Conformationally constrained anesthetic steroids that modulate GABA(A) receptors

Article abstract of DOI:10.1021/jm000977e

Various cyclic ether and other 3α-hydroxyandrostane derivatives bearing a conformationally constrained hydrogen-bonding moiety were prepared. Their anesthetic potency and their binding affinity for GABA(A) receptors, measured by intravenous administration to mice and inhibition of [³⁵S]TBPS binding to rat whole brain membranes, were compared with that of known anesthetic 3α-hydroxypregnan-20-ones. Synthetic steroids with similar in vitro and in vivo activities to the endogenous 3α-hydroxypregnan-20-ones all had an ether oxygen on the β-face of the steroid D-ring. These results suggest that for optimal GABA(A) receptor modulation, the hydrogen bond-accepting substituent should be near perpendicular to the plane of the D-ring on the β-face of the steroid.

Full text of DOI:10.1021/jm000977e

4118
J . Med. Chem. 2000, 43, 4118-4125
Con for m a tion a lly Con str a in ed An esth etic Ster oid s Th a t Mod u la te GABAA
Recep tor s
Alison Anderson, Andrew C. Boyd, J ohn K. Clark, Lee Fielding, David K. Gemmell, Niall M. Hamilton,*
Maurice S. Maidment, Valerie May, Ross McGuire, Petula McPhail, Francis H. Sansbury, Hardy Sundaram, and
Robert Taylor
Research and Development, Organon Laboratories Ltd., Newhouse, Motherwell, Lanarkshire ML1 5SH, Scotland, U.K.
Received May 8, 2000
Various cyclic ether and other 3R-hydroxyandrostane derivatives bearing a conformationally
constrained hydrogen-bonding moiety were prepared. Their anesthetic potency and their binding
affinity for GABAA receptors, measured by intravenous administration to mice and inhibition
of [35S]TBPS binding to rat whole brain membranes, were compared with that of known
anesthetic 3R-hydroxypregnan-20-ones. Synthetic steroids with similar in vitro and in vivo
activities to the endogenous 3R-hydroxypregnan-20-ones all had an ether oxygen on the â-face
of the steroid D-ring. These results suggest that for optimal GABAA receptor modulation, the
hydrogen bond-accepting substituent should be near perpendicular to the plane of the D-ring
on the â-face of the steroid.
In tr od u ction
tion of GABAA receptors. Replacement of the 17-acetyl
side chain of allopregnanolone (1) with a nitrile is known
to retain activity, but inversion of the configuration at
C17 reduces activity.9 The effect of introducing a 16,
17-double bond into steroids, including 1, 3, and 4, has
also been studied and generally found to be deleterious
for in vitro and in vivo activity.10
It is well-established that neurosteroids such as
(3R,5R)-3-hydroxypregnan-20-one (1) and (3R,5R)-3,21-
dihydroxypregnan-20-one (2) and neuroactive steroids
such as (3R,5â)-3-hydroxypregnan-20-one (pregnanolo-
ne, 3) modulate γ-aminobutyric acidA (GABAA) receptor
In this paper we report the syntheses of several
allosteric GABAA modulators where the 20-keto group
is replaced by an alternative hydrogen bond-accepting
moiety and discuss their potency as inhibitors of TBPS
binding, proposing from our results a preferred pregnan-
20-one side chain conformation for receptor modulation.
In addition we report the anesthetic effect of the
compounds upon intravenous administration to mice,
finding the hypnotic potency to be comparable in certain
instances to that of the endogenous steroids.
Ch em istr y
function.1 γ-Aminobutyric acid (GABA) is the major
inhibitory neurotransmittter within the CNS (central
nervous system), and steroids that potentiate the effects
of GABA at GABAA receptors have therapeutic potential
as anticonvulsants,2 anxiolytics,3 sedatives, and hyp-
notics.4 In addition to the above endogenous steroids, a
number of synthetic analogues including alfaxalone (4)
have been shown to modulate GABAA receptors allo-
sterically.5 The in vitro activity of steroids at GABAA
receptors can be evaluated through their ability to
inhibit the specific binding of [35S]-tert-butyl bicyclo-
phosphorothionate ([35S]TBPS) to rat whole brain mem-
branes.6 Almost all potent modulators have a 3R-ol and
20-keto function with the natural 17â-pregnane side
chain. While the conformation of pregnan-20-ones has
been extensively studied for optimal binding to proges-
terone receptors,7,8 there are few reports regarding the
preferred side chain conformation for allosteric modula-
To try to establish the preferred side chain conforma-
tion of pregnan-20-ones for binding to GABAA receptors,
a number of conformationally constrained steroids bear-
ing a hydrogen bond-accepting substituent on the D-ring
were synthesized. These included epoxides 5-7, 10, and
11, ketones 8 and 9, oxetanes 12 and 13, a nitrile 14,
an alcohol 15, and an ether 16.
The 16â,17â-epoxide 5 was prepared by treating the
halohydrin 20 with base. The halohydrin was synthe-
sized following the method of Fajkosˇ and Sˇorm: i.e.,
androsterone acetate 17 was converted to the enol ester
18 and brominated to give the 16R-bromo ketone 19
which was reduced to the 17â-ol 20 (Scheme 1).11
The 16R,17R-epoxide 612 and the 16â,17â-epoxide 7
were prepared by stereoselective epoxidation of the
relevant androstenes 21 and 24 (Scheme 2).13 The
alkene 24 was synthesized from 5â-androstan-17-one 22
via the tosylhydrazone 23.
Androsterone 8, like the 3R,17â-diol 15, is com-
mercially available. The 16-keto analogue 914 was
* Author for correspondence. Tel: +44 1698 736000. Fax: +44 1698
736187. E-mail: n.hamilton@organon.nhe.akzonobel.nl.
10.1021/jm000977e CCC: $19.00 © 2000 American Chemical Society
Published on Web 10/19/2000

Anesthetic Steroids Modulate GABAA Receptors
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4119
Sch em e 1a
Sch em e 3a
a
Reagents: (a) BzOH, PPh3, 1,1′-(azodicarbonyl)dipiperidine;
(b) NaOH, MeOH.
Sch em e 4a
a
Reagents: (a) Ac2O, pyr; (b) H2CdCMeOAc, TsOH; (c) Br2,
CCl4; (d) LiAlH4; (e) KOH, MeOH.
Sch em e 2a
a
Reagents: (a) Me3SI, KOBut; (b) Me3SOI, KOBut.
Sch em e 5a
a
Reagents: (a) NaOMe, HCO2Et; (b) NaBH4; (c) TsCl, pyr; (d)
NaOMe.
a
For synthesis of the 17R-nitrile 14 (Scheme 6), 13R-
androsterone 32 was prepared following the method of
Boar et al.16 Reaction of 13R-androsterone 32 with
tosylmethyl isocyanide (TOSMIC) selectively gave the
desired 17R-acetonitrile 14 in low yield. The stereo-
chemistry of both structures was confirmed by NMR
experiments. For structure 14 NOEs were observed
between the 18-Me protons and the 14R-, 15R-, 16R-,
and 17â-protons, as well between the 11â-proton and
the 11R-, 12R-, and 12â-protons. For structure 32 similar
NOEs were observed between the 18-Me protons and
the 14R- and 15R-protons.
Reagents: (a) AcOOH, NaOAc; (b) TsNHNH2, TsOH; (c) MeLi;
(d) NBS, DMSO, H2O then KOH.
prepared by epimerizing 3â-hydroxy 16-ketone 25 via
the benzoate 26 (Scheme 3).
Synthesis of spiro-epoxides 1015 and 11 required
stereoselective insertion of a methylene unit. This was
accomplished by treating ketones 8 and 22 with a
sulfonium ylide (Scheme 4). Further reaction of epoxide
10 with a sulfoxonium ylide gave the spiro-oxetane 12.
The fused-oxetane 13 (Scheme 5) was prepared by
formylation of androsterone 8 followed by stereoselective
borohydride reduction to give the triol 28. Selective
tosylation of the primary alcohol, followed by cyclization
with sodium methoxide, afforded the 16â,17â-oxetane
13.
