Efficient enzymatic synthesis of ampicillin using mutant Penicillin G acylase with bio-based solvent glycerol

Article abstract of DOI:10.1016/j.catcom.2016.02.014

To fulfill the industry demand of ampicillin enzymatic synthesis, immobilized mutant Penicillin G acylase and bio-based solvent glycerol were employed at high substrate concentration and low acyl donor/nucleophile ratio. After process optimization, good yield and low operation costs were achieved.

Full text of DOI:10.1016/j.catcom.2016.02.014

Catalysis Communications 79 (2016) 31–34
Contents lists available at ScienceDirect
Catalysis Communications
journal homepage: www.elsevier.com/locate/catcom
Short communication
Efficient enzymatic synthesis of ampicillin using mutant Penicillin G
acylase with bio-based solvent glycerol
c
Senwen Deng a,c,1, Xiaoqiang Ma c,1, Ming Sun c, , Dongzhi Wei , Erzheng Su
⁎
b,d,⁎⁎
a
School of Life Science, Hunan University of Science and Technology, Xiangtan 411201, Hunan, PR China
b
Enzyme and Fermentation Technology Laboratory, College of Light Industry Science and Engineering, Nanjing Forestry University, Nanjing 210037, PR China
State Key Laboratory of Bioreactor Engineering, New World Institute of Biotechnology, East China University of Science and Technology, Shanghai 200237, PR China
State Key Laboratory of Pulp and Paper Engineering, South China University of Technology, Guangzhou 510640, PR China
c
d
a r t i c l e i n f o
a b s t r a c t
Article history:
To fulfill the industry demand of ampicillin enzymatic synthesis, immobilized mutant Penicillin G acylase and
bio-based solvent glycerol were employed at high substrate concentration and low acyl donor/nucleophile
ratio. After process optimization, good yield and low operation costs were achieved.
© 2016 Elsevier B.V. All rights reserved.
Received 18 December 2015
Received in revised form 5 February 2016
Accepted 15 February 2016
Available online 17 February 2016
Keywords:
Penicillin G acylase
Antibiotic
Biocatalysis
Bio-based solvent
1. Introduction
Protein engineering is commonly used to tailor wild PGAs for SSBAs
synthesis [6,11–13]. For example, βF24G mutant of Penicillin G acylase
Semi-synthetic β-lactam antibiotics (SSBAs) are the most important
family of β-lactam antibiotics in the world market [1,2]. They were pro-
duced by the condensation of β-lactam moiety with the acyl side chain
[3]. Chemical synthesis of SSBAs has dominated the industrial produc-
tion of SSBAs for high yield (90%) [4]. But the complicated process
steps, harsh reaction conditions and large volume of organic solvent
needed, make them environmentally unsustainable [3,5]. Enzymatic
synthesis of SSBAs is an environmental friendly alternative, which is
mainly catalyzed by Penicillin G acylase (PGA: EC 3.5.1.11) [4], and
can be carried out under kinetically control (Fig. 1). But its relatively
low yield (40–60%), high process costs (raw material and catalyst),
complicated downstream processing (isolation and recycle) needed,
make it hard to fulfill the industry demand at present [3,5].
from Alcaligenes faecalis (Af PGA), in which the 24th Phenylalanine of
the β-subunit was replaced by Glycine was isolated for high S/H ratio
in this lab [13]. Also, this process can be improved by the addition of
water-miscible organic solvents, excess acyl donor and high substrate
concentration [6,14,15]. But these measures have to be taken carefully
in consideration of nucleophile conversion, raw material and down-
stream purification process costs. For example, the molar ratio acyl
donor/nucleophile should be as low as possible to reduce the costs of
raw materials and downstream processing. In the case of ampicillin,
amoxicillin and cephalexin, synthesis at high substrate concentration
has been proven to be beneficial [16–18]. However, this method could
be limited by substrate solubility, which could influence the conversion
of the nucleophile significantly. Medium engineering could improve
synthesis by increasing the S/H ratio. This has been proven for the
synthesis of cephaloglycine, ampicillin and cephalexin in methanol or
ethylene glycol [19–21]. But toxicity and biohazard of these solvents
have to be considered seriously. Ionic liquids were regarded as the sub-
stitution of organic solvents: e.g. Pereira et al. and Zhu et al. reported
enzymatic synthesis at the presence of ionic liquids [22,23]. But the
use of ionic liquids is still limited by high prices and lack of data about
the toxicity and bio-compatibility. As a bio-based solvent, glycerol
may combine the advantages of water and ionic liquids (low toxicity,
low price, large availability, renewability, high boiling point, low vapor
pressure), allows its use in the synthesis of pharmaceutically active in-
gredients [24,25].
