Development of an acid-washable tag for the separation of enantiomers from bioresolutions

Article abstract of DOI:10.1021/op900028d

A separation tool involving the use of a vinyl ester amino acid as acyl donor in a bioresolution has been developed. The acidwashable acyl group acts as a removable tag, facilitating separation of the secondary alcohol from the bioresolution product mixtu

Full text of DOI:10.1021/op900028d

Organic Process Research & Development 2009, 13, 706–709
Development of an Acid-Washable Tag for the Separation of Enantiomers from
Bioresolutions
Maude Brossat, Thomas S. Moody,* Florian de Nanteuil, Stephen J. C. Taylor, and Fatima Vaughan
Biocatalysis Group, Almac Sciences, Almac House, 20 Seagoe Industrial Estate, CraigaVon BT63 5QD, Northern Ireland
Scheme 1. Selective acylation of a secondary alcohol using
a lipase enzyme
Abstract:
A separation tool involving the use of a vinyl ester amino acid as
acyl donor in a bioresolution has been developed. The acid-
washable acyl group acts as a removable tag, facilitating separation
of the secondary alcohol from the bioresolution product mixture.
The use of these acyl donors from cheap, commercially available
amino acids has been demonstrated in conjunction with the
hydrolase enzyme Candida antarctica lipase B in the selective
acylation of a variety of secondary alcohols.
(basic workup).4 The resulting hemisuccinate product can be
separated from the unreacted alcohol by an aqueous base
workup. A summary of purification techniques used for the
separation of enantiomers from a bioresolution is shown in
Figure 1.
To our knowledge there are no reported acid-washable acyl
donors used in the separation of enantiomers from a bioreso-
lution in the literature. The present work describes an easy and
efficient method for the separation of products from a bioreso-
lution of secondary alcohols using vinyl ester amino acids as
acyl donors which lead to esters that are aqueous acid soluble.
This methodology can be applied to the industrial preparation
of valuable chiral secondary alcohols where the other techniques
fail in the separation workup.
It has been reported that proteases catalyse transesterifications
of alcohols with some N-protected vinyl ester amino acids.5
We decided to investigate this biotransformation using N-Boc-
protected vinyl ester amino acids for the bioresolution of
secondary alcohols using the amino moiety as an acid-washable
tag as shown in Scheme 2.
Introduction
Despite the dramatic improvements in enantioselective
synthesis and chromatographic separation methods, bioresolu-
tion still remains one of the most inexpensive and operationally
simple methods for producing pure enantiomers on a large scale
in the chemical industry.1 Typically the reactions are carried
out under ambient and neutral conditions. For example, hydro-
lase-catalysed kinetic resolutions have been widely used for the
synthesis of enantiopure secondary alcohols as shown in Scheme
1. These are important intermediates for the preparation of active
pharmaceutical ingredients (APIs).2
Following identification of the biocatalyst used in the
selective bioresolution, the separation of the modified enanti-
omers becomes the processing problem. The kinetic bioreso-
lution of racemic secondary alcohols is usually carried out by
a selective transesterification using acyl donors such as the enol
esters, vinyl acetate or isopropenyl acetate. The use of these
vinyl esters as acyl donors has the drawback of the separation
of the ester (product) and the alcohol (substrate) by column
chromatography, which can be costly and troublesome espe-
cially at large scale. To avoid this problem, an extractive workup
using the difference in polarity between the two products, by
choosing a long-chain fatty acid (vinyl butyrate, vinyl decanoate)
as acyl donor, can be used (neutral workup).3 An alternative
strategy involves cyclic anhydrides, such as succinic anhydride
Results and Discussion
To develop a process involving the bioresolution of second-
ary alcohols followed by an acid wash separation of products,
the hydrolase enzyme Candida antarctica lipase B (CAL-B)
was chosen since it is well recognized and accepted as a useful
catalyst in the chemical industry.6
The vinyl ester amino acids were synthesised from the
corresponding N-Boc-amino acids by treatment with vinyl
acetate, according to the method of Lobell and Schneider,7 in
(4) (a) Goswami, A.; Howell, J. M.; Hua, E. Y.; Mirfakhrae, K. D.;
