Iodinated and fluorinated steroid 2'-aryl-[3,2-c] pyrazoles as potential glucocorticoid receptor imaging agents

Article abstract of DOI:10.1016/S0039-128X(98)00069-5

We have synthesized several halogenated steroids as potential glucocorticoid receptor mediated imaging agents. These compounds are analogs of aryl-pyrazolo steroids, similar to the potent glucocorticoid, cortivazol. Compounds containing the halogens, iodine, bromine, and fluorine, as well as the E- and Z-iodovinyl side chain at the para position of 2'-phenyl- 11β,17,21-trihydroxy-16α-methyl-20-oxo-pregn-4-eno[3,2-c] pyrazole were prepared. They were tested as ligands for the glucocorticoid receptor by competition for the binding of [³H]dexamethasone and for glucocorticoid potency by the induction of alkaline phosphatase in HeLa cells. None of the iodinated steroids were good ligands for the glucocorticoid receptor or potent glucocorticoids. The bromo analog was only slightly better than the iodinated steroids as a ligand, and it had a potency in the HeLa cell assay about half that of dexamethasone. The fluoro analog showed good binding tO the glucocorticoid receptor and was a very potent glucocorticoid, approximately seven times that of dexamethasone. Consequently, it appears that the fluoro steroid, 2'-(4-fluorophenyl)-11β,17,21-trihydroxy-16α- methyl-20-pregn-4-eno[3,2-c] pyrazole, when labeled with ¹⁸F, would make an excellent glucocorticoid receptor-mediated imaging agent for positron emission tomography.

Full text of DOI:10.1016/S0039-128X(98)00069-5

Iodinated and fluorinated steroid 2؅-
aryl-[3,2-c] pyrazoles as potential
glucocorticoid receptor imaging agents
Robert M. Hoyte,* David C. Labaree,† Jean-Marie Fede,* Cathy Harris,* and
Richard B. Hochberg†
*Department of Chemistry, State University of New York, Old Westbury, New York, USA;
†Department of Obstetrics and Gynecology, Yale University School of Medicine, New Haven,
Connecticut, USA
We have synthesized several halogenated steroids as potential glucocorticoid receptor mediated imaging agents.
These compounds are analogs of aryl-pyrazolo steroids, similar to the potent glucocorticoid, cortivazol.
Compounds containing the halogens, iodine, bromine, and fluorine, as well as the E- and Z-iodovinyl side chain
at the para position of 2Ј-phenyl-11␤,17,21-trihydroxy-16␣-methyl-20-oxo-pregn-4-eno[3,2-c] pyrazole were
prepared. They were tested as ligands for the glucocorticoid receptor by competition for the binding of
[3H]dexamethasone and for glucocorticoid potency by the induction of alkaline phosphatase in HeLa cells. None
of the iodinated steroids were good ligands for the glucocorticoid receptor or potent glucocorticoids. The bromo
analog was only slightly better than the iodinated steroids as a ligand, and it had a potency in the HeLa cell assay
about half that of dexamethasone. The fluoro analog showed good binding to the glucocorticoid receptor and was
a very potent glucocorticoid, approximately seven times that of dexamethasone. Consequently, it appears that the
fluoro steroid, 2Ј-(4-fluorophenyl)-11␤,17,21-trihydroxy-16␣-methyl-20-oxo-pregn-4-eno[3,2-c] pyrazole, when
labeled with 18F, would make an excellent glucocorticoid receptor-mediated imaging agent for positron emission
tomography. (Steroids 63:595–602, 1998) © 1998 by Elsevier Science Inc.
Keywords: glucocorticoids; glucocorticoid imaging agents; glucocorticoid receptor; halogenated glucocorticoids; steroid
pyrazoles
nally by imaging would also be useful in the study of
receptor-mediated aspects of lymphocytic cancers, such as
leukemia. There have been several attempts to synthesize
18F-labeled steroids as glucocorticoid-receptor–directed im-
aging agents through positron emission tomography.6–9 Al-
though some were good ligands for the glucocorticoid re-
ceptor, none were suitably concentrated in target tissues in
vivo, probably because of metabolic inactivation and/or
inability to cross the blood–brain barrier.
Cortivazol (Figure 1) and similar arylpyrazolo steroids
have been extensively investigated as anti-inflammatory
agents10,11 and have been found to be powerful glucocorti-
coids, rivaling dexamethasone and cortisol.12,13 [3H]Corti-
vazol has been prepared and demonstrated to be a high-
affinity ligand for the glucocorticoid receptor.14 The ability
of these compounds to bind strongly to glucocorticoid re-
ceptor proteins suggests that they might be excellent models
for a ␥-emitting ligand for these receptors. The 2Ј-aryl
group of cortivazol and its analogs is a potential site for
labeling with the ␥-emitting isotopes of iodine and fluorine.
We have, therefore, prepared several of these potential
radioligands and have evaluated their properties as glu-
cocorticoids.
Introduction
The availability of ligands for the glucocorticoid receptor
that are labeled with appropriate isotopes will be useful for
mapping and imaging of glucocorticoid-receptor–rich re-
gions of the brain. Because the specific neurophysiological
mechanisms by which corticosteroids influence neuronal
function are not fully understood, such studies are of in-
creasing importance in defining and understanding the role
and function of these compounds in the brain.1 For example,
corticosteroids and their receptors in the hippocampal re-
gion of the brain, are postulated to be involved in the
management of stress, and defects in this system may have
important clinical implications for psychiatric disorders,
such as severe depression.2–4 There is also evidence that
corticosteroids are involved in the neurobiological effects of
aging.5 A glucocorticoid receptor ligand that can be de-
tected exter-
Address reprint requests to Dr. Richard B. Hochberg, Department of
Obstetrics and Gynecology, Yale University School of Medicine, New
Haven, CT 06510, USA. Tel.: 203-785-4001; E-mail: richard.hochberg@
yale.edu
Received June 18, 1998; accepted August 18, 1998.
Steroids 63:595–602, 1998
© 1998 by Elsevier Science Inc. All rights reserved.
