Novel steroid mimics directed towards the estradiol skeleton

Article abstract of DOI:10.1016/j.tetlet.2005.10.005

A series of non-symmetrical tri- and tetra-substituted ureas have been prepared to mimic the rigid tetracyclic core of estradiol. Tetra-substituted ureas were prepared by a five-step protocol, involving activation and displacement of a carbonyldiimidazole

Full text of DOI:10.1016/j.tetlet.2005.10.005

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 8521–8524
Novel steroid mimics directed towards the estradiol skeleton
Eve Bridgeman, Julie L. Cavill, Daniel J. Schofield, Derek S. Wilkins and
Nicholas C. O. Tomkinson*
School of Chemistry, Main Building, Cardiff University, Park Place, Cardiff CF10 3AT, UK
Received 2 September 2005; revised 29 September 2005; accepted 4 October 2005
Available online 19 October 2005
Abstract—A series of non-symmetrical tri- and tetra-substituted ureas have been prepared to mimic the rigid tetracyclic core of estra-
diol. Tetra-substituted ureas were prepared by a five-step protocol, involving activation and displacement of a carbonyldiimidazole
adduct in 48–67% overall yield. This method was unsuccessful for tri-substituted ureas, which were prepared in 42–74% yields by an
alternative four-step method, which included a one-pot 4-nitrophenyl carbamate formation/displacement sequence.
Ó 2005 Elsevier Ltd. All rights reserved.
Since the first complete structural characterisation of
cholesterol in 1932, steroids have distinguished them-
selves as a class of molecule with an unsurpassed chem-
ical and biological significance.1 The subsequent
discovery that the cyclopentanoperhydrophenanthrene
core of cholesterol was shared by a whole family of
steroid hormones including the estrogens, androgens,
progestins, mineralocorticoids and glucocorticoids as
well as the structurally related vitamin D and the bile
acids has since established the extreme chemical, clinical
and scientific significance of this ring system, which
plays a pivotal role in some of the most fundamental
biological signalling pathways known.2
with relatively few functional groups, their stereochemi-
cal and architectural complexity renders them an
exceedingly difficult target for chemical synthesis, which
has stretched the minds and ingenuity of numerous sci-
entists over the years.3 However, from a medicinal point
of view, the synthesis of analogues, specifically targeted
to probe the structure activity relationship (SAR) of the
structure is a slow laborious process. With frequently up
to eight contiguous chiral centres it is not surprising that
relatively few synthetic drugs have been developed that
target this class of receptor. Regarding the steroid
nucleus as a rigid molecular scaffold we have initiated
a programme to develop a series of steroid mimics aimed
at probing nuclear receptor space, targeting known
nuclear receptors such as the estrogen receptor,4 the
androgen receptor5 and the progesterone receptor,6 with
the ultimate goal of finding new ligands and hence the
biological significance of some of the orphan nuclear
receptors that have been discovered more recently.7 In
this letter we report the preparation of a series of
Steroids elicit their diverse biological actions via differ-
ent functionality located around the periphery of their
rigid tetracyclic core. For example, estradiol 1 can be
looked upon as a phenolic and a secondary alcohol that
is spatially fixed by a central molecular scaffold.
Although steroids are chemically quite simple molecules,
NMe2
O
OH
O
OH
H
D
A
H
H
HO
HO
HO
Estradiol 1
2
Diethylstilbestrol 3
Tamoxifen 4
*
Corresponding author. Tel.: +44 029 20874068; fax: +44 029 20874030; e-mail: tomkinsonnc@cardiff.ac.uk
0040-4039/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.10.005

