Determining a synthetic approach to pierisformaside C

Article abstract of DOI:10.1021/ol202147r

An efficient route detailing the construction of the central core of pierisformaside C, the first grayanane-type diterpene to possess three central double bonds, is reported.

Full text of DOI:10.1021/ol202147r

ORGANIC
LETTERS
2011
Vol. 13, No. 19
5286–5289
Determining a Synthetic Approach to
Pierisformaside C
Sharon Chow, Christoph Kreß, Nanna Albæk, Carsten Jessen, and
Craig M. Williams*
School of Chemistry and Molecular Biosciences, University of Queensland,
Brisbane 4072, Queensland, Australia
c.williams3@uq.edu.au
Received August 8, 2011
ABSTRACT
An efficient route detailing the construction of the central core of pierisformaside C, the first grayanane-type diterpene to possess three central
double bonds, is reported.
Pieris formosa D. Don (Ericaceae) is a well-known
poisonous evergreen shrub found in southwest China.1 It
has been reported that poultry fall into a coma after
accidental consumption of the leaves, and even the tradi-
tional Chinese name “Mei-Li-Ma-Zui-Mu” described
horse enebria after consumption.1 The juice of the fresh
leaves has been used in folk medicine as an insecticide or as
a lotion for the treatment of tinea and scabies.1
It was to this latter feature that we were most attracted. We
reasoned that if a route to 1 could be forged, access to closely
related family members (2,3 3,4 4,5 52,6),7 some of which
display prominent biological activity,6,8 might be possible via
judicious implementation of hydration reactions (Figure 1).
To date, most grayanane-type diterpenes have been
isolated from this species. However, in 2000, Qin isolated
pierisformaside C (1),2 which remains the only example
possessing three double bonds in a grayanane skeleton.
(1) Chen, J. S.; Zheng, S. Chinese Poisonous Plants; Science Press:
Beijing, 1987.
(2) Wang, L.-Q.; Chen, S.-N.; Cheng, K. F.; Li, C.-J.; Qin, G.-W.
Phytochemistry 2000, 54, 847.
(3) Wang, L.-Q.; Qin, G.-W.; Chen, S.-N.; Li, C.-J. Fitoterapia 2001,
72, 779.
(4) Sakakibara, J.; Shirai, N.; Kaiya, T.; Nakata, H. Phytochemistry
1979, 18, 135.
(5) (a) Sakakibara, J.; Shirai, N.; Kaiya, T.; Iitaka, Y. Phytochem-
istry 1980, 19, 1495. (b) Wang, L.-Q.; Ding, B.-Y.; Wang, P.; Zhao,
W.-M.; Qin, G.-W. Nat. Prod. Sci. 1998, 4, 68. (c) Wang, L.-Q.; Ding, B.-Y.;
Zhao, W.-M.; Qin, G.-W. Chin. Chem. Lett. 1998, 9, 465.
(6) Zhang, H.-P.; Wang, L.-Q.; Qin, G.-W. Bioorg. Med. Chem. 2005,
13, 5289.
(7) For additional related family members, including aglycons not
containing C14 hydroxylation, see: (a) Wang, L.-Q.; Chen, S.-N.; Qin,
G.-W.; Cheng, K.-F. J. Nat. Prod. 1998, 61, 1473. (b) Sakakibara, J.;
Kaiya, T.; Shirai, N. Yakugaku Zasshi 1980, 100, 540. (c) Kaiya, T.;
Sakakibara, J. Chem. Pharm. Bull. 1985, 33, 4637. (d) Sakakibara, J.;
Shirai, N. Phytochemistry 1980, 19, 2159. (e) Sakakibara, J.; Ikai, K.;
Yasue, M. Yakugaku Zasshi 1974, 94, 1534.
Figure 1. Diterpene glucosides 1À5 and grayanotoxins 6 and 7.
r
10.1021/ol202147r
Published on Web 09/12/2011
2011 American Chemical Society

