Effects of 3α-amino-5α-pregnan-20-one on GABAA receptor: Synthesis, activity and cytotoxicity

Article abstract of DOI:10.1135/cccc20041506

The 3α-hydroxy function has been considered essential for the biological activity of neurosteroids at the GABA_A receptor. It was found that 3α-amino-5α-pregnan-20-one (3) increased the binding of [³H]flunitrazepam at the GABA_A receptor in the primary culture of cortical neurons. This derivative did not display cytotoxicity at relevant neuroactive concentrations, and its structure enabled us to gain further insight into possible functional group modifications in position 3α. Various synthetic methods were investigated in search for the most suitable synthetic approach.

Full text of DOI:10.1135/cccc20041506

1506
Matyáš, Kasal, Babot Riera, Sunol:
EFFECTS OF 3α-AMINO-5α-PREGNAN-20-ONE ON GABAA RECEPTOR:
SYNTHESIS, ACTIVITY AND CYTOTOXICITY+
b1
Libor MATYÁŠa1, Alexander KASALa2,*, Zoila BABOT RIERA and
b2
Cristina E. SUNOL
a
Institute of Organic Chemistry and Biochemistry, Academy of Sciences of the Czech Republic,
Flemingovo nám. 2, 166 10 Prague 6, Czech Republic; e-mail: matyasl@centrum.cz,
kasal@uochb.cas.cz
Departament de Neuroquímica, Institut d’Investigacions Biomédiques de Barcelona,
CSIC (IDIBAPS), E-08036 Barcelona, Spain; e-mail: zbrnqi@iibb.csic.es, csenqi@iibb.csic.es
1
2
b
1
2
Received Jan uary 21, 2004
Accepted May 10, 2004
Th e 3α-h ydroxy fun ction h as been con sidered essen tial for th e biological activity of n euro-
steroids at th e GABAA receptor. It was foun d th at 3α-am in o-5α-pregn an -20-on e (3) in creased
th e bin din g of [3H]flun itrazepam at th e GABAA receptor in th e prim ary culture of cortical
n euron s. Th is derivative did n ot display cytotoxicity at relevan t n euroactive con cen tration s,
an d its structure en abled us to gain furth er in sigh t in to possible fun ction al group m odifica-
tion s in position 3α. Various syn th etic m eth ods were in vestigated in search for th e m ost
suitable syn th etic approach .
Keyw ord s: GABAA receptor; Steroids; Neurosteroids; Allopregn an olon e; [3H]Flun itrazepam ;
3α-Am in o-5α-pregn an -20-on e; Cytotoxicity; MoO3.
In search 2–8 for n ew an alogues of th e n atural n eurosteroid, 3α-h ydroxy-
5α-pregn an -20-on e (1, allopregn an olon e), we h ave tested a n um ber of
5α-pregn an e derivatives, m ostly usin g in vitro tests, wh ich m easured th e
bin din g of radiolabelled ligan ds to γ-am in obutyric acid receptors (GABAA
receptor) in th e absen ce an d presen ce of th e tested com poun ds. Th e la-
belled ligan ds used were [3H]m uscim ol an d [35S]tert-butylbicyclo[2.2.2]-
ph osporoth ion ate (TBPS)4. Th e com poun ds tested were ch osen from th e
ban k of products prepared in our laboratory over th e past few decades. On e
of th e com poun ds – 3α-fluoro-5α-pregn an -20-on e (2) – exh ibited surprisin g
activity in an an im al beh avioral test8. Th e result seem ed to con tradict th e
h ith erto h eld axiom of th e essen tiality of a 3α-h ydroxy group for n euron al
+ Part CDXVII in th e Series On Steroids; Part CDXVI see lit.1
Collect. Czech. Chem. Commun. (Vol. 69) (2004)
doi:10.1135/cccc20041506

GABAA Receptor
1507
activity. Th is prom pted us to evaluate th e n euron al activity of oth er
3α-substituted9–13 pregn an -20-on es by determ in in g th eir activity on
n euron al GABAA receptors expressed on livin g cultured n euron al cells first.
It h as been reported th at th e active n eurosteroid – 3α-h ydroxy-5α-pregn an -
20-on e (1) – allosterically in creases [3H]flun itrazepam bin din g in rat brain
m em bran es14. We h ave previously reported th at th e prim ary n euron al cul-
ture, expressin g GABAA receptors, respon ds to GABAA receptor m odulators
by in creasin g [3H]flun itrazepam bin din g15–17. Th us, th e use of such culture
would allow us to test n euroactive com poun ds actin g on GABAA receptors
as well as to assess th eir cytotoxicity in th e sam e in vitro system .