The 17â-ether 1617 was prepared by methylation of
the diol 1518 (Scheme 7).

4120 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Anderson et al.
Ta ble 1. Activity of Steroids in the Radioligand ([35S]TBPS)
Binding Assay and as Intravenous Hypnotics in the Mouse
Sch em e 6a
compd
TBPSa IC50 (nM) [N]b
HD50c (µmol‚kg-1
)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
155 ( 6 [4]
192 ( 28 [3]
145 ( 31 [5]
423 ( 36 [4]
255 ( 65 [3]
5350 ( 450 [2]
159 ( 29 [3]
1367 ( 203 [3]
1767 ( 186 [3]
55 ( 13 [3]
221 ( 71 [3]
232 ( 36 [3]
2533 ( 371 [3]
1236 ( 283 [3]
1867 ( 433 [3]
210 ( 76 [3]
10.0 (8.2-12.2)
18.5
9.5 (8.1-11.1)
6.3 (5.6-7.1)
17.3 (15.6-19.4)
34.2-68.4
38.0 (31.8-45.4)
68.9-103.3
134.7
19.8 (16.8-23.5)
22.6 (20.2-25.4)
31.4-62.8
146.9
114.9
239.4 (205.2-273.5)
23.1
a
a
IC50: concentration (nM) of steroid required to inhibit 50% of
binding of [35S]TBPS from rat whole brain membranes; values are
Reagents: (a) NH2OH‚HCl, pyr; (b) Ac2O, pyr; (c) HCl, MeOH;
(d) KOBut, TOSMIC.
b
means of experimental determinations ( SEM. Number of
experimental determinations. c Hypnotic dose50 (µmol‚kg-1): dose
required to cause a loss of righting reflex for a minimum period of
30 s in 50% of treated mice after intravenous administration.
Sch em e 7a
to the problems of studying steroid-receptor interac-
tions, novel steroids that allosterically modulate GABAA
receptors have been identified to date through structural
modification of known ligands. Following the discovery
that alfaxalone modulates GABAA receptors, other
steroid anesthetics were tested and found to possess
similar in vitro activity.5,9,23 Several papers have ap-
peared in the literature describing such compounds, but
all potent derivatives retain a 3R-ol, an androstane
nucleus with either 5R- or 5â-configuration, and a
hydrogen bond-accepting substituent at the 17â-posi-
tion. Steroids without an A-ring also have similar
potency.24 So far derivatives with the natural pregnan-
20-one side chain (e.g. endogenous neurosteroids) or a
17â-carbonitrile substituent have been shown to produce
the greatest potentiation of GABAA receptor function.9
The effect of hydrogen bond-donating substituents at
the 17â-position has also been examined, but such
compounds generally exhibit reduced efficacy at GABAA
receptors.25 We sought to further broaden the range of
modulators by synthesizing conformationally constrained
analogues in which the hydrogen bond-accepting sub-
stituent is part of a cyclic ether, usually on the â-face
of the D-ring.
a
Reagents: (a) NaH, MeI.
P h a r m a cology
The methods used have been previously reported.5e
The anesthetic potency of the steroids was determined
upon their intravenous administration to mice. In each
case the dose required to cause a loss of righting reflex
for a minimum period of 30 s in 50% of treated mice
was determined by probit analysis. This dose is termed
the HD50 (hypnotic dose50). The in vitro effect of the
compounds at GABAA receptors was also assessed, by
determination of their ability to inhibit [35S]TBPS
binding to rat whole brain membranes. In each case the
concentration of drug required to inhibit 50% binding
of this radioligand was determined (TBPS IC50). Both
in vitro and in vivo results are shown in Table 1.
Discu ssion
The nature of the binding sites of steroids that
allosterically modulate GABAA receptors has still to be
established. Progress has been hindered because GABAA
receptors are membrane-bound proteins which are not
conducive to structural determination by methods such
as X-ray crystallography and NMR spectroscopy. More-
over all functional mammalian receptors that comprise
common subunits (such as R, â, and γ)19 appear to
possess steroid binding sites, although significant dif-
ferences in response from different brain areas have
been noted.20 A study, which involved constructing
chimeras between alfaxalone-sensitive GABAA receptor
R2 or â1 subunits and the alfaxalone-insensitive glycine
R1 subunit, suggested the site of action for neurosteroids
is on the N-terminal side of the middle of TM2 of GABAA
receptors.21 Steroids modulate GABAA receptors at sites
that are distinct from other general anesthetics includ-
ing barbiturates, benzodiazepines, and etomidate.22 Due
The analogues were oxiranes such as 5-7, 10, and
11, oxetanes such as 12 and 13, ketones 8 and 9, an
alcohol 15, and an ether 16. In addition, the nitrile 14
was prepared to see whether simultaneous inversion of
the D-ring stereochemistry at C13 and C17 was toler-
ated.
The [35S]TBPS assay measures binding to GABAA
receptors in whole rat brains, and the results in Table
1 are therefore an average of binding to a large number
of different GABAA receptor subtypes. It is not known
which of these subtypes are most important for anes-
thesia though a recent report showed the R1-subunit is
crucial for sedation.26 The affinity of a steroid for GABAA
receptors varies with subunit composition, while anes-
thetic activity is liable to depend on modulation of
specific GABAA receptor subtypes. For these reasons in
vitro and in vivo results are unlikely to correlate

Anesthetic Steroids Modulate GABAA Receptors
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4121
F igu r e 1. (a) Side chain of pregnan-20-ones indicating
rotation about the C17-C20 bond. (b) Newman projection C20
f C17 illustrating the conformation of the side chain in which
the C16-C17-C20-O20 torsion angle φ is about 25°.
directly. Examination of Table 1 confirms this is the case
though some generalizations can be drawn.
Steroids with a [35S]TBPS IC50 value in the range 50-
500 nM have an HD50 e 50 µmol‚kg-1, while those with
an IC50 value > 1000 nM have an HD50 g 50 µmol‚kg-1
.
The most active allosteric GABAA receptor modulators,
i.e., steroids 5, 7, 10-12, and 16, have a hydrogen bond-
accepting ether function on the â-face. The excellent in
vitro and in vivo activity of steroids 10-12 is surprising
since 17R-substitution is generally deleterious to activ-
ity.9,27,28 Steric constraint within a three- or four-
membered ring reduces protrusion of the substituent
below the R-face of the steroids. This constraint favors
inhibition of [35S]TBPS binding and allosteric modula-
tion of GABAA receptors. The poorer in vitro activity of
the R-epoxide 6 is consistent with the hydrogen bond-
accepting substituent being on the less favored R-face.
The anesthetic activity of R-epoxide 6 is also lower than
that of â-epoxide 5; this agrees with a previous study
which reported 17R-acetyl derivatives to be inferior
anesthetics compared with the natural 17â-isomers.29
The relationship between the anesthetic activity and
the conformation of the A-ring of the steroids has been
discussed before, and the results confirm that the
configuration at C5 does not markedly affect potency.5e,29
Comparison of oxetane 13 with epoxide 5 indicates
the additional methylene unit is deleterious for in vitro
and in vivo activity. The nitrile 14 retains some in vitro
and in vivo activity, but it is apparent that the natural
configuration at C13 and C17 is preferred. The alcohol
15 is also a relatively poor GABAA receptor modulator
and has low anesthetic potency. This confirms hydrogen
bond-accepting substituents (e.g. the methyl ether 16)
are preferable to hydrogen bond-donating substituents
and is consistent with the observation that pregnan-
20-ols have ‘partial agonist’ activity at GABAA recep-
tors.9,30
Computer modeling of steroids 1-4 demonstrated
there are two conformational minima for the 17â-side
chain: one with a C16-C17-C20-O20 torsion angle φ
varying from -22° to -35° (Figure 1) and the other with
C21 eclipsing the C16-C17 bond usually at a slightly
higher energy. The normal range in crystal structures
for φ has been reported as 0° to -41° in pregnan-20-
ones, with an average of -22°.7
The minimized structures of steroids 1-4 were over-
laid to align the B- and C-rings as closely as possible,
and structures 5 and 10 were then added (Figure 2).
