Preparation and biological evaluation of radioiodine-labeled triphenylphosphine derivatives as mitochondrial targeting probes

Article abstract of DOI:10.1002/jlcr.3910

The positive-charged lipophilic triphenylphosphonium cations (TPPs⁺) have been served as mitochondrial targeting vehicles for the delivery of various probes. In this study, we developed a new method for the preparation of radioiodine-labeled TPPs⁺. Four ¹²⁵I-labeled TPPs⁺, [¹²⁵I] 9–[¹²⁵I] 12, were prepared from the corresponding triphenylphosphine phenylborate precursors of B 5–B 8 via an optimized copper-catalyzed one-step procedure in high radiochemical yield (>95%). After radio-HPLC purification, the final products could be obtained with high specific activity. Their physicochemical properties, in vitro cellular uptake, and ex vivo mice biodistribution were investigated. The results suggested the ¹²⁵I-labeled TPPs⁺ were lipophilic and could specifically accumulate in the mitochondrial-rich myocardial cells through the mitochondrial membrane potential.

Full text of DOI:10.1002/jlcr.3910

Received: 25 February 2021
DOI: 10.1002/jlcr.3910
Revised: 12 April 2021
Accepted: 13 April 2021
R E S E A R C H A R T I C L E
Preparation and biological evaluation of radioiodine-labeled
triphenylphosphine derivatives as mitochondrial targeting
probes
Shuyu Shi1,2
| Zelan Liu1,2
| Zhenmin Wu1,2 | Hang Zhou1,2 | Jie Lu1,2
1Key Laboratory of Radiopharmaceuticals,
Beijing Normal University, Ministry of
Education, Beijing, China
2College of Chemistry, Beijing Normal
University, Beijing, China
The positive-charged lipophilic triphenylphosphonium cations (TPPs+) have
been served as mitochondrial targeting vehicles for the delivery of various
probes. In this study, we developed a new method for the preparation of
radioiodine-labeled TPPs+. Four 125I-labeled TPPs+, [125I] 9–[125I] 12, were
prepared from the corresponding triphenylphosphine phenylborate precursors
of B 5–B 8 via an optimized copper-catalyzed one-step procedure in high
radiochemical yield (>95%). After radio-HPLC purification, the final products
could be obtained with high specific activity. Their physicochemical properties,
in vitro cellular uptake, and ex vivo mice biodistribution were investigated.
The results suggested the 125I-labeled TPPs+ were lipophilic and could
specifically accumulate in the mitochondrial-rich myocardial cells through the
mitochondrial membrane potential.
Correspondence
Jie Lu, Key Laboratory of
Radiopharmaceuticals, Beijing Normal
University, Ministry of Education; or
College of Chemistry, Beijing Normal
University, Beijing 100875, China.
Email: ljie74@bnu.edu.cn
Funding information
National Natural Science Foundation of
China, Grant/Award Number: 21976019
K E Y W O R D S
125I, copper-catalyzed radioiodination, mitochondrial membrane potential,
triphenylphosphonium cations
1 | INTRODUCTION
highly lipid soluble. In addition, the positive charge on
the phosphorus atom can be delocalized to the three ben-
Mitochondria are recognized as the powerhouse of the
cell. Their central function is the synthesis of ATP by
oxidative phosphorylation.1 Mitochondrial dysfunction
can lead to a wide range of debilitating or life-threatening
diseases such as cancer, heart failure, and neurodegener-
ative diseases.2–5 Therefore, mitochondria are regarded as
an important target for early diagnosis and treatment of
these diseases.
However, the highly hydrophobic inner membrane
and negative membrane potential of mitochondria result
in limited diffusive transport. The majority of
mitochondria-targeted probes commonly contain lipo-
philic, cationic moieties, including dequalinium (DQA),
rhodamine, triphenyl phosphonium cations (TPPs+), and
peptide sequences.6 Among them, the TPPs+ contain
three phenyl groups, which make the whole molecule
zene rings. These properties make it easy for TPPs+ to
pass through phospholipid bilayers into the cytoplasm
bilayers by passive diffusion.7–9 Driven by the mitochon-
drial membrane potential (ΔΨ), TPPs+ can be further
accumulated from the cytoplasm into mitochondria. The
uptake of lipophilic cations into mitochondria increases
10-fold for every 61.5 mV of membrane potential at 37ꢀC,
which lead to 100- to 500-fold accumulation of TPPs+
within mitochondria.6 Thus, TPPs+ have been used to
attach to various small molecules for the diagnosis or
treatment of these cancer and cardiovascular disease,
such as the antioxidants quinone,10 Vitamin E,11 and
ebselen,12 as well as the radionuclides.13–15
Despite the widespread clinical applications of
99mTc-MIBI and 99mTc-tetrofosmin in the field of nuclear
cardiology, the two 99mTc-based monocationic complexes
J Label Compd Radiopharm. 2021;1–11.
wileyonlinelibrary.com/journal/jlcr
© 2021 John Wiley & Sons, Ltd.
1

2
SHI ET AL.
still have the drawbacks, such as low first-pass myocar-
dial extraction, underestimation of the myocardial blood
flow at high flow rates, and relatively high liver
uptake.16,17 With the development of myocardial molecu-
lar imaging probes, radionuclide-labeled TPPs+ deriva-
tives for mitochondria targeting have become the focus of
research.14,15,18,19 Compared with 99mTc, the chemistry
of iodine is well defined and more suited for labeling the
small molecules with biological activity. Four radioiodine
isotopes are commonly used in nuclear medicine.
123I (EC, Eγ = 159 keV, T1/2 = 13.2 h) is used for single
photon emission computed tomography (SPECT).
124I (EC/β+, T1/2 = 4.18 d) is currently applied in positron
emission tomography (PET). 131I (βÀ, T1/2 = 8.04 d)
therapy is an effective measure for the radiation
treatment. 125I (EC, Eγ = 35 keV, T1/2 = 59.4 d) is
particularly well suited for kinds of researches such as
receptor binding assays and biodistribution studies.
