Optimization of the alkyl side chain length of fluorine-18-labeled 7α-alkyl-fluoroestradiol

Article abstract of DOI:10.1016/j.nucmedbio.2016.05.008

Introduction: Several lines of evidence suggest that 7α-substituted estradiol derivatives bind to the estrogen receptor (ER). In line with this hypothesis, we designed and synthesized ¹⁸F-labeled 7α-fluoroalkylestradiol (Cn-7α-[¹⁸F]FES) derivatives as molecular probes for visualizing ERs. Previously, we successfully synthesized 7α-(3-[¹⁸F]fluoropropyl)estradiol (C3-7α-[¹⁸F]FES) and showed promising results for quantification of ER density in vivo, although extensive metabolism was observed in rodents. Therefore, optimization of the alkyl side chain length is needed to obtain suitable radioligands based on Cn-7α-substituted estradiol pharmacophores. Methods: We synthesized fluoromethyl (23; C1-7α-[¹⁸F]FES) to fluorohexyl (26; C6-7α-[¹⁸F]FES) derivatives, except fluoropropyl (C3-7α-[¹⁸F]FES) and fluoropentyl derivatives (C5-7α-[¹⁸F]FES), which have been previously synthesized. In vitro binding to the α-subtype (ERα) isoform of ERs and in vivo biodistribution studies in mature female mice were carried out. Results: The in vitro IC₅₀ value of Cn-7α-FES tended to gradually decrease depending on the alkyl side chain length. C1-7α-[¹⁸F]FES (23) showed the highest uptake in ER-rich tissues such as the uterus. Uterus uptake also gradually decreased depending on the alkyl side chain length. As a result, in vivo uterus uptake reflected the in vitro ERα affinity of each compound. Bone uptake, which indicates de-fluorination, was marked in 7α-(2-[¹⁸F]fluoroethyl)estradiol (C2-7α-[¹⁸F]FES) (24) and 7α-(4-[¹⁸F]fluorobutyl)estradiol (C4-7α-[¹⁸F]FES) (25) derivatives. However, C1-7α-[¹⁸F]FES (23) and C6-7α-[¹⁸F]FES (26) showed limited uptake in bone. As a result, in vivo bone uptake (de-fluorination) showed a bell-shaped pattern, depending on the alkyl side chain length. C1-7α-[¹⁸F]FES (23) showed the same levels of uptake in uterus and bone compared with those of 16α-[¹⁸F]fluoro-17β-estradiol. Conclusions: The optimal alkyl side chain length of ¹⁸F-labeled 7α-fluoroalkylestradiol was the shortest: C1-7α-[¹⁸F]FES. Our results indicate that shorter chain lengths within the 4-? ligand binding cavities of ERα are suitable for 7α-fluoroalkylestradiol derivatives.

Full text of DOI:10.1016/j.nucmedbio.2016.05.008

Nuclear Medicine and Biology 43 (2016) 512–519
Contents lists available at ScienceDirect
Nuclear Medicine and Biology
journal homepage: www.elsevier.com/locate/nucmedbio
Optimization of the alkyl side chain length of fluorine-18-labeled
7α-alkyl-fluoroestradiol
Mayumi Okamoto a,b, Hiromitsu Shibayama c,d, Kyosuke Naka c, Yuya Kitagawa c, Kiichi Ishiwata a,e,f
,
Isao Shimizu c,†, Jun Toyohara a,
⁎
a
Research Team for Neuroimaging, Tokyo Metropolitan Institute of Gerontology, Tokyo, Japan
b
c
d
e
f
Research Institute for Science and Engineering, Waseda University, Tokyo, Japan
School of Advanced Science and Engineering, Waseda University, Tokyo, Japan
Fine Organic Chemistry, Institute for Chemical Research, Kyoto University, Kyoto, Japan
Institute for Cyclotron and Drug Discovery Research, Southern Tohoku Research Institute of Neuroscience, Koriyama, Japan
Department of Biofunctional Imaging, Fukushima Medical University, Fukushima, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Introduction: Several lines of evidence suggest that 7α-substituted estradiol derivatives bind to the estrogen
receptor (ER). In line with this hypothesis, we designed and synthesized 18F-labeled 7α-fluoroalkylestradiol
(Cn-7α-[18F]FES) derivatives as molecular probes for visualizing ERs. Previously, we successfully synthesized
7α-(3-[18F]fluoropropyl)estradiol (C3-7α-[18F]FES) and showed promising results for quantification of ER
density in vivo, although extensive metabolism was observed in rodents. Therefore, optimization of the alkyl
side chain length is needed to obtain suitable radioligands based on Cn-7α-substituted estradiol
pharmacophores.
Received 14 April 2016
Received in revised form 16 May 2016
Accepted 19 May 2016
Keywords:
Estrogen receptor
C-7α-[18F]FES
Methods: We synthesized fluoromethyl (23; C1-7α-[18F]FES) to fluorohexyl (26; C6-7α-[18F]FES) derivatives, ex-
cept fluoropropyl (C3-7α-[18F]FES) and fluoropentyl derivatives (C5-7α-[18F]FES), which have been previously
synthesized. In vitro binding to the α-subtype (ERα) isoform of ERs and in vivo biodistribution studies in mature
female mice were carried out.
Structure–activity relationship
Positron emission tomography
Results: The in vitro IC50 value of Cn-7α-FES tended to gradually decrease depending on the alkyl side chain
length. C1-7α-[18F]FES (23) showed the highest uptake in ER-rich tissues such as the uterus. Uterus uptake
also gradually decreased depending on the alkyl side chain length. As a result, in vivo uterus uptake reflected
the in vitro ERα affinity of each compound. Bone uptake, which indicates de-fluorination, was marked in 7α-
(2-[18F]fluoroethyl)estradiol (C2-7α-[18F]FES) (24) and 7α-(4-[18F]fluorobutyl)estradiol (C4-7α-[18F]FES)
(25) derivatives. However, C1-7α-[18F]FES (23) and C6-7α-[18F]FES (26) showed limited uptake in bone. As a
result, in vivo bone uptake (de-fluorination) showed a bell-shaped pattern, depending on the alkyl side chain
length. C1-7α-[18F]FES (23) showed the same levels of uptake in uterus and bone compared with those of
16α-[18F]fluoro-17β-estradiol.
Conclusions: The optimal alkyl side chain length of 18F-labeled 7α-fluoroalkylestradiol was the shortest: C1-7α-
[
18F]FES. Our results indicate that shorter chain lengths within the 4-Å ligand binding cavities of ERα are suitable
for 7α-fluoroalkylestradiol derivatives.
© 2016 Elsevier Inc. All rights reserved.
1. Introduction
them, 16α-[18F]fluoro-17β-estradiol (16α-[18F]FES) [1] is currently
used as the standard radioligand for imaging both primary and meta-
static estrogen receptor-positive tumors [2,3].
