S.D. Taylor, J. Harris / Steroids 76 (2011) 1098–1102
1099
Fig. 1. Structures of 17-amino-1,3,5(10)-estratrien-3-ol (1), 17-amino-5␣-androstan-3-ol (2) and, 17-amino-3-hydroxyandrost-5-ene (3).
Milwaukee, Wisconsin, USA) were distilled before use. Sodium
triacetoxyborohydride (Aldrich, Milwaukee, Wisconsin, USA) was
used as is. Flash chromatography was performed using silica gel
3.88 (AB system, overlapping dd, J = 14.0 Hz, 2H), 4.49 (bs, 2H), 6.49
(s, 1H), 6.57 (d, J = 8.3 Hz, 1H), 7.0 (d, J = 8.4 Hz, 1H), 7.20–7.40 (m,
5H); 13C NMR 154.0, 140.1, 138.0, 132.0, 128.5, 128.3, 127.1, 126.3,
115.6, 113.0, 68.1, 52.5, 52.3, 44.0, 43.2, 38.8, 38.1, 29.7, 29.4, 27.4,
26.5, 23.5, 12.0; LREIMS (70 eV, m/z): 361 (M+, 100%), 270 (15%),
146 (94%), 91 (48%), HREIMS calcd. for C25H31NO (M+) 361.2406,
found 361.2408.
˚
60 A (234–400 mesh) (Silicycle, Quebec, Canada). TLC was per-
formed using silica coated TLC plates (Silicycle, Quebec, Canada).
All 1H and 13C NMR spectra were recorded on a Bruker Avance
300 spectrometer. Chemical shifts (ı) for 1H NMR spectra recorded
in CDCl3 are reported in parts per million (ppm) relative to Me4Si
(0.0 ppm) while those recorded in DMSO-d6 or DMSO-d6-benzene-
d6 (1:1) are reported in parts per million (ppm) relative to the
solvent residual peak (2.49 ppm) and are reported as follows:
chemical shift (ppm), multiplicity (s, singlet; d, doublet; t, triplet;
q, quartet; m, multiplet; br, broadened), coupling constant in Hz,
integration. Chemical shifts (ı) for 13C spectra are reported in ppm
relative to CDCl3 (ı 77.0, central peak) or DMSO-d6 (ı 39.5, central
peak).
2.4. 17ˇ-Benzylamino-5˛-androstan-3ˇ-ol (5)
Prepared using the general procedure described above. White
amorphous solid; Rf = 0.30 (5% MeOH/0.5% aq. NH4OH/94.5%
CH2Cl2); 1H NMR (CDCl3) 0.65–2.05 (m, 24H), 0.71 (s, 3H), 0.79 (s,
3H), 2.54 (t, J = 8.2 Hz, 1H), 3.55 (m, 1H), 3.78 (AB system, overlap-
ping dd, J = 14.0 Hz, 2H), 7.12–7.35 (m, 5H); 13C NMR (CDCl3) 141.0,
128.3, 128.0, 126.8, 70.9, 68.2, 54.6, 53.3, 52.7, 45.0, 42.9, 38.2, 37.1,
35.6, 35.5, 31.9, 31.5, 29.6, 28.7, 23.8, 21.1, 12.4, 12.1; LREIMS (70 eV,
m/z): 381 (M+, 47%), 290 (12%), 146 (100%), 91 (20%); HREIMS calcd.
for C26H39NO (M+), 381.3032, found 381.3039.
2.2. General procedure for the reductive amination of E1, EA and
DHEA
2.5. 17ˇ-Benzylamino-3ˇ-hydroxyandrost-5-ene (6)
The steroid was dissolved in dry THF (approximately
0.1 mmol steroid/mL THF) and then an equivalent volume of
1,2-dichloroethane was added. To this was added 4 eq. of benzy-
lamine or allylamine, 4 eq. glacial acetic acid and 2.5 eq. STAB-H.
The mixture was stirred for 24–48 h. An equivalent volume of aq.
saturated Na2CO3 was added and the mixture stirred for 10 min.
Two volumes of EtOAc were added and the organic layer was
obtained. The organic layer was washed with aq. saturated Na2CO3
(3×), water (10×) and then brine (1×). The organic layer was
dried (Na2SO4), filtered and concentrated leaving a white solid.
