Stereoselective synthesis of the glycosidase inhibitor australine through a one-pot, double-cyclization strategy

Article abstract of DOI:10.1021/ol062570v

(Chemical Equation Presented) A stereocontrolled, convergent synthesis of the alkaloid australine, a glycosidase inhibitor of the pyrrolizidine class, is described. The chiral starting materials were ketone 3, derived from L-erythrulose, and α-alkoxy aldehyde 4, prepared from L-malic acid. A key step of the synthesis was the highly stereoselective aldol reaction between 4 and a Z boron enolate derived from 3. Another key step was the one-pot construction of the bicyclic pyrrolizidine system by means of a three-step sequence of S_N2 displacements induced by benzylamine on a trimesylate precursor.

Full text of DOI:10.1021/ol062570v

ORGANIC
LETTERS
2007
Vol. 9, No. 1
77-80
Stereoselective Synthesis of the
Glycosidase Inhibitor Australine through
a One-Pot, Double-Cyclization Strategy†
Celia Ribes,‡ Eva Falomir,*,‡ Miguel Carda,‡ and J. Alberto Marco*,§
Departament de Qu´ımica Inorga´nica y Orga´nica, UniVersitat Jaume I, Castello´n,
E-12080 Castello´n, Spain, and Departament de Qu´ımica Orga´nica, UniVersitat de
Valencia, E-46100 Burjassot, Valencia, Spain
alberto.marco@uV.es
Received October 18, 2006
ABSTRACT
A stereocontrolled, convergent synthesis of the alkaloid australine, a glycosidase inhibitor of the pyrrolizidine class, is described. The chiral
starting materials were ketone 3, derived from -erythrulose, and -alkoxy aldehyde 4, prepared from -malic acid. A key step of the synthesis
L
r
L
was the highly stereoselective aldol reaction between 4 and a Z boron enolate derived from 3. Another key step was the one-pot construction
of the bicyclic pyrrolizidine system by means of a three-step sequence of SN2 displacements induced by benzylamine on a trimesylate precursor.
Pyrrolizidine alkaloids have been isolated from species of
several plant families.1 Polyhydroxylated representatives have
been found in some genera belonging to the Leguminosae
and a few other families. For instance, australine 1 and
alexine 2 (Figure 1), its epimer at C-7a, were isolated from
and of some glycoprotein-processing enzymes.4 Furthermore,
it displays anti-HIV activity.5 The diverse array of potentially
useful biological activities6 and the high degree of function-
ality embedded in these alkaloids make them attractive targets
for total synthesis.7,8
There has been formerly some confusion in the literature
about the true structure of australine. While that originally
(1) For general reviews on pyrrolizidine alkaloids, see: Liddell, J. R.
Nat. Prod. Rep. 2002, 19, 773-781 and the previous papers in this series.
(2) (a) Molyneux, R. J.; Benson, M.; Wong, R. Y.; Tropea, J. E.; Elbein,
A. D. J. Nat. Prod. 1988, 51, 1198-1206. (b) An X-ray diffraction analysis
of australine tetraacetate has been reported: Gainsford, G. J.; Furneaux, R.
H.; Mason, J. M.; Tyler, P. C. Acta Crystallogr. Sect. C 2000, C56, E477-
E478.
(3) Nash, R. J.; Fellows, L. E.; Dring, J. V.; Fleet, G. W. J.; Derome, A.
E.; Hamor, T. A.; Scofield, A. M.; Watkin, D. J. Tetrahedron Lett. 1988,
29, 2487-2490.
Figure 1. Structures of australine 1 and alexine 2.
(4) (a) Tropea, J. E.; Molyneux, R. J.; Kaushal, G. P.; Pan, Y. T.;
Mitchell, M.; Elbein, A. D. Biochemistry 1989, 28, 2027-2034. (b) Zeng,
Y. C.; Elbein, A. D. Arch. Biochem. Biophys. 1998, 355, 26-34.
(5) Vlietinck, A. J.; De Bruyne, T.; Apers, S.; Pieters, L. A. Planta Med.