However, to date, only a handful of radioiodine-labeled
TPPs+ methods have been reported. Typical examples
include (1) the radiosynthesis of 125I-labeled TPPs+ that
is based on trans-vinylboronic acid reported by
Srivastava's group18 and (2) a two-step method using the
tributylstannyl as the labeling precursors reported by
Yasuhiro Magata's group.19 These methods mentioned
above still hold the drawbacks of multi-step reactions,
long reaction time, and low radiochemical yield, which
hinder their further application. In order to provide
clinicians more choices, we developed an efficient
one-step method to obtain radioiodine-labeled TPPs+
using corresponding phenylboronic-TPP as labeling
precursors (Scheme 1). In theory, the method of radio-
labeling has commonalities between the four radioiodine
nuclides. With the establishment of this new synthetic
approach, we prepared a series of meta-125I-labeled
TPPs+ with or without methoxy group, and evaluated the
potential application of these compounds as
mitochondria targeting probes.
2 | EXPERIMENTAL
2.1 | General
All reagents used for chemistry and biology were
purchased from commercial sources and used without
further purification.
[
125I]NaI was purchased from
PerkinElmer. Reactions were controlled by thin
layer chromatography on aluminum TLC plates,
silica gel 60 coated with fluorescent indicator F254
SCHEME 1 Synthetic route of benzyl TPP cations [125I] 9–[125I] 12 and the corresponding nonradioactive reference compounds 13–16.
a. [125I]NaI, Cu2O, 1,10-phenamthroline, rt, 1 h

SHI ET AL.
3
(Merck, German), and visualized under ultraviolet light
2.2.2 | (3-Boronobenzyl)-tris
(2-methoxyphenyl)phosphonium cation (B-6)
1
(254 nm). H and 13C NMR spectra were detected by a
JEOL JNM-ECZ (600 MHz) spectrometer and a Bruker
(400 MHz) spectrometer. 31P NMR spectra were detected
on a Bruker (400 MHz) spectrometer. High-resolution
mass spectra (HRMS) were obtained by a Bruker
microTOF-QII or AB SCIEX Triple TOF™ 5600+ mass
spectrometer. High-performance liquid chromatography
(HPLC) was achieved on a Shimadzu SCL-20 AVP HPLC
system equipped with a SPD-M20AUV-Vis detector and a
Bioscan Flow Count 3200 NaI/PMT γ-radiation
scintillation detector. The HPLC methods were shown as
following: A: water; B: 0.1% trifluoroacetic acid in
acetonitrile, 0–2.0 min, 10% B; 2.0–15.0 min, 10%–90% B;
15.0–20.0 min, 90%–10% B, flow rate: 1 ml/min for
analytical column (Kromasil 100-5-C18, 250 Â 4.6 mm).
The rat embryonic cardiomyoblast cell line (H9c2)
and mouse normal fibroblast cell line (NIH/3T3) were
obtained from China Infrastructure of Cell Line
Resources and cultured in DMEM medium containing
10% fetal bovine serum (Gibco) and 1% penicillin-
streptomycin. Cell lines were all incubated as a mono-
layer at 37ꢀC with saturated humidity and 5% CO2.
Radioactive count was measured on a PerkinElmer
Automatic Gamma Counter system (WIZARD2, 2480).
The Kunming mice (18–22 g, female) were purchased
from Beijing Vital River Laboratories. All protocols
related to the use of animals were approved by the
Institutional Animal Care Committee of Beijing Normal
University.
B-6 (white solid, 0.498 g, 78.7%) was prepared as the
same way described above. 1H NMR (600 MHz,
MeOD-d4, chemical shift δ in ppm relative to TMS):
δ 7.75 (t, J = 7.9 Hz, 3H), 7.46–7.30 (m, 5H), 7.21–7.16
(m, 6H,), 7.12 (s, 2H), 4.70 (s, 2H), 3.63 (s, 9H). 13C NMR
(101 MHz, MeOD-d4):
δ 161.62 (d, Jcp = 2.4 Hz),
136.78 (d, Jcp = 2.2 Hz), 135.74, 135.07 (d, Jcp = 8.2 Hz),
133.81 (d, Jcp = 3.7 Hz), 132.71, 132.48, 131.29, 127.42,
121.51 (d, Jcp = 12.8 Hz), 112.47 (d, Jcp = 6.8 Hz), 106.18
(d, Jcp = 91.8 Hz), 55.15, 30.18 (d, Jcp = 53.0 Hz).
31P NMR (400 MHz, MeOD-d4): δ 25.1112. HRMS
(+TOF MS): m/z calculated for C28H29BO5P+ [M]+
487.1840, found 487.1850.
2.2.3 | (3-Boronobenzyl)-tris
(3-methoxyphenyl)phosphonium cation (B-7)
B-7 (white solid, 0.523 g, 82.6%) was prepared as the
same way described above. 1H NMR (600 MHz,
MeOD-d4, chemical shift δ in ppm relative to TMS): δ
7.61 (td, J = 8.0, 4.5 Hz, 4H), 7.42 (d, J = 8.5 Hz, 3H),
7.29 (s, 1H), 7.25 (t, J = 7.5 Hz, 1H), 7.15 (dd, J = 12.5,
7.7 Hz, 3H), 7.08 (dd, J = 13.8, 1.7 Hz, 4H), 4.91
(d, J = 14.8 Hz, 2H), 3.76 (s, 9H). 13C NMR (101 MHz,
MeOD-d4): δ 160.61 (dd, Jcp = 15.9, 0.9 Hz), 133.76,
133.73 (d, Jcp = 0.9 Hz), 132.15, 131.44 (d, Jcp = 15.2 Hz),
128.19, 126.11 (d, Jcp = 9.5 Hz), 120.83, 119.39, 118.89
(d, Jcp = 10.2 Hz), 118.54, 55.07, 29.48 (d, Jcp = 47 Hz).
31P NMR (400 MHz, MeOD-d4): δ 23.9460. HRMS (+TOF
MS): m/z calculated for C28H29BO5P+ [M]+ 487.1840,
found 487.1846.
