The synthesis of androstane brassinosteroid analogues with α-azido acid ester groups in position 17β

Article abstract of DOI:10.1016/j.steroids.2010.06.012

Androstane brassinosteroid analogues with α-azido acid ester groups in position 17β were synthesized from 2α,3α,17β- trihydroxy-5α-androstan-6-one after the protection of the 2α,3α-diols upon treatment with the corresponding α-azido acid and the subsequent deprotection of the diol grouping. Six new castasterone analogues were prepared. The biological activities were evaluated in two bioassays: a rice lamina inclination test and bean second internode bioassays. The activities of the new analogues differ in these two bioassays.

Full text of DOI:10.1016/j.steroids.2010.06.012

Steroids 75 (2010) 1005–1010
Contents lists available at ScienceDirect
Steroids
journal homepage: www.elsevier.com/locate/steroids
The synthesis of androstane brassinosteroid analogues with ␣-azido acid ester
groups in position 17␤
Jaroslava Hnilickovaa, Ladislav Kohouta, Enric Capdevilab, Ana Esteveb, Marc Vilaplanab,
Meritxell Molistb, Carme Brosab, Jana Swaczynova-Oklestkovac, Barbora Slavikovaa,∗
a Institute of Organic Chemistry and Biochemistry, Academy of Sciences of the Czech Republic, Flemingovo nám. 2, 166 10 Prague 6, Czech Republic
b Department of Organic Chemistry and Biochemistry, Institut Químic de Sarrià, C.E.T.S., Universitat Ramon Llull, Via Augusta 390, 08017 Barcelona, Spain
c Laboratory of Plant Growth Regulators, Institute of Experimental Botany, Academy of Sciences of the Czech Republic, Slechtitelu 11, 783 71 Olomouc, Czech Republic
a r t i c l e i n f o
a b s t r a c t
Article history:
Androstane brassinosteroid analogues with ␣-azido acid ester groups in position 17␤ were synthesized
from 2␣,3␣,17␤-trihydroxy-5␣-androstan-6-one after the protection of the 2␣,3␣-diols upon treatment
with the corresponding ␣-azido acid and the subsequent deprotection of the diol grouping. Six new
castasterone analogues were prepared. The biological activities were evaluated in two bioassays: a rice
lamina inclination test and bean second internode bioassays. The activities of the new analogues differ
in these two bioassays.
Received 23 March 2010
Received in revised form 22 June 2010
Accepted 23 June 2010
Available online 1 July 2010
Keywords:
Brassinosteroids
© 2010 Elsevier Inc. All rights reserved.
Androstane analogues
Plant growth regulators
Rice lamina inclination test
Bean second internode bioassay
1. Introduction
with various ester groups [5] in the position 17␤, e.g. compound
3, or the analogue of castasterone with only a 17␤ hydroxyl group
Brassinosteroids [1] (BRs) are plant growth-promoting sub-
stances that induce cell elongation and division, increase plant
yields and especially decrease or even remove plant stress evoked
by such conditions as drought, herbicidal or pesticidal overdoses,
salinity, high or low temperature, etc. [2]. Some of our latest
studies [3] have proven that brassinosteroids are able to man-
age plant water economy during a drought period by decreasing
plant activity, e.g. by reducing photosynthesis with a simultane-
ous conservation of the whole plant for more favorable conditions.
Brassinosteroid-treated plants are then able to overcome the
drought period in a much better condition than non-treated plants.
We have also found that the application of brassinosteroids at any
stage of plant growth leads to a hormonal equilibration of the other
plant hormones [4], which contributes not only to better plant con-
ditions but finally also to increased plant yields.
[6], i.e. compound 4.
Brassinolide activity is usually determined in the literature by
two basic bioassays: a bean second internode bioassay (BSIB) [7]
and a rice lamina inclination test (RLIT) [8]. The BSIB measures the
prolongation of the second internode, whereas the RLIT measures
the angle on the inclination of rice lamina. Both of these effects
are given by the elongation and/or multiplication of the plant cells.
Compounds 3 and 4 elicited a high activity in BSIB with compound
3 cleaving the second internode, but neither of them exhibited any
activity in the RLIT. The androstane analogues 3 and 4 with a high
activity in BSIB also elicited a high activity in the field application.
To understand this problem in greater detail, we applied compu-
tational modeling [9], based on which we designed new structures
where high activity in the RLIT would be expected. The designed
structures had a keto group in the region of the 22-hydroxyl group
and an azido group in the region of the 23-hydroxyl group.
In the course of our studies on the brassinosteroid
structure–activity relationship, we have described androstane
analogues of brassinolide (1, Fig. 1) and castasterone (2) without
any carbon atom in the side chain, i.e. an analogue of brassinolide
2. Experimental
The melting points were determined on a Kofler block. The opti-
cal rotations were measured at 25 ◦C in chloroform (unless stated
otherwise) and the [˛]D values are given in 10−1 deg cm2 g−1. The
infrared spectra were recorded on a Bruker IFS 88 spectrometer in
tetrachloromethane (unless stated otherwise), the wavenumbers
∗
Corresponding author.
E-mail addresses: brosa@iqs.url.es (C. Brosa),
barbora@uochb.cas.cz (B. Slavikova).
