A biomimetic approach to C-nor-D-homo-steroids

Article abstract of DOI:10.1021/ja103152k

A biomimetic three-step transformation of classical 6-6-6-5 - steroids into their C-nor-D-homo-counterparts gives an easy and fast access to this highly important substructure of natural products, as it is found in cyclopamine, and nakiterpiosin. A novel reagent combination allows for the rearrangement even of 17-keto steroids with high endoselectivity. In several examples the broadness of this strategy is outlined.

Full text of DOI:10.1021/ja103152k

Published on Web 07/02/2010
A Biomimetic Approach to C-nor-D-homo-Steroids
Philipp Heretsch, Sebastian Rabe, and Athanassios Giannis*
Institut fu¨r Organische Chemie, Fakulta¨t fu¨r Chemie und Mineralogie, UniVersita¨t Leipzig, Johannisallee 29,
D-04103 Leipzig, Germany
Received April 14, 2010; E-mail: giannis@uni-leipzig.de
Scheme 1
Abstract: A biomimetic three-step transformation of classical “6-
6-6-5”-steroids into their C-nor-D-homo-counterparts gives an
easy and fast access to this highly important substructure of
natural products, as it is found in cyclopamine, and nakiterpiosin.
A novel reagent combination allows for the rearrangement even
of 17-keto steroids with high endoselectivity. In several examples
the broadness of this strategy is outlined.
C-nor-D-homo-Steroids or [14(13f12)]-abeo-steroids constitute
a class of natural products that possess a steroid skeleton with an
unusual 6-6-5-6 ring pattern. Although they are a rather small
bond isomer 2 to byproduct 3 in almost quantitative yield and a
reaction time of less than 30 min (Scheme 1).
The Comins reagent has been used before only for the formation
subgroup compared to all known steroids, many biologically highly
active members recruit themselves from this class. Among them
are the structurally diverse Veratrum steroidal alkaloids1 with
of vinyltriflates from enolizable ketones. The use of less reactive
reagents for this rearrangement proved to be unsuccessful. Mc-
Murry’s reagent (N-phenyl triflimide)11 gave no reaction. Among
several solvents screened for this procedure12 toluene proved to
be best suited; the amount of DMAP was also varied in a broad
range showing that an excess was necessary.
cyclopamine as the most renowned one. Cyclopamine is a highly
selective antagonist of transmembrane protein Smoothened and
therefore has emerged as a novel and potent candidate for treatment
of cancers dependent on hedgehog signaling.2 Also nakiterpiosin,
a C-nor-D-homo-steroid of marine origin, proved to be a potent
antimitotic addressing the hedgehog pathway.3
With this powerful method in hand our attention was drawn to the
class of 12-hydroxy-17-keto-steroids. These compounds have withstood
all known procedures of rearrangement (Bamford-Stevens conditions
as well as Wagner-Meerwein conditions using tosylates, mesylates,
or phosphorus species13 as a leaving group) since the ketone provides
a strong deactivating influence on the C-14 migration terminus.
Typically, these substrates give only elimination products like 6 or
extensive degradation.14 When we subjected hydroxy ketone 4 to our
method we were pleased to isolate the desired C-nor-D-homo-steroid
enone 5 together with elimination byproduct 6.
Since the necessary starting materials have been shown to be
easily accessible using Scho¨necker’s C-H activation/hydroxylation
sequence,15 we started to acquire more examples of this type.
The 12ꢀ-hydroxy-dehydroepiandrosterone derivative 13 (see
Table 1) gave a similar result as 4, the corresponding i-steroid 14
(accessible in four steps from dehydroepiandrosterone) also pro-
vided the desired rearranged product. The combination of Scho¨n-
ecker’s C-H activation/hydroxylation protocol and our reagent
combination for rearrangement constitutes an efficient and biomi-
metic synthesis of the C-nor-D-homo-steroid skeletonsa building
block for the synthesis of virtually any C-nor-D-homo-steroid that
was only prepared in tedious sequences over several steps before.16
A byproduct that was isolated in small amounts in the rear-
rangement of 13 when using toluene but in higher yield when using
xylenes under reflux conditions as a solvent was triflate 7. On
treatment of purified 7 in refluxing toluene with an excess of DMAP
and Comins reagent, no rearrangement product 8 was obtained
(Scheme 2). We therefore assume that 7 is not an intermediate in
the formation of rearrangement products in this reaction. In another
A general synthetic strategy for C-nor-D-homo-steroids is still
elusive though several different strategies have been utilized
including intramolecular Horner-Wadsworth-Emmons4 and aldol
condensations5 using the (+)-Wieland-Miescher ketone as a
starting material, the Nazarov cyclization,3 and predominantly the
degradation and rearrangement of abundantly available hecogenine.6
Especially the last strategy attracted our attention since the same
rationale that is also known from the biosynthesis7 of these natural
products is applied and therefore renders this approach biomimetic.
During our work on the synthesis of cyclopamine8 we faced the
problem of rearranging the 12ꢀ-hydroxy dehydroepiandrosterone
derivative 1 to yield the desired C-nor-D-homo-system 2. Although
Hirschmann as early as 19526 and later Mitsuhashi9 reported on similar
rearrangements of hecogenine, the reaction conditions they applied
were not successful in our case because transformation of the
12-hydroxy group into a mesylate and subsequent treatment with
several bases at elevated temperature led to degradation and only traces
of the desired product. Also the use of Bamford-Stevens conditions
to the related 12-hydrazone derivative proved to be ineffective.
When we changed the leaving group for a triflate, the rearrange-
ment occurred in one pot under rather mild conditions (triflic
anhydride, pyridine, 50 °C) but to our dismay to yield the undesired
isomer 3 with an exocyclic double bond as the main product and
only small amounts of the desired endoisomer 2.
We now examined other conditions to access predominantly the
endocyclic product and report here an unprecedented procedure that
involves treatment of 12-hydroxy steroids with the Comins reagent
(N-(5-chloro-2-pyridyl) triflimide)10 and DMAP in refluxing tolu-
ene. This method leads to a ratio of 7:2 of desired endocyclic double
9
9968 J. AM. CHEM. SOC. 2010, 132, 9968–9969
10.1021/ja103152k  2010 American Chemical Society