2.2 | Synthesis procedure and
characterization
2.2.1 | (3-Boronobenzyl)
triphenylphosphonium cation (B-5)
2.2.4 | (3-Boronobenzyl)-tris
A mixture of triphenylphosphine (0.472 g, 1.8 mmol) and
(3-(bromomethyl)phenyl)boronic acid (0.279 g, 1.3 mmol)
in 4 ml of toluene was heated at 60ꢀC for 3 h. Then,
toluene was removed by vacuum. The crude product was
purified via column chromatography (silica gel, 20:1 [v/v]
dichloromethane/methanol) to obtain B-5 as a white
solid (0.386 g, 74.8%). 1H NMR (600 MHz, MeOD-d4,
chemical shift δ in ppm relative to TMS): δ 7.89–7.85
(m, 3H), 7.71–7.67 (m, 6H), 7.65–7.59 (m, 6H), 7.43
(d, J = 7.6 Hz, 2H,), 6.96 (m, 2H), 4.89 (d, J = 15.1 Hz,
2H). 13C NMR (150 MHz, MeOD-d4): δ 135.15, 135.13,
134.07, 134.00, 133.79, 130.06, 129.97, 118.10, 117.53,
29.56. 31P NMR (400 MHz, MeOD-d4): δ 23.3978. HRMS
(+TOF MS): m/z calculated for C25H23BO2P+ [M]+
397.1523, found 397.1367.
(4-methoxyphenyl)phosphonium cation (B-8)
B-8 (white solid, 0.563 g, 88.9%) was prepared as the
same way described above. 1H NMR (600 MHz,
MeOD-d4, chemical shift δ in ppm relative to TMS): δ
7.66 (s, 1H), 7.48 (m, 6H), 7.19–7.17 (m, 8H), 7.05
(d, J = 7.2 Hz, 1H), 4.66 (s, 2H), 3.90 (s, 9H). 13C
NMR (150 MHz, MeOD-d4): δ 166.41 (d, Jcp = 2.6 Hz),
137.99, 137.33 (d, Jcp = 7.4 Hz), 134.81, 133.74, 133.47,
129.26 (d, Jcp = 17.1 Hz), 128.58 (d, Jcp = 27.1 Hz),
116.94 (d, Jcp = 25.8 Hz), 109.58 (d, Jcp = 13.7 Hz),
55.21, 31.07 (d, Jcp = 60.6 Hz). 31P NMR (400 MHz,
MeOD-d4):
δ
21.4810. HRMS (+TOF MS): m/z
calculated for C28H29BO5P+ [M]+ 487.1840, found
487.1846.

4
SHI ET AL.
2.2.5 | (3-Iodobenzyl)-
trisphenylphosphonium (13)
7.18 (d, J = 2.3 Hz, 1H), 7.04 (d, J = 8.0 Hz, 1H), 7.00 (t,
J = 7.8 Hz, 1H), 4.62 (d, J = 17.9 Hz, 2H). 3.91 (s, 9H).
31P NMR (400 MHz, MeOD-d4): δ 21.7485. HRMS (+TOF
MS): m/z calculated for C28H27IO3P+ [M]+ 569.0737,
found 569.0735.
A mixture of triphenylphosphine (0.340 g, 1.3 mmol) and
3-iodobenzyl bromide (0.504 g, 1.7 mmol) in 4 ml of tolu-
ene was heated at 60ꢀC for 7 h. After cooling to ambient
temperature, the solvents were removed by evaporation
under vacuum, and the obtained residue was washed
with diethyl ether for twice, dried, and directly obtained
2.3 | Radiochemistry
1
13 as a white solid (0.601 g, 96.6%). H NMR (600 MHz,
The meta-125I-TPPs+, [125I] 9–[125I] 12 were synthesized
from the corresponding phenylboronic acid labeling pre-
cursors via an optimized copper-catalyzed method
(shown in Scheme 1).20 Briefly, a freshly prepared mix-
ture of Cu2O (4 μmol), 1,10-phenanthroline (8 μmol) in
0.8 ml of acetonitrile and 0.2 ml of water was stirred for
10 min at room temperature. In a reaction vial, the
corresponding phenylboronic precursor (0.5 mg, 1 μmol)
was dissolved in 100 μl of acetonitrile followed by the
addition of 100 μl of Cu2O/1,10-phenanthroline mixture
and 125I-NaI (14.8 MBq, 4–5 μl). The vial was not sealed
to allow air entering through a simple activated carbon
adsorption unit and kept at room temperature for
another 60 min. The final products were purified by
HPLC, and their radiochemical purities and specific
activities were also determined.
MeOD-d4, chemical shift δ in ppm relative to TMS): δ
7.91–7.87 (td, J = 7.4, 1.3 Hz, 3H), 7.69–7.73 (m, 6H),
7.67–7.61 (m, 7H), 7.22 (dd, J = 3.9, 1.8 Hz, 1H), 7.03
(d, J = 7.7 Hz, 1H), 6.99 (t, J = 7.8 Hz, 1H), 4.85
(d, J = 15.1 Hz, 2H). 31P NMR (400 MHz, MeOD-d4):
δ 23.7438. HRMS (+TOF MS): m/z calculated for
C25H21IP+ [M]+ 479.0420, found 479.0417.
2.2.6 | (3-Iodobenzyl)-tris(2-methoxyphenyl)
phosphonium (14)
14 (white solid, 0.704 g, 95.2%) was prepared as the same
way described above. 1H NMR (600 MHz, MeOD-d4,
chemical shift δ in ppm relative to TMS): δ 7.79–7.75
(m, 3H), 7.49–7.45 (m, 2H), 7.44–7.40 (m, 3H), 7.23–7.17
(m, 7H), 6.89 (t, J = 8.1 Hz, 1H), 4.67 (d, J = 16.1 Hz,
2H), 3.64 (s, 9H). 31P NMR (400 MHz, MeOD-d4):
δ 25.4596. HRMS (+TOF MS): m/z calculated for
C28H27IO3P+ [M]+ 569.0737, found 569.0735.