The minimum energy position for O20 in pregnan-20-
ones 1-4 lies between the epoxide oxygen atoms of
androstanes 5 and 10. Other orientations of the 5â-
F igu r e 2. (Top) Structures of steroids 1-5 and 10. (Bottom)
Steroids 1-5 and 10 superimposed.
steroid 3 that involve improved overlay of the 3R-
hydroxyl group with 5R-steroids rather than overlay of
the B- and C-rings are also valid.9 For such orientations
it is still beneficial for O20 to lie between the epoxide
oxygen atoms of androstanes 5 and 10.
The in vitro results and the molecular modeling
suggest that for optimal GABAA receptor modulation the
hydrogen bond-accepting unit should be near perpen-
dicular to the plane of the D-ring on the â-face of the
steroid. This can also account for the poorer activity of
the keto androstanes 8 and 9 relative to pregnan-20-
one 1 and the good in vitro and in vivo activity of 17â-
nitrile-substituted steroids.9,29
Con clu sion
A number of steroids with a hydrogen bond-accepting
substituent attached to the D-ring were prepared and
retain good in vitro and in vivo activity relative to the
endogenous steroids 1-4. Furthermore, for optimal
GABAA receptor modulation the hydrogen bond-accept-
ing substituent should be near perpendicular to the
plane of the D-ring on the â-face of the steroid.
Exp er im en ta l Section
P h a r m a cology. In Vitr o Exp er im en ts ([35S]TBP S a s-
sa y) a n d in Vivo Exp er im en ts (h yp n otic p oten cy). The
methods used are identical to those reported previously for
water-soluble 2â-morpholinyl steroids.5e
Ch em istr y. Gen er a l. Melting points were taken with
either a Gallenkamp capillary melting point apparatus or a
Reichert hot plate apparatus and are uncorrected. Optical
rotations were determined at room temperature for solutions
in chloroform and c refers to concentration in g/100 mL. 1H
NMR (200 or 400 MHz) spectra were obtained using a Bruker
AM200 or Bruker DRX400 instrument; chemical shifts (δ) are
relative to tetramethylsilane as internal standard. Only
discrete or characteristic signals are reported. Coupling con-
stants are given in hertz. IR spectra were obtained with a
Perkin-Elmer 16PC FT-IR spectrometer. Elemental analyses
were determined on a Perkin-Elmer 2400 CHN elemental
analyzer and are within 0.4% of theory unless otherwise noted.

4122 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Anderson et al.
Ma ter ia ls. Reagents were used as supplied from com-
mercial sources. Steroids 1-4, 8 and 15 are commercially
available.
δ 0.77 (s, 3H, CH3), 0.81 (s, 3H, CH3), 3.17 (d, 1H, J ∼ 3), 3.47
(d, 1H, J ∼ 3), 4.04 (s, 1H); IR (KBr) 3284 (OH), 2931, 2853
(CH) cm-1. Anal. (C19H30O2‚0.11H2O) C, H.
Syn th eses. Steroid 15 was also isolated as a byproduct from
the preparation of steroid 5. Compounds 5,11 6,12 9,14 10,15 and
1617 have been previously reported.
(3r,5r,16r,17r)-16,17-E p oxya n d r ost a n -3-ol (6). To a
stirred solution of (3R,5R)-3-hydroxyandrost-16-ene (21) (500
mg, 1.82 mmol) in dichloromethane (25 mL) was added a
solution of sodium acetate (0.1 g, 1.22 mmol) and peracetic
acid (1.0 mL) in water (1.0 mL). After 1 h the solution was
washed with aqueous sodium sulfite, water, aqueous sodium
carbonate and then water. After drying over anhydrous sodium
sulfate, the solvent was removed under reduced pressure and
the residue was crystallized from acetone to give the title
compound 612 (337 mg, 1.16 mmol, 64%): mp 144.5-146.5 °C;
1H NMR (CDCl3) δ 0.73 (s, 3H, CH3), 0.79 (s, 3H, CH3), 1.89
(dd, 1H, J ∼ 12, 6), 3.09 (d, 1H, J ∼ 3), 3.34 (d, 1H, J ∼ 3),
(3r,5r,16â,17â)-16,17-Ep oxya n d r osta n -3-ol (5). Andro-
sterone (8) (10 g, 34 mmol) was stirred in pyridine (50 mL)
and acetic anhydride (10 mL, 106 mmol) for 3 h then left to
stand for 2 days. The product was precipitated with water (600
mL), filtered off, washed with water, taken up in diethyl ether,
washed with water, dried over sodium sulfate and the solvent
evaporated to give androsterone acetate31 (17) (11.3 g, 34
mmol), which was used in the next step without further
purification: 1H NMR (CDCl3) δ 0.82 (s, 3H, CH3), 0.87 (s, 3H,
CH3), 2.05 (s, 3H, CH3), 5.01 (m, 1H).
4.04 (m, 1H); IR (KBr) 3490 (OH), 3024, 2979, 2920 (CH) cm-1
Anal. (C19H30O2) C, H.
.
p-Toluenesulfonic acid (1.13 g, 5.94 mmol) was added to
androsterone acetate 17 (11.3 g, 34 mmol) in isopropenyl
acetate (20 mL, 182 mmol) and the mixture heated in an oil
bath at 130-140 °C for 6 h, collecting the distillate, and
replacing each 25 mL of distillate with fresh isopropenyl
acetate (7 × 25 mL). After standing overnight, more p-
toluenesulfonic acid (1.13 g, 5.94 mmol) and isopropenyl
acetate were added, and a Vigreux column added to aid
fractionation of acetone distillate from isopropenyl acetate. The
reaction mixture was heated for a further 8 h. The reaction
was cooled to 70 °C, triethylamine (0.85 mL) added, then
allowed to cool to room temperature and water (50 mL) added.
The solvent was evaporated under reduced pressure until free
of isopropenyl acetate and the product was filtered off, washed
and dried in vacuo to give crude (3R,5R)-androst-16-ene-3,17-
diol diacetate11 (18) (18.3 g), which was used in the next step
without further purification: 1H NMR (CDCl3) δ 0.82 (s, 3H,
CH3), 0.89 (s, 3H, CH3), 2.06 (s, 3H, CH3), 2.15 (s, 3H, CH3),
5.00 (m, 1H), 5.45 (m, 1H).
A solution of crude diacetate 18 (18.3 g) in carbon tetra-
chloride (360 mL) was cooled to -5 °C and a solution of
bromine (6.4 g, 40 mmol) in carbon tetrachloride (200 mL)
added dropwise with stirring over 10 min. The mixture was
stirred for a further 20 min and washed with aqueous sodium
thiosulfate, aqueous sodium bicarbonate then water, dried over
sodium sulfate and the solvent evaporated to give crude (16R)-
bromoandrosterone acetate11 (19) (13.6 g, 33.1 mmol): 1H NMR
(CDCl3) δ 0.82 (s, 3H, CH3), 0.91 (s, 3H, CH3), 2.05 (s, 3H, CH3),
4.53 (m, 1H), 5.01 (m, 1H).
A solution of crude acetate 19 (13.6 g, 33.1 mmol) in dry
tetrahydrofuran (186 mL) was added over 15 min to a stirred
suspension of lithium aluminum hydride (6.8 g, 179 mmol) in
dry diethyl ether (2040 mL). The reaction was stirred at -5
to 0 °C for 1 h, then quenched with ethyl acetate and then
water (50 mL). Sodium hydroxide (4 N, 20 mL) was added and
the mixture filtered. The filtrate was dried over sodium sulfate
and the solvent evaporated to give crude (3R,5R,16R,17â)-16-
bromoandrostane-3,17-diol32 (20) (11.1 g, 29.9 mmol): 1H NMR
(CDCl3) δ 0.67 (s, 3H, CH3), 0.71 (s, 3H, CH3), 4.00 (m, 1H).
NMR also indicated the presence of some (3R,5R,17â)-andro-
stane-3,17-diol (15).