655 Avenue of the Americas, New York, NY 10010
0039-128X/98/$19.00
PII S0039-128X(98)00069-5

Papers
bound [3H]steroid was determined by counting the radioactivity in
a liquid scintillation counter. The relative binding affinity (RBA)
was determined by a computerized curve-fitting program (Prism,
GraphPad Software, Inc.). Each steroid was tested in three exper-
iments performed in duplicate.
HeLa cell assay
The potency of the steroids as glucocorticoids was determined by
the induction of alkaline phosphatase (AlkP) in HeLa S3 cells
grown in 96-well microtiter plates. The cells, HeLacid, were a gift
from Dr. John Cidlowski. AlkP is induced in HeLa S3 cells by
glucocorticoids16 and the HeLacid cells are a clone that show a
very large and specific glucocorticoid induction of AlkP. They
were isolated on a flow cytometer by John Cidlowski. The induc-
tion of AlkP in these cells has been used in a microtiter plate
bioassay for glucocorticoids.17 We have confirmed that the induc-
tion of AlkP is specifically mediated by glucocorticoids and no
other steroid hormones (unpublished report). This induction by
glucocorticoids is inhibited by RU486. The HeLa cell assay is
performed as described,17 with some minor modifications. The
cells are grown in Jokliks Minimum Essential Medium with 3%
serum (a mixture of fetal calf serum and calf serum), antibiotics
and L-glutamine. After harvesting with trypsin, the cells (25,000
per well) are added to the microtiter plates in 100 ␮L of the growth
medium. (It is not necessary to use charcoal-stripped serum, since
the endogenous steroids in the diluted bovine serum are at a
concentration that is insufficient to have noticeable effect.) Test
steroids dissolved in ethanol, are diluted appropriately with the
medium so that the final concentration of ethanol in all wells,
including controls is 0.1%. Immediately after the cells, the glu-
cocorticoids are added to the wells in 50 ␮l of each solution (final
volume 150 ␮l). The cells are grown in a humidified atmosphere
containing 5% CO2 for 3 days. Afterward, the induction of AlkP is
measured using the chromogenic substrate p-nitrophenyl phos-
phate exactly as we have previously described for estrogen stim-
ulation of Ishikawa cells.18 The relative stimulatory activity (RSA)
of each steroid was tested three times in triplicate.
Figure 1 The structure of the potent glucocorticoid, cortivazol.
Experimental
Melting points were obtained in a Mel-Temp apparatus or in a
Koffler hot stage. Infrared spectra were recorded in potassium
bromide disks on a MIDAC Prospect FT-IR instrument. 1H NMR
spectra were obtained at 100 MHz or at 600 MHz with Bruker
WP100SY or AMX 600 FT instruments. EI mass spectra were
obtained using a Hewlett-Packard Model 5985A spectrometer at
70 eV with a direct insertion probe. FAB mass spectra were
obtained with a Kratos MS890 spectrometer. High-performance
liquid chromatography (HPLC) was performed on a Beckman
Model 334 gradient system equipped with Model 421 controller,
Altex CR-1A integrator-recorder, and Hitachi Model 100-10 vari-
able wavelength detector, or on a Beckman System Gold with
Model 168 detector using the following systems: H-1, CH3CN/
H2O (3:1); H-2, CH3OH/H2O (3:1), H-3, CH3OH/H2O (7:3).
These systems employed a 25 cm ϫ 1 cm Ultrasphere-ODS, RP
Column (Beckman) at a flow rate of 3 mL/min. Analytical HPLC
was performed on 25 cm ϫ 4.6 mm columns at a flow rate of 1
mL/min with the following systems: Ultrasphere-ODS column
(Beckman): H-4, CH3OH/H2O (3:1); H-5, CH3CN/H2O (3:2);
H-6, CH3OH/H2O (7:3); H-7, LiChrosorb Diol (EM Science)
CH2Cl2/isooctane/2-propanol (74:25:1); H-8, Partisil-10 (What-
man) CH2Cl2/2-propanol (9:1). Thin layer chromatography was
performed on silica-gel–coated glass plates (Merck F254 type)
using the following systems: T-1, toluene/ethyl acetate (3:1); T-2,
toluene/ethyl acetate (2:3). Ultraviolet (UV)-absorbing spots were
detected using a short wavelength (254 nm) lamp. Others were
detected by visualization with iodine vapors or with cerium
sulfate-sodium molybdate spray and heating. Flash chromatogra-
phy was performed on columns of silica gel-H (Merck) as de-
scribed below.
Chemical syntheses
2؅-(4-Bromophenyl)-11␤-hydroxy-16␣-methyl-17, 20:20 , 21-
bis-(methylenedioxy)-pregn-4-eno-[3,2-c]-pyrazole 3. A solu-
tion of 11␤-hydroxy-2-hydroxymethylene-16␣-methyl-17, 20 : 20,
21-bis- (methylenedioxy)-pregn-4-en-3-one 2 (1.96 g, 4.39 mmol),
prepared from 16␣-methylcortisol 1 as described,10 anhydrous
sodium acetate (0.74 g), p-bromophenylhydrazine⅐HCl (1.96 g ,
8.76 mmol) in absolute ethanol (120 mL) was stirred and heated at
reflux for 2 h. After cooling to room temperature, insoluble ma-
terial was removed by filtration through a coarse fritted funnel.
