Enzymatic desymmetrization of cis-1,3-cyclohexanedicarboxylic acid diesters

Article abstract of DOI:10.1016/S0957-4166(99)00060-9

cis-1,3-Cyclohexanedicarboxylic acid (1,3-CHDA) monoesters were prepared in high overall yield and high enantiomeric purity using a three step process from cis/trans-1,3-CHDA. The asymmetry is induced by an enzymatic hydrolytic desymmetrization of a meso

Full text of DOI:10.1016/S0957-4166(99)00060-9

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 10 (1999) 813–816
Enzymatic desymmetrization of cis-1,3-cyclohexanedicarboxylic
acid diesters
Neil W. Boaz ∗
Research Laboratories, Eastman Chemical Company, Kingsport, TN 37662, USA
Received 5 January 1999; accepted 9 February 1999
Abstract
cis-1,3-Cyclohexanedicarboxylic acid (1,3-CHDA) monoesters were prepared in high overall yield and high
enantiomeric purity using a three step process from cis/trans-1,3-CHDA. The asymmetry is induced by an
enzymatic hydrolytic desymmetrization of a meso cis-1,3-CHDA diester. A judicious choice of ester substituents
and enzyme can provide >80% overall yield of either enantiomer in >94% ee. © 1999 Elsevier Science Ltd. All
rights reserved.
Enzyme-catalyzed reactions are a powerful method for the generation of asymmetry, but are often
inefficient because they are usually used to perform resolutions of racemic mixtures. Desymmetrization
of meso compounds, where an achiral species is optimally converted to a single enantiomer of a chiral
molecule, avoids the 50% yield limits and difficult separations often associated with resolutions. This is
probably one of the simplest methods for the generation of an optically active material, provided that the
substrate is readily available and the enzymatic reaction proceeds with sufficient enantioselectivity.
Enantioselective hydrolyses of meso cis-1,2-cyclohexanedicarboxylic acid diesters have been reported
in a number of publications, invariably using pig liver esterase as the enzyme.1 The desymmetrized
monoester products have been used for the synthesis of both pharmaceutically active materials2
and natural products.3 There have also been reports of the desymmetrization of both cis-1,3-
cyclopentanedicarboxylic anhydride (via enzymatic alcoholysis) and cis-1,3-cyclopentanedicarboxylic
acid diesters (via enzymatic hydrolysis).4 There have been no reports of the desymmetrization of
cis-1,3-cyclohexanedicarboxylic acid (cis-1,3-CHDA) derivatives, although the enantiomerically
enriched monoester products, if available, should be useful chiral synthons. For example, cis-3R-3-
aminocyclohexanecarboxylic acid, a known GABA uptake inhibitor5 and pharmaceutical building
block,6 should be readily available from the cis-1,3-CHDA monoester by Hofmann rearrangement of the
corresponding amide.
The investigation of the enzymatic desymmetrization of a cis-1,3-CHDA species requires an all-cis
precursor. Unfortunately, the parent diacid is only available as a mixture of cis and trans isomers. It is
∗
E-mail: nwboaz@eastman.com
0957-4166/99/$ - see front matter © 1999 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(99)00060-9