Crude bromodiol 20 (11.1 g, 29.9 mmol) was refluxed with
methanol (150 mL) and potassium hydroxide (10 N, 15 mL)
for 2.5 h. The solution was cooled, reduced in volume and the
product precipitated by addition of water and filtered off. The
solid was taken up in diethyl ether, washed with water, dried
over sodium sulfate and the solvent evaporated to give a crude
solid. NMR indicated ∼40-50% 5, ∼25% 15 and ∼25-35% 8.
The crude solid was dissolved in methanol (100 mL), sodium
borohydride (500 mg, 13.2 mmol) added and the mixture
stirred for 30 min. The mixture was quenched with acetic acid,
poured into water and the product extracted with diethyl ether,
dried and the solvent evaporated to give a solid (8.3 g). This
was taken up in a small amount of methanol and columned
on silica, eluting with dichloromethane. Recolumning followed
by recrystallization from diethyl ether/heptane gave the title
compound 511 (2.33 g, 8.0 mmol, 23% from 8): mp 196-197
°C (lit. mp 190-191 °C); [R]D +33.9° (c 0.8); 1H NMR (CDCl3)
(3r,5â,16â,17â)-16,17-Ep oxya n d r osta n -3-ol (7). 3R-Hy-
droxy-5â-androstan-17-one (22) (2.0 g, 6.89 mmol), p-toluene-
sulfonyl hydrazide (1.42 g, 7.62 mmol) and p-toluenesulfonic
acid (20 mg, 0.1 mmol) were refluxed in ethanol overnight.
Further p-toluenesulfonyl hydrazide (700 mg, 3.76 mmol) and
p-toluenesulfonic acid (20 mg, 0.1 mmol) were added and the
mixture refluxed for a further 5 h. The solvent was evaporated
under reduced pressure and the residue chromatographed on
silica gel (gradient elution, toluene/ethyl acetate) to give crude
3R-hydroxy-5â-androstan-17-one p-toluenesulfonyl hydrazone
(23) (4.2 g, 9.2 mmol). To a solution of this steroid in
tetrahydrofuran (50 mL) (cooled in an ice-water bath) was
added methyllithium (1.4 N in tetrahydrofuran, 21.5 mL, 30
mmol) under nitrogen over 5 min. A precipitate initially formed
which redissolved as the addition of methyllithium continued.
The dark orange-red solution was stirred for 2 h, then left
overnight at room temperature. The reaction was quenched
by the slow addition of water, then diethyl ether was added
to give two layers and the mixture acidified with 2 N
hydrochloric acid. The organic layer was separated, washed
with water then brine, dried and the solvent evaporated to
give an orange oil (3.1 g), which was chromatographed on silica
gel (toluene/ethyl acetate 4/1) to give (3R,5â)-3-hydroxyandrost-
16-ene13 (24) which solidified on standing (1.14 g, 4.15 mmol,
60% from 22): 1H NMR (CDCl3) δ 0.74 (s, 3H, CH3), 0.96 (s,
3H, CH3), 3.62 (m, 1H), 5.18 (m, 1H), 5.82 (m, 1H).
To a stirred solution of the androstene 24 (1.1 g, 4.0 mmol)
in dimethyl sulfoxide (12 mL) at 5 °C were added water (1.2
mL) and then N-bromosuccinimide (1.1 g, 6.2 mmol) portion-
wise over 5 min. The mixture was stirred for 20 min after
which cold water was added and the resulting solid filtered
off. This solid was dissolved in ethanol and aqueous potassium
hydroxide (10 M, 0.9 mL) added. After heating to 75 °C for 10
min, the ethanol was removed under reduced pressure. After
cooling, the residue was diluted with water and the brown
gummy precipitate filtered off. This material was dissolved
in dichloromethane and the solution washed with water and
brine, and then dried over anhydrous sodium sulfate. The
solvent was removed under reduced pressure and the residue
(1.2 g) was chromatographed on silica gel (gradient elution,
toluene/ethyl acetate). Crystallization from diethyl ether gave
the title compound 7 (230 mg, 0.79 mmol, 20%): mp 146-150
1
°C; [R]D +37.7° (c 0.7); H NMR (CDCl3) δ 0.80 (s, 3H, CH3),
0.92 (s, 3H, CH3), 3.17 (d, 1H, J ∼ 3), 3.47 (d, 1H, J ∼ 3), 3.63
(br s, 1H); IR (KBr) 3328 (OH), 2864 (CH) cm-1. Anal.
(C19H30O2‚0.13H2O) C, H.
(3r,5r)-3-Hyd r oxya n d r osta n -16-on e (9). To a stirred
mixture of (3â,5R)-3-hydroxyandrostan-16-one (25) (244 mg,
0.84 mmol), benzoic acid (206 mg, 1.69 mmol) and tri-n-
butylphosphine (683 mg, 3.38 mmol) in toluene (12.2 mL) was
added 1,1′-(azodicarbonyl)dipiperidine (426 mg, 1.69 mmol).
After 3.5 h at 60 °C the reaction mixture was diluted with
water and extracted with diethyl ether. The combined extracts
were washed with water, dried over anhydrous sodium sulfate
and the solvent removed under reduced pressure. The residue
was chromatographed on silica gel (gradient elution, toluene/
diethyl ether) to give (3R,5R)-3-hydroxyandrostan-16-one ben-

Anesthetic Steroids Modulate GABAA Receptors
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4123
zoate (26) (318 mg, 0.81 mmol, 96%): 1H NMR (CDCl3) δ 0.89
(s, 6H, 2 × CH3), 5.30 (m, 1H), 7.50 (m, 3H), 8.07 (m, 2H).
phase was acidified with dilute hydrochloric acid and the
precipitated solid filtered off and washed with water. Evapora-
tion of residual water under reduced pressure afforded (3R,5R)-
3-hydroxy-16-hydroxymethyleneandrostan-17-one33 (27) (4.47
g, 14.0 mmol, 83%), which was sufficiently pure for the next
stage: 1H NMR (CDCl3) δ 0.76 (s, 6H, 2 × CH3), 0.97 (m, 1H),
1.78 (m, 1H), 1.88 (m, 1H), 2.48 (m, 1H), 3.80 (m, 1H, CHOH),
4.20 (br s, 1H, OH), 7.35 (s, 1H, CdCHOH), 10.50 (br s, 1H,
OH); IR (KBr) 3373 (OH), 2915 (CH), 1695, 1665, 1629, 1597
A mixture of the benzoate 26 (307 mg, 0.78 mmol) and
aqueous sodium hydroxide (0.9 mL) in methanol (6 mL) was
heated under reflux for 4 h, then left to stand overnight. The
reaction mixture was diluted with water and extracted with
diethyl ether. The combined extracts were washed with water,
dried over anhydrous sodium sulfate and the solvent removed
under reduced pressure. The residue was chromatographed
on silica gel (gradient elution, dichloromethane/diethyl ether)
and crystallized from diethyl ether/heptane to give the title
compound 914 (91 mg, 0.31 mmol, 40%): mp 173-175 °C (lit.
(CdO, CdC) cm-1
.
To a stirred solution of steroid 27 (4.3 g, 13 mmol) in
methanol (150 mL) at ∼5 °C was added sodium borohydride
(1.0 g, 26 mmol). After stirring for 5 h at room temperature,
acetone was added to decompose any residual sodium boro-
hydride, and the mixture was then concentrated to half volume
under reduced pressure. Water was added and the precipitated
off-white solid filtered off, washed with water and dried under
reduced pressure to give 16-hydroxymethylandrostane-3,17-
diol (4.10 g) as a mixture of isomers. This was flash chromato-
graphed on silica gel (toluene/ethyl acetate, gradient elution
1/1 to 1/4) to give (3R,5R,16â,17â)-16-hydroxymethylandro-
stane-3,17-diol (28) (3.5 g, 11 mmol, 80%), which was suf-
ficiently pure for the next stage: 1H NMR (DMSO) δ 0.64 (s,
3H, CH3), 0.73 (s, 3H, CH3), 3.26 (m, 1H), 3.59 (m, 2H), 3.79
(m, 1H).
1
mp 153-154 °C); H NMR (CDCl3) δ 0.82 (s, 3H, CH3), 0.88
(s, 3H, CH3), 2.11 (d, 1H, J ∼ 17), 2.20 (m, 1H), 4.06 (m, 1H);
IR (KBr) 3551 (OH), 2924 (CH), 1740 (CdO) cm-1. Anal.
