Neurosteroid analogues. 4. The effect of methyl substitution at the C-5 and C-10 positions of neurosteroids on electrophysiological activity at GABA(A) receptors

Article abstract of DOI:10.1021/jm960304p

A series of analogues of the neuroactive steroids 3α-hydroxy-5α- pregnan-20-one and 3α-hydroxy-5β-pregnan-20-one were studied to elucidate the mode of binding of 5α- and 5β-reduced steroids to steroid binding sites on GABA(A) receptors. Analogues which were either 3α-hydroxy-20-ketosteroids or 3α-hydroxysteroid-17β-carbonitriles and which contained various methyl group substitution patterns at C-5 and C-10 were prepared. Evaluations utilized whole-cell patch clamp electrophysiological methods carried out on cultured rat hippocampal neurons, and the results obtained with the rigid 17β-carbonitrile analogs were analyzed using molecular modeling methods. The molecular modeling results provide a rationale for the observation that the configuration of the hydroxyl group at C-3 is a greater determinant of anesthetic potency than the configuration of the A,B ring fusion at C-5. The electrophysiological results identify steric restrictions for the space that can be occupied in 5α- and 5β-reduced steriod modulators of GABA(A) recepters in the regions of space proximate to the steroid C-5, C-10, and possibly C-4 positions. This information is useful for the development of nonsteroidal analogues that can modulate GABA(A) receptors via interactions at steroid binding sites.

Full text of DOI:10.1021/jm960304p

4218
J . Med. Chem. 1996, 39, 4218-4232
Neu r oster oid An a logu es. 4. Th e Effect of Meth yl Su bstitu tion a t th e C-5 a n d
C-10 P osition s of Neu r oster oid s on Electr op h ysiologica l Activity a t GABAA
Recep tor s
Mingcheng Han,† Charles F. Zorumski,‡,§ and Douglas F. Covey*,†
Departments of Molecular Biology & Pharmacology, Anatomy & Neurobiology, and Psychiatry, Washington University School
of Medicine, 660 South Euclid Avenue, St. Louis, Missouri 63110
Received April 24, 1996X
A series of analogues of the neuroactive steroids 3R-hydroxy-5R-pregnan-20-one and 3R-hydroxy-
5â-pregnan-20-one were studied to elucidate the mode of binding of 5R- and 5â-reduced steroids
to steroid binding sites on GABAA receptors. Analogues which were either 3R-hydroxy-20-
ketosteroids or 3R-hydroxysteroid-17â-carbonitriles and which contained various methyl group
substitution patterns at C-5 and C-10 were prepared. Evaluations utilized whole-cell patch
clamp electrophysiological methods carried out on cultured rat hippocampal neurons, and the
results obtained with the rigid 17â-carbonitrile analogs were analyzed using molecular modeling
methods. The molecular modeling results provide a rationale for the observation that the
configuration of the hydroxyl group at C-3 is a greater determinant of anesthetic potency than
the configuration of the A,B ring fusion at C-5. The electrophysiological results identify steric
restrictions for the space that can be occupied in 5R- and 5â-reduced steriod modulators of
GABAA receptors in the regions of space proximate to the steroid C-5, C-10, and possibly C-4
positions. This information is useful for the development of nonsteroidal analogues that can
modulate GABAA receptors via interactions at steroid binding sites.
The discovery that anesthetic steroids are potent
allosteric modulators of GABAA receptor (γ-aminobu-
tyric acid type A receptor) function1 has led to major
advances in understanding the pharmacology of anes-
thetic steroids.2 However, most of the medicinal chem-
istry studies that established the structure-activity
relationships for steroid-induced anesthesia relied on
correlations of structure with anesthetic activity in mice
or rats and were completed prior to the discovery that
these compounds altered GABAA receptor function.3
Hence, new opportunities for an increased understand-
ing of the earlier derived structure-activity relation-
ships now exist.
Among the more striking structure-activity relation-
ships established in the earlier animal behavioral
studies of anesthetic steroids was one indicating that
the stereochemistry of the steroid A,B ring fusion had
a relatively minor effect on anesthetic activity. For
example, 3R-hydroxy-5R-pregnan-20-one (1) and 3R-
hydroxy-5â-pregnan-20-one (2) have equivalent anes-
molecular recognition of these steroids at their binding
sites would be useful. Accordingly, we report the results
of a study involving the electrophysiological evaluation
of analogues of steroids 1 and 2 which contain various
methyl group substitution patterns at the C-5 and C-10
positions. Additional analogues in which the 17â-
subsitutent has been changed to a carbonitrile group
have also been studied. The goal of the study was to
determine how steric bulk in the region of the steroid
A,B ring fusion altered the pharmacological actions of
5R- and 5â-reduced steroid modulators of the GABAA
receptors found in cultured rat hippocampal neurons so
that a better understanding of the mode of binding of
these steroids to GABAA receptors can be achieved.5
Ch em istr y
P r ep a r a tion of 3r-Hyd r oxy-19-n or -5â-ster oid s.
Using reaction conditions similar to those reported
previously,6 catalytic hydrogenation of 19-norsteroid 3
in alkaline 2-propanol gave steroid 5 in 60% yield after
purification by column chromatography (Scheme 1).
Lesser amounts of purified products 4 (7%) and 7 (16%)
along with a mixture of products 5 and 6 (10%) also were
obtained. In general, the ratio of 5â:5R-reduced steroids
was ∼9:1. The combined 5â-reduced steroids 5-7 were
then oxidized with J ones reagent7 in acetone to known
dione 8,6 and the dione was reduced with K-Selectride
(Aldrich) in THF at -78 °C to give the known 3â-
hydroxy-5â-estran-17-one (9a )6 in 93% yield. The hy-
droxyl group of steroid 9a was acetylated with AcOAc/
pyridine to yield known product 9b6 (91%), and this
product was reacted with diethyl cyanophosphonate and
LiCN in THF to yield a mixture of the uncharacterized
intermediate cyanophosphates 10a and 10b (97%) which
were reduced with SmI2/THF and MeOH to give a 1:3
ratio of carbonitriles 11a and 11b (90%). An attempt
to introduce the carbonitrile group into steroid 9a by
this method was unsatisfactory due to the reaction of
thetic activity in mice and rats even though the confor-
mations of the steroids differ greatly.3f Explanations
for how this conformational difference could be accom-
modated upon binding of these steroids to a common
site thereby producing this behavioral result have
appeared,3g,4 but additional information regarding the
† Department of Molecular Biology & Pharmacology.
‡ Department of Anatomy & Neurobiology.
§ Department of Psychiatry.
X Abstract published in Advance ACS Abstracts, September 15, 1996.
S0022-2623(96)00304-4 CCC: $12.00 © 1996 American Chemical Society

Neurosteroid Analogues
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21 4219
Sch em e 1a
a
(a) KOH, i-PrOH, Pd-C, H2; (b) J ones reagent, acetone; (c) K-Selectride, THF, -78 °C; (d) AcOAc, pyridine, 120 °C; (e) NCP(O)(OEt)2,
LiCN, THF; (f) SmI2, MeOH, THF; (g) NaOH, K2CO3, MeOH, H2O, reflux; (h) NaBH4, EtOH; (i) CH3MgCl, THF, reflux.
Ch a r t 1
the hydroxyl group with the diethyl cyanophosphonate.
Since the originators of this method reported that the
sterically hindered 17â-hydroxy group of 17â-hydroxyan-
drostan-3-one was unreactive,8 it appears that the
absence of steric hindrance explains the reactivity of the
hydroxyl group of steroid 9a .
The mixture of steroids 11a and 11b was saponified
to steroids 11c and 11d (84%), and the products were
separated by HPLC. The 17â-carbonitrile 11d was then
oxidized with J ones reagent in acetone to steroid 12
(97%), and the 3-keto group of this product was reduced
with NaBH4 in EtOH to give a mixture of 17â-carboni-
triles 13a and 11d . Purification by column chromatog-
raphy gave the reduction products in isolated yields of
69% (13a ) and 24% (11d ). It was anticipated that
steroid 13a would be the major product because NaBH4
reduction of the 3-keto group of (5â,17R)-17-hydroxy-
19-norpregn-20-yn-3-one was shown previously to give
(3R,5â,17R)-19-norpregn-20-yne-3,17-diol as the major
product.9 Finally, reaction of 17â-carbonitrile 13a with
CH3MgBr in THF gave an 80% yield of an ∼1:10
mixture of product 14a and previously reported product
14b,3f respectively. These products were separated by
HPLC.
lated to give known steroid 15b10 (98%), and the 17-
carbonitrile group was introduced using diethyl cyan-
ophosphonate followed by treatment with SmI2/THF
and MeOH. In this instance, partial hydrolysis of the
acetyloxy group occurred during isolation of the products
so that a mixture of products 16a , 16b and 17a , 17b
was obtained. Column chromatogaphy was used to
separate the mixture of products 16a and 16b (28%)
from products 17a (16%) and 17b (20%). Pure samples
of 16a and 16b were obtained by HPLC, and these
separated products were then hydrolyzed to 17a and
P r ep a r a tion of 3r-Hyd r oxy-5â-ster oid s a n d 3r-
Hydr oxy-19-n or -5r-ster oids. The same methods used
to prepare the steroids described in Scheme 1 were
applied for the preparation of the compounds shown in
Chart 1. Commercially available steroid 15a was acety-

4220 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21
Han et al.
Sch em e 2a
Ch a r t 2
reaction. Although it was possible to obtain pure
samples of steroids 6 and 19b by HPLC for product
characterization, this mixture was not readily separated
and it was used without purification. Reaction of the
product mixture with CH2I2, Zn-Cu couple gave a
complex mixture containing starting materials, byprod-
ucts, and the desired 4,5-methanosteroids 20a and 20b.
Since attempts to isolate the 4,5-methanodiols by re-
crystallization or HPLC on a silica column were unsuc-
cessful, this product mixture was oxidized with J ones
reagent to a product mixture containing 4,5-methano-
diones 21a and 21b. Treatment of this product mixture
in CH2Cl2 with O3 at -78 °C converted compounds in
the mixture containing double bonds to more polar
uncharacterized products that were easily removed by
column chromatography. Chromatography also yielded
the saturated diketone 8 (21% overall yield from steroid
3) and an unresolved mixture of 4,5-methanodiones 21a
and 21b (14% overall yield from steroid 3). Recrystal-
lization of the 4,5-methanodiones gave a pure sample
of steroid 21a . A pure sample of steroid 21b was not
obtained. A mixture of 4,5-methanodiones 21a and 21b
was opened with Li-liquid NH3 to give a mixture of
5-methyl steroids 22a -c. Steroids 22a and 22b were
isolated as a mixture (74% yield) by column chroma-
tography along with pure steroid 22c (9% yield). Simi-
larly, the 5R-methyl steroid 22a was obtained by
reductive opening of steroid 21a with Li-liquid NH3.
J ones oxidation of 22c also gave 5R-methyl steroid 22a
(82% yield), thus confirming that the 5-methyl group
had the same configuration in both compounds 22a and
22c and that both of these compounds were derived from
4,5-methanosteroid 21a . The evidence upon which the
assignments of configuration were made for the 4,5-
methano groups and the 5-methyl groups in these
steroids (as well as other 4,5-methano and 5-methyl
steroids shown in Charts 2 and 3) is reported following
the presentation of all synthetic methods.
a
(a) K-Selectride, THF, -78 °C; (b) CH2I2, Cu-Zn, I2, EtOEt;
(c) J ones reagent, acetone; (d) O3, CH2Cl2, -78 °C; (e) Li, liquid
NH3, THF; (f) AcOAc, pyridine, 120 °C.
17b, respectively. The 3â-hydroxysteroid 17c was
prepared in 95% yield by carrying out a modified
Mitsunobu reaction on 3R-hydroxysteroid 17b.11
The 19-norsteroid 18a was prepared as described
previously.12 Reaction of 19-norsteroid 18a with CH3-
MgBr in THF gave the previously reported 19-norsteroid
18b13 (64%) as well as unreacted starting material
(27%). No product having the C-17 side chain in the R
configuration was isolated.
P r ep a r a tion of 5r- a n d 5â-Meth yl 3r-Hyd r oxy-
19-n or ster oid s. The methodology used for the intro-
duction of 5-methyl groups into the steroids required
for this study is based on that reported by Dauben et
al.14,15 These investigators reported that the reaction
of Simmons-Smith reagent16 (CH2I2, Zn-Cu couple)
with 3-hydroxy 4-ene steroids in the androstene and
cholestene series gave 3-hydroxy-3′,4-dihydrocycloprop-
[4,5]steroids (trivially referred to as 4,5-methanosteroid
derivatives). The 4,5-methano group was demonstrated
to form syn to the hydroxy group in the 3-hydroxy 4-ene
steroid substrate. Oxidation of the 3-hydroxy group and
reductive opening of the resultant 3-keto-4,5-methanos-
teroid using Li-liquid NH3 was demonstrated to give
3-keto-5-methylsteroid products. Thus, this overall
sequence yields 3-keto-5R-methylsteroids from 3R-hy-
droxy 4-ene steroids and 3-keto-5â-methylsteroids from
3â-hydroxy 4-ene steroids.
Reduction of a mixture of compounds 22a and 22b
with K-Selectride at -78 °C for 6 h gave a mixture of
3-hydroxysteroids 23a and 23b in 94% yield. Although
compound 23a can be isolated from this mixture in
minor amounts by recrystallization, it is more conve-
nient to acetylate the mixture and separate the acety-
loxy compounds 24a (50% yield) and 24b (33% yield)
Steroid 3 was reduced with K-Selectride at -78 °C
for 15 h to afford a mixture of known unsaturated diols
19a and 19b,17 saturated alcohol 6, and recovered
steroid 3 (Scheme 2). Steroid 19b was the major
product (∼60% of the total material isolated) from the

