Heterodimeric bis(amino thiol) complexes of oxorhenium(V) that mimic the structure of steroid hormones. Synthesis and stereochemical issues

Article abstract of DOI:10.1021/jo951995k

We have prepared a series of bis-bidentate complexes of rhenium that mimic the size, shape, and peripheral functionality of steroidal androgens. In a model system, we used 2D NMR and X-ray crystallographic analysis to show that adjacent N-methyl and oxo substitutents adopt an anti configuration during the coordination reaction. We have synthesized a bis-bidentate oxorhenium(V) complex whose structure and peripheral functionality mimic 5α-dihydrotestosterone. 2D-NMR analysis indicates that the N-methyl and oxo substituents are driven into the steroidal anti configuration (β-N-methyl, α-oxo) by the β-orientation of the methyl group equivalent to C-18. Thus, this metal complex provides a remarkable structural and stereochemical mimic of a steroid. Its in vivo stability, however, appears to be limited.

Full text of DOI:10.1021/jo951995k

2624
J . Org. Chem. 1996, 61, 2624-2631
Heter od im er ic Bis(a m in o th iol) Com p lexes of Oxor h en iu m (V) Th a t
Mim ic th e Str u ctu r e of Ster oid Hor m on es. Syn th esis a n d
Ster eoch em ica l Issu es
Roy K. Hom,† Dae Yoon Chi,†,‡ and J ohn A. Katzenellenbogen†,*
Department of Chemistry, University of Illinois, 600 S. Mathews Avenue, Urbana, Illinois 61801
Received November 9, 1995X
We have prepared a series of bis-bidentate complexes of rhenium that mimic the size, shape, and
peripheral functionality of steroidal androgens. In a model system, we used 2D NMR and X-ray
crystallographic analysis to show that adjacent N-methyl and oxo substitutents adopt an anti
configuration during the coordination reaction. We have synthesized a bis-bidentate oxorhenium(V)
complex whose structure and peripheral functionality mimic 5R-dihydrotestosterone. 2D-NMR
analysis indicates that the N-methyl and oxo substituents are driven into the steroidal anti
configuration (â-N-methyl, R-oxo) by the â-orientation of the methyl group equivalent to C-18. Thus,
this metal complex provides a remarkable structural and stereochemical mimic of a steroid. Its in
vivo stability, however, appears to be limited.
In tr od u ction
conjugate design as ligands for the progesterone receptor.
Although the in vitro binding of these compounds was
quite high (up to three times that of progesterone itself),4a
in vivo, these compounds exhibited high nonspecific
binding.4b Even when their surprisingly high lipophilic-
ity was moderated, the compounds did not provide a
satisfactory target to nontarget tissue biodistribution,4c
likely due to the large overall bulk and mass of these
systems.
Technetium-99m radiopharmaceuticals have played a
vital role in diagnostic medical imaging. Largely due to
its availability via the molybdenum-99/technetium-99m
generator,1 this radionuclide is used in the vast majority
of routine imaging procedures. There has also been
interest in developing rhenium-containing compounds,
not only because they are structurally related to their
technetium-99m analogues, but because the radioactive
rhenium-188 is similarly attainable from a tungsten-188/
rhenium-188 generator, which may make this radio-
nuclide available for therapeutic applications.2 As a
result, a class of these agents, the N2S2 or bisamine-
bisthiol complexes, have been the subject of a great deal
of recent research.3
While most technetium and rhenium N2S2 complexes
are used to measure blood flow or metabolic activities in
vivo, there is a growing interest in developing radio-
pharmaceuticals labeled with these metals that act as
ligands for receptors. Two general strategies have been
used in the design of such receptor-binding radiophar-
maceuticals, the conjugate or pendant design and the
integrated design. In the pendant design, a conjugate
is prepared by attaching a metal chelate system via a
tether to a sterically tolerant site on a compound of
known receptor binding affinity.4 We have prepared
technetium- and rhenium-containing components of this
The relatively low polarity of the metal complex in
these conjugate systems suggested an alternative ap-
proach, that of integrating the metal complex within the
structure of the steroid itself. There are many ways in
which this integration might be done; the one that we
have investigated first results in the replacement of the
B and C rings of a steroid by the metal-chelate system
(see below).5 In this fashion, the steric contour of the
steroid is mimicked well, and the functionalities neces-
sary for receptor binding are preserved. The desired
consequence of the integrated design is that both the size
and the lipophilicity of the integrated complexes would
be lower than those of the conjugated series, which may
facilitate their delivery to the desired target tissue.
Our continuing interest in developing technetium- and
rhenium-containing compounds as ligands for steroid
receptors has led us to the synthesize and evaluate such
integrated complexes as potential diagnostic imaging
agents for steroid hormone-responsive cancers. In previ-
ous work, we have described the selective formation of
prototypical hetero-bisbidentate integrated systems, and
we have made a preliminary evaluation of their in vivo
stability.5 In the present study, we describe the synthesis
of a series of oxorhenium(V) complexes including the first
hetero-bisbidentate complex to mimic the steric structure
† University of Illinois.
‡ Current address: Inha University, Department of Chemistry, 253
Yong Hyun Dong Nam Gu, Inchun, Korea 402-751.
X Abstract published in Advance ACS Abstracts, March 15, 1996.
(1) (a) Muenze, R.; Berger, R. Radiochim. Acta 1987, 41 (2-3), 97.
(b) Boyd, R. E. Radiochim. Acta 1987, 41 (2-3), 59.
(2) Deutsch, E.; Libson, K.; Vanderheyden, J . L. In Technetium and
Rhenium in Chemistry and Nuclear Medicine; Nicolini, M., Bandoli,
G., Mazzi, U., Eds.; Raven Press: New York, 1990; pp 14-22.
(3) (a) Oya, S.; Kung, M.-P.; Frederick, D.; Kung, H. F. Nucl. Med.
Biol. 1995, 22, 749. (b) Charbonnel-J obic, F.; Guemas, J .-P.; Adelaere,
R.; Parrain, J .-L.; Quintard, J .-P. Bull. Soc. Chim. Fr. 1995, 132, 624.
(c) Francesconi, L. C.; Yang, Y. Y.; Kung, M.-P.; Zhang, X. X.; Billings,
J . J .; Guo, Y.-Z.; Kung, H. F. J . Med. Chem. 1994, 37, 3282. (d) Cros,
G.; Belhadj Tahar, H.; de Montauzon, D.; Gleizes, A.; Coulais, Y.;
Guiraud, R.; Bellande, E.; Pasqualini, R. Inorg. Chim. Acta 1994, 227,
25. (e) Coulais, Y.; Cros, G.; Darbieu, M. H.; Tafani, J . A. M.; Belhadj
Tahar, H.; Bellande, E.; Pasqualini, R.; Guiraud, R. Nucl. Med. Biol.
1994, 21, 263. (f) Stepniak-Biniakiewicz, D.; Chen, B.; Deutsch, E. J .
Med. Chem. 1992, 35, 274. (g) Marzilli, L. G.; Banaszczyk, M. G.;
Hansen, L., Kuklenyik, Z.; Cini, R.; Taylor, A., J r. Inorg. Chem. 1994,
33, 4850.
(4) (a) DiZio, J . P.; Fiaschi, R.; Davison, A.; J ones, A. G.; Katzenel-
lenbogen, J . A. Bioconjugate Chem. 1991, 2, 353. (b) DiZio, J . P.;
Anderson, C. J .; Davison, A.; Ehrhardt, G. J .; Carlson, K. E.; Welch,
M. J .; Katzenellenbogen, J . A. J . Nucl. Med. 1992, 33, 558. (c) O'Neil,
J . P.; Carlson, K. E.; Anderson, C. J .; Welch, M. J .; Katzenellenbogen,
J . A. Bioconjugate Chem. 1994, 5, 182. (d) Lever, S. Z.; Baidoo, K. E.;
Mahmood, A.; Matsumura, K.; Scheffel, U.; Wagner, H. N. Nucl. Med.
Biol. 1994, 21, 157. (e) Del Rosario, R. B.; J ung, Y.-W.; Baidoo, K. E.;
Lever, S. Z.; Wieland, D. M. Nucl. Med. Biol. 1994, 21, 197.
(5) (a) Chi, D. Y.; Katzenellenbogen, J . A. J . Am. Chem. Soc. 1993,
115, 7045. (b) Chi, D. Y.; O'Neil, J . P.; Anderson, C. J .; Welch, M. J .;
Katzenellenbogen, J . A. J . Med. Chem. 1994, 37, 928.