(C19H30O2‚0.06H2O) C, H.
(3r,5r,17â)-Sp ir o[a n d r osta n e-17,2′-oxir a n ]-3-ol (10). To
a stirred solution of (3R,5R)-3-hydroxyandrostan-17-one (8) (5.0
g, 17.2 mmol) in dimethylformamide (50 mL) were added
trimethylsulfonium iodide (5.27 g, 25.8 mmol) and potassium
tert-butoxide (2.9 g, 25.8 mmol) under nitrogen. After 45 min
water was added with stirring and the precipitated pale yellow
solid filtered off and washed with water. The crude product
was crystallized from aqueous acetone, then flash chromato-
graphed on silica gel (toluene/ethyl acetate 1/1). Crystallization
from ethyl acetate gave the title compound 1015 (1.58 g, 5.20
mmol, 30%): mp 227-231 °C; [R]D +7.0° (c 0.7); 1H NMR
(CDCl3) δ 0.79 (s, 3H, CH3), 0.87 (s, 3H, CH3), 2.00 (m, 1H),
2.60 (d, 1H, J ∼ 5), 2.90 (d, 1H, J ∼ 5), 4.05 (br s, 1H); IR
(KBr) 3446 (OH), 2881 (CH) cm-1. Anal. (C20H32O2) C, H.
To a stirred solution of steroid 28 (3.0 g, 9.3 mmol) in
pyridine (30 mL) was added p-toluenesulfonyl chloride (1.8 g,
9.4 mmol). After 4 h at room temperature the reaction mixture
was poured into ice/water and left to stand overnight. The solid
was filtered off, washed with water and dissolved in dichlo-
romethane. This solution was washed with water, dried over
anhydrous sodium sulfate and the solvent removed under
reduced pressure. The residue (1.2 g) was triturated with
diethyl ether to give [(3R,5R,16â,17â)-3,17-dihydroxyandrostan-
17-yl]methyl p-toluenesulfonate (29) (1.10 g, 2.31 mmol, 25%),
as an off-white solid which was sufficiently pure for the next
stage: 1H NMR (CDCl3) δ 0.68 (s, 3H, CH3), 0.77 (s, 3H, CH3),
2.45 (s, 3H, CH3), 3.73 (d, 1H, J ∼ 10), 4.03 (m, 2H), 4.22 (dd,
1H, J ∼ 10, 7), 7.34 (d, 2H, J ∼ 8), 7.79 (d, 2H, J ∼ 8); IR
(KBr) 3521, 3346 (OH), 2928 (CH), 1598 (Ar), 1355, 1173 (SO2)
(3r,5â,17â)-Sp ir o[a n d r osta n e-17,2′-oxir a n ]-3-ol (11). To
a stirred solution of (3R,5â)-3-hydroxyandrostan-17-one (22)
(2.0 g, 6.9 mmol) in dimethylformamide (10 mL) and tetrahy-
drofuran (10 mL) were added trimethylsulfonium iodide (5.27
g, 25.8 mmol) and potassium tert-butoxide (2.9 g, 25.8 mmol)
under nitrogen. After 2 h more trimethylsulfonium iodide (5.27
g, 25.8 mmol) and potassium tert-butoxide (2.9 g, 25.8 mmol)
in dimethylformamide (10 mL) was added and stirring main-
tained for a further 2 h. The reaction mixture was poured into
ice/water and after stirring for 15 min the precipitated pale
yellow solid was filtered off and washed with water. The solid
was flash chromatographed on silica gel (toluene/ethyl acetate
1/1). Crystallization from diethyl ether/heptane gave the title
compound 11 (460 mg, 1.51 mmol, 22%): mp 155-158 °C; [R]D
cm-1
.
To a stirred suspension of steroid 29 (0.95 g, 2.0 mmol) in
methanol (25 mL) was added sodium methoxide (0.47 g, 8.7
mmol) and the mixture heated under reflux for 4 h. The
resulting solution was concentrated to low volume under
reduced pressure and diluted with water. The precipitated
solid was filtered off, washed with water and dried under
reduced pressure. The crude product was flash chromato-
graphed on silica gel (toluene/ethyl acetate 2/1) to give the title
compound 13 (130 mg, 0.43 mmol, 21%): mp 183-185 °C; [R]D
+4.8° (c 0.7); 1H NMR (CDCl3) δ 0.81 (s, 3H, CH3), 1.05 (s,
3H, CH3), 3.15 (m, 1H), 4.05 (m, 1H), 4.22 (t, 1H, J ∼ 6), 4.56
(d, 1H, J ∼ 8), 4.74 (dd, 1H, J ∼ 8, 6); IR (KBr) 3469 (OH),
2910 (CH) cm-1. Anal. (C20H32O2) C, H.
1
+18.1° (c 0.6); H NMR (CDCl3) δ 0.87 (s, 3H, CH3), 0.94 (s,
3H, CH3), 2.00 (m, 1H), 2.61 (d, 1H, J ∼ 3), 2.90 (d, 1H, J ∼
3), 3.63 (br s, 1H); IR (KBr) 3381 (OH), 2934 (CH) cm-1. Anal.
(C20H32O2) C, H.
(3r,5r,17r)-17,21-Ep oxyp r egn a n -3-ol (12). To a stirred
suspension of trimethylsulfoxonium iodide (1.45 g, 6.59 mmol)
in tert-butanol (14.5 mL) at 50 °C was added a solution of
potassium tert-butoxide (0.74 g, 6.59 mmol) in tert-butanol
(10.5 mL). After 30 min
a solution of (3R,5R,17â)-spiro-
[androstane-17,2′-oxiran]-ol (10) (1.0 g, 3.3 mmol) in tert-
butanol (20 mL) was added and the mixture heated under
reflux for 72 h. The reaction mixture was concentrated to low
volume under reduced pressure and diluted with dichlo-
romethane. This solution was washed with water, dried over
anhydrous sodium sulfate and the solvent removed under
reduced pressure. The residue (1.0 g) was flash chromato-
graphed on silica gel (toluene/ethyl acetate 3/1). Crystallization
from heptane gave the title compound 12 (250 mg, 0.78 mmol,
16%): mp 192-197 °C; [R]D -29.3° (c 1.0); 1H NMR (CDCl3) δ
0.87 (s, 3H, CH3), 0.94 (s, 3H, CH3), 2.02 (m, 2H), 2.27 (m,
1H), 2.78 (m, 1H), 4.05 (br s, 1H), 4.30 (m, 1H), 4.42 (m, 1H);
IR (KBr) 3436, 3353 (OH), 2929 (CH) cm-1. Anal. (C21H34O2‚
0.25H2O) C, H.
(3r,5r,13r,17r)-3-H yd r oxya n d r ost a n e -17-c a r b on i-
tr ile (14). To androsterone (8) (15 g, 52 mmol) in pyridine (150
mL) and ethanol (150 mL) was added hydroxylamine hydro-
chloride (15 g, 216 mmol) and the solution heated under reflux
for 5 h. The solution was concentrated to approximately 200
mL then cooled and water added to precipitate a colorless solid
which was washed with water and dried over anhydrous
sodium sulfate, affording androsterone oxime31 (30) (16.18 g,
52 mmol): 1H NMR (CDCl3) δ 0.80 (s, 3H, CH3), 0.90 (s, 3H,
CH3), 4.03 (s, 1H); IR (KBr) 3248 (OH), 2926 (CH), 1680 (Cd
N) cm-1
.
The oxime 30 (15 g, 49 mmol) was refluxed gently in acetic
anhydride (450 mL) and pyridine (750 mL) with stirring under
nitrogen for 12 h. The solution rapidly developed a dark color.
It was then allowed to cool, the solvent evaporated and the
residue azeotroped with toluene to give a black oily residue.