COCH3
1, R = OH
2, R = F
3, R = NH2
R
H
A 3α-am in o group would probably n ot m ake th e an alogue m ore stable to
m etabolic processes. However, th e group would ren der it m ore polar an d
h en ce m ore soluble in aqueous body fluids. 3α-Am in o-5α-pregn an -20-on e
(3) is a kn own alkaloid called fun tum in , first isolated from Funtumia latifola
an d Holarrhena febrifuga18,19. It was prepared from 3-h ydroxyim in o-5α-
pregn an -20-on e (4; Sch em e 1) by catalytic h ydrogen ation an d subsequen t
oxidation 20. Th e h ydrogen ation reportedly yielded 75% of th e 3α-am in o
20β-alcoh ol. Sin ce Dave21 described th e selective con version of 5α-pregn an -
3,20-dion e (5) in to th e 3-h ydroxyim in o derivative 4, a straigh tforward re-
duction of th is oxim e in to fun tum in (3) appeared to be th e best way
(Sch em e 1). Alth ough oth er auth ors (e.g.22) h ave also described th e prefer-
en tial form ation of 3α-am in o-5α-steroids in th e catalytic h ydrogen ation of
3-oxim es, in our h an ds th e desired am in e 3 was accom pan ied by a con sid-
erable am oun t (1:1) of its 3β isom er (am in e 6, as determ in ed by 1H NMR
after oxidation of th e C-20 h ydroxy group).
Altern atively, th e reduction of oxim e 4 was attem pted with oth er reduc-
in g agen ts (Table I). Th e reduction was evaluated after oxidation of th e
C-20 h ydroxy group (form ed in th e course of th e reaction ) in to a m ixture
of 3α- an d 3β-am in es 3 an d 6. Table I sh ows th at lith ium alum in ium h y-
dride yielded m ain ly equatorial 3β-am in e 6, boroh ydrides com pletely failed
to reduce th e h ydroxyim in o group, yieldin g a m ixture of h ydroxyim in o al-
coh ols or h ydrolytic products. Even tually, th e h ydroxyim in o derivative was
sm ooth ly reduced with sodium boroh ydride in th e presen ce of som e in or-
Collect. Czech. Chem. Commun. (Vol. 69) (2004)

1508
Matyáš, Kasal, Babot Riera, Sunol:
gan ic salts, such as m olybden um oxide23 or copper ch loride. Neverth eless,
un der th ese con dition s th e equatorial isom er 6 prevailed again .
A truly stereospecific route was based on solvolysis of th e 3β-m eth an e-
sulfon yloxy derivative 7, wh ich was carried out in th e presen ce of sodium
1
azide. H NMR spectrum of product 8 (i.e. m ultiplicity of th e H-3 sign al, see
Experim en tal) proved th e in version of con figuration in th e reaction . Azide
8 was th en reduced with lith ium alum in ium h ydride an d m ixture of th e
am in o alcoh ols (20α an d 20β) form ed was oxidised with ch rom ium trioxide
in aqueous acetic acid in to am in o keton e 3. Th e overall yield of th e
four-step reaction startin g with th e 3β-alcoh ol 9 was aroun d 30%. Com -
poun d 9 was taken from th e laboratory stocks.
COCH3
R
H
5, R = O
NH2OH
4, R = NOH
1. reduction (see Table I)
2. Jones reagent
or
1. H2/PtO2
COCH3
COCH3
H2N
R
3
H
H
6, R = NH2
1. LiAlH4
10, R = NH3+CH3COO-
2. Jones reagent
or
1. Ph3P
COCH3
NaN3
COCH3
R
N3
8
H
H
7, R = MsO
MsCl
9, R = OH
SCHEME 1
Collect. Czech. Chem. Commun. (Vol. 69) (2004)

GABAA Receptor
1509
A straigh tforward reduction of azide 8 with triph en ylph osph in e (i.e. th e
Staudin ger reaction 24) yielded directly am in o keton e 3; th e yield was, h ow-
ever, low (11%).
