High yield synthesis of d-phenylglycine and its derivatives by nitrilase mediated dynamic kinetic resolution in aqueous-1-octanol biphasic system

Article abstract of DOI:10.1016/j.tetlet.2014.01.044

A strategy of nitrilase mediated dynamic kinetic resolution toward the synthesis of d-phenylglycine was developed, using aqueous-1-octanol biphasic system. Due to the efficient suppression of the decomposition of phenylglycinonitrile, a maximum yield of 81% is obtained. This result indicates that the nitrilase mediated dynamic kinetic resolution is a promising method toward the synthesis of d-phenylglycine and its derivatives.

Full text of DOI:10.1016/j.tetlet.2014.01.044

Tetrahedron Letters 55 (2014) 1448–1451
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
High yield synthesis of D-phenylglycine and its derivatives
by nitrilase mediated dynamic kinetic resolution in
aqueous-1-octanol biphasic system
a
a,
Jian Qiu a, Erzheng Su a,b, , Wei Wang , Dongzhi Wei
⇑
⇑
a State Key Laboratory of Bioreactor Engineering, New World Institute of Biotechnology, East China University of Science and Technology, Shanghai 200237, PR China
b Enzyme and Fermentation Technology Laboratory, College of Light Industry Science and Engineering, Nanjing Forestry University, Nanjing 210037, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
A strategy of nitrilase mediated dynamic kinetic resolution toward the synthesis of D-phenylglycine was
Received 29 October 2013
Revised 9 January 2014
Accepted 13 January 2014
Available online 18 January 2014
developed, using aqueous-1-octanol biphasic system. Due to the efficient suppression of the decomposi-
tion of phenylglycinonitrile, a maximum yield of 81% is obtained. This result indicates that the nitrilase
mediated dynamic kinetic resolution is a promising method toward the synthesis of
its derivatives.
D-phenylglycine and
Ó 2014 Elsevier Ltd. All rights reserved.
Keywords:
Nitrilase
Dynamic kinetic resolution
Biphasic system
D-Phenylglycine
D
-Phenylglycine is an important chiral building block in the
synthesis of antibiotics, such as ampicillin and cefalexin.1 The con-
ventional process for making enantio-pure -phenylglycine is the
crystallization of its diastereomeric salts racemate and (+)-cam-
phor-8-sulfonic acid.2 The other enantiomer,
-phenylglycine is
phenylglycinonitrile at pH 10.6. This process is attractive with
two features: (i) The method uses
a-aminonitriles as raw materi-
D
als, which are easier to be obtained than
a-amino acid derivatives
used in other chemo-enzymatic routes; (ii) unlike kinetic resolu-
tion routes mentioned above, recovery and racemization of the
undesired enantiomer in separated steps are not required. This
‘nitrilase process’ affords a 100% theoretical yield through continu-
ous in situ racemization under alkaline condition. Although it is a
great potential approach, there are few reports on amino acid pro-
duction involving nitrilases.12,13 Presumably, this is due to the labi-
lization of a-aminonitriles and the lack of high enantioselective
nitrilase. In our previous work (data unpublished), a novel nitrilase
SWRW1 mined from Sphingomonas wittichii RW1 was used for high
L
racemized in a separate step and then recycled. In recent years,
many chemo-enzymatic methods have been developed for prepar-
ing
penicillin G acylase,6 and
hydantoinase/carbamoylase process has been commercially used
for the preparation of
-p-hydroxyphenylgly-cine.8 However,
-phenylglycine has not yet been produced through this process
on an industrial scale.9
D
-amino acid,3 such as the use of aminoacylase,4 amidase,5
D
-aminotransferase.7 The well-known
D
D
Nitrilase (EC 3.5.5.1) is a kind of hydrolase found in many bac-
teria, fungi, and plants, which can convert nitriles into the corre-
sponding carboxylic acids and ammonia. To date, nitrilases have
already been shown to be versatile biocatalysts for the production
of chiral pharmaceutical intermediates as well as bulk products,
such as (R)-mandelic acid, (R)-4-cyano-3-hydroxybutyric acid,
and nicotinic acid.10 Chaplin et al.11 reported two novel approaches
of nitrilase-catalyzed dynamic kinetic asymmetric synthesis of
aromatic amino acids. One approach was the enantioselective
enantioselective synthesis of D-phenylglycine by kinetic resolution
of phenylglycinonitrile and 46% yield with 95% ee was obtained at
pH 6.0. Herein, we reported improved alkaline conditions, under
which the nitrilase mediated enantioselective hydrolysis is com-
bined with in situ racemization in aqueous and biphasic aque-
ous–organic system.
