Concise asymmetric syntheses of (-)-lentiginosine and of its pyrrolizidinic analogue.

Article abstract of DOI:10.1039/b212217k

(-)-Lentiginosine and its pyrrolizidinic analogue have been prepared in a straightforward five-step sequence from a versatile chiral cis-alpha,beta-epoxyamine.

Full text of DOI:10.1039/b212217k

Concise asymmetric syntheses of (2)-lentiginosine and of its
pyrrolizidinic analogue
Tahar Ayad, Yves Génisson,* Michel Baltas* and Liliane Gorrichon
LSPCMIB,UMR 5068, Université Paul Sabatier, 118 route de Narbonne, 31062 Toulouse, France.
E-mail: baltas@chimie.ups-tlse.fr, genisson@chimie.ups-tlse.fr; Fax: +33 (0)5 61 55 82 45; Tel: +33 (0)5
61 55 62 99
Received (in Cambridge, UK) 11th December 2002, Accepted 20th January 2003
First published as an Advance Article on the web 29th January 2003
(2)-Lentiginosine and its pyrrolizidinic analogue have been
prepared in a straightforward five-step sequence from a
versatile chiral cis-a,b-epoxyamine.
implicates an additional carbonyl reduction step.5 Although
most of the conditions tested proved unsuccessful, treatment of
2 with butenyl triflate gratifyingly afforded the desired tertiary
amine 3 in 67% yield. Use of proton-sponge® as base proved to
be determining for completion of the reaction. The allylation of
the secondary amine 2 was achieved in 85% yield under
standard conditions to afford 4.
With the dienic precursors 3 and 4 in hand, we were in a
position to test the key RCM step. Preliminary studies were
conducted on the partially protected (1,2)-acetonide derivative
of 3 (not shown). They indicated that the second generation
1,3-bis(mesityl)-2-imidazolidinylidene-substituted ruthenium
Grubbs catalyst (Grubbs II) was superior to the classical
benzylidene-bis(tricyclohexylphosphine)dichlororuthenium
(Grubbs I) in forming the tetrahydropyridine ring. To our
satisfaction, treatment of the fully unprotected precursor 3 with
the Grubbs II ruthenium complex in toluene at 70 °C for less
than an hour led to the desired cyclised compound 5 in 66%
yield (Scheme 1). Formation of the dihydropyrrole ring from the
N-allylated free triol 4 was best accomplished with the Grubbs
I catalyst, giving 6 in 70% yield. Both derivatives 5 and 6
possess an intracyclic olefin well positioned for further
oxidative functionalisation. They therefore represent valuable
intermediates for the synthesis of various polyhydroxylated
bicyclic alkaloids.
Completion of the synthesis required formation of the
dihydroxylated five-membered ring of the lentiginosine frame-
work. Catalytic hydrogenation first ensured N-debenzylation
with concomitant olefin saturation, yielding piperidine 7 and
pyrrolidine 8 (Scheme 2). Appel cyclisation of 7 then delivered,
in 67% yield, the all-R-lentiginosine (9). This compound
([a]25D 22.0° (c 1.0, MeOH); lit.3b,4b,4f,4j 21.6, 22.6, 23.05,
24.5) had spectral data in agreement with that reported in the
