Construction of protected hydroxylated pyrrolidines using nitrogen-centered radical cyclizations

Article abstract of DOI:10.1039/b914757h

Nitrogen-centered radical cyclizations onto silyl enol ethers were utilized for the syntheses of protected polyhydroxylated pyrrolidines 2-hydroxymethyl-3-hydroxypyrrolidine and 1,4-dideoxy-1,4-imino-l-ribitol.

Full text of DOI:10.1039/b914757h

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Construction of protected hydroxylated pyrrolidines using
nitrogen-centered radical cyclizationsw
Huimin Zhai, Maria Zlotorzynska and Glenn Sammis*
Received (in College Park, MD, USA) 22nd July 2009, Accepted 10th August 2009
First published as an Advance Article on the web 2nd September 2009
DOI: 10.1039/b914757h
Nitrogen-centered radical cyclizations onto silyl enol ethers
were utilized for the syntheses of protected polyhydroxylated
pyrrolidines 2-hydroxymethyl-3-hydroxypyrrolidine and 1,4-
dideoxy-1,4-imino-L-ribitol.
Scheme 1 Nitrogen-centered radical cyclization to form pyrrolidine 9.
Polyhydroxylated alkaloids are a diverse class of pyrrolidine,
piperidine, pyrrolizidine, indolizidine, and nortropane natural
products isolated from both plants and microorganisms.1
These compounds can be potent competitive and reversible
inhibitors of glycosidases by serving as furanose and pyranose
mimics.2 The few polyhydroxylated alkaloids that are
commercially available have promising therapeutic potential,
including efficacy as anticancer, antiviral, and antidiabetic
agents.1c However, the full medicinal potential of the poly-
hydroxylated alkaloids is limited by compound availability
and toxicity.1c A general method for their synthesis would
facilitate the testing of existing compounds as well as provide
access to potent unnatural analogs.
proline derivatives, this nitrogen-centered radical approach
can readily access pyrrolidines substituted at the 3- and
5-positions. Furthermore, this radical route should enable
the facile synthesis of 2-hydroxymethylpyrrolidine analogs.
There have been no reports of 5-exo nitrogen-centered
radical3 cyclizations onto silyl enol ethers.4 We have recently
found that oxygen-centered radicals rapidly cyclize with silyl
enol ethers.5 However, nitrogen-centered radicals are more
nucleophilic and, thus, may not cyclize as readily. We,
therefore, began our investigations by examining the key
nitrogen-centered radical cyclization with a synthesis of simple
2-hydroxymethylpyrrolidine 9 (Scheme 1).
Many of the polyhydroxylated alkaloids have
a
While many methods have been developed to generate
nitrogen-centered radicals,3,6 we started our investigations
with aminyl radical formation from azides.7 Azides are
synthetically versatile as they can be introduced easily, carried
through a multiple step synthesis, and can provide direct
access to unprotected pyrrolidines. Treatment of azide 7 with
tributyltin hydride and azobisisobutyronitrile (AIBN) resulted
in the formation of aminyl radical 8, which rapidly cyclized to
form the desired 2-siloxymethyl pyrrolidine 9 in 75% isolated
yield. While the reaction goes through a tin-bound nitrogen-
centered radical, free secondary amines are always isolated
after cyclization.7 As a control experiment, we heated a
solution of azide 7 in benzene in the absence of tributyltin
hydride. This resulted in no reaction, excluding a possible
cycloaddition pathway.
2-hydroxymethyl pyrrolidine core (Fig. 1, 1–3). We envisaged
that the core pyrrolidine (Fig. 2, 4) could be synthesized
utilizing a nitrogen-centered radical cyclization onto a silyl
enol ether (5). This cyclization sets the stereocenter at the
2-position and simplifies the pyrrolidine to a simple acyclic
precursor (6). Unlike methods starting from amino acid or
Fig. 1 Representative polyhydroxylated alkaloids.
With a successful procedure for the key nitrogen-centered
radical cyclization in hand, we next examined the utility of
the cyclization in the stereoselective synthesis of the poly-
hydroxylated alkaloid CYB-3 (1). This compound, first isolated
from the seeds and leaves of Castanospermum australe,8,9 is
the least oxygenated of the polyhydroxylated pyrrolidines.