This was shaken for several min with diethyl ether and 5%
aqueous sodium carbonate, then filtered through Celite. The
granular black residue was washed thoroughly with several
portions of diethyl ether, and the combined organic extracts
(3r,5r,16â,17â)-16,17-(4H)-Dih yd r o-3-h yd r oxya n d r ost-
16-en o[17,16-b]oxeta n e (13). A stirred solution of (3R,5R)-
3-hydroxyandrostan-17-one (8) (4.91 g, 16.9 mmol), sodium
methoxide (3.65 g, 67.6 mmol), ethyl formate (41 mL) and
toluene (100 mL) was heated under reflux for 7 h, after which
a white solid precipitated. The reaction mixture was allowed
to cool and water (500 mL) added. After separation the aqueous

4124 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Anderson et al.
washed with brine, dried over sodium sulfate and the solvent
evaporated. The residue was taken up in diethyl ether,
adsorbed on alumina and left for 1 h. Elution with diethyl
ether and subsequent evaporation gave a creamy froth,
presumed to be crude (13R)-17-acetamidoandrost-16-ene ac-
etate (31) (11.5 g, 30.8 mmol).
The crude acetate 31 (11.5 g, 30.8 mmol) was refluxed in
methanol (1150 mL) and hydrochloric acid (2 N, 450 mL) for
2 h. The solvent was evaporated to half volume and cooled.
Addition of water gave a gum, which was washed by decanta-
tion with water. The product was taken up in diethyl ether
and washed with water then brine, and the solvent evaporated
to give a gum, which upon crystallization from diethyl ether/
heptane gave (3R,5R,13R)-3-hydroxyandrostan-17-one34 (32) as
creamy prisms (3.74 g, 12.9 mmol, 26% from 31): 1H NMR
(CDCl3) δ 0.62 (s, 3H, CH3), 0.97 (s, 3H, CH3), 2.33 (m, 1H),
4.04 (s, 1H); 13C NMR (CDCl3) δ 222.7, 66.3, 51.6, 50.8, 50.1,
38.5, 37.8, 36.1, 35.6, 33.8, 32.9, 32.1, 31.9, 28.8, 28.4, 25.2,
22.2, 21.2, 10.9.
torsional values of 165.6°, 160.0° and 153.2° for structures 1,
3 and 4, respectively.
For structure 2 additional rotations were also performed on
the C21-O21 bond, which was rotated through 360° in 30°
increments, leading to the generation of 864 conformations.
In this case four low-energy conformations were detected.
Conformations with C16-C17-C20-O20 torsional values of
-23.2° and 155.5° had equal lowest energies, while conforma-
tions, with C16-C17-C20-O20 torsional values of 84.9° and
44.0° had energy values ca. 0.5-1.5 kcal‚mol-1 higher.
The alignments shown in Figure 2 were produced by fitting
on all carbons in the B- and C-rings of each structure, using
the Chem-X rigid fitting procedure. (The various ways of
overlaying 5R- and 5â-steroids have been previously dis-
cussed.9) The structures were transferred via MACCS SD files
to Cerius2,36 which was used solely to render Figure 2 graphi-
cally.
Distance measurements were performed in Chem-X on the
alignments shown in Figure 2. The carbonyl oxygens of the
minima of structures 1-4 in Figure 2 are separated from each
other by distances < 0.6 Å. The epoxide oxygen of 10 is 0.8-
1.0 Å from the carbonyl oxygens, while the epoxide oxygen of
5 is 1.6-1.8 Å from the carbonyl oxygens.
To a stirred solution of potassium tert-butoxide (11.2 g, 100
mmol) in tert-butyl alcohol (110 mL) was added a solution of
steroid 32 (2.90 g, 10 mmol) in diglyme (29 mL) under nitrogen.
A solution of p-toluenesulfonylmethyl isocyanide (3.91 g, 20
mmol) in diglyme (40 mL) was then added over 1 h. After 4 h
more p-toluenesulfonylmethyl isocyanide (3.91 g, 20 mmol) in
diglyme (40 mL) was added over 1 h, and the mixture left to
stand overnight. The mixture was treated with dilute aqueous
sodium chloride followed by hydrochloric acid (2 M) until
acidic. The aqueous liquors were decanted from the resulting
brown gum which was dissolved in dichloromethane. This
solution was washed with brine, dried over anhydrous sodium
sulfate and the solvent removed under reduced pressure. The
brown residue (2.1 g) was flash chromatographed on silica gel
(toluene/ethyl acetate 2/1). Crystallization from acetonitrile
gave the title compound 14 (145 mg, 0.48 mmol, 5%): mp 167-
170 °C; [R]D -91.6° (c 0.6); 1H NMR (CDCl3) δ 0.70 (s, 3H,
CH3), 1.07 (s, 3H, CH3), 2.88 (t, 1H, J ∼ 9), 4.05 (s, 1H); 13C
NMR (CDCl3) δ 121.6, 66.3, 52.6, 52.2, 45.0, 38.6, 37.8, 36.1,
35.6, 33.5, 33.2, 32.5, 31.8, 28.9, 28.4, 26.4, 26.0, 19.9, 11.1;
The default Chem-X parameter files, Gasteiger charges, and
molecular mechanics force field were used throughout.
Ack n ow led gm en t. We thank Dr. Alexander Kasal
for providing a sample of steroid 5 and Organon Teknika
for financial support. Physical chemistry data were
provided by the Department of Analytical Chemistry,
Organon Laboratories Ltd., Newhouse U.K.
Refer en ces
(1) (a) Majewska, M. D.; Harrison, N. L.; Schwartz, R. D.; Barker,
J . L.; Paul, S. M. Steroid Hormone Metabolites Are Barbiturate-
Like Modulators of the GABA Receptor. Science 1986, 232,
1004-1007. (b) Majewska, M. D. Neurosteroids: Endogenous
Bimodal Modulators of the GABAA Receptor. Mechanism of
Action and Physiological Significance. Prog. Neurobiol. 1992, 38,
379-395. (c) Lambert, J . J .; Belelli, D.; Hill-Venning, C.; Peters,
J . A. Neurosteroids and GABAA receptor function. Trends
Pharmacol. Sci. 1995, 16, 295-303. (d) Orchinik, M.; McEwen,
B. Novel and Classical Actions of Neuroactive Steroids. Neu-
rotransmissions 1993, IX, 1-6.
(2) (a) Belelli, D.; Lan, N. C.; Gee, K. W. Anticonvulsant Steroids
and the GABA/Benzodiazepine Receptor-Chloride Ionophore
Complex. Neurosci. Biobehav. Rev. 1990, 14, 315-322. (b)
Myslobodsky, M. S. Pro- and Anticonvulsant Effects of Stress:
the Role of Neuroactive Steroids. Neurosci. Biobehav. Rev. 1993,
17, 129-139. (c) Kokate, T. G.; Svensson, B. E.; Rogawski, M.
A. Anticonvulsant Activity of Neurosteroids: Correlation with
γ-Aminobutyric Acid-Evoked Chloride Current Potentiation. J .
Pharmacol. Exp. Ther. 1994, 270, 1223-1229. (d) Gasior, M.;
Carter, R. B.; Goldberg, S. R.; Witkin, J . M. Anticonvulsant and
Behavioral Effects of Neuroactive Steroids Alone and in Con-
junction with Diazepam. J . Pharmacol. Exp. Ther. 1997, 282,
543-553. (e) Carter, R. B.; Wood, P. L.; Wieland, S.; Hawkinson,
J . E.; Belelli, D.; Lambert, J . J .; White, H. S.; Wolf, H. H.;
Mirsadeghi, S.; Tahir, S. H.; Bolger, M. B.; Lan, N. C.; Gee, K.
W. Characterization of the anticonvulsant properties of ganax-
olone (CCD 1042; 3R-hydroxy-3â-methyl-5R-pregnan-20-one), a
selective, high-affinity, steroid modulator of the gamma-ami-
nobutyric acid(A) receptor. J . Pharmacol. Exp. Ther. 1997, 280,
1284-1295.
(3) Rodgers, R. J .; J ohnson, N. J . Behaviorally Selective Effects of
Neuroactive Steroids on Plus-Maze Anxiety In Mice. Pharmacol.
Biochem. Behav. 1998, 59, 221-32.
(4) (a) Norberg, L.; Backstrom, T.; Wahlstrom, G. Anaesthetic
Effects of Pregnanolone in Combination with Allopregnanolone,
Thiopental, Hexobarbital and Flurazepam: an EEG Study in
the Rat. Br. J . Anaesth. 1999, 82, 731-737. (b) Sear, J . W.