Neurosteroid Analogues
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21 4221
Ch a r t 3
P r ep a r a tion of 5r- a n d 5â-Meth yl-3r-h yd r ox-
yp r egn a n -20-on es. These compounds were prepared
from progesterone by a route analogous to that outlined
previously in Scheme 2. Reduction of progesterone (34,
Chart 3) with K-Selectride in THF at -78 °C for 5 h
gave a 98% yield of a 1:1 mixture of the known diols
35a and 35b.18,19 A portion of this diol mixture was
separated by HPLC for product characterization, and
the remaining material was converted with CH2I2 and
Zn-Cu couple in refluxing EtOEt for 3 d into a mixture
containing 4,5-methanodiols 36a and 36b and other
byproducts. This product mixture was not purified or
characterized, but was oxidized with J ones reagent and
treated with O3 in CH2Cl2 to facilitate isolation of the
desired 4,5-methanodiones 37a and 37b as an insepa-
rable mixture in 27% yield. Reductive opening of the
4,5-methanodione mixture with Li in liquid NH3 gave
the 20-hydroxy 3-ones 38a , 38b and 39a , 39b and the
3,20-diones 40 and 41. Separation of the 20-hydroxy
3-ones (38% yield) from the 3,20-diones (53% yield) was
readily achieved by column chromatography. Pure
samples of compounds 39a and 39b were obtained by
this procedure along with additonal mixtures of 20-
hydroxy 3-ones. Pure samples of compounds 40 and 41
were obtained by further purification using HPLC. The
remaining mixture of 20-hydroxy 3-ones was oxidized
with J ones reagent and then reduced with K-Selectride
to obtain compounds 42 and 43b in yields of 12% and
54%, respectively. Steroid 41 was reduced with NaBH4
in EtOH for 1 h at room temperature to obtain a mixture
of the epimeric 3-hydroxy 20-ones which was separated
by HPLC to give 43a and 43b in yields of 22% and 57%,
respectively.
by HPLC. Pure samples of steroids 23a and 23b are
then readily obtained by hydrolysis after separation of
acetyloxy compounds 24a and 24b, respectively.
Using the earlier described two-step cyanation pro-
cedure, compound 24a was converted in 93% overall
yield into a mixture of the 17R- and 17â-carbonitriles
25a and 25b, respectively (Chart 2). Removal of the
acetyloxy group from these unresolved epimeric 17-
carbonitriles by saponification gave a 1:1 mixture of the
epimeric 17-carbonitriles 26a and 26b (93% yield),
which was separated into its two pure components by
HPLC. A 1:1 mixture of the epimeric 17-carbonitriles
25a and 25b also was reacted with CH3MgBr in THF
at reflux for 48 h to give a 1:4 mixture (90% yield) of
the C-17 epimeric 3R-hydroxy-5R-methyl-19-norpreg-
nan-20-ones 27a and 27b. These 20-ketosteroids were
separated by HPLC.
Similarly, the two-step cyanation procedure was used
to convert compound 24b in 95% overall yield into a 1:1
mixture of the 17R- and 17â-carbonitriles 28a and 28b,
respectively (Chart 2). Saponification of these com-
pounds (97% yield), followed by HPLC separation, gave
17R-carbonitrile 29a and 17â-carbonitrile 29b. J ones
oxidation of compound 29a gave 3-ketosteroid 30a (91%)
and J ones oxidation of compound 29b gave 3-ketosteroid
30b (95%). The NaBH4 reduction of steroid 30a gave a
mixture of C-3 epimeric alcohols from which compounds
29a and 31 were isolated by HPLC in yields of 25% and
45%, respectively. Likewise, NaBH4 reduction of steroid
30b gave a mixture of C-3 epimeric alcohols from which
compounds 29b and 32 were isolated by HPLC in yields
of 46% and 50%, respectively. A 1:1 mixture of carbo-
nitriles 29b and 32 also was reacted with CH3MgBr in
THF at reflux for 24 h to give a mixture (64% yield) of
the C-3 epimeric 3-hydroxy-5â-methyl-19-norpregnan-
20-ones 33a and 33b. These 20-ketosteroids were
separated by HPLC.
Assigm en t of 5-Meth yl a n d C-3, C-17, a n d C-20
Con figu r a tion s. The assignments for the configura-
tion at C-5 of those steroids containing a 5-methyl group
are based on the NMR data summarized in Table 1. In
the NMR spectra of the known 5â-reduced steroids 5
and 8, the resonance for the C-4 axial proton is observed
as an apparent triplet (J ) 14.2 Hz) at δ 2.59 and 2.60,
respectively. The position of the resonance is due to the
deshielding effect of the C-3 carbonyl group and the
multiplicity of the resonance is explained by the identi-
cal geminal and vicinal (H4ax-H5) coupling constants.
Since no resonance is observed in this region of the NMR
spectra for the known 5R-reduced steroid 4 or 5R-
estrane-3,17-dione, the presence of this resonance in any
of the 3-keto-5-methylsteroids described herein indicates
that the methyl group has the 5â-configuration. As
required, the presence of a 5â-methyl group in the
steroid causes the C-4 axial proton resonance to appear
as a doublet. The results shown in Table 1 also indicate
that the 5â-methyl group shifts the C-4 axial proton
downfield by ∼0.1 ppm. The presence of 5â- and 19-
methyl groups results in a downfield shift for the C-4
axial proton of ∼0.4 ppm.
Deductive reasoning based on the NMR data sum-
marized in Table 1 and the earlier established stereo-
chemical correlations for the preparation and reductive
opening of 3-keto-4,5-methanosteroids by Dauben et al.
(vide supra) was used to correlate the configurations of
the 5-methyl steroids with their corresponding 4,5-
methanosteroid precursors. The 5R-steroids containing
3R-substituents were readily distinguished from the 5R-
steroids containing 3â-substitutents by the presence of
the vicinal trans-diaxial couplings found in the NMR

4222 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21
Han et al.
Ta ble 1. Chemical Shifts and Multiplicities of NMR Resonances Assigned as 4-Hax in 3-Keto-5â-steroids
chemical shift
4-Hax (δ)
multiplicity
4-Hax
coupling constant
4-Hax, J (Hz)
compd
R1
R2
R3
5
8
12
22b
30a
30b
39a
39b
41
H
H
H
H
H
H
Me
Me
Me
H
H
H
Me
Me
Me
Me
Me
Me
R-H, â-OH
dO
R-H, â-CN
dO
â-H, R-CN
R-H, â-CN
â-(20S)-Xa
â-(20R)-Xa
â-acetyl
2.59
2.60
2.57
2.71
2.69
2.69
3.01
2.94
2.95
triplet
triplet
triplet
doublet
doublet
doublet
doublet
doublet
doublet
14.2
14.2
14.2
14.3
14.3
14.7
14.7
14.4
14.9
a
X ) CH(OH)CH3.
Ta ble 2. Electrophysiological Effects of 3R-Hydroxysteroids on
resonances of the C-3 protons present in the latter class
of steroids. Thus, for the 5R-steroids, the 3R-substitu-
ents have the axial configuration and the equatorial C-3
protons appear as a broadened singlet with shoulders
(W1/2 ∼13 Hz). Conversely, for the 5â-steroids, the 3R-
substituents have the equatorial configuration and the
C-3 protons appear as complex multiplets (W1/2 ∼33 Hz).
As discussed previously for other neurosteroid ana-
logs,12,20 the configurational assignments for the C-17
substituents were based on an NMR analysis of the
coupling constants of the C-17 protons. For the purified
steroid diastereomer pair 38a and 38b, the assignment
of configuration at C-20 was based on the relative
chemical shifts of the C-18 methyl groups. In ac-
cordance with literature precedents,21 the diastereomer
having the C-18 methyl group resonance at the lower
field was assigned the 20R configuration.
GABAA Receptor Function
compd (10 µM) compd (10 µM)
c
compd R1
R2
R3
potentiationa gated currentb
N
1
Me R-H
acetyl 1111 ( 178d
26 ( 5
46 ( 13
0
6
18b
27b
42
H
H
R-H
acetyl
926 ( 101
110 ( 6
4
4
6
4
6
4
6
5
6
5
4
5
6
R-Me acetyl
Me R-Me acetyl
294 ( 19
1829 ( 272
1721 ( 182
92 ( 3
3 ( 3
44
Me R-H
CN
CN
32 ( 7
42 ( 9
0
18a
26b
2
14b
33a
43a
17b
13a
32
H
H
R-H
R-Me CN
Me â-H
H
H
Me â-Me acetyl
Me â-H
acetyl 1023 ( 192
acetyl
49 ( 8
110 ( 12
12 ( 2
24 ( 5
68 ( 18
41 ( 16
28 ( 5
â-H
â-Me acetyl
754 ( 84
652 ( 83
Electr op h ysiology
498 ( 121
1297 ( 251
1823 ( 360
452 ( 75
CN
CN
Voltage clamp recordings were obtained from cultured
postnatal rat hippocampal neurons using whole-cell
patch clamp methods.22 The results obtained with 3R-
hydroxysteroids are summarized in Table 2. The 3â-
hydroxysteroids, with the exception of compound 17c,
were not evaluated. Each compound was evaluated at
10 µM for its ability to potentiate 2 µM GABA-mediated
current and also evaluated at 10 µM for its ability to
initiate (gate) a current in the absence of GABA. The
concentrations of GABA and test compound used in the
potentiation experiments are different from those used
in our previous publications. The GABA concentration
was increased from 1 to 2 µM to facilitate the recording
of control currents. The test compound concentration
was increased from 1 to 10 µM for two reasons: (1) to
increase the sensitivity for the detection of, and dis-
crimination between, the activities of weakly potentiat-
ing compounds; and (2) because this concentration of
steroid 44 gives a maximal response. No other steroids
or benz[e]indene analogues that we have studied give
responses larger than those of steroid 44 under the
conditions used in this study. Recordings that are
representative of the potentiation responses observed
for the compounds are shown in Figure 1A-C for
compounds 44, 26b, and 32. A recording representative
of the direct gating caused by steroid 14b is shown in
Figure 1D. Compound 17c neither potentiated GABA-
H
H
â-H
â-Me CN
a
Percent response relative to current produced by GABA (2 µM).
To calculate the percentage response, the magnitude of the peak
current produced by 2 µM GABA plus 10 µM compound was
normalized with respect to the peak current produced by 2 µM
GABA alone on the same cell. A percentage response of 100%
reflects no change in the current compared to 2 µM GABA alone.
2 µM GABA is a concentration at the foot of the dose-response
curve in cultured postnatal rat hippocampal neurons. These
experiments were conducted at -60 mV, and compounds were
b
applied by pressure ejection for 500 ms. The compound gated
current reflects the peak current directly gated by 10 µM com-
pound in the absence of GABA compared to the response obtained
from the same cell in response to 2 µM GABA alone. c N ) Number
d
of cells examined. Values are the mean ( SEM.
mediated current (102 ( 5% relative to control ) 100%,
N ) 5) nor directly gated a current (0%, N ) 5).
Discu ssion
Reviews by Phillipps and co-workers on the structure-
activity relationships (SARs) of anesthetic steroids
indicate that the configuration of the hydroxyl group
at C-3 is a greater determinant of anesthetic potency
than the configuration of the A,B ring fusion at C-5.3g,h
A recent study in which the abilities of four different
diastereomers of 3-hydroxypregnan-20-one were evalu-

Neurosteroid Analogues
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21 4223
F igu r e 1. Effect of steroids in cultured hippocampal neurons.
(A-C) The traces show currents activated by 2 µM GABA in
the absence and presence of 10 µM 44 (A), 26b (B), and 32
(C). These compds were selected to show the range of steroid
effects on GABA currents. (D) The traces show currents
activated by 10 µM 14b in the absence of GABA. The response
of the same neuron to 2 µM GABA is shown for comparison.
Note the slow time course of the steroid activated current and
the marked increase in current fluctuations. In all traces,
hippocampal neurons were voltage clamped at -60 mV and
drugs were applied by pressure ejection for 500 ms as described
in the Experimental Section.
F igu r e 2. Molecular modeling analysis for steroids 44, 17b,
and 17c. A least squares fit program was used the superimpose
O-3, H-17, C-17, and the carbonitrile group atoms of steroids
17b and 17c to steroid 44. All five pairs of atoms for the fit
were weighted equally. The hydrogen atoms have been re-
moved from the structures for clarity. (A) Superimposition of
steroids 44 (yellow) and 17b (green). The intermolecular
distance between the oxygen atoms, nitrogen atoms, and C-19
atoms are 0.29, 0.11, and 1.42 Å, respectively. (B) Superim-
position of steroids 44 (yellow) and 17c (green). The intermo-
lecular distances between the oxygen atoms, nitrogen atoms,
and C-19 atoms are 0.23, 0.06, and 0.41 Å, respectively.
ated as inhibitors of [35S]TBPS binding to GABAA-
receptors reveals similar SARs. Thus, 3R-hydroxy-5R-
steroid 1 and 3R-hydroxy-5â-steroid 2 are both potent
displacers of [35S]TBPS with similar IC50s of 17 and 61
nM, respectively, and the corresponding 5R- and 5â-
steroid diastereomers containing a 3â-hydroxy group are
both weak displacers having IC50s of >105 and 1734 nM,
respectively.23 Finally, similar SARs are also apparent
from electrophysiological studies of steroid potentiation
of GABA-mediated chloride current24 and from studies
of steroid potentiation of muscimol-stimulated radioac-
tive chloride uptake in rat synaptoneurosomes.4
stitution patterns at C-5 and C-10, whose electrophysi-
ological activities at the heterogeneous GABAA receptors
found in cultured rat hippocampal neurons are re-
ported25 in Table 2, were used to investigate this
hypothesis.
Figure 3A shows the region of space occupied by the
5R-methyl group in the 5R-reduced 19-norsteroid 26b
and also shows that only minor perturbations in ring
conformation are caused by the 5R-methyl group. Ad-
ditionally, Figure 3B shows that the C-5 methyl group
of steroid 26b and C-4 of steroid 17c are similarly
located in three-dimensional space. Thus, the finding
that both steroids 26b and 17c are inactive in the
electrophysiological assays can be rationalized on the
basis of an unfavorable interaction between the steroids
and GABAA receptors in the region of space below the
A ring of the active steroid 44. Apparently, no realign-
ment that permits this steric interaction to be relieved
while maintaining an acceptable alignment of the
hydrogen bonding groups at either end of the steroid
with their bonding partners on the GABAA receptors of
cultured rat hippocampal neurons is possible for these
rigid compounds.
Steric factors in the region of space above the C-19
methyl group of 5R-steroid 44 also appear to be impor-
tant determinants of electrophysiological activity. As
indicated in Table 2, while 5R-steroid 44 and 5â-steroid
17b are both active compounds, steroid 44 is substan-
tially more active as a potentiator of GABA-mediated
current. Since the superimposition of steroids 44 and
17b shown in Figure 2A clearly shows that the C-19
methyl of steroid 17b is located above the C-19 methyl
of steroid 44, we hypothesized that the decreased
activity of steroid 17b might result from an unfavorable
Explanations for the above SARs remain to be elabo-
rated. In this regard, the results presented here have
been interpreted according to the molecular modeling
presented for the steroid 17â-carbonitriles shown in
Figures 2 and 3. In each panel of these figures two
steroids have been superimposed using a least squares
fit program to superimpose O-3, H-17, C-17, and the
carbonitrile group atoms of each steroid. All five pairs
of atoms for each fit were weighted equally, and 3R-
hydroxy-5R-steroid 44 is always the reference structure
to which other steroids were fit. As shown in Figure
2A,B, superimpositions of either the 3R-hydroxy-5â-
steroid 17b or the 3â-hydroxy-5â-steroid 17c with
steroid 44 in which the oxygens and carbonitriles occupy
nearly identical positions in three-dimensional space are
possible. An obvious spacial difference resulting from
the alignments shown in panels A and B is that the A
rings of steroid 17b and 17c are largely located on
opposite sides of the A ring of steroid 44. This difference
suggested to us that the availability of space at a steroid
binding site above, but not below, the A ring of steroid
44 might explain why steroid 17b is an active compound
whereas steroid 17c is an inactive compound. The
series of steroids with the various methyl group sub-