Surprisingly, these compounds and their aglycons have
received only sparse synthetic attention. For example,
there is only one article in the C-14 deoxy series (e.g., 1À6)
describing a partial synthesis of 14-deoxygrayanotoxin III
(6).7e In the related C-14 oxy series, only four articles
discussing the synthesis of three family members, grayano-
toxin (partial synthesis),9 grayanotoxin II (relay synthesis),10
and (À)-grayanotoxin III (7)11,12 (total synthesis), have
been published.
Considering the biological activity, limited synthetic
attention, and possibility of accessing a number of the
pierisformaside family members, we devised and executed
a plausible route to a common advanced intermediate, the
results of which are reported herein.
Our retrosynthetic analysis of 1 was based on late-stage
construction of the central seven-membered ring (8), ob-
tained via an aldol or Claisen cyclization and controlled by
the cis-double bond seen in 9 (Scheme 1). The key inter-
mediate 9 arises from a five-membered ring left-hand
fragment and a bicyclo[3.2.1]octane right-hand fragment.
both bicyclo[3.2.1]octane systems (10 and 11) would be
available from methyl 3-methoxybenzoate 15 (Scheme 1).
Ofthe twoproposed avenues for arriving atintermediate
9, we decided to pursue the Sonogashira route, as previous
experience with bicyclo[3.2.1]octanes of type 11 and silver-
(I) acetylide chemistry had not proven fruitful.13 However,
we found that construction of the bicyclo[3.2.1]octane 10,
in the first instance taking the form of a TBS-protected
alcohol at position 3 (i.e., 19), was readily achievable from
17 obtained via Birch reduction/alkylation of 15 and sub-
sequent radical cyclization of 16 as reported14 (Scheme 2).
Treatment of 17 with sodium borohydride followed by
TBS protection and reduction of the ester with diisobutyl-
aluminium hydride afforded 18 [endo/exo (3:1)] in 64%
yield over the three steps. Oxidation with TPAP15 followed
by treatment with the SeyferthÀGilbert reagent 2016
afforded 19 in 58% yield (Scheme 2).
Scheme 2. Synthesis of Bicyclo[3.2.1]octane Derivatives 19 and
27
Scheme 1. Retrosynthetic Analysis for Pierisformaside C (1)
The synthesis of 23 started with diketone 13 (Scheme 3).17
Sodium borohydride reduction of 13 in water/THF18
followed by monobenzoylation gave 21. Ketone 22 was
then obtained via TPAP oxidation (Scheme 3). At this
juncture, multiple options were available in terms of
building in the carbonyl functionality required for the
seven-membered ring closure. However, our choice was
Two options are potentially available to construct 9: (1)
Sonogashira coupling of bicyclo[3.2.1]octane 10 to vinyl
halide 12 followed by partial reduction or (2) bridgehead
alkylation of bicyclo[3.2.1]octane 11 with a suitably func-
tionalized silver(I) acetylide (14) again followed by partial
reduction. Both five-membered ring systems, 12 and 14,
could conceivably be accessed via diketone 13, whereas
(13) (a) Pouwer, R. H.; Harper, J. B.; Vyakaranam, K.; Michl, J.;
Williams, C. M.; Jessen, C. H.; Bernhardt, P. V. Eur. J. Org. Chem. 2007,
2, 241. (b) Pouwer, R. H.; Williams, C. M.; Raine, A. L.; Harper, J. B.
Org. Lett. 2005, 7, 1323. (c) Pouwer, R. H.; Williams, C. M. In Silver in
Organic Chemistry; Harmata, M., Ed.; Wiley: Hoboken, 2010; Chapter 1, p 1.
(14) Marinovic, N. N.; Ramanathan, H. Tetrahedron Lett. 1983, 24,
1871.
(8) Shirai, N.; Sakakibara, J.; Kaiya, T.; Kobayashi, S.; Hotta, Y.;
Takeya, K. J. Med. Chem. 1983, 26, 851.
(9) Hamanaka, N.; Matsumoto, T. Tetrahedron Lett. 1972, 3087.
(10) Gasa, S.; Hamanaka, N.; Matsunaga, S.; Okuno, T.; Takeda,
N.; Matsumoto, T. Tetrahedron Lett. 1976, 553.
(11) Kan, T.; Hosokawa, S.; Nara, S.; Oikawa, M.; Ito, S.; Matsuda,
F.; Shirahama, H. J. Org. Chem. 1994, 59, 5532.
(15) Ley, S. V.; Norman, J.; Griffith, W. P.; Marsden, S. P. Synthesis
1994, 639.
€
(16) For a short overview, see: Roth, G. J.; Liepold, B.; Muller, S. G.;
Bestmann, H. J. Synthesis 2004, 59.
(17) Jenkins, T. J.; Burnell, D. J. J. Org. Chem. 1994, 59, 1485.
ꢀ
(18) Molander, G. A.; Huerou, Y. L.; Brown, G. A. J. Org. Chem.
2001, 66, 4511.
(12) See also,Kan, T.; Matsuda, F.; Yanagiya, M.; Shirahama, H.
Synlett 1991, 391.
Org. Lett., Vol. 13, No. 19, 2011
5287

