Hydroxylation of the native

brassinosteroids 24-epicastasterone and

24-epibrassinolide by the fungus
Cunninghamella echinulata
Brunhilde Voigt,* Andrea Porzel,* Heidrun Naumann,t Claire H6rhoid-Schubert,t
and Giinter Adam*
*Institute of Plant Biochemistry, Halle, Weinbergweg, and tlnstitute of Microbiology and
Experimental Therapy, Jena, Beutenberg, Germany

24-Epicastasterone and 24-epibrassinolide, two naturally occurring phytohormones of the brassinoster-

oid type, were transformed by the fungus Cunninghamella echinulata to give the corresponding 12fl-
hydroxylated compounds. The structures of these compounds were determined by spectroscopic meth-
ods, especially heterocorrelated two-dimensional nuclear magnetic resonance investigations. In the
rice lamina inclination test the 12fl-hydroxylation lowered the bioactivity at 0.1 ppm to 10% in compari-
son with that of the corresponding parent hormones. The described hydroxylations represent the first
biotransformations of native brassinosteroids. (Steroids 58:320-323, 1993)

Keywords:biotransformations; fermentation; brassinosteroids; Cunninghamella echinulata; 12fl-hydroxy-24-epicas-

tasterone; 12/3-hydroxy-24-epibrassinolide; 12/3-hydroxylation; two-dimensional nuclear magnetic resonance, 24-
epicastasterone; 24-epibrassinolide

Introduction is the m o s t important c o m p o u n d for practical biological

The brassinosteroids r e p r e s e n t a new class of steroidal
p h y t o h o r m o n e s with high growth-promoting and anti-
stress activity. 1-3 W h e r e a s investigations on the chem-
istry and physiology of such c o m p o u n d s constitute an
area of t r e m e n d o u s activity,4,5 in the field of microbial
transformations only the 12/3-hydroxylation of the syn-
thetic c o m p o u n d s (22S,23S)-homocastasterone and
(22S,23S)-homobrassinolide by the fungus Cunningha-
mella echinulata has b e e n hitherto reported. 6 Further-
more, the microbial side chain degradation of 2a, 3a-
dihydroxy-5o~-cholestan-6-one, bearing the substitu-
tion pattern of brassinosteroids in the A/B ring moiety
by Mycobacterium vaccae, was reported to give 2a,
3or, 6 a - t r i h y d r o x y - 5 a - a n d r o s t a n - 1 7 - o n e and 2a-hydro-
xyandrost-4-ene-3,17-dione. 7 H o w e v e r , microbial
transformations of native p h y t o h o r m o n e s o f the brassi-
nosteroid type were not k n o w n until now.
O f the hitherto described 27 naturally occurring
brassinosteroids, 24-epibrassinolide (4) 8 b e c a u s e of its
s m o o t h synthetic accessibility as well as high activity,
Address reprint requests to Professor Dr. Giinter Adam, Institut for
Pflanzenbiochemie, Weinberg 3, Halle/Saale, Germany 0-4050.
Received August 20, 1992; accepted March 2, 1993.
investigations. 4 Therefore, we are interested in study-
ing biotransformations o1,4 as well as its native precur-
sor, 24-epicastasterone (1), 9 which could o p e n new
p a t h w a y s to additionally functionalized m e m b e r s for
structure-activity relationships and provide informa-
tion about possible metabolic and degradation pro-
cesses. This report describes the h y d r o x y l a t i o n o f 1
and 4 by the fungus C. echinulata (Figure 1).
Experimental
24-Epicastasterone (1) and 24-epibrassinolide (4) were synthe-
sized from ergosterol using a combination of the methods pub-
lished by Thompson et al. ~° and Anastasia et al. 11
Melting points were determined on a Boetius hot stage micro-
scope and are corrected, Specific rotations were determined in
methanol on a Zeiss instrument Polamat A. Infrared spectra
were recorded on a Zeiss spectrometer IR 75, and ultraviolet
(UV) spectra were recorded on a Zeiss Specord UV-VIS instru-
ment. ORD measurements were made on a Jasco spectropolar-
imeter ORD-UV 5. Nuclear magnetic resonance (NMR) spectra
were recorded on Bruker AM 500 and AC 300 instruments. The
mass spectra were obtained with an AMD 402 instrument. Silica
gel 60 (grain size 0.040 to 0.063 mm, Merck) was used for column
chromatography. Preparative high-performance liquid chroma-
tography (HPLC) was performed on a Knauer instrument on

320 Steroids, 1993, vol. 58, July © 1993 Butterworth-Heinemann

 M i c r o b i a l transformation o f brassinosteroids: Voigt et al.
R= .

