A concise and stereoselective synthesis of hydroxypyrrolidines: Rapid synthesis of (+)-preussin

Article abstract of DOI:10.1021/ol101631e

A convergent and stereoselective synthesis of 2,5-disubstituted 3-hydroxypyrrolidines has been developed that involves reductive annulation of β-iminochlorohydrins, which are readily available from β- ketochlorohydrins, and provides rapid access to a variety of 2,5-syn-pyrrolidines. Application of this process to the concise (three-step) synthesis of the fungal metabolite (+)-preussin and analogues of this substance is reported.

Full text of DOI:10.1021/ol101631e

ORGANIC
LETTERS
2010
Vol. 12, No. 18
4034-4037
A Concise and Stereoselective
Synthesis of Hydroxypyrrolidines: Rapid
Synthesis of (+)-Preussin
Jason A. Draper and Robert Britton*
Department of Chemistry, Simon Fraser UniVersity, Burnaby,
British Columbia V5A 1S6, Canada
rbritton@sfu.ca
Received July 14, 2010
ABSTRACT
A convergent and stereoselective synthesis of 2,5-disubstituted 3-hydroxypyrrolidines has been developed that involves reductive annulation of
ꢀ-iminochlorohydrins, which are readily available from ꢀ-ketochlorohydrins, and provides rapid access to a variety of 2,5-syn-pyrrolidines. Application
of this process to the concise (three-step) synthesis of the fungal metabolite (+)-preussin and analogues of this substance is reported.
A number of biologically active alkaloids incorporate an
oxygenated pyrrolidine as a key skeletal feature.1 For
example, the pyrrolizidine (e.g., hyacinthacine A4 (1))2
alkaloids possess a dihydroxypyrrolidine core and exhibit
inhibitory activity toward various carbohydrate processing
enzymes,3 and the 3-hydroxypyrrolidine salinosporamide C
(2)4 is biogenetically related to the potent 20S proteasome
Figure 1. Hydroxypyrrolidine-containing natural products hya-
inhibitor salinosporamide A.5 In addition, preussin (3), which
was originally isolated from a liquid fermentation broth of
Aspergillus ochraceus,6,7 displays broad spectrum antifungal
activity7 and has shown growth-inhibitory and cytotoxic
effects on human cancer cells.8 Despite the varied and po-
tentially useful biological activities attributed to these and
other pyrrolidine alkaloids, defining a concise, general, and
cinthacine A4 (1), salinosporamide C (2), and preussin (3).
stereoselective strategy for their synthesis remains a signifi-
cant challenge.9 Thus, while at least 25 strategically unique
syntheses of preussin (3) have been reported,10-12 owing to
their overall length (up to 22 steps) and/or reliance on amino
acids or carbohydrates as starting materials, many of these
synthetic routes are not well suited for the production of
(1) (a) Michael, J. P. Nat. Prod. Rep. 2007, 24, 191. (b) Liddell, J. R.
Nat. Prod. Rep. 2002, 19, 773. (c) O’Hagan, D. Nat. Prod. Rep. 2000, 17,
435.
(2) Yamashita, T.; Yasuda, K.; Kizu, H.; Kameda, Y.; Watson, A. A.;
Nash, R. J.; Fleet, G. W. J.; Asano, N. J. Nat. Prod. 2002, 65, 1875.
(3) Watson, A. A.; Fleet, G. W. J.; Asano, N.; Molyneux, R. J.; Nash,
R. J. Phytochemistry 2001, 56, 265.
(7) For the relative and absolute stereochemical assignment of preussin, see:
Johnson, J. H.; Phillipson, D. W.; Kahle, A. D. J. Antibiot. 1989, 42, 1184.
(8) Achenbach, T. V.; Slater, E. P.; Brummerhop, H.; Bach, T.; Mu¨ller,
R. Antimicrob. Agents Chemother. 2000, 44, 2794.
(4) Williams, P. G.; Buchanan, G. O.; Feling, R. H.; Kauffman, C. A.;
Jensen, P. R.; Fenical, W. J. Org. Chem. 2005, 70, 6196.