0022-3263/96/1961-2624$12.00/0 © 1996 American Chemical Society

Metal Complexes That Mimic Steroid Hormones
J . Org. Chem., Vol. 61, No. 8, 1996 2625
centers in DHT, it was difficult to predict the configura-
tions of the metal oxo and quaternary amine centers in
3, since these stereocenters are only formed upon complex
formation. In order to address this issue, we first
synthesized a simpler model complex 4 (Figure 3).
The synthesis of the amino thiol ligands necessary to
form model system 4 is illustrated in Scheme 1. The
steroid A-ring mimic, N-methyl-2-(mercaptomethyl)-
piperidine (6), was prepared by Mitsunobu thiolation of
the commercially available racemic amino alcohol 5
followed by cleavage of the acetate in base.6 We have
already described the synthesis of the D-ring mimic (S)-
2-(mercaptomethyl)pyrrolidine hydroiodide (8‚HI) from
L-prolinol (7).5b
Complex 4 was prepared by subjecting a 1:1:1 mixture
of bidentate ligands 6 and 8‚HI with trichlorobis(tri-
phenylphosphine)oxorhenium(V) under weak alkaline
conditions (Figure 4). A 1:1 diastereomeric mixture of
the hetero-bisbidentate complexes 4a and 4b, isolated in
16% unoptimized yield, was separated by flash chroma-
tography.
F igu r e 1. Template design of steroid-mimic integrated metal
complexes.
The structures of both diastereomers 4a and 4b were
elucidated by 1H, 13C, and 2D (1H-1H and 1H-13C
correlated) NMR spectrometry. The most prominent
features in the 1H NMR spectrum of each of these
compounds, the methyl resonances, were also the most
informative regarding the relative configurations of the
metal and quaternary nitrogen. The methyl protons in
4a (δ 2.23 ppm) and 4b (δ 2.29) appear at about the same
chemical shift as does the methyl group resonance in the
free ligand 6 (δ 2.32). If the NCH3 and RedO substitu-
ents were syn, one would expect a large downfield shift
of the methyl protons due to the pronounced deshielding
effect that is experienced at the periphery of the RedO
bond. This effect has been previously noted in syn vs
anti N-benzyl oxorhenium complexes,7 and also in syn
vs anti N-substituted oxotechnetium(V) species.8 Since
the effect is not observed, an anti disposition of the
N-methyl and oxo substituents can be deduced.
The OdRe deshielding effect was also vital in identify-
ing each diastereomer. The anti C-5 methine proton in
4a (δ 2.35-2.49) was significantly farther upfield com-
pared to the deshielded syn methine proton in 4b (δ
3.72-3.80). Because the axial proton on the pyrrolidine
ligand was known, from X-ray structures of similar
compounds, to form in a syn disposition with the oxo
substituent,5 the absolute configurations of 4a and 4b
were determined to be as shown in Figure 4.
X-ray crystallographic data20 collected on each of the
diastereomers confirmed the absolute configuration as-
signed by 1H NMR analysis (Figure 5). Additionally, the
ORTEP diagrams show both diastereomers to be of
distorted square pyramidal geometry, similar to other
five-coordinate oxorhenium(V) and oxotechnetium(V)
compounds previously studied.9
Although the quaternary nitrogen in the model com-
plexes 4a and 4b was of the opposite configuration
desired for mimicry of DHT, these results indicated that
the steric forces that govern the complexation might, in
fact, aid in the formation of the desired, steroid-like
of an androgen. In particular, our investigation has
focused on critical stereochemical issues which arise upon
formation of these complexes and determine the degree
to which they mimic the configuration of a steroid. We
have examined this issue both in a set of model com-
plexes, as well as in a complex that is the mimic of 5R-
dihydrotestosterone. The preferred stereochemistries
that we find can be rationalized on the basis of the
minimization of steric interaction during formation of the
hetero-bisbidentate N2S2 oxometal complex, with the
result that the steroid mimic adopts the desired stereo-
chemistry, the one that most corresponds to that of the
steroid, while for the same reason, the model system
adopts an alternate stereochemistry.
Resu lts a n d Discu ssion
Design of Oxom eta l(V) Com p lexes Th a t Mim ic
th e Str u ctu r e of An d r ogen s. Our approach to the
design of oxometal(V) compounds that mimic the struc-
ture of steroids has been previously described and is
illustrated in Figure 1.5 Because the metal-heteroatom
bonds in N2S2 oxometal complexes are significantly longer
than the C-C bonds in a steroidal skeleton (1), one can
envision the replacement of the B and C ring decalin
system with the metal-N2S2 complex, the metal atom
being centered on the C8-C9 bond. The steroid-metal
complex overlay 2, thus, suggests the design of a complex
3 whose structure closely mimics that of the androgen
receptor ligand 5R-dihydrotestosterone (DHT) (1).
The accuracy of this rather simple template overlay
approach may be verified by more quantitative modeling
of the proposed complex (Figure 2). In a space filling
model generated from Syby, one can see that the overall
bulk of the proposed complex is very similar to that of
the steroid (in fact, the calculated volume of complex 3
is within 3% of that determined for dihydrotestosterone).
Furthermore, from the stereo overlay view, it is evident
that the relative positions of the keto, alcohol, and methyl
groups critical to high affinity binding of DHT to the
androgen receptor are very well preserved in complex 3.
Syn th esis of a Mod el Com p lex 4. While complex 3
appears to be an excellent steric mimic of a steroid, a
critical stereochemical question needed to be addressed
regarding its formation: While appropriate synthetic
methods can be used to obtain steroid-like configurations
at the stereocenters that correspond to the 5R and 13â
(6) Hughes, D. L. In Organic Reactions; Paquette, L. A., Ed.-in-Chief;
J ohn Wiley & Sons: New York, 1992; Vol. 42, Chapter 2.
(7) O’Neil, J . P.; Wilson, S. R.; Katzenellenbogen, J . A. Inorg. Chem.
1994, 33, 319.
(8) (a) Lever, S. Z.; Baidoo, K. E.; Mahmood, A. Inorg. Chim. Acta
1990, 176, 183. (b) Mahmood, A.; Halpin, W. A.; Baidoo, K. E.;
Sweigart, D. A.; Lever, S. Z. In Technetium and Rhenium in Chemistry
and Nuclear Medicine; Nicolini, M, Bandoli, G., Mazzi, U., Eds.; Raven
Press: New York, 1990; pp 113-118.

2626 J . Org. Chem., Vol. 61, No. 8, 1996
Hom et al.
F igu r e 2. Top: Crossed stereoview of superposition of dihydrotestosterone (1) (light shading) and proposed mimic 3 (dark shading).
Bottom: Space filling models of 1 (left) and 3 (right).
F igu r e 3. Rationale for the choice of model complex.
F igu r e 4. Coordination reaction of model complex.
configuration in the real complex 3: The exclusive anti
relative configuration of the N-methyl and oxo substit-
uents in the model systems 4a and 4b, which we could
rationalize on the basis of minimizing steric repulsion,
clearly led us to expect that these two groups will be anti
in 3. However, in the model complexes 4a and 4b the
metal oxo center formed exclusively syn to the methine
proton of the pyrrolidine ligand. We thought that this
preference might be reversed in the real complex 3, where
this methine proton would be replaced with a sterically
more demanding axial methyl substituent. Thus, mini-
mization of steric repulsion led us to expect that complex
3 would have the anti,anti disposition of the N-methyl,
oxometal, and axial methyl moieties, as required for
mimicry of the 19â-methyl group in DHT.
F igu r e 5. ORTEP diagrams of 4a (top) and 4b (bottom).
Sch em e 1
Syn th esis of Liga n d s 9 a n d 19 for th e Oxom eta l
Dih yd r otestoster on e Str u ctu r a l Mim ic. The A-ring
mimic, N-methyl-2-(mercaptomethyl)-4-pyridone hydro-
chloride (9‚HCl), was prepared according to the route
shown in Scheme 2. From commercially available 2-
picoline N-oxide (10), nitration provided 11 in good yield.
Nucleophilic displacement of the nitro group with benzyl
oxide generated 4-(benzyloxy)-2-picoline N-oxide (12),
which then could be treated with acetic anhydride to
undergo a novel rearrangement to provide pyridine
acetate 13.10 Saponification gave alcohol 14 and Mit-
(9) (a) Schultze, L. M.; Todaro, L. J .; Baldwin, R. M.; Byrne, E. F.;
McBride, B. J . Inorg. Chem. 1994, 33, 5579. (b) Chi, D. Y.; Wilson, S.
R.; Katzenellenbogen, J . A. Inorg. Chem. 1995, 34, 1624. (c) Rao, T.
N.; Adhikesavalu, D.; Camerman, A.; Fritzberg, A. R. J . Am. Chem.