4224 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21
Han et al.
The reduced potentiating activity reported for the 5â-
methyl-19-norsteroid 32 in Table 2 indicates that an
unfavorable steric interaction results from adding a
methyl group in this position. Indeed, this unfavorable
interaction is stronger than that caused by the presence
of a 10â-methyl group in a 5â-steroid (compare the
results given in Table 2 for steroids 17b and 32). Thus,
the results obtained with steroids 17b and 32 suggest
that both C-5 and C-10 in 5â-steroids are in close contact
with the receptors.
An alignment for steroids 44 and 32 is shown in
Figure 3C. It is interesting to note in Figure 3C that
the 5â-methyl group in steroid 32 occupies a region of
space similar to that which would be occupied by a 4â-
substituent on steroid 44. Hence, we would predict that
the introduction of a 4â-methyl group into 5R-steroid
44 would produce a compound with diminished activity.
The previously reported result showing that the addition
of a 4â-OH group into the potent anesthetic steroid
alfaxalone (3R-hydroxy-5R-pregnane-11,20-dione) gave
an inactive compound3g,h supports the conclusion that
space for 4â-substituents may be restricted. However,
the difference in hyrophobicities between a methyl
group and a hydroxyl group, as well as the difference
in the methods of evaluation used, precludes a definitive
conclusion.
Comparisons similar to those made for these 17â-
carbonitriles can also be made for the less potent 20-
ketosteroids reported in Table 2. These comparisons
are, however, complicated by the ambiguities resulting
from the extra degree of freedom (rotation about the
C-17, C-20 bond) present in the latter series of com-
pounds. For example, whereas 17â-carbonitriles 44 and
17b differ in activity as potentiators of GABA-mediated
current, this activity is not significantly different for 20-
ketosteroids 1 and 2. The loss of this difference for the
20-ketosteroids could be reasonably explained by slightly
different conformations of the C-17 side chain which
make it possible for each steroid to achieve slightly
different, but similarly optimized, interactions with the
receptors.
F igu r e 3. Molecular modeling analysis for steroids 44, 26b,
17c, and 32. The modeling parameters were as described in
the legend to Figure 2. (A) Superimposition of steroids 44
(yellow) and 26b (green). The intermolecular distances be-
tween the oxygen atoms and nitrogen atoms are 0.45 and 0.14
Å, respectively. (B) Superimposition of steroids 26b (white)
and 17c (green) with steroid 44 (structure not shown). The
intermolecular distances between the oxygen atoms and
nitrogen atoms in steroids 26b and 17c are 0.22 and 0.09 Å,
respectively. The intermolecular distance between the 5R-
methyl carbon in steroid 26b and C-4 in steroid 17c is 0.52 Å.
(C) Superimposition of steroids 32 (green) and 44 (yellow). The
intermolecular distances between the oxygen atoms and
nitrogen atoms are 0.24 and 0.09 Å, respectively. The inter-
molecular distance between the 5â-methyl carbon in steroid
32 and C-19 in steroid 44 is 2.86 Å.
The potentiating effects of the 5R-reduced 20-keto-
steroids shown in Table 2 closely parallel those found
for the corresponding 17â-carbonitriles. The presence
of a methyl group at C-10 has no significant effect on
the activity of 5R-steroids (compare steroids 1 and 18b),
whereas the presence of methyl group at C-5 either
eliminates (steroid 27b) or greatly diminishes (steroid
42) this activity. A compelling reason for why com-
pound 42 is significantly more active than compound
27b is unclear. However, the conformational freedom
of the C-17 acetyl group found in these compounds may
be involved in the explanation of the results. It is
possible that the effect of a weak hydrophobic contact
provided by a methyl group at C-10 could be optimized
if the carbonyl group in the C-17 side chain is capable
of attaining a conformation that enhances this hydro-
phobic contact without diminishing its own hydrogen-
bonding interaction with the receptors.
steric interaction between the C-19 methyl group and
the receptors. To evaluate this hypothesis the 19-nor-
5â-steroid 13a was studied. The results reported in
Table 2 indicate that steroids 44 and 13a are equipotent
as potentiators of current and support the hypothesis
that only limited space is available for hydrophobic
groups above the region of space occupied by the C-19
methyl group of steroid 44. Thus, accommodation of the
C-19 methyl group in 5â-steroid 17b probably results
in less than optimal alignment of the 3R-hydroxy and/
or 17â-carbonitrile with their receptor hydrogen bonding
partners, thereby resulting in a decrease in biological
activity. These results should not be further interpreted
to imply that the C-19 methyl group of steroid 44 is a
critical determinant of activity. Clearly, the loss of a
hydrophobic contact involving this methyl group and a
GABAA receptor would not be expected to have as large
an effect on activity as would requiring the C-19 methyl
group in steroid 17b to be in a region of space occupied
by the receptor. Consequently, it is perhaps not sur-
prising that the 19-nor-5R-steroid 18a is equipotent to
5R-steroid 44 as a potentiator of GABA currents.
The effect of a methyl group substituent at C-10 on
the potentiation of GABA-mediated current is opposite
for 5â-reduced 17â-carbonitriles and the corresponding
20-ketosteroids (Table 2). Whereas for the 17â-carbo-
nitriles removing this C-10 substituent increases activ-
ity, for the 20-ketosteroids, removing this substituent
weakly decreases activity (compare steroids 2 and 14b).

Neurosteroid Analogues
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21 4225
otherwise) with a Perkin-Elmer 1710 FT-IR spectrophotom-
eter. Elemental analyses were carried out by M-H-W Labo-
ratories, Phoenix, AZ. Solvents were either used as purchased
or dried and purified by standard methodology. Flash chro-
matography was performed using silica gel (32-63 µm)
purchased from Scientific Adsorbents, Atlanta, GA. The 17â-
hydroxyestr-4-en-3-one was purchased from Steraloids, Wilton,
NH, and pregn-4-ene-3,20-dione was purchased from Sigma
Chemical Co., St. Louis, MO. The Econosil HPLC column was
purchased from Alltech Associates, Inc., Deerfield, IL. Steroid
44 was prepared as described previously.12
The reason for this result is not apparent. By contrast,
the decrease in GABA potentiation caused by the
presence of a 5â-methyl group is similar in both series
of compounds (steroids 32 and 33a ). Additionally, in
the 20-ketosteroid series, where the effect of the com-
bined C-5 and C-10 methyl substitutents was evaluated,
the C-10 methyl group did not significantly alter the
effect on potentiation caused by the C-5 methyl group
(compare steroids 33a and 43a ).
The results shown in Table 2 for the direct gating
effects of the steroids indicate that the effects of the
methyl substituents on the potentiation of GABA-
mediated currents are not always paralleled by similar
effects on the direct gating actions of the steroids. We
have observed a similar lack of parallel changes in these
two functional assays in a previous study of benz[e]-
indene analogues of neurosteroids, and we continue to
interpret this lack of parallelism as suggestive evidence
for the hypothesis that the potentiation and gating
activities occur at different steroid binding sites on the
same receptor.19 Without attempting to interpret fur-
ther the results from the gating assay, it is interesting
to note that (1) steroid 14b gates an amount of current
equivalent to that gated by 2 µM GABA and (2) a 5R-
methyl substituent either eliminates or reduces to a
very low level both the potentiation and gating effects
of neurosteroids.
As noted earlier in this discussion, cultured rat
hippocampal neurons contain multiple subtypes of
GABAA receptors. Accordingly, the responses to the
compounds evaluated in this study represent average
responses recorded from multiple subtypes of GABAA
receptors. Previous studies carried out with heteroge-
neous populations of GABAA receptors have provided
evidence for the existence of heterogeneity in the
binding sites for anesthetic steroids.26 In particular, the
finding that 3R-hydroxy-5â-pregnan-20-one, but not 3R-
hydroxy-5R-pregnan-20-one, shows two-component al-
losteric modulation of benzodiazepine and tert-butylbi-
cyclophosphorothionate binding in some regions of
bovine brain has been interpreted to indicate that the
5â-reduced steroid may show selective binding to dif-
ferent types of GABAA receptors.26f The extent to which
the active anesthetic steroid analogues reported here
may have selective actions on different subtypes of
GABAA receptors remains to be investigated.
In summary, the results of this study provide new
information about the mode of binding of 5R- and 5â-
reduced steroids to GABAA receptors. The data reported
clearly indicate that there are steric restrictions on the
hydrophobic space that can be occupied between the
hydrogen bond donating group at C-3 and the hydrogen
bond accepting group at C-17. We are particularly
interested in obtaining a better understanding of these
steric restrictions because we believe that this informa-
tion will be crucial for the development of nonsteroidal
analogues that can modulate GABAA receptors via
interactions at a steroid binding site.
17â-Hyd r oxy-5r-estr a n -3-on e (4), 17â-Hyd r oxy-5â-es-
tr a n -3-on e (5), a n d 5â-Estr a n e-3r,17â-d iol (7). A 10%
Pd-C catalyst (0.18 g) was added to a solution of 17â-
hydroxyestr-4-en-3-one (3, 3.0 g, 10.9 mmol) and KOH (0.8 g)
in 2-propanol (125 mL) in a glass bottle and hydrogenated (40-
45 psi, overnight, room temperature) in a Parr hydrogenation
apparatus. The catalyst was filtered off, the solvent was
removed under reduced pressure, and brine (200 mL) and
ether (200 mL) were added. The heterogeneous solvent
mixture was neutralized with 6 N HCl while being cooled with
an ice-water bath. The organic layer was separated, washed
with brine (2 × 100 mL), dried over Na2SO4, and concentrated
under reduced pressure to give a solid which was purified by
chromatography (silica gel, 30% EtOAc in hexane). In their
order of elution, the fractions contained steroid 4 (200 mg, 7%),
steroid 5 (1.8 g, 60%), a mixture (300 mg, 10%) of steroid 5
and 3â,17â-dihydroxy-5R-estrane (6), and 3R,17â-dihydroxy-
5R-estrane (7, 500 mg, 16%). Compound 4 was obtained as
colorless crystals: mp 128-130 °C (from EtOAc/hexane, lit.6
1
mp 133-134 °C); IR 3292, 1713 cm-1; H NMR δ 3.65 (t, J )
8.3 Hz, 1H, CHOH), 0.77 (s, 3H, CH3); 13C NMR δ 211.90
(CdO), 81.73 (HOCH), 11.01 (CH3), 49.87, 48.56, 47.72, 45.67,
43.60, 42.99, 41.23, 40.96, 36.51, 33.78, 30.47, 30.32, 30.12,
25.66, 23.13. Anal. (C18H28O2) C, H.
Compound 5 was obtained as colorless crystals: mp 110-
112 °C (from EtOAc/hexane, lit.6 mp 108-110 °C); IR 3427,
1709 cm-1; 1H NMR δ 3.58-3.73 (m, 1H, CHOH), 2.59 (t, J )
14.2 Hz, 1H, 4-Hax), 0.78 (s, 3H, CH3); 13C NMR δ 212.99
(CdO), 81.69 (HOCH), 11.03 (CH3), 49.88, 43.10, 42.78, 41.40,
39.64, 38.32, 38.22, 36.57, 36.30, 30.41, 30.32, 27.63, 25.39,
24.92, 23.10. Anal. (C18H28O2) C, H.
Compound 7 was obtained as colorless crystals: mp 203-
204 °C (from EtOAc/hexane, lit.6 mp 191.5-193 °C); IR 3288,
1
cm-1; H NMR δ 3.68-3.60 (m, 2H, 2 × CHOH), 0.74 (s, 3H,
CH3); 13C NMR δ 82.04 (CHOH), 71.72 (CHOH), 11.08 (CH3),
50.08, 43.15, 41.84, 39.92, 38.56, 36.76, 36.37, 35.67, 31.43,
30.55, 29.70, 26.01, 25.76, 25.24, 23.25. Anal. (C18H30O2) C,
H.
5â-E st r a n e-3,17-d ion e (8). To a solution of compounds
5-7 (2.35 g) in acetone (100 mL) at room temperature was
added J ones reagent dropwise until a yellow color persisted.
After 30 min, 2-propanol was added to destroy excess J ones
reagent, and brine (200 mL) was added. The reaction mixture
was extracted with EtOAc (3 × 150 mL), and the combined
organic solvents were dried over Na2SO4. Solvent removal
gave compound 8 (2.24 g, 97%) as colorless crystals: mp 182-
183 °C (from EtOAc/hexane, lit.6 mp 182.5-185 °C); IR 1735,
1706 cm-1; 1H NMR δ 2.59 (t, 1H, J ) 14.2 Hz, 4-Hax), 0.92 (s,
3H, CH3); 13C NMR δ 220.62 (CdO), 212.11 (CdO), 13.59
(CH3), 50.16, 47.66, 42.58, 40.66, 39.46, 38.24, 37.97, 36.10,
35.59, 31.30, 30.15, 27.42, 24.90, 24.10, 21.40. Anal. (C18H26O2)
C, H.
3â-Hyd r oxy-5â-estr a n -17-on e (9a ). K-Selectride (11.7
mL, 11.7 mmol, 1.0 M solution in THF) was added to a cooled
(-78 °C), stirred solution of compound 8 (2.14 g, 7.8 mmol) in
dry THF (200 mL) under nitrogen. After 6 h, the reaction was
stopped by the addition of 10% aqueous NaOH (20 mL),
followed by 30% H2O2 (30 mL). The reaction mixture was
allowed to warm to room temperature, and stirring was
continued for another 30 min. The mixture was extracted with
EtOAc (2 × 200 mL). The combined organic solvents were
washed with brine (2 × 100 mL) and dried over Na2SO4.
Solvent removal gave compound 9a (2.01 g, 93%) as colorless
crystals: mp 166-167 °C (from EtOAc/hexane, lit.6 mp 165-
166.5 °C); IR 3471, 1725 cm-1; 1H NMR δ 4.11 (m, 1H, CHOH),
0.87 (s, 3H, CH3); 13C NMR δ 221.46 (CdO), 66.47 (HOCH),
Exp er im en ta l Section
Gen er a l Meth od s. Melting points were determined on a
Kofler micro hot stage and are uncorrected. NMR spectra were
recorded at ambient temperature in CDCl3 with a 5 mm probe
on either a Varian Gemini 300 operating at 300 MHz (1H) or
1
75 MHz (13C). For H NMR and 13C NMR spectra, the internal
references were TMS (δ 0.00) and CDCl3 (δ 77.00), respectively.
IR spectra were recorded as films on a NaCl plate (unless noted