1998, 64, 97-109. See also: Taylor, D. L.; Nash, R. J.; Fellows, L. E.;
Kang, M. S.; Tyms, A. S. AntiViral Chem. Chemother. 1992, 3, 273-277.
(6) For further recent reviews on the chemistry and biology of pyrrolizi-
dine alkaloids, see: (a) Asano, N.; Nash, R. J.; Molyneux, R. J.; Fleet, G.
W. J. Tetrahedron: Asymmetry 2000, 11, 1645-1680. (b) Mroczek, T.;
Glowniak, K. Proc. Phytochem. Soc. Eur. 2002, 47, 1-46. (c) Fu, P. P.;
Xia, Q.; Lin, G.; Chou, M. W. Drug Metabol. ReV. 2004, 36, 1-55. (d)
Pyne, S. G.; Tang, M. Curr. Org. Chem. 2005, 9, 1393-1418.
Castanospermum australe2 and Alexa leiopetala,3 respec-
tively. Australine is an inhibitor of fungal amyloglucosidase
† Dedicated to the memory of Dr. Jasmin Jakupovic (deceased September
6, 2005).
‡ Universitat Jaume I.
§ Universitat de Valencia.
10.1021/ol062570v CCC: $37.00
© 2007 American Chemical Society
Published on Web 12/06/2006

published for the natural compound relied on an X-ray
diffraction analysis2 and was thus unambiguous, the reported
NMR data2a were actually those of the epimer at C-1. These
inconsistencies were resolved in later publications as a
consequence of accurate NMR studies9 and stereoselective
syntheses.7
Scheme 1. Retrosynthetic Analysis of Australine 1
The published synthetic strategies for pyrrolizidine deriva-
tives vary widely.10 Considering only the total synthesis of
australine, five different approaches have been reported.
Denmark’s synthesis was based on an asymmetric tandem
[4 + 2]/[3 + 2] cycloaddition methodology,7b whereas
Pearson’s synthesis started from a sugar precursor7c and
Wong and Romero made use of a chemoenzymatic strategy.7d
In all three cases, one C-N bond was created first and the
other two in a subsequent, one-pot process. In contrast with
these methodologies, White’s and Madsen’s syntheses of
australine relied upon the ruthenium-catalyzed ring-closing
metathesis (RCM).7e,f This reaction was used to generate
hexahydroazocine epoxides, which underwent transannular
cyclizations to the desired pyrrolizidine system. RCM played
also an important role in Pyne’s attempted synthesis,11 in
which it was employed to create one of the two five-
membered rings. The strategy, however, proved suitable for
the preparation of 1-epiaustraline but failed in the case of
australine itself.
stereoselectively to (S)-R-alkoxyaldehydes to yield adducts
of general structure A (Scheme 2).12 The retrosynthetic
Scheme 2. Aldol Reaction of Ketones III with
R-Alkoxyaldehydes
Our own retrosynthetic concept is depicted in Scheme 1.
As compared with the aforementioned syntheses, a distinctive
feature is the simultaneous disconnection of all C-N bonds
by means of three consecutive SN2 reactions with a nitrogen
nucleophile in a single precursor. This unveils the polyhy-
droxylated chain of I (OX ) leaving group), in which the
stereogenic centers of C-2 and C-7a carbons show a
configuration opposite to the corresponding carbons in
australine. Intermediate I could be derived from ketone II
which in turn can be disconnected, via retroaldol cleavage
of the C1-C7a bond, to ketone III and chiral aldehyde IV.
This represents a very convergent strategy in which the eight
carbon atoms of the target molecule are joined from two
four-carbon precursors.
scheme proposed for australine relied upon this finding. Thus,
after a careful choice of appropriate protecting groups, III
and IV (Scheme 1) became 3 (III, P2 ) TES) and 4 (IV, P3
) Bn, P5 ) TPS), respectively (Scheme 3).
