Chemoenzymatic syntheses of cis- and trans-3-hydroxy-5-methylpiperidin- 2-ones

Article abstract of DOI:10.1016/S0040-4039(99)02069-9

An approach to the enantioselective synthesis of cis- and trans-3- hydroxy-5-methylpiperidin-2-ones from racemic methyl 4-methyl-5-nitro-2- oxopentanonate 8 is described via the first example of a lactate dehydrogenase catalysed combined kinetic resolution and reduction of a 2-oxo acid. In an alternative approach, stereoselective conjugate addition of nitromethane to a crotonyl camphorsultam gave access to the enantiopure 2-oxo esters (S)-8 and (R)-8 which, in turn, may be converted to the 3-hydroxy-5- methyl δ-lactams.

Full text of DOI:10.1016/S0040-4039(99)02069-9

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 397–401
Chemoenzymatic syntheses of cis- and
trans-3-hydroxy-5-methylpiperidin-2-ones
Stuart R. Crosby, Martin J. Hateley and Christine L. Willis ∗
School of Chemistry, University of Bristol, Cantock’s Close, Bristol BS8 1TS, UK
Received 30 September 1999; accepted 2 November 1999
Abstract
An approach to the enantioselective synthesis of cis- and trans-3-hydroxy-5-methylpiperidin-2-ones from
racemic methyl 4-methyl-5-nitro-2-oxopentanonate 8 is described via the first example of a lactate dehydrogenase
catalysed combined kinetic resolution and reduction of a 2-oxo acid. In an alternative approach, stereoselective
conjugate addition of nitromethane to a crotonyl camphorsultam gave access to the enantiopure 2-oxo esters (S)-8
and (R)-8 which, in turn, may be converted to the 3-hydroxy-5-methyl δ-lactams. © 2000 Elsevier Science Ltd. All
rights reserved.
Hydroxylated six-membered ring nitrogen containing heterocycles are common features of many
natural products and biologically active compounds;1 we have reported an efficient method for the
synthesis of 3-hydroxypiperidin-2-ones, valuable building blocks to these molecules.2 Our strategy
involved an enzyme catalysed reduction of 5-nitro-2-oxopentanoic acid to the corresponding 2-hydroxy
acids (S)-2 and (R)-2 followed by esterification, catalytic hydrogenation of the nitro group using a
platinum(IV) oxide catalyst and spontaneous intramolecular cyclisation as shown in Scheme 1.
Scheme 1.
∗
Corresponding author.
0040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(99)02069-9
tetl 15997