The filtrate was evaporated under vacuum and the residue was
purified by flash chromatography with toluene/ethyl acetate (3:1)
Glucocorticoid receptor assay
to give a product weighing 2.01 g (77% yield): UV: ␭
ϭ 263
The binding affinity for the glucocorticoid receptor was deter-
mined as previously described by competition for the binding of
[3H]dexamethasone to the glucocorticoid receptor in rat liver cy-
tosol.15 Rat liver cytosol (160 ␮L) prepared in TEDM buffer (50
mM Tris-HCl, 1 mM ethylenediaminetetraacetic acid, 1.5 mM
dithiothreitol, and 10 mM Na2MoO4, pH 7.5 was incubated with 2
nM [3H]dexamethasone added in 20 ␮L dimethylformamide-
TEDM (1/1) (DMF-TEDM) and the steroid competitor or control
also added in 20 ␮L DMF-TEDM, final volume 200 ␮L, overnight
in an ice bath. The free steroid was adsorbed with charcoal by
adding 200 ␮L of a slurry consisting of 1% Norite-A and 0.1%
Dextran T-70 in 50 mM Tris-HCl, pH 7.5. After thorough mixing,
the tubes were incubated for 10 min in an ice bath and centrifuged
for 10 min at 3,000 ϫ g. The supernatant was decanted and the
max
1
nm, ␧
ϭ 14,200; IR(KBr): 3400 cmϪ1, 1590, 1100, 700; H
max
NMR (CDCl3): ␦ 7.59 and 7.39 (AB pattern, 4H, aromatic-H),
7.43 (s, 1H, pyrazole C-H), 6.09 (s, 1H, H-4), 5.10 (m, 4H,
-OCH2O-), 4.46 (m,1H, H-11␣), 4.00 (s,2H, H-21), 1.32 (s, 3H,
H-19), 1.19 (s, 3H, H-18), 0.96 (d, 3H, J ϭ 7Hz, 16␣-CH3); MS:
m/z 596 and 598 (Mϩ); Anal. (C31H37O5N2Br) C: Calcd, 62.29;
Found, 62.10. H: Calcd, 6.25; Found, 6.35. N: Calcd, 4.69; Found,
4.55. Br: Calcd, 13.38; Found, 13.60.
2؅-(4-Bromophenyl)-11␤-methoxymethyl-16␣-methyl-17, 20 :
20, 21-bis-(methylenedioxy)-pregn-4-eno-[3,2-c]-pyrazole 4. To
an ice-cold solution of 0.66 g (1.1 mmol) of 3 in 16.5 mL (95
mmol) of diisopropylamine was added 16.5 mL (217 mmol) of
596 Steroids, 1998, vol. 63, November

Potential glucocorticoid receptor imaging agents: Hoyte et al.
chloromethylmethyl ether. The solution was stirred at 0°C for 1 h,
2؅-[4-(3-Methyl-3-hydroxy-1-butynyl)-phenyl]-11␤-hydroxy-
16␣-methyl-17, 20: 20, 2-bis-(methylenedioxy)-pregn-4-eno-
[3,2-c]-pyrazole 7. To a solution of pyrazole 3 (1.99 g , 3.33 mmol)
in diethylamine (28.0 mL, freshly distilled over CaO) under N2
was added sequentially, tetrakis(triphenylphosphine)-Palladium(0)
(TTP-Pd), (116 mg, 0.1 mmol, freshly opened), CuI (433 mg, 2.27
mmol) and 2-methyl-3-butyn-2-ol (MEBYNOL), (0.8 mL, 8.25
mmol). The mixture stirred for 72 h at room temperature. Addi-
tional quantities of the reagents, TTP-Pd (232 mg), CuI (866 mg),
(MEBYNOL) (1.6 mL) were added and the reaction was stirred for
an additional 6 days. The reaction mixture was diluted with meth-
ylene chloride (70 mL), washed sequentially with saturated aque-
ous NH4Cl (4ϫ, 36 mL), H2O (2ϫ, 18 mL), brine (36 mL), dried
over Na2SO4, filtered, and the solvent evaporated. Purification of
the residue by flash chromatography using toluene/ethyl acetate
(3:1) gave 0.8 g (40%) of 7: 1H NMR (CDCl3): ␦ 7.48(m, 4H,
Ar-H), 6.12(s, 1H, H-4), 5.09(m, 4H, -OCH2O-), 4.00(s, 2H,
H-21), 1.63(s, 6H, -C(OH)(CH3) 2), 1.33(s, 3H, H-19), 1.18(s, 3H,
H-18), 0.95(d, 3H, J ϭ 7Hz, 16␣-CH3); FAB-MS: m/z 601.
then at room temperature overnight. After cooling to 0°C saturated
sodium bicarbonate (20 mL) was slowly added to hydrolyze the
excess reagent. The solution was diluted with an additional 100
mL of the sodium bicarbonate solution and extracted with meth-
ylene chloride. The extracts were dried over sodium sulfate, fil-
tered, and evaporated to a dark oily residue. Flash chromatography
using toluene/ethyl acetate (6:1) gave 195 mg (28%) of 4 as an oil.
1H NMR (CDCl3): ␦ 7.57 and 7.36 (AB pattern, 4H, aromatic-H),
7.42 (s, 1H, pyrazole C-H), 6.07 (s, 1H, H-4), 5.2 - 4.8 (m, 6H,
-OCH2O- and 11␤-OCH2O-), 4.2 (m,1H, H-11␣), 3.98 (s,2H,
H-21), 3.46 (s,3H, -OCH3), 1.26 (s, 3H, H-19), 1.12 (s, 3H, H-18),
0.93 (d, 3H, J ϭ 7Hz, 16␣-CH3); MS: m/z 640 and 642 (Mϩ).
2؅-(4-Trimethylstannylphenyl)-11␤-methoxymethyl-16␣-methyl-
17, 20 : 20 , 21-bis-(methylenedioxy)-pregn-4-eno-[3,2-c]-pyrazole
5. To a stirred ice-cold solution of 4 (195 mg, 0.305 mmol) in 0.75 mL
of tetraethylene glycol dimethyl ether was added 5.5 mL of a 0.0965
M solution of trimethylstannyl sodium (0.53 mmol) in tetraethylene
glycol dimethyl ether (prepared and standardized according to previ-
ously described procedures19). The reaction mixture was allowed to
warm slowly to room temperature and stirred overnight, then cooled
in an ice bath and hydrolyzed with H2O (1.8 mL) followed by
saturated aqueous NH4Cl (0.6 mL). The product was extracted into 40
mL of benzene, dried over Na2SO4, filtered, and evaporated to give a
reddish-brown oily residue. Flash chromatography (15-cm column)
using toluene/ethyl acetate (5:1) gave 118 mg (53%) of 5 as an oil. A
portion was crystallized in methylene chloride/petroleum ether; mp
2؅-(4-Ethynylphenyl)-11␤-hydroxy-16␣-methyl-17, 20 : 20, 21-
bis- (methylenedioxy)-pregn-4-eno-[3,2-c]-pyrazole 8. A solu-
tion of 7 (0.8 g, 1.33 mmol) and NaOH (2.45 g) in n-butanol (96
mL) was heated at reflux for 1 h. After cooling to room temper-
ature, the reaction mixture was diluted with methylene chloride (1
L), washed sequentially with H2O (3ϫ, 250 mL), 20% aqueous
acetic acid (255 mL), and H2O (2ϫ, 250 mL). After drying over
sodium sulfate, the organic phase was filtered and the residual
1-butanol removed by azeotropic distillation with cyclohexane.
