Effect of chirality at C-20 of methyl 11β,17α,20-trihydroxy-3-oxo-1,4-pregnadien-21-oate derivatives on antiinflammatory activity

Article abstract of DOI:10.1016/S0039-128X(01)00189-1

In an effort to determine the C-20 chirality effect on the antiinflammatory activity of 17β-glycolate esters, methyl 11β,17α,20-trihydroxy-3-oxo-1,4-pregnadien-21-oate and its 9α-fluoro analog, their acetonide and their carbonate derivatives were synthesized and evaluated. The agents were tested for their binding potency to the macrophage glucocorticoid receptor, and their effect on LPS-induced nitric oxide generation in RAW 264.7 cells. The acetonide derivatives showed the highest binding affinity while the triols and carbonates bound rather poorly to the receptors. With the exception of the triols, the α-isomer in each pair of the agents exhibited higher binding affinity to the receptor than its corresponding β-isomer, clearly indicating that C-20 chirality has a significant effect on antiinflammatory activity. In addition, the α-isomers of the acetonides showed substantially higher binding affinity than the parent compound, prednisolone. In contrast to the high binding activity exhibited by some of the acetonides, all of the agents showed weak inhibitory effect on NO generation. Metabolic inactivation during assessment of NO inhibition may play a role in the divergence noted between receptor affinity and the measured biologic activity resulting from the binding.

Full text of DOI:10.1016/S0039-128X(01)00189-1

Steroids 67 (2002) 353–359
Effect of chirality at C-20 of methyl 11␤,17␣,20-trihydroxy-3-oxo-1,4-
pregnadien-21-oate derivatives on antiinflammatory activity
Zhengqing You, Ann S. Heiman, Charles E. Hudson, Henry Joung Lee*
College of Pharmacy and Pharmaceutical Sciences, Florida A & M University, Tallahassee, FL 32307, USA
Received 13 June 2001; received in revised form 27 August 2001; accepted 30 August 2001
Abstract
In an effort to determine the C-20 chirality effect on the antiinflammatory activity of 17␤-glycolate esters, methyl 11␤,17␣,20-
trihydroxy-3-oxo-1,4-pregnadien-21-oate and its 9␣-fluoro analog, their acetonide and their carbonate derivatives were synthesized and
evaluated. The agents were tested for their binding potency to the macrophage glucocorticoid receptor, and their effect on LPS-induced nitric
oxide generation in RAW 264.7 cells. The acetonide derivatives showed the highest binding affinity while the triols and carbonates bound
rather poorly to the receptors. With the exception of the triols, the ␣-isomer in each pair of the agents exhibited higher binding affinity to
the receptor than its corresponding ␤-isomer, clearly indicating that C-20 chirality has a significant effect on antiinflammatory activity. In
addition, the ␣-isomers of the acetonides showed substantially higher binding affinity than the parent compound, prednisolone. In contrast
to the high binding activity exhibited by some of the acetonides, all of the agents showed weak inhibitory effect on NO generation. Metabolic
inactivation during assessment of NO inhibition may play a role in the divergence noted between receptor affinity and the measured biologic
activity resulting from the binding. © 2002 Elsevier Science Inc. All rights reserved.
Keywords: Anti-inflammation; Steroid methyl ester; Acetonide; Chirality; Antedrug
1
. Introduction
iment suggested that the ␣-acetonide might have a much
more complementary structure with the receptor, and thus a
much higher binding affinity than its ␤-epimer. A binding
study of these epimers, as well as similar acetonides derived
from other steroids, thus became highly desirable. It was
also very interesting to see if carbonates made from the
triols, which would render similar structural rigidity to the
acetonides, would show any ␣/␤ differentiation in activity.
A plan was therefore made to compare the antiinflammatory
activity of the ␣/␤ isomers of the acetonides, the carbonates
and the triols derived from both prednisolone and 9␣-flu-
oroprednisolone.
Our continuing search for new potent antiinflammatory
steroids devoid of systemic side effects has been based on
the antedrug concept. An antedrug is a compound that exerts
desirable local effects and rapidly biotransforms to an inac-
tive metabolite by a predictable enzymatic reaction upon
entry into the circulation. The chief strategy involved has
been incorporation of metabolically labile alkyl carboxy-
lates, at various positions of corticosteroids [1–3]. It had
been observed in the mid-1980s that chirality at C-20 of
methyl 11␤,17␣,20-trihydroxy-3-oxo-1,4-pregnadien-21-
oate, synthesized from prednisolone, has an influence on the
antiinflammatory activity [4]. The two triols showed mod-
erate antiinflammatory activities, with the 20␣-isomer [or
2. Experimental
2
0(R)-epimer], weaker than the 20␤-isomer [or 20(S)-
epimer]. However, the acetonides synthesized from these
triols were more active, and acetonide of the ␣-isomer was
substantially stronger than that of the ␤-isomer. The exper-
2.1. Synthesis
Prednisolone and isoflupredone acetate were purchased
from The Upjohn Company, Kalamazoo, MI. All other
reagents were ordered from Aldrich chemical company,
Milwaukee, WI. Solvents used, for reaction, extraction and
chromatography, were obtained from Fisher Scientific, Fair
*
Corresponding author. Tel.: ϩ1-850-599-3301; fax: ϩ1-850-599-
3
0
347.
