Stereoretentive nitrile hydratase-catalysed hydration of d-phenylglycine nitrile

Article abstract of DOI:10.1021/op000055l

The hydration of D-phenylglycine nitrile to the corresponding amide, mediated by nitrile hydratase-containing microorganisms, was studied. Batch and fed-batch reactions were compared with regard to degradation and racemisation of the chemically labile substrate. A batch process gave satisfactory results and at up to 25 mM D-phenylglycine nitrile (D-1), D-phenylglycine amide was obtained in 94% yield with 92% ee using an immobilised Rhodococcus sp. (NOVO SP 361). The enzyme could be reused, although it slowly lost its activity. When the concentration of D-phenylglycine nitrile was increased to 100 mM in a batch reaction rapid decomposition of the substrate was observed and D-phenylglycine amide was obtained in only 37% yield. A fed-batch reaction afforded an improved yield, although the decomposition of the substrate could not be avoided completely. Lowering the temperature stabilised the substrate, and a fed-batch reaction at 5 °C resulted in a 96% yield of D-phenylglycine amide with 95% ee. A number of other whole-cell hydratase/amidase systems also hydrated D-1 in nearly quantitative yield and >94% ee. Moreover, the ee was further increased to >99% upon prolonged reaction times with minimal loss in yield due to the action of the L-specific amidase that is present in these biocatalysts.

Full text of DOI:10.1021/op000055l

Organic Process Research & Development 2000, 4, 318−322
Stereoretentive Nitrile Hydratase-Catalysed Hydration of D-Phenylglycine Nitrile
M. A. Wegman,† U. Heinemann,‡ A. Stolz,‡ F. van Rantwijk,† and R. A. Sheldon*,†
Laboratory of Organic Chemistry and Catalysis, Delft UniVersity of Technology, Julianalaan 136, 2628 BL Delft,
The Netherlands, and Institut fu¨r Mikrobiologie der UniVersita¨t Stuttgart, Allmandring 31, 70569 Stuttgart, Germany
Scheme 1. Enzyme-catalysed coupling of D-phenylglycine
amide and 6-aminopenicillic acid
Abstract:
The hydration of D-phenylglycine nitrile to the corresponding
amide, mediated by nitrile hydratase-containing microorgan-
isms, was studied. Batch and fed-batch reactions were compared
with regard to degradation and racemisation of the chemically
labile substrate. A batch process gave satisfactory results and
at up to 25 mM D-phenylglycine nitrile (D-1), D-phenylglycine
amide was obtained in 94% yield with 92% ee using an
immobilised Rhodococcus sp. (NOVO SP 361). The enzyme
could be reused, although it slowly lost its activity. When the
concentration of
mM in a batch reaction rapid decomposition of the substrate
was observed and -phenylglycine amide was obtained in only
D-phenylglycine nitrile was increased to 100
D
37% yield. A fed-batch reaction afforded an improved yield,
although the decomposition of the substrate could not be
avoided completely. Lowering the temperature stabilised the
substrate, and a fed-batch reaction at 5 °C resulted in a 96%
of a carbonyl compound.5,6 A 100% yield of enantiopure
D-phenylglycine amide isstheoreticallyspossible, but the
necessary recycle of L-mandelic acid complicates the process.
We consider an alternative route to D-phenylglycine amide
directly from the R-amino nitrile. Enantiopure D-phenyl-
glycine nitrile tartaric acid salt is readily accessible via an
asymmetric transformation with tartaric acid (TA) as resolv-
ing agent based on described procedures.7-9 We investigated
the stereoretentive nitrile hydratase-catalysed hydration of
D-phenylglycine nitrile (D-1) into D-phenylglycine amide (D-
2), (Scheme 3).
Most published work on enzymatic conversions of
R-amino nitriles has been directed at the production of opti-
cally pure amino acids via a nitrilase-catalysed conversion.10-13
Only a few groups reported an R-aminonitrile hydrolysis via
a nitrile hydratase/amidase system, yielding D-amide and
L-acid.14-17
yield of
D-phenylglycine amide with 95% ee. A number of other
whole-cell hydratase/amidase systems also hydrated
D-1 in
nearly quantitative yield and >94% ee. Moreover, the ee was
further increased to >99% upon prolonged reaction times with
minimal loss in yield due to the action of the
L-specific amidase
that is present in these biocatalysts.
