Dithiol Aryl Arsenic Compounds as Potential Diagnostic and Therapeutic Radiopharmaceuticals

Article abstract of DOI:10.1021/acs.inorgchem.6b01175

Arsenic-72 (⁷²As) and ⁷⁷As have nuclear properties useful for positron emission tomography (PET) and radiotherapy, respectively. The thiophilic nature of arsenic led to the evaluation of dithioarylarsines for potential use in radiopharmaceuticals. Several dithioarylarsines were synthesized from their arylarsonic acids and dithiols and were fully characterized by NMR, ESI-MS, and X-ray crystallography. This chemistry was translated to the no-carrier-added (nca) ⁷⁷As level. Because arsenic was available at the nca nanomolar level only as [⁷⁷As]arsenate, this required addition of an aryl group directly to the As to form the [⁷⁷As]arylarsonic acid. The [⁷⁷As]arsenate was reduced from ⁷⁷As (V) to ⁷⁷As (III), and a modified Bart reaction was used to incorporate the aryl ring onto the ⁷⁷As, which was followed by dithiol addition. Various modifications and optimizations resulted in 95% radiochemical yield of nca [⁷⁷As]p-ethoxyphenyl-1,2-ethanedithiolatoarsine.

Full text of DOI:10.1021/acs.inorgchem.6b01175

Article
pubs.acs.org/IC
Dithiol Aryl Arsenic Compounds as Potential Diagnostic and
Therapeutic Radiopharmaceuticals
Anthony J. DeGraffenreid,†,§,∥ Yutian Feng,†,§ Donald E. Wycoff,† Ryan Morrow,† Michael D. Phipps,†
Cathy S. Cutler,‡,∥ Alan R. Ketring,‡ Charles L. Barnes,† and Silvia S. Jurisson*,†,‡
†Department of Chemistry and ‡Research Reactor Center (MURR), University of Missouri, Columbia, Missouri 65211, United States
S
* Supporting Information
ABSTRACT: Arsenic-72 (72As) and 77As have nuclear properties
useful for positron emission tomography (PET) and radiotherapy,
respectively. The thiophilic nature of arsenic led to the evaluation
of dithioarylarsines for potential use in radiopharmaceuticals. Several
dithioarylarsines were synthesized from their arylarsonic acids and
dithiols and were fully characterized by NMR, ESI-MS, and X-ray
crystallography. This chemistry was translated to the no-carrier-added
(nca) 77As level. Because arsenic was available at the nca nanomolar level only as [77As]arsenate, this required addition of an aryl
group directly to the As to form the [77As]arylarsonic acid. The [77As]arsenate was reduced from 77As (V) to 77As (III), and a
modified Bart reaction was used to incorporate the aryl ring onto the 77As, which was followed by dithiol addition. Various
modifications and optimizations resulted in 95% radiochemical yield of nca [77As]p-ethoxyphenyl-1,2-ethanedithiolatoarsine.
Arsenic-72 (72As) is a β+-emitter with a 26-h half-life and is of
INTRODUCTION
■
particular interest because it is the daughter product of 72Se
The well-established toxicity of arsenic and its compounds
has led to their use in insecticides, herbicides, and fungicides,
as well as significant concern regarding environmental con-
(t1/2 8.5 d). The development of a 72Se/72As generator would
make it readily available.15−20 The β−-emitting 77As with its
38-h half-life would be the therapeutic analogue for 72As. The
tamination of water supplies and food sources.1,2 Despite this
half-lives of 72As and 77As make them candidates for targeting
toxicity, arsenic and its derivatives have found uses as poultry
receptors involving antibody−antigen interactions, as these can
feed additives and in medicines. Arsenic trioxide has been used
as a medicinal compound for over a century, and was approved
by the U.S. Food and Drug Administration (FDA) in 2000 for
require days to reach maximum accumulation at the target
site and clear from the blood.21 Arsenic-72, 89Zr (t1/2 78.41 h),
and 124I (t1/2 4.17 d) are among the few positron emitters with
the treatment of acute promyelocytic leukemia under the trade
sufficiently long half-lives for PET imaging using antibodies as
name Trisenox.3,4
targeting vectors.
The medical use of radioisotopes of arsenic dates back to the
Arsenic-72 and 77As are considered “matched pair” radio-
1950s and 1960s focusing on brain trauma and tumors, uti-
nuclides since their chemistry is identical, one useful for PET
lizing 72As, 74As, or 76As.5,6 Targeting tumors with radioarsenic
imaging and the other for radiotherapy. True “matched pairs”
labeled proteins and antibodies was first reported in 1963
such as 72As/77As are rare, with 64Cu/67Cu and 123/124/131I being
using carrier-added 74As labeled arsanilic acid coupled to
other examples.22 Most so-called “matched pairs” involve
mouse albumin.6 More recently, no-carrier-added (nca) radio-
radionuclides of different elements having similar chemistries
arsenic (74As and 77As) labeled sulfhydryl modified antibodies
such as 111In/90Y, which do not necessarily behave identically
(Rituximab and bavituximab) were evaluated in tumor-bearing
mice.7,8 N-(2-Hydroxypropyl)-methacrylamide (HMPA) poly-
in vivo.
Arsenic-72 can be produced directly by proton irradiation of
mers containing sulfhydryl groups were labeled with 72/74As
72Ge.23 It is also the daughter product of 72Se, which can be
with reasonable yields and stability in water, making these poten-
3
produced by He or alpha irradiation of 70Ge, proton irradia-
tion of natBr or 75As, or deuteron irradiation of 75As.16,24 The
β−-emitting 77As is reactor-produced by neutron irradiation
of 76Ge.22,24 Separation of 72As and 77As from their parent and
target, respectively, yield high specific activity radioarsenic, also
known as nca, meaning that virtually all atoms of arsenic are of
the desired radioisotope. This is particularly important when
targeting limited numbers of receptors at an in vivo target site
tially useful in vivo for following polymer-based therapeutics.9
Radionuclides are routinely used in diagnostic and ther-
apeutic nuclear medicine, with 99mTc (6 h, 140 keV γ) and 18F
(110 min, β+) the most commonly used diagnostic radio-
nuclides for single photon emission computed tomography
(SPECT) and positron emission tomography (PET) imaging,
respectively.10,11 Therapeutic radionuclides in approved radio-
pharmaceuticals include the β−-emitters 131I, 153Sm, and 90Y,
and the recently approved α-emitter 223Ra.11−13
Arsenic has several radioisotopes that have nuclear properties
Received: May 18, 2016
suitable for radiopharmaceutical applications (Table 114).