Low yield of kinetically controlled synthesis can be mainly ascribed
to enzyme-catalyzed initial hydrolysis of the activated acyl donor and
the secondary hydrolysis of antibiotic product (Fig. 1) [6]. For natural
PGAs, low S/H (synthesis to hydrolysis) ratio is often observed [7–10].
⁎
Corresponding author.
⁎⁎ Correspondence to: E. Su, Enzyme and Fermentation Technology Laboratory, College
of Light Industry Science and Engineering, Nanjing Forestry University, Nanjing 210037,
PR China.
E-mail addresses: sunmingecust@163.com (M. Sun), ezhsu@njfu.edu.cn (E. Su).
These authors contributed equally to this work.
1
http://dx.doi.org/10.1016/j.catcom.2016.02.014
1566-7367/© 2016 Elsevier B.V. All rights reserved.

32
S. Deng et al. / Catalysis Communications 79 (2016) 31–34
Fig. 1. Scheme of enzymatic synthesis of ampicillin.
In this work, we tried to establish a process for ampicillin synthesis
method [29]. HPLC analysis was performed by a previously reported
method [13].
from an industrial view, through the use of immobilized mutant Penicil-
lin G acylase and bio-based solvent glycerol at the condition of high sub-
strate concentration and low acyl donor/nucleophile ratio.
3. Results and discussion
2. Experimental
Approximately 73% of the total protein was covalently linked to the
support (Table 1). The activity recovery was 45%. The activity of the
immobilized βF24G mutant acylase was 37 U/g (Table 1). The S/H
ratio of immobilized acylases was 4.0, slightly lower than 4.26 of the
free enzymes under the same condition (0.1 M 6-APA) [13], which can
be mainly ascribed to the mass transfer problem of the immobilized
enzymes [15].
2.1. General
All reagents were purchased from commercial sources. Plasmid pET-
28a-Af PGA-24G (βF24G mutant) was reserved in this lab [13]. The
transformation, cultivation, harvest and disruption of recombinant
cells followed the methods reported previously [13,26]. The obtained
supernatant was collected and fractionated with ammonium sulfate.
Protein that precipitated between 20 and 30% (w/v) ammonium sulfate
was redissolved by potassium phosphate buffer (100 mM, pH 8.0) as
partly purified enzyme for immobilization. The partly purified acylases
were immobilized onto an epoxy carrier LH-EP under the optimized
conditions as described before [27].
At first, the reaction was performed without the addition of glycerol
(Fig. 2A). The observed increase of initial S/H ratio (from 4.0 to 11.5),
could be mainly ascribed to the rise of 6-APA concentration (from 0.1
to 0.4 M), in accordance with the previous reports [16–18]. However,
as the process continued, we unexpectedly found that reaction solution
was concreting as byproduct D-phenylglycine (D-PG) with a poor solu-
bility accumulated and precipitated [3]. Increased kcat/Km value of
βF24G mutant Af PGA for ampicillin and high concentration of produced
ampicillin [3,13] might lead to more second hydrolysis and byproduct
D-PG accumulation. Gradient concentration of glycerol (5–20%, v/v)
was added at the beginning of the reaction to reduce the side reaction
of hydrolysis. The conversion of 6-APA increased from 66 to 87% and
the concretion time of reaction system were delayed as glycerol concen-
tration increased. This could be ascribed to the increased initial S/H ratio
(from 11.5 to 17.3) and decreased second hydrolysis as results of
reduced water activity. But the reaction speed was also significantly re-
duced by the glycerol addition, which might be the results of changed
enzyme structure [24,25]. When high concentration of glycerol was
employed (20%, v/v), dramatically reduced reaction speed was unfavor-
able for 6-APA conversion. In consideration of 6-APA conversion, 15%
(v/v) glycerol is suitable.