R20 , . . . . ~ R20 , . . . . E ~ ~

R20 ,'" v R20 ..... ~ 77-0

0 0

1: R 1 = R2 = H 4: R 1 = R2 = H
2: R 1 = OH, R2 = H 5: R 1 = OH, R2 = H
3: R 1 = OAc, R2 = Ac 6: R 1 = OAc, Rz = Ac
Figure 1 Compounds 1-6.

Table I 13C chemical shifts (ppm) of 24-epicastasterone (1) and After 48 hours, 25 nag of brassinosteroids 1 and 4, respectively,
12jS-hydroxy-24-epicastasterone (2) (75.5 MHz; solvent, CDCI3/ dissolved in 1 ml of methanol were added. The flasks were reincu-
CD30D 95 : 5) bated for 5 days. After this incubation period, the entire medium
was extracted for the isolation of the transformation products
C 1 2 with n-butanol. The purification of the raw material was per-
formed by column chromatography (SiOz) and by HPLC.
1 39.8 39.8
2 67.9 68.0
3 68.0 68.1 12fl-Hydroxy-24-epicastasterone (2)
4 26.2 26.2
5 50.7 49.5 Elution of the silica gel column with ethyl acetate/methanol 9 : 1
6 213.1 212.3 v/v first afforded 22 mg starting compound 1. Further elution
7 46.6 46.0 gave crude 2, which was purified by preparative HPLC, leading
8 37.7 36.2 to 14 nag (14%, calculated for reacted 1) of the more polar com-
9 52.5 50.6 pound 2 with melting point (mp) 242-245 C (needles from metha-
10 42.5 41.9 nol); [c~]~4-7.3° (c = 0.241, MeOH); IR (KBr), v ~ : 3,400 (OH),
11 21.1 30.3
12 39.3 77.8 1,705 cm -1 (ketone); UV (MeOH), Xm~: 282 nm (8 190); ORD:
13 42.7 48.5 [(I)]280 + 343°, [(I)]307 - 229°, a = - 6 ; ELMS (m/z): 480 (M +,
14 53.6 55.0 CzsH~O6, 20%), 361 (480-C6H130-H20 , 10%), 343 (361-H20,
15 23.8 22.0 100%), 325 (343-H20, 62%), 303 (480-C9I-I1902-H20,50%); FAB-
16 27.5 28.0 MS (m/z): 481 (M + + 1, 100%), 463 (481-18, 17%), 445 (481-2
17 56.4 56.3 H20, 50%); IH NMR (500 MHz, CDCI3), (8): 0.748 (s, 18-H3),
18 11.7 8.5 0.754 (s, 19-H3), 0.827 (d, J = 7.1 Hz, 28-H3), 0.844 and 0.909
19 13.4 13.4 (d, J = 6.9 Hz and d, 6.9 Hz, 26-H3 and 27-H3), 1.072 (d, J =
20 40.0 39.2 7.2 Hz, 21-H3), 2.68 (dd, J -- 12.5/2.7 Hz, 5~x-H), 3.35 (dd, J =
21 12.2 11.6
22 72.3 75.2 6.4/4.9 Hz, 23-H), 3.56 (dd, J = 10.9/4.6 Hz, 12~x-H), 3.71 (ddd,
23 75.8 76.7 J = 11.8/4.9/3.1 Hz, 2/3-H), 3.75 (dd, J = 4.9/1.2 Hz, 22-H), 4.01
24 41.4 41.2 (ddd, J = 2.5/2.5/2.5 Hz, 3/3-H). 13C NMR: Table 1.
25 26.8 26.8
26 17.1 16.9
27 22.0 22.0 Pentaacetyl derivative 3
28 10.7 10.6
Compound 2 (10 rag) was acetylated with Ac20-pyridine (each
0.5 ml) for 20 hours at room temperature to give 9 mg (76%) 3
with mp 158 - 161 C (from CHC13); laiD24-13.6"(C = 0.214, MeOH);
IR (CHCI3), Vmax:1,720 cm -1 (CO); UV (MeOH), kmax: 279 nm,
LiChrospher 100 RP 18.5/~m with MeOH/H20 7 : 3 v/v and UV (8 200); ORD: [(I)]272 + 1,210°, [(I)]3o2 - 962°, a = - 2 2 ; EI-MS
detection at 210 nm. (m/z): 690 (M ÷ , 1%), 630 (M+-AcOH, 100%); 588 (8%); 570 (M ÷-
A culture of fungus strain C. echinulata IMET 43918 taken 2AcOH, 27%); 555 (14%); 528 (588-AcOH, 18%); 511 (555-CO 2,
from the collection of the Institute of Microbiology and Experi- 35%); 487 (35%); 445 (32%); 427 (50%); 415 (13%); 385 (32%);
mental Therapy, Jena, was maintained on agar slants: malt ex- 367 (12%); 343 (45%); 325 (38%). 1H NMR (300 MHz, CDC13)
tract 4%, yeast extract 0.3%, agar 1.5%. Cultures inoculated (8): 0.815 and 0.826 (2 x s, 18-H3and 19-H3), 0.831, 0.848, 0.934,
onto this medium were incubated at 25 C for 5 to 7 days. For and 0.980 (4 x d, J = 6.9 Hz, 6.8 Hz, 6.9 Hz; 6.9 Hz, 21-H3,
the precultivation, the cells were washed off into 500-ml shake 26-H3, 27-H3; and 28-H3), 1.987, 2.037, 2.043, 2.070, and 2.087,
flasks containing 100 ml of medium as described above with the (5 x s, 5 x OAc) 2.339 (dd, J = 13.4/4.4 Hz, 7a-H), 2.568 (dd,
exception of agar. The flasks were incubated on a rotary shaker J = 11.8/3.9 Hz, 5ct-H), 4.664 (dd, J = 10.8/4.6 Hz, 12t~-H), 4.915
(180 rpm) at 28 C for 2 days. Ten milliliters of the preculture (ddd, J = 10.2/6.8/3.0 Hz, 2/3-H), 5.029 (dd, J = 6.4/5.8 Hz, 23-
was transferred into shake flasks with 100 ml of the same medium. H), 5.283 (dd, J = 6.5/1.2 Hz, 22-H), 5.373 broad (s, 3~-H).