814
N. W. Boaz / Tetrahedron: Asymmetry 10 (1999) 813–816
likely that the separation of the isomers would be troublesome, and it would certainly be more efficient
if the mixture could be converted into an all-cis species.
Fortunately, a method was developed some years ago7 that converted the cis/trans diacid mixture to
the all-cis anhydride by the action of acetic anhydride on the diacid. This apparently succeeds because
the cis arrangement is the only accessible geometry for cyclization to the anhydride and isomerization
of the trans isomer to cis is feasible under the reaction conditions. Indeed, heating a mixture of cis-
and trans-1,3-CHDA with acetic anhydride followed by removal of the volatiles at atmospheric pressure
afforded the all-cis anhydride, which could be readily recrystallized to high purity in >90% yield. This
cyclic anhydride appeared to be an attractive starting point for enzymatic investigation.
The enzyme-catalyzed alcoholysis of cis-1,3-CHDA anhydride (1) was the most obvious approach
to an optically pure monoester. This is a type of enzymatic reaction that is most often performed
in an organic solvent and is often a more flexible reaction than the opposite (hydrolysis) reaction. A
fairly hindered alcohol, isopropanol, was chosen as the alcohol reactant to minimize any non-enzymatic
alcoholysis (a major concern with anhydrides) that would compromise the enantiomeric purity of the
product. Unfortunately, few solvents were found that would dissolve the anhydride 1 and, once dissolved,
the enzymatic reactions of 1 were extremely sluggish. In addition, a survey of enzymes and solvents
afforded at best only moderate enantioselectivity (maximum 78% ee as determined by chiral capillary
GC) for the preparation of the monoisopropyl ester 3d.
The converse reaction, enzymatic hydrolysis of a cis-1,3-CHDA diester, was also investigated. cis-
1,3-CHDA anhydride (1) was readily converted in essentially quantitative yield to the various diesters
by reaction with the desired alcohol using acid catalysis with water removal. Somewhat surprisingly,
all of the diesters were oils as compared to the highly crystalline anhydride. Oils are highly desirable
for enzymatic hydrolysis reactions since, unlike solid substrates, they form good emulsions in water, a
requirement for a successful enzymatic hydrolysis. It was also important to note that the conversion to
the diester proceeded in all cases without any observable cis/trans isomerization as determined by control
reactions using cis/trans-1,3-CHDA.
In contrast to the anhydride, most of the diesters were readily hydrolyzed with a variety of enzymes
to afford clean conversion to the corresponding monoester. Conveniently, none of the enzymes examined
would produce the diacid even at extended reaction times. Screening a number of enzymes with a
variety of cis-1,3-CHDA diesters demonstrated that several of these enzymatic hydrolyses proceeded with
high enantioselectivity (enantiomeric excesses were determined by chiral capillary GC on a Cyclodex-
B column [J&W Scientific]). A selection of these results is presented in Table 1. There are three
important points to be learned from this data. First, the enzymatic hydrolyses of the diesters were
much more selective than the alcoholyis of the anhydride, with several of the reactions affording
enantioselectivities in the ‘useful’ range (>95% ee). Second, reactions exhibiting high but opposite
(and essentially equal) enantioselectivities were observed using lipases from Pseudomonas species and
Candida species, respectively. For example, Lipase PS-30 from Pseudomonas cepacia afforded (1R,3S)-
3-alkoxycarbonylcyclohexanecarboxylic acid [(1R,3S)-3] while Lipase AY-30 from Candida rugosa
afforded (1S,3R)-3. Thus either enantiomer of the 1,3-CHDA monoester could be obtained in high
enantiomeric purity through a judicious choice of enzyme. This complementary high enantioselectivity
from different commercial enzymes is quite unusual. Third, pig liver esterase, which had shown high
enantioselectivity for the hydrolysis of cis-1,2-CHDA diesters, exhibited poor enantioselectivity in this
1,3-system.

N. W. Boaz / Tetrahedron: Asymmetry 10 (1999) 813–816
815
Table 1
Enantiomeric excessi of cis-3 from the enzymatic hydrolysis of cis-2
Variation of the ester substituents can often afford a profound difference in the course of these types
of reactions. In this case the enantioselectivity peaked with the diethyl ester as a substrate, falling off
rapidly upon going to the dimethyl ester and gradually with increasing diester length.
A rapid reaction rate is also important for the practical application of this type of methodology. The
diethyl ester was superior in this respect as well, as the reaction slowed significantly as the ester length
increased. Surprisingly, the dimethyl ester exhibited lower reactivity than the diethyl ester.
The use of standard soluble enzymes for these reactions resulted in significant emulsions during
product isolation, causing both operational difficulties and lower yields. Control reactions indicated that
the culprit was the residual enzyme, thus removal of the enzyme prior to isolation could significantly
improve the process. Enzyme immobilization on a support is a standard method to allow enzyme removal
(by filtration) and has the added benefit of potentially allowing reuse of an enzyme that would normally
be discarded. Commercially available crystallized, cross-linked enzymes termed ChiroCLECs™ from
Altus Biologics were examined in these reactions. These systems are advantaged in that the enzyme
activity and selectivity are retained and lipases from the desired two microorganisms (Pseudomonas
and Candida species) are commercially available. Performing the enzymatic hydrolysis reactions with
the diethyl ester under standard conditions but using the ChiroCLECs™ afforded high enantioselectivity
(ChiroCLEC-PC™ [from Pseudomonas cepacia] gave 1R,3S-3b with 94% ee; ChiroCLEC-CR™ [from
Candida rugosa] gave 1S,3R-3b with >95% ee) and good rates, with no emulsion observed during
product isolation. Indeed, the product monoesters were obtained in nearly quantitative yield.
The absolute configuration of these cis-1,3-CHDA monoesters was determined by correlation with 3-
methyl-1-cyclohexanemethanol, which was separable by chiral GC analysis (30 m Cyclodex-B) into all
four stereoisomers. (R)-3-Methyl-1-(RS)-cyclohexanemethanol of known configuration at carbon three
was prepared from (R)-3-methylcyclohexanone (available from Aldrich Chemical Co.) by olefination
(Ph3P=CHOMe), hydrolysis of the enol ether (aq. HClO4), and reduction of the resulting (R)-3-methyl-1-