Despite its seemingly simple structure, the oxygenation and
stereochemistry make it a challenging synthetic target.10
Cyclization precursor 13 was synthesized in five steps
from the known (S)-methyl-3,5-dihydroxypentanoate11 (10)
(Scheme 2). Cyclization of silyl enol ether 13 provided the
desired pyrrolidine as a 60 : 40 mixture of trans to cis isomers
(Scheme 3). The majority of the mass balance corresponded to
amine 15, presumably arising from trapping of the nitrogen
radical prior to cyclization.
Fig. 2 Nitrogen-centered radical cyclization for the synthesis of
substituted 2-hydroxymethyl pyrrolidines.
Department of Chemistry, University of British Columbia,
2036 Main Mall, Vancouver, British Columbia, Canada V6T 1Z1.
E-mail: gsammis@chem.ubc.ca
w Electronic supplementary information (ESI) available: Experimental
procedures and full characterization for the intermediates and cyclized
products. See DOI: 10.1039/b914757h
ꢀc
This journal is The Royal Society of Chemistry 2009
5716 | Chem. Commun., 2009, 5716–5718

View Article Online
20 cyclized to provide the desired product (17) in excellent
isolated yield as a single diastereomer with the remainder of
the mass balance corresponding to uncyclized Z-silyl enol
ether 18. This confirms that the Z-silyl enol ether cyclizes
with greater stereoselectivity than the E-silyl enol ether.
Comparison of the cyclization of azides 13, 16 and 20
suggests that sterics alone do not justify the change in the
diastereomeric ratios.
With a successful method for the synthesis of protected
CYB-3, we next investigated a synthesis of the more challenging
unnatural polyhydroxylated alkaloid, 1,4-dideoxy-1,4-imino-
L-ribitol (2).14 1,4-Dideoxy-1,4-imino-L-ribitol possesses an
especially challenging substitution pattern and will lead to
significantly more steric interactions in the transition state
and, thus, may lead to a slower rate of cyclization.
Scheme 2 Synthesis of cyclization precursor 13.
Cyclization precursor 25, required for the synthesis of
polyhydroxylated alkaloid 2, was readily synthesized in four
steps from triol 22 (Scheme 5). Cyclization of silyl enol ether
25 afforded a mixture of the desired pyrrolidine 26, imine 27
and uncyclized amine 28 in a 2 : 1 : 1 ratio as determined by
1H NMR spectroscopy. Compounds 26 and 27 emerged as a
single diastereomer. Imine 27 is too unstable to be isolated,
but can readily be reduced to 26, e.g., with DIBAL. The
conversion to protected 1,4-dideoxy-1,4-imino-L-ribitol can,
therefore, be improved by treating the crude cyclization
mixture with DIBAL (Scheme 6), providing pyrrolidine 26
to primary amine 28 as a 3 : 1 mixture. The pyrrolidine was
then separated from the primary amine by treatment with
acid and crystallization of the salt to afford TBS protected
1,4-dideoxy-1,4-imino-L-ribitol (29) in 58% isolated yield.
We hypothesize that the imine may be formed according to
the mechanism presented in Fig. 3. Initial formation of the
tin-bound aminyl radical and subsequent cyclization provides
pyrrolidine 30. Under the cyclization conditions, the radical
(30) has time to undergo a subsequent hydrogen transfer to
afford radical 31. Elimination of tributyltin radical then
provides imine 27. The elimination of tributyltin radical from
a carbon-centered radical a to a tin-bound amine has been
previously observed by Kim and co-workers.7b We surmised
Scheme 3 Cyclization of silyl enol ethers 13 and 16.
In an effort to increase the yields to the cyclized products,
we next examined by decreasing the steric bulk of the
substrate. Decreasing the steric size of the silyl groups from
TBS (13) to TES (16) provided increases in both the ratio of
cyclized product (17) to primary amine (18) as well as the
diastereoselectivity of the cyclization. Amines 15 and 18 were
obtained in a slightly higher ratio of Z- to E-isomers for the
cyclization of both silyl enol ethers 13 and 16, suggesting that
the rate of cyclization of the Z-silyl enol ether is slightly slower
than the cyclization rate of E-silyl enol ether.
Despite the improved yields using cyclization precursor 16,
the diastereoselectivity was still significantly lower than
comparable cyclizations with oxygen-centered radicals.5,12
This discrepancy could arise from the difference in selectivities
for the E and Z-silyl enol ethers.13 To test this hypothesis, we
selectively synthesized both E-silyl enol ether 19 and Z-silyl
enol ether 20 (Scheme 4). Silyl enol ether 19 cyclized with no
diastereoselectivity, providing pyrrolidine 21 as a 50 : 50
mixture of cis to trans isomers. In contrast, silyl enol ether
Scheme 4 Cyclization of E- and Z-silyl enol ethers 19 and 20.