0039-128X/$ – see front matter © 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.steroids.2010.06.012

1006
J. Hnilickova et al. / Steroids 75 (2010) 1005–1010
Fig. 1. Brassinolide, castasterone and their analogues.
of which are given in cm−1. The 1H NMR spectra were taken in
deuteriochloroform on Bruker AVANCE-400 (at 400 MHz) instru-
ment in deuteriochloroform solutions with tetramethylsilane as
an internal reference (unless stated otherwise). Chemical shifts are
given in ppm (ı-scale), using coupling constants (J) in Hz. All of
the values were obtained by first-order analyses. The mass spectra
were obtained with a ZAB-EG spectrometer at 70 eV. The identity
of the prepared samples was checked by melting points, thin-layer
chromatography (TLC) on a silica gel G (ICN Biochemicals, detec-
tion done by spraying with sulfuric acid and heating) as well as
the IR and 1H NMR spectra. The preparative TLC was carried out
on 200 mm × 200 mm plates coated with a 0.7 mm layer of silica
gel (Woelm DC), with detection done by spraying the plates with
a 0.2% methanolic morin solution, by UV detection or by spraying
with sulfuric acid and heating the side strips of the plates (unless
stated otherwise). For column chromatography, a neutral silica gel
of 60–120 ␮m was used (Service Laboratories of the Institute).
Light petroleum refers to a fraction boiling at 40–65 ◦C.
pH 5.7). The pots were placed in a light-controlled cultivation room
(25–27 ◦C, light: 48 W/m2, light/dark period: 16 h/8 h). Groups of six
to eight 7-day-old bean seedlings with 1–2 mm long second intern-
odes were treated with different amounts of the tested compounds
in lanolin (0.25 ml). The control plants were treated with lanolin
alone. The measurements were made after 5 days. The difference
in the length (ꢀl) of the second internode of the treated and control
plants was used as a measure of the activity.
2.3. General procedure
2.3.1. Preparation of azido acid
2.3.1.1. A
solution
of
trifluoromethanesulfonyl
azide
in
dichloromethane. Sodium azide (4.9 g, 75 mmol) was dissolved
under stirring with a magnetic stirrer in 12.5 ml of water and 20 ml
of dichloromethane. The mixture was then stirred for 20 min in
an ice bath. Trifluoromethanesulfonic anhydride (2.5 ml, 15 mmol)
was slowly added and the obtained solution was stirred for 2 h in
an ice bath. The organic and water phases were separated, and
the water phase was extracted with dichloromethane (3 × 10 ml).
The organic phase and dichloromethane extracts were combined
and washed with a solution of potassium hydroxide (5% solution,
2 × 10 ml). This solution was used directly for the synthesis of the
azido acids.
2.1. Rice lamina inclination test
Seeds of Bahia rice cultivar were soaked in water and incubated
in a growth chamber under a 16 h light/8 h darkness photoperiod at
30 ◦C for 2 days. The germinated seeds were planted on the surface
of a 0.5% aqueous agar medium and incubated under the above
conditions for 4 days. Selected seedlings were treated with the
brassinosteroid test solutions (in 95% ethanol), which were applied
with a microsyringe (0.5 ml) into the second lamina joint of the
plant sheath (10–12 plants). Both treated and untreated (control)
seedlings were returned to the growth chamber at 30 ◦C in the dark
for 2 days. The interior angle between the leaf lamina of the sec-
ond leaf and its leaf sheath was then measured. This procedure
was repeated 3 or 4 times for each dose tested, and the statistical
parameters were calculated.
2.3.1.2. Azido acid. Potassium carbonate (792 mg), copper sulfate
powder (30 mg), and methanol (24 ml) were added to a solution of
amino acid (500 mg) in water (12 ml), after which a freshly prepared
solution of trifluoromethanesulfonyl azide (containing 15 mmol of
trifluoromethanesulfonyl azide) in dichloromethane from the pre-
vious reaction was added. To make the mixture homogeneous,
methanol was added (100 ml) and the resulting mixture was stirred
for 8 h while monitoring the pH – it was to be between 9 and 10
but at least 8. The pH was adjusted with a solution of potassium
hydroxide (5%).
2.2. Bean second internode bioassay
After the addition of water and dichloromethane (100 ml each),
the phases were separated. The organic phase was washed three
times with water, the water phases were combined and then they
were acidified with 2 N sulfuric acid to pH 3.5. The acidified water
phases were extracted with ethyl acetate. The ethyl acetate phase
Bean seeds (Phaseolus vulgaris L., cv. Pinto) were germinated for
2 days. Selected germinated seeds were planted into pots contain-
ing perlite and modified Hoagland’s solution (half concentration,

J. Hnilickova et al. / Steroids 75 (2010) 1005–1010
1007
was dried with sodium sulfate and evaporated with benzene to
dryness. After the addition of chloroform, the crystals of triflu-
oromethanesulfonamide [10] were removed from the obtained
mixture and oily mother liquor was used as the azido acid for the
next reaction.
