Nitrile hydratase from Rhodococcus erythropolis: Metabolization of steroidal compounds with a nitrile group

Article abstract of DOI:10.1016/S0039-128X(99)00028-8

The progestin dienogest (17α-cyanomethyl-17β-hydroxy-estra-4,9-dien-3-one) was metabolized by the nitrile hydratase-containing microorganism Rhodococcus erythropolis. An enzymatic hydrolysis of the nitrile group at the 17α-side chain was intended to obtain novel derivatives and to test them for progesterone receptor affinity. In contrast to the rapid enzymatic hydrolysis of nonsteroidal nitriles, the nitrile group of dienogest was cleaved very slowly. The dominant reaction was an aromatization of ring A. After prolonged fermentation, the 17α-acetamido derivatives of estradiol and of 9(11)-dehydroestradiol were formed. Three of the metabolites were also prepared synthetically. They were tested for hormonal activity by assessing their binding to progesterone and estrogen receptors in vitro. Neither the aromatized 17α-acetamido derivatives nor the dienogest derivative 17α-acetamido-17β-hydroxy-estra-4,9-dien-3-one, which was prepared synthetically only, exhibited affinity for the progesterone receptor. Copyright (C) 1999 Elsevier Science Inc.

Full text of DOI:10.1016/S0039-128X(99)00028-8

Steroids 64 (1999) 535–540
Nitrile hydratase from Rhodococcus erythropolis: Metabolization of
steroidal compounds with a nitrile group
Gu¨nter Kaufmanna,*, Horst Dautzenbergb,1, Harry Henkela, Gerd Mu¨ller,a Thomas Scha¨fer,b,2
Bernd Undeutscha, Michael Oettela
aDivision of Research and Development, Jenapharm GmbH & Co. KG, D-07745 Jena, Germany
bResearch Group Polyelectrolyte Complexes, University Potsdam, D-14513 Teltow, Germany
Received 24 November 1998; received in revised form 5 February 1999; accepted 10 February 1999
Abstract
The progestin dienogest (17␣-cyanomethyl-17␤-hydroxy-estra-4,9-dien-3-one) was metabolized by the nitrile hydratase-containing
microorganism Rhodococcus erythropolis. An enzymatic hydrolysis of the nitrile group at the 17␣-side chain was intended to obtain novel
derivatives and to test them for progesterone receptor affinity. In contrast to the rapid enzymatic hydrolysis of nonsteroidal nitriles, the nitrile
group of dienogest was cleaved very slowly. The dominant reaction was an aromatization of ring A. After prolonged fermentation, the
17␣-acetamido derivatives of estradiol and of 9(11)-dehydroestradiol were formed. Three of the metabolites were also prepared synthet-
ically. They were tested for hormonal activity by assessing their binding to progesterone and estrogen receptors in vitro. Neither the
aromatized 17␣-acetamido derivatives nor the dienogest derivative 17␣-acetamido-17␤-hydroxy-estra-4,9-dien-3-one, which was prepared
synthetically only, exhibited affinity for the progesterone receptor. © 1999 Elsevier Science Inc. All rights reserved.
Keywords: Progestin; Microbial transformation; Steroid aromatization; Nitrile hydratase; Progesterone receptor
1. Introduction
tal dose of 6.3 mg of DNG administered over 14 days
resulted in a complete secretory transformation of the en-
dometrium. In this assay, the oral progestational activity of
DNG was 2 and 7 times that of chlormadinone acetate and
medroxyprogesterone acetate, respectively [2]. On the other
hand, DNG exhibits only moderate binding affinity for the
progesterone receptor, about 10% of that of progesterone
[3,4]. This discrepancy has induced speculations that dieno-
gest may be a prodrug, a metabolite of which may be
responsible for its high biologic efficacy. Such is the case
with desogestrel, for example. In this regard, a number of
metabolites and chemically derived compounds were tested
but found to be less active than dienogest [4]. Until now, no
derivatives with a metabolized 17␣-cyanomethyl side chain
have been included in such studies. Thus, it was interesting
to find possible means of obtaining such derivatives, e.g. by
enzymatic hydrolysis of the nitrile function, and to test them
for progesterone receptor affinity.