(5) Feling, R.; Buchanan, G. O.; Mincer, T. J.; Kauffman, C. A.; Jensen,
P. R.; Fenical, W. Angew. Chem., Int. Ed. 2003, 42, 355.
(6) For the original isolation of preussin, see: Schwartz, R. E.; Liesch,
J.; Hensens, O.; Zitano, L.; Honeycutt, S.; Garrity, G.; Fromtling, R. A.;
(9) For reviews on pyrrolidine, pyrrolizidine, and indolizidine syntheses,
see: (a) Stocker, B. L.; Dangerfield, E. M.; Win-Mason, A. L.; Haslett,
G. W.; Timmer, M. S. M. Eur. J. Org. Chem. 2010, 1615. (b) Wolfe, J. P.
Eur. J. Org. Chem. 2007, 571. (c) Huang, P.-Q. Synlett 2006, 1133. (d)
Felpin, F.-X.; Lebreton, J. Eur. J. Org. Chem. 2003, 3693. (e) Yoda, H.
Curr. Org. Chem. 2002, 6, 223. (f) Pichon, M.; Figadere, B. Tetrahedron:
Asymmetry 1996, 7, 927.
Onishi, J.; Monaghan, R. J. Antibiot. 1988, 41, 1774
.
10.1021/ol101631e  2010 American Chemical Society
Published on Web 08/20/2010

structural analogues or related natural products. As notable
exceptions, Wolfe12c and Davis12a have recently reported 9-
and 10-step syntheses of preussin in which late-stage intro-
duction of the phenyl ring via Pd-catalyzed carboamination12c
or Horner-Wadsworth-Emmons12a reactions facilitates the
synthesis of derivatives of preussin with variously function-
alized aromatic rings. Herein, we describe a concise and
stereoselective synthesis of 3-hydroxypyrrolidines that pro-
vides rapid access to preussin (3) and analogues of this
substance and should be adaptable for the production of other
structurally related alkaloids (e.g., 1 and 2).
(see inset, Scheme 1) derived from reduction of the inter-
mediate ꢀ-keto- or iminochlorohydrins and/or dehydration/
isomerization of the pyrrolinium 6. Thus, our initial efforts
focused on the avoidance of these byproducts and optimiza-
tion of this general strategy for the selective production of
the 3-hydroxypyrrolidine (+)-preussin (3).
Scheme 2. Enantioselective Synthesis of (+)-Preussin (3)
Scheme 1. Strategy for Hydroxypyrrolidine Synthesis
As outlined in Scheme 1, our synthetic strategy centered
on the reductive amination of ꢀ-ketochlorohydrins,13 which
are readily available from the lithium aldol reaction of a
methyl ketone and an R-chloroaldehyde.14 Thus, it was
anticipated that reaction of a primary amine with a ꢀ-ke-
tochlorohydrin (e.g., 4) would afford a transient imine (e.g.,
5),15 the subsequent cyclization of which would lead to a
pyrrolinium intermediate (e.g., 6)16 and, following reduction,
the desired 2,5-disubstituted-3-hydroxypyrrolidine in a one-
pot process. Presumably, the diastereochemical outcome of
the pyrrolinium reduction would be governed by the sub-
stituent at C2, affording all-syn pyrrolidines (e.g., 7) from
1,2-anti-chlorohydrins.17 Despite the apparent simplicity of
this route, we were cognizant of the potential side products
As detailed in Scheme 2, treatment of the lithium enolate
derived from 2-undecanone (11) with (2R)-2-chlorohydro-
cinnamaldehyde (12)14,18 (Scheme 2) afforded a mixture of
the diastereomeric chlorohydrins 13 and 14 (dr ) 4:1).19 The
relatively low level of stereocontrol in the aldol reaction was
attributed to the ꢀ-phenyl substituent in the R-chloroaldehyde
12,20 as lithium aldol reactions of both linear or branched
aliphatic R-chloroaldehydes typically provide the corre-
sponding anti-chlorohydrins with diastereomeric ratios in
excess of 10:1.13 Notwithstanding, with optically enriched19
ꢀ-ketochlorohydrin 13 in hand, a number of experiments
were carried out to assess the viability of the imine formation/
cyclization/reduction strategy. For example, treatment of the
ꢀ-ketochlorohydrin 13 with methylamine in THF-D8 resulted
(10) For early syntheses of preussin, see: (a) Pak, C. S.; Lee, G. H. J.
Org. Chem. 1991, 56, 1128. (b) Shimazaki, M.; Okazaki, F.; Nakajima, F.;
Ishikawa, T.; Ohta, A. Heterocycles 1993, 36, 1823. (c) McGrane, P. L.;
Livinghouse, T. J. Am. Chem. Soc. 1993, 115, 11485. (d) Deng, W.;
Overman, L. E. J. Am. Chem. Soc. 1994, 116, 11241. (e) Overhand, M.;
Hecht, S. H. J. Org. Chem. 1994, 59, 4721.