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.6b01175
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 1. Nuclear Properties of Arsenic Radioisotopes14
70As
71As
72As
26.0 h
74As
76As
77As
half-life
52.6 m
ε (9)
β+ (91)
65.3 h
ε (72)
β+ (28)
17.77 d
ε (37)
β+ (29)
β− (34)
26.24 h
β− (100)
38.79 h
β− (100)
decay mode (%)
ε (12)
β+ (88)
Eβmax (MeV) (highest intensity particle (%))
Eγ in keV (%)
2.14 (36)
1039 (82)
0.816 (27.9)
175 (82)
2.50 (64.2)
834 (81)
0.945 (β+) (26.1); 1.35 (β−) (19)
2.96 (51)
559 (45)
0.683 (97)
239 (1.6)
596 (59); 635 (15)
Arsenic-77 (NaH277AsO4) was provided in a methanol solution (77As
stock solution, 1 mL, ∼ 4.8 mCi, 5.9 × 10−11 mol or 4 ng based on
measured activity, and which gives a specific activity of 3020 MBq/nmol
(81 mCi/nmol)), separated as previously reported.24 The methanol was
removed under a stream of air, and the 77As was taken up in 1 mL of
water. The pH of this reconstituted solution was 4−5.
(e.g., tumors). Examples of such separations have been
reported using distillation, chromatography, solvent extraction,
and precipitation methods.15−18,23−26 The development of
higher yielding production routes and separation methods for
arsenic radioisotopes has led to increased interest in 72As for
positron emission tomography (PET) imaging applications and
77As for potential radiotherapy. The development of chelation
systems for radioarsenic resulting in high in vivo stability are
needed for these radionuclides to be effective.
Physical Measurements. 1H and 13C NMR spectra were obtained
in D2O, CDCl3, (CD3)2SO, or CD3CN on either a Bruker ARX-300
or 500 MHz spectrometer. Electrospray ionization mass spectra
(ESI-MS) were collected on a Thermo Finnigan TSQ7000 triple-
quadrupole instrument with an API2 source. Elemental analyses were
performed by Atlantic Microlab, Inc. (Norcross, GA). An ORTEC
high-purity germanium (HPGe) detector coupled to a Canberra
amplifier, analog to digital converter, and HV power supply system
using PROcount 2000 software was used to assay 77As samples. A
Shimadzu Prominence HPLC system equipped with a pump,
controller, Prominence UV−vis detector (model SPD20-AV) set to
254 nm and coupled to a Beckman 170 NaI(Tl) radioisotope detector
with a Thermo Scientific Beta Basic column (C18, 5 μm, 150 mm ×
4.6 mm) was used to analyze 77As dithioarylarsines, with a linear
gradient from 100% A to 65% B (A = water with 0.1% TFA, and
B = ACN with 0.1% TFA) with a flow rate of 1 mL/min over 65 min.
An Eckert & Ziegler Bioscan AR-2000 Imager using LabLogic Win-Scan
imaging scanner software (Version 2.2(11)) was used for scanning
radioTLC plates.
Until very recently, radioarsenic chemistry focused on bind-
ing to proteins and polymers modified with monodentate
sulfhydryl groups, which have shown instability toward radio-
arsenic loss.7−9,27 The affinity of arsenic for sulfur is well estab-
lished and is believed to be the primary reason for its high
toxicity in biological systems through its binding to albumin,
glutathione, and other sulfhydryl-containing biological
molecules.28,29 Three dithiolates, British anti-Lewisite (BAL),
2,3-dimercaptosuccinic acid (DMSA), and 2,3-dimercaptopropane-
sulfonate (DMPS), are currently used to treat arsenic poisoning.30
The thiophilic nature of arsenic was simultaneously exploited
to develop two classes of potential arsenic radiotracers, one
with the recently reported tridentate trithiols31 and the other
with dithioarylarsines as target molecules. Incorporation of an
aryl-As bond is believed to add stability to the resultant
molecule, which is essential if the nca radioarsenic (nM to pM
range) is to be delivered to the target tissue. Several dithioary-
larsines of the form [arylAs(S-(CH2)n-S)] were synthesized and
fully characterized at the macroscopic level and this chemistry
was translated to nca [aryl77As(S-(CH2)n-S)] concentrations
starting with [77As]arsenate. Conjugation to antibodies or pep-
tides can be pursued through modification of either the aryl
ring or the dithiol.
Synthesis of Dithioarylarsines. Reduction of the As(V) aryl
arsonic acid to the corresponding As(III) aryl arsenic acid and sub-
sequent reaction with the dithiol was accomplished with modification
of a literature method (Scheme 1).36,37
Scheme 1. General Synthesis of Dithioarylarsines
(Compounds 1−3, S4−S12) and the Numbering System
Used for Their Identification
EXPERIMENTAL SECTION
■
Materials. Arsenic trioxide (CAUTION! Arsenic is highly toxic and
should be handled with care), mercaptoacetic acid, dimercaptosuccinic
acid (DMSA), 1,2-ethane dithiol, 1,3-propane dithiol, 4-amino-phenyl
arsonic acid (p-arsanilic acid), and phenyl arsonic acid were purchased
from Sigma-Aldrich, Acros, or Alfa Aesar, and used as received. All
solvents and acids were reagent grade and purchased from Fisher
Scientific or Sigma-Aldrich and used without further purification.