Introduction
D-Phenylglycine amide is a key intermediate in the
industrial enzymatic synthesis of semisynthetic â-lactam
antibiotics1 (Scheme 1).
Several routes to D-phenylglycine amide are known, all
starting with a Strecker reaction on benzaldehyde (Scheme
2). A route that involves an aminopeptidase catalysed
L-specific hydrolysis of D,L-phenylglycine amide2-4 has been
commercialised by DSM. In this kinetic resolution the yield
of D-phenylglycine amide is limited to 50%, and the
unwanted isomer, L-phenylglycine (L-PG), has to be race-
mised and recycled, which requires several steps. A more
elegant approach is an asymmetric transformation of the
diastereomeric salt of D,L-phenylglycine amide with L-
mandelic acid (MA) in the presence of a catalytic amount
(5) Boesten, W. H. J. Eur. Patent. Appl. EP 442584, 1991. [Chem. Abstr. 1992,
116: 42062r].
(6) Boesten, W. H. J. Eur. Patent. Appl. EP 442585, 1991. [Chem. Abstr. 1992,
116: 42063s].
(7) Clark, J. C.; Phillipps, G. H.; Steer, M. R. J. Chem. Soc. Perkin. Trans. I
1976, 475.
(8) Hylton, T. A.; Shenton, F. L. U.S. Patent 4072698, 1978. [Chem. Abstr.
1978, 88: 152276x].
(9) Matthews, G. J. French Patent 2141354, 1971. [Chem. Abstr. 1973, 79:
P42210e].
(10) Bhalla, T. C.; Miura, A.; Wakamoto, A.; Ohba, Y.; Furuhashi, K. Appl.
Microbiol. Biotechnol. 1992, 37, 184.
(11) Fukuda, Y.; Fukui, M.; Harada, T.; Izumi, Y. J. Ferment. Technol. 1971,
49, 1011.
(12) Layh, N.; Hirrlinger, B.; Stolz, A.; Knackmuss, H.-J. Appl. Microbiol.
Biotechnol. 1997, 47, 668.
(13) Macadam, A. M.; Knowles, C. J. Biotechnol. Lett. 1985, 7, 865.
(14) Arnaud, A,; Galzy, P.; Jallageas, J.-C. Bull. Soc. Chim. Fr. II 1980, 87.
(15) Bauer, R.; Hirrlinger, B.; Layh, N.; Stolz, A.; Knackmuss, H.-J. Appl.
Microbiol. Biotechnol. 1998, 42, 1.
* Corresponding author. Fax: (+31) 152784700. E-mail: r.a.sheldon@
tnw.tudelft.nl.
† Delft University of Technology.
‡ Institut fu¨r Mikrobiologie der Universita¨t Stuttgart.
(1) Bruggink, A.; Roos, E. S.; de Vroom, E. Org. Process. Res. DeV. 1998, 2,
128.
(2) Elferink, V. H. M.; Breitgoff, D.; Kloosterman, M.; Kamphuis, J.; van den
Tweel, W. J. J.; Meijer, E. M. Recl. TraV. Chim. Pays-Bas. 1991, 110 63
(3) Patel, R. N. StereoselectiVe Biocatalysis; Dekker: New York, 2000; p 23.
(4) Taillades, J.; Garrel, L.; Guillen, F.; Collet, H.; Commeyras, A. Bull. Soc.
Chim. Fr. 1995, 132, 119.
318
•
Vol. 4, No. 5, 2000 / Organic Process Research & Development
10.1021/op000055l CCC: $19.00 © 2000 American Chemical Society and The Royal Society of Chemistry
Published on Web 08/26/2000

Scheme 2. Several routes to D-phenylglycine amide
Scheme 3. Hydration of D-phenylglycine nitrile
The combined organic layers were acidified with 1 N HCl,
and the nitrile hydrochloric acid salt was re-extracted in the
aqueous phase, which was concentrated in vacuo, to give
racemic phenylglycine nitrile hydrochloride as a white solid.