Eltanolone: 50 Years On and Still Looking for Steroid Hypnotic
Agents! Eur. J . Anaesthesiol. 1998, 15, 129-132.
(5) (a) Harrison, N. L.; Simmonds, M. A. Modulation of the GABA
Receptor Complex by a Steroid Anaesthetic. Brain Res. 1984,
323, 287-292. (b) Hill-Venning, C.; Peters, J . A.; Callachan, H.;
Lambert, J . J .; Gemmell, D. K.; Anderson. A.; Byford, A.;
Hamilton, N.; Hill, D. R.; Marshall, R. J .; Campbell, A. C. The
Anaesthetic Action and Modulation of GABAA Receptor Activity
by the Novel Water-soluble Aminosteroid Org 20599. Neurop-
harmacology 1996, 35, 1209-1222. (c) Hogenkamp, D. J .; Tahir,
IR (KBr) 3501 (OH), 2930 (CH), 2236 (CN) cm-1. Anal. (C20H31
NO) C, H, N.
-
(3r,5r,17â)-17-Meth oxya n d r osta n -3-ol (16). To a stirred
suspension of (3R,5R,17â)-androstane-3,17-diol (15) (1.0 g, 3.4
mmol) in 1,2-dimethoxyethane (20 mL) was added sodium
hydride (164 mg, 6.8 mmol). After 10 min methyl iodide (0.22
mL, 3.5 mmol) was added and the mixture left to stir under
nitrogen for 60 h. The mixture was then poured into dilute
aqueous sodium carbonate and the precipitated solid filtered
off. This material was dissolved in dichloromethane and the
solution washed with water and then dried over anhydrous
sodium sulfate. The solvent was removed under reduced
pressure and the residue (1.2 g) was chromatographed on silica
gel (gradient elution, dichloromethane/diethyl ether). Crystal-
lization from acetone gave the title compound 1617 (180 mg,
0.59 mmol, 17%): mp 167-168 °C; 1H NMR (CDCl3) δ 0.75 (s,
3H, CH3), 0.78 (s, 3H, CH3), 1.89 (m, 1H), 2.02 (m, 1H), 3.22
(t, 1H, J ∼ 8), 4.04 (s, 1H); IR (KBr) 3299 cm-1 (OH). Anal.
(C20H34O2) C, H.
Ca lcu la tion s. All the structures studied were constructed
using Chem-X35 with the Chem-X 2D and 3D building tools.
Each structure was then subjected to geometry optimization
using the default Gasteiger charges and optimization param-
eters.
Conformational analyses were performed using Chem-X.35
For each structure a systematic grid search was carried out
by rotating the C17-C20 bond through 360° in 5° increments.
Structures were geometry optimized after each step. Global
minima were found at C16-C17-C20-O20 torsional values
of -34.3°, -34.8° and -21.8° for structures 1, 3 and 4,
respectively. (The measured angle reported for alfaxalone 4
is -21°.8) The analyses showed that torsional angles 10° higher
or lower than the minima could be achieved an energy cost of
ca. 0.5 kcal in each case. Slightly higher (+0.5-1.5 kcal‚mol-1
local minima were also detected at C16-C17-C20-O20
)

Anesthetic Steroids Modulate GABAA Receptors
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4125
S. H.; Hawkinson, J . E.; Upasani, R. B.; Alauddin M.; Kim-
brough, C. L.; Acosta-Burruel, M.; Whittemore, E. R.; Woodward,
R. M.; Lan, N. C.; Gee, K. W.; Bolger, M. B. Synthesis and in
Vitro Activity of 3â-Substituted-3R-hydroxypregnan-20-ones:
Allosteric Modulators of the GABAA Receptor. J . Med. Chem.
1997, 40, 61-72. (d) Upasani, R. B.; Yang, K. C.; Acosta-Burruel,
M.; Konkoy, C. S.; McLellan, J . A.; Woodward, R. M.; Lan, N.
C.; Carter, R. B.; Hawkinson, J . E. 3R-Hydroxy-3â-(phenylethy-
nyl)-5â-pregnan-20-ones: Synthesis and Pharmacological Activ-
ity of Neuroactive Steroids with High Affinity for GABAA
Receptors. J . Med. Chem. 1997, 40, 73-84. (e) Anderson, A.;
Boyd, A. C.; Byford, A.; Campbell, A. C.; Gemmell, D. K.;
Hamilton, N. M.; Hill, D. R.; Hill-Venning, C.; Lambert, J . J .;
Maidment, M. S.; May. V.; Marshall, R. J .; Peters, J . A.; Rees,
D. C.; Stevenson, D.; Sundaram, H. Anesthetic Activity of Novel
Water-Soluble 2â-Morpholinyl Steroids and Their Modulatory
Effects at GABAA Receptors. J . Med. Chem. 1997, 40, 1668-
1681. (f) Xue, B. G.; Wittemore, E. R.; Park, C. H.; Woodward,
R. M.; Lan, N. C.; Gee, K. W. Partial Agonism by 3R,21-
Dihydroxy-5â-pregnan-20-one at the γ-Aminobutyric AcidA Re-
ceptor Neurosteroid Site. J . Pharmacol. Exp. Ther. 1997, 281,
1095-1101. (g) Hawkinson, J . E.; Drewe, J . A.; Kimbrough, C.
L.; Chen, J .-S.; Hogenkamp, D. J .; Lan, N. C.; Gee, K. W.; Shen,
K.-Z.; Whittemore, E. R.; Woodward, R. M. 3R-Hydroxy-3â-
trifluoromethyl-5R-pregnan-20-one (Co 2-1970): A Partial Ago-
nist at the Neuroactive Steroid Site of the γ-Aminobutyric acidA
Receptor. Mol. Pharmacol. 1996, 49, 897-906. (h) Hu, Y.;
Wittmer, L. L.; Kalkbrenner, M.; Evers, A. S.; Zorumski, C. F.;
Covey, D. F. Neurosteroid analogues. Part 5. Enantiomers of
neuroactive steroids and benz[e]indenes: total synthesis, elec-
trophysiological effects on GABAA receptor function and anes-
thetic actions in tadpoles. J . Chem. Soc., Perkin Trans. 1, 1997,
3665-3671. (i) Hawkinson, J . E.; Acosta-Burruel, M.; Yang, K.
C.; Hogenkamp, D. J .; Chen, J .-S.; Lan, N. C.; Drewe, J . A.;
Whittemore, E. R.; Woodward, R. M.; Carter, R. B.; Upasani, R.
B. Substituted 3â-Phenylethynyl Derivatives of 3R-Hydroxy-5R-
pregnan-20-one: Remarkably Potent Neuroactive Steroid Modu-
lators of γ-Aminobutyric acidA Receptors. J . Pharmacol. Exp.
Ther. 1998, 287, 198-207.
(17) Babcock, J . C.; Campbell, J . A.; Lobl, T. J . Androstane deriva-
tives. Ger. Offen. DE 2940132; Chem. Abstr. 1981, 94, 65946y.
(18) Ruzicka, L.; Goldberg, M. W.; Meyer, J . Sex hormones. IV.
Derivatives of Synthetic Androsterone and its Stereoisomers.
Helv. Chim. Acta 1935, 18, 210-218.
(19) Hevers, W.; Luddens, H. The Diversity of GABAA Receptors:
Pharmacological and Electrophysiological Properties of GABAA
Channel Subtypes. Mol. Neurobiol. 1998, 18, 35-86.
(20) Schmid, G.; Sale, R.; Bonanno, G.; Raiteri, M. Neurosteroids May
Differentially Affect the Function of Two Native GABAA Receptor
Subtypes in the Rat Brain. Naunyn-Schmiedeberg’s Arch. Phar-
macol. 1998, 357, 401-407.
(21) Rick, C. E.; Ye, Q.; Finn, S. E.; Harrison, N. L. Neurosteroids
act on the GABAA receptor at sites on the N-terminal side of
the middle of TM2. NeuroReport 1998, 9, 379-383.