Dienogest (17␣-cyanomethyl-17␤-hydroxy-estra-4,9-dien-
3-one) (DNG) is a progestin derived from nortestosterone
but characterized by a cyanomethyl group in position 17␣
instead of the common ethinyl substituent. It is used as a
progestin in hormonal contraception and is under develop-
ment for further indications, such as endometriosis and
hormone replacement therapy. DNG shows a comparatively
strong progestational transformation effect on the endome-
trium. In estrogen-primed female rabbits (Clauberg/
McPhail assay), it is about five times more active than
levonorgestrel after oral administration [1]. Using the so-
called Kaufmann assay, the transformation dose in estro-
gen-primed postmenopausal women was estimated. The to-
* Corresponding author. Jenapharm GmbH & Co KG, Otto-Schott-Str.
15, D-07745 Jena, Germany.
The broad variety of microbial reactions available opens
the possibility of performing difficult reactions and prepar-
ing the corresponding products. Microorganisms containing
nitrile hydratase can transform a variety of aliphatic and
aromatic nitriles to the corresponding amides [5,6] and
1 Dr. Horst Dautzenberg, head of the Research Group Polyelectrolyte
Complexes, University Potsdam, died in August 1998 while this paper was
in preparation.
2 Present address: Morus-Verlag GmbH, Go¨tzstr. 65, D-12099 Berlin,
Germany.
0039-128X/99/$ – see front matter © 1999 Elsevier Science Inc. All rights reserved.
PII: S0039-128X(99)00028-8

536
G. Kaufmann et al. / Steroids 64 (1999) 535–540
further to the corresponding acids through the use of ami-
dase. Alternatively, microorganisms containing nitrilase can
hydrolyze nitriles directly to acids and ammonia [7,8]. Here,
a microbial transformation of DNG to the acetamido com-
pound was tried with Rhodococcus erythropolis. Because
the fermentation was only partially successful, the potential
metabolites were also prepared synthetically and tested for
their hormonal activity in vitro.
MgSO4⅐7H2O, 3 g yeast extract, 5 g alanine, 0.03 g CoSO4,
and 0.5 g acetonitrile as an inductor. The cells were stored
at 5°C in 67 mM phosphate buffer, pH 7.8 (Soerensen).
Enzyme activity was checked using the substrate 3-cyano-
pyridine. No remarkable loss of activity was found during
storage for up to two months.
The steroid fermentations were performed with cell sus-
pensions in phosphate buffer containing 20% dimethylsulf-
oxide (DMSO) or 2% acetone by shaking at 25°C. These
solvent concentrations had been shown to only minimally
influence the enzymatic hydrolysis of the reference sub-
stance 3-cyanopyridine to nicotinamide. Steroid concentra-
tions of 0.06–0.5 mM and cell concentrations of 1.7–7.4
mg/ml (calculated as dry weight) were used.
In order to monitor the fermentations directly, samples
were taken at suitable intervals, mixed with the same vol-
ume of methanol, centrifuged, and 20 ␮l of the resulting
solution were subjected to HPLC without further purifica-
tion. HPLC conditions: column C8 (Nowa-Pak, 4 ␮m, 4.6 ϫ
250 mm), 37°C, eluent methanol/water (70:30 v/v); UV
detection at 275 nm. TR of DNG: 3.8 min.
2. Experimental
2.1. Steroids
Dienogest (STS 557) is a product of Jenapharm. The
aromatic steroids STS 433 (later: J 540), STS 528, and J
1201 were synthesized by Ponsold and coworkers [9–11]
and by Jenapharm, respectively.
2.2. Synthesis of the 17␣-acetamide compounds
2.2.1. J 1113 (I)
For preparative isolation of the metabolites, fermenta-
tions were extracted with ethyl acetate and subjected to
preparative thin-layer chromatography (TLC).
Sodium hydroxide solution (6 M, 12 ml) and tetrabu-
tylammonium chloride (0.9 g, 2.3 mmol) were added to a
solution of 17␣-cyanomethyl-17␤-hydroxy-estra-4,9-dien-
3-one (2.7 g, 8.7 mmol) in methylene chloride (40 ml).
Hydrogen peroxide (30%, 15 ml) was added dropwise with
vigorous stirring at room temperature to the two-phase sys-
tem. The two phases were separated after 3 h of stirring at
room temperature. The organic phase was washed with
sodium hydrogensulphite solution and water, dried over
Na2SO4, and evaporated. Crystallization from ethanol/water
gave 17␣-acetamido-17␤-hydroxy-estra-4,9-dien-3-one (I),
m.p. 233–240°C (87% yield).