8522
E. Bridgeman et al. / Tetrahedron Letters 46 (2005) 8521–8524
OH
OH
O
OH
OH
non-symmetrical tri- and tetra-substituted ureas aimed
at mimicking molecules with estrogenic properties.
OH
HN
(-)-11
HN
HN
12
HN
( )-11
HN
13
It has been reported that the simplified steroidal struc-
ture 2, which lacks the traditional B and C rings of
estrone, shows very little activity as a lipid catabolic
agent.8 This is possibly due to the loss of rotational
degrees of freedom of the molecule on complexation
with the receptor, rendering the interaction energetically
unfavourable. This problem has been overcome by the
introduction of unsaturation across the linker, for exam-
ple, diethylstilbestrol 3 is a good agonist of the estrogen
receptor.9 Although 3 formally has two C–C single
bonds within its structure, these are subject to restricted
rotation due to the conjugation of the two aromatic
rings, rendering the scaffold more rigid. While diethyl-
stilbestrol is not suitable for use in the clinic due to its
associated toxicity, the stilbene scaffold has been
exploited in the development of Tamoxifen 4, the drug
of choice for the treatment of breast cancer.10 By consid-
eration of the structure of estradiol, four of the five chi-
ral centres could be removed, and hence the structure
greatly simplified, by preparation of urea 10. We
believed that restricted rotation about the two amide
bonds of the urea would render the central portion of
the molecule less conformationally flexible than the fully
saturated system and thus favour binding to the
receptor.
(+)-11
Figure 1. D-ring mimics.
With each of the adducts derived from the pyrrolidines
11 there are two low-energy conformations in which
the five-membered ring could sit. The first (A), which
would map onto the steroidal skeleton as desired and
the second (B), which would put the D-ring substituent
in the pseudo-16-a position of the steroid (Fig. 2). To
circumvent this problem, we also prepared (3R),(4R)-
pyrrolidinediol 13 in three steps from (À)-tartaric acid,15
by a modified route to that reported previously.16 The
inherent C2 symmetry of 13 means that in both low-
energy conformations of the five-membered ring, the
molecular structure will be such that the hydroxyl func-
tionalities can be mapped directly onto the estrogen
skeleton.
Treatment of the activated urea 8 with (3S)-pyrrolidinol
(À)-11 in dichloromethane at room temperature under
basic conditions, gave the protected estradiol mimic 9
in 81% yield. Removal of the protecting group to reveal
the phenol functionality was carried out in the presence
of tetrabutylammonium fluoride to give the desired urea
10 as a colourless solid. The other targets derived from
the D-ring mimics 11–13 were prepared in a similar
manner to give a small targeted library of compounds
for biological testing (Table 1).
Starting from commercially available 4-(methylamino)
phenol sulfate 5, the phenol functionality was protected
as its tert-butyldimethylsilyl ether 6 under standard con-
ditions (Scheme 1).11 Treatment of this protected aniline
with carbonyldiimidazole gave an intermediate urea 7,
which was activated by reaction with methyl iodide in
acetonitrile, in 84% for the two steps.12 At this point,
we were able to introduce diversity into our synthetic
sequence by reacting 8 with a variety of proposed D-ring
analogues (Fig. 1). These included (3S)-pyrrolidinol
(À)-11, which was prepared in three steps from
(S)-4-amino-2-hydroxybutyric acid,13 (3R)-pyrrolidinol
(+)-11 and 3-pyrrolidinol ( )-11, both of which were
commercially available, and 3-pyrrolidinone 12, which
was prepared in four steps from glycine ethyl ester.14
OH
OH
N
N
N
N
O
O
HO
HO
(A)
(B)
Figure 2. Possible D-ring conformations of estradiol mimic 10.
N
N
N
N
NaH, TBDMSCl,
THF, -78oC
97%
CDI, THF
r.t.
N
H
H
O
.
0.5H2SO4
TBDMSO
TBDMSO
HO
91%
5
6
7
MeI, MeCN
r.t.
92%
OH
OH
N
I
TBAF, THF
r.t.
(-)-11, NEt3
CH2Cl2, r.t.
81%
N
N
N
N
N
N
O
O
O
TBDMSO
89%
HO
TBDMSO
10
9
8
Scheme 1. Preparation of tetra-substituted ureas.

E. Bridgeman et al. / Tetrahedron Letters 46 (2005) 8521–8524
8523
This proved to be a general method for the preparation
of a variety of non-symmetrical tetra-substituted ureas,
however, for the synthesis of tri-substituted ureas (e.g.,
16) an alternative protocol was adopted (Scheme 2).
Protection of 4-nitrophenol 14 followed by reduction
of the nitro group gave the protected aniline 15. Treat-
ment of 15 with 4-nitrophenyl chloroformate in the pres-
ence of 1 equiv of triethylamine for 1 h followed by the
addition of (3S)-pyrrolidinol (À)-11 and a further equiv-
alent of base led directly to the protected mimic. This
one-pot procedure for the synthesis of the urea motif
presumably proceeds via an intermediate isocyanate
and proved ineffective in the tetra-substituted series,
even under forcing conditions. It proved necessary to
use this alternative protocol for the synthesis of the
tri-substituted mimic 16 as use of carbonyldiimidazole
in the coupling led directly to the symmetrical urea 17
as the only isolable product (Fig. 3). Deprotection of
the protected phenol in the same manner to that used
previously furnished the target compound 16. This
method once again proved to be general and was appli-
cable to the use of 11–13 as the prospective D-ring mim-
ics (Table 1).
Table 1. Ureas prepared
A-ring
mimic
D-ring
mimic
Product
Yieldc
(%)
OH
5a
(+)-11
60
N
N
O
HO
18
O
5a
12
48
63
N
N
N
O
O
HO
HO
OH
19
As can be seen from Table 1, both synthetic sequences to
access the tri- and tetra-substituted ureas proved to be
amenable to alterations in the D-ring substitute with
the target ureas 18–25 being isolated in excellent yield
over the series, providing access to good quantities of
material.
5a
(
)-11
N
20
OH
OH
5a
13
67
N
N
In summary, we have developed the synthesis of a novel
molecular scaffold aimed at mimicking the rigid tetra-
cyclic core of a steroidal skeleton. The preparation of
the tri-substituted ureas involved the development of a
new one-pot reaction sequence via the formation of an
activated carbamate, followed by the displacement of
4-nitrophenol with a secondary amine, providing rapid
access to this class of compound. This design allows
for simple modification of the reactants to introduce a
variety of A- and D-ring mimics of the steroidal skeleton
by a practically simple protocol, which should be appli-
cable to other steroidal backbones. We are currently
evaluating these compounds for biological activity at
the estrogen receptor and will report on these findings
shortly.
O
HO
21
OH
O
H
N
14b
14b
14b
14b
(+)-11
66
42
63
74
N
N
N
N
O
O
O
O
HO
HO
HO
HO
22
H
N
12
OH
OH
23
H
N
(
)-11
24
H
N
H
N
H
N
OH
13
O
TBDMSO
OTBDMS
25
a Method as in Scheme 1.
b Method as in Scheme 2.
c Yields refer to total overall yield for the whole synthetic sequence.
17
Figure 3. Symmetrical urea prepared by the CDI method.
OH
1. ClCO2C6H4NO2, NEt3,
1. TBDMSCl, NEt3.
NO2
(-)-11, NEt3, CH2Cl2, 0oC
H
NH2
CH2Cl2, r.t.
N
N
2. H2, Pd/C,
HO
2. TBAF,
THF, r.t.
MeOH, r.t.
84%
TBDMSO
O
HO
78%
14
15
16
Scheme 2. Preparation of tri-substituted ureas.