4226 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21
Han et al.
13.56 (CH3), 45.03, 47.74, 40.88, 40.48, 37.75, 35.65, 33.02,
31.39, 30.99, 29.61, 26.59, 24.97, 24.72, 21.43, 21.09. Anal.
(C18H28O2) C, H.
Compound 11c was obtained as colorless crystals: mp 143-
144 °C (from EtOAc/hexane); IR 3286, 2923, 2232, 1448, 1386,
1342, 1294, 1226, 1160, 1131, 1093, 1070, 1008 cm-1; 1H NMR
δ 4.12 (m, 1H, HOCH), 2.58 (d, J ) 8.8 Hz, 1H, CNCH), 0.82
(s, 3H, CH3); 13C NMR δ 122.19 (CN), 66.49 (HOCH), 50.99
(CHCN), 17.77 (CH3), 44.21, 41.69, 40.39, 39.91, 36.97, 34.91,
33.00, 31.10, 29.49, 26.99, 26.57, 25.99, 25.37, 24.46, 21.14.
Anal. (C19H29NO) C, H, N.
Compound 11d was obtained as colorless crystals: mp 149-
150 °C; IR 3280, 2928, 2235, 1446, 1384, 1341, 1265, 1221,
1101, 1009 cm-1; 1H NMR δ 4.12 (m, 1H, HOCH), 2.29 (t, J )
9.5 Hz, 1H, CNCH), 0.92 (s, 3H, CH3); 13C NMR δ 121.03 (CN),
66.31 (HOCH), 53.13 (CNCH), 13.99 (CH3), 44.28, 41.65, 40.27,
39.98, 37.20, 36.80, 32.89, 30.97, 29.38, 26.46, 26.21, 25.74,
25.22, 24.06, 20.98. Anal. (C19H29NO) C, H, N.
3â-(Acetyloxy)-5â-estr a n -17-on e (9b). Acetic anhydride
(1.3 g, 12.4 mmol) was added to stirred pyridine (40 mL) at
120 °C. After 30 min, steroid 9a (1.72 g, 6.2 mmol) dissolved
in pyridine (10 mL) was added, and the reaction solution was
kept at 120 °C for 4 h and then allowed to cool to room
temperature. Ice-water (100 mL) and 10% HCl (50 mL) were
added, and the mixture was extracted with EtOAc (2 × 150
mL). The combined organic solvents were washed with brine
(2 × 150 mL) and saturated NaHCO3 (2 × 150 mL) and dried
over Na2SO4. Evaporation of the combined organic solvents
under reduced pressure gave a solid (1.8 g, 91%). Compound
9b was obtained as colorless crystals: mp 128-129 °C (from
EtOAc/hexane); IR 2924, 2855, 1738, 1452, 1407, 1375, 1238,
3-Oxo-5â-estr a n e-17â-ca r bon itr ile (12). J ones reagent
was added dropwise to a room temperature solution of
compound 11d (520 mg, 1.8 mmol) in acetone (50 mL) until a
yellow color persisted. After 30 min, 2-propanol was added
to destroy excess J ones reagent, brine (100 mL) was added,
and the mixture was extracted with EtOAc (3 × 100 mL). The
combined organic solvents were dried over Na2SO4. Evapora-
tion of organic solvents gave product 12 as a solid (500 mg,
97%): mp 148-149 °C (colorless crystals from EtOAc/hexane);
IR 2920, 2874, 2234, 1709, 1450, 1387, 1340, 1298, 1267, 1228,
1
1217, 1189, 1145, 1094, 1052, 1071 cm-1; H NMR δ 5.09 (m,
1H, AcOCH), 2.05 (s, 3H, CH3COO), 0.88 (s, 3H, CH3); 13C
NMR δ 221.12 (CdO), 170.45 (CH3COO), 70.39 (AcOCH), 13.68
(CH3), 50.41, 47.81, 41.01, 40.25, 37.97, 35.75, 31.49, 30.92,
30.56, 30.28, 25.10, 24.70, 23.92, 21.97, 21.55, 21.36. Anal.
(C20H30O3) C, H.
3â-(Acetyloxy)-5â-estr a n e-17r-ca r bon itr ile (11a ) a n d
3â-(Acetyloxy)-5â-estr a n e-17â-ca r bon itr ile (11b). Com-
pound 9b (1.77 g, 5.6 mmol) was stirred with diethyl cyano-
phosphonate (2.74 g, 16.8 mmol) and LiCN (554 mg, 16.8
mmol) in THF (100 mL) at room temperature overnight. After
water (100 mL) was added, the mixture was extracted with
EtOAc (2 × 200 mL). The combined organic solvents were
washed with brine (2 × 100 mL), dried over Na2SO4, and
evaporated under reduced pressure to yield a solid which was
purified by chromatography (silica gel, 50% EtOAc in hexane)
to give the desired intermediates 10a and 10b (2.61 g, 97%)
as an oil.
A solution of SmI2 (16.2 mmol, 162 mL of 0.1 M solution in
THF) was added at room temperature to intermediates 10a
and 10b (2.6 g, 5.4 mmol) and methanol (518 mg, 16.2 mmol)
under nitrogen and stirred overnight. Then 10% HCl (20 mL)
and brine (100 mL) were added, and the mixture was extracted
with ethyl acetate (2 × 150 mL). The extracts were washed
with brine (2 × 100 mL), 5% Na2S2O3 (150 mL), and brine (150
mL), dried over Na2SO4, and evaporated under reduced
pressure to yield a mixture of products 11a and 11b (1.6 g,
90%), which was separated by HPLC (silica gel, 11% EtOAc
in hexane, 3 mL/min). The ratio of products 11a :11b was 1:3,
and carbonitrile 11a eluted first from the column.
1
1174, 1129, 1101 cm-1; H NMR δ 2.57 (t, J ) 14.2 Hz, 1H,
4-Hax), 2.33 (t, J ) 9.5 Hz, 1H, CNCH), 0.96 (s, 3H, CH3); 13
C
NMR δ 212.33 (CdO), 121.07 (CN), 53.20 (CNCH), 14.23 (CH3),
44.47, 42.67, 41.61, 40.20, 39.44, 37.93, 36.86, 36.21, 30.30,
27.50, 26.40, 25.32, 24.22. Anal. (C19H27NO) C, H, N.
3r-Hyd r oxy-5â-estr a n e-17â-ca r bon itr ile (13a ). NaBH4
(84 mg) was added to a solution of compound 12 (420 mg, 1.5
mmol) in EtOH (20 mL). After 2 h, the excess NaBH4 was
destroyed by addition of HOAc, water (100 mL) was added,
and the product was extracted with EtOAc (2 × 100 mL). The
combined organic solvents were washed with brine (2 × 100
mL), dried over Na2SO4, and evaporated to give a solid which
was purified by chromatography (silica gel, 30% EtOAc in
hexane) to give carbonitrile 13a (290 mg, 69%) as the first
fraction and carbonitrile 11d (100 mg, 24%) as the second
fraction. Compound 13a was obtained as colorless crystals:
mp 138-139 °C (from EtOAc/hexane); IR 3397, 2926, 2860,
2235, 1451, 1387, 1374, 1300, 1266, 1247, 1223, 1160, 1090,
1060, 1039 cm-1; 1H NMR δ 3.63 (m, 1H, HOCH), 2.29 (t, J )
9.4 Hz, 1H, CNCH), 0.92 (s, 3H, CH3); 13C NMR δ 121.30 (CN),
71.36 (HOCH), 53.40 (CNCH), 14.25 (CH3), 44.52, 42.01, 40.29,
39.68, 38.12, 37.05, 36.14, 35.36, 31.26, 29.50, 26.48, 26.12,
25.87, 25.14, 24.33. Anal. (C19H29NO) C, H, N.
Compound 11a was obtained as colorless crystals: mp 105-
107 °C; IR 2927, 2232, 1733, 1450, 1382, 1241, 1159, 1132,
1019 cm-1; 1H NMR δ 5.10 (m, 1H, AcOCH), 2.58 (dd, J ) 8.9
Hz, J ) 2.0 Hz, 1H, CNCH), 2.05 (s, 3H, CH3COO), 0.83 (s,
3H, CH3); 13C NMR δ 170.53 (CH3COO), 122.32 (CN), 70.52
(AcOCH), 17.94 (CH3), 51.07, 44.36, 41.85, 40.20, 40.10, 37.21,
35.04, 31.06, 30.49, 30.29, 27.16, 26.01, 25.52, 24.61, 23.95,
22.06, 21.42. Anal. (C21H31NO2) C, H, N.
(3r,5â,17r)-3-Hyd r oxy-19-n or p r egn a n -20-on e (14a ) a n d
(3r,5â,17â)-3-Hyd r oxy-19-n or p r egn a n -20-on e (14b). A so-
lution of methylmagnesium chloride (2.7 mL, 8.0 mmol, 3.0
M solution in THF) was added at room temperature to a stirred
solution of steroid 13a (240 mg, 0.8 mmol) in THF (150 mL)
under nitrogen. The mixture was refluxed overnight and
cooled to 0 °C, saturated NH4Cl (20 mL) was added, and the
product was extracted with EtOAc (3 × 150 mL). The
combined organic layers were washed with brine (2 × 150 mL)
and dried over Na2SO4, and the solvent was evaporated to give
the mixture of products 14a and 14b (203 mg, 80%), which
was separated by HPLC (silica gel, 30% EtOAc and 10%
ClCH2CH2Cl in hexane, 3 mL/min). The ratio of products 14a :
14b was 1:10, and steroid 14b eluted first from the column.
Compound 14a was obtained as colorless crystals: mp 82-
84 °C (from EtOAc/hexane); IR 3397, 2920, 2867, 1703, 1454,
Compound 11b was obtained as colorless crystals: mp 184-
186 °C; IR 2958, 2924, 2875, 2857, 2229, 1728, 1448, 1383,
1364, 1256, 1238, 1220, 1198, 1164, 1144, 1100, 1023 cm-1
;
1H NMR δ 5.08 (m, 1H, AcOCH), 2.29 (t, J ) 9.5 Hz, 1H,
CNCH), 2.05 (s, 3H, CH3COO), 0.93 (s, 3H, CH3); 13C NMR δ
170.60 (CH3COO), 121.28 (CN), 70.45 (AcOCH), 53.42 (CNCH),
14.27 (CH3), 44.56, 41.92, 40.33, 40.18, 37.60, 37.06, 31.03,
30.49, 30.33, 26.50, 25.88, 25.49, 24.33, 23.99, 22.01, 21.44.
Anal. (C21H31NO2) C, H, N.
3â-Hyd r oxy-5â-estr a n e-17r-ca r bon itr ile (11c) a n d 3â-
Hyd r oxy-5â-estr a n e-17â-ca r bon itr ile (11b). A mixture of
compounds 11a and 11b (1.51 g) and 0.18 M aqueous metha-
nolic K2CO3 (100 mL, 70% MeOH in water) and 10% NaOH
(10 mL) was refluxed for 2 h. After the mixture was cooled to
room temperature, brine (150 mL) and EtOAc (150 mL) were
added, and the aqueous layer was again extracted with EtOAc
(2 × 100 mL). The combined organic layers were washed with
brine (2 × 100 mL) and dried over Na2SO4. The solvent was
removed under reduced pressure, and the products were
purified by chromatography (20% EtOAc in hexane) to give
carbonitriles 11c and 11d (1.1 g, 84%). These carbonitriles
were separated by HPLC (silica gel, 25% EtOAc in hexane, 3
mL/min). Carbonitrile 11c eluted first from the column.
1
1359, 1171, 1091, 1058, 1038 cm-1; H NMR δ 3.64-3.54 (m,
1H, HOCH), 2.81 (dd, J ) 9.5 Hz, J ) 2.5 Hz, 1H, CHCOCH3),
2.13 (CH3CO), 0.91 (s, 3H, CH3); 13C NMR δ 212.97 (CH3CO),
71.67 (HOCH), 61.39 (CH3COCH), 20.88 (CH3), 49.35, 45.95,
41.95, 39.83, 37.84, 36.41, 35.59, 35.31, 32.84, 31.49, 29.67,
26.45, 26.00, 25.68, 25.55, 24.29. Anal. (C20H32O2) C, H.
Compound 14b was obtained as colorless crystals: mp 108-
109 °C (from EtOAc/hexane, lit.3f mp 105-106 °C); IR 3420,
2922, 2869, 1703, 1451, 1385, 1358, 1266, 1226, 1195, 1170,
1152, 1091, 1060, 1040 cm-1 1H NMR δ 3.69-3.59 (m, 1H,
;
HOCH), 2.54 (t, J ) 8.7 Hz, 1H, CHCOCH3), 2.12 (s, 3H, CH3-
CO), 0.61 (s, 3H, CH3); 13C NMR δ 209.73 (CH3CO), 71.47
(HOCH), 63.87 (CH3COCH), 13.33 (CH3), 55.65, 44.32, 41.65,