Fluorine-18-labeled steroid hormones are useful probes for positron
emission tomography visualization of receptor-positive tumors such as
breast and prostate cancer. Over the past 30 years, several derivatives of
18F-labeled 17β-estradiol have been synthesized and evaluated. Among
On the other hand, several lines of evidence suggest that 7α-
substituted estradiol derivatives bind to the estrogen receptor (ER),
even though they have a long chain with complex functionality
[4–11]. In contrast, small polar groups at the 7α-position do not bind
well [12,13]. Furthermore, substitutions with 7α-alkyl chains bearing
alcohol, carboxylic acid, and ester groups also have low affinity [6].
Thus, Anstead et al. recommended that groups at the 7α-position bind
well to ERs, even if they are rather long; however, polar functions
must be positioned away from the core of the steroid structure [14].
⁎
Corresponding author at: Research Team for Neuroimaging, Tokyo Metropolitan Insti-
tute of Gerontology, 35-2 Sakae-cho, Itabashi-ku, Tokyo 173-0015, Japan. Tel.: +81 3 3964
3241; fax: +81 3 3964 1148.
E-mail address: toyohara@pet.tmig.or.jp (J. Toyohara).
Deceased.
†
http://dx.doi.org/10.1016/j.nucmedbio.2016.05.008
0969-8051/© 2016 Elsevier Inc. All rights reserved.

M. Okamoto et al. / Nuclear Medicine and Biology 43 (2016) 512–519
513
In line with this hypothesis, we are interested in the design and synthe-
sis of 7α-fluoroalkylestradiol (Cn-7α-[18F]FES) as a molecular probe to
visualize ER function. Previously, French et al. synthesized the five-
carbon derivative 7α-(5-[18F]fluoropentyl)estradiol (C5-7α-[18F]FES)
and evaluated its biodistribution in immature rats [5]. Although C5-
7α-[18F]FES showed somewhat selective uptake in target tissues, the
levels of uptake into non-target tissues were high, possibly due to the
increased lipophilicity of the additional five-carbon chain. In a recent
follow-up study, we investigated the shorter three-carbon derivative
7α-(3-[18F]fluoropropyl)estradiol (C3-7α-[18F]FES) and characterized
its biological properties [15]. As expected, the shorter three-carbon
chain resulted in lower uptake into non-target tissues, such as fat and
blood, and the uterus-to-blood ratio at 60 min was double that of the
five-carbon chain derivative. However, the three-carbon derivative
underwent greater metabolic de-fluorination than the five-carbon
derivative. These results indicate that opportunities still exist for
further optimization of the alkyl side chain length of Cn-7α-[18F]FES
derivatives.
mature female mice compared to the previously published data for
C3-7α-[18F]FES and 16α-[18F]FES, and we discuss the optimization of
the alkyl side chain length.
2. Materials and methods
2.1. Chemical synthesis
The methods of synthesis of non-radioactive compounds are
outlined in Schemes 1–3, and the details are summarized in the
supporting information in the online version at http://dx.doi.org/10.
1016/j.nucmedbio.2016.05.008.
2.2. Radiochemical synthesis
[
18F]Fluoride was produced by proton irradiation of 18O-enriched
water (Taiyo Nippon Sanso, Tokyo, Japan) at 50 μA for 5 min using the
HM-20 cyclotron (Sumitomo Heavy Industries, Tokyo, Japan). Isolation
of [18F]fluoride from enriched water and subsequent 18F-fluorination,
de-protection, purification, and formulation were carried out auto-
matically by using a multi-purpose synthetic apparatus (MPS-200;
Sumitomo Heavy Industries). Preparative high-performance liquid
chromatography (HPLC) was done using a Prominence (Shimadzu,
Kyoto, Japan) that was equipped with an ultraviolet (UV) absorbance
detector set at 280 nm and a semiconductor radiation detector system
in the indicated conditions. Radiochemical purity and specific activity
(SA) of the labeled compounds were determined with the same HPLC
system using the indicated conditions.
Currently, the most common route for preparing 16α-[18F]FES uses
3-methoxymethyl-16β,17β-epistriol-O-cyclic sulfate [16]. However,
some difficulties remain regarding the time required for (N30 min)
acid hydrolysis of the bisulphate intermediate [17], and further optimi-
zation is required for routine clinical use in individual facilities [18]. In
contrast, di-methoxymethyl-protected groups of labeling precursors of
Cn-7α-[18F]FES derivatives may be removed more quickly and have ac-
ceptable times required for radiosynthesis.
In this study, we further synthesized fluoromethyl (22; C1-7α-[18F]
FES) to fluorohexyl (25; C6-7α-[18F]FES) derivatives of Cn-7α-[18F]FES,
except fluoropropyl (C3-7α-[18F]FES) and fluoropentyl derivatives (C5-
7α-[18F]FES), which were synthesized previously [5,15]. We character-
ized the in vitro binding and in vivo distribution of these derivatives in
[
18F]Fluoride was separated with anion exchange resin (Sep-Pak
Light Accell Plus QMA; Nihon Waters, Tokyo, Japan). Elution with
OMOM
OMOM
OMOM
H
H
H
a
b
H
O
H
H
O
H
O
H
H
OMe
OH
MOMO
MOMO
MOMO
OH
3
(α : β = 1 : 3)
1
2
OMOM
OMOM
d or e
H
H
c
H
4
H
H
H
OH
OR
MOMO
MOMO
5a, R = Ts
5b, R = Ns
OMOM
OMOM
H
H
f
5b
H
H
H
6
H
F
MOMO
MOMO
29
OMOM
OH
H
g
H
H
H
F
H
H
F
MOMO
HO
7
6
Scheme 1. Synthesis schemes for the precursor for C1-7α-[18F]FES (5a and 5b) synthesis and authentic standard (7). Reagents and conditions: (a) sodium hydride, dimethyl carbonate,
reflux, 1 h (yield: 24%; α:β = 1:3); (b) lithium aluminium hydride, tetrahydrofuran, room temperature, 3 h (yield: 87%); (c) hydrogen (1 atm), palladium hydroxide, ethanol, room
temperature, 6 h (yield: 73%); (d) p-toluenesulfonyl chloride, triethylamine, dichloromethane, room temperature, overnight (yield of 5a: 93%), (e) p-nitrobenzenesulfonyl chloride,
triethylamine, dichloromethane, room temperature, 2 h (yield of 5b: 73%); (f) tetra-n-butyl ammonium fluoride, tert-butyl alcohol, reflux, 20 min (yield of 6: 14%, yield of 29: 22%);
(g) 1 N hydrochloric acid, tetrahydrofuran, reflux, 30 min (yield: 93%).