A prelim in ary in vitro test of biological activity of product 3 was first
don e usin g GABAA receptors an d [3H]m uscim ol. Wh ile allopregn an olon e
in creased th e bin din g of m uscim ol to 123.7%, th e am in o derivative 3 de-
creased th e bin din g to 88.0%. A differen t picture, h owever, was obtain ed
upon usin g [3H]flun itrazepam in cultured cortical n euron s: th e title
com poun d 3 in creased [3H]flun itrazepam bin din g in a con cen tration -
depen den t m an n er (Fig. 1; calculated EC50 an d Em ax values were 5.1 µM an d
119.1%). For com parison , th e en dogen ous n eurosteroid allopregn an olon e 1
acted with a h igh er poten cy (calculated EC50 value 0.6 µM) an d efficacy (cal-
culated Em ax value 134.1%). Con trol specific bin din g in th e absen ce of ste-
roids was 2518 ± 66 dpm /well (n = 7). γ-Am in obutyric acid (100 µM) in -
creased [3H]flun itrazepam bin din g up to 150 ± 6% (n = 3).
Th e am in o derivative 3 did n ot produce cytotoxic effects in cultured cor-
tical n euron s after a 30-m in exposure (th e sam e con dition s as th ose used in
th e bin din g assay), as assessed by optical m icroscopy (Fig. 2) an d by deter-
m in in g th e reduction of th e tetrazolium salt, 3-(4,5-dim eth ylth iazol-2-yl)-
TABLE I
Reduction of oxim e 4 with differen t agen ts
Yield, %
Reagen t
Tem perature, °C
Tim e, h
3 (α)
7 (β)
L-Selectride
–68; th en 25
1; th en 1
0
0
0
0
NaBH4
0; th en 25
1; th en 48
LiAlH4
20
0
13
0
87
0
LiAlH(t-BuO)3
NaBH4 + CeCl3
NaBH4 + MoO3
NaBH4 + CuCl
NaBH4 + AgNO3
NaBH4 + CrCl3
NaBH4 + Zn Cl2
H2, PtO2
25; th en 100
3; th en 24
0
1.2
0.1
24
24
24
24
20
0
0
0–30
25
25
25
25
20
4
51
12
11
0
7
0
0
0
0
45
55
Collect. Czech. Chem. Commun. (Vol. 69) (2004)

1510
Matyáš, Kasal, Babot Riera, Sunol:
2,5-diph en yltetrazolium brom ide (MTT) (Fig. 3). In addition , a 24-h expo-
sure to 10 µM con cen tration of th e am in o derivative 3 (i.e., th e m in im um
con cen tration th at produced a sign ifican t in crease of [3H]flun itrazepam
bin din g) did n ot cause cytotoxicity, even th ough , a h igh er con cen tration
(60 µm ol/l) resulted in clear cytotoxic effects (Figs 2 an d 3). Furth erm ore,
after th e [3H]flun itrazepam bin din g, th e protein con ten t did n ot differ
am on g th e wells exposed to differen t con cen tration s of th e title com poun d.
140
120
100
–7
–6
–5
–4
0
10
10
10
concentration, mol/l
10
FIG. 1
Effect of pregn an e derivatives on [3H]flun itrazepam bin din g to th e GABAA receptor in prim ary
cultures of cortical n euron s. Cultured n euron s were in cubated at 25 °C for 30 m in with
[3H]flun itrazepam an d allopregn an olon e (1; referen ce com poun d, ꢀ), or am in e 3 (᭿). Data are
m ean s ± sem of 6–7 determ in ation s
FIG. 2
Ph otom icrograph s of prim ary cultures of cortical n euron s exposed to pregn an e derivatives at
25 °C for 30 m in (A, B, C) or 24 h at 36 °C (D, E, F). Con trol (A); 10 µM am in e 3 (B an d E); 60 µM
am in e 3 (C an d F); 10 µM allopregn an olon e (D). Th e bar represen ts 5 µm
Collect. Czech. Chem. Commun. (Vol. 69) (2004)

GABAA Receptor
1511
Alth ough 3α-am in o-5α-pregn an -20-on e (3) is n ot as poten t in m odulat-
in g th e [3H]flun itrazepam bin din g on th e GABAA receptor as allo-
pregn an olon e itself, th e above results clearly presen t n ew in sigh t in to possi-
ble fun ction al group variation s25,26 in position 3α of th e steroid skeleton .
Th e discrepan cy between th e [3H]m uscim ol an d [3H]flun itrazepam tests
could be un derstood in th e ligh t of th e com plex structure of th e recep-
tor27,28 an d th e idea of th e existen ce of m ultiple specific steroid bin din g
sites with in th e GABAA (see29–32). Addition al tests h ave to be perform ed be-
fore an y con clusion on th e applicability of am in e 3 can be m ade. Neverth e-
less, th e axiom of th e 3α-h ydroxy group as th e essen tial part of n euroactive
steroids h as n ow been put in to question .