Soc. 1990, 112, 5798. (d) Fietz, T.; Spies, H.; Pietzsch, H.-J .; Leibnitz,
P. Inorg. Chim. Acta 1995, 231, 233. (e) J ohn, C. S.; Kung, M.-P.;
Billings, J .; Kung, H. F. Nucl. Med. Biol. 1991, 18, 551. (f) McDonnell,
A. C.; Hambley, T. W.; Snow, M. R.; Wedd, A. G. Aust. J . Chem. 1983,
36, 253. (g) Bandoli, G.; Gerber, T. I. A. Inorg. Chim. Acta 1987, 126,
205.

Metal Complexes That Mimic Steroid Hormones
J . Org. Chem., Vol. 61, No. 8, 1996 2627
Sch em e 2
Sch em e 3
best metal precursor for this complexation, since the ratio
of 3a (desired DHT mimic) to 3b (undesired diastere-
omer) was as high as 1:3, while with other rhenium
reagents ([BnEt3N][OReCl4],9d [n-Bu4N][OReBr4],14 or
NH4ReO4/SnCl2) either a lower yield or a lower ratio of
3a to 3b was obtained. The modest yield and the fact
that the undesired diasteromer 3b is isolated in a higher
proportion may be explained by the stability of these
complexes (vide infra).
Oxotechnetium(V) complexes 27a and 27b were also
formed by ligand exchange with tetrabutylammonium
tetrachlorooxotechnetate(V). The technetium (V) com-
plexes formed in similar yield and under much milder
conditions compared with those used for the rhenium
complexes 3a and 3b (0 °C vs 80 °C), as consistent with
previous observations on similar systems.5,7 Compounds
27a and 27b, however, were not characterized owing to
their low stability (vide infra).
1H NMR analysis once again proved vital to the
assignment of the relative stereochemistry of complexes
3a and 3b. In the 1D spectra of 3a and 3b, the
prominent N-methyl and axial methyl protons were
clearly not shifted downfield compared to the free ligand
methyl resonances. As described earlier, this indicated
that both diastereomers had the anti,anti relationship
of the N-methyl, oxo, and axial methyl substituents
predicted by the consideration of steric repulsion. 2D
COSY analysis of each isomer allowed for identification
of the diastereomers, using the chemical shift data of the
C-5 methine proton, in a similar process to that used for
model complexes 4a and 4b. The cis-fused compound 3b
exhibited an upfield resonance (δ 2.87-3.00) for the C-5
methine proton, while trans-fused 3a showed a C-5
methine resonance dramatically deshielded (δ 3.95-4.02)
by the adjacent OdRe moiety. In contrast to model
compounds 4a and 4b, attempts at obtaining crystals of
compound 3a suitable for X-ray diffraction proved im-
practical, owing to the small amount of pure material
which was available.
sunobu thiolation provided protected pyridine thioacetate
15 in excellent yield.6 The N-methylpyridinium iodide
16 was reduced with sodium borohydride to give a
mixture of thiol 17 and disulfide 18, which was treated
with lithium aluminum hydride to afford only amino thiol
17. Direct reduction of the pyridinium ring with lithium
aluminum hydride also gave 17, but in lower yield (18%).
Thiol 17 was treated with HCl to provide A-ring mimic
as the crystalline hydrochloride salt 9‚HCl.
The D-ring mimic, trans-2-methyl-2-(mercaptomethyl)-
3-pyrrolidinol hydrochloride (19‚HCl), was prepared ac-
cording to the route outlined in Scheme 3. From trans-
6-hydroxy-5-methyl-1-aza-3-oxabicyclo[3.3.0]octan-3-
one (20),11 pyridinium chlorochromate (PCC) oxidation
to gave ketone 21. Reduction of the ketone allowed for
inversion of the secondary alcohol relative configuration.
Mitsunobu inversion of alcohol 20 using benzoic acid12
or 3-nitrobenzoic acid13 resulted in low yields and incom-
plete stereochemical inversion. Secondary alcohol 22 was
then benzyl protected, the amine and primary alcohol
functionalities deprotected, and the amine reprotected to
provide neopentyl alcohol 25. Mitsunobu thiolation on
this highly hindered alcohol yielded tris-protected pyr-
rolidine derivative 26 in good yield. Treatment with
sodium in ammonia allowed for simultaneous deprotec-
tion of all groups, and acidification afforded D-ring mimic
19‚HCl.
F or m a t ion of Oxom et a l Dih yd r ot est ost er on e
Mim ic 3. As with model complexes 4a and 4b , oxo-
rhenium species 3a and 3b were formed by coordination
under typical conditions, with a 1 : 1 ratio of bidentate
ligands 9‚HCl and 19‚HCl (Figure 6). Trichlorobis-
(triphenylphosphine)oxorhenium(V) was found to be the
(10) Ochiai, E. Aromatic Amine Oxides; Elsevier: New York, 1967;
pp 290-302.
(11) (a) Tamaru, Y.; Kawamura, S.; Bando, T.; Tanaka, K.; Hojo,
M.; Yoshida, Z. J . Org. Chem. 1988, 53, 5491. (b) Tamaru, Y.;
Kawamura, S.; Tanaka, K.; Yoshida, Z. Tetrahedron Lett. 1984, 25,
1063.
Deter m in a tion of Com p lex Lip op h ilicity. In order
to determine lipophilicity of 3a and 3b, the log Po/w values
(12) Mitsunobu, O. Synthesis 1981, 1.
(13) Dodge, J . A.; Trujillo, J . I.; Presnell, M. J . Org. Chem. 1994,
59, 234.
(14) Cotton, F. A.; Lippard, S. J . Inorg. Chem. 1966, 5, 9.

2628 J . Org. Chem., Vol. 61, No. 8, 1996
Hom et al.
very low binding affinity for the androgen receptor
(<0.02%; R1881, 100%). To our knowledge, no compound
of such low lipophilicity has exhibited significant binding
for the androgen receptor.
The technetium-99 complexes were very unstable
under aqueous conditions; attempted reversed-phase
HPLC analysis of the compounds resulted in total
decomposition. Owing to the lack of promise for further
development as imaging agents, the technetium com-
plexes were not investigated further.
Con clu sion
We have shown it is possible to synthesize an oxo-
metal(V) N2S2 hetero-bisbidentate complex which is a
close structural and stereochemical mimic of the andro-
gen 5R-dihydrotestosterone. The minimization of steric
repulsion during complex formation assists in the gen-
eration of the desired steroid-like stereochemistry. This
oxometal mimic, however, exhibited low stability and low
lipophilicity, which precluded its use in vivo.
F igu r e 6. Coordination reaction to form steroidal mimics 3a
and 3b.
Exp er im en ta l Section
Molecu la r Mod elin g. Molecular modeling calculations
were performed on a Silicon Graphics Indigo Elan computer.
Compound 3a was built using the SYBYL Molecular Modeling
Package (Version 6.1, Tripos Associates, St. Louis, MO) around
the X-ray crystal parameters for rhenium and the five sur-
rounding heteroatoms from 4a . The compound was energy
minimized to a gradient of <0.05 kcal/mol Å with the Tripos
force field. Molecular volumes were determined by reading
these minimized structures into Macromodel17 v.4.5 and
calculating the total van der Waals volume occupied by the
sum of the atoms.
Gen er a l. All reagents and solvents were obtained from
Aldrich, Eastman, Fisher, or Mallinckrodt. All reactions were
performed under a nitrogen atmosphere unless otherwise
indicated. The synthesis of (S)-2-(mercaptomethyl)pyrrolidine
hydroiodide has been previously described.5 Tetrahydrofuran
(THF) was distilled from sodium/benzophenone ketyl im-
mediately prior to use. Methylene chloride (CH2Cl2) was
distilled from calcium hydride prior to use in reactions unless
specified otherwise. Dimethylformamide (DMF) was distilled
from and stored over 4 Å molecular sieves. Hexanes were
distilled from calcium sulfate (Drierite) before use in column
chromatography.
F igu r e 7. In vitro stability of rhenium complexes 3a and 3b
in buffer solution.
were estimated by the reversed-phase HPLC method of
Minick.15 The log Po/w values were the following: for 3a ,
1.47; for 3b, 1.69. Compared to dihydrotestosterone (log
Po/w ) 3.90), oxorhenium(V) mimic 3a was 250 times less
lipophilic. Thus, integration of the metal complex within
the periphery of the steroid, which was done in part to
reduce the lipophilicity of the metal-containing system,
has resulted in compounds which may be too polar.