2.4 | Stability study
In stability experiment, the purified 125I-TPPs+, [125I]
9–[125I] 12, were added to physiological saline (0.9%) and
incubated at room temperature for 6 h, respectively.
Samples of the resulting solutions were analyzed by
radio-HPLC.
2.2.7 | (3-Iodobenzyl)-tris(3-methoxyphenyl)
phosphonium (15)
In vitro serum stability experiment, 50 μl of the
125I-TPPs+ ([125I]-9, [125I]-10, [125I]-11, or [125I]-12) solu-
tion was added to 0.1 ml of mouse serum and incubated
at 37ꢀC for 2 h, respectively. Then 500 μl of acetonitrile
was added to precipitate the serum proteins. After centri-
fugation for 5 min, the supernatants were filtered through
a 0.2-μm Millipore filter and analyzed by radio-HPLC.
15 (white solid, 0.694 g, 93.8%) was prepared as the same
way described above 1H NMR (600 MHz, MeOD-d4,
chemical shift δ in ppm relative to TMS): δ 7.68
(d, J = 7.8 Hz, 1H), 7.65–7.62 (m, 3H), 7.44 (d,
J = 9.4 Hz, 3H), 7.30 (s, 1H), 7.19–7.16 (m, 3H), 7.10 (d,
J = 13.9 Hz, 4H), 7.02 (t, J = 7.8 Hz, 1H), 4.88 (s, 2H),
3.80 (s, 9H). 31P NMR (400 MHz, MeOD-d4): δ 24.2542.
HRMS (+TOF MS): m/z calculated for C28H27IO3P+
[M]+ 569.0737, found 569.0739.
2.5 | Determination of the octanol–water
partition coefficient (log P)
2.2.8 | (3-Iodobenzyl)-tris(4-methoxyphenyl)
phosphonium (16)
The log P values of 125I-labeled TPPs+ were determined
by measuring the distribution of the radiotracer between
octanol (organic phase) and phosphate-buffered saline
(PBS, 0.05 mol/L, pH = 7.4, aqueous phase). Firstly,
mixing 20 μl of the purified radiotracers solution with
480 μl of PBS and 500 μl of n-octanol in a 2.0-ml EP
microtube. After vigorously vortexing for 3 min, the tube
16 (white solid, 0.712 g, 96.2%) was prepared as the same
way described above 1H NMR (600 MHz, MeOD-d4,
chemical shift δ in ppm relative to TMS): δ 7.65 (d,
J = 7.7 Hz, 1H), 7.50–7.47 (m, 6H), 7.22–7.10 (m, 6H),

SHI ET AL.
5
was centrifugated for 5 min at 10,000 rpm. The radioac-
tive counts in 100-μl aliquots of both phases were mea-
sured by a gamma counter. The octanol-water partition
coefficient (log P) was calculated using the following
equation: P = (activity in octanol-background activity)/
(activity in aqueous layer-background activity). The log
P value was reported as mean standard deviation
(SD) of triplicate samples.
Software™, La Jolla, CA, USA). Difference at the 95%
confidence level (p < 0.05) was considered to be statisti-
cally significant.
3 | RESULTS AND DISCUSSION
3.1 | Chemistry and radiochemistry
The synthetic routes of the meta-125I-labeled-TPPs+ and
the corresponding nonradioactive reference com-pounds
were shown in Scheme 1. The four radiolabeling precur-
sors, B-5, B-6, B-7, and B-8, were obtained by coupling
the (3-(bromomethyl)phenyl)boronic acid with 1 or the
ortho/meta/para-methoxy groups modified TPPs, 2, 3,
and 4, respectively. The aimed compounds were charac-
terized by 1H NMR, 13C NMR, 31P NMR, and HRMS. The
purities of all the above compounds were greater than
95%, as determined by HPLC (shown in Table 1).
2.6 | In vitro cellular uptake assay
The cellular uptake studies of 125I-labeled TPPs+ were
performed in the rat embryonic cardiomyoblast cell line
(H9c2) and mouse normal fibroblast cell line (NIH/3T3).
Cells were seeded in 24-well plates at a density of 2 Â 105
cells/well and cultured for 24 h at 37ꢀC under saturated
humidity and 5% CO2. To further evaluate the mitochon-
drial membrane potential, the H9c2 cells were allowed to
pretreat with carbonyl cyanide m-chlorophenylhydrazone
(CCCP, 10 μM in 50-mM low-K+ HEPES buffer) for
30 min prior to the experiment. The adherent cells were
washed twice with 1-ml culture medium, followed by
addition of 125I-labeled TPPs+ (11.7 KBq in 600 μl of
DMED medium per well). After incubation at 37ꢀC for
30 min, the medium was removed, and the cells were
washed with ice-cold PBS (0.01 mol/L, pH = 7.4, con-
taining 0.2% BSA) and lysed with 0.5-ml NaOH (1 mol/L),
respectively. The lysate was analyzed by γ-counter. All
assays were studied in triplicate and repeated three times.
The nonradioactive compounds (13–16), commonly
used as structure-referenced congeners of meta-125I-
labeled-TPPs+,
were
obtained
by
coupling
1-(bromomethyl)-3-iodobenzene with 1, 2, 3, and 4,
respectively. The targeted compounds were character-
ized by NMR (1H, 31P) and HRMS.
The labeling efficiency and radiochemical purity of
meta-125I-labeled-TPPs+ were determined by Radio-
HPLC. The HPLC retention time of 125I- was found to be
3.24 min, while those of [125I] 9–[125I] 12 were 16.11,
16.20, 16.29, and 16.08 min, respectively, which agreed
with the retention time of non-radioactive compounds
(As shown in Table 1 and Figure 1).