514
M. Okamoto et al. / Nuclear Medicine and Biology 43 (2016) 512–519
OMOM OMOM
OMOM
H
H
H
a
b
H
O
H
H
O
H
H
O
H
MOMO
MOMO
MOMO
OMOM
1
8
8a
OMOM
c, d
H
H
e
f
H
H
H
H
MOMO
OH
MOMO
OH
OH
9
10
OMOM
OH
g, h
H
H
H
H
H
H
MOMO
OTs
HO
F
12
11
Scheme 2. Synthesis schemes for the precursor for C2-7α-[18F]FES (11) synthesis and authentic standard (12). Reagents and conditions: (a) allyl iodide, potassium
bis(trimethylsilyl)amide, tetrahydrofuran, 0 °C, 2 h (yield: 32%); (b) sodium methoxide, methanol, reflux, overnight (yield: quant.); (c) ozone, methanol, −78 °C, 5 min, (d) sodium
borohydride, −78 °C to room temperature, overnight (yield: 78%); (e) hydrogen (1 atm), palladium hydroxide, ethanol, room temperature, 6 h (33%); (f) p-toluenesulfonyl chloride,
triethylamine, dichloromethane, room temperature, overnight (yield: 98%); (g) tetra-n-butyl ammonium fluoride, acetonitrile, reflux, 15 min; (h) 1 N hydrochloric acid, reflux, 20 min
(yield: 64%).
0.1 M aqueous potassium carbonate (0.3 mL, 30 μmol) resulted in a so-
lution of potassium [18F]fluoride. [18F]Fluoride dissolved in 66 mmol
aqueous potassium carbonate (0.3 mL, 19.8 μmol) was added to a solu-
tion of 4,7,13,16,21,24-hexaoxa-1,10-diazabicyclo[8.8.8]hexacosane
(Kryptofix 222; 15 mg, 39.8 μmol; Merck Millipore, Billerica, MA) in
1 mL acetonitrile. The solvents were removed azeotropically with aceto-
nitrile under a slight flow of helium at 100 °C. This procedure was con-
ducted three times in total with 1.0 mL dry acetonitrile. Thereafter,
2.0 mg (3.8–4.4 μmol) each tosylate (5a, 11, 21a, and 21b) in 1 mL dry
acetonitrile was added, and the reaction mixture was heated to 100 °C
a
b
HO
OH
HO
OBn
I
OBn
n
n
n
13a n = 2
13b n = 4
14a n = 2
14b n = 4
15a n = 2
15b n = 4
OH
OMOM
OMOM
H
H
H
c, d
e
H
O
H
H
O
H
H
H
OBn
OBn
n
HO
MOMO
MOMO
n
O
1
16a, n = 4
16b, n = 6
17a, n = 4
17b, n = 6
OH
OMOM
OMOM
g
f
h
H
H
H
H
H
H
H
OBn
H
H
OBn
HO
n
OH
n
MOMO
n
MOMO
18a, n = 4
18b, n = 6
19a, n = 4
19b, n = 6
20a, n = 4
20b, n = 6
OMOM
OH
i
j, k
H
H
H
H
H
H
OTs
F
n
MOMO
n
HO
21a: n = 4
21b: n = 6
22a: n = 4
22b: n = 6
Scheme 3. Synthesis schemes for the precursor for C4- and C6-7α-[18F]FES (21a and 21b) synthesis and authentic standard (22a and 22b). Reagents and conditions: (a) benzyl chloride,
potassium hydroxide, 130 °C, 2 h (yield of 14a: 52%, yield of 14b: 40%); (b) triphenylphosphine, iodine, imidazole, dichloromethane, room temperature, 5 h (yield of 15a: 73%, yield of 15b:
88%); (c) 15a or 16a, potassium tert-butoxide, tetrahydrofuran, room temperature, overnight; (d) sodium methoxide, methanol, reflux, overnight (yield of 16a: 40%, yield of 16b: 56%);
(e) 1 N hydrochloric acid, tetrahydrofuran, reflux, 3 h or 1 h (yield of 17a: 74%, yield of 17b: 84%); (f) triethylsilane, borontrifluoride–diethylether complex, dichloromethane, room
temperature, overnight (yield of 18a: 29%, yield of 18b: 79%); (g) chloromethyl methyl ether, N,N-diisopropylethylamine, dichloromethane, reflux, overnight (yield of 19a: 86%, yield
of 19b: 90%); (h) hydrogen (2.7 atm), palladium hydroxide, ethanol, room temperature, 2 h (yield of 20a: 86%, yield of 20b: 69%); (i) p-toluenesulfonyl chloride, triethylamine,
dichloromethane, room temperature, overnight (yield of 21a: 93%, yield of 21b: 83%); (j) tert-n-butyl ammonium fluoride, acetonitrile, reflux, 20 min; (k) 1 N hydrochloric acid, reflux,
20 min (yield of 22a: 59%, yield of 22b: 64%).

M. Okamoto et al. / Nuclear Medicine and Biology 43 (2016) 512–519
OMOM OMOM
515
OMOM
H
H
H
b
a
H
4
H
H
O
H
H
OH
H
MOMO
MOMO
MOMO
(α : β = 3 : 1)
27
28
Scheme 4. Synthesis of C2-7α-alcohol (10) using a homologation reaction. Reagents and conditions: (a) oxalyl chloride, dimethylsulfoxide, triethylamine, dichloromethane, −78 °C, 1 h to
room temperature for 20 min (α:β = 3:1); (b) triphenylphosphine, methyl iodide, tert-butyl alcohol, tetrahydrofuran.
length; TOSOH, Tokyo, Japan) was used, with 45% acetonitrile in
50 mmol sodium acetate buffer (tR: 4.1 min).
OH
OH
a, b
H
H
5a
3. Biological methods
H
H
H
H
F
HO
HO
3.1. ER binding assay
7
29
[2,4,6,7-3H(N)]-Estradiol ([3H]estradiol; 3300 GBq/mmol,
37 MBq/mL) was purchased from PerkinElmer (Boston, MA). 16α-
fluoro-17β-estradiol (16α-FES) was purchased from ABX GmbH
(Radeberg, Germany). C3-7α-FES was prepared from estradiol as re-
ported previously [15]. The purity of C3-7α-FES was confirmed by
HPLC (condition 1): tR = 18.2 min, purity =94.5%. Human recombi-
nant α-subtype (ERα) of the ER was purchased from Life Technolo-
gies (Carlsbad, CA).
Relative binding affinity (RBA) was determined by a competitive ra-
diometric assay [15] with recombinant human ERα. Samples were incu-
bated for 18–20 h at 4 °C, the receptor–ligand complexes were absorbed
onto hydroxyapatite (Wako Pure Chemical Industries), and unbound li-
gand was washed away. IC50 values from binding displacement by
blockers were determined using GraphPad Prism (GraphPad Software,
Scheme 5. Side reaction in [19F]fluorination of 5a. Reagents and conditions: (a) tert-n-
butyl ammonium fluoride, acetonitrile, reflux, 15 min; (b) 1 N hydrochloric acid, reflux,
5 min (yield of 7: trace, yield of 29: 47%).