Soumeillant, M. C.; Swaminathan, S.; Qian, X.; Quiroz, F. A.; Vu,
T. C.; Wang, X.; Zheng, B.; Kronenthal, D. R.; Patel, R. N. Org.
Process. Res. DeV 2001, 5, 415–420. (b) Bouzemi, N.; Debbeche, H.;
Aribi-Zouiou, L.; Fiaud, J.-C. Tetrahedron Lett. 2004, 45, 627–630.
(5) (a) Lloyd, R. C.; Dickman, M.; Jones, J. B. Tetrahedron: Asymmetry
1998, 9, 551–561. (b) Davis, B. G.; Fairbanks, A. J. Method for
preparing ester linked peptide-carbohydrate conjugates. US 2004/
0208883, A1, 2004.
* Author for correspondence. E-mail: tom.moody@almacgroup.com.
(1) (a) Ou, L.; Xu, Y.; Ludwig, D.; Pan, J.; Xu, J. H. Org. Process. Res.
DeV. 2008, 12, 192–195. (b) Park, O.-J.; Lee, S.-H.; Park, T.-Y.; Chung,
W.-G.; Lee, S.-W. Org. Process. Res. DeV. 2006, 10, 588–591. (c)
Yazbeck, D.; Derrick, A.; Panesar, M.; Deese, A.; Gujral, A.; Tao, J.
Org. Process. Res. DeV. 2006, 10, 655–660.
(2) (a) Ghanem, A.; Aboul-Enien, H. Y. Tetrahedron: Asymmetry 2004,
15, 3331–3351. (b) Gotor-Ferna´ndez, V.; Brieva, R.; Gotor, V. J. Mol.
Catal. B: Enzym. 2006, 40, 111–120.
(6) (a) Vaidyanathan, R.; Hesmondhalgh, L.; Hu, S. Org. Process. Res.
DeV. 2007, 11, 903–906. (b) Gotor, V. Org. Process. Res. DeV. 2002,
6, 420–426. (c) Reigada, J. B.; Tcacenco, C. M.; Andrade, L. H.; Kato,
M. J.; Porto, A. L. M.; Lago, J. H. G. Tetrahedron: Asymmetry 2007,
18, 1054–1058.
(3) Halle, R.; ter; Bernet, Y.; Billard, S.; Bufferne, C.; Carlier, P.; Delatre,
C.; Flouzat, C.; Humblot, G.; Laigle, J. C.; Lombard, F.; Wilmouth, S.
Org. Process. Res. DeV 2004, 8, 283–286.
(7) Lobell, M.; Schneider, M. P. Synthesis 1994, 375–377.
706
•
Vol. 13, No. 4, 2009 / Organic Process Research & Development
10.1021/op900028d CCC: $40.75  2009 American Chemical Society
Published on Web 05/18/2009

Figure 1. Purification techniques used for the separation of enantiomers after a bioresolution.
Scheme 2. Bioresolution of secondary alcohols using vinyl
ester amino acids as acyl donors
Scheme 4. Bioresolution of 1-phenylethanol (R/S)-7 using
CAL-B and N-Boc-ꢀ-alanine vinyl ester 5 as acyl donor
Table 1. Bioresolution of (R/S)-7 using acyl donors 4, 5, and 6
(S)-alcohol
(R)-ester
conversion
(%)b
acyl donor
(% ee)a
(% ee)a
Eb
5
4
6
89
48.5
4.5
99.3
96.8
100
47
33
5
>200
105
>200
a Determination of ee by HPLC using CHIRALCEL OJ-H column.