2.4. Analytics
TLC was carried out on silica gel GF254 (Merck) pre-
washed with chloroform/methanol (1:1 v/v) or on TLC
aluminium sheets of silica gel 60 F254 (Silufol, Merck)
using the solvent systems given in Table 1.
High-performance liquid chromatography (HPLC) char-
acterization of isolated DNG metabolites was done with a
Shimadzu CBM-10A equipped with an auto injector (SIL-
10A), a column oven (CTO-10A), and an UV diode-array
detector (SPD-M10A VP), using a hypersil ODS column, 5
␮m (250 ϫ 4.6 mm) at 35°C. For further conditions, see
Table 1.
NMR Spectra were recorded on a VARIAN Gemini 300
at 300 MHz and 75.4 MHz. DMSO-d6 was used as solvent.
Chemical shifts were reported as ␦-values in ppm downfield
from the internal standard tetramethylsilane; J were given in
Hz.
2.2.2. J 1358 (formula see IIIa)
17␣-Acetamido-17␤-hydroxy-estra-4,9-dien-3-one (0.5
g, 1.5 mmol) was suspended in methylene chloride (20 ml).
3-Chlorperbenzoic acid (420 mg, 2.4 mmol) was added to
the suspension at room temperature. Sodium hydrogensul-
phite solution was added to the reaction mixture after 4 h
and stirred again. The organic phase was washed with di-
luted sodium hydroxide solution and water, dried, and evap-
orated. The residue obtained was dissolved in methanol (10
ml). After addition of sodium hydroxide solution (2 M, 6
ml), the reaction mixture was stirred for 2 h at room tem-
perature. The 17␣-acetamido-estra-1,3,5(10),9(11)-tet-
raene-3,17␤-diol was precipitated by addition of HCl (6 M,
2.5 ml). The compound was purified by column chromatog-
raphy and by crystallization from tert-butylmethyl ether/
methylene chloride. Melting point (m.p.) 282–287°C (75%
yield).
High resolution mass spectra were recorded on a Joel
JMS-D 100.
2.5. Receptor affinity
Binding affinities for the progesterone receptor (PR) and
estrogen receptor (ER) were measured by competitive bind-
ing of the compounds of interest together with the tracers
[6,7-3H]ORG 2058 (5 nM, for PR) or [6,7-3H]Estradiol (3
nM, for ER) to receptors in rabbit uterus cytosol (rabbits for
PR primed with estradiol, immature rabbits for ER). Buffer:
Tris/HCl, 20 mM, pH 7.4, containing 1 mM EDTA, 2 mM
DTT, and 250 mM sucrose. Incubations were carried out for
2.3. Microorganism and fermentations
Rhodococcus erythropolis FZB 53 was grown at 30°C on
medium containing 0.5 g KH2PO4, 0.5 g K2HPO4, 0.5 g

G. Kaufmann et al. / Steroids 64 (1999) 535–540
537
Table 1
Data of DNG metabolites
Metabolite
IIa
IIb
IIIa
IIIb
(see formula, Fig. 1)
TLC
Rf in system Aa
Rf in system Ba
HPLC: tR (methodb)
0.42
0.56
9.5 min (C)
280 nm
0.42
0.56
10.1 min (C)
262; 299 nm
0.s21
0.22
14.0 min (D)
280 nm
0.21
0.22
14.6 min (D)
262; 299 nm
␭
(nm)
max
HRMS
Mϩ Found
311.1882
311.1885
309.1714
309.1729
329.1991
329.1991
327.1856
327.1834
C20H25NO3
Calculated
Formula
C
20H25NO2
C
20H23NO2
C
20H27NO3
1H-NMR
8.98 (s, 3-OH),
9.26 (s, 3-OH),
8.99 (s, 3-OH),
9.26 (s, 3-OH),
7.03 (d, 8.7, H-1),
6.51 (dd, 8.1, 2.2, H-2),
6.43 (d, 2.2, H-4),
4.94 (s, 17-OH),
0.80 (s, H-18)
7.42 (d, 8.7, H-1),
6.53 (dd, 8.7, 2.2, H-2),
6.44 (d, 2.2, H-4),
6.05 (m, H-11),
5.06 (s, 17-OH),
0.80 (s, H-18)
7.04 (d, 8.4, H-1),
6.50 (dd, 8.4, 2.3, H-2),
6.43 (d, 2.3, H-4),
5.63 (s, 17-OH),
0.81 (s, H-18)
7.42 (d, 8.7, H-1),
6.53 (dd, 8.7, 2.3, H-2),
6.44 (d, 2.3, H-4),
6.07 (m, H-11),
5.67 (s, 17-OH),
0.81 (s, H-18)