for 15 min. After removal of the solvents under a slight flow of helium at
100 °C, 0.5 N hydrochloric acid in 50% acetonitrile in water (2 mL,
1 mmol) was added, and the reaction mixture was heated to 100 °C
for 2 min. The mixture was neutralized by addition of Meylon Injection
8.4% (v/v) (1 mL, 1 mmol; Otsuka Pharmaceutical, Naruto, Japan) and
then applied to a semi-preparative HPLC column (YMC-Pack ODS-AM,
10-mm inner diameter (i.d.) × 250-mm length; YMC, Kyoto, Japan). At
a 5 mL/min flow rate, a mixture of acetonitrile and water was used for
the mobile phase: 45% acetonitrile in water for C1-7α-[18F]FES (23)
(tR: 9.5 min) and C2-7α-[18F]FES (24) (tR: 13.1 min), and 65% acetoni-
trile in water for C4-7α-[18F]FES (25) (tR: 17.5 min) and C6-7α-[18F]
FES (26) (tR: 20.2 min). Each [18F]-product fraction was collected in a
flask containing 0.1 mL of 250 mg/mL ascorbate injection (Nipro
Pharma, Osaka, Japan) and evaporated to dryness. The residue was dis-
solved in physiological saline containing 0.125% (v/v) polyoxyethylene
(20) sorbitan monoolate (polysorbate 80) (Wako Pure Chemical Indus-
tries, Osaka, Japan), and the solution was filtered through a 0.22-μm
pore size membrane filter (Millex GV; Merck Millipore). The labeled com-
pound was analyzed by HPLC with a YMC-Pack Pro C18 RS S-5 μm column
(4.6-mm i.d. × 150-mm length; YMC) at a flow rate of 1.0 mL/min: 50%
acetonitrile in water for C1-7α-[18F]FES (23) (tR: 2.2 min) and C4-7α-
San Diego, CA). The IC50 values are expressed as the mean
standard
deviation (SD) (n = 3). The binding affinities were expressed as RBA
values, with the RBA of estradiol set to 100.
3.2. Biodistribution
Seven-week-old female ddY mice were purchased from Japan SLC
(Hamamatsu, Japan) and allowed to acclimate for at least a week prior
to use. The Animal Care and Use Committee of the Tokyo Metropolitan
Institute of Gerontology approved the animal studies.
Each 18F-labeled compound (0.2 MBq) was injected intravenously
into mice, which were sacrificed by cervical dislocation 15 and 30 min
after injection (n = 5). Blood was collected by heart puncture, and
the tissues were harvested. The samples were measured for 18F-
radioactivity with an auto-gamma counter (1282 Commpugamma CS;
[
18F]FES (25) (tR: 6.2 min), and 65% acetonitrile in 50 mmol sodium ace-
tate buffer (pH 4.6) for C6-7α-[18F]FES (26) (tR: 4.3 min). In the case of
C2-7α-[18F]FES (24), TSKgel Super-ODS 2.3 μm (4.6-mm i.d. × 100-mm
Table 1
Effect of the leaving group and solvent for fluorination of C1-7α-FES.
Entry
Leaving group (R)
Solvent
Yield of 6 (%)
1
2
3
4
5
tosylate
tosylate
nosylate
nosylate
nosylate
acetonitrile
tert-butyl alcohol
acetonitrile
tert-butyl alcohol
t-amyl alcohol
trace
trace
trace
14 (6:30 = 1:1.6)
10 (6:30 = 1:5.8)

516
M. Okamoto et al. / Nuclear Medicine and Biology 43 (2016) 512–519
OH
OMOM
and synthesis of derivatives with short side chains is particularly chal-
lenging. Therefore, we aimed to establish a novel synthesis method for
Cn-7α-FES that can be used to synthesize derivatives with a variety of
substituents, rather than just an 18F label. First, we planned the synthe-
sis of an 18F-labeled estradiol derivative via an SN2 reaction on a leaving
group. We selected a nucleophilic substitution reaction with [18F]fluo-
ride, which can easily be used with the automated synthetic apparatus.
Non-radioactive 19F-compounds are synthesized with the same pro-
cedure. Our plan also included substitution of a side chain at Cn-7α
via various alkylations on the C-6 ketone (1) and subsequent functional
group replacement. We thus, selected the C-6 ketone (1) as the key
intermediate.
a, b
H
H
H
H
H
H
18F
n
OTs
HO
MOMO
n
5a: n = 1
11: n = 2
21a: n = 4
23: n = 1
24: n = 2
25: n = 4
21b: n = 6
26: n = 6
Scheme 6. Radiofluorination of Cn-7α-FES derivatives. Reagents and conditions: (a) 18F-
fluoride, 4,7,13,16,21,24-hexaoxa-1,10-diazabicyclo[8.8.8]hexacosane, potassium carbon-
ate, acetonitrile, 100 °C, 15 min; (b) 0.5 N hydrochloric acid, 50% acetonitrile, 100 °C, 2 min
(yield of 23: 0.4%, yield of 24: 9.8%, yield of 25: 43%, yield of 26: 44%).
4.2. Synthesis of the precursors for labeling
LKB Wallac, Turku, Finland) and weighed. The tissue uptake of 18F was
expressed as dimensionless standardized uptake value (SUV) (cpm
measured per gram of tissue/cpm injected per gram of body weight).
The target (uterus and ovary)-to-non-target (blood and muscle) con-
centration ratio of radioactivity was also calculated. The previously
published biodistribution data for C3-7α-[18F]FES and 16α-[18F]FES
[15] were used for comparison with current data.
The key intermediate C-6 ketone (1) was synthesized as described
previously [19]. Then, dimethyl carbonate was reacted with the C-6 ke-
tone (1) in the presence of sodium hydroxide to yield β-ketoester (2)
(Scheme 1). The synthesized β-ketoester (2) comprised a diastereomer
mixture, with a 1:3 ratio of 7α:7β. Although 7α and 7β can be separated
by column chromatography, 7α is unstable and isomerizes when incu-
bated for several days at room temperature, resulting in a mixture con-
taining mostly 7β. Thus, the next step should be performed
immediately. The primary product during β-ketoester (2) synthesis
was 7β, which is the stereoisomer of the desired product. However, by
decarboxylation in basic conditions, we were able to recover the key in-
termediate C-6 ketone (1) from 7β at high yield. Lithium aluminium hy-
dride reduction of β-ketoester (2) yielded diol 3, after which the
hydroxyl group at the benzyl-position was reduced by a hydrogenation
reaction to synthesize C1-7α-alcohol (4) (Scheme 1).
3.3. Statistical analyses
The SD of the mean was used to determine the spread of data around
the mean. Differences between the tracer uptake of 16α-[18F]FES and
each C-7α-[18F]FES derivative in organs of interest (blood, fat, ovary,
uterus, and bone), the tissue-to-blood ratio (uterus-to-blood and
ovary-to-blood), and the tissue-to-muscle ratio (uterus-to-muscle and
ovary-to-muscle) were tested for statistical significance using one-
way analysis of variance with Dunnett's multiple comparison test for in-
dependent samples. In all analyses, P b 0.05 was considered statistically
significant. The correlation between IC50 and SUV was examined using
Pearson's correlation coefficient. All statistical analyses were performed
using GraphPad Prism (GraphPad Software, San Diego, CA).
We initially planned to synthesize the C2-7α-alcohol (10) using a
homologation reaction based on the synthesis route of C1-7α-alcohol
(4). However, isomerization at the C-7 position occurred via Swern ox-
idation of alcohol 4, which resulted in a 3:1 mixture of α and β (27)
(Scheme 4).