2.2. General procedure for enzymatic synthesis of ampicillin
The 100 mL reaction mixture contained 400 mM 6-amino-
penicillanic acid (6-APA), appropriate amount of D-phenylglycine
methyl ester (D-PGME) and immobilized mutant Af PGA. The reaction
was conducted without pH control. The conversion of 6-APA was calcu-
lated by measuring the decrease of the 6-APA by HPLC.
2.3. Analysis
The enzyme activity was determined by a previously reported
method [28]. Specific activity is defined as UPGK per milligrams of
protein. Protein concentration was determined using the Bradford
Table 1
Protein expression and immobilization.
Immobilization yield(%)a
73
Activity recovery(%)b
45
Activity (UPGK/g)c
37
S/H ratioini
d
Offered enzyme (mg/g)
34.9
Specific activity (UPGK/mg)
2.37
4.0
All data were the average of three independent experiments.
a
The immobilization yield (%, w/w) was calculated as described before [27].
The activity recovery was calculated as described before [27].
Activity toward PGK hydrolysis expressed per gram of support (UPGK/g).
S/H ratioini: Initial rate of antibiotic formation to initial rate of D-PG formation.
b
c
d

S. Deng et al. / Catalysis Communications 79 (2016) 31–34
33
Fig. 2. Optimization of ampicillin enzymatic synthesis. (A) Time course of enzymatic synthesis of ampicillin at different concentrations of glycerol. Conditions: initial pH 5.8, 20 °C, 0.4 M 6-
APA, 0.42 M D-PGME, 0.02 g/mL immobilized βF24G mutant Af PGA. (B) Time course of enzymatic synthesis of ampicillin at different concentrations of immobilized βF24G mutant Af PGA.
Conditions: 15% (v/v) glycerol, initial pH 5.8, 20 °C, 0.4 M 6-APA, 0.42 M D-PGME. (C) Time course of enzymatic synthesis of ampicillin at different initial medium pH. Conditions: 15% (v/v)
glycerol, 20 °C, 0.4 M 6-APA, 0.42 M D-PGME, 0.04 g/mL immobilized βF24G mutant Af PGA. (D) Time course of enzymatic synthesis of ampicillin at different temperatures. Conditions: 15%
(v/v) glycerol, initial pH 5.8, 0.4 M 6-APA, 0.42 M D-PGME, 0.04 g/mL immobilized βF24G mutant Af PGA. (E) Time course of enzymatic synthesis of ampicillin at different substrate ratios.
Conditions: 15% (v/v) glycerol, initial pH 5.8, 20 °C, 0.4 M 6-APA, 0.04 g/mL immobilized βF24G mutant Af PGA. All data were the average of three independent experiments.
As shown in Fig. 2B, the ampicillin synthesis was significantly influ-
enced by the concentration of immobilized βF24G mutant Af PGA. In-
creased enzyme loading obviously led to faster reaction rate, as well as
the accumulation of the insoluble byproduct D-PG. The reaction would
be ended when the solution is concreting. In addition to higher enzyme
preparation costs, increased enzyme loading could also lead to mass
transfer limitations. This is unwanted for increased second hydrolysis
and reduced S/H ratio [15]. When enzyme loading increased to certain
amount (more than 0.04 g/mL), 6-APA conversion cannot be further im-
proved. In consideration of enzyme cost, 0.04 g/mL of enzyme loading
was thus chosen for further study.
Also, the pattern of ampicillin synthesis could be changed by ini-
tial reaction pH (Fig. 2C). Higher initial pH would increase the reac-
tion rate, in accordance with the slightly alkaline pH optimum of
PGAs. But when the initial pH increased from 5.8 to 6.3, the decrease
of initial S/H ratio (from 17 to 12.9) and more byproduct D-PG
accumulation could be observed at the same time [30]. When the ini-
tial pH was reduced to a certain degree, the 6-APA conversion also
decreased as the decline of reaction rate. The highest 6-APA conver-
sion was achieved at initial pH of 5.8. Meanwhile, it is noteworthy
that acidic pH is beneficial for substrate stability, especially 6-APA
[31].