117 mg (24%) crystalline 6-oxo-2␣,3␣-(2,2-propylendioxy)-5␣-
androstan-17␤-yl-2␰-azido-pentanoate (6), m.p. 159–165 ◦C
(methanol/water). IR: 2116, 2104 (azide), 1744 (CO ester), 1715
(C O), 1381, 1368 (CH3), 1275, 1265 (azide), 1242, 1218, 1183
(C–O), 1164, 1093, 1058 (dioxolane). 1H NMR: 0.69 s, 3H (3×
H-18), 0.82 s, 3H (3× H-19), 0.965 t, 3H (Jꢀ = 7.4, CH3CH2), 1.32
s and 1.48 s, 6H (2× CH3 acetonide), 2.56 dd, 1H (H-5␣), 3.59
m, 1H (H–C–N3), 4.10 m, 1H (H-2␤), 4.28 m, 1H (H-3␤), 4.72 m,
1H (H-17␣). FAB-MS, m/e: 488 (M+1). Analysis: for C27H41N3O5
(487.64) calcd. C, 66.50; H, 8.47; N, 8.62. Found. C, 66.49; H, 8.54;
N, 8.38.
2.3.2. Esterification of steroid alcohol with azido acid
Steroid alcohol (300 mg, 0.83 mmol) was dissolved in 15 ml
of benzene. A magnetic stirrer and one half of the proper, i.e.
freshly synthesized, azido acid (general procedure 2.3.1.) were
added and about 5 ml of benzene was distilled off. The mixture
was prevented from coming into contact with air moisture (via
a KOH air moisture seal). After cooling to room temperature,
N,Nꢀ-dicyclohexylcarbodiimide (DCC, 150 mg, 0.72 mmol) and 4-
dimethylaminopyridine (DMAP, 3 mg, 0.03 mmol) were added. The
mixture was stirred for 24 h at room temperature. The reaction was
checked by TLC (light petroleum/diethyl ether 1:1). In some cases,
if a starting steroid was present according to TLC, the second half of
the azido acid, and further portions of DCC (150 mg, 0.72 mmol) and
DMAP (3 mg, 0.03 mmol) were added. The mixture was stirred for
the next 24 h or longer until all the starting material was consumed
in the reaction.
2.6. 6-Oxo-2˛,3˛-(2,2-propylendioxy)
-5˛-androstan-17ˇ-yl-2R-azido-pentanoate (7)
Analogously, acetonide 5 (300 mg, 0.83 mmol) and azido acid
prepared from d-norVal (500 mg, 3.5 mmol) according to procedure
2.3.2. afforded crystals of 6-oxo-2␣,3␣-(2,2-propylendioxy)-5␣-
androstan-17␤-yl-2R-azido-pentanoate (7, 240 mg, 59%), m.p.
165–173 ◦C (methanol), [˛]20 = +28.0 (c 0.193). IR: 2115, 1265
(azide); 1735 (CO ester), 1709 (C O); 1192 (C–O); 1383, 1370, 1241,
1056 (dioxolane). 1H NMR: 0.69 s, 3H (3× H-18), 0.82 s, 3H (3×
H-19), 0.97 t, 3H (J = 7, CH3CH2), 1.34 s and 1.51 s, 6H (2× CH3 ace-
tonide), 2.55 dd, 1H (H-5␣), 3.78 m, 3.78 d, 1H (H–C–N3), 4.10 m,
1H (H-2␤) 4.29 m, 1H (H-3␤), 4.71 m, 1H (H-17␣). FAB-MS, m/e:
488 (M+1)+. Analysis: for C27H41N3O5 (487.64) calcd. C, 66.50; H,
8.47; N, 8.62. Found. C, 66.49; H, 8.54; N, 8.38.
The mixture was subsequently poured into water, extracted
with diethyl ether, washed with water, saturated by a KHCO3 solu-
tion, dried over sodium sulfate and evaporated. The obtained ester
was purified by column chromatography, it underwent elution with
light petroleum/diethyl ether (3:1), and the product was crystal-
lized from methanol to yield white needles in all of the cases.
2.7. 6-Oxo-2˛,3˛-(2,2-propylendioxy)
-5˛-androstan-17ˇ-yl-2S-azido-pentanoate (8)
2.3.3. Deprotection of 2˛,3˛-diol
Acetonide (35 mg, 0.07 mmol) was dissolved in chloroform
(5 ml) and 37% hydrochloric acid (0.1 ml). The obtained mixture
was shaken for 1 min. The solvents were evaporated in vacuum
until dryness at 30 ◦C.
Acetonide
5 (300 mg, 0.83 mmol) and azido acid derived
from l-norVal (500 mg, 3.5 mmol) according to general pro-
cedure 2.3.2. afforded 130 mg (32%) of white-needle crystals
of
2␣,3␣-(2,2-propylendioxy)-dihydroxy-6-oxo-5␣-androstan-
2.4. 17ˇ-Hydroxy-2˛,3˛-(2,2-propylendioxy)
-5˛-androstan-6-one (5)
17␤-yl-2S-azido-pentanoate (8), m.p. 156–159 ◦C (methanol),
[˛]20 = +25.0 (c 0.230). IR: 2115, 1275 (azide), 1735 (CO ester),
1709 (C O), 1192 (C–O), 1383, 1056, 1015 (dioxolane). 1H NMR:
0.69 s, 3H (3× H-18), 0.82 s, 3H (3× H-19), 0,97 t, 3H (CH3CH2),
1.34 s, 3H (CH3 acetonide), 1.51 s, 3H (CH3 acetonide), 2.55 dd, 1H
(H-5), 3.78 dd, 1H (H–C–N3), 4.10 m, 1H (H-2␤), 4.29 m, 1H (H-3␤),
4.71 t, 1H (H-17␣). Analysis: for C27H41N3O5 (487.62) calcd. C,
66.50; H, 8.47; N, 8.62. Found. C, 66.20; H, 8.52; N, 8.59.