Thus, instead of a homologation reaction, we used an olefin synthe-
sis route with ozone oxidation via an allyl body 8 with introduction of a
4. Results and discussion
Table 2
4.1. Design consideration
Binding affinity (IC50 and RBA) of Cn-7α-FES derivatives and 16α-FES to the recombinant
α-subtype (ERα) of the human estradiol receptor and estimated alkyl length (Å) of Cn-
7α-FES derivatives.
The first reported C5-7α-[18F]FES was synthesized via a key reaction
involving addition of a 1,6 conjugate to a dienone body using the acetyl
protecting group of dihydrotestosterone as the starting material [5].
This method of synthesis cannot be universally used with other labels,
Compound
n
Estimated alkyl length (Å)
IC50 (nM)
RBA (%)
Estradiol
16α-FES
0.17
0.14
0.22
0.45
0.47
5.71
15.8
0.03
100
121
77
38
36
3
0.01
0.03
0.04
0.03
0.30†
4.18‡
C1-7α-FES (7)
C2-7α-FES (11)
C3-7α-FES
C4-7α-FES (22a)
C6-7α-FES (22b)
1
2
3
4
6
2.39
3.78
4.99
6.27
8.74
1
Relative binding affinity (RBA) values of ERα in this study were determined in a compet-
itive radiometric binding assay. Estradiol =100%. IC50 data represent the mean SD
(n = 3). RBA data represent the mean (n = 3) relative to the affinity of unlabeled estra-
diol as a standard.
Statistical analysis (one-way analysis of variance with Dunnett's multiple comparison
test) was performed between estradiol and each Cn-7α-FES derivative and 16α-FES in
IC50 values.
Fig. 1. Competitive binding assay of Cn-7α-FES for recombinant human ERα bound with
[3H]estradiol. The IC50 values were used to calculate relative binding affinity values (see
Table 2). Solid circles = estradiol; open circles = C1-7α-FES (7); solid squares = C2-
7α-FES (12); open squares = C3-7α-FES; solid triangles = C4-7α-FES (22a); open trian-
gles = C6-7α-FES (22b); solid stars =16α-FES. Data represent the means SD (n = 3).
The estimated C-F alkyl chain lengths were calculated using ChemBio3D Ultra.
†
P b 0.01 compared to estradiol.
P b 0.005 compared to estradiol.
‡

M. Okamoto et al. / Nuclear Medicine and Biology 43 (2016) 512–519
517
Table 3
Comparison of the tissue distribution of the five Cn-7α-[18F]FES derivatives with 16α-[18F]FES in mature female mice at 15 min post-injection.
16α-[18F]FES
C1-7α-[18F]FES
C2-7α-[18F]FES
C3-7α-[18F]FES
C4-7α-[18F]FES
C6-7α-[18F]FES
⁎
⁎
Tissue
Heart
Lung
Liver
Pancreas
Spleen
Small Intestine
Adrenal
Kidney
Muscle
Brain
0.22
0.68
2.77
1.36
0.24
4.02
1.82
0.75
0.27
0.19
0.19
0.62
1.24
1.77
0.28
10.23
6.65
7.22
4.67
0.03
0.15
0.48
2.92
0.36
0.20
4.68
1.03
0.65
0.26
0.11
0.12
0.47
1.00
1.62
0.34
14.16
6.44
8.47
3.89
0.03
0.06
0.22
0.03
0.03
1.43
0.12
0.08
0.03
0.02
0.03
0.05
0.09
0.58
0.04
8.56
3.22
2.69
0.84
0.42
0.76
2.78
0.54
0.30
2.27
1.08
1.17
0.29
0.17
0.51
0.60
1.15
1.88
1.71
3.72
6.60
2.25
4.00
0.03
0.02
0.33
0.03
0.01
1.47
0.36
0.64
0.03
0.01
0.04§
0.08
0.37
0.74
0.25§
1.58†
2.72
0.60¶
1.18
0.52
0.70
2.54
0.81
0.32
2.85
1.12
1.06
0.38
0.24
0.49
0.79
0.89
1.36
1.47
2.78
3.63
1.83
2.37
0.04
0.03
0.16
0.15
0.07
0.57
0.19
0.07
0.03
0.03
0.04§
0.14
0.18†
0.34
0.27§
0.66‡
1.09†
0.41¶
0.63¶
0.64
0.78
2.55
0.58
0.39
1.21
1.17
1.19
0.42
0.33
0.68
0.57
0.80
1.07
1.41
1.57
2.55
1.20
1.92
0.09
0.09
0.31
0.06
0.05
0.23
0.08
0.13
0.03
0.02
0.11§
0.21
0.10‡
0.21†
0.20§
0.12‡
0.50‡
0.29§
0.33§
0.53
0.64
2.15
0.68
0.33
2.96
1.45
0.99
0.41
0.44
0.32
0.54
0.59
0.58
0.79
1.84
1.41
1.87
1.44
0.03
0.11
0.52
0.07
0.05
0.58
0.25
0.12
0.05
0.04
0.05†
0.13
0.16¶
0.08¶
0.30†
0.18‡
0.27¶
0.46¶
0.43§
0.11
0.72
0.42
0.02
0.98
0.31
0.12
0.02
0.03
0.10
0.13
0.18
0.31
0.06
3.51
1.68
2.38
1.03
Blood
Fat
Ovary
Uterus
Bone
Uterus/Blood
Uterus/Muscle
Ovary/Blood
Ovary/Muscle
The radioactivity levels are expressed as SUV. Data represent the mean
SD (n = 5).
Statistical analysis (one-way analysis of variance with Dunnett's multiple comparison test) was performed between 16α-[18F]FES and each Cn-7α-[18F]FES derivative in organs of interest
(blood, fat, ovary, uterus, and bone), and for the tissue-to-blood ratio (uterus-to-blood and ovary-to-blood) and tissue-to-muscle ratio (uterus-to-muscle and ovary-to-muscle).
†
P b 0.05 compared to 16α-[18F]FES.
‡
P b 0.001 compared to 16α-[18F]FES.
¶
P b 0.0005 compared to 16α-[18F]FES.
§
P b 0.0001 compared to 16α-[18F]FES.
Original data expressed as % injected dose/g of 16α-[18F]FES (n = 4) and C3-7α-[18F]FES (n = 5) were taken from our previous report [15] and recalculated as SUV.
⁎
stereoselective side chain. An allylation reaction was performed with C-
6 ketone (1) using allyl iodide. 1H NMR demonstrated that the ratio of
7α to 7β was approximately 1:1 (8a). When the mixture was heated
under reflux using sodium methoxide in methanol, these diastereomers
isomerized to thermodynamically stable 7α (8) (Scheme 2). Compared
to 7β, 7α with an axially bound allyl group showed greater stability; this
is probably due to steric hindrance between the side chain and the CD
steroid ring. Various studies on the yield from an allylation reaction
were conducted regarding the type of base (potassium
hexamethyldisilazane, lithium diisopropylamide), type of reaction
solvent (tetrahydrofuran, 1,2-dimethoxyethane), and number of equiv-
alents of base and allyl iodide, but the yield did not improve further.