Scheme 5 Synthesis and cyclization of silyl enol ether 25.
Chem. Commun., 2009, 5716–5718 | 5717
ꢀc
This journal is The Royal Society of Chemistry 2009

View Article Online
3 For reviews on nitrogen-centered radicals, see: (a) R. S. Neale,
Synthesis, 1971, 1; (b) P. Mackiewicz and R. Furstoss, Tetrahedron,
1978, 34, 3241; (c) L. Stella, Angew. Chem., Int. Ed. Engl., 1983, 22,
337; (d) J. L. Esker and M. Newcomb, Adv. Heterocycl. Chem.,
1993, 58, 1; (e) S. Z. Zard, Synlett, 1996, 1148; (f) A. G. Fallis and
I. M. Brinza, Tetrahedron, 1997, 53, 17543; (g) S. Z. Zard, Chem.
Soc. Rev., 2008, 37, 1603; (h) D. P. Curran, in Comprehensive
Organic Synthesis, ed. B. M. Trost and I. Fleming, Pergamon
Press, Oxford, 1991, vol. 4, ch. 4.2, pp. 811–812; (i) L. Stella, in
Radicals in Organic Synthesis, ed. P. Renaud and M. P. Sibi,
Wiley-VCH, Weinheim, 2001, vol. 2, ch. 51, pp. 407–426;
(j) M. Minozzi, D. Nanni and P. Spagnolo, Chem.–Eur. J., 2009,
15, 7830.
Scheme 6 Synthesis of protected 1,4-dideoxy-1,4-imino-L-ribitol.
4 There is only one example of a 6-endo cyclization of a nitrogen-
centered radical onto an alkyl enol ether, which was part of a study
on lone-pair repulsion during radical cyclizations: X. Yuan, K. Liu
and C. Li, J. Org. Chem., 2008, 73, 6166.
5 For 5-exo cyclizations of oxygen-centered radicals onto silyl enol
ethers, see: M. Zlotorzynska, H. Zhai and G. M. Sammis, Org.
Lett., 2008, 10, 5083.
6 For
a representative example of nitrogen-centered radical
formation from the homolysis of sulfenamides, see:
W. R. Bowman, D. N. Clark and R. J. Marmon, Tetrahedron,
1994, 50, 1275.
Fig. 3 Proposed mechanism for the formation of imine 27.
7 For initial studies, see: (a) S. Kim, G. H. Joe and J. Y. Do, J. Am.
Chem. Soc., 1993, 115, 3328; (b) S. Kim, G. H. Joe and J. Y. Do,
J. Am. Chem. Soc., 1994, 116, 5521; For calculations indicating
that tin-bound aminyl radicals are more nucleophilic than free
aminyl radicals, see: ; (c) S. Kim, K. M. Yeon and K. S. Yoon,
Tetrahedron Lett., 1997, 38, 3919.
8 R. J. Nash, E. A. Bell, G. W. J. Fleet, R. H. Jones and
J. M. Williams, J. Chem. Soc., Chem. Commun., 1985, 738.
9 CYB-3 has been found to exhibit modest glycosidase inhibition:
A. M. Scofield, L. E. Fellows, R. J. Nash and G. W. J. Fleet, Life
Sci., 1986, 39, 645.
10 For early syntheses of 2-hydroxymethyl-3-hydroxypyrrolidine and
protected derivatives,see: (a) N. Ikota and A. Hanaki, Hetero-
cycles, 1988, 27, 2535; (b) R. C. Roemmele and H. Rapoport,
J. Org. Chem., 1989, 54, 1866; (c) P. Merino, I. Delso, T. Tejero,
F. Cardona, M. Marradi, E. Faggi, C. Parmeggiani and A. Goti,
Eur. J. Org. Chem., 2008, 2929; (d) M. Toumi, F. Couty and
G. Evano, Tetrahedron Lett., 2008, 49, 1175; (e) J. M. Schomaker,
S. Bhattacharjee, J. Yan and B. Borhan, J. Am. Chem. Soc., 2007,
129, 1996; (f) M. Toumi, F. Couty and G. Evano, Angew. Chem.,
Int. Ed., 2007, 46, 572; (g) D. J. Michaelis, M. A. Ischay and
T. P. Yoon, J. Am. Chem. Soc., 2008, 130, 6610.
that if the concentration of tributyltin hydride is minimized
during the cyclization, one may be able to favor the imine
product. Indeed slow addition of tributyltin hydride leads
to greater conversion to cyclic imine 27. Correspondingly,
increasing the concentration of tributyltin hydride through
fast addition resulted in only minimal amounts of imine
formation, which is consistent with the proposed mechanism.