literature.4 Dihydroxypyrrolizidine 10 was prepared from 8 in
the same manner and in comparable yield.† The starting
epoxyamine 1 being equally available in both optical series, this
We recently demonstrated that cis-a,b-epoxyamine 1, readily
prepared from the corresponding chiral epoxyaldehyde, is a
valuable precursor of five-membered ring azasugars (Fig. 1).1
We wish to describe here the extension of this methodology to
the asymmetric synthesis of polyhydroxylated indolizidines and
pyrrolizidines.
Fig. 1
In our previous work, the epoxyamine precursor 1 was
efficiently transformed into a polyhydroxylated pyrrolidine via
the regiocontrolled C-3 opening of the epoxide, followed by the
Appel cyclisation of the activated primary alcohol, and olefin
dihydroxylation (Fig. 1). We envisioned that the vinyl moiety,
instead of being directly oxidised, could be first engaged in a
ring closing metathesis (RCM), allowing access to bicyclic
analogues of five-membered ring azasugars (Fig. 2).2 At some
stage this would require the attachment onto the amine of a
second olefinic partner, namely a butenyl residue for in-
dolizidine syntheses and an allyl moiety for the preparation of
pyrrolizidines.
Lentiginosine was targeted to validate our synthetic plan.
Indeed this dihydroxylated indolizidine possesses some unique
features, being the least hydroxylated biologically active
bicyclic azasugar, and the most potent inhibitor of amylogluco-
sidase.3 It has therefore inspired a lot of synthetic work, mostly
based on chiral pool starting material, relying much more rarely
on asymmetric synthesis.4 We also wished to demonstrate the
flexibility of our approach in preparing a novel pyrrolizidinic
analogue of lentiginosine.
The amino triol 2, possessing the relative configuration
related to lentiginosine, was prepared in 70% yield by acidic
hydrolysis of (2S, 3R, 4R)-a,b-epoxyamine 1 (Scheme 1). Ring
opening, occurring exclusively at C-2, was accompanied by the
desilylation of the primary alcohol. Direct introduction of the
butenyl moiety via the alkylation of the unreactive secondary
amine 2 then required considerable experimentation. It was
however preferred to the alternative amine acylation which
Scheme 1 Reagents and conditions: i, 3 M H2SO4 (8 eq.), dioxane, reflux,
70%; ii, for n = 0: allylbromide (4 eq.), NaHCO3 (6 eq.), THF/H2O, rt,
85%; for n = 1: 4-butenyltrifluoromethanesulfonate (1.2 eq.), proton-
sponge® (1.5 eq.), CH2Cl2, rt, 67%; iii, for n = 0: Grubbs I cat. (8 mol%),
CH2Cl2, reflux, 70%; for n = 1: Grubbs II cat. (8 mol%), toluene, 70 °C,
66%. Grubbs
I cat.: benzylidene-bis(tricyclohexylphosphine)dichloro-
ruthenium; Grubbs II cat.: benzylidene[1,3-bis(2,4,6-trimethylphenyl)-
2-imidazolidinylidene]dichloro(tricyclohexylphosphine)ruthenium.
Fig. 2
582
CHEM. COMMUN., 2003, 582–583
This journal is © The Royal Society of Chemistry 2003