039-128X/02/$ – see front matter © 2002 Elsevier Science Inc. All rights reserved.
PII: S0039-128X(01)00189-1

3
54
Z. You et al. / Steroids 67 (2002) 353–359
Lawn, NJ. NMR spectra were recorded on a Brucker HX-
70 spectrometer and the chemical shifts in parts per million
ppm) down field from tetramethylsilane as an internal stan-
g, 47 mmol) were added. The resultant mixture was heated
to 43°C for 2 days under a reflux condenser and then diluted
with 1:1 EtOAc/Hexanes (250 ml). The solution was
2
(
dard. Mass spectra were obtained on a Finnigan 4510
GCMS spectrometer, using positive chemical ionization.
Elemental analysis was carried out by Galbraith Laborato-
ries, Knoxville, TN. Melting points were determined on a
Thomas Hoover Capillary Melting Point Apparatus and
uncorrected.
washed with saturated aqueous NaHCO (20 ml) and water
3
(20 ml). The organic solution was then dried over sodium
sulfate and condensed. The residue was purified by flash
chromatography (3:2 EtOAc/Hexanes) to yield the 20(R)-
epimeric acetonide 5␣ (355 mg, 54%) as a slightly yellow
solid after vacuum drying. Part of the solid was recrystal-
lized from MeOH/water (3/1) to afford needles for melting
point measurement. 5␣: mp 197–199°C (lit [6].
2
.1.1. Methyl 11␤,17␣,20-trihydroxy-3-oxo-1,4-
pregnadien-21-oate (4␣,␤)
199–200°C). Proton NMR (CDCl ): 7.266 (1H, d, J ϭ 10.2
3
A mixture of prednisolone (12.5 g, 34.5 mmol) and
cupric acetate monohydrate (8.2 g, 40 mmol) in anhydrous
methanol (850 ml) was stirred under a drierite tube for 2
days. An aqueous EDTA disodium (6 g in 200 ml) was
added, and the resultant mixture condensed via rotary evap-
oration to a volume of about 100 ml. More water (200 ml)
was then added and the new aqueous suspension was fil-
tered. The solid was washed with water (50 ml), and then
mixed with 2% aqueous NaOH (150 ml), and stirred for 1 h.
The reaction mixture was acidified with 10% aqueous HCl
to pH 2, and then filtered to give a white solid, a mixture of
Hz, H-1), 6.261 (1H, dd, J ϭ 10.2, 1.8 Hz, H-2), 6.006 (1H,
br s, H-4), 4.567 (1H, s, H-20), 4.445 (1H, m, H-11), 3.774
(3H, s, MeO), 1.495 (3H, s, Me), 1.440 (3H, s, Me), 1.368
(3H, s, Me), 0.945 (3H, s, H-18). 20(S)-Epimeric acetonide
5␤: 48%, mp 187–189°C (lit [6]. 188–190°C). Proton NMR
(DMSO-d6): 7.362 (1H, d, J ϭ 10 Hz, H-1), 6.152 (1H, br
d, J ϭ 10 Hz, H-2), 5.907 (1H, br s, H-4), 4.512 (1H, s,
H-20), 4.408 (1H, m, H-11), 3.746 (3H, s, MeO), 1.463 (6H,
s, 2ϫ Me), 1.298 (3H, s, Me), 1.041 (3H, s, H-18).
2.1.3. Methyl 11␤,17␣,20-trihydroxy-3-oxo-1,4-
2
0(R) (3␣), and 20(S)-epimer (3␤). The filtrate was ex-
pregnadien-21-oate 17,20-carbonate (6␣,␤)
tracted with ethyl acetate (3 ϫ 150 ml) and the combined
organic solution condensed to yield more white solid. The
combined solid was dissolved in ethyl acetate (300 ml), and
treated with excess diazomethane in diethyl ether. The
methyl ester reaction was complete within a few minutes of
stirring and its crude products, the 20(R) (4␣) and 20(S)-
epimer (4␤), were obtained by condensation of the solution
to yield a slightly yellow solid. The solid was subjected to
repeated recrystallizations from 3:1 EtOAc/methanol to af-
ford 4␣ as a white solid (6.9 g, 51%). The combined mother
liquors were condensed and then chromatographed through
silica gel with 3:2 chloroform/acetone solvent system to
yield 4␤ as a while solid (1.4 g, 10%). A mixture of roughly
The triol (4␣ or 4␤, 140 mg, 0.35 mmol) was mixed with
carbonyldiimidazole (300 mg, 1.85 mmol) in ethyl acetate
(30 ml). The mixture was heated to reflux for 2 h and then
diluted with more EtOAc (100 ml). The solution was then
washed with 3% aqueous HCl (2 ϫ 25 ml) and water (25
ml). The solution was dried over sodium sulfate and then
condensed to remove the solvent. The residue was triturated
with EtOAc to give the 20(R)-epimeric carbonate 6␣ (142
mg, 94%) as a white solid: mp 266–268°C. Proton NMR
(CDCl ): 7.189 (1H, d, J ϭ 10.1 Hz, H-1), 6.189 (1H, dd,
3
J ϭ 10.1, 1.8 Hz, H-2), 5.952 (1H, dd, J ϭ 1.8, 1.3 Hz, H-4),
4.801 (1H, s, H-20), 4.357 (1H, m, H-11), 3.805 (3H, s,
1
3
MeO), 1.390 (3H, s, H-19), 1.116 (3H, s, H-18). C NMR
1
:1 ratio of the epimeric isomers (1.5 g, 11%) remained
after the separation procedure. 4␣: mp 265–267°C (lit [5].
