ga‐67 complexes
.
12. Kim YS, Yang CT, Wang J, et al. Radiolabeled triphenylphosphonium cations
as highly tumor‐selective imaging agents: Effects of radiometals, bifunctional
chelators and molecular charge. J Med Chem. 2008;51:2971‐2984.
1
3. Madar I, Weiss L, Izbicki G. Preferential accumulation of (3)H‐
tetraphenylphosphonium in non‐small cell lung carcinoma in mice: compari-
son with (99m)Tc‐MIBI. J Nucl Med. 2002;43:234–238.
1
1
4. Madar I, Anderson JH, Szabo Z, et al. Enhanced uptake of [11C]‐TPMP in
canine brain tumor: a PET study. J Nucl Med. 1999; 40:1180–1185.
5. Shoup TM, Elmaleh DR, Brownell AL, et al. Evaluation of (4‐[18F]
Fluorophenyl)triphenylphosphonium ion. A potential myocardial blood flow
agent for PET. J Mol Imaging Biol. 2011;13:511–517.

KARDASHINSKY ET AL.
11
1
1
1
1
6. Sakai T, Saito Y, Takashima M, Ogawa M, Magata Y. Development of
radioiodinated lipophilic cationic compounds for myocardial imaging. Nucl
Med Biol. 2015;42:482–487.
20. Armarego WLF, L CCL. Purification of Laboratory Chemicals. 7th ed.
Waltham, MA: Butterworth‐Heinemann;2013.
7. Wadas TJ, Wong EH, Weisman GR, Anderson CJ. Coordinating radiometals
of copper, gallium, indium, yttrium, and zirconium for PET and SPECT imag-
ing of disease. Chem Rev. 2010;110:2858–2902.
How to cite this article: Kardashinsky, M., Lengkeek,
N., and Rendina, L. M. Synthesis and stability studies of
Ga‐67 labeled phosphonium salts. J Label Compd
Radiopharm 2017;60:4–11. doi: 10.1002/jlcr.3448.
8. Morrison D, Aitken J, De Jonge M, Ioppolo J, Harris HH, Rendina LM. High
mitochondrial accumulation of new gadolinium(III) agents within tumour
cells. Chem Commun. 2014;50:2252–2254.
9. Morrison D, Aitken J, De Jonge M, Issa F, Harris HH, Rendina LM. Synthesis
and biological evaluation of a class of mitochondrially-targeted gadolinium
(III) agents. Chem Eur J. 2014;20:16602–16612.