13C-NMR
154.8 (3), 137.0 (5),
130.1 (10), 125.9 (1),
119.9 (-CN), 114.8 (4),
112.6 (2), 80.5 (17),
48.6 (14), 46.1 (13),
42.9 (9), 39.3 (8),
14.1 (18)
157.3 (3), 138.7 (5),
136.5 (10), 127.4 (9),
126.2 (1), 120.3 (-CN),
117.2 (11), 115.9 (4),
114.7 (2), 82.2 (17),
14.9 (18)
175.7 (-CONH2), 154.8 (3),
137.0 (5), 130.2 (10),
125.9 (1), 114.8 (4),
112.6 (2), 81.2 (17),
48.6 (14), 46.0 (13),
43.3 (9), 22.8 (15),
13.9 (18)
175.6 (-CONH2), 155.9 (3),
136.9 (5), 134.2 (10),
125.2 (9), 124.8 (1),
116.5 (4), 114.7 (2),
113.6 (11), 80.9 (17),
46.3 (14), 44.4 (13),
39.5 (8), 23.4 (15),
14.0 (18)
a TLC system A: chloroform/methanol (92:8 v/v); TLC B: benzene/ethyl acetate/ethanol/acetic acid (50:42:8:0.1 v/v).
b HPLC method C: eluent acetonitrile/water (40:60 v/v), 1 ml/min. Method D: eluent methanol/sodium phosphate buffer (25 mM, pH 7.5) (55:45 v/v); 0.5
ml/min.
18 h at 0–4°C. Separation of free and bound steroid was
NMR spectra clearly showed that it consisted of a mixture
of the estra-1,3,5(10)-triene and the estra-1,3,5(10),9(11)-
tetraene compounds IIa and IIb. There were also two me-
tabolites in product III, IIIa, and IIIb, which were shown
by 13C-NMR to have lost the cyano group and to contain a
carbonyl group instead. Mass spectra showed the Mϩ peaks
of the estratriene and estratetraene derivatives. Resolution
of both substance pairs was achieved by HPLC. The struc-
tures were proved by UV spectra and by comparison with
the synthetic steroids STS 433 (later code J 540), J 1201,
and J 1358. The data of the four metabolites are given in
Table 1; for formulas, see Fig. 1.
A trace of an even more polar product was found also in
prolonged fermentations. The expected carboxylic acid de-
rivative with an aromatized ring A was extracted from ether
solution into aqueous sodium carbonate or sodium bicar-
bonate. However, the amount was insufficient for charac-
terization.
The velocity of the metabolization of the nitrile group of
DNG to the acetamido compounds was estimated from
samples taken at suitable intervals from different fermenta-
tions. Furthermore, the aromatic compound STS 433 (iden-
tical with IIa) was compared to DNG with respect to me-
tabolization velocity. Both compounds, when compared to
the reference 3-cyanopyridine, showed that the steroidal
17␣-CH2CN group was metabolized very slowly, differing
achieved by charcoal/dextran (1%/0.1%) treatment.
The molar IC50s were evaluated from a series of concen-
trations, and the relative molar binding affinities (RBA)
were calculated as the quotient of the IC50 of the reference
compound and that of each test substance (ϫ100%).
2. Results
Although Rhodococcus erythropolis cells were able to
enzymatically hydrolyze the reference compound 3-cyano-
pyridine to nicotinamide within minutes, the steroid sub-
strate was metabolized relatively slowly. No metabolite
with the intact estra-4,9-diene system was found in a num-
ber of incubations with different concentrations of dienogest
and cells. Instead of the intended hydrolysis of the nitrile
group, an aromatization of ring A was noticed. After pro-
longed fermentation of DNG with Rhodococcus erythropo-
lis, a second, more polar type of metabolite, was detected.
Both of these products were isolated by preparative TLC
from the culture extract of a 27-day fermentation, together
with a residue of about 5% DNG.