Variant (a): 4-Toluenesulfonic acid (250 mg, 0.69 mmol) was
added to a solution of 2␣,3␣,17␤-trihydroxy-5␣-androstan-6-one
4 (1 g, 3.1 mmol) in acetone (200 ml). The solution was stirred with
a magnetic stirrer overnight. Potassium carbonate (2 g) was then
added and 3/4 of the acetone was quickly distilled off in a vacuum.
The residue was extracted into chloroform, washed with water,
dried (sodium sulfate) and evaporated. The residue was crystallized
from methanol to afford 820 mg (75%) white needles of acetonide
5, m.p. 220–222 ◦C. IR: 3614 (OH), 1707 (C O), 1383, 1370, 1241,
1057, 1014 (dioxolane). 1H NMR: 0.69 s, 3H (3× H-18), 0.74, 3H (3×
H-19), 1.34 s and 1.50 s, 6H (2× CH3 acetonide), 2.56 dd, 1H (H-5␣),
3.68 m, 1H (H-17␣), 4.10 m, 1H (H-2␤), 4.28 m, 1H (H-3␤). Anal-
ysis: for C22H34O4 (362.5) calcd. C, 72.89; H, 9.45. Found. C, 72.86;
H, 9.46.
Variant (b): A solution of 2␣,3␣,17␤-trihydroxy-5␣-androstan-
6-one 4 (2 g, 6.2 mmol) in dimethoxypropane (25 ml) was treated
with 4-toluenesulfonic acid (140 mg). The mixture was stirred for
20 h and then taken into diethyl ether, washed with water, potas-
sium hydrogen carbonate, dried (sodium sulfate), and evaporated.
Crystallization from methanol afforded compound 5 (1.7 g, 77%),
which is identical to the acetonide obtained under (a).
2.8. 6-Oxo-2˛,3˛-(2,2-propylendioxy)
-5˛-androstan-17ˇ-yl-2␰-azido-3-methylpentanoate (9)
Acetonide 5 (300 mg, 0.83 mmol) and azido acid derived from
d,l-isoLeu (500 mg, 3.2 mmol) according to general procedure
2.3.2. afforded 80 mg (19%) of 6-oxo-2␣,3␣-(2,2-propylendioxy)-
5␣-androstan-17␤-yl-2␰-azido-3-methylpentanoate (9), m.p.
170–178 ◦C (methanol) IR: 2110, 1262 (azide); 1732, 1709 (C O);
1241 (C–O); 1056 (dioxolane); 1383 (CH3). 1H NMR: 0.69 s, 3H
(3× H-18), 0.82 s, 3H (3× H-19), 0,97 t, 3H (CH3CH2), 0.93 d, 3H
(CH3CH), 1.34 s, 3H and 1.50 s, 3H (2× CH3 acetonide), 2.55 dd, 1H
(H-5␣), 3.81 d, 1H (H–C–N3), 4.10 m, 1H (H-2␤), 4.28 m 1H (H-3␤),
4.72 t, 1H (H-17␣). MS EI, m/e: 486 (M CH3)+, 444 (M−N3)+.
Analysis: for C28H43N3O5 (501.66) calcd. C, 67.04; H, 8.64; N, 8.38.
Found. C, 67.22; H, 8.54; N, 8.01.
2.5. 6-Oxo-2˛,3˛-(2,2-propylendioxy)
-5˛-androstan-17ˇ-yl-2␰-azido-pentanoate (6)
2.9. 6-Oxo-2˛,3˛-(2,2-propylendioxy)
-5˛-androstan-17ˇ-yl-2R-azido-3-methylpentanoate (10)
The treatment of acetonide
5 (300 mg, 0.83 mmol) with
azido acid (prepared from d,l-norVal, general procedure 2.3.1.,
500 mg, 3.5 mmol) following general procedure 2.3.2. afforded
Acetonide 5 (300 mg, 0.83 mmol) and azido acid derived from
d-isoLeu (500 mg, 3.2 mmol) according to general procedure

1008
J. Hnilickova et al. / Steroids 75 (2010) 1005–1010
2.3.2. afforded 100 mg (24%) of 6-oxo-2␣,3␣-(2,2-propylendioxy)-
5␣-androstan-17␤-yl-2R-azido-3-methylpentanoate (10) as white
needles (methanol), m.p. 163–172 ◦C. [˛]20 = +53 (c 0.076). IR: 2109,
1264 (azide); 1732 (CO ester), 1709 (C O); 1191 (C–O), 1383, 1241,
1056 (dioxolane). 1H NMR: 0.69 s, 3H (3× H-18), 0.82 s, 3H (3× H-
19), 0,92 t, 3H (CH3CH2), 0.97 d, 3H (CH3CH), 1.34 s, 3H and 1.51 s,
3H (2× CH3 acetonide), 2.55 dd, 1H (H-5␣), 3.69 d, 1H (H–C–N3),
4.10 m, 1H (H-2␤), 4.28 m, 1H (H-3␤), 4.72 t, 1H (H-17␣). FAB-MS,
m/e: 502 (M+1)+. Analysis: for C28H43N3O5 (501.66) calcd. C, 67.04;
H, 8.64; N, 8.38. Found. C, 66.99; H, 8.71; N, 7.99.