Neurosteroid Analogues
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21 4227
39.76, 38.99, 38.18, 36.22, 35.46, 31.46, 31.38, 29.53, 26.09,
25.86, 25.54, 24.18, 22.70. Anal. (C20H32O2) C, H.
1448, 1377, 1339, 1240, 1167, 1034, 1005 cm-1; 1H NMR δ 4.11
(m, 1H, HOCH), 2.27 (t, 1H, J ) 9.6 Hz, CNCH), 0.98 (s, 3H,
CH3), 0.91 (s, 3H, CH3); 13C NMR δ 121.31 (CN), 66.83
(HOCH), 14.27 (CH3), 54.48, 44.49, 40.26, 39.59, 37.28, 36.31,
36.00, 35.13, 33.35, 29.86, 27.76, 26.56, 26.36, 26.23, 24.52,
23.74, 20.67. Anal. (C20H31NO) C, H, N.
3r-(Acetyloxy)-5â-a n d r osta n -17-on e (15b). Using a pro-
cedure similar to that described for the preparation of steroid
9b, 3R-hydroxy-5â-androstan-17-one (15a , 500 mg, 1.72 mmol)
was converted into compound 15b (560 mg, 98%). Compound
15b was obtained as colorless crystals: mp 90-91 °C (from
3r-Hyd r oxy-19-n or -5r-p r egn a n -20-on e (18b). Methyl-
magnesium chloride (3 mL, 9.0 mmol, 3.0 M solution in THF)
was reacted with previously prepared steroid 18a 12 (59 mg,
0.2 mmol) in THF (20 mL) at reflux for 2 days, followed by a
workup procedure analogous to that described for the prepara-
tion of steroids 14a and 14b. HPLC (silica gel, 20% EtOAc,
10% ClCH2CH2Cl in hexane, 3 mL/min) separation of the
reaction products gave the initially eluted unreacted steroid
18a (10 mg) followed by the product 18b (40 mg, 64%), which
was recovered as colorless crystals: mp 176-177 °C (from
1
EtOAc/hexane, lit.10 mp 96-97 °C); IR 1739 cm-1; H NMR δ
4.73 (m, 1H, AcOCH), 2.03 (s, 3H, CH3COO), 0.96 (s, 3H, 19-
CH3), 0.86 (s, 3H, 18-CH3); 13C NMR δ 221.28 (CdO), 170.61
(CH3COO), 74.06 (AcOCH), 13.74 (18-CH3), 51.44, 47.82, 41.77,
40.68, 35.88, 35.31, 34.98, 34.72, 32.13, 31.66, 26.67, 26.52,
25.23, 23.21, 21.75, 21.42, 20.05. Anal. (C21H32O3) C, H.
3r-(Acet yloxy)-5â-a n d r ost a n e-17r-ca r b on it r ile (16a ),
3r-(Acetyloxy)-5â-a n d r osta n e-17â-ca r bon itr ile (16b), 3r-
Hyd r oxy-5â-a n d r osta n e-17r-ca r bon itr ile (17a ), a n d 3r-
Hyd r oxy-5â-a n d r osta n e-17â-ca r bon itr ile (17b). Using a
two-step cyanation procedure similar to that described for the
preparation of steroids 11a and 11b, compound 15a (550 mg,
1.7 mmol) was converted into a mixture (343 mg) of compounds
16a and 16b, along with hydrolysis products 17a and 17b.
The partial loss of the acetyloxy group occurred after the
second step of the procedure. Column chromatography (silica
gel, 25% EtOAc in hexane) was performed to separate the
mixture of products 16a and 16b (160 mg, 28%) from product
17a (80 mg, 16%) and product 17b (100 mg, 20%). Products
16a (first component) and 16b were separated by HPLC (silica
gel, 7% EtOAc in hexane, 3 mL/min).
EtOAc/hexane); IR 3307, 2918, 2859, 1705, 1442, 1353, 1200,
1
1160, 1099, 1001 cm-1
; H NMR δ 4.09 (m, 1H, HOCH), 2.55
(t, J ) 8.7 Hz, CHCOCH3), 2.12 (s, 3H, CH3CO), 0.62 (s, 3H,
CH3); 13C NMR δ 209.81 (COCH3), 66.40 (HOCH), 63.93
(CHCOCH3), 13.41 (CH3), 55.87, 47.90, 46.93, 44.31, 41.21,
40.53, 39.02, 35.92, 33.55, 32.92, 31.52, 30.91, 25.61, 24.17,
23.66, 22.70. Anal. (C20H32O2) C, H.
3′,4â-Cyclop r op [4,5]-5r-estr a n e-3,17-d ion e (21a ). A so-
lution of K-Selectride (55 mL, 55 mmol, 1.0 M solution in THF)
was added to a cooled (-78 °C), stirred solution of steroid 3
(7.5 g, 27.3 mmol) in dry THF (100 mL) under nitrogen. After
15 h, the reaction mixture was allowed to warm to -10 °C,
and stirring was continued for another 3 h. The reaction was
stopped by the addition of 10% aqueous NaOH (15 mL) and
brine (150 mL), and the mixture was extracted with EtOAc (2
× 200 mL). The combined organic layers were washed with
brine (2 × 150 mL) and dried over Na2SO4. Evaporation of
organic solvent gave a mixture of compounds 3, 6, 19a , and
19b (7.2 g). HPLC (silica gel, 50% EtOAc in hexane, 3 mL/
min) separation of a portion of the product mixture gave in
order of elution steroid 6, steroid 19b, and a mixture of steroids
3 and 19a in the relative ratio of 14:57:29, respectively. The
structures of known steroids 66 and 19b17 were confirmed by
combustion analysis and NMR and IR spectroscopic analysis.
Diiodomethane (24 g) was added to a stirred mixture of
zinc-copper couple (6 g) and a small crystal of iodine in ether
(50 mL). The mixture was warmed with a tungsten lamp until
the reaction started, and the reaction was continued for 30
min on a water bath at 60 °C. A solution of the mixture of
compounds 3, 6, 19a , and 19b (7 g) was added over 10 min,
and the mixture was stirred for 3 days at reflux. The ice-cooled
mixture was diluted with saturated NH4Cl, and the precipitate
was filtered and washed twice with ether. The combined ether
extracts were washed with brine, dried over Na2SO4, and
concentrated under reduced pressure to give a crude mixture
containing products 20a and 20b. This crude mixture was
used without product characterization.
Compound 16a was obtained as an oil: IR 2938, 2866, 2233,
1734, 1451, 1383, 1363, 1244, 1167, 1030 cm-1; 1H NMR δ 4.72
(m, 1H, AcOCH), 2.57 (dd, J ) 8.9 Hz, J ) 1.8 Hz, 1H, CNCH),
2.04 (s, 3H, CH3COO), 0.94 (s, 3H, 19-CH3), 0.81 (s, 3H, 18-
CH3); 13C NMR δ 170.56 (CH3COO), 122.21 (CN), 73.87
(AcOCH), 17.86 (18-CH3), 51.99, 44.18, 41.52, 39.93, 39.83,
35.97, 35.13, 34.93, 34.45, 31.93, 27.18, 26.64, 26.32, 24.76,
23.07, 21.33, 20.37. Anal. (C22H33NO2) C, H, N.
Compound 16b was obtained as colorless crystals: mp 132-
134 °C (from EtOAc/hexane); IR 2936, 2868, 2235, 1735, 1451,
1383, 1363, 1244, 1029 cm-1; 1H NMR δ 4.72 (m, 1H, AcOCH),
2.28 (t, J ) 9.7 Hz, 1H, CNCH), 2.03 (s, 3H, CH3COO), 0.95
(s, 3H, CH3), 0.90 (s, 3H, CH3); 13C NMR δ 170.50 (CH3COO),
121.24 (CN), 74.01 (AcOCH), 54.34 (CNCH), 14.23 (18-CH3),
44.42, 41.59, 40.23, 37.18, 36.11, 34.95, 34.57, 32.08, 26.69,
26.53, 26.49, 26.25, 24.49, 23.17, 21.38, 20.39. Anal. (C22H33
NO2) C, H, N.
-
Compound 17a was obtained as colorless crystals: mp 105-
106 °C (from EtOAc/hexane); IR 3386, 2933, 2863, 2236, 1450,
1387, 1266, 1171, 1108, 1088, 1066, 1036 cm-1; 1H NMR δ 3.62
(m, 1H, HOCH), 2.58 (dd, J ) 8.8 Hz, J ) 1.8 Hz, 1H, CNCH),
0.93 (s, 3H, 19-CH3), 0.80 (s, 3H, 18-CH3); 13C NMR δ 122.27
(CN), 71.79 (HOCH), 52.04, (CNCH), 17.95 (18-CH3), 23.19 (19-
CH3), 44.26, 41.84, 40.00, 39.91, 36.31, 36.12, 35.34, 35.21,
34.53, 30.47, 27.25, 26.90, 26.50, 24.87, 20.43. Anal. (C20H31
NO) C, H, N.
-
J ones reagent was added at room temperature to a stirred
solution of the above crude mixture dissolved in acetone (100
mL) until a yellow color persisted for 10 min, and 2-propanol
was added to destroy any excess J ones reagent. EtOAc (200
mL) and water (200 mL) were added, and the aqueous layer
was extracted with EtOAc (3 × 200 mL). The combined
organic layers were washed with brine (3 × 200 mL) and dried
over Na2SO4. Following solvent removal, an oil was obtained.
Partial purification by chromatography (silica gel, 30% EtOAc
in hexane) gave a product mixture containing steroids 21a and
21b. This product mixture was dissolved in dichloromethane
(50 mL) and HOAc (1 mL), cooled to -78 °C, and a stream of
O3 was bubbled through the solution until a blue color
persisted. The excess O3 was removed with a stream of O2.
Additional dichloromethane (100 mL) was added, and the
solution was washed with 5% Na2CO3, water, and brine and
dried over Na2SO4. Solvent removal gave a solid which was
purified by chromatography (silica gel, 35% EtOAc in hexane)
to give steroid 8 (1.57 g) and a mixture of steroids 21a and
21b (1.1 g, 14% overall from compound 3). A pure sample of
product 21a was obtained upon crystallization from EtOAc/
hexane. A pure sample of product 21b was not obtained.
Compound 21a was obtained as colorless crystals: mp 207-
209 °C (from EtOAc/hexane); IR 2926, 1738, 1682, 1452, 1378,
Compound 17b was obtained as colorless crystals: mp 161-
162 °C (from EtOAc/hexane); IR 3365, 2929, 2861, 2234, 1451,
1
1383, 1268, 1167, 1092, 1068, 1041 cm-1; H NMR δ 3.64 (m,
1H, HOCH), 2.28 (t, J ) 9.5 Hz, 1H, CNCH), 0.94 (s, 3H, CH3),
0.90 (s, 3H, CH3); 13C NMR δ 121.29 (CN), 71.46 (HOCH),
54.35 (CNCH), 14.23 (18-CH3), 23.20 (19-CH3), 44.43, 41.77,
40.24, 40.21, 37.21, 36.17, 35.25, 34.54, 30.35, 26.87, 26.54,
26.34, 24.52, 20.38. Anal. (C20H31NO) C, H, N.
3â-Hyd r oxy-5â-a n d r osta n e-17â-ca r bon itr ile (17c). Tri-
fluoroacetic acid (78 mg, 0.68 mmol) and then solid triph-
enylphosphine (178 mg, 0.68 mmol) were added to diethyl
azodicarboxylate (118 mg, 0.68 mmol) and compound 17b (50
mg, 0.17 mmol) in stirred dry THF (2 mL). After 5 min,
sodium benzoate (200 mg, 1.4 mmol) was added and the
mixture was stirred 2 h. EtOH (5 mL) and 10% NaOH (2 mL)
was added, and stirring was continued for 1 h. The reaction
solution was poured into water (100 mL) and extracted with
EtOAc (2 × 50 mL). The combined organic solvents were
washed with brine (2 × 50 mL) and dried over Na2SO4.
Solvent removal gave a solid that was purified by chromatog-
raphy (silica gel, 30% EtOAc in hexane) to yield product 17c
(48 mg, 95%) as colorless crystals: mp 193-194 °C (from
EtOAc/hexane); IR 3333, 2931, 2874, 2860, 2233, 1508, 1471,

4228 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21
Han et al.
1330, 1253, 1213, 1046 cm-1; 1H NMR δ 0.92 (s, 3H, CH3), 0.82
(q, J ) 4.5 Hz, 1H cyclopropyl-H); 13C NMR δ 220.79 (CdO),
208.59 (CdO), 13.70 (CH3), 15.10 (cyclopropyl-CH2), 50.22,
47.75, 47.61, 40.79, 39.84, 36.10, 35.80, 35.66, 33.82, 31.30,
30.88, 28.90, 24.89, 21.68, 21.55. Anal. (C19H26O2) C, H.
5-Meth yl-5r-estr a n e-3,17-d ion e (22a ) a n d 5-Meth yl-5â-
estr a n e-3,17-d ion e (22b). Lithium wire (450 mg, 64.8 mmol)
was added to stirred liquid NH3 (ca. 160 mL). After 30 min,
a solution of steroids 21a and 21b (1.2 g, 4.2 mmol) in THF
(50 mL) was added slowly, stirring was continued for 2 h, and
1,2-dibromoethane was added dropwise to discharge the blue
color. A solution of HOAc (4 mL) in MeOH (16 mL) was added
dropwise over 15 min, and the liquid NH3 was allowed to
evaporate at room temperature. The mixture was diluted with
water (100 mL) and EtOAc (150 mL), and the organic layer
was separated. The water layer was extracted with EtOAc (2
× 150 mL), and the combined organic layers were washed with
brine (2 × 150 mL) and dried over Na2SO4. The solvent was
removed to give a residue which was purified by chromatog-
raphy (silica gel, 30% EtOAc in hexane) to give a first fraction
that contained a mixture of products 22a and 22b (900 mg,
74%) and a second fraction that contained steroid 22c (110
mg, 9%). Steroid 22c [IR 3275, 1706 cm-1; 1H NMR δ 3.66 (t,
J ) 8.5 Hz, 1H, CHOH), 0.79 (s, 3H, CH3), 0.77 (s, 3H, CH3)]
was oxidized with J ones reagent in acetone as described above
to give a solid that was purified by chromatography (30%
EtOAC in hexane) to give product 22a (90 mg, 82%). Pure
compound 22b was obtained by J ones oxidation of steroid 23b
(vide infra).
3r-(Acetyloxy)-5-m eth yl-5r-estr a n -17-on e (24a ) a n d 3â-
(Acetyloxy)-5-m eth yl-5â-estr a n -17-on e (24b). A solution
of steroids 23a and 23b (830 mg, 2.9 mmol) in pyridine (10
mL) was added to a stirred solution of (AcO)2O (2.9 g, 29.6
mmol) and pyridine (40 mL) that had been preheated to 120
°C for 30 min. The reaction was continued for 2 h, the mixture
was cooled to room temperature, and ice-water (100 mL) and
10% HCl (50 mL) were added. The mixture was extracted with
EtOAc (2 × 150 mL). The combined organic solvents were
washed with brine (2 × 150 mL) and dried over Na2SO4.
Evaporation of organic solvent under reduced pressure gave
a solid, which was purified by chromatography (silica gel, 15%
EtOAc in hexane) and separated by HPLC (silica gel, 15%
EtOAc in hexane, 3 mL/min) into initially eluted steroid 24a
(470 mg, 50%) and steroid 24b (310 mg, 33%).
Compound 24a was obtained as colorless crystals: mp 145-
147 °C (from EtOAc/hexane); IR 2928, 2859, 1733, 1434, 1377,
1242, 1230, 1162, 1132, 1108, 1049, 1021 cm-1; 1H NMR δ 5.02
(m, 1H, HOCH), 2.03 (s, 3H, CH3COO), 0.97 (s, 3H, 5-CH3),
0.87 (s, 3H, 18-CH3); 13C NMR δ 220.68 (CdO), 170.06
(CH3COO), 69.65 (CHOAc), 13.54 (18-CH3), 50.22, 49.08, 47.53,
44.52, 41.60, 41.49, 40.81, 35.47, 33.46, 31.27, 30.79, 24.95,
24.47, 21.21, 19.39, 18.32. Anal. (C21H32O3) C, H.
Compound 24b was obtained as colorless crystals: mp 163-
164 °C (from EtOAc/hexane); IR 2925, 1737, 1454, 1376, 1241,
1
1153, 1113, 1019 cm-1; H NMR δ 5.08 (m, 1H, HOCH), 2.04
(s, 3H, CH3COO), 1.13 (s, 3H, 5-CH3), 0.87 (s, 3H, 18-CH3; 13
C
NMR δ 221.09 (CdO), 170.26 (CH3COO), 70.44 (CHOAc), 17.65
(5-CH3), 13.55 (18-CH3), 50.45, 47.48, 45.43, 40.97, 40.53,
39.39, 35.71, 34.08, 32.67, 31.41, 30.45, 25.65, 25.33, 23.49,
21.46. Anal. (C21H32O3) C, H.
Compound 22a was obtained as colorless crystals: mp 176-
178 °C (from EtOAc/hexane); IR 2922, 1739, 1700, 1453, 1378,
1
1295, 1251, 1197, 1104, 1050, 1009 cm-1; H NMR δ 0.91 (s,
3r-Hydr oxy-5-m eth yl-5r-estr an e-17r-car bon itr ile (26a)
a n d 3r-H yd r oxy-5-m et h yl-5r-est r a n e-17â-ca r b on it r ile
(26b). Using a two-step cyanation procedure similar to that
described for the preparation of steroids 11a and 11b, com-
pound 24a (460 mg, 1.38 mmol) was converted into a mixture
of compounds 25a and 25b (440 mg, 98%) which was saponi-
fied in stirred THF (5 mL) with 0.18 M aqueous methanolic
K2CO3 (20 mL, 70% MeOH in water) and 10% NaOH (0.5 mL)
at reflux for 2 d. After the mixture was cooled to room
temperature, brine (150 mL) and EtOAc (150 mL) were added
and the aqueous layer was extracted further with EtOAc (2 ×
100 mL). The combined organic layers were washed with brine
(2 × 100 mL) and dried over Na2SO4. Solvent removal gave a
mixture of products 26a and 26b (360 mg, 93%), which was
separated by HPLC (silica gel, 20% EtOAc in hexane, 3 mL/
min). The product ratio was 1:1 and steroid 26a eluted first
from the column.
3H, CH3), 0.82 (s, 3H, CH3); 13C NMR δ 220.86 (CdO), 211.47
(CdO), 17.75 (CH3), 13.73 (CH3), 56.74, 50.20, 47.76, 47.66,
42.60, 41.40, 41.07, 40.70, 38.56, 35.70, 31.42, 25.71, 25.37,
25.19, 21.47. Anal. (C19H28O2) C, H.
Compound 22b was obtained as colorless crystals: mp 141-
143 °C (from EtOAc/hexane); IR 2923, 2855, 1739, 1712, 1456,
1378, 1259, 1231, 1191, 1107, 1060, 1020 cm-1; 1H NMR δ 2.71
(d, J ) 14.3 Hz, 4-Hax), 0.97 (s, 3H, CH3), 0.91 (s, 3H, CH3);
13C NMR δ 220.96 (CdO), 212.81 (CdO), 13.76 (18-CH3), 50.61,
48.11, 47.67, 45.50, 40.85, 40.03, 39.46, 39.02, 36.22, 35.86,
31.55, 29.27, 25.89, 25.55, 22.79, 21.60. Anal. (C19H28O2) C,
H.
3r-Hyd r oxy-5-m eth yl-5r-estr a n -17-on e (23a ) a n d 3â-
Hyd r oxy-5-m eth yl-5â-estr a n -17-on e (23b). K-Selectride
(6.2 mL, 6.2 mmol, 1.0 M solution in THF) was added to a
cooled (-78 °C), stirred solution of compounds 22a and 22b
(900 mg, 3.1 mmol) in dry THF (150 mL) under nitrogen. After
6 h, the reaction was stopped by the addition of 10% NaOH
(10 mL), followed by 30% H2O2 (10 mL). The reaction mixture
was allowed to warm to room temperature, and stirring was
continued for another 30 min. The mixture was extracted with
EtOAc (2 × 200 mL), and the combined organic solvents were
washed with brine (2 × 100 mL) and dried over Na2SO4.
Evaporation of organic solvent gave a solid (23a and 23b, 850
mg, 94%), which when washed with EtOAc/methanol gave a
small amount (ca. 20 mg) of pure product 23a . The remainder
of the product mixture was characterized as the acetyloxy
derivatives 24a and 24b. The pure sample of steroid 23b was
obtained by hydrolysis of steroid 24b (vide infra). Compound
23a was obtained as colorless crystals: 247-255 °C (from
EtOAc); IR 3488, 3278, 2918, 2868, 1720, 1451, 1401, 1371,
Compound 26a was obtained as colorless crystals: mp 220-
221 °C (from EtOAc/hexane); IR 3278, 3133, 2919, 2858, 2232,
1448, 1380, 1335, 1259, 1229, 1163, 1121, 1071, 1046 cm-1
;
1H NMR δ 4.10 (m, 1H, HOCH), 2.57 (dd, J ) 8.8 Hz, J ) 2.1
Hz, 1H, CNCH), 1.08 (s, 3H, 5-CH3), 0.82 (s, 3H, 18-CH3); 13
C
NMR δ 122.37 (CNCH), 67.30 (AcOCH), 51.22, 49.80, 48.24,
44.37, 42.73, 41.58, 41.20, 40.05, 35.12, 34.19, 33.66, 27.24,
25.95, 25.66, 24.54, 19.37, 19.12, 18.12. Anal. (C20H31NO) C,
H, N.
Compound 26b was obtained as colorless crystals: mp 229-
230 °C (from EtOAc/hexane); IR 3333, 2917, 2239, 1449, 1362,
1
1328, 1273, 1232, 1195, 1104, 1041 cm-1; H NMR δ 4.11 (m,
1H, HOCH), 2.28 (t, J ) 9.7 Hz, 1H, CNCH), 1.06 (s, 3H,
5-CH3), 0.92 (s, 3H, 18-CH3); 13C NMR δ 121.39 (CNCH), 67.29
(AcOCH), 14.39 (18-CH3), 53.57, 49.75, 48.20, 44.54, 42.76,
41.60, 41.51, 40.29, 37.13, 34.19, 33.69, 26.54, 25.78, 25.60,
24.21, 19.35, 19.10. Anal. (C20H31NO) C, H, N.
1
1328, 1259, 1228, 1198, 1110, 1020 cm-1; H NMR δ 4.11 (m,
1H, HOCH), 1.08 (s, 3H, 5-CH3), 0.87 (s, 3H, 18-CH3); 13C
NMR) δ 221.52 (CdO), 67.33 (CHOH), 13.87 (CH3), 50.57,
49.87, 48.24, 47.94, 41.99, 41.89, 41.46, 35.85, 34.22, 33.84,
31.57, 25.28, 24.71, 21.54, 19.39, 19.10. Anal. (C19H30O2) C,
H.
(3r,5r,17r)-3-H yd r oxy-5-m e t h yl-19-n or p r e gn a n -20-
on e (27a ) a n d 3r-Hyd r oxy-5-m eth yl-19-n or -5r-p r egn a n -
20-on e (27b). As described earlier for the preparation of
compounds 14a and 14b, a mixture of steroids 26a and 26b
(80 mg, 0.26 mmol) was reacted with CH3MgBr (8 mL, 24.0
mmol, 3.0 M solution in THF) in THF (40 mL) to give a mixture
of products 27a and 27b (76 mg, 90%), which was separated
by HPLC (silica gel, 10% EtOAc, 10% ClCH2CH2Cl in hexane,
3 mL/min). The product ratio was 1:1 and steroid 27b eluted
first from the column.
Compound 23b was obtained as colorless crystals: mp 192-
194 °C (from EtOAc/hexane); IR 3503, 2921, 1728, 1452, 1377,
1
1323, 1268, 1155, 1101, 1055, 1021 cm-1; H NMR δ 4.16 (m,
1H, HOCH), 1.22 (s, 3H, 5-CH3), 0.87 (s, 3H, 18-CH3); 13C NMR
δ 221.62 (CdO), 67.60 (CHOH), 17.10 (5-CH3), 13.71 (18-CH3),
50.71, 47.71, 46.01, 41.43, 40.69, 39.58, 37.49, 35.89, 32.84,
31.61, 31.43, 26.57, 25.89, 25.44, 21.62. Anal. (C19H30O2) C,
H.
Compound 27a was obtained as colorless crystals: mp 198-
200 °C (from EtOAc/hexane); IR 3340, 2964, 2912, 2868, 1698,