Cu0 nanoparticles (<20 nm in acetone at 100 mg/mL (surfactant and
reactant-free)) were purchased from Strem Chemicals, Inc. SORB-
TECH Silica Gel TLC (thin layer chromatography) plates were
purchased from Sorbtech (Norcross, GA) and used as received.
p-Aminophenyldichloroarsine, p-nitrophenylarsonic acid, p-ethoxyphe-
nyldiazonium tetrafluoroborate, and p-ethoxyphenylarsonic acid were
prepared using literature methods.32−35
2-(4-Ethoxyphenyl)-1,3,2-dithiaarsolane [CH3CH2OC6H4As-
(SCH2CH2S)], 1. p-Ethoxyphenylarsonic acid (200 mg, 0.81 mmol)
was dissolved in ethanol (100%, 20 mL) in a round-bottom flask
equipped with a stir bar, and heated to 55 °C in a water bath. Aque-
ous ammonium mercaptoacetate (5.5 M, 890 μL, 4.9 mmol) was
added, and heating continued with vigorous stirring. After 60 min, the
flask was removed from heat and 1,2-ethanedithiol (69 μL, 77 mg,
0.82 mmol) was added. The resultant reaction mixture was stirred for
30 min, at which time water (∼50 mL) was added to precipitate the
product. After cooling in the freezer (−15 °C) for 2 h, the solids were
collected by vacuum filtration, washed with cold ethanol, and dried
in vacuo to obtain the product as a light yellow powder. Yield: 47%,
110 mg. 1H NMR (CDCl3, 500 MHz) δ ppm: 1.41 (t, 3H, CH3), 3.19
(m, 2H, SCH2), 3.35 (m, 2H, SCH2), 4.02 (q, 2H, OCH2), 6.88
CAUTION! 77As (Table 1) and 77Ge (2.7 MeV β−, 211, 215, and
264 keV γ, 11.3-h half-life) are radioactive and must be handled in
laboratories outfitted and approved for work with radioactive materials.
Arsenic-77 was produced by neutron irradiation of 96.2% or
98.6% enriched 76GeO2 (Trace Sciences International, Richmond
Hill, ON) in the flux trap at a thermal flux of ∼2.4 × 1014 n/cm2-s
at the University of Missouri Research Reactor Center (MURR).
B
DOI: 10.1021/acs.inorgchem.6b01175
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(d, 2H, ArH), 7.54 (d, 2H, ArH). 13C NMR (CDCl3, 125.8 MHz) δ
ppm: 14.92 (CH3), 41.84 (SCH2), 63.57 (OCH2), 114.66 (ArC),
132.24 (ArC), 133.21 (ArC), 160.05 (ArC). Elem. Anal. Calcd
(found) for C10H13OS2As: C, 41.67 (41.84); H, 4.55 (4.45); S, 22.24
(22.40). ESI/APCI MS (m/z): 305.11 (304.96 calcd for
[C10H14O2S2As] [M + OH]+).
Rf = 0.44; [77As]p-ethoxyphenyl-1,2-ethanedithioarsine, Rf = 0.72.
[77As]p-ethoxyphenyl-1,2-ethanedithioarsine was analyzed by an
additional radioTLC method (silica gel, 50/50 CHCl3/CH2Cl2):
Rf = 0.93; no other species migrated from the origin. The radio-
chemical yield of [77As]p-ethoxyphenyl-1,2-ethanedithioarsine was
determined to be 95% by radioTLC. Reversed-phase HPLC comparison
of the nonradioactive standard and [77As]p-ethoxyphenyl-1,2-ethanedi-
thioarsine confirmed that the desired product had been synthesized
(both had retention times of 28.35 min) in 95% radiolabeling yield.
2-(4-Ethoxyphenyl)-1,3,2-dithiaarsolane-4,5-dicarboxylic acid,
Diammonium salt (NH4)2[CH3CH2O C6H4As(SCH(COO)CH(COO)S)],
2. p-Ethoxyphenylarsonic acid (200 mg, 0.81 mmol), 5.5 M ammonium
mercaptoacetate (890 μL, 4.9 mmol), and meso-2,3-dimercaptosuc-
cinic acid (149 mg, 0.82 mmol) were reacted in ethanol (100%,
10 mL) as described above for 1. The product began to precipitate
shortly after the addition of DMSA. After stirring for 30 min, the
reaction mixture was placed in the freezer at −15 °C to further
precipitate the product. Solids were collected by vacuum filtration,
washed with cold ethanol, and dried in vacuo to obtain the product as
a light yellow powder. Yield: 32.8%, 100 mg.1H NMR (D2O d2,
500 MHz) δ ppm: 1.41 (t, 3H, CH3), 4.18 (q, 2H, OCH2), 4.45 (s,
2H, SCH), 7.08 (d, 2H, ArH), 7.77 (d, 2H, ArH). 13C NMR (D2O d2,
125.8 MHz) δ ppm: 13.83 (CH3), 63.33 (CH2), 64.33 (SCH), 115.04
(ArC), 132.38.66 (ArC), 134.36 (ArC), 159.08 (ArC), 175.11
(COOH). Elem. Anal. Calcd (found) for C12H19AsN2O5S2: C, 35.13
(34.72); H, 4.67 (4.36); N, 6.83 (6.25); S, 15.63 (15.64). ESI/APCI
MS (m/z): 393.15 (393.16 calcd for [C12H14O6S2As] [M + OH]+).
2-(4-Ethoxyphenyl)-1,3,2-dithiaarsinane [CH3CH2OC6H4As-
(SCH2CH2CH2S)], 3. p-Ethoxyphenylarsonic acid (200 mg, 0.81 mmol),
5.5 M ammonium mercaptoacetate (890 μL, 4.9 mmol), and 1,3-
propanedithiol (82 μL, 88 mg, 0.82 mmol) were reacted in ethanol
(100%, 10 mL) as described above for 1. The reaction mixture
was treated with water (∼50 mL) to precipitate the product and
cooled in the freezer (−15 °C) for 2 h. Solids were collected by
vacuum filtration, washed with cold ethanol, and dried in vacuo to
obtain the product as a light yellow powder. Yield: 61.3%, 150 mg. 1H
NMR (CDCl3, 500 MHz) δ ppm: 1.44 (t, 3H, CH3), 2.14 (m, 1H,
CH2CH2CH2), 2.17 (m, 1H, CH2CH2CH2), 2.71 (m, 2H, SCH2),
2.87 (m, 2H, SCH2), 4.08 (q, 2H, OCH2), 7.01 (d, 2H, ArH), 7.78 (d,
2H, ArH). 13C NMR (CDCl3, 125.8 MHz) δ ppm: 14.96 (CH3),
26.32 (SCH2CH2), 28.67 (SCH2), 63.62 (OCH2), 115.44 (ArC),
128.66 (ArC), 134.01 (ArC), 160.12 (ArC). Elem. Anal. Calcd
(found) for C11H15AsOS2: C, 41.67 (41.84); H, 4.55 (4.45); S, 22.24
(22.40). ESI/APCI MS (m/z): 317.15 (317.98 calcd for
[C11H15O2S2As] [M + O]+).