plates afforded 36 mg of the product, which was crystallized
from the acetone/n-heptane mixture to afford 25 mg (45%)
of pure 2␣,3␣-dihydroxy-6-oxo-5␣-androstan-17␤-yl-2S-azido-
pentanoate (14), m.p. 153–157 ◦C. IR: 3628, 3585, 3527, 3410, 3350
(OH); 2115, 2103 (azide); 1743 (CO ester); 1715 (C O); 1701 (CO);
1264, 1246, 1185 (C–O); 1041 (e), 1013 (a) (C–OH). 1H NMR: 0.77
s, 3H (3× H-18), 0.83 s, 3H (3× H-19), 0.97 t, 3H (CH3CH2), 2.70 dd,
1H (H-5␣, J = 6.4 and 3.5 Hz), 3.76 m, 2H (H–C–N3 and H-2␤), 4.06
m, 1H (H-3␤), 4.735 t, 1H (H-17␣). EI-MS, m/e: 447 (M+). Analysis:
for C24H37N3O5 (447.58) calcd. C, 64.41; H, 8.33; N, 9.39. Found. C,
64.13; H, 8.19; N, 9.41.
2.10. 6-Oxo-2˛,3˛-(2,2-propylendioxy)
-5˛-androstan-17ˇ-yl-2S-azido-3-methylpentanoate (11)
2.14. 2˛,3˛-Dihydroxy-6-oxo-5˛-androstan
-17ˇ-yl-2␰-azido-3-methylpentanoate (15)
Acetonide 5 (300 mg, 0.83 mmol) and azido acid prepared
from l-isoLeu (500 mg, 3.2 mmol) according to general procedure
2.3.2. afforded 120 mg (29%) of 6-oxo-2␣,3␣-(2,2-propylendioxy)-
5␣-androstan-17␤-yl-2S-azido-3-methylpentanoate (11), m.p.
187–193 ◦C (methanol), [˛]20 = +33.3 (c 0.048). IR: 2110, 1264
(azide); 1732, 1709 (C O); 1241 (C–O); 1095, 1056 (dioxolane);
1383 (CH3). 1H NMR: 0.69 s, 3H (3× H-18), 0.83 s, 3H (3× H-19),
0,92 t, 3H (CH3CH2), 0.97 d, 3H (CH3CH), 1.34 s, 3H and 1.50 s, 3H
(2× CH3 acetonide), 2.55 dd, 1H (H-5␣), 3.69 d, 1 H (H–C–N3), 4.10
m, 1H (H-2␤), 4.28 m, 1H (H-3␤), 4.72 t, 1H (H-17␣). FAB-MS, m/e:
502 (M+1)+. Analysis: for C28H43N3O5 (501.66) calcd. C, 67.04; H,
8.64; N, 8.38. Found. C, 66.83; H, 9.01; N, 8.38.
Compound 9 (45 mg, 0.09 mmol) was hydrolyzed following
general procedure 2.3.3. The residue (36 mg, 88%) was crys-
tallized from the acetone/n-heptane mixture to afford 17 mg
(41%) of 2␣,3␣-dihydroxy-6-oxo-5␣-androstan-17␤-yl-2␰-azido-
3-methylpentanoate (15), m.p. 71–74 ◦C. IR: 3626, 3579, 3415 (OH);
2108, 660 (azide); 1738 (CO ester); 1714 (C O); 1262, 1248, 1187
(C–O); 1041 (e), 1013 (a) (C–OH). 1H NMR (400 MHz): 0.78 s, 3H
(3× H-18), 0.84 s, 3H (3× H-19), 0.92 t, 3H (CH3CH2, J = 7.4 Hz), 0.98
d, 3H (CH3CH, J = 6.8), 2.70 dd, 1H (H-5␣), 3.66 m, 1H (H–C–N3),
3,74 m, 1H (H-2␤), 4.04 m, 1H (H-3␤), 4.73 t, 1H (H-17␣). FAB-
MS, m/e: 462 (M+1)+, 444 (M−H2O)+, 426 (M−2×H2O)+. Analysis:
for C25H39N3O5 (461.61) calcd. C, 65.05; H, 8.52; N, 9.10. Found. C,
65.11; H, 8.43; N, 9.40.
2.11. 2˛,3˛-Dihydroxy-6-oxo-5˛
-androstan-17ˇ-yl-2␰-azido-3-pentanoate (12)
2.15. 2˛,3˛-Dihydroxy-6-oxo-5˛-androstan
-17ˇ-yl-2R-azido-3-methylpentanoate (16)
Compound 6 (35 mg, 0.07 mmol) was hydrolyzed according to
general procedure 2.3.3. The residue was purified on two prepar-
ative TLC plates, undergoing elution with light petroleum/ethyl
acetate (1:1, v/v). Working up the plates afforded 27 mg (84%)
of pure 2␣,3␣-dihydroxy-6-oxo-5␣-androstan-17␤-yl-2␰-azido-
3-pentanoate (12), m.p. 133–137 ◦C (acetone/n-heptane). IR: 3628,
3585, 3416 (OH); 2116, 1275, 1040, 1013 (azide); 1743 (CO ester);
1715 (C O); 1264, 1244 (C–O); 1040 (e), 1013 (a) (C–OH). 1H NMR:
0.77 s, 3H (3× H-18), 0.83 s, 3H (3× H-19), 0,96 t, 3H (CH3CH2),
2.69 dd, 1H (H-5␣), 3.77 m, 2H (H–C–N3 and H-2␤), 4.05 m, 1H
(H-3␤), 4.73 t, 1H (H-17␣). EI-MS, m/e: 447 (M+), 432 (M−CH3)+,
429 (M−H2O)+. Analysis: for C24H37N3O5 (447.58) calcd. C, 64.41;
H, 8.33; N, 9.39. Found. C, 64.11; H, 8.42; N, 9.50.