8524
E. Bridgeman et al. / Tetrahedron Letters 46 (2005) 8521–8524
12. Batey, R. A.; Santhakumar, V.; Yoshina-Ishii, C.; Taylor,
Acknowledgements
S. D. Tetrahedron Lett. 1998, 39, 6267.
13. Harris, B. D.; Bhat, K. L.; Joullie, M. M. Synth. Commun.
1986, 16, 1815.
14. Roglans, A.; Marquet, J.; Moreno-Manas, M. Synth.
Commun. 1992, 22, 1249.
15. Skarzewski, J.; Gupta, A. Tetrahedron: Asymmetry 1997,
8, 1861.
The authors wish to thank Dr. Timothy M. Willson
(GSK) for helpful discussions, the BBSRC (72/18337)
and the Wellcome Trust for financial support, and the
HRMS service at Swansea for analyses.
16. In the reduction of (3S),(4S)-dihydroxy-1-benzyl-pyrroli-
din-2,5-dione we found a modified work-up procedure
more convenient than that previously reported (Ref. 15):
A solution of (3S),(4S)-dihydroxy-1-benzylpyrrolidin-2,5-
dione (11.54 g, 52.2 mmol) and lithium aluminium hydride
(9.92 g, 0.26 mol) in THF (100 mL) was heated at reflux
for 12 h. The reaction mixture was allowed to cool to
room temperature and was cooled further using an ice
bath. The mixture was quenched with water (3 mL) and
5 M NaOH (3 mL) and diluted with ether (50 mL). The
resulting mixture was filtered through CeliteÒ and the
solids collected were washed with dichloromethane
(100 mL). The filtrate was dried over Na2SO4, filtered
and the solvent removed under reduced pressure to give
the desired compound (10.01 g, 99%) as a colourless
References and notes
1. Since its initial discovery at least 13 Nobel prizes have
been awarded that are associated with cholesterol:
Spencer, T. A. Acc. Chem. Res. 1994, 27, 83.
2. OꢀMalley, B. W. Hormones and Signalling; Academic
Press: San Diego, 1998.
3. Groen, M. B.; Zeelen, F. J. Recl. Trav. Chim. Pays-Bas
1986, 105, 465.
4. Parker, M. G. Vitam. Horm. 1995, 15, 267.
5. Chang, C.; Saltzman, A.; Yeh, S.; Young, W.; Keller, E.;
Lee, H.; Wang, C.; Mizokami, A. Crit. Rev. Eukaryot.
Gene Express. 1995, 5, 97.
6. Tsai, J. J.; OꢀMalley, B. W. Annu. Rev. Biochem. 1994, 63,
451.
7. (a) Kliewer, S. A.; Lehmann, J. M.; Willson, T. M. Science
1999, 284, 757; (b) Xie, W.; Evans, R. M. J. Biol. Chem.
2001, 276, 37739.
8. Jones, G. J. Chem. Soc. (C) 1971, 922.
9. Korach, K. S. Mol. Struct. Biol. Act. Steroids 1992, 209.
10. Grainger, D. J.; Metcalfe, J. C. Nature Med. 1996, 2, 381.
11. All compounds were characterised by Mpt, [a]D, IR ,
1H NMR, 13C NMR, MS and HRMS.
crystalline solid, mp 98–99 °C (lit. mp 98–99 °C);15 mmax
1
(Nujol)/cmÀ1 3136 (OH); H NMR(400 MHz, CDCl ) d
3
7.28–7.12 (5H, m), 4.03 (2H, dd, J = 5.3, 3.4 Hz), 3.61
(1H, d, J = 12.9 Hz), 3.57 (1H, d, J = 12.9 Hz), 2.92 (2H,
dd, J = 10.2, 5.3 Hz), 2.40 (2H, dd, J = 10.2, 3.4 Hz), 2.36
(2H, br s); dC (100 MHz, CDCl3) d 138.5, 129.2, 128.7,
127.6, 78.9, 60.6, 60.2; m/z (APCI) 194 (MH+, 100%);
HRMS calcd for C11H15NO2 194.1183 [M+H]Æ+, found
194.1180.