When an equivalent of base was used, the raw material remained
unreacted, and when the number of equivalents of base was increased,
diallyl bodies were generated (data not shown). After ozone oxidation
of the terminal alkene in compound 8, a reduction reaction was per-
formed using sodium borohydride to yield diol 9. The hydroxyl group
at the C-6 position was then reduced by a hydrogenation reaction to
yield 10 (Scheme 2). In the conversion reaction from 9 to 10, isomeriza-
tion occurred at the C-7 position, and the ratio of 7α to 7β was 6:1. We
Table 4
Comparison of the tissue distribution of the five Cn-7α-[18F]FES derivatives with 16α-[18F]FES in mature female mice at 30 min post-injection.
16α-[18F]FES
C1-7α-[18F]FES
C2-7α-[18F]FES
C3-7α-[18F]FES
C4-7α-[18F]FES
C6-7α-[18F]FES
⁎
Tissue
Heart
Lung
Liver
Pancreas
Spleen
Small Intestine
Adrenal
Kidney
Muscle
Brain
0.10
0.30
1.43
0.42
0.16
3.15
0.87
0.45
0.19
0.08
0.14
0.33
0.99
2.66
0.30
31.05
13.72
10.61
5.14
0.01
0.04
0.27
0.03
0.01
1.50
0.25
0.08
0.01
0.01
0.14
0.06
0.20
1.09
0.02
20.75
5.45
5.30
1.41
0.10
0.24
1.52
0.18
0.11
4.79
0.72
0.43
0.24
0.06
0.07
0.34
1.01
2.44
0.39
35.46
10.49
14.83
4.27
0.04
0.01
0.16
0.02
0.01
0.91
0.33
0.09
0.04
0.00
0.01
0.04
0.22
0.45
0.05
7.75
2.68
4.12
0.58
0.23
0.33
1.13
0.26
0.17
2.17
0.52
0.69
0.16
0.10
0.23
0.36
0.82
1.73
2.71
7.36
10.11
3.57
4.92
0.03
0.05
0.14
0.04
0.04
0.96
0.08
0.59
0.02
0.06
0.03
0.07
0.21
0.95
0.70†
3.31‡
4.36
0.50‡
0.80
0.27
0.36
1.21
0.67
0.35
1.16
0.72
0.50
0.25
0.11
0.23
0.51
0.49
1.29
2.47
5.58
5.50
2.14
2.06
0.02
0.03
0.06
0.21
0.25
0.81
0.62
0.08
0.04
0.02
0.02
0.14
0.20‡
0.46
0.75¶
1.94‡
2.55†
0.91‡
1.02¶
0.40
0.48
1.70
0.31
0.27
1.14
0.89
0.92
0.28
0.16
0.43
0.83
0.80
0.63
2.75
1.46
2.21
1.86
2.80
0.03
0.03
0.48
0.02
0.02
0.17
0.08
0.41
0.01
0.01
0.02§
0.19§
0.10
0.15¶
0.55†
0.36¶
0.57§
0.26¶
0.27‡
0.32
0.37
1.27
0.35
0.23
2.70
1.22
0.61
0.24
0.29
0.19
0.80
0.72
0.66
0.91
3.57
2.78
3.89
3.04
0.04
0.04
0.22
0.05
0.04
0.76
0.23
0.06
0.02
0.03
0.03
0.18§
0.07
0.11¶
0.36
1.02‡
0.62¶
0.76‡
0.43‡
Blood
Fat
Ovary
Uterus
Bone
Uterus/Blood
Uterus/Muscle
Ovary/Blood
Ovary/Muscle
The radioactivity levels are expressed as SUV. Data represent the mean
SD (n = 5).
Statistical analysis (one-way analysis of variance with Dunnett's multiple comparison test) was performed between 16α-[18F]FES and each Cn-7α-[18F]FES derivative in organs of interest
(blood, fat, ovary, uterus, and bone), and for the tissue-to-blood ratio (uterus-to-blood and ovary-to-blood) and tissue-to-muscle ratio (uterus-to-muscle and ovary-to-muscle).
†
P b 0.05 compared to 16α-[18F]FES.
‡
P b 0.001 compared to 16α-[18F]FES.
¶
P b 0.0005 compared to 16α-[18F]FES.
§
P b 0.0001 compared to 16α-[18F]FES.
Original data expressed as % injected dose/g of 16α-[18F]FES (n = 4) and C3-7α-[18F]FES (n = 5) were taken from our previous report [15] and recalculated as SUV.
⁎

518
M. Okamoto et al. / Nuclear Medicine and Biology 43 (2016) 512–519
using a tertiary alcohol as a solvent as previously described by Kim
et al. [20] (Table 1). [19F]Fluorination of nosylate 5b in tert-butyl alcohol
improved the nucleophilic substitution reaction (6, 14%). After column
purification of 6 and the following de-protection, 7 (93%) was produced
(Scheme 1). When all steps through de-protection were performed in a
single reaction, separating 7 and 29 by column chromatography was dif-
ficult. Therefore, isolation and purification were performed once after
fluorination.
4.4. Radiochemical synthesis
Fig. 2. Correlation between ERα affinity (IC50) and uterus standardized uptake value
(SUV) at 15 min (A) and 30 min (B) after injection of [18F]fluoroestradiol derivatives.
Open circles = C1-7α-FES (7); solid squares = C2-7α-FES (12); open squares = C3-
7α-FES; solid triangles = C4-7α-FES (22a); open triangles = C6-7α-FES (22b); solid
stars =16α-FES. Data represent the means (n = 3 for IC50 and n = 4–5 for SUV). The cor-
relations between IC50 and SUV were examined using Pearson's correlation coefficient.
[
18F]Fluorination of the tosylates was carried out using the 18F-
Kryptofix222-potassium carbonate system in acetonitrile at 100 °C for
15 min, followed by de-protection with 0.5 N hydrochloric acid in 50%
acetonitrile at 95 °C for 5 min to yield Cn-7α-[18F]FES derivatives
(Scheme 6).
The desired product was purified by reversed-phase HPLC. For C2-
7α-[18F]FES (24), preparative HPLC revealed two products, correspond-
ing to the 7α- and 7β-diastereomers. The total synthesis time was with-
believe that this was due to the production of methoxide ions in the sys-
tem by sodium borohydride, resulting in a partial enolate before the C-6
position was reduced and the stereo structure of the C-7 position was
broken down. However, because C2-7α-FES (eluted first) and C2-7β-
FES (eluted later) could be separated by chromatography, we did not
examine conditions to prevent isomerization.