Steroids, 1993, vol. 58, J u l y 321

Papers
a) Me-19 ~ . ~ ] ~ ~ ~ Me-18
c)

19 18

_ ..j-J \ ..........

0 0 • Ill'
35

III iIi
AI rPM
40
I'PM
. •13C
45

50

b)
. . . .

6O

!I,ill
65

2 ~.6i ~:~l 70
0 3.10
22
3 75
4.|11

|" !
~-~ ~ C-12
i'i ~'~ ~'s 7's 71 7~
PPil
7'i 7'I ]J s'9
.80 .75
.. 813C
.,, 5 II1

Figure 2 1H, 13C heteronuclear shift correlation 2D NMR spectra of 12/3-hydroxy-24-epicastasterone (2).
a: 1H, 13C one-bond correlation spectrum, part of methyl region.
b: 1H, 13Cone-bond correlation spectrum, low-field region.
c: 1H, 13Cmultiple-bond correlation spectrum, part of 1H methyl region.

12fl-Hydroxy-24-epibrassinolide (5)
Elution of the silica gel column with ethyl acetate/methanol 9 : 1
v/v gave 19 mg starting material 4 and crude 5, which was purified
by preparative HPLC leading to 13 mg (12%) of pure 5 with mp
218 to 221 C (needles from MeOH); [a]26 + 25.7 ° (c = 0.219,
MeOH); IR (KBr), Vmax:3,400 (OH), 1,725 cm -i (lactone); El-
MS (m/z): 496 (M +, C28I~I4807, 1%), 395 (M+-CrHi3 O, 32%), 377
(395-H20, 62%), 341 (377-H20, 64%); FAB-MS (m/z): 497
(M + + 1, 48%), 479 (497-H20, 24%), 461 (479-HZO, 75%), 443
(461-H20, 33%), 377 (M+-C6Hi30-H2 O, 17%).
P e n t a a c e t y l derivative 6
Compound 5 (10 rag) was acetylated with Ac20/pyridine (each
0.5 ml) for 20 hours to give 6 mg (65%) 6 with mp 139 to 142 C
(needles from CHCLJn-hexane); [ct]2D s + 29.6 ° (C = 0.227,
MeOH); IR (CHCI3), Vmax:1,730, 1,720 cm- 1(acetyl and lactone);
EI-MS (m/z): 706 (M +, 1%), 647 (M +-CH3COO -, 2%), 586 (M +-
2AcOH, 4%), 564 (6%), 544 (10%), 527 (12%), 503 (25%), 461
(100%), 443 (21%), IH NMR (300 MHz, CDC13) (8): 0.833, 0.848,
0.938, and 0.973 (4 x d, J = 6.8 Hz, 6.7 Hz, 6.9 Hz, and 6.8
Hz, 21-H3, 26-H 3, 27-H 3, and 28-H3), 0.845 (s, 18-H3), 0.976 (s,
19-H3), 2.000, 2.030, 2.041, 2.077, and 2.133 (5 x s, 5 x OAc),
2.970 (dd, J = 12.3/4.5 Hz, 5a-H), 4.018 (dd, J --- 12.6/9.2 Hz,
7c~-H), 4.146 (br d, J = 12.6 Hz, 7fl-H), 4.606 (dd, J = 10.4/4.6
Hz, 12c~-H), 4.844 (ddd, J = 12.4/4.7/2.6 Hz, 2fl-H), 5.019 (dd,
J = 6.3/5.8 Hz, 23-H), 5.268 (dd, J = 6.4/1.3 Hz, 22-H), 5.359,
broad (s, 3/3-H).
Results and discussion
W e h a v e studied the microbial t r a n s f o r m a t i o n o f the
t w o naturally o c c u r r i n g b r a s s i n o s t e r o i d s 24-epicastas-
t e r o n e (1) and 24-epibrassinolide (4) b y the f u n g u s C.
echinulata I M E T 43918. This f u n g u s is usually k n o w n
for h y d r o x y l a t i o n s in 6fl-, l l a - , llfl-, 15fl-, a n d 17a-
positions. 12 U p o n e x p o s u r e o f I to the g r o w i n g culture
o f this f u n g u s after silica gel c h r o m a t o g r a p h y and
H P L C purification, the h i t h e r t o not d e s c r i b e d 12fl-
h y d r o x y l a t e d c o m p o u n d 2 w a s isolated in 14% yield.
T h e U V a b s o r p t i o n m a x i m u m at 282 n m a n d the nega-