1
Stability by H NMR Spectroscopy. An initial step in the sta-
bility assessment of the ethyl versus propyl backbones of the dithio-
arylarsines was investigated qualitatively for compounds S7, S8, and S9
1
by H and 13C NMR by dissolving 20−30 mg of the compound in
deuterated aqueous (S8) or 20/80 aqueous/acetonitrile solutions
(S7 and S9) based on solubility. NMR spectra were obtained multiple
times over 30 days.
X-ray Crystal Structures. Intensity data for compounds 1, 3, S4,
S6−S10, and S12 were obtained at −100 °C or −173 °C on a Bruker
SMART CCD Area Detector system using the ω scan technique with
Mo Kα radiation from a graphite monochromator. Intensities were
corrected for Lorentz and polarization effects. Equivalent reflections
were merged, and absorption corrections were made using the
multiscan method. The structures were solved by direct methods with
full-matrix least-squares refinement, using the SHELX package.38 All
non-hydrogen atoms were refined with anisotropic thermal parame-
ters. The hydrogen atoms were placed at calculated positions and
included in the refinement using a riding model, with fixed isotropic
U. Data were corrected for decay and absorption using the program
SADABS.39 The final difference maps contained no features of chem-
ical significance.
RESULTS AND DISCUSSION
■
Translation of macroscopic chemistry to tracer level concen-
trations is often not straightforward, especially when going to
nca levels. The concentrations of short half-life, nca radio-
nuclides are often nanomolar or less, and this concentration
difference can significantly affect the synthetic chemistry.
Three main methods for synthesizing aryl arsonic acids from
arsenite, AsCl3, or arsenate have been reported, namely the
Bart, Scheller, and Bechamp reactions.32 The separation of nca
77As from the parent 77Ge yields [77As]arsenate. The Bart reac-
tion was chosen as most suitable for translation to the nca
radiotracer level. The Bechamp reaction starts with arsenate but
requires high temperature and sometimes pressure, not ideal for
reactions involving radiotracers.32 The Scheller reaction uses
AsCl3 as its starting compound.32 Bart synthesized arylarsonic
acids by reaction of aryldiazonium ions formed in situ with
arsenite in the presence of a copper catalyst (Scheme 2).32
Synthesis of No-Carrier-Added 77As 2-(4-Ethoxyphenyl)-
1,3,2-dithiaarsolane [CH3CH2OC6H477As(SCH2CH2S)], [77As]1.
[77As]2-(4-ethoxyphenyl)-1,3,2-dithiaarsolane was synthesized at the
nca radiotracer level from [77As]arsenate, by optimizing the amounts
of reducing agent (mercaptoacetate, 1−550 mM, pH 4−5),
aryldiazonium salt (p-ethoxybenzenediazaonium tetrafluoroborate,
20−85 mM), dithiol (1,2-ethanedithiol, 9.8−100 mM), Cu catalyst
(polished Cu (pieces (5 mm × 5 mm × 2 mm) and Cu nanoparticles),
solvent (ethanol, acetonitrile), temperature, and time. Each parameter
was optimized independently and then the final formulation was
fine-tuned. The optimal synthesis is as follows: Acetonitrile (650 μL)
was added to a screw-cap vial containing 250 μL of the aqueous
nca 77As stock solution (1.05 mCi, 1.3 × 10−11 mol, ∼1 ng (mass
calculated based on activity only) followed by addition of ammo-
nium mercaptoacetate (50 μL of 500 mM) while stirring in a 60 °C
water bath. After 30 min, a 10-μL aliquot of Cu0 nanoparticle solu-
tion (100 mg/mL in acetone) was added, followed by 100 μL of
p-ethoxybenzenediazonium tetrafluoroborate (0.1 mg/ μL) in
acetonitrile. The reaction mixture was removed from the water
bath and stirred at room temperature for 45 min. Ammonium
mercaptoacetate (100 μL of 5.5 M) was added and the reaction
mixture was placed in a 60 °C water bath (45 min). Ethanedithiol
(10 μL; 1.123 g/mL) was added, and the reaction mixture was stirred
at room temperature for 30 min. Silica gel TLC using 9% acetone in
methanol and radioisotope detection (Bioscan AR-2000) was used to
determine yield and follow the progress of the reaction: 77AsO43−(V),
Rf = 0; 77As(mercaptoacetate)3, Rf = 0.64; [As77]p-ethoxyphenylarsonic
acid, Rf = 0.24; (2-(4-ethoxyphenyl))-77As(mercaptoacetate)2,
Scheme 2. Synthesis of Phenyl Arsonic Acids Using the Bart
Method
Improved reaction yields were reported when starting with the
diazonium tetrafluoroborate salt, and this method was selected
for the initial radiosynthesis.32
Synthesis and Characterization of Dithioarylarsines.
As the first step in the development of potential radiopharma-
ceuticals for PET imaging with 72As or therapy with 77As, a
number of dithioarylarsines were synthesized, characterized,
and evaluated for aqueous stability. The reaction shown in
Scheme 1 was used to synthesize the various dithioarylarsines
C
DOI: 10.1021/acs.inorgchem.6b01175
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(1−3, Supporting Information S4−S12). Two arylarsonic acids
(p-nitro, p-ethoxy) were not commercially available and were
synthesized following literature methods.32 The intermediates
in the reaction, the bis(mercaptoacetato)arylarsines, were
prepared from their respective aryl arsonic acids by reduction
and complexation with an aqueous solution of ammonium
mercaptoacetate. Subsequent reaction in situ with a bidentate
dithiol formed stable five- or six-membered chelates of the
arylarsines, often in excellent yields. Electrospray ionization
mass spectrometry (ESI-MS) and elemental analyses confirmed
the compositions and purities of the various dithioarylarsines.