Using 1,4-butanediol (13a) and 1,6-hexanediol (13b) as starting
materials, the hydroxyl group on one side was protected with a benzyl
group, and the other hydroxyl group was replaced with iodine by an
Appel reaction, to produce side chains 15a and 15b (Scheme 3). Then,
an alkylation reaction was performed on C-6 ketone (1) using 15a and
15b to produce 16a and 16b, respectively. Then, 19a and 19b were pro-
duced from 16a and 16b, respectively, in three steps. Subsequently,
through a hydrogenation reaction [19], the benzyl group of the side
chain was de-protected to yield alcohols 20a and 20b, respectively
(Scheme 3).
in
2 h from the end of bombardment. The decay-corrected
radiochemical yields of C1-7α-[18F]FES (23), C2-7α-[18F]FES (24), C4-
7α-[18F]FES (25), and C6-7α-[18F]FES (26) were 0.4%, 9.8%, 43%, and
44%, respectively. The radiochemical purity percentages of C1-7α-[18F]
FES (23), C2-7α-[18F]FES (24), C4-7α-[18F]FES (25), and C6-7α-[18F]
FES (26) were 98.1%, 99.2%, 100%, and 100%, respectively. Cn-7α-[18F]
FES derivatives showed inconspicuous UV absorbance. Calculating
their SA was difficult, which may indicate a higher SA than previously
reported for C3-7α-[18F]FES [15]. The highest calculated SA for C3-7α-
[
18F]FES was 55.3 GBq/μmol at the end of synthesis. Therefore,
N55.3 GBq/μmol, which is equivalent to b3.6 pmol per mouse head,
was used for the biodistribution study. The difference in SA between
C3-7α-[18F]FES and other 18F-compounds may be due to the difference
in the cyclotron used, the target system, irradiation conditions (current,
energy, time), etc.
Tosylation or nosylation of alcohols (4, 10, 20a, and 20b) produced
precursors for labeling (5a, 5b, 11, 21a, and 21b) (Schemes 1–3).
4.3. Examination of fluorination conditions
4.5. ER binding assay
[
19F]Fluorination of tosylates (11, 21a, and 21b) using the tetra-n-
The displacement curve of [3H]estradiol by C1- to C3-7α-FES was
parallel to the curve obtained with unlabeled estradiol, suggesting com-
petitive binding with recombinant human ERα (Fig. 1).
C1-7α-FES (7) had comparable affinity as estradiol and 16α-FES for
ERα (Table 2). The results also demonstrated that binding affinity to
ERα increased as the carbon number decreased. The results also demon-
strated that when the carbon number was ≥4, affinity dropped marked-
ly to about 1/100 of that of estradiol.
butyl ammonium fluoride (TBAF) system in acetonitrile at reflux follow-
ed by de-protection with acid successfully produced fluorinated stan-
dards (12, 22a, and 22b). In contrast, when TBAF was reacted with
tosylate 5a under reflux in the presence of acetonitrile, the elimination
reaction occurred preferentially, and the product that resulted was
compound 29 as shown with 1H NMR, in which olefin was isomerized
by subsequent de-protection (Scheme 5).
Fluorination reactions with 5b, in which the leaving group was con-
verted from tosyl to nosyl, with high leaving ability were examined
In our present experimental conditions, the RBA values of C3-7α-FES
(77%) and 16α-FES (121%) were different from those reported in
Fig. 3. Comparison of bone (A) and fat (B) uptake of [18F]fluoroestradiol derivatives at 30 min post-injection. Data represent the means SD (n = 4–5). X-axis; a = 16α-[18F]FES; b = C1-
7α-[18F]FES (23); c = C2-7α-[18F]FES (24); d = C3-7α-[18F]FES; e = C4-7α-[18F]FES (25); f = C6-7α-[18F]FES (26). Statistical analysis (one-way analysis of variance with Dunnett's mul-
18
tiple comparison test) was performed between 16α-[18F]FES and each Cn-7α-[18F]FES derivative. P b 0.05 compared to 16α-[ F]FES. P b 0.0005 compared to 16α-[18F]FES.
⁎
⁎⁎
P b 0.0001 compared to 16α-[18F]FES. The biodistribution data for 16α-[18F]FES (n = 4) and C3-7α-[18F]FES (n = 5) were taken from our previous report [15] and recalculated as SUV.
⁎⁎⁎

M. Okamoto et al. / Nuclear Medicine and Biology 43 (2016) 512–519
519
previous studies: 93% and 67%, respectively [15]. These differences may
be due to the different sources of recombinant ERα [21].
advice. This work was supported in part by a Grant-in Aid for Scientific
Research (B) 25293271 from the Japan Society for the Promotion of
Science (JSPS) and an Early Bird support project for young researchers
at Waseda Research Institute for Science and Engineering.
Some preformed pockets may be present within 4 Å at the Cn-7α-
position around the ligand [22,23]. Our data suggest that when the
side chain becomes longer, steric hindrance may occur. As a result, the
ligand does not easily fit into the ERα pocket, and the affinity to ERα
may be reduced. Furthermore, the estimated C-F alkyl chain length
(Å) of Cn-7α-FES (see Table 2) strongly supports these suggestions.
References
[1] Kiesewetter DO, Kilbourn MR, Landvatter SW, Heiman DF, Katzenellenbogen
JA, Welch MJ. Preparation of four fluorine-18-labeled estrogens and their se-
lective uptakes in target tissues of immature rats. J Nucl Med 1984;25:
1212–21.
4.6. Tissue distribution in normal mice
The tissue distribution data for C-7α-[18F]FES (23, 24, 25, and 26) in
female mice are summarized along with previously published results
(C3-7α-[18F]FES and 16α-[18F]FES) in Tables 3 and 4.
[2] Mintun MA, Welch MJ, Siegel BA, Mathias CJ, Brodack JW, McGuire AH, et al. Breast
cancer: PET imaging of estrogen receptors. Radiology 1988;169:45–8.
[3] McGuire AH, Dehdashti F, Siegel BA, Lyss AP, Brodack JW, Mathias CJ, et al. Positron
tomographic assessment of 16α-[18F]fluoro-17β-estradiol uptake in metastatic
breast carcinoma. J Nucl Med 1991;32:1526–31.
[4] Redeuilh G, Secco C, Baulieu E-E. The use of botinyl estradiol-avidin system for the
purification of “nontransformed” estrogen receptor by biohormonal affinity chroma-
tography. J Biol Chem 1985;260:3996–4002.
[5] French AN, Wilson SR, Welch MJ, Katzenellenbogen JA. A synthesis of 7α-substituted
estradiols: synthesis and biological evaluation of a 7α-pentyl-substituted BODIPY
fluorescent conjugate and a fluorine-18-labeled 7α-pentylestradiol analog. Steroids
1993;58:157–69.
[6] Bucourt R, Vignau M, Torelli V, Richard-Foy H, Geynet C, Secco-Millet C, et al. New
biospecific absorbents for the purification of estradiol receptor. J Biol Chem 1978;
253:8221–6.
The uterus uptake at 30 min post-injection gradually decreased as
the carbon number increased: SUV = 2.66 1.09 for C1-7α-[18F]FES
(23), 1.73 0.95 for C2-7α-[18F]FES (24), 1.29 0.46 for C3-7α-[18F]
FES, 0.63 0.15 for C4-7α-[18F]FES (25), and 0.66 0.11 for C6-7α-
[
18F]FES (26) (Table 4). The relationship between IC50 values obtained
from in vitro ERα affinity studies and SUVs obtained from the
biodistribution study is given in Fig. 2.