Racemic phenylglycine amide was chemically synthesised
via ammonolysis of the corresponding methyl ester.19
D-Phenylglycine nitrile was used as the free base. There-
fore the pH of a 25 mM solution of D-phenylglycine nitrile
tartaric acid in water was adjusted to pH 5 with 2 N sodium
hydroxide. The aqueous phase was extracted with CH2Cl2,
dried over MgSO4, and concentrated in vacuo to give D-1 as
a yellowish oil, which was used directly.
During the enzymatic reactions there is a potential risk
of producing cyanide via retro-Strecker reaction of D-
phenylglycine nitrile. Care must be taken by performing all
reactions in a fume cupboard provided with an efficient
draught.
Microorganisms, Culture Conditions and Growth
Media. Bacterial cultures were isolated by enrichment from
soil taken from the area of Stuttgart (Germany). Nitrogen-
free media were prepared in a mineral medium containing
per liter: 11.2 g of Na2HPO4, 4 g of KH2PO4, 0.05 g of
CaCl2‚2H2O, 0.03 g of Fe3+-citrate, 0.02 g of MgSO4‚7 H2O,
and 1 mL of a standard trace element solution20 without iron
salts and EDTA. With R. globerulus MAWA fructose (18
g‚L-1) was added as the sole source of carbon and energy,
and phenylacetonitrile (1 g‚L-1) was added as the sole source
of nitrogen. The other strains were cultured with glycerol
(10 g‚L-1) as the sole source of carbon and energy.
Additionally, urea21 (1 g‚L-1) and 1 mL of a vitamin solution
were added.22 In all cases phenylacetonitrile (1 g‚L-1) was
added as the source of nitrogen.
The bioconversion of R-amino nitriles is hampered by
their spontaneous degradation into aldehyde and ammonia
via a retro-Strecker reaction. This reaction takes place readily
under the conditions that are optimal for the enzymatic
reaction, that is, pH 7,13,18 with low yields of amides as a
result. Decomposition of the nitrile could partially be avoided
by performing the reaction at concentrations of 0.5-1.5 mM,
which have low synthetic significance.15-17
We have undertaken to improve the nitrile hydratase-
catalysed conversion of D-1 into D-2 by reaction engineering.
As biocatalyst we selected Novo SP 361, an immobilised
Rhodococcus sp. that has been widely used. Herein we report
two different concepts for the SP 361-catalysed hydration
of D-1 with the aim of achieving high yields of D-phenyl-
glycine amide (D-2) by minimising competing racemisation
and degradation of D-1.
Experimental Section
Chemicals and Catalysts. Immobilised Rhodococcus sp.
SP 361 was a gift from Novo Nordisk A/S (Bagsværd,
Denmark). Enantiomerically pure D-phenylglycine, D-phen-
ylglycine amide, and D-phenylglycine nitrile tartaric acid salt
(Alpha Drug, India LTD) were kindly donated by DSM
(Geleen, The Netherlands). Racemic phenylglycine nitrile
hydrochloride (technical grade) and racemic phenylglycine
were obtained from Acros. The nitrile was purified by
neutralisation of the R-aminonitrile hydrochloric acid salt
with base, followed by an extraction with dichloromethane.
Cells were grown in 300-ml flasks with baffles with 50
mL of medium at 25 °C with shaking (125 rpm). After 24 h
incubation cultures were transferred (1:20 v/v) to fresh
medium with the same composition. After 24-48 h, the cells
(19) Reilen, H.; Kno¨pfe, J. Liebigs Ann. Chem. 1936, 523, 199.
(20) Pfennig, M.; Lippert, K. D. Arch. Mikrobiol. 1966, 55, 245.
(21) Kobayashi, M.; Fujita, T.; Shimizu, S. Appl. Microbiol. Biotechnol. 1996,
45, 176.