(22) Belelli, D.; Pistis, M.; Peters, J . A.; Lambert, J . J . The Interaction
of General Anaesthetics and Neurosteroids with GABAA and
Glycine Receptors. Neurochem. Int. 1999, 34, 447-452.
(23) Harrison, N. L.; Majewska, M. D.; Harrington, J . W.; Barker, J .
L. Structure-Activity Relationships for Steroid Interaction with
the γ-Aminobutyric AcidA Receptor Complex. J . Pharmacol. Exp.
Ther. 1987, 241, 346-353.
(24) (a) Rodgers-Neame, N. T.; Covey, D. F.; Hu, Y.; Isenberg, K. E.;
Zorumski, C. F. Effects of a Benz[e]indene on GABA-Gated
Chloride Currents in Cultured Postnatal Rat Hippocampal
Neurons. Mol. Pharmacol. 1992, 42, 952-957. (b) Covey, D. F.;
Hu, Y.; Bouley, M. G.; Holland, K. D.; Rodgers-Neame, N. T.;
Isenberg, K. E.; Zorumski, C. F. Modulation of GABAA Receptor
Function by Benz[e]indenes and Phenanthrenes. J . Med. Chem.
1993, 36, 627-630. (c) Hu, Y.; Zorumski, C. F.; Covey, D. F.
Neurosteroid Analogues: Structure-Activity Studies of Benz-
[e]indene Modulators of GABAA Receptor Function. 1. The Effect
of 6-Methyl Substitution on the Electrophysiological Activity of
7-Substituted Benz[e]indene-3-carbonitriles. J . Med. Chem.
1993, 36, 3956-3967.
(25) (a) Belelli, D.; Lambert, J . J .; Peters, J . A.; Gee, K. W.; Lan, N.
C. Modulation of Human Recombinant GABAA Receptors by
Pregnanediols. Neuropharmacology 1996, 35, 1223-1231. (b)
McCauley, L. D.; Liu, V.; Chen, J .-S.; Hawkinson, J . E.; Lan, N.
C.; Gee, K. W. Selective Actions of Certain Neuroactive Preg-
nanediols at the γ-Aminobutyric Acid Type A Receptor Complex
in Rat brain. Mol. Pharmacol. 1995, 47, 354-362.
(26) Rudolph, U.; Crestani, F.; Benke, D.; Brunig, I.; Benson, J . A.;
Fritschy, J .-M.; Martin, J . R.; Bluethmann, H.; Mohler, H.
Benzodiazepine actions mediated by specific γ-aminobutyric
acidA receptor subtypes. Nature 1999, 401, 796-800.
(6) Hawkinson, J . E.; Kimbrough, C. L.; Belelli, D.; Lambert, J . J .;
Purdy, R. H.; Lan, N. C. Correlation of Neuroactive Steroid
Modulation of [35S]tert-Butylbicyclophosphorothionate and [3H]-
Flunitrazepam Binding and γ-Aminobutyric AcidA Receptor
Function. Mol. Pharmacol. 1994, 46, 977-985.
(7) Duax, W. L. In Molecular Structure and Biological Activity of
Steroids; Bohl, M., Duax, W. L., Eds.; CRC Press: Boca Raton,
1992; pp 2-7.
(8) Duax, W. L.; Griffin, J . F.; Rohrer D. C. Conformation of
Progesterone Side Chain: Conflict between X-ray Data and
Force-Field Calculations. J . Am. Chem. Soc. 1981, 103, 6705-
6712.
(27) Gyermek, L.; Iriarte, J .; Crabbe, P. Steroids. CCCX. Structure-
Activity Relation of Some Steroidal Hypnotic Agents. J . Med.
Chem. 1968, 11, 117-125.
(9) Purdy, R. H.; Morrow, A. L.; Blinn J . R.; Paul, S. M. Synthesis,
Metabolism, and Pharmacological Activity of 3R-Hydroxy Ste-
roids Which Potentiate GABA-Receptor-Mediated Chloride Ion
Uptake in Rat Cerebral Cortical Synaptoneurosomes. J . Med.
Chem. 1990, 33, 1572-1581.
(10) Bolger, M. B.; Wieland, S.; Hawkinson, J . E.; Xia, H.; Upasani,
R.; Lan, N. C. In Vitro and In Vivo Activity of 16,17-Dehydro-
epipregnanolones: 17,20-Bond Torsional Energy Analysis and
D-ring Conformation. Pharm. Res. 1996, 13, 1488-1494.
(11) Fajkosˇ, J .; Sˇorm, F. On Steroids. XXXV. The Synthesis of
Epimeric 16-Bromo Derivatives of Androsterone and Testoster-
one. Collect. Czech. Chem. Commun. 1959, 24, 766-785.
(12) Gustafsson, J . A. Formation of 16,17-Dihydroxylated C19-
Steroids from 16-Dehydro C19-Steroids in Liver Microsomes
from Male and Female Rats. Biochim. Biophys. Acta 1973, 296,
179-188.
(13) Prelog, V.; Ruzicka, L.; Meister, P.; Wieland, P. Studies on the
Synthesis of 16-Dehydro Steroids. Helv. Chim. Acta 1945, 28,
618-627.
(14) Fajkosˇ, J .; Sˇorm, F. Steroids. IX. Synthesis of 3R-hydroxy-16-
oxoandrostane and 3,16-dioxo-4-androstene. Collect. Czech. Chem.
Commun. 1954, 19, 349-356.
(15) De Marcano, D.; Del Giorgio, J . F.; Evans, J . M.; Hurtado, E. J .;
Kohout, L.; Osorio, E.; Vitolo, M. J . D-Homoandrostanes. 3.
Incubation of some D-Homo-5R-androstanes with Aspergillus
Ochraceus. J . Org. Chem. 1978, 43, 3960-3963.
(16) Boar, R. B.; J etuah, F. K.; McGhie, J . F.; Robinson, M. S.; Barton,
D. H. R. An Improved Synthesis of 13-epi-Androstanes and of
13-epi-Oestranes. J . Chem. Soc., Perkin Trans. 1 1977, 2163-
2165.
(28) Im, W. B.; Blakeman, D. P.; Davis, J . P.; Ayer, D. E. Studies on
the Mechanism of Interactions between Anesthetic Steroids and
γ-Aminobutyric AcidA Receptors. Mol. Pharmacol. 1990, 37,
429-434.
(29) Phillipps, G. H. Structure-Activity Relationships in Steroidal
Anaesthetics. J . Steroid Biochem. 1975, 6, 607-613.
(30) (a) Belelli, D.; Gee, K. W. 5R-Pregnan-3R,20R-diol Behaves Like
a
Partial Agonist in the Modulation of GABA-Stimulated
Chloride in Uptake by Synaptoneurosomes. Eur. J . Pharmacol.
1989, 167, 173-176. (b) Bolger, M. B.; Gee, K. W.; Lan, N. T.
Y.; Belelli, D.; Mirsadeghi, S.; Purdy, R. H.; Tahir, S. H.
Preparation of Hydroxysteroid Modulators of the γ-Aminobutyric
Acid Receptor-Chloride Ionophore Complex. PCT Int. Appl. WO
9303732; Chem. Abstr. 1993, 119, 49738m.
(31) Butenandt, A.; Dannenbaum. Androsterone, a crystalline male
sex hormone. II. Its chemical characterization. Z. Physiol. Chem.
1934, 229, 185-191.
(32) Zamecnik, J .; Barbe, G.; Moger, W. H.; Armstrong, D. T.
Radioimmunoassays for androsterone, 5R-androstane-3R,17â-
diol and 5R-androstane-3â,17â-diol. Steroids 1977, 30, 679-689.
(33) Ringold, H. J .; Mancera, O. Androstenopyrimidines. U.S. Patent
US 3071577; Chem. Abstr. 1964, 61, 4430g.
(34) Butenandt, A.; Poschmann, L. Lumiandrosterone. Chem. Ber.
1944, 77B, 394-397.
(35) The Chem-X Suite of Programs; CDL, Oxford Molecular Group,
Oxford, U.K.
(36) Cerius2; Molecular Simulations Inc., San Diego, CA.
J M000977E