Purification of the residue by flash chromatography using toluene/
ethyl acetate (3:1) gave 273 mg of product, which was crystallized
with methanol to give 200 mg of 8 (28% yield): mp 149–151°C; IR
(KBr): 3427 cmϪ1 (OH, C-11), 3290, 2105(-C'C-H), 1090
1
137–140°C; H NMR (CDCl3): ␦ 7.63–7.34 (m, 5H, phenyl-H and
pyrazolyl-H), 6.15 (s, 1H, H-4), 5.1 (m, 4H, -OCH2O-), 4.77 (m, 2H,
11␤-OCH2O-), 4.22 (m, 1H, H-11␣), 4.00 (s, 2H, H-21), 3.47 (s, 3H,
-OCH3) 1.28 (s, 3H, H-19), 1.13 (s, 3H, H-18), 0.95 (d, 3H, J ϭ 7Hz,
16␣-CH3); 0.33 (s, 9H, -Sn(CH3)3); FAB-MS: m/z 727 (Mϩ1). Anal.
(C36H50O6N2Sn) C: Calcd, 59.59; Found, 59.75. H: Calcd, 6.94;
Found, 6.47. N: Calcd, 3.86; Found, 4.06.
1
(-OCH2O-); H NMR (CDCl3): ␦ 7.53(m, 4H, Ar-H), 6.12(s, 1H,
H-4), 5.09(m, 4H, -OCH2O-), 4.00(s, 2H, H-21), 3.14(s, 1H,
ethynyl-H), 1.31(s, 3H, H-19), 1.18(s, 3H, H-18), 0.95(d, 3H,
16␣-CH3). HRMS (FAB): Calculated for C33H39N2O5: 543.2859
(MHϩ), Found 543.2881 (MHϩ).
2؅-(4-Iodophenyl)-11␤,17,21-trihydroxy-16␣-methyl-20-oxo-pregn-
4-eno[3,2-c] pyrazole 6. A solution of pyrazole 5 (37 mg, 0.051
mmol) in chloroform (10 mL) was stirred at room temperature as
iodine (6.0 mL of a 3 mg/mL solution in chloroform) was added
dropwise. The reaction mixture was stirred for 30 min, washed
with aqueous 10% sodium bisulfite-1% potassium fluoride solu-
tion (15 mL), washed with H2O (3ϫ, 10 mL), dried over Na2SO4,
filtered, and evaporated to give an oily residue (38 mg). The
residue was treated with formic acid solution (5 mL, 60% aqueous)
and heated in a boiling water bath for 1 h, cooled to room
temperature, treated with 10 ␮L of concentrated H2SO4 and heated
for 20 min at 100°C. The H2SO4 was quenched with sodium
methoxide (0.375 mL of a 0.5M solution in methanol) and the
solution evaporated under vacuum. To saponify traces of the
21-formate (a side-product of the formic acid hydrolysis), the
residue was stirred with sodium methoxide (0.8 mL of 0.5M in
methanol) in methanol (2.5 mL) for 1 h at room temperature. The
reaction mixture was neutralized with glacial acetic acid (29 ␮L),
evaporated, diluted with methylene chloride (15 mL), washed with
H2O (3ϫ, 5 mL), dried over Na2SO4, filtered, and evaporated.
Purification of the residue (26.3 mg) by flash chromatography
using toluene/ethyl acetate (2:3) as eluent gave 6 as a colorless
film (20.5 mg, 67%), which was crystallized with methylene
chloride/petroleum; mp 162–165°C; UV: ␭max ϭ 265 nm, ␧max ϭ
18,800; 1H NMR (CDCl3): ␦ 7.8 and 7.4(AB pattern, 4H,
aromatic-H), 7.21(s, 1H, pyrazole C-H), 6.10(s, 1H, H-4), 4.7 -
4.3(m, 3H, H-11␣ and H-21), 1.3(s, 3H, H-19), 1.06 (s, 3H, H-18),
0.94(d, 3H, J ϭ 7 Hz, 16␣-CH3); FAB-MS: m/z 603 (Mϩ1). Anal.
(C29H35N2O4I): C: Calcd, 57.81; Found, 57.32. H: Calcd, 5.86;
Found, 5.72. N: Calcd, 4.65; Found, 4.38.
2؅-(4-E- and Z-Iodovinyl phenyl)-11␤,17,21-trihydroxy-16␣-
methyl-20 oxo-pregn-4-eno [3,2-c] pyrazole 10a, 10b. A solution
of 7 (74 mg, 0.136 mmol), Bu3SnH (70 ␮L, 0.26 mmol) AIBN (20
mg), in freshly distilled benzene (4.5 mL) was stirred for 5 min at
room temperature in a 16 ϫ 100 mm screw-capped tube under N2.
