The preparation of (-)-grandisine B from (+)-grandisine D; A biomimetic total synthesis or formation of an isolation artefact?

Article abstract of DOI:10.1021/ol2014939

An efficient new alkyne-acetal cyclization procedure has been developed to prepare enantiopure indolizidine building blocks from l-proline and then applied to prepare the Elaeocarpus-derived alkaloids grandisine B and grandisine D in an efficient manner.

Full text of DOI:10.1021/ol2014939

ORGANIC
LETTERS
2011
Vol. 13, No. 15
3976–3979
The Preparation of (ꢀ)-Grandisine B from
(þ)-Grandisine D; A Biomimetic Total
Synthesis or Formation of an Isolation
Artefact?
James D. Cuthbertson,† Andrew A. Godfrey,‡ and Richard J. K. Taylor*,†
Department of Chemistry, University of York, Heslington, York YO10 5DD, U.K., and
AstraZeneca, Chemical Sciences, Pharmaceutical Development, Silk Road Business
Park, Charter Way, Macclesfield, Cheshire SK10 2NA, U.K.
rjkt1@york.ac.uk
Received June 2, 2011
ABSTRACT
An efficient new alkyne-acetal cyclization procedure has been developed to prepare enantiopure indolizidine building blocks from L-proline and
then applied to prepare the Elaeocarpus-derived alkaloids grandisine B and grandisine D in an efficient manner. However, evidence is presented
which indicates that grandisine B does not occur naturally but is formed by reaction of grandisine D with ammonia during the extraction/
purification process.
The Elaeocarpaceae plant family has been the source of
numerous, structurally diverse alkaloids over the years.1
Carroll and co-workers recently reported the isolation of
the Elaeocarpus-derived indolizine alkaloids grandisines A
1 and B 2 (Figure 1) from the Australian rainforest tree
Elaeocarpus grandis as part of a high throughput drug
discovery program.2a Subsequently, further studies re-
vealed the presence of five additional members of this
family, grandisines CꢀG (3ꢀ7).2b These indolizidine alka-
loids have attracted considerable attention as they display
human δ-opioid receptor affinity.2 First, Danishefsky and
Maloney reported a total synthesis of (þ)-grandisine A 1,3
and then Tamura’s group published a total synthesis of
(þ)-grandisine D 4 (as its TFA salt).4 Most recently, the
latter group reported the conversion of grandisine B 2 into
grandisine D 4 by the double addition of ammonia5
(following the biosynthetic proposal by Carroll2).
Our interest also focused on grandisine B 2, particularly
in view of its structural novelty; it contains an unprece-
dented combination of both indolizidine and isoquinucli-
dinone units linked by a Csp2ꢀCsp2 bond. Although many
indolizidine alkaloids have been isolated,6 isoquinuclidi-
none alkaloids are very rare indeed. In fact, mearsine 8,7,8
isolated from another member of the Elaeocarpaceae
family, Peripentadenia mearsii, appears to be the only
other published example. We recently reported a facile
route to a range of isoquinuclidinones from 6-acyl-cyclo-
hex-2-enones, employing aqueous ammonia in a one-pot
tandem amination/imination procedure, and applied this
† University of York.
‡ AstraZeneca.
(1) Johns, S. R.; Lamberton, J. A. The Alkaloids; Manske, R. H. F.,
Ed.; Academic Press: New York, 1993; Vol. 14, p 324.
(2) (a) Carroll, A. R.; Arumugan, G.; Quinn, R. J.; Redburn, J.;
Guymer, G.; Grimshaw, P. J. Org. Chem. 2005, 70, 1889. (b) Katavic,
P. L.; Venables, D. A.; Forster, P. I.; Guymer, G.; Carroll, A. R. J. Nat.
Prod. 2006, 69, 1295.
(5) Kurasaki, H.; Okamoto, I.; Morita, N.; Tamura, O. Chem.;Eur.
J. 2009, 15, 12754.
(6) For a recent review, see: Michael, J. P. Beilstein J. Org. Chem.
2007, 3, 27.
(3) Maloney, D. J.; Danishefsky, S. J. Angew. Chem., Int. Ed. 2007,
46, 7789.
(4) Kurasaki, H.; Okamoto, I.; Morita, N.; Tamura, O. Org. Lett.
2009, 11, 1179.
(7) Robertson, G. B.; Tooptakong, U.; Lamberton, J. A.; Geewananda,
Y. A.; Gunawardana, P.; Bick, I. R. C. Tetrahedron Lett. 1984, 25,
2695.
(8) Crouse, J. R.; Pinder, A. R. J. Nat. Prod. 1989, 52, 1227.
r
10.1021/ol2014939
Published on Web 06/30/2011
2011 American Chemical Society