(15) Our anticipation that imine formation (i.e., 4 f 5) would occur
preferentially over direct chloride displacement was predicated by the fact
that (3S,4R)-3-chloro-1-phenyl-4-octanol failed to react with methylamine
in THF after 24 h at rt.
(16) For the intramolecular displacement of a primary alkyl chloride
by an imine, see: (a) De Kimpe, N.; D’Hondt, L.; Stanoeva, E. Tetrahedron
Lett. 1991, 32, 3879. (b) Maeda, K.; Yamamoto, Y.; Tomimoto, K.; Mase,
T. Synlett 2001, 1808. (c) Tehrani, K. K.; D’hooghe, M.; De Kimpe, N.
Tetrahedron 2003, 59, 3099. (d) Breuning, M.; Steiner, M.; Mehler, C.;
Paasche, A.; Hein, D. J. Org. Chem. 2009, 74, 1407. (e) Reddy, L. R.;
Prashad, M. Chem. Commun. 2010, 46, 222.
(11) For a comprehensive review of preussin syntheses up to June 2003,
see: Basler, B.; Brandes, S.; Speigel, A.; Bach, T. Top. Curr. Chem. 2005,
243, 1.
(12) For recent syntheses of preussin, see: (a) Davis, F. D.; Zhang, J.;
Qiu, H.; Wu, Y. Org. Lett. 2008, 10, 1433. (b) Gogoi, N.; Boruwa, J.; Barua,
N. C. Eur. J. Org. Chem. 2006, 1722. (c) Bertrand, M. B.; Wolfe, J. P.
Org. Lett. 2006, 8, 2353. (d) Davis, F. A.; Deng, J. Tetrahedron 2004, 60,
5111. (e) Canova, S.; Bellosta, V.; Cossy, J. Synlett 2004, 1811. (f) Okue,
M.; Watanabe, H.; Kasahara, K.; Yoshida, M.; Horinouchi, S.; Kitahara,
T. Biosci. Biotechnol. Biochem. 2002, 66, 1093.
(17) For the reduction of related pyrrolinium species see refs 10c, e and
11. Brenneman, J. B.; Martin, S. F. Org. Lett. 2004, 6, 1329.
(18) The R-chloroaldehyde 12 was prepared following the procedure
reported by MacMillan and coworkers in ref 14c in 98% ee. The optical
purity of 12 was determined by chiral HPLC analysis (see the Supporting
Information) following conversion to the ꢀ-ketochlorohydrin 13.
(19) The ꢀ-ketochlorohydrins 13 and 14 decomposed to varying degrees
when purified by silica gel flash chromatography. Consequently, both
compounds were ultimately purified by recrystalization from hexanes.
Notably, the optical purity of the recrystallized ꢀ-ketochlorohydrin 13 was
>99.5% ee as determined by chiral HPLC analysis.
(13) Kang, B.; Mowat, J.; Pinter, T.; Britton, R. Org. Lett. 2009, 11,
1717. (b) Kang, B.; Chang, S.; Decker, S.; Britton, R. Org. Lett. 2010, 12,
1716.
(14) For the asymmetric R-chlorination of aldehydes, see: (a) Brochu,
M. P.; Brown, S. P.; MacMillan, D. W. C. J. Am. Chem. Soc. 2004, 126,
4108. (b) Halland, N.; Braunton, A.; Bachmann, S.; Marigo, M.; Jørgensen,
K. A. J. Am. Chem. Soc. 2004, 126, 4790. (c) Amatore, M.; Beeson, T. D.;
Brown, S. P.; MacMillan, D. W. C. Angew. Chem., Int. Ed. 2009, 48, 5121.
Org. Lett., Vol. 12, No. 18, 2010
4035

in complete conversion to the corresponding N-methylimine
In order to explore the scope and limitations of this
hydroxypyrrolidine synthesis, a variety of readily available
ꢀ-ketochlorohydrins13,25 were treated with methylamine in
THF followed, after 6 h, by the addition of NaCNBH3 (Table
2). In general, these reactions provided the corresponding
1
after 6 h, as observed by H and 13C NMR spectroscopy.