In some cases, acetonitrile adducts or addition of O or OH was
observed. This is interesting since the addition of an oxo group
to the dithioarylarsines with peroxide was not observed.
biological conditions. This ensures the best potential for deliv-
ery of the radiolabeled molecule to its in vivo target site. As an
initial step in this assessment, the stability of the ethyl versus
propyl backbones of the dithioarylarsines was investigated
qualitatively for compounds S4, S5, and S6 (SI) by 1H and 13
C
NMR in aqueous or mixed aqueous/acetonitrile solutions
depending on solubility. S5 (dimercaptosuccinic acid analogue)
is water-soluble, while S4 (1,2-ethanedithiol analogue) and S6
(1,3-propanedithiol analogue) are not water-soluble and were
thus run in a 20/80 aqueous/acetonitrile solution. No change
in the 1H or 13C NMR spectra was observed over 30 days indi-
cating good stability in solution at room temperature.
Structural Studies: X-ray Crystallography. Compounds
1, 3, S4, S6−S10, and S12 were characterized by X-ray crystal-
lography. Crystal refinement data, bond angles and distances
are summarized in Tables 2 and 3, and SI Tables S2 and S3.
The 1H NMR spectra of the dithioarylarsines showed the loss
of the sulfhydryl protons at 1.33 and 1.65 ppm for 1,2-ethane
and 1,3-propane dithiol, respectively. Each dithiolate back-
bone proton became unique on complexation to the arsenic
because of its trigonal pyramidal geometry, resulting in com-
Table 2. X-ray Crystal Data, Data Collection Parameters,
a
and Refinement Parameters for 1 and 3
1
plex H NMR spectra. The chemical shifts for the backbone
1
3
alkyl protons are observed downfield on complexation to
arsenic (0−0.86 ppm depending on the proton) relative to the
free dithiols (SI Table S1, Figure 1) and consistent with values
reported by others.36,37,40,41
CCDC No.
formula
fw
1436768
C10H13AsOS2
288.24
1436769
C11H15AsOS2
302.27
The 13C NMR spectra showed significant chemical shift
changes in the alkyl C signals on complexation to arsenic rela-
tive to the free dithiol (SI Table S1, Figure 1), consistent with
crystal system
space group
a (Å)
orthorhombic
Pbca
monoclinic
P21/c
6.8669(3)
10.9641(5)
30.5022(12)
90°
90°
90°
15.4565(10)
6.9759(4)
11.5428(7)
90°
91.322(2)°
90°
b (Å)
c (Å)
α (deg)
β (deg)
γ (deg)
V (Å3)
Z
2296.49(17)
8
1244.25(13)
4
r
calc, g/cm3
1.667
1.614
T, K
100(2)
7.121
100(2)
μ, mm−1
λ source (Å)
R(F)
6.601
1.54178
0.027
1.54178
0.2069
Rw(F)2
0.0594
0.0722
GOF
0.861
1.083
a
2
R = (Σ||FO| − |FC||/Σ|FO||). RW = [Σϖ(|FO2| − |FC |)2/Σϖ(|FO2|2]1/2
.
Table 3. Selected Bond Distances (Å) and Angles (o) for
1 and 3
1
3
Figure 1. Ethane- and propanedithioarylarsines showing the proton
assignments from H NMR spectrometry.
aryl
p-OEt
p-OEt
1
dithiol
ethyl
propyl
As−S1
As−S2
As−C1
S1−As−S2
S1−As−C1
S2−As−C1
2.2476(7)
2.2409(7)
1.961(2)
92.62(3)
101.56(8)
100.67(8)
2.2224(15)
2.2274(16)
1.958(6)
100.95(6)
102.11(17)
100.03(18)
values reported previously.36,37,40,41 The arylarsine−DMSA
complexes exhibited a 22−25 ppm downfield chemical shift
relative to DMSA, dependent on the particular arylarsine.
Similarly, the ethane dithiol complexes exhibited a downfield
13C shift of 12−14 ppm relative to the free dithiol, while the
propane dithiol complexes displayed downfield shifts ranging
from 2 to 4 ppm for the central carbon and upfield shifts of
8−10 ppm for the carbon adjacent to sulfur. The shifts in the
1H and 13C NMR values reported here are consistent with
values reported by others for similar compounds.33−36
These structures contain either five- or six-membered rings on
chelation of two sulfurs to the arsenic, resulting in a trigonal
pyramidal geometry (Figures 2, 3, and SI Figures S1−S7).
Comparison of the observed values with literature reports
is limited to only two other structures reported, one with
British anti-Lewisite (HSCH2CH(CH2OH)SH, BAL) to form
a 5-membered dithioarylarsine chelate ring and one with
Stability by 1H NMR Spectroscopy. Radiopharmaceutical
applications require that the radionuclide is incorporated in
a compound that is kinetically stable at high dilution under
D
DOI: 10.1021/acs.inorgchem.6b01175
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
and Sx−As−Cx, range from 91.94(2)° to 93.91(2)° and 97.49(5)°
to 101.56(8)°, respectively, for the five-membered chelate ring
complexes and from 97.27(2)° to 100.95(6)° and 97.05(6)° to
102.11(17)°, respectively, for the six-membered chelate ring
complexes. As expected, the six-membered chelate ring com-
plexes have a larger S−As−S bite angle. The lengths and angles
observed are typical and fall within or near the range reported
for other dithioarylarsine complexes.40−42
Synthesis and Initial Optimization of No-Carrier-
Added 77As 2-(4-Ethoxyphenyl)-1,3,2-dithiarsolane
[CH3CH2OC6H577As(SCH2CH2S)]. Translation of the dithioar-
ylarsine synthesis to the nca radiotracer level included four
essential steps: reduction, aryl incorporation, reduction, and
dithiol complexation (Schemes 3−5).
No-carrier-added [77As]dithioarylarsine syntheses started
from an aqueous NaH277AsO4 solution (∼2 nM) at pH 4.5−5.