Scheme 1. Synthesis of Indolizidine Alcohol 17
Figure 1. Structure of grandisines and mearsine.
utimately from alkyne 12 which we felt would be readily
accessible from L-proline.
methodologyaspartofanefficient synthesis of (ꢀ)-mearsine
8.9 Herein, we report an efficient new route to grandisine D 4
and its conversion into grandisine B 2 (with confirmation by
X-ray crystallography); we also discuss the likely nonbiolo-
gical origin of grandisine B 2.
Our retrosynthetic approach is illustrated in Figure 2.
Disconnection of the isoquinuclidinone moiety of grand-
isine B 2 by an imination/amination sequence via amine 9
leads back to grandisine D 4. The use of an aldol/oxidation
sequence involving enolate 10 derived from (S)-5-methyl-
cyclohexenone10 then leads to the requirement for indoli-
zidine 11.
The preparation of the key indolizidine building block is
shown in Scheme 1. Thus, alkyne 13 was readily accessed on a
multigram scale from commercially available N-Boc-prolinol
in ca. 70% overall yield (3 steps) using known procedures.12
Boc deprotection followed by N-alkylation with iodo-acetal
14 proceeded readily to afford acetal 15 in 79% yield over the
two steps. Thioalkynes are known to undergo hydration in
acidic media to afford the corresponding thioester,13 and
therefore, deprotonation of alkyne 15 with n-butyllithium
and trapping with ethyl disulfide afforded the cyclization
precursor 12 in 85% yield. Pleasingly, on heating a solution of
thioalkyne 12 in formic acid,14 clean cyclization was observed
giving thioester 16 as the sole product and as a single
enantiomer {[R]D ꢀ87 (c 0.97, CHCl3)}. This is an extremely
efficient route (7 steps from N-Boc-prolinol, ca. 50% overall,
unoptimised yield) which appears to be simple, robust, and
scaleable. In order to prepare the desired allylic alcohol 17,
thioester 16 was converted into the corresponding methyl
ester and then reduced using DibalH15 (direct reduction of
the thioester proved problematic).
(9) Cuthbertson, J. D.; Godfrey, A. A.; Taylor, R. J. K. Tetrahedron
Lett. 2011, 52, 2024. See also: Cuthbertson, J. D.; Godfrey, A. A.;
Taylor, R. J. K. Synlett 2010, 2805.
(10) (a) (S)-5-Methylcyclohexenone is available via a number of
procedures; see: Carpenter, R. D.; Fettinger, J. C.; Lam, K. S.; Kurth,
M. J. Angew. Chem., Int. Ed. 2008, 47, 6407 and ref 10b for lead references.
(b) Carlone, A.; Marigo, M.; North, C.; Landa, A.; Jorgensen, K. A. Chem.
Commun. 2006, 4928.
(11) (a) Gonzalez-Rodriguez, C.; Escalante, L.; Varela, J. A.; Castedo,
L.; Saa, C. Org. Lett. 2009, 11, 1531. (b) Xu, T.; Yang, Q.; Yu, Z.; Li,
D.; Dong, J.; Li, Y. Chem.;Eur. J. 2010, 16, 9264 and references
therein.
(12) Paul, A.; Bitterman, H.; Gmeiner, P. Tetrahedron 2006, 62, 8919
and references therein.
(13) Maruyama, H.; Shiozaki, M.; Oida, S. Tetrahedron Lett. 1985,
26, 4521.
Figure 2. Retrosynthetic analysis of grandisine B 2.
(14) Menashe, N.; Shvo, Y. J. Org. Chem. 1993, 58, 7434.
(15) Cordero, F. M.; Anichini, B.; Goti, A.; Brandi, A. Tetrahedron
1993, 49, 9867.
Given the recent interest in the cyclization reactions of
alkynyl aldehydes/acetals,11 we envisaged preparing 11
Org. Lett., Vol. 13, No. 15, 2011
3977