Sta bility of Oxom eta l(V) Com p lexes a n d An d r o-
gen Recep tor Bin d in g Affin ity. Studies of the in vitro
stability of 3a and 3b in androgen receptor buffer
indicated significant decomposition (Figure 7).16 After
only 1 h in buffer solution, complex 3a has nearly
completely disappeared, and after 2 h only 22% of 3b
remained. Further investigations determined that com-
petitive ligand exchange from water and not solutes in
the buffer contributed to the destruction of the bis-
bidentate complexes. The relative stability of 3b over
3a was likely due to the cis-fused piperidone ring of 3b,
which hindered the nucleophilic anti attack of water at
the metal center; 3a , lacking this steric protection,
decomposed much faster.
Reaction progress was monitored with analytical thin-layer
chromatography (TLC) on 0.25 mm Merck F-254 silica gel
glass plates. Visualization was achieved using phospho-
molybdic acid (PMA), KMnO4, or ninhydrin spray reagents,
or UV illumination. Flash chromatography was performed
according to the method of Still18 with Woelm silica gel (0.040-
0.063 mm) packing. High performance liquid chromatography
(HPLC) was performed isocratically on a Spectra-Physics
Model 8700 with an analytical 5-mm SiO2 cyanopropyl column
(4.6 mm × 30 cm, Supelco LC-CN) or a semipreparative SiO2
cyanopropyl column (10 mm × 50 cm, Phenomenex IB-SIL).
The UV/visible absorbance of the eluent was monitored at 254
or 410 nm. Melting points were determined using a Thomas
Hoover melting point apparatus and are uncorrected. Crystal-
lographic measurements were made on an Enraf-Nonius CAD4
automated κ-axis diffractometer equipped with graphite mono-
chromated Mo radiation, λ(KRj) ) 0.71073 Å.
1H and 13C NMR spectra were obtained on a General Electric
QE-300 (300 MHz) or a Varian Unity (400 MHz) spectrometer
and are reported in parts per million downfield from internal
tetramethylsilane or from proton resonances resulting from
incomplete deuteration of the NMR solvent (δ scale). COSY
and HETCOR spectra were obtained on a Varian Unity (400
As a result of the hydrophilicity and low stability of
the rhenium species, both complexes 3a and 3b exhibited
(15) Minick, D. J .; Frenz, J . H.; Patrick, M. A.; Brent, D. A. J . Med.
Chem. 1988, 31, 1923.
(16) Brandes, S. J .; Katzenellenbogen, J . A. Mol. Pharmacol. 1987,
32, 391.
(17) Mohmadi, F.; Richards, N. G. J .; Guida, W. C.; Liskamp, R.;
Lipton, M.; Caufield, C.; Chang, G.; Hendrickson, T.; Still, W. C. J .
Comput. Chem. 1990, 11, 440.
(18) Still, W. C.; Kahn, M.; Mitra, A. J . Org. Chem. 1978, 43, 2923.

Metal Complexes That Mimic Steroid Hormones
J . Org. Chem., Vol. 61, No. 8, 1996 2629
MHz) spectrometer. Low resolution electron impact (EI),
chemical impact (CI), and fast atom bombardment (FAB) mass
spectra were obtained on Finnigan MAT CH5, VG 70-VSE,
and VG ZAB-SE spectrometers, respectively. High resolution
EI, CI, and FAB mass spectra were obtained on Finnigan MAT
731, VG 70-VSE, and VG 70-SE-4F spectrometers, respec-
tively. Elemental analyses were performed by the Microana-
lytical Service of the University of Illinois.
Anal. Calcd for C12H22N2OS2Re: C, 31.22; H, 5.02; N, 6.07.
Found: C, 31.31; H, 5.13; N, 5.89.
(1R,2S)-4b: IR (cm-1, KBr pellet) 930; 1H NMR (CDCl3, 400
MHz) δ 1.47-1.55 (m, 2R, 1), 1.67-1.77 (m, 4R, 10R, 2), 1.82-
1.95 (m, 3R, 3â, 2), 1.95-2.06 (m, 2â, 4â, 9R, 3), 2.06-2.15 (m,
9â, 1), 2.29 (s, 7-CH3, 3), 2.40-2.48 (m, 10â, 1), 2.58 (t, J )
11.1 Hz, 12R, 1), 2.66 (dd, J ) 13.0, 12.0 Hz, 6R, 1), 3.11 (dd,
J ) 12.9, 4.9 Hz, 6â, 1), 3.39 (dd, J ) 11.0, 5.6 Hz, 12â, 1),
3.72-3.80 (m, 5â, 1) 3.85-3.95 (m, 11â, 1), 4.02 (dtq, J ) 13.4,
3.5, 0.6 Hz, 1R, 1), 4.10 (ddddd, J ) 13.1, 7.9, 4.2, 1.5, 0.5 Hz,
8R, 1), 4.27 (dddt, J ) 13.9, 3.5, 2.0, 0.6 Hz, 1â, 1), 4.50 (ddddd,
J ) 13.2, 8.5, 6.9, 2.1, 0.5 Hz, 8â, 1); 13C NMR (CDCl3, 100
MHz) δ 21.1, 22.9, 24.3 (C4), 27.6 (C9), 32.0 (C10), 37.6 (C7),
43.2 (C6), 48.3 (C12), 66.0 (C1), 66.4 (C8), 77.3 (C5), 86.5 (C11);
MS (EI, 10 eV) (m/z) 463 (20), 462 (M+ for 187Re, 100), 461
(11), 460 (M+ for 185Re, 56), 393 (84), 391 (52), 278 (17), 277
(23), 113 (13), 112 (63), 111 (48), 98 (83); HRMS calcd for
C12H23N2OS2185Re 460.0782, found 460.0782. Anal. Calcd for
C12H22N2OS2Re: C, 31.22; H, 5.02; N, 6.07; S, 13.88. Found:
C, 31.52; H, 5.16; N, 6.08; S, 13.56.
N-Meth yl-2-(m er ca p tom eth yl)p ip er id in e (6). Triphen-
ylphosphine (12.18 g, 46.5 mmol) was dissolved in THF (100
mL), and diisopropyl azodicarboxylate (9.1 mL, 46 mmol) was
added at rt. The solution was cooled to 0 °C, and additional
THF (50 mL) was added via syringe into the flask. N-Methyl-
2-piperidinemethanol (5, 4.96 g, 39 mmol) in THF (25 mL) was
added after 20 min, and thiolacetic acid (3.5 mL, 46.5 mmol)
was added dropwise immediately thereafter with stirring. The
reaction was allowed to stir at 0 °C for 1 h and then at room
temperature for 16 h. The mixture was concentrated, and
flash chromatography (5% MeOH/CH2Cl2, Rf 0.24 in 10%
MeOH/CH2Cl2; KMnO4) yielded slightly yellow liquid as
N-methyl-2-(thioacetyl)methylpiperidine (6.04 g, 92%).
To a solution of N-methyl-2-[(acetylthio)]methylpiperidine
(2.13 g, 12.1 mmol) in 5:1 MeOH:H2O (60 mL) was added
potassium carbonate (5.13 g, 37.1 mmol) in one portion. After
15 min, additional MeOH/H2O (50 mL/15 mL) was added to
help solubilize any remaining solid. TLC showed one spot near
the base line (20% MeOH/CH2Cl2 ; KMnO4) after 15 additional
min. The solution was concentrated to 30 mL and extracted
with CH2Cl2. The extracts were dried (MgSO4) and concen-
trated to afford the title compound (1.32 g, 75%). A small
amount of the slightly yellow oil was purified by Kugelrohr
distillation (130 °C, 0.7 torr) to yield an analytically pure
sample: 1H NMR (CDCl3, 400 MHz) δ 1.22-1.34 (m, 1), 1.42-
1.51 (m, 1), 1.51-1.63 (m, 2), 1.68-1.76 (m, 1), 1.76-1.83 (m,
1), 2.04-2.14 (m, 1), 2.14-2.21 (m, 1), 2.32 (s, NCH3, 3), 2.85
(dt, J ) 11.5, 3.9 Hz, 1), 2.92 (dd, B of AB q, J ) 12.8, 7.2 Hz,
CH2OH, 1), 3.00 (ddd, A of AB q, J ) 12.9, 4.2, 3.2 Hz, CH2OH,
1); 13C NMR (CDCl3, 100 MHz) δ 23.7, 25.5, 30.6, 43.11, 43.14,
56.6 (NCH3), 62.9 (C2); MS (CI, CH4) (m/z) 290 (10), 289 (M +
H+ of dithiol, 58), 287 (14), 146 (M + H+, 59), 145 (24), 144
(100), 142 (28), 112 (50), 98 (81); HRMS calcd for C7H16NS+
146.1003, found 146.0992. Anal. Calcd for C7H15NS: C, 57.89;
H, 10.42; N, 9.65. Found: C, 58.29; H, 10.55; N, 10.03.