Determination of the ee of the product was achieved by acidic hydrolysis in
MeOH using 5 M aqueous HCl and saponification followed by extraction of the
reaction mixture. b Calculated from eealcohol and eeester after 16 h reaction.
test the methodology and the bioresolution showed high activity
and selectivity using N-Boc-ꢀ-alanine vinyl ester 5 with an
E-value of >200 as shown in Scheme 4.
Scheme 3. Synthesis of the vinyl ester amino acids from
N-Boc-protected amino acids
The bioresolution of (R/S)-7 using acyl donors 4, 5, and 6
is shown in Table 1. The use of valine 6 as the acyl donor
resulted in a sluggish reaction, potentially due to steric hindrance
of this acyl donor.
From these results we decided to investigate the solvent
tolerance for the acylation using 5 as the acyl donor. The
bioresolution of (R/S)-7 showed high selectivity for all solvents
screened with E-values >90 as shown in Table 2.
These results show that this bioresolution can be performed
in a wide range of organic solvents with good E-values.
Temperature optimisation studies were carried out from 20 to
70 °C as shown in Figure 2, and as expected, the bioresolution
showed a temperature profile with maximum activity being
achieved at 50 °C with an E-value of >200.
the presence of potassium hydroxide and a catalytic amount of
palladium acetate as shown in Scheme 3.
The acyl donors selected for investigation were derived from
cheap, commercially available amino acids glycine 1, ꢀ-alanine
2 and valine 3, yielding the corresponding acyl donors 4, 5,
and 6. Initially, the CAL-B-catalysed transesterification of
racemic 1-phenylethanol (R/S)-7 in MTBE was investigated to
Vol. 13, No. 4, 2009 / Organic Process Research & Development
•
707

Table 2. Solvent screen for the bioresolution of (R/S)-7 using
acyl donor 5
(S)-alcohol
(R)-ester
conversion
(%)b
solvent
(% ee)a
(% ee)a
Eb
MTBE
98.6
92.7
100
58.6
68.5
99.9
98.8
38.9
99.6
93.6
96.7
94.3
100
96.8
84.4
100
51
49
52
37
41
54
5
170
>200
>200
>200
134
hexane
heptane
THF
MeTHF
toluene
acetonitrile
acetone
cyclohexane
90
>200
>200
141
99
89.8
28
53
Figure 4. Secondary alcohol substrates used to demonstrate
the bioresolution with acyl donor 5.
a Determination of ee by HPLC using CHIRALCEL OJ-H column. Dete-
rmination of the ee of the product was achieved by acidic hydrolysis in MeOH
using 5 M aqueous HCl, saponification followed by extraction of the reaction
mixture. b Calculated from eealcohol and eeester after 16 h reaction.
Table 3. Bioresolution of 7, 9-12 with acyl donor 5
(S)-alcohol
(R)-ester
conversion
(%)b
substrate
(% ee)a
(% ee)a
Eb
7
80.0
66.2
72.4
99.0
54.2
99.0
99.6
99.0
92.0
99.0
45
40
42
52
44
>200
>200
>200
125
9
10
11
12
>200
a Determination of ee by HPLC using CHIRALCEL OJ-H column.
Determination of the ee of the product was achieved by acidic hydrolysis in
MeOH using 5 M aqueous HCl and saponification followed by extraction of the
reaction mixture. b Calculated from eealcohol and eeester after 16 h reaction.
Figure 2. Temperature profile for the bioresolution of (R/S)-7
with acyl donor 5 in MTBE.
Figure 5. Reaction profile for the bioresolution of 1-phenyle-
thanol (R/S)-7.
Figure 3. Bioresolution profile of (R/S)-7 at different enzyme
loadings with acyl donor 5 in MTBE.
30 g, using the acid-washable acyl donor 4, which was readily
available to us and gave good E-value for the bioresolution.
The reaction profile is shown in Figure 5. The bioresolution
was stopped after 16 h with ee of alcohol and ester being 91.5%
and 98.9%, respectively.