C O M M U N I C A T I O N S
Table 1. Rearrangement of 12ꢀ-Hydroxy Steroids Using the
Scheme 3
Comins Reagenta
Hirschmann in his initial studies.6 These substrates proved to give
the rearrangement in high yields and selectivity for the endo-
products. All substrates shown in Table 1 could not be rearranged
using the various conditions reported by Hirschmann6 or Fu’s PCl5
system13 with the exception of hecogenine derivative 22 which
reacted readily under all conditions.
In conclusion, we developed a novel reagent combination for
the rearrangement of 12ꢀ-hydroxy steroids. This method accesses
otherwise difficult to obtain C-nor-D-homo-steroids and is char-
acterized by giving predominantly the endocyclic double bond
isomer and rendering for the first time available 17-keto steroids,
a species completely unreactive to this type of rearrangement before.
We therefore believe that the application of our method in total
synthesis can both simplify and shorten routes to natural products
bearing a C-nor-D-homo-steroid skeleton and help to enable
convergent strategies. Future work will address mechanistic studies
and biological testing of the obtained products.
Acknowledgment. Bayer Schering Pharma is acknowledged for
a gracious gift of steroids.
Supporting Information Available: Experimental procedures and
spectral data for all compounds. This material is available free of charge
via the Internet at http://pubs.acs.org.
a Reaction conditions: substrate (0.30 mmol), N-(5-chloro-2-pyridyl)
triflimide (0.90 mmol), DMAP (1.80 mmol), toluene (10 mL), 111 °C.
b Only the main isomer is shown. c The yields refer to isolated products
in the order endo-product:exo-product. The values within parentheses are
the yields of the elimination byproduct (∆-11-12-steroids). d No
endo-product was isolated. e Additionally, 9% of rearranged product
with a ∆-13-17 double bond was isolated.
References
(1) Li, H.-J.; Jiang, Y.; Li, P. Nat. Prod. Rep. 2006, 23, 735–752.
(2) Heretsch, P.; Tzagkaroulaki, L.; Giannis, A. Angew. Chem., Int. Ed. 2010,
49, 3418–3427, and citations therein.
(3) Gao, S.; Wang, Q.; Huang, L. J.-S.; Lum, L.; Chen, C. J. Am. Chem. Soc.
2010, 132, 371–383.
(4) Guile, S. D.; Saxton, J. E.; Thornton-Pett, M. J. Chem. Soc., Perkin Trans.
1 1992, 1763–1767.
(5) Brown, E.; Lebreton, J. Tetrahedron 1987, 43, 5827–5840.
(6) (a) Hirschmann, R.; Snoddy, C. S., Jr.; Wendler, N. L. J. Am. Chem. Soc.
1952, 74, 2693–2694. (b) Hiskey, C. F.; Hirschmann, R.; Wendler, N. L.
J. Am. Chem. Soc. 1953, 75, 5135–5136. (c) Hirschmann, R.; Snoddy, C. S.,
Jr.; Hiskey, C. F.; Wendler, N. L. J. Am. Chem. Soc. 1954, 76, 4013–
4025.
Scheme 2
(7) Kaneko, K.; Mitsuhashi, H.; Hirayama, K.; Yoshida, N. Phytochemistry
1970, 9, 2489–2495.
(8) (a) Giannis, A.; Heretsch, P.; Sarli, V.; Sto¨ssel, A. Angew. Chem., Int. Ed.