While the pyrrolinium intermediate 15 was not detected in
these experiments, after extended periods of time (>6 h) or
at elevated temperatures, conversion to the corresponding
N-methylpyrrole 16 was observed, indicating the N-meth-
ylimine (or enamine) is a competent precursor to the desired
pyrrolinium 15. These observations proved crucial to the
eventual success of this strategy, as the addition of reducing
agents prior to imine formation (i.e., 6 h) resulted in the
undesirable production of diols and amino chlorohydrins21
(e.g., see 8 and 9 (Scheme 1)), while further delaying the
reduction led to increased amounts of pyrrole 16. As
indicated in Table 1, however, addition of reducing agents
Table 2. Stereoselective Synthesis of 3-Hydroxypyrrolidinesa
Table 1. Reductive Amination of ꢀ-Ketochlorohydrin 13a
entry
reducing agent
yieldb
3:17c
1
2
3
4
5
6
NaBH4
DIBAL-H
NaBH(OAc)3
Me4NBH(OAc)3
H2, Pd/C
50%
<5%
69%
80%
23%
91%
1:1
1:1
1:1
3:1
6:1
6:1
NaCNBH3
a General procedure: a 0.1 M solution of 13 and CH3NH2 (10 equiv) in
THF was stirred at rt for 6 h followed by addition of reducing agent. b Combined
isolated yield of 3 and 17. c Determined by analysis of 1H NMR spectra recorded
on the crude reaction product following treatment with TFA.
after complete imine formation provided the diastereomeric
hydroxypyrrolidines 3 and 17.22 For example, when the
intermediate imine was treated with NaBH4 or NaBH(OAc)3
(entries 1 and 3), a 1:1 mixture of preussin (3) and 5-epi-
preussin (17) was produced in good yield. While the
diastereoselectivity and yield of this process was improved
through the use of Me4NBH(OAc)3, the optimal reducing
agent proved to be NaCNBH3 (entry 6), which afforded a
6:1 mixture of hydroxypyrrolidines from which (+)-preussin
(3) was isolated in 78% yield.23 The spectral data (1H, 13C
NMR, IR, HRMS)24 derived from synthetic preussin was in
complete agreement with that reported for the natural
product. Notably, this three-step synthesis of preussin (3)
compares well with those reported in the literature and
presents new opportunities for the production of structural
analogues with potentially improved pharmaceutical proper-
ties, work that is ongoing in our laboratory.
(20) Low levels of diastereocontrol have also been reported for the
addition of methyl ketone derived lithium enolates to R-benzyloxyhydro-
cinnamaldehyde. See: Evans, D. A.; Cee, V. J.; Siska, S. J. J. Am. Chem.
Soc. 2006, 128, 9433.
a General procedure: A solution of the chlorohydrin and CH3NH2 (10
equiv) in THF was stirred at rt for 6 h followed by the addition of NaCNBH3.
b Isolated yield of major diastereomer. c Determined by analysis of 1H NMR
spectra recorded on the crude reaction product in CD3OD with excess TFA.
d The minor, C5-epimer was not observed in 1H NMR spectra recorded on
the crude reaction product.
(21) Treatment of the ꢀ-ketochlorohydrin 13 with MeNH2 and
NaBH(OAc)3 in THF provided a 2:1 mixture of diastereomeric aminochlo-
rohydrins and none of the desired 3-hydroxypyrrolidine.
(22) The 1H NMR spectral data derived from 5-epi-preussin (17) differed
from that reported for this substance in refs 10b and 12f, which were also
inconsistent. For characterization purposes, 5-epi-preussin (17) was treated
with an excess amount of TFA, which produced a 1:1 mixture of
diastereomeric ammonium salts prior to NMR spectroscopic analysis. See
the Supporting Information for details.
(23) Reductive amination of the TBS-protected ꢀ-ketochlorohydrin (i.e.,
13: R1 ) H, R2 ) OTBS) employing the conditions described in entry 6
(Table 1) afforded a 2.3:1 mixture of 3-silyloxypyrrolidines. Treatment of
this mixture with TBAF in THF afforded a 2.3:1 mixture of 3:17.
2,5-disubstituted 3-hydroxypyrrolidines in good to excellent
yield. Notably, while the characterization of C2-benzyl-
substituted pyrrolidines (e.g., 3 and 27-31) by NMR
spectroscopy did not present significant challenges, inter-
1
pretation of H and 13C NMR spectroscopic data acquired
4036
Org. Lett., Vol. 12, No. 18, 2010

on the C2-alkyl analogues of these substances proved
difficult. In particular, the H NMR spectra of these later
tochlorohydrin 13 with allyl-, benzyl-, and n-heptylamine
followed by NaCNBH3 afforded the corresponding N-allyl-,
N-benzyl-, and N-alkylpyrrolidines in modest to good yield.