The first step (Scheme 3) is reduction of the 77As(V) to 77As(III),
Scheme 3. Reduction of [77As]arsenate using Ammonium
Mercaptoacetate
which is critical since incorporation of the aryl group by these
methods does not occur with pentavalent arsenic. Reduction of
77As(V) to 77As(III) was accomplished with ammonium mer-
captoacetate to presumably yield a reactive tris-mercaptoacetato
species, 77As(SR)3 (Scheme 3). Addition of p-ethoxybenze-
77
nediazonium tetrafluoroborate to a solution containing the
As(SR)3 and a copper catalyst (as Cu0 nanoparticles or polished
Cu metal) yields the phenylarsonic acid in an oxidative addition/
hydrolysis reaction (Scheme 4). This product required reduction
Figure 2. ORTEP representation of (1) (CCDC 1436768) with 50%
probability ellipsoids.
Scheme 4. Incorporation of a p-Ethoxybenzene Group to nca
77As to yield p-Ethoxyphenylarsonic Acid
back to 77As(III) with mercaptoacetate resulting in presumably
a bis(mercaptoacetato)arylarsine (Scheme 5). The addition of a
dithiol displaces the mercaptoacetates to yield the final product,
the [77As]-dithioarylarsine (Scheme 5).
The reducing agent, aryl donor, and dithiol concentrations,
solvent, copper catalyst, temperature, and time were optimized
to maximize the formation of the [77As]-dithioarylarsine. An
excess (mM vs nM) of these reagents compared to the nca 77As
is necessary, however their relative concentrations must be
balanced to minimize interference with the incorporation of the
aryl moiety onto the nca 77As(III). Potential precipitation of
byproducts formed (biphenyl, phenyl-SR) can interfere with
the desired reaction and isolation of the product.
Figure 3. ORTEP representation of (3) (CCDC 1436769) with 50%
probability ellipsoids.
dihydrolipoic acid to form a 6-membered dithioarylarsine
chelate ring.40,41
The As−Sx, and As−C1 bond lengths in these compounds
range from 2.2224(15) to 2.2573(6) Å, and 1.940(2) to 1.986(2) Å,
respectively, with no significant difference between the
five- and six-membered chelate ring complexes (Table 3 and
SI Table S3). Angles observed around the As center, Sx−As−Sy
E
DOI: 10.1021/acs.inorgchem.6b01175
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Scheme 5. Reduction and Complexation of p-
Ethoxyphenylarsonic Acid with Ammonium Mercaptoacetate
Followed by Substitution with the Bidentate 1,2-Ethane
Dithiol
Table 5. Initial Optimization of the Ammonium
Mercaptoacetate Concentration in the Reduction of [As77]
a
p-Ethoxyphenylarsonic Acid (n = 3)
concn of reducing agent (mM)
yield (%)
108
160
210
260
26
36
50
54
3
4
5
3
a
Conditions: 9:1, ACN/DI water; 60 °C; 45 min; 30 mM monothiol
in the reduction step; 15 mg of p-ethoxybenzenediazonium
tetrafluoroborate.
The first optimization (Scheme 3) involving mercaptoacetate
reduction of arsenate showed that at 1 mM concentration
the reduction yield of the nca [77As]arsenic acid fell to 36%
(Table 4). Thus, 30 mM mercaptoacetate was used during
Table 6. Initial Optimization of the Dithiol
a
(1,2-Ethanedithiol) Concentration (n = 3)
concn of 1,2-ethanedithiol (mM)
yield (%)
9.8
26
36
50
54
57
2
3
5
3
1
Table 4. Optimization of the Reducing Agent (Ammonium
Mercaptoacetate) Concentration for Arsenate Reduction
(n = 2)
19.3
28.3
37.5
100
a
concn of reducing agent (mM)
yield (%)
a
40
30
20
10
1
99
100
99
5
6
8
7
6
Conditions: 9:1, ACN/DI water; 60 °C; 45 min; 30 mM monothiol
in reduction step; 15 mg of p-ethoxybenzenediazonium tetrafluor-
oborate; 270 mM monothiol during incorporation step.
94
36
a
observed in the reaction vials: a pale yellow organic layer above
a clear aqueous layer. Analysis of the 2 layers showed the
aqueous phase contained more 77As activity but less product,
while the organic phase contained less 77As but with more
product. The separation of phases is likely due to the high NaCl
concentration originating from the 77As solution, which resulted
in salting out the acetonitrile. Modifying the acetonitrile/water
ratio to 7:3 maintained a single homogeneous reaction and
increased the product yields.
Various temperatures and times were evaluated to optimize
the various reduction, aryl incorporation, and dithiol complex-
ation steps. During the first reduction step (Scheme 3), yields
reached a plateau after 30 min; at 30 mM ammonium mercap-
toacetate and 60 °C, the reduction yield was over 90% after
30 min. In terms of temperature effects, the yield for the first
reduction step increased when the reaction temperature was
increased from 50 to 60 °C. However, the higher temperature
(60 °C) for this reduction step led to discoloration (dark
brown color) and a significant decrease in yield during the aryl
incorporation in the following step (Scheme 4). For the aryl
incorporation step, the reaction mixture was removed from the
heat reaching 90% yield after 45 min, with no further change
observed at longer reaction times. Temperature had no
significant effect on the second ammonium mercaptoacetate
reduction or the dithiol complexation (final step), and they
were run at room temperature (Scheme 5); 95% yield was
observed during the dithiol complexation after 30 min.
Conditions: 9:1, ACN/DI water; 60 °C; 30 min.
subsequent optimization studies to ensure complete reduction
of the 77As to the +3 oxidation state, which is required for sub-
sequent incorporation of the aryl group. Too low of a con-
centration of mercaptoacetate reduced the yield in the sub-
sequent step, likely due to air oxidation.
Incorporation of the aryl group (Scheme 4) involves reaction
of As(III) with an aryl diazonium in the presence of a Cu
catalyst. At the macroscopic level, Cu(s)/Cu(II) and Cu(I)
have been used. This is not feasible at the nca radiotracer level
because the Cu(I)/Cu(II) concentration would react with
monothiol present from As(V) reduction and interfere with the
desired reaction. Additionally, dithiol added in the subse-
quent step would also interfere. Thus, polished Cu metal and
Cu nanoparticles were evaluated, and both were equally good
as the catalyst. Because the nanoparticle addition was more
reproducible, it is the catalyst of choice.