54–255°C). Proton NMR (CDCl ): 7.197 (1H, d, J ϭ 10.1
(75.5 MHz, CDCl ): 186.38, 169.55, 166.96, 155.91,
3
152.97, 127.79, 122.49, 96.09, 78.83, 68.86, 54.70, 52.80,
51.90, 44.59, 43.79, 38.36, 36.99, 33.39, 31.71, 30.67,
22.64, 21.02, 15.47. Anal. Calc. for C H O : C 66.33%, H
2
3
Hz, H-1), 6.231 (1H, dd, J ϭ 10.1, 1.9 Hz, H-2), 5.991 (1H,
dd, J ϭ 1.9, 1.3 Hz, H-4), 4.406 (1H, m, H-11), 4.1675 (1H,
d, J ϭ 9.6 Hz, H-20), 3.787 (3H, s, OMe), 3.086 (1H, d, J ϭ
2
3 28 7
6.78%. Found: C 66.21%, H 6.92%. 20(S)-Epimeric car-
bonate 6␤ (92%): mp 257.5–260°C. Proton NMR (CDCl3):
7.203 (1H, d, J ϭ 10.1 Hz, H-1), 6.240 (1H, dd, J ϭ 10.1,
1.9 Hz, H-2), 5.998 (1H, br s, H-4), 4.925 (1H, s, H-20),
4.502 (1H, m, H-11), 3.822 (3H, s, MeO), 1.434 (3H, s,
9
.6 Hz, OH-20), 1.426 (3H, s, H-19), 1.111 (3H, s, H-18).
4␤: mp 174–176°C (lit [5]. 171–173°C). Proton NMR
(CDCl ): 7.244 (1H, d, J ϭ 10.1 Hz, H-1), 6.240 (1H, dd,
3
J ϭ 10.1, 1.9 Hz, H-2), 5.987 (1H, br s, H-4), 4.421 (1H, m,
H-11), 4.339 (1H, d, J ϭ 5.8 Hz, H-20), 3.778 (3H, s, OMe),
H-19), 1.130 (3H, s, H-18). C-13 NMR (CDCl , 68 MHz):
3
186.41, 169.49, 167.49, 155.80, 153.03, 127.93, 122.58,
95.28, 75.35, 69.20, 54.79, 52.97, 50.22, 46.87, 43.85,
38.38, 33.44, 31.77, 31.37, 30.36, 23.27, 21.04, 17.04. Anal.
Calc. for C H O : C 66.33%, H 6.78%. Found: C 66.19%,
3
.162 (1H, d, J ϭ 5.8 Hz, OH-20), 1.435 (3H, s, H-19),
1
.141 (3H, s, H-18).
2
3 28 7
2
.1.2. Methyl 11␤,17␣,20-trihydroxy-3-oxo-1,4-
H 6.98%.
pregnadien-21-oate 17,20-acetonide (5␣,␤)
The fluorinated derivatives of triols, acetonides and car-
bonates were synthesized by the same procedures as their
non-fluorinated counterparts and their data are as follows.
Methyl 11␤,17␣,20␣-trihydroxy-9␣-fluoro-3-oxo-1,4-
The triol (4␣ or 4␤, 0.6 g, 1.5 mmol) dissolved in
anhydrous DMF (3 ml), and p-toluenesulfonic acid mono-
hydrate (100 mg, 0.5 mmol) and 2,2-dimethoxypropane (5

Z. You et al. / Steroids 67 (2002) 353–359
355
pregnadien-21-oate (F-4a): white solid, mp 280–282°C (lit
7]. 279–281°C). Proton NMR (DMSO-d6): 7.266 (1H, d,
J ϭ 10.2 Hz, H-1), 6.194 (1H, dd, J ϭ 10.2, 1.8 Hz, H-2),
.987 (1H, br s, H-4), 5.214 (1H, m, H-11), 5.139 (1H, d,
J ϭ 7.8 Hz, H-20), 4.147 (1H, s, OH-17), 4.052 (1H, d, J ϭ
J ϭ 19 Hz), 30.05, 26.96, 23.01, 22.93, 21.77, 14.71. Anal.
[
Calc. for C H FO : C 63.58%, H 6.26%, F 4.37%. Found:
2
3
27
7
C 63.42%, H 6.54%, F 4.31%.
Methyl 11␤,17␣,20␤-trihydroxy-9␣-fluoro-3-oxo-1,4-
pregnadien-21-oate 17,20-carbonate (F-6␤): mp 252.5–
5
7
1
1
.8 Hz, OH-20), 3.610 (3H, s, MeO), 1.495 (3H, s, H-19),
.036 (3H, s, H-18). C-13 NMR (DMSO-d6, 75.47 MHz):
85.28, 172.80, 167.22, 152.90, 128.87, 124.03, 101.26 (d,
254.5°C. Proton NMR (CDCl ): 7.187 (1H, d, J ϭ 10.2 Hz,
3
H-1), 6.301 (1H, dd, J ϭ 10.2, 1.8 Hz, H-2), 6.098 (1H, br
s, H-4), 4.944 (1H, s, H-20), 4.390 (1H, m, H-11), 3.824
(3H, s, MeO), 1.536 (3H, s, H-19), 1.125 (3H, s, H-18).