The reaction mixture was filtered to remove the enzyme
followed by the addition of 10% v/v of methanol and 5 M HCl
(5 volumes). The mixture was stirred until complete disappear-
ance of N-Boc-protected amino ester due to its hydrolysis to
the amino ester. The reaction mixture was then separated and
the organic layer yielded (S)-1-phenylethanol (14.1 g, 47%
yield) with 91.5% ee. The acidic aqueous layer was then
adjusted to pH 12 using 5 M NaOH and stirred until complete
saponification had occurred. The mixture was then extracted,
With increasing pressures to lower costs, the enzyme loading
was also investigated as shown in Figure 3. The bioresolution
was possible at a 10% wt/wt loading of CAL-B with reaction
completion in less than 12 h.
The methodology was expanded to other secondary alcohol
substrates 9-12 as shown in Figure 4. The bioresolutions were
carried out at an enzyme loading of 50% wt/wt for rapid turn
around of results with N-Boc-ꢀ-alanine vinyl ester 5 as acyl
donor in MTBE.
A summary of the E-values are shown in Table 3. To our
delight the selectivity of the enzyme was maintained to >125
for the different substrates 7-12.
Scale-Up and Separation. To test the scalability of the
biotransformation, the bioresolution of (R/S)-7 was scaled to
708
•
Vol. 13, No. 4, 2009 / Organic Process Research & Development

1
and the organic layer yielded (R)-1-phenylethanol (12.7 g, 42%
N-Boc-glycine vinyl ester (4): yellow oil. 80% yield. H
yield) with 98.9% ee.
NMR (DMSO, 500 MHz) δ 7.31 (1H, t, J ) 5.0, NH), 7.20
(1H, dd, J1 ) 5.0, J2 ) 15.0, CH2)CH-O), 4.93 (1H, dd, J1
) 1.4, J2 ) 15.0, CH2)CH), 4.70 (1H, dd, J1 ) 1.4, J2 ) 5.0,
CH2)CH), 3.80 (2H, d, J ) 5.0, CH2-NH), 1.40 (9H, s,
(CH3)3).
Conclusion
This work has resulted in a new methodology that can be
added to the chemists’ tool-box for the separation of enantiomers
after a bioresolution. The acid-soluble tag fully complements
the other workup techniques and can be considered as a potential
workup solution where the other techniques fail. We are
currently investigating other protecting groups that can be
cleaved under a range of conditions so that the acyl donor can
be tuned to downstream chemistry to minimise processing time.
In addition, the presence of the amino moiety has the added
advantage for the potential ee polish through crystallization of
a corresponding salt. This work is currently under investigation.
N-Boc-ꢀ-alanine vinyl ester (5): yellow oil. 85% yield. 1H
NMR (DMSO, 500 MHz) δ 7.20 (1H, dd, J1 ) 6.3, J2 ) 14.0,
CH2)CH-O), 6.92 (1H, s, NH), 4.35 (1H, dd, J1 ) 1.5, J2 )
14, CH2)CH), 4.66 (1H, dd, J1 ) 1.5, J2 ) 6.3, CH2)CH),
3.20 (2H, t, J ) 6.8, CH2-CH2-NH), 2.56 (2H, t, J ) 6.8,
CH2-CH2-NH), 1.38 (9H, s, (CH3)3).
1
N-Boc-valine vinyl ester (6): yellow oil. 77% yield. H
NMR (DMSO, 500 MHz) δ 7.34 (1H, d, J ) 5.0, NH), 7.22
(1H, dd, J1) 5.0, J2 ) 15.0, CH2)CH-O), 4.93 (1H, dd, J1 )
1.4, J2 ) 15.0, CH2)CH), 4.71 (1H, dd, J1) 1.4, J2 ) 5.0,
CH2)CH), 3.90 (1H, m, CH-NH), 2.05 (1H, m, CH-CH3),
1.40 (9H, s, (CH3)3), 0.90 (6H, m, 2 × CH3).
Experimental section
Chemicals and Enzymes. Chemicals were purchased from
Alfa Aesar. Hydrolase enzyme CAL-B (Novozym 435) was
obtained from Enzagen Ltd.