The less polar product II corresponded to the known
estra-1,3,5(10),9(11)-tetraene compound with
CH2CN group [11]. Although it was crystallized, the H-
a 17␣-
1

538
G. Kaufmann et al. / Steroids 64 (1999) 535–540
Fig. 1. Metabolites of dienogest.
in reactivity from cyanopyridine by several orders of mag-
nitude (see Table 2).
matic nitrile compounds [5], and the intermediate amides
can also be isolated. Other species of Rhodococcus have
been used for the hydrolysis of nitriles to drugs of phar-
maceutical interest, like pyrazinoic acid [7] and (S)-
naproxen [8], which was obtained by stereoselective ni-
trile hydrolysis.
In search for possible hormonal activities of the me-
tabolites, receptor binding affinities for the progesterone
and the estrogen receptors were determined using the
corresponding synthetic compounds. The results are
listed in Table 3. Estrogen receptor affinity was found for
the ring A aromatized derivatives of DNG, corresponding
to IIa and IIb. Further, some progesterone receptor af-
finity was found for those derivatives that were charac-
terized by the unmetabolized 17␣-CH2CN group. The
17␣-acetamido compounds IIIa, IIIb, and I were nearly
or completely inactive at the receptor level. Furthermore,
compound I exhibited no progestin activity when tested
for pregnancy maintenance s.c. in mice (Elger W, 1998,
unpublished observations).
When the steroidal nitrile compound dienogest was used
as substrate, the A-ring aromatization was found to be the
dominant reaction instead of the nitrile hydratation reaction.
Although unwanted, this result correlates with aromatiza-
tion reactions of other 19-nor componds. The species
Rhodococcus rhodochrous was able to transform 3-keto-
desogestrel (13-ethyl-17-hydroxy-11-methylene-18,19-di-
nor-17␣-pregn-4-en-20-yn-3-one) to the corresponding A-
ring aromatized compound [12]. DNG itself was
metabolized to aromatized metabolites by Mycobacterium
smegmatis [13]. Interestingly, the aromatized metabolite IIb
was also formed by mammalian and human metabolism
[14,15]. The estrogenic effect of this metabolite was one
order of magnitude lower than that of estriol s.c. in mice,
that of IIa even lower [4]. Nevertheless, the antiprogesta-
tional activity of DNG in some animal experiments may be
mimicked by the aromatized metabolite [16].
3. Discussion
Normally, the strain Rhodococcus erythropolis FZB
53 can rapidly hydrolyze a variety of aliphatic or aro-
The desired microbial transformation of the nitrile
group of DNG, in contrast to that of nonsteroidal nitriles,
e.g. cyanopyridine, was seen only in prolonged fermen-
tations, where ring A of the metabolites had already been
aromatized (see Fig. 1). Thus, the 17␣-acetamido analog
of dienogest was only synthetically prepared (J 1113) and
included in the receptor binding tests, see below. The
17␣-cyanomethyl group of DNG seems to be relatively
stable against the microbial nitrile hydratase, an enzyme
of low substrate specificity. The stability of the 17␣-
cyanomethyl group was also seen in the metabolization
of DNG in humans. Of the variety of metabolites pro-
duced, only one was found that had lost the nitrogen of
Table 2
Comparison of metabolization rates in fermentations with Rhodococcus
rhodochrous at 25°C
Substrate
Substrate
Biomass
Time for 50%
metabolization by
Concentr. Conc.a
(mmol/l)
(mg/ml)
Aromatization Nitrile
Hydratase
DNG
DNG
0.1
0.5
7.4
1.7
7.4
2.2
7.4
1.5 h
8 h
48 h
264 h
24 h
24 min
15 min
STS 433 (ϭ IIa) 0.06
Cyanopyridine
Cyanopyridine
480
1260
aBiomass as dry weight.