The aldol addition of ketone 313 to aldehyde 414 was
accomplished under the described conditions.12 After oxida-
tive workup, the polyoxygenated ketone 5 was obtained as
a single stereoisomer in 72% yield (Scheme 3). In order to
achieve a stereoselective reduction of the ketone carbonyl
(12) Marco, J. A.; Carda, M.; D´ıaz-Oltra, S.; Murga, J.; Falomir, E.;
Ro¨per, H. J. Org. Chem. 2003, 68, 8577-8582.
We have recently demonstrated that Z boron enolates
derived from L-erythrulose derivatives such as III add
(13) For the preparation of variously protected D- and L-erythrulose
derivatives using chiral precursors other than erythrulose itself, see: Marco,
J. A.; Carda, M.; Gonza´lez, F.; Rodr´ıguez, S.; Murga, J. Liebigs Ann. Chem.
1996, 1801-1810. For an improved preparation of silylated L-erythrulose
acetonides III (P2 ) TES, TBS, TPS) from L-erythrulose hydrate, see: Carda,
M.; Rodr´ıguez, S.; Murga, J.; Falomir, E.; Marco, J. A.; Ro¨per, H. Synth.
Commun. 1999, 29, 2601-2610.
(7) Previous syntheses of australine: (a) Furneaux, R. H.; Gainsford, G.
J.; Mason, J. M., Tyler, P. C. Tetrahedron 1994, 50, 2131-2160. (b)
Denmark, S. E.; Martinborough, E. A. J. Am. Chem. Soc. 1999, 121, 3046-
3056. (c) Pearson, W. H.; Hines, J. V. J. Org. Chem. 2000, 65, 5785-
5793. (d) Romero, A.; Wong, C.-H. J. Org. Chem. 2000, 65, 8264-8268.(e)
White, J. D.; Hrnciar, P. J. Org. Chem. 2000, 65, 9129-9142. (f) Lauritsen,
A.; Madsen, R. Org. Biomol. Chem. 2006, 4, 2898-2905.
(8) For recent reviews on syntheses of this compound class, see: (a)
Yoda, H. Curr. Org. Chem. 2002, 6, 223-243. (b) Ayad, T.; Genisson, Y.;
Baltas, M. Curr. Org. Chem. 2004, 8, 1211-1233.
(9) (a) Wormald, M. R.; Nash, R. J.; Hrnciar, P.; White, J. D.; Molyneux,
R. J.; Fleet, G. W. J. Tetrahedron: Asymmetry 1998, 9, 2549-2558. (b)
Kato, A.; Kano, E.; Adachi, I.; Molyneux, R. J.; Watson, A. A.; Nash, R.
J.; Fleet, G. W. J.; Wormald, M. R.; Kizu, H.; Ikeda, K.; Asano, N.
Tetrahedron: Asymmetry 2003, 14, 325-331.
(10) (a) Hanzawa, Y.; Ito, H.; Taguchi, T. Synlett 1995, 299-305. (b)
Oppolzer, W. Gazz. Chim. Ital. 1995, 125, 207-213. (c) Broggini, G.;
Zecchi, G. Synthesis 1999, 905-917. (d) Pandit, U. K.; Overkleeft, H. S.;
Borer, B. C.; Biera¨ugel, H. Eur. J. Org. Chem. 1999, 959-968. (e) Bowman,
W. R.; Fletcher, A. J.; Potts, G. B. S. J. Chem. Soc., Perkin Trans. I 2002,
2747-2762. (f) Pyne, S. G.; Davis, A. S.; Gates, N. J.; Hartley, J. P.;
Lindsay, K. B.; Machan, T.; Tang, M. Synlett 2004, 2670-2680.
(11) Tang, M.; Pyne, S. G. Tetrahedron 2004, 60, 5759-5767.