Neurosteroid Analogues
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21 4229
1439, 1360, 1278, 1232, 1171, 1086, 1045 cm-1; 1H NMR δ 4.09
(m, 1H, HOCH), 2.80 (dd, J ) 8.5 Hz, J ) 2.3 Hz, 1H,
CHCOCH3), 2.12 (s, 3H, CH3CO), 1.03 (s, 3H, 5-CH3), 0.91 (s,
3H, 18-CH3); 13C NMR δ 212.93 (COCH3), 67.38 (HOCH), 61.32
(CHCOCH3), 49.84, 49.43, 48.35, 45.94, 42.64, 41.72, 41.20,
35.37, 34.17, 33.67, 32.82, 26.05, 25.93, 25.50, 24.21, 20.99,
19.36, 19.09. Anal. (C21H34O2) C, H.
Compound 27b was obtained as colorless crystals: mp 262-
263 °C (from EtOAc/hexane); IR 3325, 2919, 2869, 1704, 1471,
1447, 1379, 1360, 1267, 1226, 1206, 1194, 1173, 1104, 1085,
1045, 1005 cm-1; 1H NMR δ 4.10 (m, 1H, HOCH), 2.54 (t, J )
8.4 Hz, 1H, CHCOCH3), 2.12 (s, 3H, CH3CO), 1.07 (s, 3H,
5-CH3), 0.60 (s, 3H, 18-CH3); 13C NMR δ 209.71 (COCH3), 67.15
(AcOCH), 63.80 (CHCOCH3), 13.40 (18-CH3), 55.73, 49.73,
48.17, 44.26, 42.31, 41.58, 38.98, 34.08, 33.61, 31.40, 25.92,
25.68, 23.98, 22.61, 19.31, 19.02. Anal. (C21H34O2) C, H.
3â-(Acet yloxy)-5-m et h yl-5â-est r a n e-17r-ca r b on it r ile
(28a ) a n d 3â-(Acetyloxy)-5-m eth yl-5â-estr a n e-17â-ca r bo-
n itr ile (28b). Using a two-step cyanation procedure similar
to that described for the preparation of steroids 11a and 11b,
compound 24b (240 mg, 0.72 mmol) was converted into a
mixture of compounds 28a and 28b (235 mg, 95%) which was
separated by HPLC (silica gel, 15% EtOAc in hexane, 3 mL/
min). The product ratio was 1:1, and steroid 28a eluted first
from the column.
38.77, 36.10, 34.92, 29.11, 27.11, 27.01, 25.81, 24.48, 22.75.
Anal. (C20H29NO) C, H, N.
5-Meth yl-3-oxo-5â-estr a n e-17â-ca r bon itr ile (30b). Us-
ing the earlier described J ones oxidation procedure, steroid
29b (85 mg, 0.28 mmol) was oxidized into product 30b (80 mg,
95%), which was obtained as colorless crystals: mp 158-160
°C (from EtOAc/hexane); IR 2920, 2235, 1709, 1450, 1384,
1348, 1316, 1274, 1233, 1175, 1127 cm-1; 1H NMR δ 2.69 (d, J
) 14.7 Hz, 1H, 4-Hax), 2.32 (t, J ) 9.3 Hz, 1H, CNCH), 0.95 (s,
6H, 2 × CH3); 13C NMR δ 212.54 (CdO), 121.00 (CN), 14.12
(18-CH3), 53.32, 47.91, 45.16, 44.15, 41.48, 40.09, 39.37, 38.69,
36.84, 36.04, 29.05, 26.80, 26.37, 25.69, 24.12, 22.61. Anal.
(C20H29NO) C, H, N.
3r-Hyd r oxy-5-m eth yl-5â-estr a n e-17r-ca r bon itr ile (31).
NaBH4 (25 mg) was added at room temperature to a solution
of compound 30a (30 mg, 0.1 mmol) in THF (10 mL) and EtOH
(5 mL). After 1 h, the excess NaBH4 was destroyed by addition
of HOAc, and then water (50 mL) was added. The mixture
was extracted with EtOAc (2 × 100 mL), and the combined
organic solvents were washed with brine (2 × 100 mL) and
dried over Na2SO4. Evaporation of organic solvent gave an
oil, which was separated by HPLC (silica gel, 40% EtOAc in
hexane, 3 mL/min) into products and 29a (first fraction, 10
mg, 25%) and 31 (second fraction, 18 mg, 45%). Compound
31 was obtained as an amorphous solidified foam: IR 3423,
2922, 2233, 1450, 1382, 1267, 1035, 1007 cm-1; 1H NMR δ 3.91
(m, 1H, HOCH), 2.58 (dd, J ) 8.8 Hz, J ) 1.8 Hz, 1H, CNCH),
0.98 (s, 3H, 5-CH3), 0.81 (s, 3H, 18-CH3); 13C NMR δ 122.25
(CN), 68.17 (HOCH), 17.90 (18-CH3), 51.28, 45.13, 44.10, 41.75,
40.93, 40.61, 40.05, 38.92, 35.27, 35.10, 29.63, 29.09, 27.22,
26.77, 25.71, 24.65, 21.28. Anal. (C20H31NO) C, H, N.
3r-Hyd r oxy-5-m eth yl-5â-estr a n e-17â-ca r bon itr ile (32).
Using the NaBH4 reduction procedure described immediately
above, compound 30b (50 mg, 0.17 mmol) was converted into
an oil which was separated by column chromatography (silica
gel, 35% EtOAc in hexane) into products 29b (first fraction,
23 mg, 46%) and 32 (second fraction, 25 mg, 50%). Compound
32 was obtained as colorless crystals: mp 161-163 °C (from
EtOAc/hexane); IR 3376, 2923, 2236, 1452, 1383, 1269, 1168,
Compound 28a was obtained as colorless crystals: mp 94-
96 °C; IR 2924, 2233, 1733, 1451, 1380, 1243, 1182, 1156, 1114,
1
1020 cm-1; H NMR δ 5.09 (m, 1H, AcOCH), 2.58 (d, J ) 8.7
Hz, 1H, CNCH), 2.04 (s, 3H, CH3CO), 1.11 (s, 3H, 5-CH3), 0.82
(s, 3H, 18-CH3); 13C NMR δ 170.49 (CH3CO), 122.38 (CN),
70.70 (AcOCH), 17.92 (CH3), 17.85 (CH3), 51.28, 45.52, 44.16,
41.52, 41.24, 40.12, 38.75, 35.11, 34.21, 32.67, 30.59, 27.24,
27.09, 25.89, 24.67, 23.65, 21.67. Anal. (C22H33NO2) C, H, N.
Compound 28b was obtained as colorless crystals: mp 117-
119 °C (from EtOAc/hexane); IR 2921, 2235, 1733, 1450, 1381,
1243, 1151, 1114, 1019 cm-1; 1H NMR δ 5.07 (m, 1H, AcOCH),
2.28 (t, J ) 9.8 Hz, 1H, CNCH), 2.04 (s, 3H, CH3CO), 1.11 (s,
3H, 5-CH3), 0.92 (s, 3H, 18-CH3); 13C NMR δ 170.54 (CH3CO),
121.33 (CN), 70.60 (AcOCH), 17.79 (CH3), 14.22 (CH3), 53.61,
45.48, 44.35, 41.55, 41.18, 40.34, 39.13, 37.12, 34.24, 32.67,
1
1040 cm-1; H NMR δ 3.91 (m, 1H, HOCH), 2.28 (t, J ) 9.5
Hz, 1H, CNCH), 0.98 (s, 3H, CH3), 0.91 (s, 3H, CH3); 13C NMR
δ 121.32 (CN), 14.20 (18-CH3), 68.16 (HOCH), 53.63, 45.10,
44.31, 41.80, 40.89, 40.52, 40.30, 39.26, 37.13, 35.30, 29.54,
29.09, 26.64, 26.54, 25.64, 24.34, 21.21. Anal. (C20H31NO) C,
H, N.
30.56, 26.96, 26.55, 25.83, 24.35, 23.66, 21.66. Anal. (C22H33
NO2) C, H, N.
-
3â-Hyd r oxy-5-m eth yl-5â-estr a n e-17r-ca r bon itr ile (29a )
a n d 3â-H yd r oxy-5-m et h yl-5â-est r a n e-17â-ca r b on it r ile
(29b). Using a hydrolysis procedure similar to that described
for the preparation of steroids 26a and 26b, a mixture of
steroids 28a and 28b (230 mg, 0.7 mmol) was converted into
a mixture of products 29a and 29b (195 mg, 97%), which was
separated by HPLC (silica gel, 30% EtOAc in hexane, 3 mL/
min). Steroid 29a eluted first from the column.
3r-Hyd r oxy-5-m eth yl-19-n or -5â-p r egn a n -20-on e (33a )
an d 3â-Hydr oxy-5-m eth yl-19-n or -5â-pr egn an -20-on e (33b).
As described earlier for the preparation of compounds 14a and
14b, a 1:1 mixture of steroids 29b and 32 (40 mg, 0.13 mmol)
was reacted (24 h at reflux) with CH3MgBr (4 mL, 12.0 mmol,
3.0 M solution in THF) in THF (20 mL) to give a mixture of
products 33a and 33b (40 mg, 64%), which was separated by
HPLC (silica gel, 35% EtOAc in hexane, 3 mL/min). Steroid
33b eluted first from the column.
Compound 29a was obtained as colorless crystals: mp 179-
181 °C (from EtOAc/hexane); IR 3358, 2920, 2693, 2233, 1474,
1451, 1383, 1337, 1268, 1224, 1174, 1103, 1068, 1005 cm-1
;
1H NMR δ 4.17 (m, 1H, HOCH), 2.58 (dd, J ) 8.9 Hz, J ) 2.0
Hz, 1H, CNCH), 1.21 (s, 3H, 5-CH3), 0.82 (s, 3H, 18-CH3); 13
C
Compound 33a was obtained as colorless crystals: mp 140-
142 °C (from EtOAc/hexane); IR 3322, 2921, 1702, 1450, 1383,
NMR δ 122.42 (CN), 67.61 (HOCH), 17.92 (CH3), 17.85 (CH3),
51.37, 45.94, 44.18, 41.55, 41.52, 40.14, 38.79, 37.48, 35.16,
1360, 1288, 1207, 1170, 1106, 1028, 1004 cm-1; 1H NMR δ 3.92
(m, 1H, HOCH), 2.54 (t, J ) 8.7 Hz, 1H, CHCOCH3), 2.12 (s,
3H, CH3CO), 0.97 (s, 3H, 5-CH3), 0.60 (s, 3H, 18-CH3); 13C
NMR δ 209.76 (COCH3), 68.34 (HOCH), 63.93 (CHCOCH3),
13.34 (18-CH3), 55.92, 45.19, 44.13, 41.47, 41.05, 40.65, 39.35,
39.11, 35.37, 31.52, 29.62, 29.16, 26.65, 26.08, 24.22, 22.77,
21.25. Anal. (C21H34O2) C, H.
32.67, 31.41, 27.25, 27.18, 26.57, 25.85, 24.67. Anal. (C20H31
NO) C, H, N.
-
Compound 29b was obtained as colorless crystals: mp 175-
177 °C (from EtOAc/hexane); IR 3367, 2920, 2233, 1736, 1452,
1
1383, 1268, 1173, 1103, 1006 cm-1; H NMR δ 4.15 (m, 1H,
HOCH), 2.27 (t, J ) 9.6 Hz, 1H, CNCH), 1.21 (s, 3H, 5-CH3),
0.92 (s, 3H, 18-CH3); 13C NMR δ 121.32 (CN), 67.56 (HOCH),
17.07 (5-CH3), 14.19 (18-CH3), 53.69, 45.88, 44.34, 41.53, 41.48,
40.34, 39.13, 37.48, 37.17, 32.69, 31.38, 27.03, 26.55, 25.78,
24.34. Anal. (C20H31NO) C, H, N.
Compound 33b was obtained as colorless crystals: mp 159-
160 °C (from EtOAc/hexane); IR 3347, 2921, 2870, 1698, 1449,
1383, 1359, 1272, 1205, 1165, 1105, 1028, 1005 cm-1; 1H NMR
δ 4.16 (m, 1H, HOCH), 2.53 (t, J ) 8.9 Hz, 1H, CHCOCH3),
2.12 (s, 3H, CH3CO), 1.20 (s, 3H, 5-CH3), 0.61 (s, 3H, 18-CH3);
13C NMR δ 209.69 (COCH3), 67.69 (HOCH), 63.96 (CHCOCH3),
17.07 (5-CH3), 13.31 (18-CH3), 55.97, 45.94, 44.14, 41.63, 41.17,
39.21, 39.14, 37.59, 32.70, 31.45, 27.01, 26.59, 26.20, 24.20,
22.75. Anal. (C21H34O2) C, H.
4-P r egn en e-3r,20(R)-d iol (35a ) a n d 4-p r egn en e-3â,20-
(R)-d iol (35b). K-Selectride (1.0 M solution in THF, 41.9 mL,
41.9 mmol) was added to a cooled (-78 °C), stirred solution of
progesterone 34 (7.5 g, 20.9 mmol) in dry THF (50 mL) under
nitrogen. After 5 h, the reaction mixture was allowed to warm
5-Meth yl-3-oxo-5â-estr a n e-17r-ca r bon itr ile (30a ). Us-
ing the earlier described J ones oxidation procedure, steroid
29a (40 mg, 0.13 mmol) was oxidized into product 30a (36 mg,
91%), which was obtained as colorless crystals: mp 140-142
°C (from EtOAc/hexane); IR 2921, 2233, 1710, 1476, 1451,
1383, 1347, 1314, 1272, 1233, 1212, 1180, 1126, 1077 cm-1
;
1H NMR δ 2.72 (d, J ) 14.3 Hz, 1H, 4-Hax), 2.62 (dd, J ) 8.9
Hz, J ) 2.0 Hz, 1H, CNCH), 0.95 (s, 3H, 5-CH3), 0.86 (s, 3H,
18-CH3); 13C NMR δ 212.60 (CdO), 122.10 (CN), 17.82 (18-
CH3), 51.13, 47.96, 45.26, 44.02, 41.48, 39.93, 39.43, 39.10,