G. Kaufmann et al. / Steroids 64 (1999) 535–540
539
Table 3
Receptor binding affinities of dienogest derivatives
Steroid (Code)
Structure
Relative molar binding affinity (RBA) [%]
(see Fig. 1)
Mean Ϯ SD (n)
Progesterone receptor
Estrogen receptor
(progesterone ϭ 100)
(estradiol ϭ 100)
J 540
J 1201
IIa
IIb
3.4 Ϯ 0.2 (4)
1.0 Ϯ 0.1 (4)
Ͻ0.1 (3)
Ͻ0.1 (4)
Ͻ0.2 (5)
30 Ϯ 3 (5)
28 Ϯ 4 (5)
0.28 Ϯ 0.04 (5)
0.12 Ϯ 0.01 (4)
Ͻ0.01 (3)
IIIa, in mixture with IIIb (c. 1:1)
J 1358
J 1113 (not found as metabolite)
For comparison: Dienogest
IIIb
I
DNG
10.5 Ϯ 2.5 (10)
0.65 Ϯ 0.25 (4)
Ͻ0.1 (5)
Ͻ0.01 (3)
STS 528 (17␣-hydroxymethyl-17␤-hydroxyestr-4-en-3-one)
the side chain. It was tentatively identified as a lactone
resulting from complete hydrolysis of the 17␣-cyanom-
ethyl group to the carboxyl acid and from 15- (or 16-)
hydroxylation [15]. This compound is analogous to a
lactone metabolite of 17␣-cyanomethylestradiol 3-methyl
ether produced in the rat [17].
References
[1] Hu¨bner M, Ponsold K, Oettel M, Freund R. Eine neue Klasse hoch-
wirksamer Gestagene: 17␣-CH2X-substituierte Gona-4,9(10)-diene.
Arzneim-Forsch-Drug Res 1980;20:401–4.
[2] Oettel M, Carol W, Kaufmann G, Moore C, Ro¨mer W, Klinger B,
Schneider B, Schro¨der J, Sobek L, Walter F, Zimmermann H. A
19-norprogestin without a 17␣-ethinyl group. II: Dienogest from a
pharmacodynamic point of view. Drugs of Today 1995;31:517–36.
[3] Juchem M, Schaffrath M, Pollow K, Hoffmann G, Kaufmann G.
Dienogest: Bindungsstudien an verschiedenen Rezeptor- und Serum-
proteinen. In: Teichmann AT, editor. Dienogest. Pra¨klinik und Klinik
eines neuen Gestagens, 2th Ed. Berlin: De Gruyter, 1995. pp. 119–33.
[4] Oettel M, Kaufmann G, Kurischko A. Das endokrinologische Profil
von Metaboliten des Gestagens Dienogest. Pharmazie 1993;48:541–5.
[5] Hu¨bner B, Bo¨hme M, Lebentrau B, Mu¨ller M, Ha¨fner B, Juhr E, FZ
Biotechnik Berlin GmbH. Mikroorganismus und Verfahren zur mik-
robiellen Herstellung eines Nitrilhydratase-/Amidase-Enzymkom-
plexes. Germ Patent DE 43 13 649 C1. 1993 April 21.
[6] Nawaz MS, Franklin W, Campbell WL, Heinze TM, Cerniglia CE.
Metabolism of acrylonitrile by Klebsiella pneumoniae. Arch Micro-
biol 1991;156:231–8.
[7] Kobayashi M, Yanaka N, Nagasawa T, Yamada H. Nitrilase-cata-
lyzed production of pyrazinoic acid, an antimycobacterial agent, from
cyanopyrazine by resting cells of Rhodococcus rhodochrous J1. J
Antibiot Tokyo 1990;43:1316–20.
[8] Effenberger F, Bo¨hme J. Enzyme-catalysed enantioselective hydro-
lysis of racemic naproxen nitrile. Bioorg Med Chem 1994;2:715–21.
[9] Menzenbach B, Hu¨bner M, Ponsold K. Untersuchungen zur Brom-
ierung/Dehydrobromierung von 17␣-Cyanomethyl-17␤-hydroxy-
o¨str-5(10)-en-3-on. J Prakt Chemie 1984;326:893–8.
[10] Menzenbach B, Hu¨bner M, Sahm R, Ponsold K. Synthese potentieller
Metaboliten der STS 557 (Dienogest). Teil 3: 17␣-Hydroxymethyl-
17␤-hydroxyestra-4,9-dien-3-on. Pharmazie 1984;39:496–7.
[11] Hu¨bner M, Menzenbach B, Ro¨hrig H, Ponsold K. Synthese potenti-
eller Metaboliten der STS 557 (Dienogest). Teil 2: 17␣-Cyanometh-
ylestra-1,3,5(10),9(11)-tetraen-3,17␤-diol. Pharmazie 1984;39:496.