398
(S)-3-Hydroxypiperidin-2-one (S)-3 was obtained using commercially available lactate dehydrogenase
(LDH) from Bacillus stearothermophilus (BS-LDH) whilst the (R)-enantiomer, (R)-3 was prepared using
LDH from Staphylococcus epidermidis (SE-LDH). Recently we have been exploring further the utility
of this approach to the enantioselective synthesis of more complex 3-hydroxypiperidin-2-ones such as
those with a substituent at C-5. These investigations have led to some unexpected results which are now
reported.
Our initial target compounds were the cis- and trans-isomers of 3-hydroxy-5-methylpiperidin-2-ones
for which (S)- and (R)-4-methyl-5-nitro-2-oxopentanoic acids were required as the enzyme substrates.
Before embarking on the enantioselective synthesis of these compounds, the racemic 2-oxo acid 9 was
prepared to ensure that it was indeed a substrate for the LDHs. Conjugate addition of nitroalkanes
to α,β-unsaturated carbonyl compounds is well precedented3 and reaction of nitromethane with tert-
butyl crotonate in the presence of DBU gave tert-butyl 3-methyl-4-nitrobutanoate 4 in excellent yield
(Scheme 2). Acid catalysed hydrolysis of the ester, conversion of the resultant acid 5 to an acid chloride 6
and coupling with cyanomethylenetriphenylphosphorane in the presence of bis(trimethylsilyl)acetamide
(BSA) as a proton scavenger4 gave β-ketocyanophosphorane 7. Ozonolysis of 7 gave the required 2-oxo
ester 8 in 60% yield over the five steps.
Scheme 2. Reagents. (i) CH3NO2, DBU; (ii) H2SO4, CHCl3; (iii) SOCl2; (iv) Ph3PCHCN, BSA, CH2Cl2; (v) O3, MeOH,
CH2Cl2
For recognition in the active site of LDH, it was necessary to convert ester 8 into the corresponding
carboxylic acid 9. We have found that a convenient method is to use a lipase to catalyse the hydrolysis
under mild conditions and then, in the same pot, to reduce the ketone using LDH.5 Thus 8 was incubated
with a lipase from Candida rugosa (CRL) prior to addition of BS-LDH, NADH and a further enzyme,
formate dehydrogenase (FDH), which is used to recycle the expensive cofactor NADH (Scheme 3).6
After work-up, the products were methylated with diazomethane giving two diastereomeric hydroxy
1
esters 10 and 11 which were inseparable by column chromatography. The H NMR spectrum revealed
that 10 and 11 had not been formed in equal amounts as expected but in a 3:1 ratio, indicating that both
a kinetic resolution and reduction had occurred. It was not clear which stage gave the kinetic resolution,
either the lipase catalysed hydrolysis or the LDH catalysed reduction. There are many reports in the
literature of lipase catalysed resolution/hydrolysis of esters7 but there are no examples of a combined
reduction and kinetic resolution of a 2-oxo acid catalysed by a lactate dehydrogenase.
Various methods could be adopted to establish the stage of kinetic resolution, and we elected
to repeat the reaction sequence changing just a single parameter, the oxidoreductase. The H205Q
mutant of D-hydroxyisocaproate dehydrogenase from Lactobacillus delbrueckii (LB-hicDH) has broader
substrate specificity than BS-LDH giving (R)-2-hydroxy acids in excellent yields and enantioselective
excesses.8 Hence, the hydrolysis/reduction/methylation protocol on 8 was repeated using LB-hicDH as
the oxidoreductase (Scheme 3). After purification by column chromatography, a 1:1 mixture (by 1H NMR
spectroscopy) of (2R,4R)- and (2R,4S)-hydroxy esters 12 and 13 was isolated in 90% overall yield. This
result confirmed that kinetic resolution is occurring with the BS-LDH catalysed reduction of 2-oxo acid
9 and not at the stage of the lipase catalysed hydrolysis of 8.

399
Scheme 3. Reagents. (i) Candida rugosa lipase; (ii) BS-LDH, NADH, FDH; (iii) CH2N2, MeOH; (iv) NADH, FDH and either
LB-hicDH or SE-LDH; (v) H2, PtO2; (vi) PhCH2OCOCl, Et3N, CH2Cl2
Next it was necessary to determine whether the syn or anti hydroxy acid was the major product from
the BS-LDH catalysed reduction of 9. Thus, the mixture of 10 and 11 was converted to the corresponding
six-membered ring lactams 14 and 15 by catalytic hydrogenation and protection (Scheme 3). In the 1H
NMR spectrum of the major diastereomer, the signal (δ1.60) assigned to 4-Hax appeared as a quartet with
coupling J 12.5 arising from a geminal coupling with 4-Heq and two axial–axial couplings to 5-H and
3-H in accord with the cis-lactam 14.9 By extrapolation, these results indicate that the BS-LDH catalysed
reduction of 9 gave, after methylation, a 3:1 mixture of (2S,4R)- to (2S,4S)-hydroxy esters 10 and 11.
Some caution is needed in the interpretation of these data as the hydroxylated lactams will not exist
in a true chair conformation. To verify these conclusions, a single enantiomer of methyl 4-methyl-5-
nitro-2-oxopentanoate 8 of known configuration at C-4 was required for the biotransformations. The
required stereogenic centre was created by treatment of (1S,2R)-N-[(E)-crotonyl]bornane-10,2-sultam
1610 with nitromethane in the presence of DBU in THF and DMPU giving a 3:1 mixture of 17 and
18 in 90% yield (Scheme 4). Although 17 and 18 were inseparable by column chromatography, the
major diastereomer 17 was readily isolated by crystallisation from methanol and its structure confirmed
by X-ray crystallography. Cleavage of the auxiliary with lithium hydroperoxide gave (S)-3-methyl-4-
nitrobutanoic acid which was converted to methyl (S)-4-methyl-5-nitro-2-oxopentanoate (S)-8 by the
same sequence of reactions as used for the racemic material (Scheme 2). The (R)-enantiomer, (R)-8 was
similarly prepared using (1S,2R)-N-[(E)-crotonyl]bornane-10,2-sultam as the starting material.
Each 2-oxo ester was then subjected to the three stage hydrolysis/BS-LDH catalysed
reduction/methylation procedure (Scheme 5). Reduction of the (S)-enantiomer was slow and eventually
gave the (2S,4S)-diastereomer 11 in 55% yield over the 3 steps. Reduction of the (R)-enantiomer
proceeded smoothly to give the (2S,4R)-2-hydroxy ester 10 in 85% yield from (R)-8. All spectral data
were in accord with those previously obtained thus confirming that the (R)-2-hydroxy acid was indeed
favoured over the (S)-enantiomer for reduction by BS-LDH.