4230 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21
Han et al.
to room temperature, and stirring was continued for another
4 h. The reaction was stopped by the addition of 10% NaOH
(100 mL). The mixture was extracted with EtOAc (3 × 200
mL). The combined organic layers were washed with brine
(2 × 150 mL) and dried over Na2SO4. Evaporation of organic
solvent gave a white solid, which was purified by chromatog-
raphy (silica gel, 25% EtOAc in hexane) to yield a mixture of
products 35a and 35b (7.4 g, 98%). A small portion of the
product mixture was separated by HPLC (silica gel, 50%
EtOAc in hexane, 3 mL/min). The product ratio was 1:1, and
steroid 35b eluted first from the column.
and 38b were oxidized (J ones reagent), without characteriza-
tion, to steroid 40.
Compound 39a was obtained as colorless crystals: mp 152-
154 °C (from EtOAc/hexane); IR 3448, 2939, 1713, 1458, 1376,
1301, 1267, 1218, 1174, 1122, 1098, 1011 cm-1; 1H NMR δ 3.74
(m, 1H, CHOH), 3.01 (d, 1H, J ) 14.7 Hz, 4-Hax), 1.15 (d, 3H,
J ) 6.0 Hz, 21-CH3), 0.93 (s, 3H, CH3), 0.88 (s, 3H, CH3), 0.77
(s, 3H, CH3); 13C NMR δ 213.48 (CdO), 70.52 (CHOH), 17.10
(CH3), 12.46 (CH3), 58.44, 56.27, 49.06, 42.26, 41.91, 41.82,
40.07, 37.81, 37.59, 35.34, 35.01, 32.09, 27.58, 25.66, 24.67,
24.43, 23.69, 21.38. Anal. (C22H36O2) C, H.
Compound 39b was obtained as colorless crystals: mp 153-
154 °C (from EtOAc/hexane); IR 3466, 2938, 1708, 1446, 1383,
Compound 35a was obtained as colorless crystals: mp 172-
173 °C (lit.18 mp 155-158 °C). The IR, 1H NMR, and 13C NMR
spectra were identical to those reported.18 Anal. (C21H34O2)
C, H.
1
1263, 1219, 1157, 1118, 1010 cm-1; H NMR δ 3.65 (m, 1H,
CHOH), 2.94 (d, 1H, J ) 14.4 Hz, 4-Hax), 2.30 (tt, 1H, J )
14.5 Hz, J ) 6.2 Hz, 17-CH), 1.17 (d, 3H, J ) 6.3 Hz, 21-CH3),
0.84 (s, 3H, CH3), 0.80 (s, 3H, CH3), 0.61 (s, 3H, CH3); 13C NMR
δ 213.36 (CdO), 70.25 (CHOH), 17.07 (CH3), 12.53 (CH3),
58.37, 56.61, 49.01, 41.84, 41.79, 41.55, 39.01, 37.77, 37.53,
35.31, 34.77, 32.08, 27.49, 25.73, 24.64, 24.05, 23.46, 21.20.
Anal. (C22H36O2) C, H.
Compound 40 was obtained as colorless crystals: mp 158-
160 °C; IR 2938, 1709, 1447, 1382, 1350, 1193, 1160 cm-1; 1H
NMR δ 2.55 (t, 1H, J ) 8.9 Hz, CHCOCH3), 2.13 (s, 3H, 21-
CH3), 1.16 (s, 3H, CH3), 0.95 (s, 3H, CH3), 0.63 (s, 3H, 18-CH3);
13C NMR δ 212.37 (CdO), 209.58 (CdO), 14.21 (CH3), 13.53
(CH3), 63.66, 56.59, 52.14, 46.12, 44.27, 39.85, 39.02, 37.77,
37.42, 35.30, 34.45, 32.66, 31.51, 26.45, 24.28, 22.76, 21.66,
20.87. Anal. (C22H34O2) C, H.
Compound 41 was obtained as colorless crystals: mp 174-
175 °C (from EtOAc/hexane); IR 2943, 1707, 1450, 1386, 1358,
1160 cm-1; 1H NMR δ 2.95 (d, 1H, J ) 14.9 Hz, 4-Hax), 2.51 (t,
1H, J ) 8.8 Hz, CHCOCH3), 2.08 (s, 3H, 21-CH3), 0.87 (s, 3H,
CH3), 0.83 (s, 3H CH3), 0.58 (s, 3H, 18-CH3); 13C NMR 213.07
(CdO), 209.46 (CdO), 17.04 (CH3), 13.30 (CH3), 63.60, 56.92,
48.94, 43.88, 41.76, 41.72, 39.01, 37.76, 37.47, 35.18, 35.07,
32.06, 31.50, 27.41, 24.60, 24.31, 22.79, 21.46. Anal. (C22H34O2)
C, H.
Compound 35b was obtained as colorless crystals: mp 178-
180 °C (lit.19 mp 163-172 °C); IR 3373, 1658 cm-1; 1H NMR δ
5.28 (s, 1H, CHdC), 4.18-4.11 (m, 1H, 3-CHOH), 3.77 -3.68
(m, 1H, 20-CHOH), 1.13 (d, J ) 6.0 Hz, 1H, 21-CH3), 1.11 (s,
3H, CH3), 0.77 (s, 3H, CH3); 13C NMR δ 147.57, 123.32 (CdC),
70.54 (HOCH), 67.94 (HOCH), 18.89 (CH3), 12.47 (CH3), 58.43,
55.58, 54.40, 42.37, 39.90, 37.33, 35.74, 35.34, 33.08, 32.15,
29.48, 25.57, 24.43, 23.62, 20.86. Anal. (C21H34O2) C, H.
3′,4â-Cyclopr op[4,5]-5r-p r egn a n e-3,20-d ion e (37a ) Con -
tain in g 3′,4r-Cyclopr op[4,5]-5â-pr egn an e-3,20-dion e (37b).
The procedure used for the preparation of these compounds
was analogous to that described earlier for the preparation of
steroid 21a . In brief, diiodomethane (8.8 g) was added to a
stirred mixture of zinc-copper couple (3 g) and a small crystal
of iodine in ether (50 mL). The mixture was warmed with a
tungsten lamp until the reaction started, and the reaction was
continued for 30 min on a water bath at 60 °C. A solution of
the mixture of compounds 35a and 35b (3.0 g, 9.4 mmol) was
added over 10 min, and the mixture was stirred for 3 days at
50 °C. The crude product that was obtained was purified by
chromatography (silica gel, 30% EtOAc in hexane) to give a
mixture of unreacted steroids 35a and 35b and products 36a
and 36b (1.87 g), which was oxidized immediately with J ones
reagent in acetone (100 mL) at room temperature. The oil that
was obtained from the oxidation was purified by chromatog-
raphy (silica gel, 30% EtOAc in hexane) to yield an unchar-
acterized solid (1.72 g). This solid was treated with O3, and
the crude product that was obtained was purified by chroma-
tography (silica gel, 30% EtOAc in hexane) to give an insepa-
rable isomeric mixture of steroids 37a and 37b (820 mg, 27%
overall from 35a and 35b) as a white solid: mp 108-110 °C;
IR 1701, 1684, 1477, 1447, 1383, 1357, 1266, 1226, 1157, 1116,
1080 cm-1; 1H NMR δ 2.51-2.44 (m, 1H, CHCOCH3), 2.06 (s,
3H, 21-CH3), 2.05 (s, 3H, 21-CH3), 1.02 (s, 3H, CH3), 0.99 (s,
3H, CH3); 13C NMR δ 209.82 (CdO), 209.35 (CdO), 209.22
(CdO), 17.79 (CH3), 17.66 (CH3), 13.29 (CH3), 13.20 (CH3),
63.65, 63.44, 56.13, 56.04, 51.70, 45.66, 44.03, 43.83, 38.76,
38.62, 38.13, 36.73, 35.72, 35.31, 34.95, 34.73, 34.31, 34.20,
32.52, 32.16, 32.11, 32.02, 31.48, 31.38, 30.62, 30.50, 29.92,
27.47, 24.26, 24.22, 22.64, 21.51, 20.48, 18.41. Anal. (C22H32O2)
C, H.
3r-Hyd r oxy-5-m eth yl-5r-p r egn a n -20-on e (42) a n d 3â-
Hyd r oxy-5-m eth yl-5â-p r egn a n -20-on e (43b). The proce-
dure used was analogous to that used for the preparation of
steroids 23a and 23b. Briefly, K-Selectride (1.4 mL, 1.5 mmol,
1.0 M solution in THF) was added to a cooled (-78 °C), stirred
solution of compounds 40 and 41 (240 mg, 73 mmol) in dry
THF (20 mL) under nitrogen. The reaction time was 4 h. The
solid obtained was purified by chromatography (silica gel, 30%
EtOAc in hexane) to give a mixture (230 mg) of steroids 40,
42, and 43b. Separation of this mixture by HPLC (silica gel,
23% EtOAc in hexane, 3 mL/min) gave pure steroids 40 (60
mg, 24%), 42 (30 mg, 12%), and 43b (130 mg, 54%).
Compound 42 was obtained as colorless crystals: mp 173-
175 °C (from EtOAc/hexane); IR 3419, 2934, 1698, 1449, 1382,
1
1357, 1260, 1203, 1156, 1046, 1006 cm-1; H NMR δ 4.10 (m,
1H, CHOH), 2.55 (t, 1H, J ) 8.5 Hz, CHCOCH3), 2.12 (s, 3H,
21-CH3), 1.24 (s, 3H, CH3), 0.95 (s, 3H, CH3), 0.60 (s, 3H, 18-
CH3); 13C NMR δ 209.58 (CdO), 14.08 (CH3), 13.67 (CH3),
66.93, 63.77, 56.97, 45.68, 44.50, 43.07, 39.24, 37.69, 35.50,
35.16, 34.37, 31.53, 29.23, 27.07, 25.31, 24.17, 22.74, 21.72,
21.33. Anal. (C22H36O2) C, H.
20(S)-Hyd r oxy-5-m eth yl-5â-p r egn a n -3-on e (39a ), 20(R)-
Hyd r oxy-5-m eth yl-5â-p r egn a n -3-on e (39b), 5-Meth yl-5r-
p r egn a n e-3,20-d ion e (40), a n d 5-Meth yl-5â-p r egn a n e-
3,20-d ion e (41). The procedure used was analogous to that
described earlier for the preparation of steroids 22a and 22b.
Briefly, Li wire (180 mg, 26 mmol) was added to stirred liquid
NH3 (ca. 70 mL). After 30 min, a solution of steroids 37a and
37b (400 mg, 1.2 mmol) in THF (10 mL) was added slowly,
stirring was continued for 2 h, and 1,2-dibromoethane was
added dropwise to discharge the blue color. After workup, the
solid that was obtained was separated into two fractions by
chromatography (silica gel, 30% EtOAc in hexane). The first
fraction contained a mixture (150 mg, 38%) of products 40 and
41. The second fraction contained a mixture (210 mg, 53%)
of steroids 38a , 38b, 39a , and 39b. The mixture of steroids
40 and 41 was separated by HPLC (silica gel, 40% EtOAc in
hexane, 3 mL/min). A portion (110 mg) of the second fraction
was purified further by HPLC (40% EtOAc, 10% ClCH2CH2Cl
in hexane, 5 mL/min) to give unseparated steroids 38a and
38b along with separated steroids 39a and 39b. Steroids 38a
Compound 43b was obtained as colorless crystals: mp 195-
196 °C (from EtOAc/hexane); IR 3378, 3322, 2937, 2874, 1704,
1
1450, 1383, 1357, 1182, 1152, 1104 cm-1; H NMR δ 4.15 (m,
1H, CHOH), 2.50 (t, 1H, J ) 8.8 Hz, CHCOCH3), 2.10 (s, 3H,
21-CH3), 1.08 (s, 3H, CH3), 0.82 (s, 3H, CH3), 0.57 (s, 3H, 18-
CH3); 13C NMR δ 209.70 (CdO), 67.85 (CHOH), 17.65 (CH3),
13.18 (CH3), 63.76, 57.06, 43.92, 40.95, 39.12, 38.94, 37.71,
36.48, 35.45, 35.01, 31.46, 27.73, 27.52, 27.43, 25.80, 24.38,
22.74, 20.97. Anal. (C22H36O2) C, H.
3r-Hydr oxy-5-m eth yl-5â-pr egn an -20-on e (43a). The pro-
cedure used was analogous to that described earlier for the
preparation of steroid 31. Briefly, NaBH4 (16 mg) was added
at room temperature to a solution of compound 41 (120 mg,
0.36 mmol) in ethanol (20 mL). The reaction time was 1 h.
The solid obtained was purified by HPLC (silica gel, 30%
EtOAc in hexane, 4 mL/min) to give compound 43b (first
fraction, 80 mg, 59%) and compound 43a (second fraction, 30
mg, 22%).