J ϭ 174 Hz, C-9), 84.42, 73.64, 70.60 (d, J ϭ 37 Hz, C-11),
5
3
1
1
1.02, 47.92 (d, J ϭ 24 Hz), 45.08, 44.01, 34.84, 34.60,
3.41 (d, J ϭ 18 Hz), 30.30, 27.35, 23.04, 22.98, 22.41,
6.21. Methyl 11␤,17␣,20␤-trihydroxy-9␣-fluoro-3-oxo-
,4-pregnadien-21-oate (F-4b): white solid, mp 257–260°C
C-13 NMR (CDCl , 75.47 MHz): 186.56, 167.41, 165.97,
3
153.03, 152.04, 129.74, 125.12, 99.84 (d, J ϭ 177 Hz, C-9),
95.17, 75.28, 71.14 (d, J ϭ 38 Hz, C-11), 52.97, 48.09 (d,
J ϭ 22 Hz), 46.41, 43.81, 34.80, 34.37 (d, J ϭ 20 Hz),
30.82, 30.44, 26.98, 23.01, 22.96, 16.59. Anal. Calc. for
C H FO : C 63.58%, H 6.26%, F 4.37%. Found: C
(lit [8]. 259–261°C). Proton NMR: (CDCl ): 7.196 (1H, d,
3
J ϭ 10.2 Hz, H-1), 6.306 (1H, dd, J ϭ 10.2, 1.8 Hz, H-2),
.092 (1H, br s, H-4), 4.30 (2H, m, H-11 & H-20), 3.774
3H, s, MeO), 3.308 (1H, br s, OH), 1.526 (3H, s, H-19),
2
3
27
7
6
(
63.33%, H 6.44%, F 4.72%.
1
.137 (3H, s, H-18).
2.2. Bioassays
Methyl 11␤,17␣,20␣-trihydroxy-9␣-fluoro-3-oxo-1,4-
pregnadien-21-oate 17,20-acetonide (F-5␣): mp 278–
81°C. Proton NMR (CDCl ): 7.203 (1H, d, J ϭ 10.2 Hz,
H-1), 6.326 (1H, dd, J ϭ 10.2, 1.8 Hz, H-2), 6.110 (1H, br
2.2.1. Cell culture
2
RAW 264.7 murine macrophages were purchased from
American Type Tissue Collection (Rockville, MD) (ATCC
TIB-71) and grown in monolayer culture in 100 mm Corn-
ing tissue culture dishes containing DMEM supplemented
with 10% newborn calf serum (NCS) and pen/strep. At
confluence, cells were passaged by scraping, washing in
sterile Hank’s Balanced Salts solution and seeding at 0.25 ϫ
3
s, H-4), 4.573 (1H, s, H-20), 4.358 (1H, m, H-11), 3.780
(
(
3H, s, MeO), 1.532 (3H, s, Me), 1.498 (3H, s, Me), 1.371
3H, s, H-19), 0.938 (3H, s, H-18). C-13 NMR (CDCl3,
75.47 MHz): 186.54, 170.18, 166.14, 152.28, 129.68,
125.04, 108.91, 100.12 (d, J ϭ 177 Hz, C-9), 93.84, 78.18,
71.84 (d, J ϭ 38 Hz, C-11), 52.05, 48.18 (d, J ϭ 23 Hz),
45.42, 44.85, 36.59, 36.46, 34.03 (d, J ϭ 19 Hz), 31.05,
27.09, 26.54, 25.85, 23.01, 22.94, 22.73, 16.34. Anal. Calc.
6
10 viable cells/ml. Viability was accessed by trypan blue
o
exclusion. Cells were incubated at 35 C in an environment
of 5% CO and 95% relative humidity.
2
for C H FO : C 66.95%, H 7.42%, F 4.23%. Found: C
2
5
33
6
6
6.81%, H 7.61%, F 3.92%.