After cessation of bubbling, the tube was capped and heated at
80°C in a heating block for 5 h. Additional reagents were added
(70 L Bu3SnH, 20 mg AIBN) and heating was continued for
an additional 20 h. The reaction mixture (containing 2Ј-(4-E-
and Z-tributylstannylvinylphenyl)-11␤-hydroxy-16␣-methyl-17,
20:20,21-bis-(methylenedioxy)-pregn-4-eno-[3,2-c]-pyrazole, 9a
and 9b) was diluted with methylene chloride (15 mL) and treated
by the dropwise addition of a solution of iodine (90 mg , 0.354
mmol in 30 mL of CH2Cl2) with stirring in a 250 mL round bottom
flask. After stirring for 4 h at room temperature the solution was
washed with 60 mL each of the following solutions: 10% Na2S2O7,
H2O, 10% NaHSO3-1% KF, twice with H2O, dried over Na2SO4,
filtered and evaporated to give a residue (33 mg) containing a
mixture of 2Ј-(4-E- and-Z-iodovinylphenyl)-11b-hydroxy-16␣-
methyl-17,20:20,21-bis-(methylenedioxy)-pregn-4-eno
[3,2-c]
pyrazole: 1H NMR (CDCl3) 6.8 (d, 1H, E-CHA); 6.58 (d, 1H,
Z-CHA). A portion of the residue (18 mg, 0.03 mmol) was
dissolved in 60% formic acid (3 mL) and heated in a boiling water
bath for 2 h to remove the bis-methylenedioxy protecting group.
After cooling, the mixture was evaporated by repeated azeotro-
pic distillation with absolute ethanol to give 10 mg (60%) of a
2:1 mixture of 10a and 10b as indicated by HPLC analysis in
Steroids, 1998, vol. 63, November 597

Papers
Scheme 1 The synthetic scheme for 2Ј-(4-iodophenyl)-11␤,17,21-trihydroxy-16␣-methyl-20-oxo-pregn-4-eno[3,2-c] pyrazole 6.
Scheme 2 The synthetic scheme for E-and-Z-2Ј-(4-iodovinyl phenyl)-11␤,17,21-trihydroxy-16␣-methyl-20 oxo-pregn-4-eno [3,2-c]
pyrazole 10a, 10b.
598 Steroids, 1998, vol. 63, November

Potential glucocorticoid receptor imaging agents: Hoyte et al.
Figure 2 Competition for the binding of [3H]dexamethesone to the glucocorticoid receptor in rat liver cytosol. The conditions for the
assay are in the Experimental section.
system H-1 (10b elutes at 27 min and 10a at 33 min). Separation
of the isomers was accomplished by semipreparative HPLC in
system H-2 (10b elutes at 42 min and 10a at 49 min). These
materials were pure by analytical scale HPLC using system H-4.
Crystallization of each in acetone/hexane gave: E-isomer 10a: mp
154–155°C; UV ␭max (aq CH3OH) ϭ 291 nm; IR(KBr): 3420 cm1,
1707, 1604, 843; 1H NMR (CDCl3): 7.50 and 6.88(AB pattern,
2H, J ϭ 14.7 Hz, E-CH ϭ CH), 7.44(m, 5H, Ar-H and pyrazole-
H), 6.14(s, 1H, H-4), 4.55(m, 2H, H-21), 4.40(m, 1H, H-11),
1.31(s, 3H, H-19), 1.07(s, 3H, H-18), 0.94(d, 3H, 16␣-CH3);
HRMS (FAB): Calculated for C31H38N2O4I: 629.1876 (MHϩ),
Found 629.1882 (MHϩ). Z-isomer 10b: mp 142–143°C; UV:
Crystallization with methylene chloride/petroleum ether gave an
off-white powder; mp 154–156°C. This material was pure by
analytical HPLC using system H-5. UV: ␭max (aq CH3OH) ϭ 262
nm, ␧max ϭ16,700; 1H NMR(CDCl3): d 7.59 and 7.39(AB pattern,
4H, aromatic-H), 7.43(s, 1H, pyrazole C-H), 6.10(s, 1H, H-4),
4.54(m, 3H, H-11 and H- 21), 1.31(s, 3H, H-19), 1.07(s, 3H,
H-18), 0.94(d, 3H, 16␣-CH3); FAB-MS: m/z ϭ 555 and 557
(Mϩ16); HRMS (FAB): Calculated for C29H36N2O4Br (MHϩ):
555.1858, Found 555.1880.
2؅-(4-Fluorophenyl)-11␤,17,21-trihydroxy-16␣-methyl-20-oxo-
pregn-4-eno[3,2-c] pyrazole 12. Compound 1211 was prepared by
condensation of 2 with p-fluorophenylhydrazine as described
above for 3 followed by formic acid hydrolysis of the protecting
group as described for 10a, 10b. The resulting material was
purified by flash chromatography using toluene/ethylacetate (2:3)
and then by semipreparative HPLC with system H-2 where 12
migrates at 17.9 min. Crystallization with methylene chloride/
petroleum ether gave product with mp 143–145°C. The isolated
material was found to be pure by analytical HPLC in three differ-
ent systems, giving a single peak at 14.7 min in system H-5, at 21.4
min in system H-7, at 6.8 min in system H-8. UV ␭max(aq
␭
max(aq CH3OH) ϭ 283. IR(KBr): 3400 cmϪ1, 1706, 1606, 850;
1H NMR (CDCl3): 7.76 and 6.62(AB pattern, 2H, J ϭ 8.5 Hz,
Z-CH ϭ CH), 7.50(m, 4H, Ar-H), 7.32(s, 1H, pyrazole-H), 6.20(s,
1H, H-4), 4.56(m, 2H, H-21), 4.39(m, 1H, H-11␣), 1.31(s, 3H,
H-19), 1.07(s, 3H, H-18), 0.95(d, 3H, 16␣-CH3). HRMS (FAB):
Calculated for C31H38N2O4I: 629.1876 (MHϩ), Found 629.1860
(MHϩ).
2؅-(4-Bromophenyl)-11␤,17,21-trihydroxy-16␣-methyl-20-
oxo-pregn-4-eno[3,2-c] pyrazole 11. Compound 11 was prepared
by formic acid hydrolysis of the protecting group of 3 (102 mg) as
described above for compounds 10a, 10b. Flash chromatography
using toluene/ethyl acetate (2:3) gave of 11 (73.9 mg, 78%).