The amount of p-ethoxybenzenediazonium tetrafluoroborate
was varied from 5 to 20 mg, in 5-mg increments, in the opti-
mization. The yield increased to 10 mg and then remained
constant after that point.
On the macroscopic level, incorporation of the aryl group
generates the arylarsonic acid, which has the arsenic in
oxidation state +5. At the nca 77As radiotracer level, both the
arylarsonic acid and a reduced arylarsine, likely the bis-
(monothio)arylarsine, were formed. An additional reduction
with mercaptoacetate was necessary to allow incorporation
of the dithiol. The reducing agent (mercaptoacetate) and
1,2-ethanedithiol concentrations were individually varied to
optimize the product yield (Tables 5 and 6). Dithiol alone did
not yield the desired product at radiotracer level and poor
yields were observed at the macroscopic level.
The optimal conditions involved a single-vial, multistep
aqueous synthesis involving reduction of 77As(V)-arsenate (∼1
mCi, ∼1 × 10−11 mol, ∼1 ng) with the water-soluble mercap-
toacetate as the first step. Following the reduction to 77As(III),
the aryl group was incorporated by addition of p-ethoxybenze-
nediazonium tetrafluorborate in the presence of Cu0 nano-
particles to generate the aryl arsonic acid. A second reduction
step was required to form the aryl bis(mercaptoacetato)arsine,
and finally dithiol substitution of the monothiols yielded the
desired product. The synthesis of nca [77As]2-(4-ethoxyphen-
The acetonitrile/water solvent ratio also played an important
role during the reduction, incorporation, and complexation
reactions. The initial optimizations were carried out in 9:1
acetonitrile/water (Tables 4−6). However, two layers were
F
DOI: 10.1021/acs.inorgchem.6b01175
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(NMR chemical shifts); Table S2 (X-ray crystallographic
information for S4, S6−S10, S12); Table S3 (selected
bond distances/angles for S4, S6−S10, S12) (PDF)
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: jurissons@missouri.edu; phone: 573-882-2107; fax:
573-882-2754.
Present Address
∥Medical Isotope Research and Production Program (MIRP),
Collider-Accelerator Department, Brookhaven National Labo-
ratory, PO Box 5000, Building 801, Upton, NY 11973.
Author Contributions
§Authors Anthony DeGraffenreid and Yutian Feng contributed
equally and are cofirst authors.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by NIBIB training grant 5T32-
EB004822 (A.J.D.) and DOE grants DE-SC0003851 and DE-
SC0010283. Special thanks to Dr. Nathan Leigh, University of
Missouri, for the collection of ESI-MS data, and Dr. Wei
Wycoff, University of Missouri, for assistance with NMR and
the respective facilities used.
REFERENCES
■
(1) Greenwood, N. N.; Earnshaw, A. Chemistry of the Elements, 2nd
ed.; Butterworth-Heinemann: Oxford, 1997; pp 547−599.
(2) Mudhoo, A.; Sharma, S. K.; Garg, V. K.; Tseng, C.-H. Crit. Rev.
Environ. Sci. Technol. 2011, 41 (5), 435−519.
Figure 4. HPLC chromatograms of (a) macroscopic standard of 1
(UV detection) and (b) nca [77As]1 (radioisotope detection) with
both exhibiting retention times of 28.67 min.
(3) Akao, Y.; Mizoguchi, H.; Kojima, S.; Naoe, T.; Ohishi, N.; Yagi,
K. Br. J. Haematol. 1998, 102, 1055−1060.
yl)-1,3,2-dithiarsolane ([77As]1) was successfully optimized to
>95% yield, and it eluted with the same HPLC retention time
as its nonradioactive standard 1 (Figure 4).
(4) Yuan, B.; Yoshino, Y.; Kaise, T.; Toyoda, H. Biological Chemistry
of Arsenic, Antimony and Bismuth; John Wiley & Sons, Ltd, 2010; pp
263−292.
CONCLUSION
(5) Sweet, W. H.; Brownell, G. L. J. Am. Med. Assoc. 1955, 157 (14),
1183−1188.
■
Several dithioarylarsines were synthesized and fully characterized
(including their crystal structures) as macroscopic standards
for the nca 77As chemistry. Successful incorporation of an aryl
group onto nca 77As, beginning with arsenate (77AsO43−), was
achieved in close to quantitative yield with various modifi-
cations to the Bart reaction to optimize the production of the
aryl arsonic acids. Subsequent incorporation of a dithiol gave
the [77As]-dithioarylarsine in 95% radiochemical yield. The
reactions were carried out in high mercaptoacetate (monothiol)
concentration and the products remained intact over several
days (based on radio thin layer chromatography (radioTLC)),
suggesting these compounds will likely be stable to challenge by
thiols found in vivo, such as glutathione and the sulfhydryl
groups of albumin. This is only the second fully characterized,
nca radioarsenic compound synthesized following our recently
reported 77As-trithiol.31 Synthesis of a bifunctional chelate for
conjugation to a suitable targeting vector is currently underway.
(6) Mego, J. L.; McQueen, J. D. J. Nucl. Med. 1964, 5, 920−928.
(7) Jennewein, M.; Lewis, M. A.; Zhao, D.; Tsyganov, E.; Slavine, N.;
He, J.; Watkins, L.; Kodibagkar, V. D.; O’Kelly, S.; Kulkarni, P.; Antich,
P. P.; Hermanne, A.; Rosch, F.; Mason, R. P.; Thorpe, P. E. Clin.
Cancer Res. 2008, 14, 1377−1385.
(8) Jennewein, M.; Hermanne, A.; Mason, R. P.; Thorpe, P. E.;
Rosch, F. Nucl. Instrum. Methods Phys. Res., Sect. A 2006, 569, 512−
̈
517.
(9) Herth, M. M.; Barz, M.; Jahn, M.; Zentel, R.; Rosch, F. Bioorg.
̈
Med. Chem. Lett. 2010, 20, 5454−5458.
(10) Pimlott, S. L.; Sutherland, A. Chem. Soc. Rev. 2011, 40, 149−
162.
(11) Carroll, V.; Demoin, D. W.; Hoffman, T. J.; Jurisson, S. S.
Radiochim. Acta 2012, 100, 653−667.