2.2.2. Whole cell receptor binding assays
Methyl 11␤,17␣,20␤-trihydroxy-9␣-fluoro-3-oxo-1,4-
Whole cell glucocorticoid receptor binding assays were
carried out using RAW 264.7 cells seeded at 0.2 ϫ 10
6
pregnadien-21-oate 17,20-acetonide (F-5␤): mp 222–
24.5°C. Proton NMR (CDCl ): 7.243 (1H, d, J ϭ 10.2 Hz,
2
cells/well in 96-well plates and allowed to attach and grow
for 24 h in the medium described above. Cell monolayers
were washed with cold HBSS then incubated for 2 h in
Ham’s F-10 with 0.1% lactalbumin hydrolysate (200 ␮l/
well). Medium was aspirated and fresh culture medium
containing 5 nM dexamethasone and various concentrations
(1–1000 nM) of unlabeled competing steroids were added
(250 ␮l/well). Nonspecific binding was determined in the
3
H-1), 6.340 (1H, dd, J ϭ 10.2, 1.6 Hz, H-2), 6.127 (1H, br
s, H-4), 4.563 (1H, s, H-20), 4.430 (1H, m, H-11), 3.786
3H, s, MeO), 1.570 (3H, s, Me), 1.520 (3H, s, Me), 1.348
3H, s, H-19), 1.078 (3H, s, H-18). C-13 NMR (CDCl3,
5.47 MHz): 186.66, 171.63, 166.43, 152.30, 129.73,
25.03, 111.40, 100.50 (d, J ϭ 180 Hz, C-9), 94.56, 77.62,
2.15 (d, J ϭ 38 Hz, C-11), 52.14, 48.31 (d, J ϭ 23 Hz),
5.76, 44.34, 36.19, 34.46 (d, J ϭ 20 Hz), 33.47, 31.05,
7.75, 27.38, 27.15, 23.50, 22.99, 22.90, 17.33. Anal. Calc.
(
(
7
1
7
4
2
3
presence of 1000-fold excess of [ H]dexamethasone. Cells
were then incubated at 35°C, 5% CO for 3 h, then washed
2
for C H FO : C 66.95%, H 7.42%, F 4.23%. Found: C
five times with 250 ␮l/well cold HBSS with a 2 min soak
between washes using a microplate washer. Following as-
piration of the final wash, cells were solubilized by adding
80 ␮l 0.1 N NaOH/well. Contents of 6 wells were pooled
for each data point and a 0.2 ml aliquot counted by scintil-
lation spectroscopy. Total cellular protein was quantitated
by the Lowry method using bovine serum albumin (BSA) as
standard.
2
5
33
6
6
6.86%, H 7.56%, F 3.88%.
Methyl 11␤,17␣,20␣-trihydroxy-9␣-fluoro-3-oxo-1,4-
pregnadien-21-oate 17,20-carbonate (F-6␣): mp 277–
79°C. Proton NMR (DMSO-d6): 7.240 (1H, d, J ϭ 10.2
2
Hz, H-1), 6.214 (1H, dd, J ϭ 10.2, 1.6 Hz, H-2), 6.014 (1H,
br s, H-4), 5.505 (1H, m, H-11), 5.419 (1H, s, H-20), 3.813
(
3H, s, MeO), 1.479 (3H, s, H-19), 1.076 (3H, s, H-18).
C-13 NMR (75.47 MHz, DMSO-d6): 185.16, 167.34,
66.57, 152.90, 152.41, 129.05, 124.22, 100.62 (d, J ϭ 178
Hz, C-9), 96.32, 77.75, 69.67 (d, J ϭ 37 Hz, C-11), 52.87,
7.64 (d, J ϭ 22 Hz), 45.42, 43.51, 35.92, 34.34, 32.88 (d,
1
2.2.3. Generation and measurement of NO
RAW 264.7 macrophages were seeded into 24 well clus-
6
4
ter plates at densities of 0.8 ϫ 10 cells/well and allowed to

3
56
Z. You et al. / Steroids 67 (2002) 353–359
attach and grow overnight. Culture medium was aspirated
and replaced with DMEM without phenol red and supple-
mented with 2.5% NCS to which indicated concentrations
of steroids were added in triplicate. Following a 3 h incu-
bation period, cells were stimulated to generate NO by
addition of 30 ng/ml LPS (Escherichia coli serotype 0111:
B4). Following incubation for the indicated times, the
amount of stable nitrite, the end product of NO generation
by activated cells, was determined by a modification of the
Griess reaction. Briefly, 50 ␮l of culture supernatants from
control or stimulated macrophages were transferred to 96-
well microtiter plates. Supernatants were mixed with 50 ␮l
of 1% sulfanilamide in 5% phosphoric acid, incubated for
Fig. 1. 20␣ and 20␤ epimer
epimers in each pair. While H-18 of the ␣-isomer showed a
small NOE response (ϳ 1%), that of the ␤-isomer gave a
significantly stronger signal enhancement of about 4%.
1
0 min at room temperature, protected from light, followed
by 50 ␮l of 0.1% aqueous N-1-naphthethylenediamine di-
hydrochloride for another light-protected 10 min incuba-
tion. Absorbances at 540 nm were determined on a Bio-Tek
microtiter plate reader. Nitrite accumulation (nmol/50 ␮l)
was determined by extrapolation from a sodium nitrite stan-
dard curve. Data from concentration-response time course
experiments were expressed as nmol nitrite/10 cells, and
data from steroid treatment experiments were expressed as
percent untreated controls [9].
3.2. Screening bioassays
Both receptor binding in a whole cell glucocorticoid
receptor binding assay and inhibition of nitric oxide (NO)
production in lipopolysaccharide-stimulated RAW 264.7
murine macrophages were used to assess anti-inflammatory
activity of the agents. The rationale for these tests is based
on reports that isoforms of nitric oxide synthase (NOS) are
found in endothelial cells, macrophages and other cells
involved in inflammatory responses. NO generated by the
inducible NOS isoform is considered an important modula-
tor of inflammatory responses and macrophages may be one
of the effector cells in this process. Following induction of
inducible NOS by treatment with bacterial products and/or
proinflammatory cytokines, macrophages generate high sus-
tained levels of NO which can be inhibited by pretreatment
of cells with glucocorticoids [13–15]. Recently, our labora-
tories reported that new 9␣-fluoroprednisolone anti-inflam-
matory antedrugs showed high affinity for macrophage glu-
cocorticoid receptors and potent inhibition of NO
generation [9]. In contrast, results of another study demon-
strated that anti-inflammatory steroids with a spiro ring
structure exhibited low affinity for macrophage glucocorti-
coid receptors and no significant inhibition of NO genera-
tion [16].
