A Single-Transaminase-Catalyzed Biocatalytic Cascade for Efficient Asymmetric Synthesis of l-Phosphinothricin

Article abstract of DOI:10.1002/cbic.202000488

A single-transaminase-catalyzed biocatalytic cascade was developed by employing the desired biocatalyst, ATA-117-Rd11, that showed high activity toward 2-oxo-4-[(hydroxy)(methyl)phosphinoyl] butyric acid (PPO) and α-ketoglutarate, and low activity against pyruvate. The cascade successfully promotes a highly asymmetric amination reaction for the synthesis of l-phosphinothricin (l-PPT) with high conversion (>95 %) and>99 % ee. In a scale-up experiment, using 10 kg pre-frozen E. coli cells harboring ATA-117-Rd11 as catalyst, 80 kg PPO was converted to ≈70 kg l-PPT after 24 hours with a high ee value (>99 %).

Full text of DOI:10.1002/cbic.202000488

Combining Chemistry and Biology
Accepted Article
Title: A single-transaminase-catalyzed biocatalytic cascade for
efficiently asymmetric synthesis of L-phosphinothricin
Authors: Feng Cheng, Ju-Mou Li, Shi-Peng Zhou, Qi Liu, Li-Qun Jin,
Ya-Ping Xue, and Yu-Guo Zheng
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemBioChem 10.1002/cbic.202000488
Link to VoR: https://doi.org/10.1002/cbic.202000488
01/2020

10.1002/cbic.202000488
ChemBioChem
COMMUNICATION
A single-transaminase-catalyzed biocatalytic cascade for efficient
asymmetric synthesis of L-phosphinothricin
Feng Cheng[a], Ju-Mou Li[a], Shi-Peng Zhou[a], Qi Liu[a], Li-Qun Jin[a], and Ya-Ping Xue*[a], Yu-Guo
Zheng[a]
[a]
Dr. F. Cheng, J. -M. Li, S. -P. Zhou, Q. Liu, Prof. L. -Q. Jin, Prof. Y. -P. Xue, Prof. Y. -G. Zheng
The National and Local Joint Engineering Research Center for Biomanufacturing of Chiral Chemicals
Zhejiang University of Technology
Hangzhou 310014, P. R. China
E-mail: xyp@zjut.edu.cn
Supporting information for this article is given via a link at the end of the document.
Abstract: A single-transaminase-catalyzed biocatalytic cascade was
developed by employing a desired biocatalyst ATA-117-Rd11 that
showed high activity toward 2-oxo-4-[(hydroxy)(methyl)phosphinoyl]
butyric acid (PPO) and α-ketoglutarate, and low activity against
pyruvate. The cascade successfully promotes highly asymmetric
amination reaction for the synthesis of L-phosphinothricin (L-PPT)
with high conversion (>95%) and > 99% e.e.. In the scale-up
experiment, using 10 kg pre-frozen E. coli cells harboring ATA-117-
Rd11 as catalyst, 80 kg PPO was converted to ≈70 kg L-PPT after 24
hours with a high e.e. value (>99%).
donor may cause a substrate inhibition, and a large amount of
unconverted amino donor will increase the difficulty of product
post-processing.
To break these bottlenecks, a biocatalytic cascade was
developed by coupling of two transaminases that are an aspartate
transaminase from Bacillus stearothermophilus (AspTA) and a
glutamate: oxaloacetate transaminase (GOT) (Figure 1B). In the
cascade, glutamate was used as an amino donor, and aspartate
(1d) was used as co-substrate that was converted into α-
ketoglutarate (1c). 1c is easily decarboxylated to pyruvate in the
presence of metal ions.[14] This biocatalytic cascade, using 1.2
equivalent 1d, reached a high reaction conversion (85%), and a
96% e.e. of 1b.[15] Since only few amino donor was remained in
the reaction system, the post-processing is relatively simple. In
spite of the success of AspTA-GOT coupled cascade, two
different transaminases were required in the cascade, resulting
high cost of biocatalyst, and redundant fermentation process.[16]
Another disadvantage of the AspTA-GOT coupled cascade is that
the pyruvate formed in the decarboxylation step can lead to a side
reaction where 1f is formed by transamination, which limited the
maximum conversion and may complicate the product purification.
Phosphinothricin (2-amino-4-[(hydroxy)(methyl)phosphonoyl]
butanoic acid, PPT, also known as glufosinate) is spectral, low
toxicity, non-selective herbicide that is widely used for genetically
modified crops.[1] However, the currently commercial PPT is a
racemic mixture containing about 50% D-PPT that is no herbicidal
activity and harmful to the environment. Therefore, it is necessary
to develop a feasible route to synthesize optically pure L-PPT.
Non-enzymatic
synthesis
of
L-PPT
by
asymmetric
hydrogenation,[2] Strecker synthesis,[3] and Michael addition,[4]
has been intensively studied. But, it often raised the problems of
complex synthesis process, high cost of catalyst, and
environmental pollution.[5] Recently, the biological method of
synthesizing L-PPT has attracted more and more attention due to
its advantages of low impact on the environmental, high atom
utilization rate, and high product yield, in which amidase,[6]
nitrilase,[7] deacetylase,[8] amino acid dehydrogenase[9] and
transaminase[10] were involved.
Transaminase (TA, EC 2.6.1.X) catalyzes a transamination
reaction between an amino acid and an α-keto acid, which is
important in the synthesis of natural and unnatural amino acids.[11]
Thus, the application of transaminase to the synthesis of L-PPT
(1b) from 2-oxo-4-[(hydroxy)(methyl)phosphinoyl] butyric acid
(PPO, 1a) is a promising route (Figure 1A). However, organic
synthesis by TAs usually suffer from the unfavorable reaction
equilibrium so that net conversion of 1a to L-PPT is
thermodynamically limited and very difficult to reach >85%.[12] In
order to shift the reaction balance to L-PPT and increase the
conversion of transamination, one way is to use a large number
of amino donors or smart amino donors.[12] For instance, in the
TA-catalyzed reaction, Schulz et al utilized a quadruple equivalent
of an amino donor (toward 1a) for synthesis of 1b, which yielded
a conversion of 90% with an e.e. >99%.[13] However, there are still
two major drawbacks of this strategy. The highly loaded amino
Figure 1. The scheme of biosynthesis of L-PPT. (A) a single-TA-catalyzed
reaction for synthesis of L-PPT, (B) couple enzymes (TA and GOT)-catalyzed
cascade for synthesis of L-PPT, (C) a single enzyme-catalyzed cascade for
synthesis of L-PPT
This article is protected by copyright. All rights reserved.