2.7 | Biodistribution studies
The meta-125I-labeled-TPPs+ were synthesized based
on a simple one-step copper-mediated radioiodination
method (Scheme 1). In order to optimize the reaction
conditions, the effects of the solvents, Cu2O/1,
10-phenanthroline concentrations and reaction tempera-
ture on the radiochemical yield of [125I]-9 were investi-
gated. The best solvent condition of Cu2O/1,
10-phenanthroline mediated radioiodination reported in
All biodistribution studies were performed under a
protocol approved by the Beijing Administration Office of
Laboratory Animal (BAOLA). The ex vivo biodistribution
studies were performed on Kunming mice (female,
4–5 weeks of age, 18–2 g, n = 4). The HPLC-purified
125I-TPPs+ ([125I]-9, [125I]-10, [125I]-11, or [125I]-12,
37 KBq diluted in 0.1 ml of saline) was administered into
the mice via the tail vein. At the selected time point,
(5, 30, 60, or 120 min), the mice were sacrificed. The
tissues and organs of interest were collected, weighed,
and counted by a γ-counter. The percentage of injected
dose per gram (%ID/g) for each sample was calculated
and expressed as mean SD.
literature
was
acetonitrile.21
However,
lower
radiochemical yield of [125I]-9 (<50%) was observed when
acetonitrile was used alone. The labeling efficiency could
be significantly improved by adding an aliquot of water.
The yield of the reaction decreased (<70%) when the reac-
tion temperature raised to 100ꢀC. In addition, a very infe-
rior radiochemical yield (<10%) was also found when the
reaction vial was sealed, which indicated that oxygen
might be required in the labeling process. It has been
reported that an oxidative process was involved in the for-
mation of iodobenzene intermediate during the copper-
catalyst Chan–Evans–Lam cross-coupling reaction.22 In
this copper-catalyst radioiodination reaction, oxygen in air
2.8 | Statistical methods
Statistical analysis was performed using the unpaired
t-test with commercial software (GraphPad Prism 8.0

6
SHI ET AL.
TABLE 1 The chemical or
radiochemical yields, the HPLC
retention times and log P values of
compounds
Entry
Compound
R1
R2
Yielda
74.8%
78.7%
82.6%
88.9%
>95%
>95%
>95%
>95%
96.6%
95.2%
93.8%
96.2%
HPLC Rt/min
14.55
Log P
1
B-5
B-6
B-7
B-8
H
B (OH)2
B (OH)2
B (OH)2
B (OH)2
125I
125I
125I
125I
127I
127I
127I
-
2
o-OMe
m-OMe
p-OMe
H
14.52
-
3
14.74
-
4
14.38
-
5
[
[
[
[
125I]-9
16.11
1.26 0.03
6
125I]-10
125I]-11
125I]-12
o-OMe
m-OMe
p-OMe
H
16.20
1.47 0.04
7
16.29
1.59 0.06
8
16.08
1.31 0.03
9
13
14
15
16
15.59
-
-
-
-
10
11
12
o-OMe
m-OMe
p-OMe
15.68
15.73
127I
15.41
aYields shown are chemical or radiochemical yield.
may play an important role as the oxidant. Thus, the opti-
mal labeling conditions of [125I]-9 were as follows: 1 μmol
of phenylboronic precursor, Cu2O/1,10-phenanthroline
(0.4 μmol/0.8 μmol) in 0.2 ml of acetonitrile/water
(9:1, v/v) mixture solvent, 5 μl of [125I]NaI, incubating in
air atmosphere at room temperature for 60 min.
radiochemical purity of 125I-TPPs+ remained in the range
of 90–95%. Only the ortho-methoxy group modified com-
pound, [125I]-10, displayed slight deiodination in mouse
serum at 37ꢀC for 2 h (Figure 1).
Under these optimized conditions, radioiodinations of
the OMe-modified triphenylphosphine-phenylboronic
3.2 | In vitro cellular uptake studies
acids,
B
6–B 8, were also investigated, and the
In vitro cellular uptake experiments of the 125I-TPPs+,
corresponding aim products, [125I] 10–[125I] 12 were
[
125I] 9-[125I] 12 were performed on H9c2 cells, a rat
obtained in high radiochemical yields (>95%). The result
embryonic cardiomyoblast cell line with overexpressed
mitochondria, and NIH/3T3 cells, a mouse normal fibro-
blast cell line as a negative control. CCCP is reported to
be a chemical inhibitor of oxidative phosphorylation that
can depolarize the mitochondria membrane potential. To
further confirm whether the radiotracer [125I] 9–[125I] 12
localized in the cells through mitochondrial membrane
potential, the cellular uptakes on CCCP pretreated H9c2
cells were also studied.
suggested
that
the
OMe-substituents
in
the
triphenylphosphine-phenylboronic acids did not obviously
affect the efficiency of the reactions. The 125I-TPPs+ could
sufficiently separate from the excess labeling precursors
with further HPLC purification and obtained the final
products with high specific activity (>46.6 GBq/μmol). The
labeling results suggested that the present procedure is a
simple, condition mildly, and efficient copper-catalyzed
method for the synthesis of 125I-labeled TPPs+ derivatives.
The log P of [125I] 9–[125I] 12 in octanol and PBS
(0.05 mol/L, pH = 7.4) were 1.26 0.03, 1.47 0.04,
1.59 0.06, and 1.31 0.03, respectively. All of these
125I-TPPs+ remain stable in saline at room temperature
for 6 h, and only intact 125I-TPPs+ were detected by
RP-HPLC. After 2 h of incubation in mouse serum, the
As shown in Figure 2, in H9c2 cell line, [125I] 9–[125I]
12 displayed a relatively high cellular uptake of 7.1%
0.4%, 7.7% 1.2%, 7.6% 0.4%, and 9.1% 0.6% of the
total added radioactivity, respectively. However,
the radioactivity accumulation of [125I] 9–[125I] 12 in
NIH/3T3 was only 56%, 51%, 50%, and 41% of those in
H9c2 cell line, respectively. After the CCCP pretreatment,

SHI ET AL.