Results of macrophage glucocorticoid receptor binding
of the present series of agents are depicted in Table 1 which
also includes the previously reported ear edema inhibition
activity [4]. All members of the two pairs of triols, 4␣/4␤
and F-4␣/F-4␤, showed much weaker receptor affinity than
prednisolone. These results are consistent with their inhib-
itory activity in the rodent croton oil ear edema bioassay.
Similarly, the two carbonate pairs, 6␣/6␤ and F-6␣/F-6␤
showed weak binding affinity with the ␣-isomer of each pair
exhibiting slightly higher affinity than the ␤-derivative. Of
the compounds tested, the acetonides, 5␣/5␤ and F-5␣/
6
3
. Results and discussions
3
.1. Synthesis
The literature procedure to prepare methyl 11␤,17␣,20-
trihydroxy-3-oxo-1,4-pregnadien-21-oate was time con-
suming and gave very low yield [5,10]. Thus a modified
procedure based on the research of Lewbard and Maddox
[11,12] was developed: cupric oxidation of prednisolone in
methanol to the 21-hemiacetal (2 & F-2), followed by
NaOH initiated Cannizzaro rearrangement to 20-hydroxy-
2
1-oic acid (3␣,␤ & F-3␣,␤), and methyl ester formation
(4␣,␤ & F-4␣,␤) by treatment with diazomethane. Separa-
tion of the ester epimers was achieved by fractional recrys-
tallization, rather than the HPLC method as described in the
literature [5,10]. The 20␣-isomers (4␣, F-4␣) were ob-
tained pure after the recrystallizations, while the 20␤-iso-
mers (4␤, F-4␤) needed further purification through silica
gel column chromatography. A combined yield of the two
methyl ester epimers of about 70% was obtained from the
starting material, with an ␣/␤ ratio of approximately 3.5/1.
Preparation of the acetonides and carbonates went
smoothly. Thus treatment of the triols with 2,2-dime-
thoxypropane under the catalysis of p-toluenesulphonic acid
in DMF gave the desired acetonide derivatives (5␣,␤ &
F-5␣,␤) in moderate yields. And reaction of the triols with
carbonyldiimidazole in ethyl acetate afforded the carbonates
in high yields (6␣,␤ & F-6␣,␤).
3
F-5␤ competitively displaced [ H]dexamethasone from the
macrophage glucocorticoid receptors with each ␣-isomer
displaying much greater affinity than its ␤-counterpart.
These results are consistent with the in vivo inhibition of
croton oil induced ear edema. It is noteworthy that the
The stereochemistry at C-20 was confirmed by melting
point comparison with literature data, and NOE experiment
with irradiation at the proton on C-20 of the carbonate

Z. You et al. / Steroids 67 (2002) 353–359
357
9
-fluorine substitution did not significantly improve glu-
cocorticoid receptor binding as exhibited by IC50 values of
5 nM and 33 nM for 5␣ and F-5␣, respectively, and that
occupancy was assessed as inhibition of NO generation in
LPS-stimulated macrophages. Results are shown in Table 2.
With the exception of F-5␣, all agents were very weak
inhibitors of NO which were assessed as the accumulation
of nitrites in cell culture supernatants. IC50 calculated from
concentration-response curves were 342 nM, 186 nM and
3
affinities of the acetonide derivatives equal that of 34 nM
obtained with 9␣-fluoroprednisolone.
Biologic activity resulting from glucocorticoid receptor
7
5 nM for F-5␣, Pred and F-Pred, respectively. There are,
however, several similarities in the data of the two tables.
The acetonides are more active than the other derivatives,
and results suggest that the ␣-isomer in each of the ace-
tonide pairs is more potent than its ␤-counterpart.
Table 1
3
Competitive displacement of [ H]dexamethasone from Raw 264.7 cell
glucocorticoid receptors
a
3
3
Steroid [ H]Dex bound [ H]Dex bound
IC50
Ear edema
ID50
Novel anti-inflammatory steroids including 4␤, 5␣/
b
d
3
3
00 nM steroid 1000 nM steroid (nM)
5␤, F-5␣/F-5␤ displaced [ H]dexamethasone from the
4
4
␣
␤
75.7
66.0
80.5
83.1
16.7
33.4
25.9
47.4
78.2
80.9
69.0
93.4
18.4
51.2
39.1
47.1
84.4
11.1
11.7
9.7
30.2
56.9
62.7
46.1
82.1
—
Ͼ1000
622
901
Ͼ1000
35.1
211
32.6
81.5
Ͼ1000
Ͼ1000
972
Ͼ1000
32.8
5.9
glucocorticoid receptors in the competitor range of 10–
1000 nM, thus IC50 values were calculated. With the
exception of F-5␣, these agents did not inhibit NO gen-
eration in LPS-stimulated macrophages in comparable
concentration ranges. These differences may be ex-
plained, at least in part, by differences in time required to
complete binding versus NO generation assays and the
intracellular events required for the assay endpoints.