EXPERIMENTAL
Ch em istry
Meltin g poin ts were determ in ed on a Boetius m eltin g poin t m icroapparatus (Germ an y) an d
are un corrected. An alytical sam ples were dried over ph osph orus pen toxide at 50 °C/100 Pa.
Optical rotation s were m easured in ch loroform ([α]D values are given in 10–1 deg cm 2/g).
IR spectra of ch loroform solution s were recorded on
a Bruker IFS 88 spectrom eter,
waven um bers are given in cm –1. NMR spectra were m easured on an FT NMR spectrom eter
Varian Un ity-200 (at 200 MHz) in CDCl3 with tetram eth ylsilan e as in tern al referen ce.
Ch em ical sh ifts are given in ppm (δ-scale), couplin g con stan ts (J) an d width s of m ultiplets
(W) in Hz. Un less oth erwise stated, th e data were in terpreted as first-order spectra. Th in -
100
75
50
*
25
0
control
10 µM
60 µM
FIG. 3
Cell viability as determ in ed by th e MTT assay. Cultured n euron s were in cubated at 25 °C for
24 h (᭿) with 10 µM, or at 36 °C for 30 m in (ᮀ) an d 24 h (᭿) with 60 µM am in e 3. Results are
m ean s ± sem of 3–4 determ in ation s. * p < 0.001
Collect. Czech. Chem. Commun. (Vol. 69) (2004)

1512
Matyáš, Kasal, Babot Riera, Sunol:
layer ch rom atograph y (TLC) was perform ed on silica gel (ICN Bioch em icals). Preparative
TLC (PLC) was carried out on 200 × 200 m m plates coated with a 0.7 m m th ick layer of th e
sam e m aterial. For colum n ch rom atograph y, 60–120 µm silica gel was used. Wh en ever aque-
ous solution s of h ydroch loric acid, potassium h ydrogen carbon ate an d potassium carbon ate
were used, th eir con cen tration was always 5%. Before evaporation on a rotary evaporator in
vacuum (bath tem perature 50 °C), solution s in organ ic solven ts were dried over an h ydrous
sodium sulfate.
Biology
GABAA receptor binding. Mem bran es12 were prein cubated at 37 °C for 10 m in in a Tris-HCl
buffer (50 m m o/l, pH 7.4) with or with out a sam ple tested. Th e m ixture was in cubated with
[3H]m uscim ol (5 n m ol/l) an d, after 10 m in , th e reaction was term in ated by rapid vacuum
filtration th rough a Wh atm an GF/B glass-fiber filter. Non specific bin din g was determ in ed in
th e presen ce of 1 m M GABA. Prelim in ary experim en ts with 5 n M [3H]m uscim ol were carried
out with 100 n M con cen tration of th e steroids tested.
Cell culture preparation. Pregn an t (gestation al day 17) NMRI rats were obtain ed from Iffa
Credo (St.-Germ ain -sur-l’Arbreste, Fran ce). Th e an im als were h an dled in com plian ce with
th e protocol (No. 95/99 of th e Un iversity of Barcelon a), approved by th e Gen eralitat de
Catalun ya, Spain , followin g EU guidelin es. Plastic 24- an d 48-well plates were purch ased
from CoStar (Corn in g Scien ce Products, Acton (MA), U.S.A.) an d NUNC Ltd. (Roskilde,
Den m ark). Foetal Calf Serum was obtain ed from Gibco (Glasgow, U.K.) an d Dulbecco’s
Min im um Essen tial Medium (DMEM) from Bioch rom (Berlin , Germ an y). Bicucullin e an d
GABA were purch ased from SIGMA Ch em ical Co. (St. Louis (MO), U.S.A.).
Radioactivity was m easured in liquid scin tillation cocktail Optiph ase “Hisafe” 2, from
Wallac Oy (Turku, Fin lan d) on a liquid scin tillation coun ter 1414 Win spectral from Wallac.