2009, 48, 7911–7914. (b) Heretsch, P.; Rabe, S.; Giannis, A. Org. Lett.
2009, 11, 5410–5412.
(9) (a) Mitsuhashi, H.; Shimizu, Y. Tetrahedron 1963, 19, 1027–1035. (b)
Mitsuhashi, H.; Shibata, K.; Sato, T.; Shimizu, Y. Chem. Pharm. Bull. 1964,
12, 1–5.
(10) Comins, D. L.; Dheghani, A. Tetrahedron Lett. 1992, 33, 6299–6302.
(11) McMurry, J. E.; Scott, W. J. Tetrahedron Lett. 1983, 24, 979–982.
(12) Solvents that gave no or a very slow reaction: benzene, dichloroethane,
nitromethane, n-octane; solvents that gave similar or slightly lower yields:
chlorobenzene, pyridine, xylenes.
(13) Fu, X.; Chan, T.-M.; Tann, C.-H.; Thiruvengadam, T. K. Steroids 2002,
67, 549–554.
(14) On treatment of 13 with PCl5 in THF at-78 °C, no elimination occurred
but 3-benzyl-12ꢀ-(4-chlorobutoxy)-dehydroepiandrosterone (24) was ob-
tained in 50% yield, presumably by ring opening of a molecule of solvent
facilitated by PCl5. A similar observation using SOCl2 at high temperatures
has been reported: Ferrari, T.; Vogel, P. Synlett 1991, 233–234.
(15) Scho¨necker, B.; Zheldakova, T.; Liu, Y.; Ko¨tteritzsch, M.; Gu¨nther, W.;
Go¨rls, H. Angew. Chem., Int. Ed. 2003, 42, 3240–3244.
(16) (a) Johns, W. F.; Salamon, K. W. J. Org. Chem. 1979, 44, 958–961. (b)
Green, M. J.; Abraham, N. A.; Fleischer, E. B.; Case, J.; Fried, J. Chem.
Commun. 1970, 234–235. (c) Johnson, W. S.; Cox, J. M.; Graham, D. W.;
Whitlock, H. W., Jr. J. Am. Chem. Soc. 1967, 89, 4524–4526.
(17) Marko, I. E.; Ates, A.; Gautier, A.; Leroy, B.; Plancher, J.-M.; Quesnel,
Y.; Vanherck, J.-C. Angew. Chem., Int. Ed. 1999, 38, 3207–3209.
control experiment we subjected the substrates to the action of triflic
acid in toluene under reflux conditions. In all cases but 4 and 13,
only excessive degradation was observed. 13 gave the elimination
product 34 (see Supporting Information) while 4 remained un-
changed. Thus, a mechanism based on acid induced cation formation
does not seem to be responsible for the observed reaction outcome.
We also examined other substituents in position 17 and thereby
explored the functional group tolerance of our method. Interestingly,
by converting the 17-keto group into an acetal (like 9) and treating
these substrates under our conditions we could isolate both
rearranged acetals 10 and 11 with a slight prevalence of the
exocyclic double bond isomer 10. On careful treatment with cerium
ammonium nitrate in a pH 9 buffer solution,17 we were able to isolate
the corresponding enones 12 and 5, respectively (Scheme 3).
Further examples included progesterone derivative 18, DHEA-
lactone 20 with inverted stereochemistry at C-17 with regard to
lactone 1, and hecogenine derivative 22 similar to the one used by
JA103152K
9
J. AM. CHEM. SOC. VOL. 132, NO. 29, 2010 9969