Compound 10 (39 mg, 0.08 mmol) was hydrolyzed accord-
ing to general procedure 2.3.3. The residue (33 mg, 92%)
was purified on two preparative TLC plates, undergoing elu-
tion with light petroleum/ethyl acetate (1:1, v/v). Working up
the plates afforded 24 mg of the product, which was crys-
tallized from the acetone/n-heptane mixture to afford 12 mg
(33%) of 2␣,3␣-dihydroxy-6-oxo-5␣-androstan-17␤-yl-2R-azido-
3-methylpentanoate (16), m.p. 106–110 ◦C. IR: 3625, 3578, 3415
(OH); 2109, 660 (azide); 1739 (CO ester); 1715 (C O); 1262, 1249,
1187 (C–O); 1040 (e), 1013 (a) (C–OH). 1H NMR (400 MHz): 0.77 s,
3H (3× H-18), 0.83 s, 3H (3× H-19), 0.92 t, 3H (CH3CH2, J = 7.4 Hz),
0.97 d, 3H (CH3CH, J = 6.8), 2.70 dd, 1H (H-5␣), 3.67 m, 1H (H–C–N3),
3,75 m, 1H (H-2␤), 4.06 m, 1H (H-3␤), 4.73 t, 1H (H-17␣). FAB-
MS, m/e: 462 (M+1)+, 444 (M−H2O)+, 426 (M−2 × H2O)+. Analysis:
for C25H39N3O5 (461.61) calcd. C, 65.05; H, 8.52; N, 9.10. Found. C,
65.10; H, 8.21; N, 9.31.
2.12. 2˛,3˛-Dihydroxy-6-oxo-5˛-androstan
-17ˇ-yl-2R-azido-pentanoate (13)
Compound 7 (20 mg, 0.04 mmol) was hydrolyzed follow-
ing general procedure 2.3.3. The residue (16 mg, 87%) was
crystallized from an acetone/n-heptane mixture to afford 7 mg
(38%) of 2␣,3␣-dihydroxy-6-oxo-5␣-androstan-17␤-yl-2R-azido-
pentanoate (13), m.p. 134–136 ◦C. IR: 3630, 3584, 3417 (OH); 2115,
2104 (azide); 1741 (CO ester); 1715 (C O); 1265, 1294, 1185 (C–O);
1041 (e), 1013 (a) (C–OH). 1H NMR: 0.77 s, 3H (3× H-18), 0.83 s,
3H (3× H-19), 0,97 t, 3H (CH3CH2), 2.69 dd, 1H (H-5␣), 3.77 m, 2H
(H–C–N3 and H-2␤), 4.06 m, 1H (H-3␤), 4.74 t, 1H (H-17␣). EI-MS,
m/e: 447 (M+). Analysis: for C24H37N3O5 (447.58) calcd. C, 64.41;
H, 8.33; N, 9.39. Found. C, 64.31; H, 8.11; N, 9.42.
2.16. 2˛,3˛-Dihydroxy-6-oxo-5˛-androstan
-17ˇ-yl-2S-azido-3-methylpentanoate (17)
Compound 11 (22 mg, 0.04 mmol) was hydrolyzed accord-
ing to general procedure 2.3.3. The residue (19 mg, 95%)
was purified on two preparative TLC plates, undergoing elu-
tion with light petroleum/ethyl acetate (3:7, v/v). Working up
the plates afforded 15 mg of the product, which was crys-
tallized from acetone/n-heptane mixture to afford 6 mg (30%)
of pure 2␣,3␣-dihydroxy-6-oxo-5␣-androstan-17␤-yl-2S-azido-
3-methylpentanoate (17), m.p. 108–110 ◦C. IR: 3627, 3580, 3416
(OH); 2110, 661 (azide); 1738 (CO ester); 1714 (C O); 1262, 1249,
1187 (C–O); 1040 (e), 1013 (a) (C–OH). 1H NMR (400 MHz): 0.77
s, 3H (3× H-18), 0.84 s, 3H (3× H-19), 0.925 t, 3H (CH3CH2,
J = 7.45 Hz), 0.98 d, 3H (CH3CH, J = 6.8), 2.70 dd, 1H (H-5␣), 3.68
m, 1H (H–C–N3), 3,76 m, 1H (H-2␤), 4.06 m, 1H (H-3␤), 4.73 t,
1H (H-17␣). FAB-MS, m/e: 462 (M+1)+. Analysis: for C25H39N3O5
2.13. 2˛,3˛-Dihydroxy-6-oxo-5˛-androstan
-17ˇ-yl-2S-azido-pentanoate (14)
Compound
8 (60 mg, 0.12 mmol) was hydrolyzed accord-
ing to general procedure 2.3.3. The residue (43 mg, 78%) was
purified on three preparative TLC plates, undergoing elution
with light petroleum/ethyl acetate (1:1, v/v). Working up the

J. Hnilickova et al. / Steroids 75 (2010) 1005–1010
1009
Table 1
Results of the rice lamina inclination test.