eventually determined by the chemistry performed. For
example, PMB protection proved difficult, and installation
of aldehyde functionality R to the ketone was not possible.
Therefore, ketone 22 was converted, with ease, into triflate
23 via sequential treatment with LDA and Mander’s
reagent19 and LDA/triflic anhydride (Scheme 3).
Reflecting on the failure depicted in Scheme 4, we
realized that a protecting group was probably not an
absolute requirement, and therefore, we decided to investigate
the use of bicyclo[3.2.1]octane 27, also readily accessible
from 17 via 28 (Scheme 2). Sonogashira coupling of 27 to
23 proceeded without adverse incident giving 29 in 72%
yield. In light of previous difficulties in our laboratories
with Lindlar reductions22 (Scheme 4 being an exception),
we decided to implement methodology described by Wei et al23
utilizing palladium acetate and sodium methoxide
(hydrogen source). In a fortuitous event, it was discovered
that when using the conditions of Wei et al., the advanced
intermediate 32 was obtained in a one-pot cascade.
Although the timing of events is unclear, mechanistically,
sodium methoxide removed the protecting group (i.e., 31),
partially reduced the triple bond (i.e., 31), and facilitated
the Claisen cyclization giving the seven-membered ring
system (Scheme 5). Unfortunately, the one-pot cascade,
which afforded only 10% yield of 32 along with over-
reduction and isomerization side products, could not
be further improved. If, however, a carefully controlled
Lindlar reduction was performed and the crude material
Scheme 3. Construction of the Left Hand Fragment 23
With both fragments in hand (i.e., 19 and 23), the
Sonogashira coupling20 was performed using standard
conditions giving the coupled product 24 in 70% yield.
Lindlar reduction proceeded remarkably smoothly, most
likely due to the steric congestion limiting over-reduction,
which afforded the (Z)-diene (25) in 60% yield. Unfortu-
nately, all attempts to remove the TBS protecting group21
from 25 either returned starting material or gave material
tentatively assigned as 26 (Scheme 4).
Scheme 5. Construction of the Advanced Intermediate 32
Scheme 4. Sonogashira Coupling of 23 and 19: Exploring the
Deprotection of 25
treated with sodium methoxide, the yield over two steps
increased to30%. Other Lindlarcatalyst systems including
P2ÀNi24 failed to offer any improvement.
Not satisfied with the above result (i.e., Scheme 5) and
realizing that the bicyclic[3.2.1] exo methylene double
(19) Mander, L. N.; Sethi, S. P. Tetrahedron Lett. 1983, 24, 5425.
(20) Negishi, E.-i.; Anastasia, L. Chem. Rev. 2003, 103, 1979.
(21) We have previously experienced difficulties with late-stage de-
protection of TBS; see, for example: Schwartz, B. D.; Denton, J. R.;
Bernhardt, P. V.; Davies, H. M. L.; Williams, C. M. Synthesis 2009,
2840.
(22) Pouwer, R. H.; Schill, H.; Williams, C. M.; Bernhardt, P. V. Eur.
J. Org. Chem. 2007, 4699.
(23) Wei, L.-L.; Wei, L.-M.; Pan, W.-B.; Leou, S.-P.; Wu, M.-J.
Tetrahedron Lett. 2003, 44, 1979.
(24) See, for example: Feutrill, J. T.; Lilly, M. J.; White, J. M.;
Rizzacasa, M. A. Tetrahedron 2008, 64, 4880 and references therein.
5288
Org. Lett., Vol. 13, No. 19, 2011

It was hoped that this tactical maneuver would avoid over-
reduction products and bring the advanced intermediate
closer to the target molecule (1, Figure 1). Treatment of 29
with m-chloroperbenzoic acid gave the desired material 33
as the major (6:1) product in 56% yield (71% brsm).
Subjecting 33 to the two-step procedure of Lindlar reduc-
tion followed by exposure to sodium methoxide gave the
deprotected cyclized advanced intermediate 34 in an ac-
ceptable yield of 47% (Scheme 6). If, however, treatment
with methoxide was performed for an extended period,
intramolecular epoxide ring-opening was observed (i.e.,
35) as the major product.
Scheme 6. Optimized Reduction/Cyclization Giving
Advanced Intermediate 34 toward a Total Synthesis of
Pierisformaside C (1)
In conclusion, an advanced intermediate 34 en route to
pierisformaside C (1) was forged in 10 linear steps (15 in
total) laying the way for an asymmetric total synthesis and
access to closely related family members (i.e., 2À6).
Acknowledgment. We thank the University of Queens-
land for financial support.
Supporting Information Available. Experimental pro-
1
cedures, compound characterization, and copies of H
bond in 29 was surprisingly susceptible to reduction under
Lindlar conditions, it was replaced with an epoxide ring.
and 13C NMR spectra. This material is available free of
charge via the Internet at http://pubs.acs.org.
Org. Lett., Vol. 13, No. 19, 2011
5289