crystallisation and from deacetylcolchicine 7 by partition
between dichloromethane and 2% aqueous tartaric acid.
245, 252; δH 2.03 (3H, s, COCH3), 2.23 (3H, s, NMe), 1.7–3.5
(6H, m, 3 × CH2), 3.6 (1H, m, H-7), 3.62 (3H, s, 1-OMe), 3.89,
3.93 and 3.94 (each 3H, s, 3 × OMe), 6.53 (1H, s, H-4), 6.8–7.3
(3H, m, ArH), 6.75 and 7.26 (each 1H, d, J 11, H-11 and H-12),
7.6 (1H, m, ArH), 8.23 (1H, s, H-8); m/z 519 (Mϩ, 28%), 371
(14), 370 (26), 342 (33), 207 (21), 107 (100).
A third fraction afforded speciosine 14 (4.8 mg) m/z 477
(Mϩ, 14%), 371 (62), 356 (24), 207 (100), 107 (14), 106 (26)
identified by full comparison with authentic material.11 It
is common experience to find some phenolic materials of
substantial molecular weight in the non-phenolic fraction
because of very favourable partition towards the organic phase.
Investigation of the metabolism of [3-14C]autumnaline in C.
autumnale
(1RS)-[3-14C]Autumnaline was fed as above to C. autumnale
in spring and separate batches of plants were harvested after 1
day (A) and after 3 days (B). The alkaloids were isolated and
separated by solvent partition as described above. The neutral
and water soluble fractions were only weakly radioactive but
the basic fractions were strongly radioactive and were analysed
by TLC on silica using benzene–ethyl acetate–diethylamine–
methanol, 5:4:1:0.8. The radioactive metabolites were located
by a radioactivity scanner (Desaga 12-20) as highly active spots
of Rf 0.17, 0.21 and 0.28 and a less active spot of higher
Rf which corresponded to O-methylandrocymbine. The basic
fraction from batch B (8 mg) was diluted with radio-inactive
O-methylandrocymbine (50 mg) and subjected to PLC as
above. The purified O-methylandrocymbine was converted into
the picrate using picric acid (35 mg) in hot methanol, and the
salt was recrystallised to constant specific activity (1.2%
incorporation), mp 155–157 ЊC. For radio-assay, the free base
was regenerated quantitatively from the picrate by passing a
chloroform solution of the salt down a short column of basic
alumina.
Hydrolysis of O-acetylspeciosine
Aqueous 10% sodium hydroxide (0.1 cm3) was added to O-
acetylspeciosine (15 mg) in methanol (1 cm3); the solution was
kept at 20 ЊC for 16 h and then largely evaporated. Water
(1 cm3) was added, followed by solid carbon dioxide, and the
mixture was extracted with ethyl acetate to give speciosine
(13.7 mg), mp 209–212 ЊC (lit.,11 mp 211–214 ЊC), identified as
above.
Hydrolysis of N-formyldemecolcine
N-Formyldemecolcine (30 mg) was heated for 20 h in con-
centrated sulfuric acid (0.2 cm3) and water (1 cm3) at 100 ЊC,
then cooled and diluted with water. The distillate (400 cm3)
from steam distillation was basified to phenolphthalein with
0.1 M aqueous lithium hydroxide and then evaporated. The
residual solid, thoroughly dried under vacuum, was dissolved
in dry DMF (0.5 cm3) and heated at 90 ЊC for 2.5 h with
p-bromophenacyl bromide (100 mg). Evaporation of the
dimethylformamide gave a solid; the part soluble in diethyl
ether was purified by PLC on silica using benzene and the ester
was eluted from the silica with anhydrous diethyl ether
(400 cm3) to give a solid (16 mg), mp 97–98 ЊC (lit.,18 mp 99 ЊC)
from light petroleum (bp 60–80 ЊC). It was identified by full
comparison with the authentic p-bromophenacyl formate. This
sensitive ester decomposed during attempts to elute it from the
silica using chloroform–methanol.
The aqueous solution after steam distillation was adjusted
to pH 8 using saturated aqueous sodium hydrogen carbonate
and extraction with chloroform yielded demecolceine (typically
40%), mp 133–135 ЊC from MeOH (lit.,19 mp 133–135 ЊC)
further identified by comparison with an authentic sample.