1
CH3OH)ϭ 262 nm.; IR(KBr): 3426 cmϪ1, 1708, 840; H NMR
(CDCl3): 7.50(m, 5H, Ar-H and Pyrazole H), 6.06(s, 1H, H-4),
4.56(m, 2H , H-21), 4.52(m, 1H, H-11), 1.30(s, 3H , H-19), 1.05(s,
Steroids, 1998, vol. 63, November 599

Papers
3 H, H-18), 0.96(d, 3 H, 16␣-CH3); HRMS: Calculated for
C29H35N2O4F: 494.2581, Found 494.2586.
was tested for its ability to compete for the binding of
[3H]dexamethasone for the glucocorticoid receptor. Nonra-
dioactive dexamethasone and cortisol were also tested for
comparison. A typical experiment is shown in Figure 2 (in
other experiments, higher concentrations of 6, up to 105M,
were used to obtain a better estimate of its RBA). The
results of all experiments are shown in Table 1. It can be
seen that the bromo analog 11, the iodo analog 6, and the
E-iodovinyl analog 10a, all had very low RBAs, on the
order of 1% or less. The fluoro analog 12, as well as the
phenylpyrazole parent 13 showed relatively good binding,
about 50% and 30% of dexamethasone, respectively. Both
were better ligands than cortisol.
The biological potency of all of the steroids was tested
by the induction of AlkP in the HeLa cell assay. A typical
experiment is shown in Figure 3, and the results of all
experiments calculated as RSA compared to dexametha-
sone, are in Table 1. All of the iodinated steroids, 6, 10a,
10b, had poor activity, about 5% of dexamethasone. The
bromo analog 11 had an RSA of 50%, which was consid-
erably better than cortisol (10%). Both the phenyl parent 13
and the fluoro analog 12 were very active with RSA over 7
times that of dexamethasone. The low activity of the iodin-
ated steroids is in agreement with their poor glucocorticoid-
receptor–binding activity. However, both 12 and 13 had
slightly lower affinity for the glucocorticoid receptor than
dexamethasone, but each had much higher biological activ-
2؅-Phenyl-␤11b,17,21-trihydroxy-16␣-methyl-20-oxo-pregn-4-
eno[3,2-c] pyrazole 13. Compound 1311 was prepared by conden-
sation of 2 and phenylhydrazine as described above for 3 followed
by formic acid hydrolysis of the protecting group as described for
10a, 10b. Semipreparative HPLC was performed with system H-3
where 13 migrated at 32 min. This material gave a single peak at
12.7 min on analytical HPLC in system H-5. Crystallization with
methylene chloride/petroleum ether gave product with mp 147–
148°C. UV: ␭max(aq CH3OH) ϭ 262 nm. IR(KBr) 3433 cmϪ1
,
1706, 763, 692. 1H NMR (CDCl3): 7.47(m, 6H, Ar-H and Pyrazole
H), 6.14(s, 1H, H-4), 4.5(m, 2H, H-21), 4.4(m, 1H, H-11), 1.31(s,
3H, H-19), 1.06(s, 3H, H-18), 0.94(d, 3H, J ϭ 7Hz,16␣-CH3);
FAB-MS: m/zϭ477 (Mϩ1); HRMS: Calculated for C29H36N2O4:
476.2675, Found 476.2675.
Results and discussion
The decision to use the arylpyrazole structure as the basis
for glucocorticoid receptor mediated imaging agents was
grounded on several factors, especially that many similar
compounds are very potent glucocorticoids10,11 and the
aromatic ring is an excellent target for substitution. Further-
more, since the aromatic pyrazole is not present in the
natural glucocorticoids it appeared unlikely that it would be
an important feature in the binding of the steroids to the
glucocorticoid receptor. Thus, we reasoned that bulky sub-
stituents on the phenyl ring should not appreciably affect
binding. We synthesized two types of halogenated steroids,
compounds, 6, 11 and 12 have direct bonds to the aromatic
ring. The other two compounds (isomers 10a and 10b,
Scheme 2) have an iodovinyl side chain in the para position
of the aromatic ring. The iodovinyl group has been shown to
possess the requisite chemical and metabolic stability for
useful radioligands in the androgen,20–22 progestin,23–25 and
estrogen26 families.27–29
Table 1 Glucorticoid actions of steroid phenylpyrazoles
The synthetic methods leading to substituted 2Ј-
arylpyrazolo steroids were originally described by Hirsch-
mann and coworkers10 and have been used by others in
similar or modified form.11 Using these methods we pre-
pared the protected 2Ј-(4-bromophenyl)- derivatives 3 and
4, which provided access to the target compounds 6
(Scheme 1) and 10a and 10b (Scheme 2). Briefly, the
4-trimethylstannylphenyl derivative 5 was prepared from 4
by treatment with trimethylstannyl sodium. Electrophilic
cleavage of the aryl stannane with I2 followed by deprotec-
tion gave the 2Ј-(4-iodophenyl)- derivative 6. The E- and
Z-4-iodovinylphenyl derivatives 10a and 10b were prepared
from 3 by a sequence using Heck coupling with 2-methyl-
3-butyne-2-ol followed by base cleavage30 to give the
4-ethynylphenyl derivative 8. Addition of tributyltin hydride to
the ethynyl group, followed by exchange with iodine and
deprotection gave the iodovinylphenyl compounds.
Glucocorticoid
receptor RBAa
HeLa cell
Steroid
AlkP RSAb
Dexamethasone
Cortisol
100
10 Ϯ 2
32 Ϯ 9
51 Ϯ 2
1 Ϯ 0.2
0.4 Ϯ 0.3
0.4 Ϯ 0.2
—
100
11 Ϯ 0.8c
703 Ϯ 115
760 Ϯ 159
50 Ϯ 39
6 Ϯ 2
R ϭ H
13
12
R ϭ fluoro
R ϭ bromo
R ϭ iodo
R ϭ (E)CH2 ϭ CH2I
R ϭ (Z)CH2 ϭ CH2I
11
6
10a
10b
5 Ϯ 1
3 Ϯ 1
aInhibition of binding of [3H]dexamethasone to glucocorticoid
receptor in rat liver cytosol. RBA, relative binding affinity com-
pared to dexamethasone.
bGlucocorticoid induction of AlkP in HeLacid cells. RSA, relative
stimulatory activity compared to dexamethasone.
cRBA Ϯ SD are from three separate experiments performed in
duplicate; RSA Ϯ SD from three separate experiments per-
formed in triplicate.