[(1R,2R)-N-Meth yl-2-(m er ca p tom eth yl)p ip er id in a to]-
[(S)-2-(m er captom eth yl)pyr r olidin ato]oxor h en iu m (V) (4a)
a n d [(1R,2S)-N-Meth yl-2-(m er ca p tom eth yl)p ip er id in a to]-
[(S)-2-(m er ca p t om et h yl)p yr r olid in a t o]oxor h en iu m (V)
(4b). A sample vial (5 mL) was charged with amino thiol 6
(30.6 mg, 0.2 mmol), a stir bar, and 1 N methanolic NaOAc
solution (2 mL, 2.0 mmol). To it was added the oxotrichlorobis-
(triphenylphosphine)rhenium(V) (166 mg, 0.2 mmol), and then,
in succession, amino thiol 8‚HCl (49.0 mg, 0.2 mmol) and
triethylamine (28 µL, 0.2 mmol). The vial was sealed and
heated to 75 °C for 20 min. From the dark burgundy red
reaction mixture, rhenium starting material (4 mg) was
isolated by filtration. The filtrate was then concentrated, and
flash chromatography (10-30% EtOAc/hexanes) yielded an
orange complex 4a (Rf 0.10 in 10% EtOAc/hexanes) and a red
complex 4b (Rf 0.056 in 10% EtOAc/hexanes) as solid products
(14.4 mg, 16% of a 1:1 ratio of isomers). Recrystallization of
each diastereomer by vapor diffusion from EtOAc/pentane at
-20 °C afforded crystals suitable for X-ray analysis.
4-Nitr o-2-p icolin e N-Oxid e (11). Note: 2-Picoline N-oxide
is extremely hygroscopic. 2-Picoline N-oxide (10, 13.1 g, 0.12
mol) and concentrated sulfuric acid (30 mL) were thoroughly
mixed and then heated to 85-95 °C. A mixture of nitric (55
mL) and sulfuric acid (44 mL) was added slowly dropwise with
stirring. The mixture was maintained at the above temper-
ature for 16 h before cooling in an ice bath. The acid was
neutralized carefully using 30% potassium hydroxide solution.
The precipitate was filtered, and the resulting solids were
washed with CH2Cl2. The remaining aqueous layer was also
extracted with CH2Cl2. The organic layers were collected,
dried (NaSO4), and concentrated. Recrystallization from
CHCl3 gave product 11 as yellow crystalline needles (14.1 g,
76%): Rf 0.28 in 2% EtOH/CH2Cl2; mp 154-156 °C; 1H NMR
(CDCl3, 300 MHz) δ 2.57 (s, CH3, 3), 8.00 (dd, J ) 7.2, 3.2 Hz,
C5-H, 1), 8.15 (d, J ) 3.1 Hz, C3-H, 1), 8.32 (d, J ) 7.2 Hz,
C6-H, 1); MS (EI, 70 eV) (m/z) 154 (M+, 100), 137 (42), 108
(21), 91 (37); Anal. Calcd for C6H6N2O3: C, 46.76; H, 3.92; N,
18.18. Found: C, 46.67; H, 3.89; N, 18.05.
4-(Ben zyloxy)-2-p icolin e N-Oxid e (12). To the nitro-
picoline N-oxide 11 (10.1 g, 65.5 mmol) in benzyl alcohol (40
mL) at 80 °C was added a premixed solution of sodium benzyl
alkoxide (1.5 g sodium, 70 mL benzyl alcohol) over 30 min.
Immediately after addition, TLC indicated complete conversion
of starting material (Rf 0.24 in 5% EtOH/CH2Cl2). Benzyl
alcohol was removed by vacuum distillation, and the mixture
was diluted with 1:1 CH2Cl2/H2O. The aqueous layer was
extracted with CH2Cl2. The organic layers were combined,
washed with brine solution, and concentrated. The material
obtained from flash chromatography (0-10% EtOH/CH2Cl2
elution) was recrystallized from acetone to give product 12 as
1
fluffy white crystals (7.4 g, 52%): mp 154-156 °C; H NMR
(CDCl3, 300 MHz) δ 2.53 (s, CH3, 3), 5.09 (s, CH2, 2), 6.77 (dd,
J ) 7.2, 3.4 Hz, C5-H, 1), 6.86 (d, J ) 3.4 Hz, C3-H, 1), 7.40 (b
s, Ar-H, 5), 8.18 (d, J ) 7.2 Hz, C6-H, 1); MS (EI, 70 eV) (m/z)
215 (M+, 4), 91 (100), 65 (12). Anal. Calcd for C13H13NO2: C,
72.54; H, 6.09; N, 6.51. Found: C, 72.45; H, 6.09; N, 6.46.
4-(Ben zyloxy)-2-(h ydr oxym eth yl)p yr idin e (14). To ben-
zyl ether 12 (1.13 g, 5.25 mmol) in acetic anhydride (6 mL)
was added concentrated sulfuric acid (3 drops) at room
temperature. This mixture was then heated to 110 °C for 3 h
and then allowed to stand at room temperature overnight. The
reaction mixture was washed (NaHCO3, H2O, brine) and
extracted with EtOAc, and the organic solution was dried
(Na2SO4) and concentrated, leaving acetate 13 as a dark brown
oil: 1H NMR (CDCl3, 300 MHz) δ 2.16 (s, CH3, 3), 5.12 (s, CH2,
2), 5.17 (s, CH2, 2), 6.81 (dd, J ) 5.7, 2.4 Hz, C3-H, 1), 6.95 (d,
J ) 2.4 Hz, C5-H, 1), 7.36-7.42 (m, Ph-H, 5), 8.41 (d, J ) 5.7
Hz, C2-H, 1).
(1R,2R)-4a : IR (cm-1, KBr pellet) 944; 1H NMR (CDCl3,
400 MHz) δ 1.40-1.49 (m, 3R, 1), 1.63-1.73 (m, 2R, 10R, 2),
1.78-1.85 (m, 3â, 1), 1.96-2.08 (m, 9R, 1), 2.08-2.19 (m, 9â,
1), 2.19-2.35 (m, 4R, 4â, 2), 2.23 (s, 7-CH3, 3), 2.35-2.49 (m,
5R, 10â, 2), 2.45 (t, J ) 11.0 Hz, 12R, 1), 2.88 (dd, A of ABC,
J ) 12.9, 5.4 Hz, 6R, 1), 2.93-3.04 (m, 6â, 1R, 2), 3.27 (dd, J
) 10.7, 5.6 Hz, 12â, 1), 3.90-4.02 (m, 11â, 1), 4.16 (ddt, J )
12.6, 4.0, 1.7 Hz, 1â, 1), 4.24 (ddddd, J ) 12.7, 8.5, 3.9, 1.3,
0.7 Hz, 8R, 1), 4.62 (ddddd, J ) 13.1, 8.8, 7.2, 1.7, 0.5 Hz, 8â,
1); 13C NMR (CDCl3, 100 MHz) δ 23.12, 23.16, 27.8 (C9), 29.0
(C3), 32.0 (C10), 45.1, 45.2, 53.2 (C7), 63.1 (C1), 65.9 (C8), 74.7
(C5), 86.1 (C11); MS (EI, 10 eV) (m/z) 464 (11), 463 (16), 462
(M+ for 187Re, 100), 461 (13), 460 (M+ for 185Re, 59), 395 (11),
394 (10), 393 (88), 391 (49), 112 (62), 111 (51), 99 (10), 98 (84);
HRMS calcd for C12H23N2OS2185Re 460.0782, found 460.0778.
Crude acetate 13 was taken up in 2:1 MeOH/H2O (30 mL),
and K2CO3 (1 g, crude excess) was added. After 15 min,
solvent was evaporated, and the residue was dissolved in
CH2Cl2. This solution was dried (Na2SO4) and concentrated,
and flash chromatography (Rf 0.42 in 5% MeOH/CH2Cl2)
yielded 14 as a solid waxy product (752 mg, 67% by 1H NMR):
mp 102.5-103.5 °C; 1H NMR (CDCl3, 300 MHz) δ 4.69 (s, CH2,
2), 5.10 (s, CH2, 2), 6.76 (dd, J ) 5.7, 2.4 Hz, C5-H, 1), 6.91 (d,

2630 J . Org. Chem., Vol. 61, No. 8, 1996
Hom et al.
J ) 2.2 Hz, C3-H), 7.35-7.41 (m, Ar-H, 5), 8.32 (d, J ) 5.7 Hz,
C6-H, 1); MS (EI, 70 eV) (m/z) 215 (M+, 4), 109 (4), 91 (100).
Anal. Calcd for C13H13NO2: C, 72.54; H, 6.09; N, 6.51.
Found: C, 72.50; H, 6.37; N, 6.45.
corresponding ketone (Rf 0.48 in 5% MeOH/CH2Cl2; KMnO4).