In summary, we have developed a general method for the
synthesis of substituted 2-hydroxymethyl pyrrolidines, a key
structural motif in the polyhydroxylated alkaloids. Our route
utilizes a nitrogen-centered radical cyclization onto a silyl enol
ether. We have demonstrated the synthetic potential of this
methodology in the synthesis of 2-hydroxymethyl-3-hydroxy-
pyrrolidine and in the synthesis of 1,4-dideoxy-1,4-diamino-L-
ribitol. Efforts to further investigate the stereochemical aspects
of the cyclization and to explore this new nitrogen-centered
radical cyclization in the context of complex polyhydroxylated
alkaloids and related analogs are currently underway.
This work was supported by the University of British
Columbia, the Natural Sciences and Engineering Research
Council of Canada (NSERC), and a doctoral fellowship from
NSERC to M.Z.
11 For
a synthesis of (S)-methyl-3,5-dihydroxypentanoate, see:
B. Loubinaux, J.-L. Sinnes, A. C. O’Sullivan and T. Winkler,
Tetrahedron, 1995, 51(12), 3549.
12 For stereoselectivity models in radical cyclizations, see: (a) A. L.
J. Beckwith and C. H. Schiesser, Tetrahedron, 1985, 41, 3925;
(b) D. C. Spellmeyer and K. N. Houk, J. Org. Chem., 1987,
52, 959.
13 Synthesis of E-enriched silyl enol ethers was accomplished using
TBSCl and DBU at 35 1C providing selectivities greater than 90 :
10 E–Z. Synthesis of Z-enriched silyl enol ethers was accomplished
using TBSOTf and diisopropylethylamine at 0 1C providing
selectivities greater than 60 : 40 Z–E. The selectivities for TES
enol ether formation were lower. See ESIw for details.
14 For early syntheses of 4-dideoxy-1,4-imino-L-ribitol and protected
derivatives, see: (a) H. Setoi, H. Kayakiri, H. Takeno and
M. Hashimoto, Chem. Pharm. Bull., 1987, 35, 3995; (b) G. W.
J. Fleet, J. C. Son, D. C. Green, I. C. Bello and B. Winchester,
Tetrahedron, 1988, 44, 2649. For recent, enantioselective syntheses
of 1,4-dideoxy-1,4-imino-L-ribitol, see: (c) S.-Y. Luo,
S. S. Kulkarni, C.-H. Chou, W.-M. Liao and S.-C. Hung,
J. Org. Chem., 2006, 71, 1226; (d) M. Marradi, S. Cicchi,
J. I. Delso, L. Rosi, T. Tejero, P. Merino and A. Goti, Tetrahedron
Lett., 2005, 46, 1287.
Notes and references
1 For recent reviews, see: (a) B. Winchester and G. W. J. Fleet,
Glycobiology, 1992, 2, 199; (b) A. E. Stutz, Iminosugars as
¨
Glycosidase Inhibitors: Nojirimycin and Beyond, Wiley-VCH,
Weinheim, 1999; (c) A. A. Watson, G. W. J. Fleet, N. Asano,
R. J. Molyneux and R. J. Nash, Phytochemistry, 2001, 56, 265;
(d) T. M. Wrodnigg, Monatsh. Chem., 2002, 133, 393;
(e) P. Compain and O. R. Martin, Curr. Top. Med. Chem., 2003,
3, 541.
2 (a) G. Legler, Adv. Carbohydr. Chem. Biochem., 1990, 48, 319;
(b) N. Asano, H. Kizu, K. Oseki, E. Tomioka, K. Matsui,
M. Okamoto and M. Baba, J. Med. Chem., 1995, 38, 2349;
(c) N. Asano, R. J. Nash, R. J. Molyneux and G. W. J. Fleet,
Tetrahedron: Asymmetry, 2000, 11, 1645; (d) T. D. Butters,
R. A. Dwek and F. M. Platt, Curr. Top. Med. Chem., 2003, 3, 561.
ꢀc
This journal is The Royal Society of Chemistry 2009
5718 | Chem. Commun., 2009, 5716–5718