(14) Aldehyde 4 was prepared through modification of a previously
published method for the synthesis of a closely related product (TES instead
of Bn: Hayashi, Y.; Yamaguchi, J.; Shoji, M. Tetrahedron 2002, 58, 9839-
9846). See details of the preparation in the Supporting Information.
(15) Greene, T. W.; Wuts, P. G. M. ProtectiVe Groups in Organic
Synthesis, 3rd ed.; John Wiley and Sons: New York, 1999; pp 45-48.
(16) The SEM group was selected for its ability to allow control of the
carbonyl reduction of 6 by means of a chelation mechanism. When other
protecting groups with the same property, such as MOM or a MEM, were
assayed, cyclic methylenedioxy derivatives (formaldehyde acetals) were
formed in the reduction process with LiBH4. Precedents of such side
reactions are known: (a) Herbert, J. M.; Knight, J. G.; Sexton, B.
Tetrahedron 1996, 52, 15257-15266. (b) Kiyooka, S.; Shahid, K. A.; Goto,
F.; Okazaki, M.; Shuto, Y. J. Org. Chem. 2003, 68, 7967-7978. (c)
Ramachandran, P. V.; Prabhudas, B.; Chandra, J. S.; Reddy, M. V. R. J.
Org. Chem. 2004, 69, 6294-6304. In some cases, however, formation of
formaldehyde acetals may be the desired reaction: Durham, T. B.;
Blanchard, N.; Savall, B. M.; Powell, N. A.; Roush, W. R. J. Am. Chem.
Soc. 2004, 126, 9307-9317.
78
Org. Lett., Vol. 9, No. 1, 2007

Eventually, treatment of compound 9 with an excess of
MeMgBr in a refluxing toluene-ether mixture not only
caused an alkylative opening of the acetonide ring18 but also
cleaved the SEM group to yield diol 10. This admittedly
unanticipated but felicitous finding increased the efficiency
of our synthetic strategy. Removal of the TPS group with
TBAF was followed by conversion of the corresponding triol
11 into trimesylate 12. Attempts at one-pot cyclization of
12 to a pyrrolizidine derivative through reaction with
ammonia19 were unsuccessful. While no reaction took place
at room temperature, decomposition was observed under
forcing conditions. Interestingly, when a solution of 12 in
DMSO was heated at 80 °C with benzylamine and catalytic
amounts of NaI, the fully protected australine derivative 13
was obtained rather than the anticipated quaternary N-
benzylammonium salt. The reaction requires an excess of
benzylamine, as it is otherwise very slow, and the presence
of catalytic amounts of iodide ion, the yield being much
lower in its absence. It is likely that I-, as a better both
nucleophile and leaving group, displaces first the primary
mesylate group in 12 and becomes then displaced by the
amine nucleophile. The second and third nucleophilic
substitutions by the nitrogen atom are intramolecular. After
formation of the third C-N bond, iodide ion attacks on the
benzylic position of quaternary ammonium salt i to yield 13
and benzyl iodide. The latter will then react with excess
benzylamine to give dibenzylamine (see proposed catalytic
cycle in Scheme 4).20 A key aspect for the success may be
the fact that one mesylate is primary while the other two
are secondary and relatively hindered. Thus, the first and
intermolecular SN2 step takes place only on the primary
mesylate. Indeed, when we studied the reaction on a
simplified model compound (the trimesylate of heptane-1,4,7-
triol), double SN2 substitution at both primary ends was
observed and no detectable amounts of pyrrolizidine were
recovered.
Scheme 3. Stereoselective Synthesis of Australine 1a
a Acronyms and abbreviations: Chx, cyclohexyl; TES, triethyl-
silyl; TPS, tert-butyldiphenylsilyl; SEM 2-(trimethylsilyl)ethoxym-
ethyl; TBAF, tetra-n-butylammonium fluoride hydrate; TFA,
trifluoroacetic acid.
Compound 13 was then hydrogenolytically debenzylated
to 14. Finally, acid-catalyzed cleavage of the tert-butyl
residue of 14 followed by alcalinization provided synthetic
australine, with physical and spectral properties identical to
those reported for the natural compound.1,7,9 The overall yield
of the 11-step sequence was 10% based on ketone 3.