It has been previously demonstrated that pH values higher than
10 were required for the racemization of phenylglycinonitrile, and
t1/2 for the racemization at pH 10.5 was estimated to be 60 min
compared to 4 h at pH 9.5.11 To evaluate the alkaline tolerance of
the nitrilase SWRW1, resting cells of a recombinant Escherichia coli
BL21 expressing nitrilase activity were incubated with phenylgly-
cinonitrile at various pH values. The maximum activity of whole
cell biocatalysts was observed at pH 8.0, which was similar to
synthesis of D-phenylglycine via dynamitic kinetic resolution of
⇑
Corresponding authors. Tel./fax: +86 25 85428906 (E.S.); tel.: +86 21 64252078;
fax: +86 21 64250068 (D.W.).
E-mail addresses: ezhsu@njfu.edu.cn (E. Su), dzhwei@ecust.edu.cn (D. Wei).
0040-4039/$ - see front matter Ó 2014 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2014.01.044

J. Qiu et al. / Tetrahedron Letters 55 (2014) 1448–1451
1449
the result observed with the purified nitrilase. About 50% residual
activities were maintained at pH 6.0 and pH 10.0, respectively
(Fig. 1), indicating that whole cell biocatalysts displayed more
alkaline stability than the purified nitrilase.
The whole cell biocatalysts were then used in the dynamic
kinetic resolution reaction optimized, and the results are
summarized in Table 1. This dynamic kinetic resolution route is
complicated, which mainly involves three reactions (Scheme 1):
spontaneous decomposition (reaction 1), enantioselective hydroly-
sis mediated by nitrilase (reaction 2), and racemization under alka-
line conditions (reaction 3). Therefore, the yields and ee values are
dependent on the competition of three reaction rates under differ-
ent conditions.
In our earlier work, kinetic resolution was investigated at pH
6.0. Under this acidic condition, phenylglycinonitrile was stable
without any racemization (Table 1, entry 1). In order to accelerate
the racemization, different alkaline conditions (pH 9.5–11) were
investigated and the ee values were higher than 92% in most cases
(Table 1, entries 3–6), except 80% ee at pH 9.5 (Table 1, entry 2).
Presumably, this is due to the inefficient racemization, which
might lead to the low ee value. A maximum yield of 68% was ob-
served at pH 10.5 (Table 1, entry 4), and the yields decreased along
with the pH increase (Table 1, entries 5 and 6). The decreasing
yield might be primarily caused by the more rapid decomposition.
Other possibilities could be the high concentration of benzalde-
hyde acting as an enzyme inhibitor,14 or the deactivation of nitri-
lase under severe alkaline conditions. The reaction temperature
is a key parameter in bioprocess, which can significantly influence
the activity, enantioselectivity, and stability of a biocatalyst as well
as the equilibrium of a reaction.15 Although lower temperature
(10 °C) was useful in enhancing the enantioselectivity, lower solu-
bility, and reaction rate also decreased the final yield (Table 1, en-
try 7). The optimal temperature for this reaction was 20 °C
(Table 1, entry 8).
Substrate concentration is another very important issue associ-
ated with nitriles biotransformation. In general, nitrile compounds
120
100
80
60
40
20
0
Scheme 1. Production of D-phenylglycine by dynamic kinetic resolution.