34
S. Deng et al. / Catalysis Communications 79 (2016) 31–34
Unlike initial pH, the change of reaction temperature exhibited
Conflict of interest
smaller effect on ampicillin synthesis (Fig. 2D). The reaction rate in-
creased slightly while temperature is rising, as well as the 6-APA con-
version. The highest 6-APA conversion was achieved at 20 °C. 6-APA
conversion cannot be further improved when the temperature was
over 20 °C due to the balance between hydrolysis and synthesis. Also,
higher temperature would speed up the chemical degradation of sub-
strates especially 6-APA [31]. In consideration of 6-APA conversion, re-
action time and substrate stability, 20 °C was chosen.
The authors declare that they have no conflict of interest.
Acknowledgments
This work was supported by “State Key Laboratory of Pulp and Paper
Engineering (201515)”, “Open Funding Project of the State Key Labora-
tory of Bioreactor Engineering”, “Six talent peaks project in Jiangsu
Province (2015-JY-016)”.
Also, the influence of changed substrate ratio on 6-APA conversion
was studied (Fig. 2E). The 6-APA concentration was kept at 400 mM
while 400–600 mM D-PGME was added to the reaction solution. In-
creased concentration of D-PGME would accelerate the reaction speed,
supporting the assumption that acyl-enzyme intermediate formation
is the rate-limiting step under kinetic control [13]. But the mass transfer
problems could be serious at the same time. Decreased S/H ratio and
more byproduct D-PG accumulation could also be observed. Though ob-
viously higher 6-APA conversion (93.5% vs 91%) was obtained at higher
D-PGME/6-APA ratio (1.05:1 vs 1.00:1) at first. But when this ratio in-
creased from 1.05:1 to 1.10:1 and 1.20:1 (20 and 60 mM extra D-
PGME were added), due to lower initial S/H ratio (decreased from 17
to 15.8 and 14.5, respectively) and the balance between hydrolysis
and synthesis, only 3.2 and 6 mM more 6-APA were converted, respec-
tively. This is not favored for the economic performance of the whole
process for much higher raw material and downstream purification
cost. When more extra D-PGME was added (180 mM), the reaction so-
lution would be concreting before higher 6-APA conversion could be
achieved. In conclusion, D-PGME/6-APA ratio of 1.05:1 could be
regarded as a better choice.
References
[1] C. Aguirre, P. Opazo, M. Venegas, R. Riveros, A. Illanes, Process Biochem. 41 (2006)
1924–1931.
[2] A. Parmar, H. Kumar, S. Marwaha, J.F. Kennedy, Biotechnol. Adv. 18 (2000) 289–301.
[3] A. Bruggink, E.C. Roos, E. de Vroom, Org. Process. Res. Dev. 2 (1998) 128–133.
[4] M.A. Wegman, M.H.A. Janssen, F. van Rantwijk, R.A. Sheldon, Adv. Synth. Catal. 343
(2001) 559–576.
[5] R.A. Sheldon, Chem. Commun. 29 (2008) 3352–3365.
[6] H. Marešová, M. Plačková, M. Grulich, P. Kyslík, Appl. Microbiol. Biotechnol. 98
(2014) 2867–2879.
[7] W.B.L. Alkema, E. de Vries, D.B. Janssen, Biochem. J. 365 (2002) 303–309.
[8] W.B.L. Alkema, A.J. Dijkhuis, E. de Vries, D.B. Janssen, Eur. J. Biochem. 269 (2002)
2093–2100.
[9] E.M. Gabor, et al., Enzym. Microb. Technol. 36 (2005) 182–190.
[10] S. Bečka, et al., Appl. Microbiol. Biotechnol. 98 (2014) 1195.
[11] V.I. Tishkov, S.S. Savin, A.S. Yasnaya, Acta Nat. 2 (2012) 47–61.
[12] A.L. Deaguero, J.K. Blum, A.S. Bommarius, Protein Eng. Des. Sel. 25 (2012) 135–144.
[13] S. Deng, E. Su, X. Ma, S. Yang, D. Wei, J. Biotechnol. 199 (2015) 62–68.
[14] A.K. Chandel, L.V. Rao, M.L. Narasu, O.V. Singh, Enzym. Microb. Technol. 42 (2008)
199–207.