In recognition of the fact that the 2Ј-(4-fluorophenyl)
derivative 12 is a potential radioligand for PET if labeled
with fluorine-18, we also prepared the nonradioactive com-
pound, as previously described,11 and have evaluated its
affinity for the glucocorticoid receptor as well as its potency
in the HeLa cell assay.
The RSA and RBA are calculated using the curve-fitting pro-
gram, Prism.
—, not done; AlKP, alkaline phosphatase.
With the exception of 10b, each of the phenylpyrazols
600 Steroids, 1998, vol. 63, November

Potential glucocorticoid receptor imaging agents: Hoyte et al.
Figure 3 Glucocorticoid induction of AlkP in HeLacid cells. The conditions for the assay are in the Experimental section.
ity, the induction of AlkP in the HeLa cell. This is not
necessarily anomalous since biological activity (gene induc-
tion) is considerably more complex, it depends not only
upon receptor binding but also on the duration of the hor-
monal signal (metabolic half-life). Clearly, compounds that
are more resistant to metabolism would have a greater
potency. Other factors influencing transcription of genes,
including effects of the steroid ligand on the conformation
of the activated receptor may also play a role. Presumably,
12 and 13 have a greater catabolic resistance than dexa-
methasone. A previous study found that, compared to dexa-
methasone, cortivazol has a lower RBA for the glucocorti-
coid receptor and a much higher glucocorticoid activity, the
induction of tyrosine aminotransferase in rat hepatoma cells
grown in culture.31 It was suggested that this might be due to
an exceptionally slow rate of dissociation of the steroid-
receptor complex that would not be accurately measured in
competition assays. Regardless, the high RSA in the HeLa cell
assay is in excellent agreement with the potency of 12 and 13
compared to cortisol that was previously seen in in vivo glu-
cocorticoid anti-inflammatory assays performed in rats.11
The weak receptor binding and biological activity of the
para-substituted iodo steroids 6, 10a, 10b is somewhat
surprising, especially when compared to the enhanced ac-
tion of 12 and 13. It is apparent from the results that our
assumption that substitutions in the phenyl ring would be
transparent to the receptor was not correct. Clearly, the
arylpyrazole structure must come into close proximity with
the receptor binding site in such a way that unfavorable
steric or electrostatic interactions caused by the iodinated
substituents adversely affect properties of the steroid that
are important for their binding to the receptor. Electrostatic
effects due to these substituents may act directly upon the
receptor or indirectly by transmission into the steroid ring
system. The results of the receptor and HeLa cell assays
show that the fluoro analog 12 is a good ligand for the
glucocorticoid receptor and a powerful glucocorticoid.
While the specificity of its binding is yet unknown, 12
appears to have the requisite properties that makes it highly
likely that this compound labeled with 18F would be an
excellent glucocorticoid-receptor–mediated imaging agent.
Compound 12 can be prepared labeled with 18F by electro-
philic fluorodestanylation of the intermediate 5, which we
have used in the synthesis of iodo derivative 6 described
above.
Acknowledgments
We gratefully acknowledge Donna Raucci for her skilled
technical assistance and Linda O’Sullivan for her excellent
assistance in the preparation of this manuscript.
Steroids, 1998, vol. 63, November 601

Papers
This grant was supported by National Institutes of Health
17.
18.
Markiewicz L, Gurpide E (1998). In vitro bioassays for hormonal
activities. In: Bulletti C, De Ziegler D, Guller S, Levitz M (eds),The
uterus: Endometrium and myometrium. Annals of the New York
Academy of Sciences.The New York Academy of Sciences, New
York, pp. 95–102.
Littlefield BA, Gurpide E, Markiewicz L, McKinley B, Hochberg
RB (1990). A simple and sensitive microtiter plate estrogen bioas-
say based on stimulation of alkaline phosphatase in Ishikawa cells:
Estrogenic action of ⌬5 adrenal steroids. Endocrinology 127:2757–
2762.
Wursthorn KR, Kuivila HG (1977). Synthesis of substituted arylt-
rimethylstannanes by the reaction of trimethylstannylsodium with
aryl bromides. J Organometallic Chem 140:29–39.
Hoyte RM, Brown TJ, MacLusky NJ, Hochberg RB (1993). 7␣-
Methyl-17␣-(E-2Ј-[125I]iodovinyl)-19-nortestosterone: A new ra-
dioligand for the detection of androgen receptor. Steroids 58:13–23.
Hoyte RM, MacLusky NJ, Hochberg RB (1990). The synthesis and
testing of E-17␣-(2-iodovinyl)-5␣-dihydrotestosterone and Z-17␣-
(2-iodovinyl)-5␣-dihydrotestosterone as emitting ligands for the
androgen receptor. J Steroid Biochem 36:125–132.
grants GM08180 (to RMH) and CA37799 (to RBH).
References
1.
2.
3.
4.
De Kloet ER, Vreugdenhil E, Oitzl MS, Joe¨ls M (1998). Brain
corticosteroid receptor balance in health and disease. Endocr Rev
19:269–301.
Sapolsky RM, Krey LC, McEwen BS (1986). The neuroendocrinol-
ogy of stress and aging: The glucocorticoid cascade hypothesis.
Endocr Rev 7:284–298.
Holsboer F (1989). Psychiatric implications of altered limbic-
hypothalamic-pituitary-adrenocortical activity. Eur Arch Psychiatr
Neurol Sci 238:302–322.
Lisansky J, Peake GT, Strassman RJ, Qualls C, Meikle AW, Risch
SC, Fava GA, Zownir-Brazis M, Hochla P, Britton D (1989).
Augmented pituitary corticotropin response to a threshold dosage of
human corticotropin-releasing hormone in depressives pretreated
with metyrapone. Arch Gen Psychiatry 46:641–649.
Kerr DS, Campbell LW, Hao S-U, Lanfield PW (1989). Cortico-
steroid modulation of hippocampal potentials: Increased effect of
aging. Science 245:1505–1509.