160
140
120
100
80
4
5
6
7
8
9
10
11
pH
Figure 1. pH stability of the whole cell biocatalysts and the purified nitrilase. The
relative activities were expressed as a percentage of maximum activities of whole
cell biocatalysts (black line) and purified nitrilase (red line), respectively. Reactions
were performed in triplicate, and error bars represented the standard error of the
mean. Sodium citrate–citric acid buffer (j), sodium phosphate buffer (ꢀ), Tris–HCl
buffer (N), and carbonate buffer (d).
60
40
20
0
Table 1
octane
control
toluene
hexane
benzene
1-butanol
1-octanol
Reaction conditions optimized in aqueous system
cyclohexane
ethyl acetate
butyl acetate
dichloromethane
Organic solvents
Entry Substrate
concentration (mM)
pH
Temperature Time Yield ee (%)
(°C)
(h)
(%)
Figure 2. Effect of various organic solvents on the activity retention of the whole
1a
2a
3a
4a
5b
6b
7b
8b
25c
6.0
9.5
20
20
5
6
6
6
10
10
8
6
6
6
8
11.5
14
46e
51
55
68
60
57
60
68
59
68
65
37
52
95
80
92
95
95
95
97
95
95
95
95
N.D.
N.D.
cells.
25c
25c
25c
25c
25c
25c
25c
25c
25c
50c
50d
75c
10.0 20
10.5 20
11.0 20
11.5 20
10.5 10
10.5 20
10.5 30
10.5 20
10.5 20
10.5 20
10.5 20
Table 2
Reaction conditions optimized in water–1-octanol biphasic system
Entry Substrate
Water
Temperature Time Yield ee (%)
concentration (mM) content (%) (°C)
(h)
(%)
9b
a
1a
2a
3a
4a
5a
6a
7a
8a
9a
50
50
50
50
20
10
2
10
10
10
10
10
20
20
20
20
20
20
20
25
30
14
20
24
48
48
48
48
48
48
70
73
80
57
78
69
78
81
73
95
95
95
97
95
N.D.
95
95
N.D.
10
11b
12b
13b
50
100
150
100
100
100
N.D.: not determined, due to low yield.
a
300 mg dry cells were used as the whole cell biocatalysts.
600 mg dry cells were used as the whole cell biocatalysts.
10% methanol as the cosolvent.
20% methanol as the cosolvent.
The reaction was terminated at the conversion near to 50% because of the
b
c
d
e
N.D.: not determined, due to low yield.
a
kinetic route of resolution (Fig. S1).
600 mg dry cells were used as the whole cell biocatalysts.

1450
J. Qiu et al. / Tetrahedron Letters 55 (2014) 1448–1451
Table 3
High yield synthesis of D-phenylglycine derivatives in aqueous-1-octanol biphasic system
Entry
Reaction system
Aqueous system
Substrate concentration (mM)
pH
Temperature (°C)
Time (h)
Yield (%)
ee (%)
1a
2a
3a
4a
25b
25c
25b
25c
6.0
6.0
10.5
10.5
20
20
25
25
20
16
48
48
42d
49d
59
91
93
90
93
Aqueous-1-octanol
Biphasic system
70
a
b
c
600 mg dry cells were used as the whole cell biocatalysts.
2-Chloro-phenylglycinonitrile was used as substrate.
4-Chloro-phenylglycinonitrile was used as substrate.