Acetylation of the crude mixtures of hydroxypyrrolidines
proved necessary for both chromatographic purification and
characterization purposes. Contrary to the reductive amina-
tions with methylamine (Table 2) and heptylamine (Table
3, entry 3), the reactions involving allyl- and benzylamine
afforded almost equal amounts of diastereomeric pyrro-
lidines. In these cases (entries 1 and 2), nonselective
reduction27 of the intermediate ꢀ-hydroxy imine may precede
pyrrolidine formation. Notwithstanding, together with the
results presented in Table 2, a wide variety of 3-hydroxy-
pyrrolidines, functionalized at N1, C2, and C5, are now
readily available in three steps (i.e., aldehyde chlorination,
aldol coupling, reductive amination) from methyl ketones
and aliphatic aldehydes, both of which are commonly
available starting materials.
1
compounds were characterized by broad signals, and the
chemical shifts of diagnostic protons (e.g., N-CH3, H-2, H-3,
and H-5) varied with concentration and source of deuterio
solvent.22 Consequently, the NMR spectroscopic character-
ization of these substances, including NOE analysis, was
carried out on the corresponding TFA salts, which exhibited
sharp resonances and reproducible spectroscopic data. As
indicated in Table 2, the diastereoselectivity of these reactions
was comparable to or better than that observed during the
production of preussin (3) (dr ) 6:1, Scheme 2) and 2,3-
anti pyrrolidines could also be accessed from the corre-
sponding syn-configured ꢀ-ketochlorohydrins26 (e.g., entries
1, 3, and 5). For example, when the optimized conditions
were employed, reaction of the ꢀ-ketochlorohydrin 14 with
methylamine afforded 3-epi-preussin (27), an equipotent
analogue of the natural product preussin (3),12f in excellent
isolated yield. In all cases, reduction of the pyrrolinium
intermediate occurred preferentially from the face opposite
the substituent at C2, affording 2,5-syn pyrrolidines 27-36
as the major products. As an apparent limitation to this
methodology, the optimized reaction conditions failed to
provide significant quantities of the desired 3-hydroxypyr-
rolidines 32 and 33 from the phenylketones 22 and 23,
respectively. Allowing less time for pyrrolinium formation
and/or lowering the reaction temperature offered little
improvement on these results and in both cases the major
byproduct was the corresponding pyrrole.
In summary, we have developed a straightforward and
convergent synthesis of 2,5-disubstituted 3-hydroxy-pyrro-
lidines that involves treating readily available ꢀ-ketochlo-
rohydrins with a primary amine followed by NaCNBH3, all
at room temperature. Notably, this operationally straightfor-
ward and protecting group-free process is selective for the
production of 2,5-syn-pyrrolidines. The efficiency of this
process was demonstrated in a short (three step, 42% overall
yield) synthesis of the natural product (+)-preussin (3) as
well as a number of analogues of this substance. The
application of this methodology to the synthesis of more
structurally complex hydroxypyrrolidine-containing natural
products (e.g., 1 and 2, Figure 1) is currently underway in
our laboratory, and the results of these efforts will be reported
in due course.
Finally, the reductive amination of ketochlorohydrin 13
with amines other than methylamine was briefly investigated
(Table 3), and the versatility of this hydroxypyrrolidine
Table 3. Reductive Amination of ꢀ-Ketochlorohydrin 13a
Acknowledgment. This research was supported by NSERC-
Canada (R.B.) and the Province of BC through the ministry
of Advanced Education (J.A.D.).
Supporting Information Available: Characterization data
and detailed experimental procedures. This material is
available free of charge via the Internet at http://pubs.acs.org.
entry
R
pyrrolidine
yieldb (dr)c
OL101631E
1
2
3
allyl
benzyl
n-heptyl
37a
37b
37c
60% (1.3:1)
37% (1.3:1)
30% (8:1)
(24) The specific rotation for (+)-preussin (3) ([R]D ) +19.0, c 0.6,
CHCl3) was consistent with that reported for the natural product in ref 7
([R]D ) +22.0, c 1.0, CHCl3).
a General procedure: A solution of the chlorohydrin 13 and amine (10
equiv) in THF was stirred at rt for 6 h followed by the addition of NaCNBH3.
b Combined isolated yield of both diastereomers. c Determined by analysis
(25) For the preparation, characterization, and stereochemical assignment
of all compounds, see the Supporting Information.
1
of H NMR spectra recorded on the crude reaction product.
(26) These compounds were produced as the minor diastereomeric
products of the aldol reactions involving 2-chloro-3-phenylpropanal.
(27) For an example of a nonselective reductive amination of a
ꢀ-hydroxyketone, see: Enders, D.; Palecek, J.; Grondal, C. Chem. Commun.
2006, 655.
synthesis was demonstrated in the synthesis of N-function-
alized analogues of preussin. Thus, treatment of the ꢀ-ke-
Org. Lett., Vol. 12, No. 18, 2010
4037