◦
Compound
1 ␮g/plant
0.1 ␮g/plant
0.01 ␮g/plant
0.001 ␮g/plant
0.0001 ␮g/plant
0.00001 ␮g/plant
Activity −log[dose]45
28-homoCS
4
6
7
8
9
10
11
12
13
14
15
16
17
99
1
8
6
2◦
2◦
2◦
3
6
6
2◦
1◦
2◦
Residuala
Residuala
Residuala
Residuala
Residuala
Residuala
Residuala
0.46
8
6
2◦
2◦
12 4◦
7
13 3◦
7
1◦
2◦
12 5◦
29 7◦
4
3◦
−3 2◦
71 5◦
94 4◦
39 8◦
35 6◦
54 5◦
19 4◦
50 7◦
62 + 7◦
29 7◦
29 7◦
33 4◦
1◦
11 + 4◦
20 5◦
1
1.50
40 8◦
23 27◦
39 7◦
Residuala
Residuala
0.70
28 5◦
20 5◦
6
5◦
Residuala
a
Residual: below detection limit.
(461.61) calcd. C, 65.05; H, 8.52; N, 9.10. Found. C, 65.17; H, 8.41;
N, 9.19.
GRID program. We had previously developed [11] and validated
a quantitative structure–activity relationship model with this
methodology, taking into account the activity values obtained by
the RLIT. It was noticed [12] that the activity of a compound in RLIT
was dependent on the ability of the four hydroxyls of BL (1) and CS
(2) to form hydrogen bonds with the putative receptor. This abil-
ity was shown by MIF around these polar functionalities described
with the water probe.
3. Results and discussion
Owing to the low activity of compounds 3 and 4 found in
the RLIT, we studied these compounds with computational mod-
eling tools using Molecular Interaction Fields (MIF) with the
Fig. 2. Synthesis of derivatives of R-, S- and R,S-azido acids (a) Me2CO, p-TosOH; (b) benzene, DCC, DMAP, azido acid; (c) CHCl3, HCl.

1010
J. Hnilickova et al. / Steroids 75 (2010) 1005–1010
Table 2
[ꢀl is the difference in the prolongation of the second internode of
Results of the bean second internode bioassay.
the treated and control (untreated) plants in millimeters]. As for the
results obtained on the BSIB (Table 2), the synthesized compounds
exhibited low activity.
Hence, we see that the activities of these new brassinosteroids
are different in RLIT and BSIB, i.e. the bioassays used up to now
to determine brassinolide activity. This result will require further
studies, especially in the context that we have previously found
[15], namely that the activity in the BSIB corresponds to the activity
in the field.
Compound
ꢀl at 10−9 M applied
24-epiBR
4
6
7
8
9
10
11
12
13
14
15
16
17
34
27
9
2
4
1
3
12
2
8
Acknowledgements
6
15
1
This work was supported by Projects 1QS510 680 561 (AVCR)
and 1M 06030 (MSMT) and Research Project Z4 055 0506 (IOCB).
9
References
We noticed that the MIF described by an H2O probe around
hydroxyl 23 of BL (1) was lacking in the androstane analogues
without a side chain like 3 and 4.
[1] Brosa C. Biological effects of brassinosteroids. In: Parish EH, Nes WD, editors.
Biochemistry and function of sterols. Boca Raton: CRC Press; 1997. p. 201–18
[chapter 15].
[2] (a) Sasse J. Physiological actions of brassinosteroids. In: Sakurai A, Yokota
Y, Clouse SD, editors. Brassinosteroids—steroidal plant hormones. Berlin:
Springer-Verlag; 2003. p. 151–3;
These results showed that the synthesis of new androstane
analogues with
a polar functionality in the ␣ position of
the carbonyl in the ester moiety of the side chain could be
interesting.
Keeping this idea in mind, we described in this paper the syn-
thesis of different androstane analogues of brassinosteroids with
␣-azido acid ester groups in position 17␤.
ˇ
(b) Rothová O, Kocˇová M, Bartáková J, Holá D, Benesˇová M, Sísˇa M, et al. Inbred
and hybrid genotypes of maize differ in their response to brassinosteroids
under conditions of cold stress. In: XV. FESPB congress. 2006. Book of Abstracts,
p. 166;
(c) Bláha L, Hnilicˇka F, Kadlec P, Smrcˇková-Jankovská, Machácˇková I, Sychrová
E et al. Influence of abiotic stresses on the physiology of sprouting plants. IX.
ESA congress, Warsaw, Poland, September 4–7, 2006. Book of proceedings,
fragmenta agromomica vol. 11; 2006. p. 57–8.