Prim ary cultures of cortical n euron s were obtain ed from th e cerebral cortices of
17-day-old rat em bryos, followin g th e m eth od described by Hertz et al.33 In brief, cells were
dissociated by m ild trypsin isation at 37 °C, followed by trituration in a DNAse solution
(0.004% w/v) con tain in g a soybean trypsin in h ibitor (0.05% w/v). Th e cells were th en sus-
pen ded in DMEM (25 m M KCl, 31 m M glucose an d 0.2 m M glutam in e) supplem en ted with
4-am in oben zoic acid, in sulin , pen icillin , an d 10% of foetal calf serum . Th e cell suspen sion
was seeded in 24- or 48-well plates pre-coated with poly-L-lysin e an d in cubated for 7–12
days in a h um idified 5% CO2/95% air atm osph ere at 36.8 °C. A m ixture of 5 µM 5-fluoro-
2′-deoxyuridin e an d 20 µM uridin e was added after 40–48 h in th e culture to preven t glial
proliferation .
Data analysis. Data are m ean s ± sem . Con cen tration -respon se data were fit by n on -
regression an alysis to sigm oid curves by usin g th e Graph Pad Prism Program (Graph Pad Soft-
ware In c., San Diego (CA), U.S.A.). Statistical an alysis was perform ed usin g th e Graph Pad
Prism Program . On e way an alysis of varian ce (ANOVA) followed by Dun n ett’s test was used.
P < 0.05 was con sidered to be sign ifican t.
3-Hydroxyim in o-5α-pregn an -20-on e (4)
A solution of dion e 5 (289.5 m g, 0.91 m m ol) in pyridin e (3.0 m l) was treated with
h ydroxylam in e h ydroch loride (76 m g, 3.08 m m ol) at 0 °C. Hydroxylam in e h ydroch loride
was added stepwise over a period of 1 h . Th e reaction was followed by TLC (am m on iacal
ch loroform ) an d quen ch ed after 3 h with water. Th e organ ics were taken up in to ch loro-
Collect. Czech. Chem. Commun. (Vol. 69) (2004)

GABAA Receptor
1513
form , th e solution was wash ed an d dried. Th e reaction yielded 280 m g (92%) of 4. M.p.
235–239 °C (m eth an ol–eth er) (lit.34 240–241 °C). [α]D +94.9 (c 0.2) (lit.34 +107 (c 0.5)). IR:
1698 (CO), 1654 (CN), 3591 (OH). 1H NMR: 0.60 s, 3 H (H-18); 0.90 s, 3 H (H-19); 2.15 s,
3 H (H-21); 2.53 t, H (H-17). For C21H33NO2 (331.5) calculated: 76.09% C, 10.03% H,
4.23% N; foun d: 76.54% C, 10.32% H, 4.11% N.
Platin um Catalysed Hydrogen ation of Oxim e 4
A solution of oxim e 4 (330 m g, 0.99 m m ol) in acetic acid (10 m l) was h ydrogen ated over
platin um oxide (100 m g) at room tem perature. After 20 h , th e catalyst was rem oved by fil-
tration an d wash ed with acetic acid. Th e filtrate was con cen trated to dryn ess in vacuum .
Th e rem ain der was dissolved in acetic acid (90%, 25 m l) an d treated at with a solution of
ch rom ium trioxide (270 m g, 2.7 m m ol) in aqueous sulfuric acid (20%, 0.4 m l) at room tem -
perature. Th e course of th e reaction was m on itored by TLC (am m on iacal ch loroform ). Th e
excess of th e oxidan t was decom posed with potassium h ydrogen sulfite (2.0 g, 16.6 m m ol).
Most of th e solven t was rem oved from th e reaction m ixture usin g a rotary evaporator, th en
am m on ia was added un der coolin g an d th e precipitate extracted in to eth yl acetate. Th e ex-
tract was wash ed with water, dried an d con cen trated in vacuum . Th e solution was applied
on 11 PLC plates wh ich were developed with am m on iacal ch loroform con tain in g 3% of
m eth an ol. Each fraction was evaporated upon th e addition of acetic acid (1 m l) an d th en
co-evaporated with an addition al dose of toluen e (15 m l).
Th e less polar fraction con tain ed 3α-am in o-5α-pregn an -20-on e (3; 166 m g, 45%). M.p.
119–122 °C (eth er–petroleum eth er) (lit.19 123 °C). [α]D +94.4 (c 0.8) (lit.35 +95 (c 1.7)).
1H NMR: 0.60 s, 3 H (H-18); 0.78 s, 3 H (H-19); 2.11 s, 3 H (H-21); 3.20 s, H (H-3).
Th e m ore polar fraction con tain ed 20-oxo-5α-pregn an -3β-yl am m on ium acetate (10;
158 m g, 42%). M.p. 157–159 °C (eth er). [α]D +11.9 (c 0.8). 1H NMR: 0.59 s, 3 H (H-18);
0.81 s, 3 H (H-19); 1.95 s, 3 H (H-21); 2.11 s, 3 H (CH3COO); 2.52 t, 1 H, J = 8.6 (H-17);
2.88 m , 1 H, W = 43.0 (H-3). For C23H39NO3 (377.6) calculated: 73.17% C, 10.41% H,
3.71% N; foun d: 73.20% C, 10.70% H, 3.77% N.