The radioactive N-formyldemecolcine from Experiment 4,
Table 2, was diluted to the specific radioactivity of 14C
1.84 × 105 and 3H 6.82 × 105 (disintegrations per 100 s per
mmol). Degradation as above yielded the p-bromophenacyl
formate with specific activity (disintegrations per 100 s per
mmol) of 14C 9.48 × 104 (51%) and 3H 4.1 × 103 (0.6%). The fall
in specific activity is due to dilution effects (see text).
The basic fraction from batch A (20 mg) was partitioned
between ethyl acetate and 0.1 M aqueous sodium hydroxide to
give a non-phenolic fraction Rf 0.28 (16 mg) and a phenolic
fraction (Rf 0.17, 0.21) (3 mg). The non-phenolic fraction was
treated with acetic anhydride (0.2 cm3) in pyridine (1 cm3) for
16 h at 20 ЊC and a radioactive product of appreciably higher
Rf value was obtained but no further work has been done on
this product.
Isolation of minor alkaloids from C. autumnale
250 Plants were harvested in spring when the seed capsules were
beginning to develop. The plants were extracted in the usual
way3 to give a total mass of alkaloid of 5 g. Part of this (2.86 g)
was partitioned between diethyl ether (150 cm3) and 0.1 M
tartaric acid (150 cm3), and the ethereal layer was further
washed with 0.1 M tartaric acid (2 × 50 cm3) and water
(2 × 25 cm3) before evaporation to give a neutral residue which
was not investigated further. The aqueous phase was adjusted
to pH 8 with saturated aqueous sodium hydrogen carbonate
and extracted with diethyl ether (6 × 100 cm3) to yield the basic
fraction (380 mg). Finally the aqueous phase was extracted with
chloroform (6 × 100 cm3) to afford water-soluble materials
(2.23 g).
The basic fraction (380 mg) in chloroform (50 cm3) and
diethyl ether (200 cm3), was extracted with 2 M aqueous sodium
hydroxide (4 × 30 cm3). Each alkaline extract was immediately
acidified, then adjusted to pH 8 as above and extracted with
chloroform (4 × 50 ml) to afford the phenolic bases (140 mg).
The organic phase after removal of phenols was washed with
water (2 × 20 cm3) and evaporated to give the non-phenolic
bases (220 mg) which were partitioned between 0.5 M phos-
phate buffer (pH 4.5) (4 cm3) and ethyl acetate (0.5 cm3). Celite
(4 g) was added, followed by light petroleum (bp 60–80 ЊC)
(1 cm3) and the mixture was shaken until homogeneous, then
packed onto a column of Celite (60 g) which had previously
been shaken with the lower layer from a mixture of 0.5 M
phosphate buffer (60 cm3), ethyl acetate (1 dm3) and light
petroleum (bp 60–80 ЊC) (1 dm3). Elution of the column with
the upper organic phase afforded O-methylandrocymbine 2
(10 mg); m/z 385 (Mϩ); [α]D24 Ϫ286 (c 0.079 in CHCl3) {lit.,17
[α]D24 Ϫ295 (c 0.127 in CHCl3)}, further characterised as the
picrate,17 mp 153–154 ЊC. Its identity was established by full
comparison with authentic material.17
Acknowledgements
We are indebted to the following colleagues for valuable gifts of
ˇ
alkaloids: Professor F. Santavý, Palacky University (rare minor
Colchicum alkaloids), Dr A. L. Morrison, Roche Products Ltd.
(demecolcine) and Drs E. Grew and F. R. Smith, Glaxo Group
(colchicine). We also thank the SRC and SERC for Research
Studentships (to A. C. B., D. R. J. and R. N. W.) together with
Mr H. Soame (Breconshire), Mr J. K. Hulme (Ness Botanic
Garden) and Mr J. Symonds (University Botanic Gardens,
Cambridge) for all Colchicum plants. Our thanks are recorded
to Roche Products Ltd. and to the EPSRC for financial
support.
References
1 Part 27. E. McDonald, R. Ramage, R. N. Woodhouse, E. W.
Underhill, L. R. Wetter and A. R. Battersby, J. Chem. Soc., Perkin
Trans. 1, 1998, 2979.
A second basic product was shown to be O-acetylspeciosine
15 (88 mg) by the following data: νmax/cmϪ1 1750; λmax/nm 213,
J. Chem. Soc., Perkin Trans. 1, 1998
2993