Whole cell receptor competition assays are completed in
F-4␣
F-4␤
5
5
␣
␤
1.4
3.0
F-5␣
F-5␤
6
6
␣
␤
F-6␣
F-6␤
Pred
3
h. The endpoint of the competitive binding assay is
c
c
85.5
34.1
0.4
binding of the agent to the cognate cytoplasmic glucocor-
ticoid receptor, the first of a long series of events which
culminate in the biologic activity resulting from receptor
occupancy. NO generation was assessed by utilizing the
Griess reaction which detects the amounts of stable ni-
trite, an endproduct of NO generation by activated cells.
Results of time course studies with RAW 264.7 macro-
phages have shown that colorimetric detection of nitrite
requires approximately 12 h with levels continuing to
FPred 13.9
—
a
3
Percent of [ H]dexamethasone bound after competitive displacement.
3
Specific [ H]dexamethasone binding was 8853 Ϯ 286 dpms/mg protein.
b
IC50 is the concentration of competing steroid which displaces 50% of
3
specifically bound [ H]dexamethasone.
c
IC50 for Pred and FPred were calculated from the concentration-
response ranges of 3–300 nM.
d
The dose (␮mol/ear) which inhibited ear edema by 50% was estimated
from a plot of percent inhibition versus dose.

3
58
Z. You et al. / Steroids 67 (2002) 353–359
Table 2
boosters or inhibitors, respectively [20]. Effective inter-
actions of coactivators and corepressors have been stud-
ied using classic glucocorticoids and are thought to occur
exclusively with steroid-bound glucocorticoid receptors
where their associations result in further conformational
differences. The current emerging picture suggests coac-
tivators/corepressors are flexible, but precise, coordina-
tors of complex and dynamic networks in which tran-
scriptional regulation by glucocorticoid receptor
complexes and other nuclear receptors is linked to the
final signaling pathway [21]. Subtly different conforma-
tional changes induced by binding of the novel antedrugs
to the glucocorticoid receptors could influence the asso-
ciations of either coactivators or corepressors and mark-
edly alter receptor-mediated transcriptional activity.
Taken together, these results suggest that the ␣-iso-
mers of the acetonides (5␣, F-5␣) which bind to
glucocorticoid receptors with high affinity and possess
a metabolically labile methyl ester group appear to pro-
vide a solid lead to ideal anti-inflammatory steroid ante-
drugs. Structural modifications which retain their binding
affinity while modestly enhancing the ester stability
against metabolic hydrolysis could yield very potent anti-
inflammatory agents with greatly reduced systemic side
effects.
Effects of steroids on LPS-induced nitric oxide generation in Raw 264.7
a
cells
Steroid
% NO
Inhibition–
% NO
Inhibition–
300 nM
% NO
Inhibition–
1000 nM
IC50
(nM)
c
1
00 nM
b
4
4
␣
␤
NI
NI
NI
NI
NI
39.8
14.8
50.5
11.4
NI
7.3
1.6
NI
54.5
69.2
NI
NI
NI
NI
44.9
17.8
53.1
12.7
6.7
8.6
4.6
NI
69.2
80.6
Ͼ1000
Ͼ1000
Ͼ1000
Ͼ1000
Ͼ1000
Ͼ1000
NI
NI
NI
35.8
8.9
48.0
19.1
NI
NI
NI
NI
41.5
57.8
F-4␣
F-4␤
5
5
␣
␤
d
F-5␣
F-5␤
342
Ͼ1000
Ͼ1000
Ͼ1000
Ͼ1000
Ͼ1000
6
6
␣
␤
F-6␣
F-6␤
Pred
FPred
d
186
d
75.4
a
6
Attached cells (1 ϫ 10 cells/well) in 24-well plates were treated with
LPS (10 ng/ml) in DMEM/5% fetal bovine serum 3 h after treatment with
3
–1000 nM steroids. Supernatant nitrite, the stable end product of NO
generation, was assayed by the Griess reaction at 18 h. Experiments were
conducted in triplicate. Untreated controls generated 12.9 Ϯ 0.89 nmol
6
NO/1 ϫ 10 cells.
b
NI ϭ no inhibition.
IC50 is the concentration of steroid which inhibits generation of NO by
c
5
0%.
d
IC50s for F-5␣ and Pred were calculated from the concentration-
response range of 10–1000 nM, and the IC50 for FPred was calculated
from the range of 3 to 300 nM.
Acknowledgments
This work was supported by NIH 21627, RR 08111 and
RCMI 0330.
increase up to 72 h following addition of the stimulant
[
9]. In contrast to Pred and F-Pred, the steroidal ante-
drugs being screened in these investigations may not be
stable enough to endure the 18 h treatment. Substantial
hydrolysis of the methyl ester group may have occurred
within this time, thus resulting in a loss of NO inhibition.