Large-Scale Bioresolution of 1-Phenylethanol Using N-
Boc-glycine Vinyl Ester. To a stirred ((400 rpm) solution of
30 g (0.24 mol) of 1-phenylethanol and 36 g (0.18 mol) of
N-Boc-glycine vinyl ester in 600 mL of MTBE (20 vol) was
added 7.5 g of CAL-B (25 wt %/wt). After stirring for 16 h at
30 °C, a sample was taken and HPLC analysis indicated 91.5%
ee for the alcohol. After filtering off the enzyme, 60 mL (10%
v/v) of methanol and 150 mL of 5 M HCl (5 vol) were added
to the solution. The solution was stirred at 30 °C for 1 h. The
disappearance of the Boc-amino ester was checked by TLC (8:2
hexane/EtOAc). At the end of the Boc-deprotection, the
remaining 1-phenylethanol was extracted three times with 150
mL (5 vol) of MTBE. The solvent was removed to yield (S)-
1-phenylethanol (14.1 g, 47% yield) with 91.5% ee (97% purity
by NMR). The pH of the aqueous phase was then adjusted to
pH 12 using 5 M NaOH. The solution was stirred at 30 °C for
1 h, resulting in complete saponification. The mixture was then
extracted two times with 150 mL (5 vol) of MTBE. The organic
phase was dried over Na2SO4, filtered, and solvent evaporated
to afford 12.7 g (42% yield) of (R)-1-phenylethanol with 98.9%
ee (98% purity by NMR).8
Analytical Methods. 1H NMR spectra were recorded at 500
MHz on a Bruker AV-500 spectrometer; shifts are relative to
internal TMS.
The enantiomeric excesses were measured by chiral station-
ary phase HPLC on Chiracel OJ-H column (250 mm × 4.6
mm × 10 µm, Daicel Chemical Industries) with UV detection
(λ ) 210 nm). For substrate 7 1-phenylethanol, eluent hexane/
2-propanol (9:1); flow rate 0.5 mL/min; typical retention times
were 17.0 min (S)-1-phenylethanol, 19.3 min (R)-1-phenyle-
thanol. For substrate 9 1-(4-trifluoromethylphenyl)ethanol eluent
hexane/2-propanol (9:1); flow rate 0.25 mL/min; typical reten-
tion times were 20.0 min (S), 21.2 min (R). For substrate 10
1-(4-methoxyphenyl)ethanol eluent hexane/2-propanol (9:1);
flow rate 0.8 mL/min; typical retention times were 18.1 min
(S), 19.2 min (R). For substrate 11 1-(naphthalen-2-yl)ethanol
eluent hexane/2-propanol (9:1); flow rate 1 mL/min; typical
retention times were 16.5 min (S), 21.3 min (R). For substrate
12 1-(1,2,3,4-tetrahydronaphthalen-2-yl)ethanol eluent hexane/
2-propanol (9:1); flow rate 0.5 mL/min; typical retention times
were 14.0 min (S), 17.1 min (R).
General Method for the Preparation of the N-Boc-vinyl
Ester Amino Acids. A mixture of N-Boc-amino acid (2.65
mmol), Pd(OAc)2 (0.0265 mmol, 0.01 equiv), KOH (0.265
mmol, 0.1 equiv) and 5 mL of vinyl acetate (10 vol) was stirred
for 24 h at room temperature under N2. The reaction mixture
was filtered on Celite, and the solvent was evaporated under
reduced pressure. The crude product was purified by flash
chromatography (95:5 EtOAc/hexane) to give the corresponding
N-Boc-vinyl ester amino acid.
Acknowledgment
We thank Prof. Tony McKervey for reviewing the manu-
script and providing valuable suggestions.
Received for review February 10, 2009.
OP900028D
(8) No trace of diketopiperazine was observed.
Vol. 13, No. 4, 2009 / Organic Process Research & Development
•
709