The reaction was terminated at the conversion near to 50%.
d
have poor solubility in aqueous media, and the polar low-molecu-
lar weight solvents, such as methanol and DMSO, are used as cosol-
vents.16 However, high concentration of nitriles and cosolvents are
both detrimental to nitrilases. To access better efficacy in this ap-
proach, substrate concentrations were thus optimized. Conducting
the reaction in 50 mM of phenylglycinonitrile (65%, Table 1, entry
11) was as effective as conducting the reaction in 25 mM of sub-
strate (Table 1, entry 10). However, further increasing the substrate
concentration made solution turbid immediately, indicating that
10% methanol could not completely dissolve the substrate. 20%
methanol was then used in the further test. Unfortunately, this le-
vel of methanol inhibited or inactivated the enzyme dramatically
and low yield (37%) was obtained within 11.5 h (Table 1, entry
12). When the substrate concentration was increased up to
75 mM, the yield (52%) was not improved further (Table 1, entry
13). This result might be attributed to enzyme inactivated by the
substrate or benzaldehyde.
decreased the yield (Table 2, entry 6), which might be due to the
fact that high concentration of the substrate assembling at the bi-
phasic interface areas deactivated the enzyme. The effect of reac-
tion temperature on yield was also investigated (Table 2, entries
7–9) and showed that 25 °C was the optimal temperature (Table 2,
entry 8). It was found that the selectivity of the nitrilase was inde-
pendent of all the factors mentioned above, and the enantiomeric
excess value was greater than 95% in all cases.
Having established the optimal conditions for the hydrolysis of
phenylglycinonitrile, other derivatives, such as 2-chloro- and
4-chloro-phenylglycinonitrile were also investigated in the aque-
ous system and aqueous-1-octanol biphasic system. As shown in
Table 3, in both cases, the activity and enantioselectivity decreased
(Table 3, entries 1 and 2), which might be due to the electronic and
steric effects. Compared to the aqueous system, this biphasic sys-
tem also improved the yield effectively (42% vs 59%; 49% vs 70%).
In conclusion, a new nitrilase mediated dynamic kinetic resolu-
In order to suppress the decomposition of the substrate and im-
prove the yield, aqueous–organic reaction system was investi-
gated. Ten water-immiscible organic solvents, such as ethyl
acetate (0.68), 1-butanol (0.8), dichloromethane (0.93), butyl ace-
tate (1.7), benzene (2.0), toluene (2.5), 1-octanol (2.9), cyclohexane
(3.2), hexane (3.5), and octane (4.5) were examined for the effect
on the activity retention of the whole cells and the results are pre-
sented in Figure 2. The activity retentions of more than 100% were
obtained in the solvents with the logP value ranging from 2.0 to
4.5, except in toluene. It may be due to the more rigid and stable
conformational structure of the enzyme in this solvents.17,18
The maximum activity retention was observed in 1-octanol,
which was 1.5 times greater than that in neat aqueous buffer.
Although with similar logP value, the activity retention in toluene
was much lower, which was opposite to the result observed by
Zhang et al.19 This result indicated that the structure of enzyme
is an important factor for its performance in different organic sol-
vents and toluene was harmful to SWRW1 in this work. Accounting
for the solubility of the substrate and the product, 1-octanol was
chosen as the best organic solvent for further experiments.
To evaluate the performance of the whole cells in aqueous-1-
octanol biphasic system, many important factors such as water
content, substrate concentration, and reaction temperature were
tested as summarized in Table 2. Water content is a vital factor
for suppressing the decomposition of the substrate. Conducting
the reaction in 50% water content failed to suppress the decompo-
sition reaction, and only 70% yield was obtained, which was similar
to that in the aqueous system (Table 1, entries 1 and 2). Decreasing
the water content from 50% to 10% significantly enhanced the yield
to 80% (Table 2, entry 3). However, further decreasing the water
content to 2% lowered the yield to 57%, even when the reaction
time was extended to 48 h (Table 2, entry 4). This might be due
to the diffusion limitation. Thus, further investigations were car-
ried out in water–1-octanol (1:9, v/v) biphasic system. Substrate
concentration also plays an important role in the reaction outcome.
Similar yield was observed when the concentration was double to
100 mM, (Table 2, entry 5). Further increasing the concentration
tion toward the synthesis of D-phenylglycine and its derivatives in
aqueous-1-octanol biphasic system was developed. Several impor-
tant parameters were evaluated in detail, and showed that 10%
water content was essential for the suppression of the decomposi-
tion of substrate. The 81% maximum yield was obtained under the
optimized conditions, which is 1.2 times greater than that in the
aqueous system. These results indicated that the nitrilase mediated
dynamic kinetic resolution could be used as a promising approach
toward D-phenylglycine and its derivatives.