Lith ium Alum in ium Hydride Reduction of Oxim e 4
A solution of oxim e 4 (50 m g, 0.15 m m ol) in tetrah ydrofuran (5.0 m l) was treated with lith -
ium alum in ium h ydride (approxim ately 100 m g) at room tem perature for 2 h . Th e excess of
th e reagen t was destroyed with eth yl acetate (10 m l) an d th e reaction m ixture was filtered
th rough a layer of sodium sulfate. Th e product was oxidised as above (3.0 m l of acetic acid,
60 m g of ch rom ium trioxide in 10 drops of water). Usual work-up yielded a m ixture of 3
an d 6, th e ratio of 3:6 bein g 1:3, as determ in ed by 1H NMR.
Sodium Boroh ydride Reduction of Oxim e 4 in th e Presen ce of MoO3
To a solution of oxim e 4 (12.2 m g, 0.04 m m ol) in m eth an ol (1.0 m l), m olybden um (VI) ox-
ide (50 m g, 0.35 m m ol) was added an d th e solution was carefully h eated to dissolve oxim e
4. Sodium boroh ydride (200 m g, 5.28 m m ol) was slowly in troduced in to th e reaction m ix-
ture at room tem perature. After 10 m in , th e excess of sodium boroh ydride was decom posed
with water, an d m eth an ol was distilled off on a vacuum evaporator. Th e organ ics were taken
up in to am m on iacal ch loroform an d th e solution was dried. Th e product was oxidised as
Collect. Czech. Chem. Commun. (Vol. 69) (2004)

1514
Matyáš, Kasal, Babot Riera, Sunol:
above to yield 9.1 m g (78%) of a m ixture of 3 an d 6, th e ratio of 3:6 bein g 1:12.8, as deter-
m in ed by 1H NMR.
3β-O-(Meth an esulfon yl)-5α-pregn an -20-on e (7)
To a solution of h ydroxy derivative 9 (204.4 m g, 0.63 m m ol) in pyridin e (5.0 m l), cooled
down to 0 °C, m eth an esulfon yl ch loride (0.15 m l, 1.93 m m ol) was slowly added. Th e reac-
tion was followed by TLC (25% eth er in ben zen e) an d quen ch ed with ice after on e day. Th e
organ ics were taken up in to ch loroform an d wash ed with 5% h ydroch loric acid, saturated
sodium h ydrogen carbon ate, water an d dried. Derivative 7 was crystallised from ch loroform –
eth er an d th en from ch loroform –petroleum eth er to give 63.9 m g (25%) of 7. Th e overall
yield after ch rom atograph y was 193.7 m g (76%). M.p. 139–141 °C (ch loroform –petroleum
eth er). [α]D +116.5 (c 0.4). 1H NMR: 0.6 s, 3 H (H-18); 0.83 s, 3 H (H-19); 2.11 s, 3 H (H-21);
3.00 s, 3 H (CH3S); 4.63 m , 1 H, W = 40.89 (H-3). For C22H36O4S (396.6) calculated:
66.63% C, 9.15% H, 8.08% S; foun d: 66.45% C, 9.46% H, 7.81% S.
3α-Azido-5α-pregn an -20-on e (8)
A solution of com poun d 7 (402.1 m g, 1.01 m m ol) in dim eth ylform am ide (4.0 m l) was
treated with sodium azide (200 m g, 3.08 m m ol) at room tem perature, an d th e m ixture was
th en h eated to 80 °C for 5 h . Th e organ ics were taken up in to eth yl acetate an d th e solution
was dried. Crystallisation from eth yl acetate yielded 164.6 m g (47%) of 8. M.p. 162–164 °C
(eth yl acetate). [α]D +67.3 (c 0.8). IR: 1710 (CO), 2110 (N3). 1H NMR: 0.60 s, 3 H (H-18);
0.79 s, 3 H (H-19); 2.12 s, 3 H (H-21); 2.53 t, H (H-17); 3.89 m , 1 H, W = 24.41 (H-3). For
C
21H33N3O (343.5) calculated: 73.43% C, 9.68% H, 12.23% N; foun d: 72.54% C, 9.75% H,
11.59% N.