Feliu AL, Rottenberg DA (1987). Synthesis and evaluation of
fluorine-18, 21-fluoroprednisone as a potential ligand for neuro-
PET studies. J Nucl Med 28:998–1005.
Pomper MG, Kochanny MJ, Thieme AM, Carlson KE, VanBrocklin
HF, Mathias CJ, Welch MJ, Katzenellenbogen JA (1992). Fluorine-
substituted corticosteroids: Synthesis and evaluation as potential
receptor-based imaging agents for positron emission tomography of
the brain. Int J Rad Appl Instrum[B] 19:461–480.
19.
20.
21.
5.
6.
7.
22.
Ali H, Rousseau J, Van Lier JE (1994). Synthesis of (17␣,20E/Z)
iodovinyl testosterone and 19-nortestosterone derivatives as poten-
tial radioligands for androgen and progesterone receptors. J Steroid
Biochem Mol Biol 49:15–29.
23.
24.
25.
Hochberg RB, Hoyte RM, Rosner W (1985). E-17␣-(2-[125I]iodovi-
nyl)-19-nortestosterone: The synthesis of a gamma-emitting ligand
for the progesterone receptor. Endocrinology 117:2550–2552.
Hoyte RM, Rosner W, Johnson IS, Zielinski J, Hochberg RB
(1985). Synthesis and evaluation of potential radioligands for the
progesterone receptor. J Med Chem 28:1695–1699.
8.
9.
DaSilva JN, Crouzel C, Stulzaft O, Khalili-Varasteh M, Hantraye P
(1992). Synthesis, tissue distribution in rats and PET studies in
baboon brain of no-carrier-added [18F]RU 52461: in vivo evaluation
as a brain glucocorticoid receptor radioligand. Int.J.Rad.Appl.In-
strum.[B] 19:167-173.
Visser GM, Krugers HJ, Luurtsema G, Van Waarde A, Elsinga PH,
deKloet ER, Groen MB (1995). Synthesis and organ distribution of
[18F]fluoro-Org 6141 in the rat: A potential flucocorticoid receptor
ligand for positron emission tomography. Nucl Med Biol 22:915–
920.
Grill HJ, Heubner A, Pollow K, Manz B, Schulze P-E, Hofmeister
H, Laurent H, Wiechert R, Elger W (1987). (Z)-17␤-Hydroxy-17␣-
(2-[125I]iodovinyl)-4-estren-3-one: A new specific gamma-emitting
ligand for determination of progesterone receptor. J Clin Chem Clin
Biochem 25:107–112.
26.
27.
Hanson RN, Seitz DE, Botarro JC (1982). E-17␣-[125I]iodovi-
nylestradiol: An estrogen-receptor-seeking radiopharmaceutical.
J Nucl Med 23:431-436.
Hanson RN, Franke LA (1984). Preparation and evaluation of
17␣-[125I]iodovinyl-11␤-methoxyestradiol as a highly selective ra-
dioligand for tissues containing estrogen receptors. J Nucl Med
25:998–1002.
Jagoda EM, Gibson RE, Goodgold H, Ferreira NL, Francis BE,
Reba RC, Rzeszotarski WJ, Eckelman WC (1984). [I-125] 17␣-
iodovinyl 11␤-methoxyestradiol: In vivo and In vitro properties of a
high-affinity estrogen-receptor radiopharmaceutical. J Nucl Med
25:472–477.
Ali H, Ghaffari RMA, Van Lier JE (1991). Synthesis, receptor
binding, and tissue distribution of 7␣- and 11␤-substituted (17␣,
20E)- and (17␣,20Z)-21-[125I]iodo-19-norpregna-1,3,5(10),20-
tetraene-3,1 7-diols. J Med Chem 34:854–860.
Melissaris AP, Litt MH (1992). Economical and convenient synthe-
sis of p-ethynylbenzoic acid and p-ethynylbenzoyl chloride. J Org
Chem 57:6998–6999.
Simons SS, Thompson EB, Johnson DF (1979). Anti-inflammatory
pyrazolo-steroids: Potent glucocorticoids containing bulky A-ring
substituents and no C3-carbonyl. Biochem Biophys Res Commun
86:793–800.
Coenen HH, Moerlein SM (1987). Regiospecific aromatic flu-
orodemetallation of group IVb metalloarenes using elemental fluo-
rine or acetyl hypofluorite. J Fluorine Chem 36:63–75.
10.
Hirschmann R, Buchschacher P, Steinberg NG, Fried JH, Ellis R,
Kent GJ, Tishler M (1964). Synthesis and structure of steroidal
pregn-4-eno-and 5␣-pregnano[3, 2, c]pyrazoles. A novel class of
potent anti-inflammatory steroids. J Am Chem Soc 86:1520–1527.
Hannah J, Kelly K, Patchett AA (1975). Substituted pyrazolo cor-
ticoids as topical antiinflammatory agents. J Med Chem 18:168–
172.
Dausse JP, Duval D, Meyer P, Gaignault JC, Marchandeau C,
Raynaud JP (1977). The relationship between glucocorticoid struc-
ture and effects upon thymocytes. Mol Pharmacol 13:948–955.
Harmon JM, Schmidt TJ, Thompson EB (1981). Deacylcortivazol
acts through glucocorticoid receptors. J Steroid Biochem 14:273–
279.
Schlechte JA, Simons SS, Jr., Lewis DA, Thompson EB (1985).
[3H]Cortivazol: A unique high affinity ligand for the glucocorticoid
receptor. Endocrinology 117:1355–1362.
Petrazzuoli M, Pahuja SL, Larner JM, Hochberg RB (1990). Bio-
logical activity of the fatty acid ester metabolites of corticoids.
Endocrinology 127:555–559.
28.
29.
11.
12.
13.
14.
15.
16.
30.
31.
32.
Littlefield BA, Cidlowski NB, Cidlowski JA (1980). Modulation of
glucocorticoid effects and steroid receptor binding in butyrate-
treated HeLa S3 cells. Arch Biochem Biophys 201:174–184.
602 Steroids, 1998, vol. 63, November