Acknowledgments
This work was supported by the National Basic Research Pro-
gram of China (No. 2012CB721003), the Fundamental Research
Funds for the Central Universities and the National Major Science
and Technology Projects of China (No.2012ZX09304009).
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.tetlet.2014.01.
044.
References and notes
1. Ma, J. S. Chim. Oggi 2003, 21, 65–68.
2. Wegman, M. A.; Janssen, M. H. A.; Rantwijk, F.; Sheldon, R. A. Adv. Synth. Catal.
2001, 343, 559–576.
3. Leuchtenberger, W.; Huthmacher, K.; Drauz, K. Appl. Microbiol. Biotechnol. 2005,
69, 1–8.
4. Wakayama, M.; Yoshimune, K.; Hirose, Y.; Moriguchi, M. J. Mol. Catal. B Enzym.
2003, 23, 71–85.
5. Yamaguchi, S.; Komeda, H.; Asano, Y. Appl. Environ. Microbiol. 2007, 37, 5370–
5373.
6. Gong, X. Y.; Su, E. Z.; Wang, P. X.; Wei, D. Z. Tetrahedron Lett. 2011, 52, 5398–
5402.
7. Kobayashi, J.; Shimizu, Y.; Mutaguchi, Y.; Doi, K.; Ohshima, T. J. Mol. Catal. B:
Enzym. 2013, 94, 15–22.
8. May, O.; Verseck, S.; Bommarius, A.; Drauz, K. Org. Process Res. Dev. 2002, 6,
452–457.

J. Qiu et al. / Tetrahedron Letters 55 (2014) 1448–1451
1451
9. Synthesis of b-Lactam Antibiotics: Chemistry, Biocatalysis & Process Integration;
Bruggink, A., Ed.; Kluwer Academic Publishers: Dordrecht, 2001.
10. Gong, J. S.; Lu, Z. M.; Li, H.; Shi, J. S.; Zhou, Z. M.; Xu, Z. H. Microb. Cell Fact. 2012,
11, 142–159.
11. Chaplin, J. A.; Levin, M. D.; Morgan, B.; Farid, N.; Li, J.; Zhu, Z. L.; McQuaid, J.;
Nicholson, L. W.; Rand, C. A.; Burk, M. J. Tetrahedron: Asymmetry 2004, 15,
2793–2796.
12. Choi, S. Y.; Goo, Y. M. Arch. Pharm. Res. 1986, 9, 45–47.
13. Bhalla, T. C.; Miura, A.; Wakamoto, A.; Ohba, Y.; Furuhashi, K. Appl. Microbiol.
Biotechnol. 1992, 37, 184–190.
14. Zhang Z. J. Study on nitrilase catalyzed preparation of enantiopure mandelic acid
and rational design of hydroxynitrile lyase. Ph.D. Thesis, East China University of
Science and Technology, 2013, p 42.
15. Zhang, Z. J.; Xu, J. H.; He, Y. C.; Ouyang, L. M.; Liu, Y. Y.; Imanaka, T. Process
Biochem. 2010, 45, 887–891.
16. Laane, C.; Boeren, S.; Vos, K.; Veeger, C. Biotechnol. Bioeng. 1987, 30, 81–87.
17. Zaks, A.; Klibanov, A. M. Proc. Natl. Acad. Sci. 1985, 82, 192–196.
18. Ohya, Y.; Miyaoka, J.; Ouchi, T. Macromol. Rapid Commun. 1996, 17, 871–874.
19. Zhang, Z. J.; Pan, J.; Liu, J. F.; Xu, J. H.; He, Y. C.; Liu, Y. Y. J. Biotechnol. 2011, 152,
24–29.