816
N. W. Boaz / Tetrahedron: Asymmetry 10 (1999) 813–816
(RS)-cyclohexanecarboxaldehyde (NaBH4). A single stereoisomer of 3-methyl-1-cyclohexanemethanol
was prepared from monoester 3b by reduction of the acid group to afford 3-ethoxycarbonyl-1-
cyclohexanemethanol (BH3·SMe2), alcohol tosylation, and exhaustive reduction of both the tosylate
and ester groups (LiAlH4). Comparison of the materials prepared above by chiral GC indicated that
the enantiomer prepared via Lipase PS-30 hydrolysis possessed the (R) configuration at carbon three of
3-methyl-1-cyclohexanemethanol (which corresponds to carbon one of compound 3b). Thus, Lipase PS-
30 hydrolysis of diethyl cis-1,3-CHDA afforded (1R,3S)-3-ethoxycarbonylcyclohexanecarboxylic acid
[(1R,3S)-3b] and that derived from Lipase AY-30-catalyzed hydrolysis was the (1S,3R)-isomer. These
configurational assignments should likely hold true for all of the other monoesters investigated.
With this last piece of data we now can prepare either 1,3-CHDA monoethyl ester enantiomer in high
enantiomeric purity (>94% ee) and in high overall yield (>80%) with known configuration by a three-step
process involving two chemical steps and one enzymatic step.
References
1. (a) Jones, J. B.; Sabbioni, G. J. Org. Chem. 1987, 52, 4565. (b) Gais, H. J.; Lukas, K. L.; Ball, W. A.; Braun, S.; Lindner,
H. J. Liebigs Ann. Chem. 1986, 687. (c) Lam, L. K. P.; Brown, C. M.; De Jeso, B.; Lym, L.; Toone, E. J.; Jones, J. B. J.
Am. Chem. Soc. 1988, 110, 4409. (d) Schneider, M.; Engel, N.; Honicke, P.; Heinemann, G.; Gorisch, H. Angew. Chem.,
Int. Ed. Engl. 1984, 23, 67.
2. (a) Turbanti, L.; Cerbai, G.; Di Bugno, C.; Giorgi, R.; Garzelli, G.; Criscuoli, M.; Renzetti, A. R.; Subissi, A.; Bramanti,
G.; DePriest, S. A. J. Med. Chem. 1993, 36, 699. (b) Turbanti, L; Cerbai, G.; Criscuoli, M. Eur. Pat. Appl. EP 337,348,
1989. (c) Hamilton, G. S.; Huang, Z.; Yang, X. J.; Patch, R. J.; Narayaran, B. A.; Ferkany, J. W. J. Org. Chem. 1993, 58,
7263. (d) Hamilton, G. S. PCT Intl. Appl. WO 92 10,099, 1992. (e) Brion, F.; Marie, C.; Mackiewicz, P.; Roul, J. M.;
Buendia, J. Tetrahedron Lett. 1992, 34, 4889.
3. Borzilleri, R. M.; Weinreb, S. M.; Parvez, M. J. Am. Chem. Soc. 1994, 116, 9789.
4. (a) Chenevert, R.; Martin, R. Tetrahedron: Asymmetry 1992, 3, 199. (b) Chenevert, R.; Lavoie, M.; Courchesne, G.; Martin,
R. Chem. Lett. 1994, 93.
5. Allan, R. D.; Johnston, G. A. R.; Twitchin, B. Aust. J. Chem. 1981, 34, 2231.
6. Jones, D. F.; Oldham, K. Eur. Pat. Appl. EP14057, 1980.
7. Corfield, G. C.; Crawshaw, A. J. Macromol. Sci., Chem. 1971, A5, 3.