(16) Choi, S. Y.; Goo, Y. M. Arch. Pharmacal Res. 1986, 9, 45.
(17) Stolz, A.; Trott, S.; Binder, M.; Bauer, R.; Hirrlinger, B.; Layh, N.;
Knackmuss, H.-J. J. Mol. Catal. B: Enzymatic 1998, 5, 137.
(18) Ospiran, I.; Jarret, C.; Strauss, U.; Kroutil, W.; Orru, R. V. A.; Felfer, U.;
Willetts, A. J.; Faber, K. J. Mol. Catal. B: Enzymatic 1999, 6, 555.
(22) Hunter-Cevera, J. C.; Belt, A. Manual of Industrial Microbiology and
Biotechnology; ASM Press: Washington, 1999; p 3.
Vol. 4, No. 5, 2000 / Organic Process Research & Development
•
319

Table 1. Nitrile hydratase activity of several strains
wet total
H2SO4 was added in 1-5 h tosunless mentioned otherwises
0.5 g of SP 361 in 5 mL of sodium phosphate buffer (10
mM). The reaction mixture was shaken at the selected tem-
perature at pH 7. The pH was kept constant using an auto-
matic titrator (Metrohm 665 Dosimat and a Metrohm 614
Impulsomat). The reactions were monitored by periodically
withdrawing 0.1-mL samples from the reaction mixture. To
stop the reaction in the sample 0.5 mL of 0.1 N HClO4 was
added, followed by HPLC analysis. The experiments were
performed in duplo and were reproducible within the margin
of error. The symbols in Figure 3 indicate the error margin.
specific
culture biomass activity activity
strain
time (h)
(g/L)
(U)
(U/g)
R. globerulus MAWA
R. rhodochrous MAWB
R. rhodochrous MAWE
R. erythropolis MAWF
48
16
16
24
3.0
6.0
7.2
2.7
213
228
403
59
71
38
56
22
were, in the late exponential growth phase, harvested by
centrifugation (Table 1).
Initial Taxonomic Characterisation and Classification
of the Bacterial Strains. For the taxonomic description the
almost complete 16S rDNA was amplified by PCR; the
resulting fragment was cloned and partially sequenced using
the universal primers 27f and 519r.23 The nucleotide se-
quences obtained were compared with the NCBI database
using the program BLASTN. Strains MAWB and MAWE
showed the highest degree of sequence identity (more than
99% sequence identity) with strain R. rhodochrous NCIMB
13064 (this strain is listed in the NCBI database as R.
erythropolis). A BLAST search with the sequence obtained
for the strain MAWF showed the highest degree of similarity
(>99%) with a sequence (X80618) deposited for Nocardia
calcarea. Because N. calcarea is generally considered as a
synonym for R. erythropolis,24 this strain was designated as
R. erythropolis MAWF. The strains are deposited at the
Deutsche Sammlung von Mikroorganismen (DSM) (Braun-
sweig, Germany).
Enzyme Assay and Definition of Units. An appropriate
amount of whole cells was incubated with 2% (v/v) Triton
X-100 for 30 min. After washing with 0.9% NaCl, 2 mL of
54 mM sodium phosphate buffer pH 7.4 and 4 µmol
phenylpropionitrile (400 mM stock solution in ethanol) were
added and shaken at room temperature. Samples were
periodically withdrawn, the cells were removed by centrifu-
gation, and the supernatant was analysed by HPLC. One unit
(U) of nitrile hydratase activity is defined as the amount of
enzyme (g wet biomass) that converts 1 µmol phenylpropio-
nitrile per min.
Batch-Wise Hydration of D-Phenylglycine Nitrile. D-
Phenylglycine nitrile (0.25 mmol, 33 mg) and 0.5 g of SP
361 in 10 mL of sodium phosphate buffer (10 mM) were
shaken in 50-mL reaction vessels at room temperature at a
selected pH, which was kept constant using an automatic
titrator (Metrohm 665 Dosimat and a Metrohm 614 Impul-
somat). The reactions were monitored by periodically
withdrawing 0.1-mL samples from the reaction mixture. To
stop the reaction in the sample 0.5 mL of 0.1 N HClO4 was
added, followed by HPLC analysis. The experiments were
performed in duplo and were reproducible within the margin
of error. The symbols in Figures 1 and 2 indicate the error
margin.