(12) Cutler, C. S.; Hennkens, H. M.; Sisay, N.; Huclier-Markai, S.;
Jurisson, S. S. Chem. Rev. 2013, 113 (2), 858−883.
(13) Ogawa, K.; Ishizaki, A. BioMed Res. Int. 2015, 2015, 1−12.
(14) National Nuclear Data Center (www.nndc.bnl.gov).
(15) Wycoff, D. E.; Gott, M. D.; DeGraffenreid, A. J.; Morrow, R. P.;
Sisay, N.; Embree, M. F.; Ballard, B.; Fassbender, M. E.; Cutler, C. S.;
Ketring, A. R.; Jurisson, S. S. J. .Chromatogr. A 2014, 1340, 109−114.
(16) Ballard, B.; Wycoff, D.; Birnbaum, E. R.; John, K. D.; Lenz, J.
W.; Jurisson, S. S.; Cutler, C. S.; Nortier, F. M.; Taylor, W. A.;
Fassbender, M. E. Appl. Radiat. Isot. 2012, 70, 595−601.
(17) Ballard, B.; Nortier, F. M.; Birnbaum, E. R.; John, K. D.; Phillips,
D. R.; Fassbender, M. E. Curr. Radiopharm. 2012, 5, 264−270.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the ACS
Publications website at DOI: 10.1021/acs.inorgchem.6b01175.
■
S
Synthesis and characterization of S4−S12; X-ray crystal
structures of compounds S4, S6−S10, and S12; Table S1
G
DOI: 10.1021/acs.inorgchem.6b01175
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(18) Jennewein, M.; Qaim, S. M.; Kulkarni, P. V.; Mason, R. P.;
Hermanne, A.; Rosch, F. Radiochim. Acta 2005, 93, 579−583.
̈
(19) Al-Kouraishi, S. H.; Boswell, G. G. Int. J. Appl. Radiat. Isot. 1978,
29, 607−609.
(20) Phillips, D. R.; Hamilton, V. T.; Taylor, M. D.; Farnham, J. E.;
Emraan, A. M.; Rowe, R. W.; Patel, D. Radioact. Radiochem. 1992, 3,
53−58.
(21) Nayak, T. K.; Brechbiel, M. W. Bioconjugate Chem. 2009, 20,
825−841.
(22) Bockisch, A. Eur. J. Nucl. Med. Mol. Imaging 2011, 38, 1−3.
(23) Shehata, M. M.; Scholten, B.; Spahn, I.; Coenen, H. H.; Qaim, S.
M. J. Radioanal. Nucl. Chem. 2011, 287, 435−442.
(24) Jennewein, M.; Qaim, S. M.; Hermanne, A.; Jahn, M.; Tsyganov,
E.; Slavine, N.; Seliounine, S.; Antich, P. A.; Kulkarni, P. V.; Thorpe, P.
E.; Mason, R. P.; Rosch, F. Appl. Radiat. Isot. 2005, 63, 343−351.
(25) Gott, M. D.; DeGraffenreid, A. J.; Feng, Y.; Phipps, M. D.;
Wycoff, D. E.; Embree, M. F.; Cutler, C. S.; Ketring, A. R.; Jurisson, S.
S. J. Chromatogr. A 2016, 1441, 68−74.
(26) Jennewein, M. Ph.D. dissertation. Production, Radiochemical
Separation and Chemical Coupling of Radioactive Arsenic Isotopes to
Synthesize Radiopharmaceuticals for Molecular Imaging; Johannes
Gutenberg University, Mainz, Germany, 2005.
(27) Ellison, P. A.; Barnhart, T. E.; Chen, F.; Hong, H.; Zhang, Y.;
Theuer, C. P.; Cai, W.; Nickles, R. J.; DeJesus, O. T. Bioconjugate
Chem. 2016, 27, 179−188.
(28) Kitchin, K. T. Biological Chemistry of Arsenic, Antimony and
Bismuth; John Wiley & Sons, Ltd, 2010; pp 19−51.
(29) Kitchin, K. T.; Wallace, K. J. Inorg. Biochem. 2008, 102, 532−
539.
(30) Aposhian, H. V. Annu. Rev. Pharmacol. Toxicol. 1983, 23, 193−
215.
(31) DeGraffenreid, A. J.; Feng, Y.; Barnes, C. L.; Ketring, A. R.;
Cutler, C. S.; Jurisson, S. S. Nucl. Med. Biol. 2016, 43, 288−295.
(32) Hamilton, C. S.; Morgan, J. F. Chapter 10: The Preparation of
Aromatic Arsonic and Arsinic Acids by the Bart, Bechamp, and
Rosenmund Reactions. In Organic Reactions; John Wiley & Sons, Inc.,
2004; pages 415−454 (DOI: 10.1002/0471264180.or002.10).
(33) Heredia-Moya, J.; Kirk, K. L. Bioorg. Med. Chem. 2008, 16,
5743−5746.
(34) Lloyd, N. C.; Morgan, H. W.; Nicholson, B. K.; Ronimus, R. S. J.
Organomet. Chem. 2008, 693, 2443−2450.
(35) Starkey, E. B. Org. Synth. 1943, 2, 225.
(36) Loiseau, P. M.; Lubert, P.; Wolf, J.-G. Antimicrob. Agents
Chemother. 2000, 44, 2954−2961.
(37) Gibaud, S.; Alfonsi, R.; Mutzenhardt, P.; Fries, I.; Astier, A. J.
Organomet. Chem. 2006, 691, 1081−1084.
(38) Sheldrick, G. Acta Crystallogr., Sect. A: Found. Crystallogr. 2008,
64 (1), 112−122.
(39) Sheldrick, G. SADABS; University of Gottingen: Germany,
̈
1996.
(40) Adams, E.; Jeter, D.; Cordes, A. W.; Kolis, J. W. Inorg. Chem.
1990, 29, 1500−1503.
(41) von Dollen, A.; Strasdeit, H. Eur. J. Inorg. Chem. 1998, 1998,
̈
61−66.
(42) Shaikh, T. A.; Bakus, R. C.; Parkin, S.; Atwood, D. A. J.
Organomet. Chem. 2006, 691, 1825−1833.
H
DOI: 10.1021/acs.inorgchem.6b01175
Inorg. Chem. XXXX, XXX, XXX−XXX