According to the above results, the optimal reaction conditions were
determined: 0.04 g/mL immobilized mutant Af PGA, 15% (v/v) glycerol,
initial pH 5.8, 20 °C, D-PGME/6-APA ratio of 1.05:1. The conversion of 6-
APA reached 93.5% by a 600 min procedure, higher than 90% of the
chemical method. It is noted that this yield was achieved at low D-
PGME/6-APA ratio (1.05:1). Less than 1.12 mol D-PGME was required
to produce 1 mol ampicillin, which is much lower than the reported
values [12,32,33]. All those results met the industrial requirement for
reduced raw material and downstream processing costs. The catalytic
performance of immobilized βF24G mutant Af PGA was in accordance
with its kinetic parameters and catalysts style. High speed of ampicillin
synthesis using free mutant enzyme has been observed in our previous
report [13]. However, when it was using in an immobilized form at high
substrate concentration, the mass transfer problems can be serious. Side
reaction of hydrolysis and byproduct accumulation made the glycerol
addition to be a critical choice. Besides its benefits for catalytic perfor-
mance, glycerol is a cheap, nontoxic, bio-degradable and nonflammable
solvent, allows an eco-efficient, bio-compatible and easily implemented
process [24,25].
[15] A.I. Kallenberg, F. van Rantwijk, R.A. Sheldon, Adv. Synth. Catal. 347 (2005)
905–926.
[16] C.G. Schröen, V.A. Nierstrasz, H.M. Moody, M.J. Hoogschagen, R. Bosma, H.H.
Beeftink, A.E. Janssen, J. Tramper, Biotechnol. Bioeng. 73 (2001) 171–178.
[17] M.I. Youshko, L.M. van Langen, E. de Vroom, F. van Rantwijk, R.A. Sheldon, V.K.
Švedas, Biotechnol. Bioeng. 73 (2001) 426–430.
[18] M.I. Youshko, H.M. Moody, A.L. Bukhanov, W.H. Boosten, V.K. Švedas, Biotechnol.
Bioeng. 85 (2004) 323–329.
[19] R. Fernandez-Lafuente, C. Rosell, B. Piatkowska, J. Guisan, Enzym. Microb. Technol.
19 (1996) 9–14.
[20] A. Illanes, S. Anjarí, R. Arrieta, C. Aguirre, Appl. Biochem. Biotechnol. 97 (2002)
165–179.
[21] A. Illanes, L. Wilson, C. Altamirano, Z. Cabrera, L. Alvarez, C. Aguirre, Enzym. Microb.
Technol. 40 (2007) 195–203.
[22] S.C. Pereira, R. Bussamara, G. Marin, R.L.C. Giordano, J. Dupont, R.C. Giordano, Green
Chem. 14 (2012) 3146–3156.
[23] S. Zhu, X. Ma, E. Su, D. Wei, Green Chem. 17 (2015) 3992–3999.
[24] A. Wolfson, C. Dlugy, D. Tavor, J. Blumenfeld, Y. Shotland, Tetrahedron Asymmetry
17 (2006) 2043–2045.
[25] Y.L. Gu, F. Jérôme, Chem. Soc. Rev. 42 (2013) 9550–9570.
[26] T. Wang, H. Zhu, X. Ma, Z. Fei, Y. Ma, D. Wei, Appl. Microbiol. Biotechnol. 72 (2006)
953–958.
[27] X. Gong, E. Su, P. Wang, D. Wei, Tetrahedron Lett. 52 (2011) 5398–5402.
[28] G.S. Jaiprakash, K.K. Kamalesh, R.A. Gangadhar, Biotechnol. Tech. 1 (1987) 69–72.
[29] M.M. Bradford, Anal. Biochem. 72 (1976) 248–254.
In conclusion, the use of immobilized mutant Penicillin G acylase
and bio-based solvent glycerol at high substrate concentration and
low acyl donor/nucleophile ratio was tested in this study. After process
optimization, this combined strategy had been demonstrated to be eas-
ily implemented, economically efficient and environmentally benign,
providing a promising model for the industrial production of SSBAs.
[30] C.G. Schroën, V.A. Nierstrasz, P.J. Kroon, R. Bosma, A.E.M. Janssen, H.H. Beeftink, J.
Tramper, Enzym. Microb. Technol. 24 (1999) 498–506.
[31] A. Kumar, A. Prabhune, G.C. Suresh, A. Pundle, Process Biochem. 43 (2008) 961–967.
[32] D. Wei, L. Yang, J. Chem. Technol. Biotechnol. 78 (2003) 431–436.
[33] P. Xue, X. Song, X. Cao, Chin. Chem. Lett. 21 (2010) 765–768.