7
FIGURE 1 The radio-HPLC chromatograms of the 125I-labeled TPPs+ before and after incubating at room temperature in saline for 6 h
and at 37ꢀC in the mice serum for 2 h. The retention times of [125I] 9–[125I] 12 were 16.11, 16.20, 16.29, and 16.08 min, respectively
FIGURE 2 Cellular uptakes
of the 125I-labeled TPPs+ in
H9c2 (with or without CCCP)
and NIH/3T3 cells at 37ꢀC for
30 min. *p < 0.05 shows
significant differences between
groups
the cellular uptakes of [125I] 9–[125I] 12 in H9c2 cell line
decreased to 58%, 55%, 62%, and 45%, respectively
(p < 0.05). It demonstrated that these 125I-TPPs+ accu-
mulated in the cardiomyoblast cells through the specific
mitochondrial membrane potential pathway.
pharmacokinetic properties. The results of the bio-
distribution studies are shown in Tables 2–5.
125I]-9 displayed a moderate initial accumulation in
[
heart (6.31 0.59%ID/g at 5 min p.i.) and good retention
(5.44 0.25% ID/g at 120 min p.i.). The uptake in lung
and blood at 5 min p.i. was 4.13 0.56%ID/g and
1.34 0.16%ID/g, respectively. The heart/lung and
heart/blood ratios were 2.78 and 28.63 at 120 min p.i.,
respectively. The radioactivity accumulations of [125I]-9
in the liver, intestine, and kidneys were remarkably high,
which indicated the radioactivity of [125I]-9 might be
3.3 | Ex vivo biodistribution study
The ex vivo biodistribution studies of 125I-TPPs+ were
performed on Kunming mice to investigate the

8
SHI ET AL.
TABLE 2 Biodistribution data of
125I]-9 in female Kunming mice
(n = 4, mean SD, %ID/g)
Uptake
[
5 min
30 min
60 min
120 min
Organ
Heart
6.31 0.59
34.85 3.26
4.13 0.56
80.07 1.00
5.11 1.69
2.44 0.65
1.79 0.33
1.76 0.16
4.29 0.67
1.34 0.16
0.13 0.04
5.81 0.44
31.79 2.71
2.85 0.38
76.82 1.85
3.42 0.42
3.21 0.13
2.17 0.08
1.61 0.21
4.96 0.30
1.07 0.15
0.43 0.04
5.54 0.08
20.53 2.26
2.73 0.44
77.13 2.37
3.85 0.61
5.31 1.11
2.15 0.44
1.77 0.17
17.86 0.80
0.36 0.02
0.62 0.12
5.44 0.25
12.30 1.48
1.96 0.16
64.23 3.10
3.28 0.27
4.99 0.30
1.49 0.15
1.72 0.12
19.92 1.47
0.19 0.00
0.47 0.04
Liver
Lung
Kidneys
Spleen
Stomach
Bone
Muscle
Intestine
Blood
Thyroid (%ID)
TABLE 3 Biodistribution data of
125I]-10 in female Kunming mice
Uptake
[
5 min
30 min
60 min
120 min
(n = 4, mean SD, %ID/g)
Organ
Heart
7.76 0.50
47.18 3.50
11.26 3.44
77.90 2.38
19.42 3.04
2.96 0.20
4.57 0.55
2.03 0.49
6.40 0.40
3.68 0.53
0.16 0.07
5.70 0.28
11.17 0.35
5.50 1.22
80.56 1.12
8.69 0.22
6.35 0.62
3.68 0.30
2.41 0.44
28.32 2.41
1.07 0.15
0.19 0.09
5.69 0.33
6.64 0.26
4.75 0.25
58.46 2.17
5.64 0.98
5.72 1.16
3.49 0.22
2.24 0.17
16.68 0.97
0.65 0.06
1.01 0.02
4.49 0.19
4.31 0.24
1.59 0.04
35.64 1.18
3.66 0.40
5.00 0.68
2.67 0.18
1.02 0.31
5.88 0.93
0.34 0.05
0.92 0.01
Liver
Lung
Kidneys
Spleen
Stomach
Bone
Muscle
Intestine
Blood
Thyroid (%ID)
TABLE 4 Biodistribution data of
Uptake
[
125I]-11 in female Kunming mice
5 min
30 min
60 min
120 min
(n = 4, mean SD, %ID/g)
Organ
Heart
6.01 0.51
86.47 2.15
8.51 0.22
81.44 1.88
12.78 0.92
4.82 0.36
3.66 0.32
1.21 0.16
10.89 1.29
3.19 0.24
0.17 0.06
5.14 0.29
31.49 1.95
5.30 0.48
56.51 2.14
11.00 0.53
10.08 0.52
3.17 0.52
1.49 0.13
79.19 0.80
1.54 0.24
0.46 0.10
4.98 0.36
22.31 1.64
4.17 0.57
46.20 2.49
8.46 0.56
10.47 1.20
3.23 0.23
1.46 0.23
29.60 2.14
0.92 0.09
0.51 0.13
4.50 0.36
14.75 1.67
2.35 0.14
35.02 1.54
6.22 0.44
9.69 0.92
2.59 0.56
1.08 0.13
17.44 1.04
0.51 0.02
0.66 0.38
Liver
Lung
Kidneys
Spleen
Stomach
Bone
Muscle
Intestine
Blood
Thyroid (%ID)

SHI ET AL.
9
TABLE 5 Biodistribution data of
Uptake
[
125I]-12 in female Kunming mice
5 min
30 min
60 min
120 min
(n = 4, mean SD, %ID/g)
Organ
Heart
2.70 0.15
24.55 1.78
5.16 0.62
32.12 1.00
7.94 0.65
2.00 0.91
1.71 0.35
0.65 0.13
8.50 1.78
1.45 0.15
0.32 0.47
2.43 0.30
9.86 1.16
2.76 0.24
30.31 1.35
6.02 1.29
2.61 0.50
1.47 0.09
0.62 0.20
11.72 2.01
0.49 0.21
0.21 0.10
2.38 0.18
5.84 0.82
1.37 0.43
19.47 0.78
3.15 0.89
2.89 1.54
1.39 0.20
0.64 0.12
8.89 1.57
0.12 0.10
0.20 0.08
2.36 0.25
5.79 0.30
1.19 0.32
19.31 0.26
2.57 0.14
3.94 0.38
1.35 0.37
0.74 0.03
8.18 1.45
0.08 0.04
0.20 0.18
Liver
Lung
Kidneys
Spleen
Stomach
Bone
Muscle
Intestine
Blood
Thyroid (%ID)
FIGURE 3 Comparison of organ uptake (heart, liver, lung, kidneys, and blood) between [125I] 9 and [125I] 12 in Kunming mice (%ID/g,
60 min post-injection, n = 4)
excreted from the body via both the hepatobiliary and
renal systems. Actually, for the mitochondrial membrane
potential based imaging agent, the high liver uptake is a
drawback, which will interfere with the imaging quality
of myocardium (low signal-to-noise ratio), as well as trac-
ing tumors in chest and abdomen.