Analysis and Equipment
The reaction mixtures of the stereoretentive hydration
were analysed by chiral HPLC on a Daicel Chemical
Industries Ltd. 4.6 × 150 mm 5 µ Crownpak CR (+) column
using a Waters 625 LC pump and a Waters 486 UV detector.
The eluent was aqueous HClO4, pH 1.0 at a flow of 0.6
mL‚min-1, the column temperature was 18 °C.
The reaction mixtures of the nitrile hydratase activity
assays were analysed by HPLC on a custom-packed Sym-
metry C18 cartridge (Waters Radial-Pak, 8 × 100 mm, 5 µm)
contained in a Waters RCM 8 × 10 compressing unit. A
Waters 590 pump and a Waters 486 UV detector were used.
The eluent was acetonitrile/water 40/60 with 0.1% trifluo-
roacidic acid at a flow of 1.0 mL‚min-1.
Results and Discussion
SP 361-Catalysed Batch Reaction at Low pH. Although
pH 7 is optimal for SP 361, initial experiments were
conducted at pH 5 in an attempt to minimise racemisation
and degradation of D-1. Even at such a low pH D-1 was
hydrated to D-2 by the non-selective nitrile hydratase, and
racemisation as well as decomposition of D-1 were negligible.
The amidase appeared to be slow and highly L-specific, and
no formation of phenylglycine was observed. However, when
the enzyme was reused in a second cycle, its performance
had decreased considerably. This effect was more pro-
nounced when the substrate was used as its tartaric acid salt
rather than as the free base (Table 2). We tentatively
concluded that tartaric acid deactivated SP 361; hence, further
experiments were performed with D-1 as the free base.
Table 2. Hydration of D-phenylglycine nitrile at pH 5a
cycle 1
cycle 2
time yield D-2 ee D-2 yield D-2 ee D-2
(h)
substrate
(%)
(%)
(%)
(%)
tartaric acid salt
free base
3
3
5
65
70
77
95
99
99
11
31
38
86
94
92
a Reaction conditions: 0.25 mmol R-aminonitrile D-1 and 500 mg SP 361
in 10 mL of 10 mM phosphate buffer pH 5 were shaken at room temperature.
Fed-Batch Hydration of D-Phenylglycine Nitrile. D-
Phenylglycine nitrile (1.0 mmol, 132 mg) in 5 mL of 0.2 N
To gain more insight into the deactivation of SP 361, the
effect of the buffer on the enzymatic hydration of D-1 was
briefly investigated. A 10 mM phosphate buffer had a slight
accelerating effect compared with reaction in the absence
(23) Gerhardt, P.; Murray, R. G. E.; Wood, W. A.; Krieg, N. R. Methods for
General and Molecular Bacteriology; A.S.M Press: Washington, 1994.
(24) Rainey, F. A.; Burghardt, J.; Kroppenstedt, R. M.; Klatte, S.; Stackebrandt,
E. Microbiology 1995, 141, 523.
320
•
Vol. 4, No. 5, 2000 / Organic Process Research & Development

Figure 3. Study of inhibition effect on SP 361 in a (2) fed-
batch reaction (feeding in 1 h) of D-1 (100 mM) at pH 7 in the
presence of (9) 50 mM D-phenylglycine amide, ([) 100 mM
sulfate, (b) 0.5 mM benzaldehyde.
Figure 1. SP 361-catalysed hydration of D-1 (25 mM) at pH 5
in various buffers: (2) no buffer, (b) 10 mM phosphate, (9)
100 mM phosphate, ([) 100 mM acetate.