This mixture was then added slowly to a rapidly stirring
suspension of Celite/Et2O (800 mL), and filtered through a pad
of silica gel, which was washed with 4:1 Et2O/CH2Cl2. Con-
centration of the organic layers and flash chromatography
yielded the title compound (1.36 g, 69%) as a white solid: mp
102.5-103.3 °C; IR (cm-1, KBr pellet) 3480, 1777, 1753, 1745,
1406, 1390, 1342, 1184, 1077, 1031; 1H NMR (CDCl3, 300 MHz)
δ 1.37 (s, CH3, 3), 2.45-2.64 (m, C7-H, 2), 3.54 (ddd, J ) 12.8,
9.1, 8.0 Hz, C8-H 1), 4.13 (d, J ) 8.9 Hz, 1), 4.24 (ddd, J )
12.8, 4.0, 3.6 Hz, 1), 4.45 (d, J ) 8.9 Hz, 1); MS (CI, CH4) (m/
4-Ben zyloxy-2-[(a cetylth io)m eth ylp yr id in e (15). To a
flask containing triphenylphosphine (1.02 g, 3.9 mmol) in THF
(10 mL) at 0 °C was added diisopropyl azodicarboxylate (761
µL, 3.9 mmol). This was allowed to stir at 0 °C for 30 min
after which the mixture became a white suspension.
A
solution of crude alcohol 14 (416 mg, 1.93 mmol) in THF (5
mL) was then added. After another 30 min, thiolacetic acid
(277 µL, 3.9 mmol) was added, causing the mixture to turn
dark orange. This mixture was allowed to stir at 0 °C for 1 h
and then at rt for 1 h. Concentration and flash chromatog-
raphy of the residue (Rf 0.41 in 60% EtOAc/hexanes) afforded
product 15 as a slightly yellow oil (500 mg, 95%): 1H NMR
(CDCl3, 400 MHz) δ 2.36 (s, CH3, 3), 4.20 (s, CH2, 2), 5.09 (s,
CH2, 2), 6.75 (dd, J ) 5.6, 2.4 Hz, C5-H, 1), 6.96 (d, J ) 2.4
Hz, C3-H, 1), 7.34-7.41 (m, Ar-H, 5), 8.34 (d, J ) 5.6 Hz, C6-
H, 1); 13C NMR (CDCl3, 100 MHz) δ 30.2, 35.4, 69.8, 109.2,
109.8, 127.431, 127.438, 127.5, 128.3, 128.7, 135.4, 158.9,
165.3, 195.0; MS (EI, 70 eV) (m/z) 273 (M+, 5), 231 (15), 91
(100), 65 (13). Anal. Calcd for C15H15NO2S: C, 65.92; H, 5.53;
N, 5.12; S, 11.72. Found: C, 65.73; H, 5.40; N, 5.27; S, 11.50.
N-Meth yl-2-(m er ca p tom eth yl0-4-p yr id on e Hyd r och lo-
r id e (9‚HCl). To a solution of thiolacetate 15 (1.93 g, 7.1
mmol) in acetone (7 mL) at rt was added by syringe methyl
iodide (6.6 mL, 106 mmol). This was allowed to stir at rt for
2.5 hand then at reflux for 1 h. Concentration in vacuo
provided the methiodide as a hygroscopic, slightly yellow foam.
This material was taken immediately to the next reaction: 1H
NMR (CD3OD, 400 MHz) δ 2.30 (s, COCH3, 3), 4.08 (s, N-CH3,
3), 4.41 (s, CH2S, 1), 4.78 (s, CH2S, 1), 5.35 (s, CH2O, 2), 7.23-
7.37 (m, ArH, 3), 7.38-7.45 (m, ArH, 3), 7.54 (d, J ) 2.9 Hz,
C3-H, 1), 8.56 (d, J )7.6 Hz, C6-H, 1); MS (FAB) (m/z) 288 (M+,
100), 214 (39); HRMS calcd for C16H18NO2S+ 288.1058, found
288.1057.
+
z) 156 (M + H+, 100), 127 (12). HRMS calcd for C7H10NO3
156.0661, found 156.0654. Anal. Calcd for C7H10NO2: C,
54.19; H, 5.85; N, 9.03. Found: C, 54.18; H, 5.88; N, 9.07.
cis-6-Hydr oxy-5-m eth yl-1-a za -3-oxa bicyclo[3.3.0]octa n -
2-on e (22). To ketone 21 (1.36 g, 8.7 mmol) in THF (50 mL)
at 0 °C was added, in portions, LiAl(O-tert-Bu)3H (2.74 g, 10.4
mmol). This reaction was allowed to stir for 2 h after which
TLC analysis indicated complete conversion to the alcohol (Rf
0.14 in 5% MeOH/CH2Cl2; KMnO4). Ether was then added,
and reaction was quenched slowly with addition of saturated
sodium potassium tartrate solution. The precipitated salts
were filtered through Celite, and the filtrate was concentrated.
Flash chromatography (5% MeOH/CH2Cl2) yielded the desired
alcohol 22 (1.31 g, 95% yield, 94% de; minor isomer separable
by chromatography) as a white solid: 1H NMR (CDCl3, 400
MHz) δ 1.30 (s, CH3, 3), 1.83-2.00 (m, Hâ of C-7, 1), 2.35 (ddt,
J ) 12.5, 9.0, 3.5 Hz, HR of C-7, 1), 3.31 (ddd, J ) 12.0, 10.1,
3.7 Hz, Hâ of C-8, 1), 3.37 (b s, OH, 1), 3.58 (ddd, J ) 12.0,
9.0, 7.6 Hz, HR of C-8, 1), 4.058 (d, J ) 9.0 Hz, Hâ of C-4, 1),
4.061 (t, J ) 8.8 Hz, HR of C-6, 1), 4.36 (d, J ) 9.0 Hz, HR of
C-4, 1); 13C NMR (CDCl3, 100 MHz) δ 17.3 (CH3), 31.8, 42.0,
66.5, 73.1, 74.9, 161.6 (CdO); MS (CI, CH4) (m/z) 159 (9), 158
+
(M+H+, 100), 85 (5); HRMS calcd for C7H12NO3 158.0817,
found 158.0820. Anal. Calcd for C7H11NO3: C, 53.49; H, 7.05;
N, 8.91. Found: C, 53.20; H, 7.25; N, 8.97.
cis-6-(Ben zyloxy)-5-m et h yl-1-a za -3-oxa b icyclo[3.3.0]-
octa n -2-on e (23). To a solution of alcohol 22 (1.18 g, 7.5
mmol) in DMF (15 mL) at room temperature was added NaH
(60% dispersion in mineral oil, 382 mg, 9.6 mmol) in one
portion. This mixture was allowed to stir for 15 min to permit
H2 gas evolution to proceed to completion. Benzyl bromide
(1.1 mL, 9.3 mmol) was then added by syringe, in the process,
warming the solution. A catalytic amount of n-BuNI (32 mg)
was added, and the mixture was allowed to stir at room
temperature for 90 min. Reaction was quenched by addition
of MeOH (5 mL). The mixture was concentrated and flash
chromatography (Rf 0.19 in 2:1 hexanes/ethyl acetate; 33-50%
EtOAc/hexanes elution; KMnO4) yielded product 23 (1.67 g,
90%) as a white solid: mp 71.0-71.5 °C; IR (cm-1, KBr pellet)
2985, 2974, 2912, 1764, 1456, 1395, 1343, 1308, 1125, 1101,
1067, 761; 1H NMR (CDCl3, 300 MHz) δ 1.33 (s, CH3, 3), 1.87-
2.02 (m, HC-7, 1), 2.30-2.44 (m, HC-7, 1), 3.32 (ddd, J ) 11.9,
10.3, 3.6 Hz, HC-8, 1), 3.62 (ddd, J ) 11.9, 9.2, 7.6 Hz, HC-8, 1),
3.79 (t, J ) 8.5 Hz, HC-6, 1), 4.04 (d, J ) 8.9 Hz, Hâ of C-4, 1),
4.20 (d, J ) 8.9 Hz, HR of C-4, 1), 4.45 (d, J ) 11.8 Hz, CH2 of
Bn, 1), 4.58 (d, J ) 11.8 Hz, CH2 of Bn, 1), 7.27-7.40 (m, Ar-
H, 5); MS (CI, CH4) (m/z ) 276 (22), 249 (21), 248 (M + H+,
100), 91 (24). Anal. Calcd for C14H17NO3: C, 68.00; H, 6.93;
N, 5.66. Found: C, 68.00; H, 6.92; N, 5.69.