Triph en ylph osph in e Reduction of Azide 8
A solution of azide 8 (75 m g, 0.22 m m ol) in tetrah ydrofuran (2.0 m l) was treated with
triph en ylph osph in e (234 m g, 0.88 m m ol) at room tem perature. Th e reaction was quen ch ed
with water after two days. Tetrah ydrofuran was evaporated on a vacuum evaporator. Th e
organ ics were taken up in to am m on iacal ch loroform . Th e reaction m ixture was separated us-
in g preparative TLC (20% m eth an ol in ch loroform ) to afford 21.6 m g (30%) of com poun d 3.
[3H]Flun itrazepam Bin din g Assay
A sam ple of am in o derivative 3 was dissolved in DMSO. A con cen tration scale of 0.1, 1, 3,
10, 30, an d 60 µm ol/l of derivative 3 in a HEPES (0.5% DMSO) was prepared. As an in tern al
stan dard, GABA (100 µm ol/l) in HEPES was used. Diazepam (20 µm ol/l) was used to deter-
m in e th e n on specific bin din g. A 6-day-old culture of cortical n euron s cultivated on
a
48-well plate (for preparation , see above) was wash ed 3 tim es with approxim ately 0.5 m l of
HEPES buffer prewarm ed to 37 °C. After wash in g, th e cells were m ain tain ed in 0.5 m l of
HEPES buffer un til th e assay solution was added. Th e HEPES buffer was replaced with 150 µl
of th e assay solution s (steroid, 100 µM GABA or 20 µM diazepam ). 0.5% DMSO was presen t
in HEPES buffer con trol cells. [3H]Flun itrazepam (50 µl) was added in to each well. Th e fin al
con cen tration of [3H]flun itrazepam in a well, after fin e-tun in g th e m eth od, was 5 n m ol/l.
Th e plate was th en in cubated at 25 °C for 30 m in an d th e wells were th en wash ed 4 tim es
Collect. Czech. Chem. Commun. (Vol. 69) (2004)

GABAA Receptor
1515
with 0.5 m l of HEPES buffer. NaOH (0.2 m ol/l, 150 µl) was added in to each well an d th e
plates were stirred for 4 h . Th e wh ole con ten t of each well was tran sferred in to a vial an d
scin tillation liquid (4 m l) was added. Th e vials were agitated, an d radioactivity m easured.
Cell Viability Assay
Cultured cells were exposed to pregn an e derivatives in HEPES buffer at 25 °C for 24 h (th e
sam e con dition s as in th e [3H]flun itrazepam bin din g assay) or at 36 °C for 30 m in an d 24 h
in th e culture m edium . DMSO (0.5%) was presen t in th e in cubation m edium of con trol
cells. At th e en d of th e exposure, th e cultures were rin sed with HEPES buffer an d in cubated
at 37 °C for 15 m in with HEPES buffered salin e con tain in g 250 µg/m l of oxidised
3-(4,5-dim eth ylth iazol-2-yl)-2,5-diph en yltetrazolium brom ide (MTT). MTT is in corporated
in to viable cells an d reduced by m itoch on dria to a coloured tetrazolium com poun d. Follow-
in g a careful rem oval of excess MTT, th e cells were disin tegrated with 5% sodium dodecyl-
sulfate. Th e coloured solution was m easured at 560 n m . Th e results are expressed as th e per-
cen tage of th e con trol values after blan k subtraction . Th e MTT assay m easures m itoch on -
drial activity an d is curren tly used as th e cell viability assay36
.
The authors are indebted to Dr Z. Krištofíková for the measurement of the binding to GABAA recep-
tors ([3H]muscimol). This work was supported by grants No. S 5011007 and No. Z4 055 905 and by
grants No. 01/1318 from FIS (Ministry of Health, Spain) and No. 2001SGR00356 from CIRIT
(Generalitat de Catalunya, Spain). Z. Babot and L. Matyáš are recipients of fellowships from the Min-
istry of Science and Technology and the Foreign Office Ministry (Spain), respectively.
REFERENCES
1. Franěk F., Eckschlager T., Kohout L.: Collect. Czech. Chem. Commun. 2003, 68, 2190.
2. Chodounská H., Kasal A.: Collect. Czech. Chem. Commun. 1998, 63, 1543.
3. Slavíková B., Kasal A.: Collect. Czech. Chem. Commun. 1999, 64, 1479.
4. Slavíková B., Kasal A., Chodounská H., Krištofíková Z.: Collect. Czech. Chem. Commun.
2002, 67, 30.