Accordingly, we carried out fed-batch reactions by adding
a 200 mM solution of D-1, resulting in a final concentration
of 100 mM reaction mixture. When the feedstock was added
in 1 h, 50 mM D-2 was obtained (whereas the batch reaction
yielded only 37 mM D-2). However, attempts to increase
the yield by prolonging the addition time from 1 to 5 h failed;
in all cases the reaction stopped at 50 mM D-2.
A turnover effect would be a possible explanation, but
when the amount of catalyst was doubled, only 79 mM D-2
was obtained instead of the expected 100 mM D-2. This
prompted us to investigate inhibition effects on SP 361
(Figure 3). First, product inhibition was examined. When
the reaction was performed in the presence of 50 mM D-2,
no decrease in reaction rate was observed; therefore, we
concluded that SP 361 is not inhibited by the product. The
effect considered next was ionic strength, because in all fed-
batch processes D-1 had been added to the reaction mixture
as its sulphuric acid salt, whereas in the batch reactions at
pH 5 the enzyme activity decreased at high buffer concentra-
tions. Possible inhibition by sulfate was examined by
performing the reaction in the presence of 100 mM Na2-
SO4, but only a small decrease in enzyme activity was
observed. Moreover, when D-1 was added as the free base,
the reaction also stopped at 50 mM D-2. Finally, the possible
inhibition by the retro-Strecker product, benzaldehyde, was
investigated. We found that benzaldehyde caused a drastic
decrease in enzyme activity at concentrations as low as 0.5
mM. Hence, we conclude that the decomposition of D-1 is
the main cause of deactivation of the nitrile hydratase, besides
being detrimental to the yield.
SP 361-Catalysed Reactions at Low Temperature. On
the basis of the above findings it is clear that to increase the
yield of D-2, the stability of D-1 towards benzaldehyde
formation via a retro-Strecker reaction must be increased.
We found that decompositionsand racemisationsof D-1 is
less at low temperatures. After 24 h at 5 °C and pH 7
spontaneous degradation of D-1 (100 mM) to benzaldehyde
was 13%, and the ee of the remaining D-1 was 82% compared
to 75% and complete racemisation at room temperature.
At 5 °C a 100 mM batch reaction yielded 63 mM D-2,
compared with 37 mM at room temperature. This prompted
us to carry out a fed-batch reaction at low temperature. The
feedstock (200 mM D-1 in 0.2 N H2SO4) was added over 3
h. The reaction was followed in time, and after 7 h we
obtained 96% yield of D-2 (102 mM) with 95% ee. Hence,
Figure 2. SP 361-catalysed batch-wise hydration of D-1 (25
mM) at pH 7 (9) cycle 1, (2) cycle 2, (b) cycle 3.
of buffer, presumably because of the unavoidable pH
excursions in the latter case. Phosphate (100 mM) caused a
decrease in rate as well as in yield; acetate (100 mM) had
an even more pronounced effect (Figure 1).
In an attempt to reactivate the catalyst FeCl3 (1 mM) as
well as CoCl2 (1 mM) were added to the reaction mixture,
since it is known that some nitrile hydratases contain iron
or cobalt as cofactor,25,26 but no increase in enzyme activity
was observed. We conclude that D-1 can be hydrated to D-2
at pH 5 without racemisation or decompostion, but recycling
of the catalyst is not possible. In view of these results all
further experiments were carried out at pH 7.
SP 361-Catalysed Batch Reaction at pH 7. At pH 7
D-1 spontaneously racemises and decomposes, but this can
be minimised by keeping its concentration low. Surprisingly
we found that up to 25 mM D-1 was converted to D-2 in
high yield with negligible decomposition or racemisation.
Again no formation of acid was observed. The enzyme could
be recycled several times, although its activity slowly
declined (Figure 2).
When the substrate concentration was increased to 100
mM, the above procedure yielded only 37 mM D-2. Race-
misation of D-1 was negligible (5.5%), but its decomposition
into benzaldehyde and ammonia amounted to 43%. In
conclusion, this simple methodology is not feasible at
practically relevant concentrations.