Pyridinium iodide 16 was dissolved in MeOH (20 mL), and
to it was added sodium borohydride (490 mg, 13.0 mmol) in
portions at 0 °C. After gas evolution ceased, the reaction was
heated to reflux for 30 min. Flash chromatography (Rf 0.46
in 10% MeOH/CH2Cl2; 2-5% MeOH/CH2Cl2 elution; KMnO4)
yielded the tetrahydropyridine 17 and 18 as a slightly yellow
oil (1.13 g, 64%). The residue was dissolved in THF (20 mL)
and added to a suspension of lithium aluminum hydride (175
mg, 4.6 mmol) in THF (15 mL) at 0 °C. After 10 min, TLC
indicated only free thiol in solution. This reaction was diluted
with diethyl ether and quenched by addition of 1 :1 sodium
sulfate decahydrate/Celite until a white precipitate settled.
The suspension was filtered through Celite and concentrated
to give amino thiol 17 as a slightly yellow oil (887 mg, 78%):
1H NMR (CHCl3, 300 MHz) δ 1.60-1.80 (m, 1), 2.05-2.25 (m,
1), 2.39 (s, CH3, 3), 2.55-2.70 (m, 2), 2.75-3.40 (m, 3), 4.62 (s,
CdCH, 1), 4.79 (s, ArCH2, 2), 7.20-7.45 (m, ArH, 5); MS (CI,
CH4) (m/z) 250 ((M + H)+, 33), 248 (37), 202 (100), 126 (53),
91 (96), 90 (56).
Amino thiol 17 (887 mg, 3.6 mmol) was taken up in THF
(50 mL), and water (64 µL, 3.6 mmol) was added. To this
solution was added 1 N HCl in ether (7 mL, 7 mmol) at 0 °C.
This solution was allowed to stir at 0 °C for 1 h. The mixture
was then concentrated and washed with dry dietheyl ether to
give product 9‚HCl as an off-white solid (893 mg, 128%,
contaminated by BnOH and H2O). Silica gel chromatography
afforded a small sample of free amine 9: 1H NMR (CD3OD,
400 MHz) δ 1.67-1.77 (m, 1), 1.86 (d, B of AB q, J ) 11.6 Hz,
1), 1.90-2.02 (m, 1), 2.14 (dt, A of AB q, J ) 11.7, 3.7 Hz, 1),
2.29 (s, CH3, 3), 2.80-2.90 (m, 2), 2.95-3.10 (m, 2), 3.42-3.48
(m, 1); MS (CI, CH4) (m/z) 161 (26), 160 ((M + H)+, 100), 142
(32), 112 (26), 102 (39); HRMS calcd for C7H14NOS+ 160.0796,
found 160.0795.
t r a n s-3-(Be n zyloxy)-N -(b e n zyloxyca r b on yl)-2-(h y-
d r oxym eth yl)-2-m eth ylp yr r olid in e (25). According to a
modified procedure of Hirai et al.,19 a solution of 23 (1.67 g,
6.8 mmol) in 1 N KOH in MeOH (60 mL) was heated to gentle
reflux for 16 h. TLC then indicated complete conversion (Rf
0.26 in 18:2:80 MeOH:NEt3:CH2Cl2; ninhydrin). This reaction
was concentrated and flash chromatography (10-35% MeOH/
CH2Cl2) yielded trans-3-(benzyloxy)-2-(hydroxymethyl)-2-
methylpyrrolidine (24) as a clear, slightly yellow gel (1.15 g,
77%): 1H NMR (CD3OD, 400 MHz) δ 1.19 (s, CH3, 3), 1.78-
1.87 (m, HC-4, 1), 1.94-2.04 (m, HC-4, 1), 2.96-3.10 (m, HC-5
,
5-Meth yl-1-aza-3-oxabicyclo[3.3.0]octan e-2,6-dion e (21).
Compound 2011 (2.0 g, 12.7 mmol) was dissolved in CH2Cl2
(200 mL), and solid NaOAc (10.7 g, 130 mmol) was added. The
reaction vessel was then covered with aluminum foil. To this
vigorously stirred suspension was added pyridinium chloro-
chromate (PCC) (13.6 g, 63 mmol) in portions. After 3 h, TLC
indicated complete conversion of the starting alcohol to the
(19) Hirai, Y.; Terada, T., Amemiya, Y.; Momose, T. Tetrahedron
Lett. 1992, 31, 7893.
(20) The author has deposited atomic coordinates for this structure
with the Cambridge Crystallographic Data Centre. The coordinates
can be obtained, on request, from the Director, Cambridge Crystal-
lographic Data Centre, 12 Union Road, Cambridge, CB2 1EZ, UK.

Metal Complexes That Mimic Steroid Hormones
J . Org. Chem., Vol. 61, No. 8, 1996 2631
2), 3.25 (s, CH2-OH, 2), 3.68 (dd, J ) 5.6, 4.9 Hz, CH-OBn, 1),
4.18 (d, J ) 11.7 Hz, CH2 of Bn, 1), 4.30 (d, J ) 11.7 Hz, CH2
of Bn, 1), 6.92-7.04 (m, ArH, 5); 13C NMR (CD3OD, 100 MHz)
δ 15.9 (CH3), 30.1 (C-3), 43.2 (C-4), 65.0, 71.2, 73.0, 80.8, 125.0,
128.9, 129.5, 139.3; MS (CI, CH4) (m/z) 222 (M + H+, 100),
ca p t om et h yl-4-p yr id on a t o][tr a n s-3-h yd r oxyl-2-m et h yl-
2-(m er ca p tom eth yl)p yr r olid in a to]oxor h en iu m (V) (3b).
Amino thiol compound 19‚HCl (150 mg, 0.82 mmol) was
dissolved in MeOH (10 mL). To this solution were added, in
succession, amino thiol 9‚HCl (156 mg, 0.80 mmol), 1 N
NaOAc/MeOH (10 mL, 10 mmol), and trichlorobis(triphenyl-
phosphine)oxorhenium(V) (950 mg, 1.14 mmol). The suspen-
sion was then refluxed for 20 min and then allowed to cool to
rt. Concentration and flash chromatography on a cyanopropyl
silica gel column yielded an orange solid product (53.2 mg,
13%) as a mixture of diastereomers, which could be separated
by normal phase HPLC (30% (5% i-PrOH/CH2Cl2)/70% hexane
elution; 3b, tR ) 13.87 min; 3a , tR ) 17.20 min).
+
114(22), 107(29), 91(47), 61 (62); HRMS calcd for C13H20NO2
222.1494, found 222.1493.
To amino alcohol 24 (613 mg, 2.8 mmol) in 1:1 THF:H2O
(60 mL) was added K2CO3 (2.33 g, 16.9 mmol). This mixture
was stirred vigorously as benzyl chloroformate (1.7 mL, 11.9
mmol) was added via syringe at room temperature. The
reaction was allowed to stir for 1 h, after which TLC (Rf 0.32
in 40% EtOAc/hexanes; KMnO4) indicated complete conversion
of starting material. The layers were separated, and the
aqueous phase was back-extracted with CH2Cl2. The organic
layers were dried over MgSO4 and concentrated. Flash
chromatography (40% EtOAc/hexanes) yielded the title com-
pound 25 as a clear, colorless oil (835 mg, 84%): 1H NMR
(CDCl3, 300 MHz) δ 1.30 (s, CH3, 3), 1.65-1.92 (m, C4-H, 1),
2.05-2.15 (m, C4-H, 1), 3.30 (dt, J ) 7.6, 10.8 Hz, C5-H, 1),
3.55-3.70 (m, C5-H, one CH2-OH, and C3-H, 3), 3.83 (d, J )
11.7 Hz, CH2-OH, 1), 4.30 (br s, OH, 1), 4.51 (d, J ) 11.9 Hz,
CH2 of Bn, 1), 4.62 (d, J ) 11.9 Hz, CH2 of Bn, 1), 5.11 (s, CH2
of Cbz, 2), 7.25-7.43 (m, ArH, 10); 13C NMR (CDCl3, 100 MHz)
δ 14.8 (CH3), 26.4, 43.7, 66.5, 66.9, 68.0, 72.0, 80.1, 127.4,
127.7, 128.0, 128.4, 128.5, 136.5, 137.9, 155.6 (CdO); MS (CI,
CH4) (m/z) 356 (M + H+, 47), 324 (25), 313 (23), 312 (86), 276
For the cis compound 3b: 1H NMR (CD2Cl2, 400 MHz) δ
1.01 (d, J ) 0.7 Hz, 13-CH3, 3), 1.24-1.35 (m, 2R, 1), 2.04 (ddq,
J ) 12.0, 9.7, 0.7 Hz, 9â, 1), 2.16 (ddd, M of AMX, J ) 16.0,
3.8, 2.6 Hz, 4â, 1), 2.33 (s, N-CH3, 3), 2.37-2.50 (m, 2â, 9R, 2),
2.77 (dd, B of ABX, J ) 11.2, 0.7 Hz, 12â, 1), 2.82 (d, B of
ABX, J ) 13.2 Hz, 6â, 1), 2.87-3.00 (m, 5â, 1), 3.03 (dd, A of
ABX, J ) 11.0 Hz, 12R, 1), 3.27 (ddd, A of AMX, J ) 15.9,
13.0, 0.7 Hz, 4R, 1), 3.33-3.47 (m, 1â, 1), 4.15-4.32 (m, 10â,
8â, 1R, 3), 4.60 (ddd, J ) 13.6, 9.3, 7.9 Hz, 8R, 1); HRMS calcd
for C13H24N2O3S2187Re+ 507.0786, found 507.0779.