In vitro ERα affinity showed a significant negative correlation with
the uterus uptake at 15 (r = 0.9118, p = 0.0134) and 30 min (r =
0.9579, p = 0.0026). This negative correlation was higher at 30 min
than at 15 min. This phenomenon indicates that the tissue accumulation
in the receptor-rich uterus may be affected by the permeability of the li-
gands, which is likely to be related to the relative lipophilicities.
Because radioactivity accumulation in bone reduces image contrast
in positron emission tomography and is a factor that complicates
image analysis, an 18F label that does not readily undergo de-
fluorination is desirable. As shown in Fig. 3A, radioactivity accumulation
in bone at 30 min post-injection, which is thought to be due to in vivo
metabolic de-fluorination [24], exhibited a bell-shaped curve in which
C3-7α-[18F]FES was at the apex and C1-7α-[18F]FES (23) was at the
lowest point. This indicates that C1-7α-[18F]FES (23) had the highest
in vivo 18F label stability among the Cn-7α-[18F]FES derivatives.
As expected, a shorter carbon chain resulted in lower uptake in fat at
[7] Mühlenbruch B, Kirmeier F, Roth HJ. Synthesis and properties of fluorescent estradi-
ol derivatives. Arch Pharm (Weinheim) 1986;319:196–203.
[8] Dasilva JN, van Lier JE. Synthesis and structure-affinity of
a series of 7α-
undecylestradiol derivatives: a potential vector for therapy and imaging of
estrogen-receptor-positive cancers. J Med Chem 1990;33:430–4.
[9] Bowler J, Lilley TJ, Pittman JD, Wakeling AE. Novel steroidal pure antiestrogens. Ste-
roids 1989;54:71–99.
[10] Weathrill PJ, Wilson APM, Nicholson RI, Davis P, Wakeling AE. Interaction of the an-
tiestrogen ICI 164,384 with the oestrogen receptor. J Steroid Biochem 1988;30:
263–6.
[11] Wakeling AE, Bowler J. ICI 182,780, a new antiestrogen with clinical potential. J Ste-
roid Biochem 1992;43:173–7.
[12] Müller RE, Woitz HH. Post-coital contraceptive activity and estrogen receptor bind-
ing affinity of phenolic steroids. Endocrinology 1977;100:513–9.
[13] Peters RH, Crowe DF, Avery MA, Chong WKM, Tanabe M. 17-Desoxy estrogen ana-
logues. J Med Chem 1989;32:1642–52.
[14] Anstead GM, Carlson KE, Katzenellenbogen JA. The estradiol pharmacophore: ligand
structure-estrogen receptor binding affinity relationships and a model for the recep-
tor binding site. Steroids 1997;62:268–303.
[15] Okamoto M, Naka K, Kitagawa Y, Ishiwata K, Yoshimoto M, Shimizu I, et al. Syn-
thesis and evaluation of 7α-(3-[18F]fluoropropyl) estradiol. Nucl Med Biol
2015;42:590–7.
[16] Lim JL, Zheng L, Berridge MS, Tewson TJ. The use of 3-methoxymethyl-16β,17β-
epiestriol-O-cyclic sulfone as the precursor in the synthesis of F-18 16α-
fluoroestradiol. Nucl Med Biol 1996;23:911–5.
30 min post-injection: SUV = 0.34
0.04 for C1-7α-[18F]FES (23),
0.36 0.07 for C2-7α-[18F]FES (24), 0.51 0.14 for C3-7α-[18F]FES,
0.83 0.19 for C4-7α-[18F]FES (25), and 0.80 0.18 for C6-7α-[18F]
FES (26) (Fig. 3B).
[17] Kil HS, Cho HY, Lee SJ, Oh SJ, Chi DY. Alternative synthesis for the preparation of
16α-[18F]fluoroestradiol. J Label Compd Radiopharm 2013;56:619–26.
[18] Zhou D, Lin M, Yasui N, Al-Qahtani MH, Dence CS, Schwarz S, et al. Optimization
of the preparation of fluorine-18-labeled steroid receptor ligands 16alpha-
5. Conclusion
[
18F]fluoroestradiol (FES), [18F]fluoro furanyl norprogesterone (FFNP), and
In conclusion, we optimized the alkyl side chain length of Cn-7α-
[
18F]FES derivatives. The current study revealed the impact of the alkyl
16beta-[18F]fluoro-5alpha-dihydrotestosterone (FDHT) as radiopharmaceuti-
cals. J Label Compd Radiopharm 2014;57:371–7.
side chain length of Cn-7α-[18F]FES on the in vitro binding affinity,
biodistribution, and in vivo metabolism. C1-7α-[18F]FES (23) showed
uterus-to-blood and ovary-to-blood ratios that were comparable to
those of 16α-[18F]FES. Combined with in vivo stability as indicated by
low bone uptake and reasonably good affinity for ERα and its high
target-to-non-target ratio, we suggest that C1-7α-[18F]FES (23) is a
good candidate for a novel ER probe.
[19] Jiang XR, Sowell W, Zhu BT. Synthesis of 7α-substituted derivatives of 17β-estradiol.
Steroids 2006;71:334–42.
[20] Kim DW, Jeong HJ, Lim ST, Sohn MH, Katzenellenbogen JA, Chi DY. Facile nucleophil-
ic fluorination reactions using tert-alcohols as a reaction medium: significantly en-
hanced reactivity of alkali metal fluorides and improved selectivity. J Org Chem
2008;73:957–62.
[21] Freyberger A, Wilson V, Weimer M, Tan S, Tran H-S, Ahr H-J. Assessment of a ro-
bust model protocol with accelerated throughput for a human recombinant full
length estrogen receptor-α binding assay: protocol optimization and
intralaboratory assay performance as initial steps toward validation. Reprod
Toxicol 2010;30:50–9.
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.nucmedbio.2016.05.008.
[22] Katzenellenbogen JA, Muthyala R, Katzenellenbogen BS. The nature of the ligand-
binding pocket of estrogen receptor α and β: the search for subtype-selective li-
gands and implications for the prediction of estrogenic activity. Pure Appl Chem
2003;75:2397–403.
Acknowledgments
[23] Minutolo F, Macchia M, Katzenellenbogen BS, Katzenellenbogen JA. Estrogen recep-
tor β ligands: recent advances and biomedical applications. Med Res Rev 2011;31:
364–442.
We thank Mr. Kunpei Hayashi (SHI Accelerator Service) for his
technical support with the cyclotron operation and radiosynthesis and
Dr. Seijiro Hosokawa (Waseda University) and Dr. Kazuo Nagasawa
(Tokyo University of Agriculture and Technology) for their valuable
[24] Fiserova-Bergerova V. Changes of fluoride content in bone: an index of drug
defluorination in vivo. Anesthesiology 1973;38:345–51.