SP 361-Catalysed Fed-Batch Reactions at pH 7. A fed-
batch procedure would present an option to combine a low
actual concentration of D-1sto retard its decompositions
with a high over-all throughput.
(25) Sugiura, Y.; Kuwahara, J.; Nagasawa, T.; Yamada, H. J. Am. Chem. Soc.
1987, 109, 5848.
(26) Nagasawa, T.; Takeuchi, K.; Yamada, H. Eur. J. Biochem. 1991, 196, 581.
Vol. 4, No. 5, 2000 / Organic Process Research & Development
•
321

Table 3. Fed-batch hydration of D-phenylglycine nitrile by
running the reaction for a longer time (Table 3, entry 5).
The amidase present in the microorganism converted the
unwanted L-amide slowly into the corresponding L-acid,
resulting in 95% yield of D-2 with >99% ee.
different strainsa
time
(h)
yield
(%)
ee
entry
enzyme (U)
SP 361 (-)
R. globerules MAWA (69)
R. erythropolis MAWF (44)
R. rhodochrous MAWE (56)
R. rhodochrous MAWB (38)
(%)
1
2
3
4
5
7
4
7
7
6
96.0
98.7
95.6
98.6
95.4
95.1
95.0
94.6
97.4
96.8
98.3
99.5
Conclusions
Notwithstanding the instability of D-phenylglycine nitrile
(D-1), its hydration to the corresponding amide (D-2) is
feasible in high yield. Up to 25 mM D-1, a batch reaction at
room temperature gives D-2 in 94% yield and 92% ee. At
higher substrate concentrations a fed-batch mode of operation
at lower temperatures is necessary to achieve high yields
and enantiopurities; for example, at 5 °C and a final
concentration of 100 mM, D-2 was obtained in 95% yield
and 94% ee. This concept appeared to be a generally
applicable method, as other nitrile hydratase/amidase systems
gave similar results. Moreover, the L-amidase yielded an
increased ee up to 99.5% with minimal loss in yield.
22
a Reaction conditions: 1 mmol R-amino nitrile D-1 in 5 mL of 0.2 N H2SO4
was added in 3 h to the microorganism in 5 mL of sodium phosphate buffer (10
mM). The reaction mixture was shaken at 5 °C.
by performing the reaction in a fed-batch mode at low
temperatures decomposition is avoided, and D-2 can be
obtained in nearly quantitative yield.
Other Nitrile Hydratase/Amidase Systems. Because of
uncertainty regarding the commercial availability of SP 361
other nitrile hydratase/amidase systems were also tested in
our set-up in order to demonstrate the general applicability
of our fed-batch procedure at low temperature for the
conversion of chemically labile D-1. We selected several
microorganisms which hydrated D,L-phenylglycine nitrile 1
efficiently.27 The bacterial cultures were grown on defined
media; the whole cells were permeabilised by incubation with
Triton X-100 and were used without further treatment in the
fed-batch hydration of D-1.
Acknowledgment
Generous donations of nitrile hydratase/amidase SP 361
by Novo Nordisk A/S (Bagsværd, Denmark) are gratefully
acknowledged. DSM Life-Sciences is acknowledged for a
gift of D-phenylglycine nitrile tartaric acid salt. We thank
Dr. P. Porskamp of DSM Deretil (Almeria, Spain) for his
encouragement and Professor Dr. L. Fischer (University of
Stuttgart-Hohenheim, Germany) for his help with cultivating
the several nitrile hydratase/amidase systems. This work was
financially supported by The Netherlands Ministry of
Economic Affairs and coordinated by DSM Life Sciences
(Geleen, The Netherlands).
In all cases D-1 was hydrated to D-2 with high yield and
ee (Table 3). Moreover, the enantiomeric purity of D-2
could be increasedswith a minimum loss in yieldsby
Received for review May 19, 2000.
OP000055L
(27) Wegman, M. A.; Heinemann, U.; van Rantwijk, F.; Stolz, A.; Sheldon, R.
A. J. Mol. Catal. B: Enzymatic 2000, in press.
322
•
Vol. 4, No. 5, 2000 / Organic Process Research & Development