In many cases, the methoxy group introduced in
TPPs+ radiotracers could act as a positive effect on their
pharmacokinetic properties by accelerating the radioac-
tivity clearance from liver and other non-targeting
organs.22,23 In this study, three methoxy group intro-
duced 125I-TPPs+, [125I]-10, [125I]-11, and [125I]-12 were
designed and synthesized based on the architecture of
[
125I]-9. The methoxy groups were introduced into the
phenyl groups of triphenylphosphine moieties on para,
ortho, and meta positions, respectively. As expected, the
liver clearance of three methoxy group modified 125I-
TPPs+ was significantly quicker than that of [125I]-9. At
60 min post-injection, the liver uptakes of [125I]-10,
[
125I]-11, and [125I]-12 were rapidly decreased to 14.1%,
25.8%, and 23.8% from the maximum values, respectively.
And the liver accumulation of [125I]-9 still remained
58.9% of the maximum uptakes at 60 min post-injection.
Moreover, focusing on the uptake of 60 min post-
injection (Figure 3), it was found that not only the intro-
duction of methoxy groups, but also the position of

10
SHI ET AL.
methoxy groups of 125I-TPPs+ functioned a significant
effect on their pharmacokinetic properties. The
meta-OMe modified [125I]-11 displayed a higher liver
uptake (22.31 1.64%ID/g), while [125I]-10 and [125I]-12
contained methoxy groups on the ortho and para
positions, respectively, displayed a decreased liver uptake
(6.64 0.26 and 5.84 0.82%ID/g, respectively) than
those of [125I]-11 and [125I]-9 (20.53 2.26%ID/g). The
para-OMe modified 125I-TPPs+, [125I]-12 also displayed a
significant lower radioactivity accumulation in heart,
lungs and blood than those of other 125I-TPPs+.
was successfully developed. The method uses
Cu2O/1,10-phenanthroline as the catalyst system, oxygen
in the air as the oxidant, triphenylphosphine
phenylborate compounds as the labeling precursors, to
obtain the radioiodine-labeled TPPs+ by a one-step reac-
tion under mild hydrous condition. Four meta-125I-
labeled-TPPs+, [125I] 9–[125I] 12 were synthesized in high
radiochemical yield (>95%). Results of biological evalua-
tions indicated the four meta-125I-labeled-TPPs+ could
accumulate in the mitochondrial-rich myocardial cells
through the mitochondrial membrane potential. Although
further structural modification is still needed to improve
their pharmacokinetic properties, the highly efficient
radioiodination method offers a new strategy to design
and prepare novel 125I-labeled-TPPs cations, which lays a
prospective foundation for further study and application.
Compared with the [125I]-9, the lipophilicity of [125I]
10–[125I] 12 was increased by the introduction of met-
hoxy groups and was slightly different due to the position
of methoxy groups, on the order of [125I]-12 < [125I]-
10 < [125I]-11. As shown in Tables 3–5, the liver uptakes
in 5, 30, 60, and 120 min post-injection of three
OMe-modified 125I-TPPs+ were also increased on the
order of [125I]-12 < [125I]-10 < [125I]-11. It indicated that
the difference of lipophilicity might be the vital factor
that influence the pharmacokinetic properties of methoxy
groups modified 125I-labeled TPPs cations.
ACKNOWLEDGEMENT
This work was financially supported by the National Nat-
ural Science Foundation of China (21976019).
ORCID
Among three OMe-modified 125I-labeled TPPs cations,
the ortho-OMe modified 125I-TPPs+, [125I]-10 had the
fastest liver clearance, and comparable accumulation in
heart, resulted in the highest heart/liver ratio (1.04) at
120 min post-injection. However, [125I]-10 displayed the
highest thyroid uptakes at 60 and 120 min post-injection
(1.01 0.02 and 0.92 0.01%ID, respectively), and slight
deiodination was also found in the result of in vitro
serum stability study. Therefore, further structural opti-
mization is needed to obtain the 125I-TPPs+-based mito-
chondrial membrane potential targeting probes with
improved physicochemical and pharmacokinetic proper-
ties. Besides cardiomyocytes, mitochondria are also
enriched in many carcinoma cells with mitochondrial
dysfunction.1 Compared with healthy cells, cancer cells
have an increase of approximately 60 mV in negative
mitochondrial membrane potential (ΔΨ), which cause a
higher accumulation of radiolabeled TPPs+ in cancer
cells.24,25 For example, 18F-BnTP has been reported as a
potential non-invasive probe to functionally profile mito-
chondrial membrane potential and mitochondrial hetero-
geneity within non-small cell lung cancer (NSCLC).26
Therefore, potential application of these 125I-labeled
TPPs+ in noninvasive monitoring of mitochondrial mem-
brane potential in cancers can also be considered.
Shuyu Shi https://orcid.org/0000-0001-7510-5120
Zelan Liu https://orcid.org/0000-0002-5522-9115
Jie Lu https://orcid.org/0000-0001-6351-0508
REFERENCES
1. Ma C, Xia F, Kelley SO. Mitochondrial targeting of probes and
therapeutics to the powerhouse of the cell. Bioconjug Chem.
2020;31(12):2650-2667.
2. Lin MT, Beal MF. Mitochondrial dysfunction and oxidative
stress in neurodegenerative diseases. Nature. 2006;443(7113):
787-795.
