A.-H.A. Hamdy et al. / Steroids 74 (2009) 927–930
929
Fig. 4. Selected 1H NMR chemical shifts of 1, 10 and 11.
Fig. 3. Selected NOEs of compound 1.
agreement with the chemical shift of C-20 (ı 42.6) [12] and with
those reported for H3-21 and H-22, for the synthetic steroids 10
[13] and 11 [14] whose stereochemistries are 20S, 22R and 20R, 22R,
respectively (Fig. 4). Thus, the absolute configuration of 1 can be
The IC50 values on the cell growth were obtained from the drug
dose–response curves using the software Origin 6.1.
3. Results and discussion
Investigation on the natural occurrence of 22-hydroxylated
cation, for the first time in the Plant Kingdom, of the C27-BRs 3–6
(Fig. 2). Both 3 and 6 partially rescued the det2 mutant phenotype
in the dark which supports the presence of an early C-22 oxida-
tion pathway [6]. Although little is known about the biosynthetic
pathways of C27-BRs such as 28-norcastasterone (Fig. 1), the find-
ing of compounds 3–6 supports the idea that C27-sterols are direct
precursors of the corresponding C27-BRs, and suggests that 3–6
3-Keto-22-epi-28-nor-cathasterone 1 is a C27-polyoxygenated
steroid with the following features: (a) like natural brassinosteroids
it possesses a common 5␣-cholestane skeleton; (b) it contains only
one hydroxyl group in the side chain like compounds 3–7 [6], and
cathasterone 8 [7], both identified in cultured cells of Catharan-
thus roseus; (c) A/B rings of 1 bear a 3,6-diketo functionality, which
is uncommon in C27-steroids. At present, only a few compounds,
in addition to 1 and 2, have been reported with that functional-
ity [8]. Among BRs only one compound bearing a 3,6-diketo group,
3-dehydroteasterone 9, has been identified [9].
derive from cholesterol via a biosynthetic pathway similar to C28
-
BRs biosynthesis [15]. In fact, cholesterol is the best substrate for
cytochrome P450, CYP90B1, that catalyses C-22 hydroxylation and,
according to kinetic analysis, the substrate specificity of CYP90B1
indicated that sterols with a C-5–C-6 double bond are preferred
substrates when compared with stanols, which have no double
bond at that position [16].
Compound 1 was active against human liver (HEPG-2) and colon
(HCT116) cancer cells with an IC50 2.96 M and IC50 12.38 M,
respectively. Compound 2 was found to be significantly cytotoxic
to both cell lines, with an IC50 of 5.63 and 1.16 M, respectively. It
exhibited a particularly high cytotoxicity to liver cancer cells, where
the cell number of HEPG-2 fell below 10% that of the control at
3.79 M (IC10).
Compound 1 was isolated as a white powder. NMR data coupled
with a molecular ion at m/z 416 (HREIMS) suggested a molecular
formula of C27H44O3, indicating six degrees of unsaturation. The 13
C
NMR data and a DEPT NMR experiment (Table 1) were consistent
for a C-27 sterol having two ketone and one hydroxyl groups. Com-
parison of 1H and 13C chemical shifts of 1 with those reported for
steroids possessing a diketo-3,6 nucleus such as stigmastane-3,6-
dieneone [10] indicated that 1 must be a 3,6-diketocholestane. 2D
NMR experiments, HSQC, HMBC, and COSY, allowed us to confirm
the position of the ketone groups at C-3 and C-6. The presence of
a hydroxyl group at C-22 was established by COSY correlations and
by the HMBC correlation of H3-21 (ıH 0.92) with the carbon that
bears the hydroxyl group (ıC 73.9, C-22).
The relative stereochemistry of 1 was assigned on the basis of a
2D NOESY experiment. The C-5/C-10; C-8/C-9 and C-13/C-14 trans
fused ring junctions were established by NOE correlations of H-
8 with H3-18 and H3-19. NOEs observed between H3-18 and H-20
indicateda dispositionforthesidechain(Fig. 3). Theabsolutecon-
figuration of C-22 was established by derivatization with (R)- and
(S)-␣-methoxy-␣-phenylacetic acids (MPA). NMR analysis [11] of
the ꢁı values for the two MPA esters 1a and 1b gave clear evidence
to assign the absolute stereochemistry at C-22 as R (Table 2). The
observed NOEs between H-22 and H2-16 and between H3-21 and
H-12 indicated C-21 must be in ␣ disposition, therefore, consider-
ing that the absolute stereochemistry of C-20 has been established
as R, the absolute stereochemistry of C-20 must be S. This is in good
Acknowledgments
This work was supported by the Academy of Science and Tech-
nology (Egypt) and Ministerio de Educación y Ciencia (Spain)
(BIO2007-61745, SAF2006-03004). A.R.D.-M. acknowledges finan-
cial support from Programa Juan de la Cierva (Ministerio de
EducaciónyCienciaofSpain). Mr. E. PalmerandMr. D. Lilesprovided
technical support. The Government of Egypt granted permission for
the collection of the samples.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
Table 2
References
1H NMR ꢁı (ıR − ıS) values (CDCl3, ppm, recorded at 500 MHz) of the diastere-
omeric mPA esters 1a and 1b.
[1] Clouse SD. Brassinosteroids plant counterparts to animal steroid hormones?
Vitam Horm 2002;65:195–223.
[2] Bajguz A, Tretyn A. The chemical characteristic and distribution of brassinos-
teroids in plants. Phytochemistry 2003;62:1027–46.
[3] Yokota T, Kim SK, Fukui Y, Takahashi N, Takeuchi Y, Takematsu T. Brassinos-
teroids and sterols from a green alga, Hydrodictyon reticulatum: configuration
at C-24. Phytochemistry 1987;26:503–6.
[4] Kontiza I, Abatis D, Malakate K, Vagias C, Roussis V. 3-Keto steroids from
the marine organisms Dendrophyllia cornigera and Cymodocea nodosa. Steroids
2006;71:177–81.
#
ıR
ıS
ꢁıRS (ıR − ıS)
H-20
1.53
0.64
1.46
0.82
0.83
1.81
0.89
1.24
0.64
0.65
−0.28
−0.25
+0.22
+0.18
+0.18
Me-21
H-25
Me-26a
Me-27a
a
Interchangeable values.