400
Scheme 4. Reagents. CH3NO2 (6 equiv.), DBU (6 equiv.), THF, DMPU
Finally, the reduction of racemic 4-methyl-5-nitro-2-oxopentanoic acid 9 with SE-LDH was examined
(Scheme 3). This reaction also led to both a kinetic resolution and reduction but the resolution was less
pronounced than in the case of BS-LDH giving a 1.8:1 mixture of the (2R,4S)- and (2R,4R)-diastereomers
12 and 13. Interestingly, in contrast to the reaction with BS-LDH, the (S)-enantiomer was the preferred
substrate for SE-LDH.
Scheme 5. Reagents. (i) Candida rugosa lipase then BS-LDH, NADH, FDH; (ii) CH2N2
In conclusion, we have reported the first examples of a lactate dehydrogenase catalysed combined
kinetic resolution and reduction of a 2-oxo acid. These reactions have interesting potential in organic
synthesis, enabling the creation of two asymmetric centres in a single pot process as demonstrated here
by the enantioselective synthesis of the cis- and trans-3-hydroxy-5-methylpiperidin-2-ones 14 and 15. In
addition, these 2-oxo acids (S)-9 and (R)-9 are valuable substrates to further probe the active sites of both
SE- and BS-LDHs.

401
Acknowledgements
We are very grateful to the following for financial support: BBSRC (S.R.C.), EPSRC (M.J.H.) and
Genzyme (M.J.H.). We thank Dr. J. Charmant for the X-ray analysis and Professor J. J. Holbrook for a
gift of LB-hicDH.
References
1. For a review see, for example: Michael, J. P. Nat. Prod. Rep. 1997, 14, 619.
2. Gibbs, G.; Hateley, M. J.; McLaren, L.; Welham, M.; Willis, C. L. Tetrahedron Lett. 1999, 40, 1069 and references cited
therein.
3. For example: Sebti, S.; Boukhal, H.; Hanafi, N.; Boulaajaj, S. Tetrahedron Lett. 1999, 40, 6207. Backvall, J.-E; Ericsson, A.
M.; Plobeck, N. A.; Juntunen, S. K. Tetrahedron Lett. 1992, 33, 131.
4. Wasserman, H. H.; Ho, W.-B. J. Org. Chem. 1994, 59, 4364.
5. Macritchie, J. A.; Silcock, A.; Willis, C. L. Tetrahedron: Asymmetry 1997, 8, 3895.
6. Shaked, Z.; Whitesides, G. M. J. Am. Chem. Soc. 1980, 102, 7104.
7. Turner, N. J. Nat. Prod. Reports 1994, 11, 1
8. Sutherland, A.; Willis, C. L. J. Org. Chem. 1998, 63, 7764 and references cited therein.
1
9. H NMR data: Lactam 14, δ (300 MHz, CDCl3) 1.11 (3H, d, J 6.5, 5-CH3), 1.60 (1H, q, J 12.5, 4Hax), 2.30–2.40 (2H, m,
5-H and 4-Heq), 3.40 (1H, t, J 11, 6-Hax), 3.63 (1H, ddd, J 11, 5, 6-Heq), 4.24 (1H, dd, J 12.5, 6, 3-H), 5.28 (2H, s, PhCH2),
7.38 (5H, m, aromatics). Lactam 15, δ (300 MHz, CDCl3) 1.19 (3H, d, J 6.5, 5-CH3), 1.96 (2H, m, 4-H2), 2.51 (1H, oct, J
6.5, 5-H), 3.41 (1H, dd, J 11, 6.5, 6-HH), 3.79 (1H, dd, J 11, 6.5, 6-HH), 4.37 (1H, t, J 6, 3-H), 5.28 (2H, s, PhCH2), 7.38
(5H, m, aromatics).
10. Vandewalle, M.; Van der Eycken, J.; Oppolzer, W.; Vullioud, C. Tetrahedron 1986, 42, 4035