322 Steroids, 1993, vol. 58, July

 Microbial transformation of brassinosteroids: Voigt et al.
tive cotton effect with a = - 6 shows the remained 6-
carbonyl function in 2. The E1 and FAB mass spectra
indicate the elemental composition of C28H480 6. The
~H N M R spectrum o f 2 shows in comparison with that
o f 1 an additional double doublet in the low-field region
(at 8 3.56). According to the found coupling constants
(10.9 Hz, 4.6 Hz), this methine proton has to be axial;
thus, the corresponding O H group must be equatorial.
The assignment of the 1H N M R signals of H3-18 and
H3-19 was done by the direct heteronuclear ill, ~3C
shift correlation two-dimensional (2D) N M R spectrum
(XHCORRD) using the known 13C chemical shifts of
C-18 and C-19 (Figure 2a). Furthermore, in the
X H C O R R D spectrum the proton signal at 3.56 ppm
correlates with the carbon signal at 77.8 ppm (Figure
2b). Assignment of this 13C signal was done using a
proton-detected multiple bond ill, ~3C shift correlation
2D N M R experiment (HMBC), which gives correlation
peaks for protons and carbons in the case of ~H-~3C
couplings via two or three bonds. The found correlation
o f the ~3C signal at 77.8 ppm with the proton signal of
the 18 methyl group in the H M B C spectrum (Figure
2c) indicates the assignment C-12 for this carbon signal.
In an H M B C spectrum, the IH 18-methyl signal is ex-
pected to give correlations with C-12, C-13, C-14, and
C-17. The additional O H group is bound to one of these
four carbons, but only C-12 gives a methine signal in
this case. Thus, the position o f the additional hydroxyl
group is 12/3. This is supported by the upfield shift of
C-18 (8 8.5; A8-3,2) in 2 compared with 1 (8 11.7), which
is a result of the y-gauche effect of 12/3-OH on methyl
group 18. Assignment of 13C N M R signals of 1 and 2
(Table 1) was done according to the strategy used for
brassinosteroids o f the lactone series.13 The introduc-
tion of the additional O H function is also confirmed
b y formation of the pentaacetyl derivative 3 with a
molecular ion at m/z 690 for the elemental composition
of C38H58Otl and the appearance of five acetyl singlets
in its IH N M R spectrum. The double doublet at 8 4.66
(J = 10.8/4.6 Hz) for the axial 12a-H confirms the 12/3
configuration of the newly introduced oxygen function.
In the same m a n n e r as described for 24-epicastaster-
one (1), 24-epibrassinolide (4) was also microbially
transformed to give after purification 12/3-hydroxy-24-
epibrassinolide (5) with a molecular ion at m/z 496 in
the EI-MS spectrum for C28H4807. Compound 5 was
characterized as its pentaacetyl derivative 6 (m/z 706),
with N M R data corresponding with the given structure.
The double doublet (J = 10.4/4.6 Hz) at 4.61 ppm in
the 1H N M R spectrum of 6 (H-12a) especially verifies
the presence of the 12/3-function.
F o r studies on structure-activity relationships, the