This is supported by results of recent investigations on
inhibition of macrophage NO production by arachido-
nate-cascade inhibitors. Inhibitory activity of the classic
anti-inflammatory steroid dexamethasone was much
higher in continuous exposure experiments than in glu-
cocorticoid pulsing experiments [17]. To exert steroid-
regulated biologic effects, the current working model
involves binding of steroid to the cognate receptor, acti-
vation and/or transformation of the receptor-steroid com-
plex, nuclear translocation and dimerization, complex
binding to the hormone response element, recruitment of
coactivators vs corepressors and interaction with auxil-
iary transcription factors followed by a series of events
which initiate or inhibit gene transcription [18,19]. These
are the intracellular events which precede measurement
of the ability of glucocorticoids to inhibit NO generation
in stimulated cells. Diverse mechanisms of steroid-recep-
tor complex actions have been partially clarified as a
consequence of discovery of steroid receptor coactivators
and corepressors which function as transcriptional power
References
[1] Lee HJ, Soliman MRI. Anti-inflammatory steroids without pituitary-
adrenal suppression. Science 1982;215:989–91.
[
2] Lee HJ, Heiman AS, Taraporewala IB. New steroidal anti-inflamma-
tory drugs. In: Rainsford KD, Velo GP, editors. New Developments
in Anti-Rheumatic Therapy. MPT Press, Lancaster, 1989:III:153–86.
3] Khalil MA, Kwon T, Lee HJ. A novel approach to the development
of safer anti-inflammatory steroids: Antedrug. Current Topics in Med
Chem 1993;1:173–202.
[
[4] Bird J, Kim HP, Lee HJ. Topical anti-inflammatory activity of esters
of steroid-21-oic acids. Steroids 1986;47:35–40.
[
5] Kim HP, Bird J, Heiman AS, Hudson GF, Taraporewala IB, Lee HJ.
Synthesis of new antiinflammatory steroidal 20-carboxamides: (20R)-
and (20S)-(N-substituted amino)-11,17,20-trihydroxy-3,21-dioxo-
1
,4-pregnadiene. J Med Chem 1987;30:2239–44.
[
6] Heiman AS, Taraporewala IB, Lee HJ. Local anti-inflammatory
activity of steroid-21-oate esters in the carrageenan pleurisy model
of acute inflammation. Drug Develop Res 1989;17:153–60.
th
[7] Steraloids, Inc., Wilton, NH. Steroids (11 Ed) p. 166.
th
[8] Steraloids, Inc., Wilton, NH. Steroids (11 Ed) p. 167.
[
9] Heiman AS, Hickman F, Ko D, Lee HJ. New steroidal anti-inflam-
matory antedrugs bind to macrophage glucocorticoid receptors and
inhibit nitric oxide generation. Steroids 1997;63:644–9.
[
10] Khalil MA, Lay JC, Lee HJ. Synthesis of new anti-inflammatory
steroidal acid esters: methyl 11-hydroxy-3,20-dioxo-1,4-pregnadiene-
21-oate. J Pharm Sci 1985;74:180–3.

Z. You et al. / Steroids 67 (2002) 353–359
359
[
[
[
11] Lewbart ML, Mattox VR. Glycolic acids and esters from cortisone. J
[17] Ryoyama K, Nomura T, Nakamura S. Inhibition of macrophage nitric
oxide production by arachidonate-cascade inhibitors. Cancer Immu-
nol Immunother 1993;37:385–91.
[18] Szapary D, Huang Simons SS. Opposing effects of corepressor and
coactivators in determining the dose-response curve of agonists,
and residual agonist activity of antagonists for glucocorticoid
receptor-regulated gene expression. Molec Endocrinol 1999;13:
2108–21.
Org Chem 1963;28:1773–9.
12] Lewbart ML, Mattox VR. Conversion of steroid-17-yl glyoxals to
epimeric glycolic esters. J Org Chem 1963;28:1779–85.
13] DiRosa M, Radomski M, Carnuccio R, Moncada S. Glucocorticoids
inhibit the induction of nitric oxide synthase in macrophages. Bio-
chem Biophys Res Commun 1990;172:1246–52.
[14] Szabo C, Thiemermann C. Regulation of the expression of the induc-
ible isoform of nitric oxide synthase. Adv Pharmacol 1995;34:113–
[19] Barnes PJ. Molecular mechanisms of steroid action in asthma. J
Allergy Clin Immunol 1996;97:159–68.
5
3.
[
15] Simmons WW, Ungureanu-Longrois D, Smith GK, Smith TW, Kelly
RA. Glucocorticoids regulate inducible nitric oxide synthase by in-
hibiting tetrahydrobiopterin synthesis and L-arginine transport. J Biol
Chem 1996;271:23928–37.
[20] Kurihara I, Shibata H, Suzuki T, Ando T, Kobayashi S, Hayashi M,
Saito I, Saruta T. Transcriptional regulation of steroid receptor coac-
tivator-1 (SRC-1) in glucocorticoid action. Endocr Res 2000;26:
1033–8.
[
16] You Z, Heiman AS, Chen M, Lee HJ. Novel steroid spiro enones:
condensation of prednisolone derivatives with diethyl oxalate. Ste-
roids 2000;65:109–15.
[21] Jenkins BD, Pullen CB, Darimont BD. Novel glucocorticoid receptor
coactivator effector mechanisms. Trends Endocrinol Metab 2001;3:
1222–6.