In summary, the pyrrolizidine alkaloid australine has been
synthesized using our recently reported double asymmetric
aldol addition as one key step. This allowed us to build up
the C1-C7a bond in one step with the desired configuration
at these two stereocenters. As a second key feature, the
pyrrolizidine ring system was constructed in a one-pot
procedure by means of three consecutive inter/intramolecular
group, aldol 5 was treated with SEMCl15,16 to yield the fully
protected ketone 6. Reduction of 6 with LiBH4 at -90 °C
stereoselectively furnished alcohol 7 (dr > 95:5) in 80%
chemical yield. The TES group was then selectively removed
with the aid of catalytic amounts of DDQ in aqueous THF.17
This gave diol 8, which was then converted into the
perbenzylated derivative 9. In order to set the polyhydroxy-
lated chain in a situation suitable for the foreseen SN2
displacements, the TPS, SEM, and acetonide groups had to
be removed with parallel protection of the primary alcohol
group liberated after acetonide hydrolysis. A number of
methods were assayed to accomplish this goal, but all
attempts at selective cleavage of the acetonide moiety were
plagued with formation of side products and low yields.
(18) Cheng, W.-L.; Yeh, S.-M.; Luh, T.-Y. J. Org. Chem. 1993, 58,
5576-5677.
(19) (a) We have found only one example in the recent literature of this
method of creation of pyrrolizidine systems: Dong, H. Q.; Shi, Z. C.; Lin,
G. Q. Chin. Chem. Lett. 1997, 8, 773-776. (b) An old publication reported
the nonstereoselective formation of racemic 3-methylpyrrolizidine in the
reaction of 1,4,7-tribromo-3-methylheptane with ammonia under harsh
conditions (130 °C): Seiwerth, R.; Orescanin-Majhofer, B. Monatsh. Chem.
1952, 83, 1298-1300.
(17) Tanemura, K.; Suzuki, T.; Horaguchi, T. J. Chem. Soc., Perkin
Trans. 1 1992, 2997-2998. These unconventional conditions were found
to be the only suitable means for the selective removal of the TES group
without cleaving the acetonide moiety and the TPS group.
(20) The yield of 13 (60% overall from triol 11) is excellent, taking into
account that it is the result of seven consecutive steps (three mesylations,
three nucleophilic substitutions, and the N-debenzylation step).
Org. Lett., Vol. 9, No. 1, 2007
79

the application of our methodology to the synthesis of other
members of the australine/alexin family.
Scheme 4. Mechanistic Proposal for the Formation of 13 from
12
Acknowledgment. Financial support has been granted by
the Spanish Ministry of Science and Technology (Project
Nos. CTQ2005-06688-C02-01 and CTQ2005-06688-C02-
02), by the BANCAJA-UJI foundation (projects P1-1A-2005-
15 and P1-1B-2005-30), and by the AVCyT (Project Nos.
GVACOMP2006/006, GVACOMP2006/008, and GV05/52).
C.R. thanks the Spanish Ministry of Education and Science
for a predoctoral fellowship (FPI program). We further thank
Prof. G. W. J. Fleet, Department of Chemistry, University
of Oxford, and Dr. R. J. Molyneux, Western Regional
Research Center, for kindly providing a sample of australine,
and Prof. S. G. Pyne, University of Wollongong, Australia,
for kindly providing reviews on pyrrolizidines.
Supporting Information Available: Experimental pro-
cedures for the preparation and tabulated spectral data of all
new compounds. Graphical NMR spectra of all new com-
pounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
SN2 displacements on a trimesylate mediated by benzylamine.
The formation of the bicyclic ring system took place with
concomitant N-debenzylation of the presumed quaternary
ammonium salt intermediate. Future studies will focus on
OL062570V
80
Org. Lett., Vol. 9, No. 1, 2007