Neurosteroid Analogues
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21 4231
140. (c) Olsen, R. W.; Sapp, D. W. Neuroactive Steroid Modula-
tion of GABAA Receptors. Adv. Biochem. Psychopharmacol. 1995,
48, 57-74. (d) Gee, K. W.; McCauley, L. D.; Lan, N. C. A Putative
Receptor for Neurosteroids on the GABAA Receptor Complex:
the Pharmacological Properties and Therapeutic Potential of
Epalons. Crit. Rev. Neurobiol. 1995, 9, 207-227.
Compound 43a was obtained as colorless crystals: mp 176-
178 °C (from EtOAc/hexane); IR 3389, 2932, 1702, 1451, 1383,
1360, 1236, 1211, 1185, 1157, 1123, 1043 cm-1 1H NMR δ
;
3.87-3.94 (m, 1H, CHOH), 2.53 (t, 1H, J ) 8.9 Hz, CHCOCH3),
2.12 (s, 3H, 21-CH3), 0.91 (s, 3H, CH3), 0.80 (s, 3H, CH3), 0.58
(s, 3H, 18-CH3); 13C NMR δ 209.72 (CdO), 68.30 (CHOH),
17.15 (CH3), 13.23 (CH3), 63.77, 56.97, 43.92, 42.04, 41.21,
39.12, 37.76, 37.20, 36.33, 35.40, 31.52, 30.40, 30.37, 27.13,
25.09, 24.42, 22.80, 20.94. Anal. (C22H36O2) C, H.
(3) (a) Selye, H. Correlations between the Chemical Structure and
the Pharmacological Actions of the Steroids. Endocrinology 1942,
30, 437-453. (b) Figdor, S. K.; Kodet, M. J .; Bloom, B. M.;
Agnello, E. J .; P’An, S. Y.; Laubach, G. D. Central Activity and
Structure in a Series of Water-Soluble Steroids. J . Pharmacol.
Exp. Ther. 1957, 119, 299-309. (c) Witzel, H. Steroids with
Anesthetic Action. Z. Vitam. Horm. Fermentforsch 1959, 10, 46-
75. (d) Cocker, J . D.; Elks, J .; May, P. J .; Nice, F. A.; Phillipps,
G. H.; Wall, W. F. Action of Some Steroids on the Central
Nervous System of the Mouse. I. Synthetic Methods. J . Med.
Chem. 1965, 8, 417-425. (e) Atkinson, R. M.; Davis, B.; Pratt,
M. A.; Tomich, E. G. Action of Some Steroids on the Central
Nervous System of the Mouse. II. Pharmacology. J . Med. Chem.
1965, 8, 426-432. (f) Gyermek, L.; Iriarte, J .; Crabbe´, P.
Steroids. CCCX. Structure-Activity Relationship of Some Ste-
roidal Hypnotic Agents. J . Med. Chem. 1968, 11, 117-125. (g)
Phillipps, G. H. Structure-Activity Relationships in Steroidal
Anaesthetics. In Molecular Mechanisms in General Anaesthesia;
Halsey, M. J ., Millar, R. A., Sutton, J . A., Eds.; Churchill
Livingston: New York, 1974; pp 32-47. (h) Phillips, G. H.
Structure-Activity Relationships in Steroidal Anaesthetics. J .
Steroid Biochem. 1975, 6, 607-613. (i) Phillipps, G. H.; Ayres,
B. E.; Bailey, E. J .; Ewan, G. B.; Looker, B. E.; May, P. J . Water-
Soluble Steroidal Anaesthetics. J . Steroid Biochem. 1979, 11,
79-86.
(4) Purdy, R. H.; Morrow, A. L.; Blinn, J . R.; Paul, S. A. 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.
(5) A preliminary account of this work was presented (Abstract 157,
Medicinal Chemistry Division) at the 210th ACS National
Meeting in Chicago, IL. August 20-24, 1995.
(6) Counsell, R. E. Isomeric Estrane Derivatives. Tetrahedron 1961,
15, 202-211.
(7) Bowden, K.; Heilbron, I. M.; J ones, E. R. H.; Weedon, B. C. L.
Acetylenic Compounds. I. Preparation of Acetylenic Ketones by
Oxidation of Acetylenic Carbinols and Glycols. J . Chem. Soc.
1946, 39-45.
(8) Yoneda, R.; Harusawa, S.; Kurihara, T. Cyanophosphate: An
Efficient Intermediate for Conversion of Carbonyl Compounds
to Nitriles. Tetrahedron Lett. 1989, 30, 3681-3684.
(9) Curts, S. W.; Wren, D. L.; Cooper, G. F. Synthesis of 5â,17R-
19-Norpregn-20-yne-3â,17-diol and of 5â,17R-19-Norpregn-20-
yne-3R,17-diol, Human Metabolites of Norethindrone. Steroids
1991, 56, 8-11.
(10) Ruzicka, L.; Goldberg, M. W.; Meyer, J . Sex Hormones. II. The
Synthesis of Testicular Hormone (Androsterone) and Its Stere-
oisomers. Helv. Chim. Acta 1934, 17, 1395-1406.
(11) Varasi, M.; Walker, K. A.; Maddox, M. L. A Revised Mechanism
for the Mitsunobu Reaction. J . Org. Chem. 1987, 52, 4235-4238.
(12) Hu, Y.; Zorumski, C. F.; Covey, D. F. Neurosteroid Analogues:
Structure-Activity Studies of Benz[e]indene Modulators of
GABAA Receptor Function. I. 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.
(13) Clayton, J . C. H.; Faulkner, P. J .; J ones, W. R.; Phillipps, G. H.;
Bailey, E. J . 3R-Hydroxy-5R-steroids. Ger. Offen. 2,255,108,
1973.
Electr op h ysiology. Hippocampal cultures were prepared
from 1-2-day-old albino rat pups and maintained as described
previously.27 Experiments were carried out at room temper-
ature (∼22 °C) using cultures that had been maintained in
vitro for 3-10 days. At the time of an experiment the growth
media was exchanged for a solution containing the following
(in mM): 140 NaCl, 5 KCl, 2 CaCl2, 2 MgCl2, 10 glucose, 10
hydroxyethylpiperazine ethanesulfonic acid (HEPES), and
0.001 tetrodotoxin (TTX) with pH adjusted to 7.3. TTX was
included to block voltage-gated Na+ currents and to diminish
spontaneous synaptic currents. Voltage clamp recordings were
obtained using whole-cell patch clamp methods.22 Recording
electrodes were fashioned from 1.2 mm borosilicate glass
capillaries (World Precision Instruments) using a Flaming-
Brown P-87 horizontal pipet puller (Sutter Instruments) and
had resistances of 5-8 MΩ when fire-polished and filled with
a solution containing the following (in mM): 140 CsCl, 4 NaCl,
4 MgCl2, 0.5 CaCl2, 10 HEPES, and 5 ethyleneglycol-bis(â-
aminoethyl ether)-N,N,N ′,N ′-tetraacetic acid (EGTA) with pH
adjusted to 7.3 using CsOH. Currents were filtered at 1.5 kHz
and were digitized at 0.125 kHz using pCLAMP V 5.5 (Axon
Instruments). Data were analyzed using pCLAMP V 5.5,
Sigmaplot for Windows V 2.0, and routines written in Axoba-
sic. The data in this paper are presented as the mean ( SEM.
GABA stock solutions were prepared in the extracellular
solution. Test compound stock solutions were prepared in
DMSO and were diluted with the extracellular solution at the
time of an experiment. The final DMSO concentration was
<0.2%, a concentration that does not alter GABA currents in
hippocampal neurons. Compounds were applied by pressure
ejection from pipets positioned within 5 µm of the recorded
neuron using a 500 ms jet of compressed air at 10-20 psi. This
system allows no discernable drug leakage between applica-
tions and affords reliable repeated drug delivery. The con-
centrations of drugs reported are those in the pipet. The actual
concentration at the cell is likely to be less due to diffusion
and the fact that the entire cell is not uniformly exposed to
the pipet contents.
Molecu la r Mod elin g. Molecular modeling was performed
on a SiliconGraphics Iris Indigo Elan 4000 computer using the
Sybyl Molecular Modeling Software, Version 6.2 from Tripos
Associates, Inc., St. Louis, MO. Minimizations were performed
using the Powell minimization option (without charges) in the
Sybyl Program.
Ack n ow led gm en t. The authors thank Ms. Ann
Benz and Mr. J ian Que for technical assistance. The
work was supported by NIH Grant GM 47969, Research
Scientist Development Award MH00964 (C.F.Z.), and
a fellowship from the Bantly Foundation. Assistance
was also provided by the Washington University High
Resolution NMR Facility supported in part by NIH 1
S10 RR00204 and a gift from Monsanto Co.
(14) Dauben, W. G.; Laug, P.; Berezin, G. H. Preparation and
Dimerization of 4,5-Methano-3-keto Steroids. J . Org. Chem.
1966, 31, 3869-3871.
(15) Dauben, W. G.; Deviny, E. J . Reductive Opening of Conjugated
Cyclopropyl Ketones with Lithium in Liquid Ammonia. J . Org.
Chem. 1966, 31, 3794-3797.
(16) Simmons, H. E.; Smith, R. D. A New Synthesis of Cyclopropanes
J . Am. Chem. Soc. 1959, 81, 4256-4264.
(17) Hartman, J . A. The Preparation of Four 3,17â-Dihydrox-
yestrenes. J . Am. Chem. Soc. 1955, 77, 5151-5154.
(18) Weibe, J . P.; Deline, C.; Buckingham, K. D. Synthesis of the
Allylic Gonadal Steroids, 3R-Hydroxy-4-pregnen-20-one and 3R-
Hydroxy-4-Androsten-17-one, and of 3R-Hydroxy-5R-pregnan-
20-one. Steroids 1985, 45, 39-51.
(19) J ulia, S.; Simon, P. VIII. 4-Vinyl- and 4-Ethynylprogesterone.
Bull. Soc. Chim. Fr. 1964, 331-333.
(20) Han, M.; Hu, Y.; Zorumski, C. F.; Covey, D. F. Neurosteroid
Analogues. 3. The Synthesis and Electrophysiological Evaluation
of Benz[e]indene Congeners of Neuroactive Steroids Having the
5â-Configuration. J . Med. Chem. 1995, 38, 4548-4556.
(21) (a) Lee, H.; Wolfe, M. E. C-18 Functional Steroids and D-Homo
Steroids. J . Org. Chem. 1967, 32, 192-196. (b) Covey, D. F.
Synthesis of 17â-[(1S)-1-hydroxy-2-propynyl]- and 17â-[(1R)-1-
hydroxy-2-propynyl]androst-4-en-3-one. Potential Suicide Sub-
strates of 20R- and 20â-Hydroxysteroid Dehydrogenases. Ste-
roids 1978, 34, 199-206.
Refer en ces
(1) (a) Harrison, N. L.; Simmonds, M. A. Modulation of the GABA
Receptor Complex by a Steroid Anaesthetic. Brain Res. 1984,
323, 287-292. (b) 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.
(2) For recent reviews, see: (a) Lambert, J . J .; Belelli, D.; Hill-
Venning, C.; Peters, J . A. Neurosteroids and GABAA Receptor
Function. Trends Pharmacol. Sci. 1995, 16, 295-303. (b) Mc-
Carthy, M. M. Functional Significance of Steroid Modulation of
GABAergic Neurotransmission: Analysis at the Behavioral,
Cellular, and Molecular Levels. Horm. Behav. 1995, 29, 131-

4232 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 21
Han et al.
(22) Hamill, O. P.; Marty, A.; Neher, E.; Sakmann, B.; Sigworth, F.
J . Improved Patch-clamp Techniques for High Resolution Cur-
rent Recordings from Cells and Cell-Free Membrane Patches.
Pflugers Arch. 1981, 391, 85-100.
Pharmacol. 1990, 37, 263-270. (c) Gee, K. W.; Lan, N. C.
γ-Aminobutyric AcidA
Receptor Complexes in Rat Frontal Cortex
and Spinal Cord Show Differential Responses to Steroid Modu-
lation. Mol. Pharmacol. 1991, 40, 995-999. (d) Sapp, D. W.;
Witte, U.; Turner, D. M.; Longoni, B.; Kokka, N.; Olsen, R. W.
Regional Variation in Steroid Anesthetic Modulation of [35S]-
TBPS Binding to γ-Aminobutyric AcidA Receptors in Rat Brain.
J . Pharmacol. Exp. Ther. 1992, 262, 801-808. (e) Prince, R. J .;
Simmonds, M. A. Differential Antagonism by Epipregnanolone
of Alphaxalone and Pregnanolone Potentiation of [3H]Fluni-
trazepam Binding Suggests More Than One Class of Binding
Site for Steroids at GABAA Receptors. Neuropharmacology 1993,
32, 59-63. (f) McCauley, L. D.; Liu, V.; Chen, J .-S.; Hawkinson,
J . E.; Lan N. C.; Gee, K. W. Selective Actions of Certain
Neuroactive Pregnandiols at the γ-Aminobutyric Acid Type A
Receptor Complex in Rat Brain. Mol. Pharmacol. 1995, 47, 354-
362. (g) Hawkinson, J . E.; Kimbrough, C. L.; McCauley, L. D.;
Bolger, M. B.; Lan, N. C.; Gee, K. W. The Neuroactive Steroid
3R-Hydroxy-5â-Pregnan-20-one is a Two-Component Modulator
of Ligand Binding to the GABAA Receptor. Eur. J . Pharmacol.
1994, 269, 157-163.
(23) Belelli, D.; Lan, N. C.; Gee, K. W. Anticonvulsant Steroids and
the GABA/Benzodiazepine Receptor-Chloride Ionophore Com-
plex. Neurosci. Biobehav. Rev. 1990, 14, 315-322.
(24) 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.
(25) (a) Killisch, I.; Dotti, C. G.; Laurie, D. J .; Luddens, H.; Seeburg,
P. H. Expression Patterns of GABAA Receptor Subtypes in
Developing Hippocampal Neurons. Neuron 1991, 7, 927-936.
(b) Wisden, W.; Laurie, D. J .; Monyer, H.; Seeburg, P. H. The
Distribution of 13 GABAA Receptor Subunit mRNAs in the Rat
Brain. I. Telencephalon, Diencephalon, Mesencephalon. J . Neu-
rosci. 1992, 12, 1040-1062. (c) Laurie, D. J .; Wisden, W.;
Seeburg, P. H. The Distribution of Thirteen GABAA Receptor
Subunit mRNAs in the Rat Brain. III. Embryonic and Postnatal
Development. J . Neurosci. 1992, 12, 4151-4172.
(26) (a) Turner, J . P.; Simmonds, M. A. Modulation of the GABAA
Receptor Complex by Steroids in Slices of Rat Cuneate Nucleus.
Br. J . Pharmacol. 1989, 96, 409-417. (b) Morrow, A. L.; Pace,
J . R.; Purdy, R. H.; Paul, S. M. Characterization of Steroid
Interactions with γ-Aminobutyric Acid Receptor-Gated Ion
Channels: Evidence for Multiple Steroid Recognition Sites. Mol.
(27) Thio, L. L.; Clifford, D. B.; Zorumski, C. F. Characterization of
Quisqualate Receptor Desensitization in Culture Postnatal Rat
Hippocampal Neurons. J . Neurosci. 1991, 11, 3430-3441.
J M960304P