For the trans compound 3a : 1H NMR (CD2Cl2, 400 MHz) δ
0.99 (d, J ) 0.7 Hz, 13-CH3, 3), 2.01 (dq, J ) 12.0, 9.2 Hz, 1),
2.33-2.50 (m, 9â+2, 3), 2.46 (s, NCH3, 3), 2.73-2.79 (m, 1),
2.78 (d, B of AB q, J ) 13.0 Hz, 12R, 1), 2.84 (dd, B of AB q,
J ) 11.0, 0.7 Hz, 12â, 1), 2.85-3.00 (m, 9R, 1), 3.20 (d, A of
AB q, J ) 11.0 Hz, 12R, 1), 3.23 (dd, J ) 13.0, 5.0 Hz, 6R, 1),
3.95-4.02 (m, 5R, 1), 4.00-4.10 (m, 1), 4.16-4.26 (m, 2), 4.44
(ddd, J ) 13.8, 9.3, 7.8 Hz, 8â, 1), 4.52 (ddd, J ) 13.9, 6.7, 1.6
Hz, 8R, 1); HRMS calcd for C13H24N2O3S2187Re+ 507.0786,
found 507.0779.
HP LC Deter m in a tion of Sta bility. Stability of complexes
3a and 3b were determined by the following procedure.
Samples of 0.2 mg of rhenium complex, 0.2 mg estriol (as
internal standard), and 0.1 mg bovine albumin (fraction V)
were combined in a vial with 1 mL androgen receptor assay
buffer (0.01 M Tris, 0.0015 M EDTA, 0.02% NaN3, 0.01 M
thioglycerol, 20 mM sodium molybdate, 10% (v/v) glycerol; pH
7.4)16 and incubated at 0 °C. At various time intervals samples
were then extracted with CH2Cl2 (3 × 1 mL), and concentrated
to 0.5 mL, and an aliquot (10 µL) was injected into an
analytical HPLC column with isocratic elution (35% (5%
i-PrOH/CH2Cl2)/65% hexane).
An d r ogen Recep tor Bin d in g Affin ity. These values
were determined by a competitive radiometric binding assay,
using rat prostate cytosol as a source of receptor, tritium-
labeled R1881 as tracer, according to a previously published
method.16
+
(15), 248 (71), 91 (100), 79 (13); HRMS calcd for C22H26NO4
356.1862, found 356.1861. Anal. Calcd for C21H25NO4: C,
70.96; H, 7.23; N, 3.94. Found: C, 71.00; H, 7.23; N, 4.03.
tr a n s-3-(Ben zyloxy)-N-(ben zyloxyca r bon yl)-2-m eth yl-
2-[(a cetylth io)m eth yl]p yr r olid in e (26). To triphenylphos-
phine (589 mg, 2.2 mmol) in THF (30 mL) was added diethyl
azodicarboxylate (354 µL, 2.2 mmol) at 0 °C. After 30 min,
alcohol 25 (662 mg, 1.9 mmol) in THF (10 mL) was added to
this solution, followed by thiolacetic acid (170 µL, 2.3 mmol).
This mixture was allowed to stir at 0 °C for 1 h and then at rt
for 32 h. TLC showed partial conversion to product (Rf 0.39
in 20% EtOAc/hexanes; UV/KMnO4). Flash chromatography
(20-30% EtOAc/hexanes) yielded product 26 (364 mg, 87%
based on consumed starting material) and unreacted alcohol
25 (302 mg, 48%): 1H NMR (CDCl3, 400 MHz) δ 1.31 (s, CH3
of a rotamer, 1), 1.41 (s, CH3 of a rotamer, 2), 1.72-1.84 (m,
CH2CH, 1), 2.00-2.10 (m, CH2CH, 1), 2.32 (s, SCOMe, 3),
3.20-3.30 (m, NCH2, 1), 3.58 (s, CH2S, 2), 3.56-3.68 (m, NCH2,
1), 3.74 (dd, J ) 8.1, 6.4 Hz, CHOBn, 1), 4.57 (s, CH2 of Cbz,
2), 5.08-5.14 (m, CH2 of Bn, 2), 7.25-7.42 (m, Ar-H, 10); MS
(CI, CH4) (m/z) 414 (M + H+, 9), 370 (19), 354 (39), 328 (15),
324 (21), 295 (17), 264 (28), 119 (15), 91 (100); HRMS calcd
for C23H28NO4S+ 414.1739, found 414.1731. Anal. Calcd for
C23H27NO4S: C, 66.80; H, 6.58; N, 3.39; S, 7.75. Found: C,
66.40; H, 6.85; N, 3.26; S, 7.40.
Ack n ow led gm en t. This work was supported by a
grant from the Department of Energy (DE FG02
84ER60410). We are grateful to Dr. Scott Wilson and
Ms. Teresa Prussak-Wieckowski for obtaining the X-ray
crystallographic data of 4a and 4b, and Ms. Kathryn
E. Carlson for performing the binding assays. NMR
spectra at 300 and 400 MHz were obtained on instru-
ments supported by a grant from the National Institutes
of Health (PHS 1S10 RR02299) and the National
Science Foundation (CHE 90001438 EQ), respectively;
mass spectra were obtained on instruments supported
by grants from the National Institute of General Medi-
cal Sciences (GM 27029), the National Institutes of
Health (RR 01575 for FAB; RR 04648 for CI), and the
National Science Foundation (PCM 8121494 for FAB).
t r a n s-3-H yd r oxy-2-m e t h yl-2-(m e r ca p t om e t h yl)p yr -
r olid in e Hyd r och lor id e (19‚HCl). Into a solution of acetate
26 (653 mg, 1.6 mmol) in THF (5 mL), ammonia (45 mL) was
condensed at -78 °C in a dry ice-2-propanol bath. The solution
was then allowed to warm to reflux (dry ice-cold finger
condenser), at which time sodium (234 mg, 10.2 mmol) was
added in pieces. The blue solution was allowed to stir at reflux
for 20 min, and reaction was then quenched by slow addition
of degassed methanol (4 mL). Solvent was evaporated by
nitrogen overflow, and the residue was passed through a silica
gel plug (2 cm) with 20% MeOH/CH2Cl2 elution. The resulting
solution was then acidified with 1 N HCl in ether (5 mL) and
concentrated under vacuum leaving 19‚HCl as a white gum
(330 mg, 114%, contaminated with MeOH). This material was
used in the complexation reaction without further purifica-
tion: 1H NMR (CD3OD, 400 MHz) δ 1.29 (s, CH3, 3), 1.89
(dddd, J ) 13.7, 9.1, 5.8, 4.5 Hz, C4-H, 1), 2.24 (dddd, J ) 13.8,
9.2, 7.2, 5.9 Hz, C4-H, 1), 2.68 (AB q, ∆ν ) 22.0 Hz, J ) 14.7
Hz, CH2-S, 2), 3.18-3.33 (m, C5-H, 2), 4.05 (dd, J ) 5.8, 4.4
Hz, CH-OH, 1); MS (CI, CH4) (m/z) 148 ((M + H+), 100), 146
(20), 130 (68), 116 (27), 114 (67), 100 (92), 70 (42); HRMS calcd
for C6H14NOS+ 148.0796, found 148.0795.
Su p p or tin g In for m a tion Ava ila ble: Proton NMR spectra
for compounds 3a , 3b, 9, and 16 (4 pages). This material is
contained in libraries on microfiche, immediately follows this
article in the microfilm version of the journal, and can be
ordered from the ACS; see any current masthead page for
ordering information.
[tr a n s-N-Met h yl-2-(m er ca p t om et h yl)-4-p yr id on a t o]-
[t r a n s-3-h yd r oxyl-2-m e t h yl-2-(m e r ca p t om e t h yl)p yr -
r olid in a to]oxor h en iu m (V) (3a ) a n d [cis-N-Meth yl-2-m er -
J O951995K