We were now in a position to explore the final steps of
the grandisine synthesis (Scheme 2). Swern oxidation of
alcohol 17 gave aldehyde 11 which was reacted with the
lithium enolate derived from (S)-5-methyl-cyclohexenone
10 (90:10 er; obtained from tert-butyl acetoacetate and
crotonaldehyde by an organocatalytic procedure).10b This
aldol reaction produced allylic alcohol 19 along with the
diastereoisomer 18 derived from (R)-5-methyl-cyclohexe-
none, which was readily removed by chromatography.
Alcohol 19 was obtained as a single diastereomer, tenta-
tively assigned as the trans-isomer about the ring. Oxida-
tion using Swern conditions gave grandisine D 4 in
80% yield {[R]D þ73.7 (c 0.1, MeOH); published values:
Scheme 3. Synthesis of Grandisine B 2a
2
4,5
[R]D þ34.6 (c0.09, MeOH); [R]D þ65.7 (c0.09, MeOH)}.
Compound 4 was assigned as the trans-isomer, as evi-
denced by the large coupling constant (11.5 Hz) although
minor traces of the enol tautomer/cis-isomer were also
observed in the 1H NMR spectrum.
a X-ray structure of compound 2 (picric acid)2 depicted using
ORTPE-3 (CCDC 815228); picrate anions omitted for clarity.
3
partitioned with CH2Cl2.”2 These conditions were ex-
tremely close to the ones we had employed for the
conversion of grandisine D 4 into grandisine B 2; we
therefore conjectured about the origin of grandisine B
2. One distinct possibility appeared to be that grand-
isine D 4 is a true natural product but that on extraction
using ammonia it is converted into grandisine B 2. If
true, then grandisine B 2 is not a natural product and is
actually an artefact of the extraction procedure.16 To
gain greater understanding, we contacted Professor
Carroll who replied stating “Yes, we have certainly
speculated about whether some of these compounds
might be artefacts of the extraction and purification
process. Grandisines B, F, and G in particular are not
observed by (þ) ESI MS in crude methanol extracts of
the leaves suggesting that these compounds at least are
artefacts formed on treatment with ammonia.” It
would therefore appear that grandisine B 2 is not
naturally occurring but is formed by reaction of grand-
isine D 4 with ammonia during the extraction/purifica-
tion process.
In summary, an efficient new alkyne cyclization proce-
dure has been developed to prepare enantiopure indolizi-
dine building blocks from L-proline. Using this
methodology, the natural product grandisine D 4 has been
prepared in an efficient manner (9 steps, 14% overall yield
from the known alkynyl-pyrrolidine 13; 13 steps, 10%
overall yield from prolinol); this route compares well with
the procedure recently published by Tamura et al. (15
steps, 12% overall yield from (S)-malic acid).4
Scheme 2. Synthesis of Grandisine D 4
With grandisine D 4 in hand, we were in a position to
investigate the one-pot tandem amination/imination
sequence to generate grandisine B 2 (Scheme 3). This
reaction had also been utilized by Tamura’s group in
their studies.5 Upon treatment of diketone 4 with 35%
aq ammonia, the target compound 2 was obtained
stereoselectively in 72% yield; key spectroscopic data
were consistent with those published (see Supporting
Information). The isolation paper2 and the publication
by Tamura et al.5 were inconsistent concerning the
optical rotation data {[R]D þ11 (c 0.1, CH2Cl2)2; [R]D ꢀ159
(c 0.08, CH2Cl2)5}. Our sample had [R]D ꢀ177.5
(c 0.08, CH2Cl2), which agreed well with Tamura’s
value. A further indication of the purity was that
compound 2, as its dipicrate salt, was readily crystal-
lized and, for the first time, an X-ray crystal structure
was obtained which fully confirmed the published
structure (Scheme 3).
In addition, a tandem imination/amination sequence
has been employed for the assembly of the isoquinuclidi-
none moiety in the conversion of grandisine D 4 into
grandisine B 2 (and the first X-ray of grandisine B as its
dipicrate salt has been obtained). Perhaps most
On close reading of the original publications describing
the isolation of grandisine B, our attention was drawn to
the extraction conditions: “The aqueous layer (400 mL)
was basified with 27% NH4OH (2 ꢁ 200 mL) and
(16) For other examples of ammonia-induced artefact formation
during the isolation of natural products, see: (a) Wenkert, E.; Fuchs,
A.; McChesney, J. D J. Org. Chem. 1965, 30, 2931 (rosmaricine). (b)
Suliman, M.; Martin, M. -T.; Pais, M.; Hadi, H. A.; Awang, K.
Phytochem. 1998, 49, 2192 (desmosine).
3978
Org. Lett., Vol. 13, No. 15, 2011

significantly, evidence is presented which indicates that
grandisine B 2 does not occur naturally but is formed by
reaction of grandisine D 4 with ammonia during the
extraction/purification process.
Professor Tony Carroll (Griffith University, Brisbane,
Australia) for valuable discussions.
Supporting Information Available. Experimental pro-
1
cedures, characterization data, and H and 13C NMR
spectra for all novel compounds. Crystallographic data
for 2 (CCDC815228). This material is available free of
charge via the Internet at http://pubs.acs.org.
Acknowledgment. We are grateful to the EPSRC and
AstraZeneca for studentship support (J.D.C.) and to
Org. Lett., Vol. 13, No. 15, 2011
3979