Our synthesis started with 2␣,3␣,17␤-trihydroxy-5␣-
androstan-6-one (4). The 2,3-diol grouping of 4 was protected
as acetonide in the presence of 4-toluensulfonic acid to afford
the protected compound 5. Subsequently, we synthesized azido
acids. First, trifluoromethanesulfonyl azide (triftyl azide, TfN3) was
prepared by a modification of the published procedure [13,14]
from NaN3 and Tf2O (trifluoromethanesulfonic anhydride) in a
mixture of water and dichloromethane. The amino acid dissolved
in water was subsequently treated with potassium carbonate and
copper sulfate and TfN3 in dichloromethane, and, if necessary,
methanol was added to homogeneity. After working up and
separating the trifluoromethanesulfonamide, oily mother liquid
remains as the azido acid. Compound 5 (Fig. 2) was condensed
with the corresponding ␣-azido acid using the DCC and DMAP
in benzene to yield esters 6–11. These compounds were fur-
ther treated with hydrochloric acid and chloroform to free the
2,3-diols, thus yielding the desired analogues of castasterone
12–17.
ˇ
[3] (a) Holá D, Bartáková J, Kocˇová M, Rothová O, Benesˇová M, Kohout L, Sísˇa M.
Brassinosteroids improve the response of photosynthetic apparatus in maize
leaves to low-temperature conditions. In: XV. FESPB congress. 2006. Book of
Abstracts, p. 153;
(b) Bláha L, Kadlec P, Machácˇková I, Kohout L, Hnilicˇka F. Klícˇivost, vzcházivost
a vyuzˇití vody semeny. Úroda 2006;3:69–71.
[4] Kohout L, Hnilickova J, MachJ. Hormonal ratios in plants: treatment with brassi-
nosteroids equilibrates hormonal levels in the plant; in press.
[5] Strnad M, Kohout L. A simple brassinolide analogue 2␣,3␣-dihydroxy-17␤-(3-
methyl-butyryloxy)-7-oxa-B-homo-5␣-androstan-6-one which induces bean
second internode splitting. Plant Growth Regul 2003;40:39–47.
[6] Kohout L, Kasal A, Chodounská H, Slavíková B, Hnilicˇková J. The use of
2␣,3␣,17␤-trihydroxy-5␣-androstan-6-one for plant development regulation.
European Patent No. 1401278 (28.9.2005).
[7] Mitchell JW, Livingstone GA. Methods of studying plant hormones and growth
regulating substances. In: Agricultural handbook no. 336. Washington, DC: US
Government Printing Office; 1968. p. 26.
[8] Takeno K, Pharis RP. Brassinosteroid-induced bending of the leaf lamina
of dwarf rice seedlings—an auxin-mediated phenomenon. Plant Cell Physiol
1982;23:1275–81.
[9] (a) MOE 2001.01—Chemical Computing Group (www.chemcomp.com).;
(b) GRID22b—Molecular Discovery Ltd. (www.moldiscovery.com);
(c) Molist M, Ardévol A, Kohout L, Brosa C. Synthesis and bioactivity of
androstane brassinosteroid analogs having commercially available ester side
chain, selected by molecular modeling techniques. In: 21st conference on iso-
prenoids. 2005. Book of Abstracts, p. 58;
(d) Brosa C, Zamora I, Terricabras E, Kohout L. The effect of electrostatic prop-
erties and ability to form hydrogen-bonds on the activity of brassinosteroid
side-chain analogues. Collect Czech Chem Commun 1998;6:1635–45.
[10] Burdon J, Farazmand J, Stacey M, Tatlow JC. Fluorinated sulphonic acids.
1. Perfluoro-methane-sulphonic, perfluoro-octane-sulphonic, and perfluoro-
decane-sulphonic acids and their simple derivatives. J Chem Soc 1957:2574–8.
[11] Brosa C, Zamora I, Terricabras E, Soca L, Peracaula R, Rodríguez-Santamarta C.
Synthesis and molecular modeling: related approaches to progress in brassi-
nosteroid research. Lipids 1997;32(12):1341–7.
4. Activity evaluation
4.1. Rice lamina inclination test (RLIT)
A highly sensitive modified rice lamina inclination test, using
Bahia as a cultivar, based on the procedure developed by Takeno
and Pharis [7], has been used to evaluate the activity data of the
new, synthesized compounds (Table 1).
Table 1 shows the activity of castasterone analogues in RLIT. As
predicted, compounds 13 and 16, prepared from d-amino acids, are
the most active. Compounds synthesized from l-amino acids (14
and 17) are less active. This is caused by the fact that the compounds
prepared from d-amino acids have an azido group sterically in the
same position as 23-hydroxyl in castasterone.
[12] Brosa C. Structure–activity relationship. In: Sakurai A, Yokota T, Clouse SD, edi-
tors. Brassinosteroids: steroidal plant hormones. Tokyo: Springer-Verlag; 1999.
p. 191–222.
[13] Alper PB, Hung Shang-Cheng, Wong Chi-Huey. Metal catalyzed diazo
transfer for the synthesis of azides from amines. Tetrahedron Lett
1996;37(34):6029–32.
[14] Zaloom J, Roberts DC. Preparation of azido derivatives from amino acids and
peptides by diazotransfer. J Org Chem 1981;46:5173–6.
[15] Hnilicˇková J, Sísˇa M, Capdevila E, Esteve A, Brosa C, Kohout L. Synthesis
4.2. Bean second internode bioassay (BSIB)
ˇ
of androstane brassinosteroid analogues with ester group in position 17␤
designed by molecular modeling techniques. In: 15th European symposium
on organic chemistry (ESOC 15). Ireland: University College of Dublin; 2007.
Abstracts, p. 224.
Activity in the bean second internode bioassay, which has been
used for the evaluation of brassinolide activity since the discovery
of brassinolide, was utilised to compare the activities with the RLIT