3. Ballinger SW. Mitochondrial dysfunction in cardiovascular dis-
ease. Free Radic Biol Med. 2005;38(10):1278-1295.
4. Rosca MG, Hoppel CL. Mitochondrial dysfunction in heart fail-
ure. Heart Fail Rev. 2013;18(5):607-622.
5. Exner N, Lutz AK, Haass C, Winklhofer FK. Mitochondrial
dysfunction in Parkinson's disease: molecular mechanisms and
pathophysiological consequences. EMBO J. 2012;31(14):
3038-3062.
6. Murphy MP. Targeting lipophilic cations to mitochondria.
Biochim Biophys Acta Biomembr. 2008;1777(7):1028-1031.
7. Azzone GF, Pietrobon D, Zoratti M. Determination of the
proton electrochemical gradient across biological membranes.
In: Lee CP, ed. Current Topics in Bioenergetics. Vol 13. Elsevier;
1984:1-77.
8. Ross MF, Kelso GF, Blaikie FH, et al. Lipophilic
triphenylphosphonium cations as tools in mitochondrial bioen-
ergetics and free radical biology. Biochim Biophys Acta. 2005;
70(2):222-230.
9. Ross MF, Da RT, Blaikie FH, et al. Accumulation of lipophilic
dications by mitochondria and cells. Biochem J. 2006;400(1):
199-208.
4 | CONCLUSION
In this study, a simple and efficient radioiodination
method for the preparation of radioiodine-labeled TPPs+
10. Kelso GF, Porteous CM, Coulter CV, et al. Selective targeting
of a redox-active ubiquinone to mitochondria within cells:

SHI ET AL.
11
antioxidant and antiapoptotic properties.
2001;276(7):4588-4596.
11. Smith RAJ, Porteous CM, Coulter CV, Murphy MP. Selective
targeting of an antioxidant to mitochondria. Eur J Biochem.
1999;263(3):709-716.
J
Biol Chem.
20. Yang H, Li Y, Jiang M, Wang J, Fu H. General copper-
catalyzed transformations of functional groups from
arylboronic acids in water. Chem A Eur J. 2011;17(20):
5652-5660.
21. Zhang P, Zhuang R, Guo Z, Su X, Chen X, Zhang X. A highly
efficient copper-mediated radioiodination approach using aryl
boronic acids. Chem A Eur J. 2016;22(47):16783-16786.
22. Kim Y-S, Yang C-T, Wang J, et al. Effects of targeting moiety,
linker, bifunctional chelator, and molecular charge on biologi-
cal properties of 64Cu-labeled triphenylphosphonium cations.
J Med Chem. 2008;51(10):2971-2984.
12. Filipovska A, Kelso GF, Brown SE, Beer SM, Smith AR,
Murphy MP. Synthesis and characterization of
triphenylphosphonium-conjugated peroxidase mimetic:
insights into the interaction of ebselen with mitochondria.
J Biol Chem. 2005;280(25):24113-24126.
a
13. Madar I, Ravert H, DiPaula A, Du Y, Dannals RF, Becker L.
Assessment of severity of coronary artery stenosis in a canine
model using the PET agent 18F-fluorobenzyl triphenyl phos-
phonium: comparison with 99mTc-tetrofosmin. J Nucl Med.
2007;48(6):1021-1030.
23. Chen S, Zhao Z, Zhang Y, Fang W, Lu J, Zhang X. Effect of
methoxy group position on biological properties of 18F–labeled
benzyl triphenylphosphonium cations. Nucl Med Biol. 2017;49:
16-23.
14. Shoup TM, Elmaleh DR, Brownell AL, Zhu A, Guerrero JL,
24. Moura C, Mendes F, Gano L, Santos I, Paulo A. Mono- and
dicationic Re(I)/99mTc(I) tricarbonyl complexes for the
targeting of energized mitochondria. J Inorg Biochem. 2013;
123(8):34-45.
Fischman
AJ.
Evaluation
of
(4-[18F]fluorophenyl)
triphenylphosphonium ion. A potential myocardial blood flow
agent for PET. Mol Imaging Biol. 2011;13(3):511-517.
15. Kim DY, Kim HJ, Yu KH, Min JJ. Synthesis of [18F]-labeled
25. Zhou Y, Liu S. 64Cu-labeled phosphonium cations as PET
radiotracers for tumor imaging. Bioconjug Chem. 2011;22(8):
1459-1472.
(2-[2-fluoroethoxy]ethyl)triphenylphosphonium cation as
a
potential agent for myocardial imaging using positron emission
tomography. Bioorg Med Chem Lett. 2012;22(1):319-322.
16. Guludec D, Lautamäki R, Knuuti J, Bax JJ, Bengel FM. Present
and future of clinical cardiovascular PET imaging in Europe—
a position statement by the European Council of Nuclear
Cardiology (ECNC). Eur J Nucl Med Mol. 2008;35(9):1709-1724.
17. Li J, Lu J, Zhou Y. Mitochondrial-targeted molecular imaging
in cardiac disease. Biomed Res Int. 2017;2017(2):1-11.
18. Srivastava PC, Knapp FF. [(E)-1-[123I]Iodo-1-penten-5-yl]
triphenylphosphonium iodide: convenient preparation of a
potentially useful myocardial perfusion agent. J Med Chem.
1984;27(8):978-981.
26. Momcilovic M, Jones A, Bailey ST, et al. In vivo imaging of
mitochondrial membrane potential in non-small-cell lung
cancer. Nature. 2019;575(7782):380-384.
How to cite this article: Shi S, Liu Z, Wu Z,
Zhou H, Lu J. Preparation and biological
evaluation of radioiodine-labeled
triphenylphosphine derivatives as mitochondrial
targeting probes. J Label Compd Radiopharm.
2021;1–11. https://doi.org/10.1002/jlcr.3910
19. Sakai T, Saito Y, Takashima M, Magata Y. Development of
radioiodinated lipophilic cationic compounds for myocardial
imaging. Nucl Med Biol. 2015;42(5):482-487.