5. Kasal A.: Collect. Czech. Chem. Commun. 1999, 64, 1471.
6. Kasal A.: Tetrahedron 2002, 56, 3559.
7. Chodounská H., Kasal A.: Collect. Czech. Chem. Commun. 2002, 65, 1156.
8. Slavíková B., Kasal A., Uhlířová L., Kršiak M., Chodounská H., Kohout L.: Steroids 2001,
66, 99.
9. Harrison N. L., Simmonds M. A.: Brain Res. 1984, 323, 287.
10. Cottrell G. A., Lambert J. J., Peters J. A.: Br. J. Pharmacol. 1987, 90, 491.
11. Gee K. W., Bolger M. B., Brinton R. E., Coirini H., McEwen B. S.: J. Pharmacol. Exp. Ther.
1988, 246, 803.
12. Harrison N. L., Majewska M. D., Harrington J. L.: J. Pharmacol. Exp. Ther. 1987, 241, 346.
13. Peters J. A., Kirkness E. F., Callachan H., Lambert J. J., Turner A. J.: Br. J. Pharmacol.
1988, 94, 1257.
14. Zhong Y., Simmonds M. A.: Neuropharmacology 1996, 35, 1193.
15. Vale C., Pomes A., Rodriguez-Farre E., Sunol C.: Eur. J. Pharmacol. 1997, 29, 343.
16. Vale C., Damgaard I., Sunol C., Rodriguez-Farre E.: J. Neurosci. Res. 1998, 52, 276.
Collect. Czech. Chem. Commun. (Vol. 69) (2004)

1516
Matyáš, Kasal, Babot Riera, Sunol:
17. Vale C., Vilaro M. T., Rodriguez-Farre E., Sunol C.: J. Neurosci. Res. 1999, 57, 95.
18. Nelle S., Charles G., Cave A., Goutarel R.: C. R. Acad. Sci., Ser. II 1970, 271, 153.
19. Dadoun H., Conreur Ch., Cave A.: Phytochemistry 1973, 12, 923.
20. Janot M. M., Khuong-Huu Q., Lusenchi X., Goutarel R.: Bull. Soc. Chim. Fr. 1960, 1,
1669.
21. Dave V., Stothers J. B., Warnhoff E. W.: Can. J. Chem. 1980, 58, 2666.
22. Knox L. H.: J. Org. Chem. 1964, 29, 2187.
23. Szendi Z., Dombi G., Vincze I.: Monatsh. Chem. 1996, 127, 1189.
24. Dieter A., Martin F.: Tetrahedron Lett. 1994, 35, 565.
25. Hawkinson J. E., Kimbrough C. L., Belelli D., Lambert J. J., Purdy R. H., Lan N. C.: Mol.
Pharmacol. 1994, 46, 977.
26. Purdy R. H., Morrow A. L., Blinn J. R., Paul S. M.: J. Med. Chem. 1990, 33, 1572.
27. Mohler H.: Pharmacology of GABA and Glycine Transmission. Handbook of Experimental
Pharmacology, Vol. 150, p. 79. Springer, Berlin 2001.
28. Pirker S., Schwartzer C., Wieselthaler A., Sieghart W., Sperk G.: Neuroscience 2000, 101,
815.
29. Barnard E. A., Skolnick P., Olsen R. W., Mohler H., Sieghart W., Biggio G., Braestrup C.,
Bateson A. N., Langer S. Z.: Pharmacol. Rev. 1998, 50, 291.
30. Wisden W., Laurie D. J., Monyer H., Seeburg P. H.: J. Neurosci. 1992, 12, 1581.
31. Park-Chung M., Malayev A., Purdy R. H., Gibbs T. T., Farb D. H.: Brain Res. 1999, 830,
72.
32. Han M., Zorumski C. F., Covey D. F.: J. Med. Chem. 1996, 39, 4218.
33. Hertz E., Yu A., Hertz L., Juurlink B., Schousboe A. in: Preparation of Primary Culture of
Mouse Cortical Neurons (A. Shahar, J. de Vellis, A. Vernadakis and B. Haber, Eds), p. 183.
Alan R. Liss, Inc., New York 1989.
34. Nace H. R., Watterson A. C.: J. Org. Chem. 1966, 31, 2109.
35. Janot M. M.: C. R. Acad. Sci., Ser. II 1960, 251, 559.
36. Lobner D.: J. Neurosci. Methods 2000, 96, 147.
Collect. Czech. Chem. Commun. (Vol. 69) (2004)