In the case of allylamine 7, the crude product is >98% pure by
1H and 13C and no further purification was necessary for further
manipulations. However, if highly pure 7 is desired, it can be
subjected to silica gel flash chromatography using (5% MeOH/0.5%
aq. NH4OH/94.5% CH2Cl2 as eluent) (Rf = 0.25). In the case of
identified by 1H NMR as N-benzylacetamide. Careful silica gel
chromatography (run by gravity, 40% EtOAc–60% benzene then
65% EtOAc–35% benzene) is required to remove this impurity
(see Schemes 1 and 2 for the yields of 4–7). N-benzylacetamide is
slightly soluble in water and can also be removed by dissolving
the crude material in EtOAc and washing extensively with water.
Alternatively, the N-benzylacetamide impurity can be carried
through the hydrogenolysis reaction as we found that it did not
affect the hydrogenolysis and could be very easily removed by
chromatography after the hydrogenolysis reaction.
Prepared using the general procedure described above. White
amorphous solid; Rf = 0.30 (5% MeOH/0.5% aq. NH4OH/94.5%
CH2Cl2); 1H NMR (CDCl3) 0.74 (s, 3H), 1.00 (s, 3H), 0.65–1.63 (m,
13H), m (1.74–2.06 (m, 5H), 2.56 (t, J = 8.5 Hz, 1H), 3.41–3.52 (m,
1H), 3.79 (AB dd, J = 13.8 Hz, 2H)), 5.32 (d, J = 3.2 Hz, 1H), 7.15–7.38
(m, 5 H); 13C NMR (CDCl3) 141.1, 140.8, 128.3, 128.0, 126.7, 121.5,
71.7, 68.1, 53.6, 52.7, 50.4, 42.6, 42.3, 38.0, 37.3, 36.6, 31.9, 31.7,
29.6, 23.9, 20.8, 19.4, 11.8; LR + ESIMS: 380.3 (M+H); HR + ESIMS
calcd. for C26H38NO (M+H), 380.2961, found 380.2953.
2.6. 17ˇ-Allylamino-3ˇ-hydroxyandrost-5-ene (7)
Prepared using the general procedure described above. White
amorphous solid; Rf = 0.25 (5% MeOH/0.5% aq. NH4OH/94.5%
CH2Cl2); 1H NMR (CDCl3) 0.67 (s, 3H), 0.96 (s, 3H), 0.7–1.60 (m,
12H), 1.71–2.01 (m, 6 H), 2.51 (t, J = 8.0 Hz, 1H), 3.21 (d, J = 5.6 Hz,
1H), 3.43 (m, 1H), 5.01 (d, J = 10.2 Hz, 1H), 5.10 (d, J = 17.1 Hz, 1H),
5.76–5.92 (m, 1H); 13C NMR (CDCl3) 140.9, 137.3, 121.3, 115.6,
71.4, 68.2, 53.5, 51.4, 50.3, 42.5, 42.3, 37.9, 37.3, 36.5, 31.8, 31.6,
29.5, 23.8, 20.8, 19.4, 11.7; LREIMS (70 eV, m/z): 329 (M+, 15%), 314
(100%), 96.1 (30%); HREIMS calcd. for C22H35NO (M+), 329.2719,
found 329.2716.
2.7. General procedure for the hydrogenolysis of 4 and 5
Compound 4 or 5 was added to absolute ethanol (approximately
15 mg 4 or 5 per mL of solvent) and heated gently with a heat gun
until the steroid completely dissolved. The mixture was allowed
to cool and then 10% Pd/C (13 wt%) was added and the flask fitted
with a balloon filled with H2 and the mixture was stirred for 24 h.
The mixture was filtered through Celite and the filtrate concen-
trated to give a white solid. If any N-benzylacetamide was present
2.3. 17ˇ-Benzylamino-1,3,5(10)-estratrien-3-ol (4)
Prepared using the general procedure described above. White
amorphous solid; Rf = 0.30 (5% MeOH/0.5% aq. NH4OH/94.5%
CH2Cl2); 1H NMR (CDCl3) 0.81 (s, 3H), 1.17–1.58 (m, 7H), 1.65–1.76
(m, 1H), 1.79–1.90 (m, 1H), 2.01–2.30 (m, 4 H), 2.65–2.85 (m, 3H),