[12] Groh H, Scho¨n R, Ritzau M, Kasch H, Undisz K, Hobe G. Prepara-
tion of 3-ketodesogestrel metabolites by microbial transformation and
chemical synthesis. Steroids 1997;62:437–43.
The metabolization of the side chain of DNG to 17␣-
acetamido compounds, as found here, is a possible, but
probably not important pathway. It does not result in hor-
monally active metabolites which would exhibit high recep-
tor binding affinities. On the contrary, the 17␣-acetamido
compounds exhibit negligible receptor binding, and even
the synthetically prepared 17␣-acetamido analog of dieno-
gest J 1113 (compound I), is inactive at the hormone recep-
tor level and in mice. This finding is not surprising. The
mammalian progesterone receptor preferentially binds nort-
estosterone derivatives bearing hydrophobic 17␣-substitu-
ents. Thus, 17␣-methyl and ethinyl [18], even 17␣-chlo-
romethyl [19], or 17␣-iodovinyl substituents [20] were
tolerated. More polar 17␣-substituents were barely toler-
ated. Correspondingly, the 17␣-hydroxymethyl compound
STS 528, another potential DNG metabolite tentatively
characterized in rabbit urine [14], exhibited only marginal
receptor affinity.
Thus, the high efficacy of dienogest as a progestin in
vivo cannot be explained by hormonal activity of metabo-
lites with a hydrolyzed 17␣-cyanomethyl side chain. Rather,
it may be explained by the high levels of free, meaning
active, DNG in plasma [16], by the influence of the conju-
gated 9(10) double bond inhibiting some metabolization
pathways [21], and by the pharmacokinetic parameters of
this progestin.
[13] Hobe G, Scho¨n R, Ho¨rhold C, Hu¨bner M, Schade W, Schubert K.
Microbial transformation of 17␣-cyanomethyl-17-hydroxy-4,9-estra-
dien-3-one (STS 557) and 17␣-cyanomethyl-19-nortestosterone by
Mycobacterium smegmatis. Steroids 1982;39:399–409.
[14] Hobe G, Scho¨n R, Wehrberger K, Schade W, Ritter P, Hillesheim
HG. Studies on the biotransformation of the progestagen dienogest in
the rabbit. Exp Clin Endocrinol 1989;94:203–10.
Acknowledgments
We are most grateful to Dr Barbara Ha¨fner, Research
Centre of Biotechnics GmbH, Berlin, for growing and pro-
viding us with the Rhodococcus erythropolis cells.

540
G. Kaufmann et al. / Steroids 64 (1999) 535–540
[15] Hobe G, Scho¨n R, Schade W, Wehrberger K, Knappe R, Claussen C,
Carol W, Klinger G. Zur Biotransformation von Dienogest beim
Menschen. In: Teichmann AT, editor. Dienogest. Pra¨klinik und
Klinik eines neuen Gestagens, 2th Ed. Berlin: De Gruyter, 1995. pp.
105–17.
[16] Oettel M, Bervoas-Martin B, Elger W, Golbs S, Hobe G, Kaufmann
G, Mathieu M, Moore C, Puri C, Ritter P, Reddersen G, Scho¨n R,
Strauch G, Zimmermann H. A 19-norprogestin without a 17␣-ethinyl
group. I: Dienogest from a pharmacokinetic point of view. Drugs
Today 1995;31:499–516.
[18] Delettre´ J, Mornon JP, Lepicard G, Ojasoo T, Raynaud JP. Steroid
flexibility and receptor specificity. J Steroid Biochem 1980;13:45–59.
[19] Bohl M, Simon Z, Vlad A, Kaufmann G, Ponsold K. MTD calcula-
tions on quantitative structure-activity relationships of steroids bind-
ing to the progesterone receptor. Z. Naturforsch 1987;42c:935–40.
[20] Ali H, Rousseau J, van Lier JE. Synthesis of (17␣,20E/Z)iodovinyl
testosterone and 19-nortestosterone derivatives as potential radioli-
gands for androgen and progesterone receptors. J Steroid Biochem
Mol Biol 1994;49:15–29.
[21] Schubert K, Schumann G, Kaufmann G. Influence of a 9-double bond
on stereospecific microbial 4,5-reductions. J Steroid Biochem 1983;
18:75–80.
[17] Hobe G, Scho¨n R, Wehrberger K. Species differences in biotransfor-
mation of 17␣-cyanomethylestradiol 3-methyl ether. Steroids 1980;
36:131–47.