View Article Online
competitive inhibitor (Ki = 121 mM), as indicated by the
Lineweaver–Burk plot shown in Fig. 3.
The enzyme was kindly provided by Pr. P. Monsan, INSA,
Toulouse, France. We are grateful to S. Gavalda, R. Braga and
Dr C. Blonski for their help in enzymatic inhibition experi-
ments. T. A. acknowledges the ‘Ministère de l’Education
Nationale de la Recherche et des Technologies’ for a grant.
Notes and references
† Characteristic data for compound 10: [a]25 +7.6° (c 1.3, MeOH); dH
D
Scheme 2 Reagents and conditions: i, 12 bar H2, 10% Pd/C (10% w/w),
MeOH, 12 M HCl (cat.), for n = 0, 1: 90%; ii, PPh3 (2 eq.), CCl4 (2 eq.),
Et3N (2 eq.), DMF, rt, for n = 0, 1: 68%.
(400 MHz, CD3OD): 4.23–4.13 (1H, m, H-2), 3.90 (1H, br s, H-1),
3.63–3.55 (1H, m, H-7a), 3.19 (2H, AB of ABX, Jab = 11.6 Hz, Jax = 5.2
Hz, Jbx = 4.4 Hz (da 2 db = 230 Hz), H-3 and H-3A), 3.28–3.25 (1H, m, H-
5), 3.23–2.90 (1H, m, H-5A), 2.26–2.10 (1H, m, H-7), 2.09–1.93 (2H, m, H-
7A and H-6), 1.83–1.78 (1H, m, H-6A); dC (100 MHz, CD3OD): 82.8 (C-1),
80.0 (C-2), 75.0 (C-7a), 61.0 (C-3), 58.9 (C-5), 32.1 (C-7), 28.0 (C-6); mass
spectrum (DCI): m/z 144 (MH+, 100); HRMS: m/z (M+) calc. for C7H13NO2
143.0946, found 143.0951.
‡ Conditions were essentially identical to those used in ref. 3b.
1 T. Ayad, Y. Génisson, M. Baltas and L. Gorrichon, Synlett, 2001, 866.
2 For a review of the use of RCM in synthesis of bicyclic azasugars, see: U.
K. Pandit, H. S. Overkleeft, B. C. Borer and H. Bieräugel, Eur. J. Org.
Chem., 1999, 959; for selected recent examples, see: T. Subramanian, C.-
C. Lin and C.-C. Lin, Tetrahedron Lett., 2001, 42, 4079; C. F. Klitzke and
R. A. Pilli, Tetrahedron Lett., 2001, 42, 5605; K. B. Lindsay and S. G.
Pyne, J. Org. Chem., 2002, 67, 7774.
3 (a) Isolation: I. Pastuszak, R. J. Molyneux, L. F. James and A. D. Elbein,
Biochemistry, 1990, 29, 1886; (b) activity: A. Brandi, S. Cicchi, F. M.
Cordero, R. Frignoli, A. Goti, S. Picasso and P. Vogel, J. Org. Chem.,
1995, 60, 6806.
4 (a) For chiral pool syntheses, see: H. Yoda, H. Kitayama, T. Katagiri and
K. Takabe, Tetrahedron: Asymmetry, 1993, 4, 1455; (b) M. K. Gurjar, L.
Ghosh, M. Syamala and V. Jayasree, Tetrahedron Lett., 1994, 35, 8871;
(c) R. Giovannini, E. Marcantoni and M. Petrini, J. Org. Chem., 1995, 60,
5706; (d) A. Goti, F. Cardona and A. Brandi, Synlett, 1996, 761; (e) H.
Yoda, M. Kawauchi and K. Takabe, Synlett, 1998, 137; (f) A. E. McCaig,
K. P. Meldrum and R. H. Wightman, Tetrahedron, 1998, 54, 9429; (g)
D.-C. Ha, C.-S. Yun and Y. Lee, J. Org. Chem., 2000, 65, 621; (h) H.
Yoda, H. Katoh, Y. Ujihara and K. Takabe, Tetrahedron Lett., 2001, 42,
2509; (i) J. Rabiczko, Z. Urbanczyk-Lipkowska and M. Chmielewski,
Tetrahedron, 2002, 58, 1433; (j) K. L. Chandra, M. Chandrasekhar and
V. K. Singh, J. Org. Chem., 2002, 67, 4630; for asymmetric syntheses,
see (k) S. Nukui, M. Sodeoka, H. Sasai and M. Shibasaki, J. Org. Chem.,
1995, 60, 398; (l) M. O. Rasmussen, P. Delair and A. E. Greene, J. Org.
Chem., 2001, 66, 5438; (m) S. H. Lim, S. Ma and P. Beak, J. Org. Chem.,
2001, 66, 9056; (n) Z.-X. Feng and W.-S. Zhou, Tetrahedron Lett., 2003,
44, 497.
Fig. 3 Effect of substrate (p-nitrophenyl-a-D-glucopyranoside) concentra-
tion on amyloglucosidase (Aspegillus niger) inhibition by various concen-
trations of 10.
new route also offers a direct access to the other enantiomer of
these bicyclic azasugars.
This synthetic work was complemented by enzymatic
inhibition experiments with respect to the amyloglucosidase
from Aspergillus niger.‡ An IC50 of 25.5 mg mL21 was
measured for our sample of (2)-lentiginosine. This value is in
agreement with the IC50 (17 mg mL21) that Brandi et al found
for the laevorotatory isomer, proposed to be the unnatural
enantiomer of lentiginosine.3b The novel dihydroxylated pyrro-
5 This stepwise formation of tetrahydropyridine ring has been used in
several syntheses of polyhydroxylated indolizidines. For lentiginosine,
see: refs. 4j and 4m. For related alkaloids, see: C. Paolucci and L.
Mattioli, J. Org. Chem., 2001, 66, 4787.
lizidine 10 was found to display an IC50 of 27.3 mg mL21
comparable with that of (2)-lentiginosine, and to behave as a
,
CHEM. COMMUN., 2003, 582–583
583