highly sensitive rice lamina inclination test ~4 using the
cultivar Koshirikari x5 was applied as the assay for the
new brassinosteroids 2 and 5. Preliminary experiments
showed that, similar to the 12fl-hydroxylated
(22S,23S)-homobrassinolide, 5'6 the activity o f 12/3-
hydroxy-24-epicastasterone (2) and 12/3-hydroxy-24-
epibrassinolide (5) is lowered to 10% at 0.1 ppm in
comparison with that o f the corresponding parent hor-
mones 1 and 4.
Acknowledgments
We thank Drs. Jiirgen Schmidt and G e r n o t Schneider
for the measurements o f mass spectra and H P L C , re-
spectively. This work was supported by the Ministry of
Research and Technology, Germany, and is gratefully
acknowledged.
References
1. AdamG, Marquardt V (1986). Brassinosteroids. Phytochemis-
try 25:1787-1799.
2. Hamada K (1986). Brassinolide in crop cultivation. FFTC
Book Ser 34:188-197.
3. Meudt WJ (1987). Chemical and biological aspects of brassi-
nolide. ACS Syrup Ser 325:53-75.
4. MarquardtV, Adam G (1991). Recent advances in brassino-
steroid research. Chem Plant Protect 7:103-139.
5. Cutler HG, Yokota T, Adam G (¢ds.) (1991). Brassinoste-
roids--chemistry, bioactivity, application. ACS Syrup Ser
474.
6. Adam G, Vorbrodt HM, H6rhold C, B6hme KH, D~inhardt
S, Porzel A, Zeigan D (1989). Procedure for production of
brassinosteroid analogs. DD 287 957 AS (German patent).
7. Vorbrodt HM, Adam G, Porzel A, H6rhold C, D~nhardt S,
B6hme KH ( 1991).Microbial degradation of 2t~,3t~-dihydroxy-
5a-cholestan-6-one by Mycobacterium vaccae. Steroids
56:586-588.
8. Ikekawa N, Nishiyama F, Fujimoto Y (1988). Identification
of 24-epibrassinolide in bee pollen of the broad bean Vicia
faba L. Chem Pharm Bull (Tokyo) 36:405-407.
9. Yokota T, Kim SK, Fukui Y, Takahashi N, Takeuchi Y,
Takematsu T (1987). Brassinosteroids and sterols from a green
alga, Hydrodictyon reticulatum: configuration at C-24. Phyto-
chemistry 26:503-506.
10. Thompson MH, Mandava NB, Flippen-Anderson JL, Worley
JF, Dutky SR, Robbins WE, Lusby W (1979). Synthesis of
brassinosteroids: new plant growth promoting steroids. J Org
Chem 44:5002-5004.
11. AnastasiaM, Allevi P, Ciuffreda P, Oleotti A (1985). A conve-
nient isomerization of 6-oxo-3t~,5-cyclo-5c~-steroidsto 6-oxo-
&2-5a-steroids. Steroids 45:561-564.
12. MahatoSB, Banerjee S, Podder S (1989). Steroid transforma-
tions by microorganisms--Ill. Phytochemistry 7,8:7--40.
13. Porzel A, Marquardt V, Adam G, Massiot G, Zeigan D (1992).
1H and 13C NMR analysis of brassinosteroids. Magn Reson
Chem 30:651-657.
14. WadaK, Marumo S, Abe H, Morishita T, Nakamura K, Uchi-
yama M, Mori K (1984). A rice lamina inclination test--a
microquantitative bioassay for brassinosteroids. Agric Biol
Chem 48:719-726.
15. ArimaM, Yokota T, Takahashi N (1984). Identification and
quantificationof brassinolide-related steroids in the insect gall
and healthy tissues of the chestnut plant. Phytochernistry
23:1587-1591.
Steroids, 1993, vol. 58, July 323