10.1002/cbic.202000488
ChemBioChem
COMMUNICATION
In this work, to solve the above-mentioned problems, we
developed a single-transaminase-catalyzed cascade, in which a
desired transaminase displaying high activity toward PPO and α-
KG and low activity toward pyruvate was obtained by virtual
screening and experiments. Firstly, 50 different TAs, sharing
20%-90% similarities, were selected from NCBI database by a
three-step procedure (see method section in supporting
informatino). Virtual screening was performed by homology
low possibility for pyruvate binding. Finally, the 9 transaminases
were selected for the further experimental investigation.
The three ligands were used as substrate for testing TA
activities. All the selected nine TAs exhibited high activities of
more than 50 U/g toward 1a (Table 1), showing good agreement
with the docking results that, the binding energies between 1a and
the nine TAs were all lower than -5 kcal/mol. Among them, two
TAs (ATA117-Rd11 and ABAT1) displayed the highest activity of
modeling and molecular docking toward PPO, α-KG and pyruvate. >100 U/g. Using 20 mM 1a as substrate and glutamate as amino
Table S1 summarized the docked binding energy of these 50
transaminases toward the three ligands. Only 10 transaminases
showed low binding energy (<-5 kcal/mol) to both PPO and α-KG,
which indicated that the desired biocatalyst is likely to be found
from the 10 TAs. Among them, ATA-117-Rd11 showed the lowest
binding energy, which is -5.9 and -6.7 kcal/mol to PPO and α-KG,
respectively. The evolutionary conservation of the active site
region of the 50 TAs were evaluated by ConSurf server (colored
turquoise to maroon according to the conservation grades)[17] The
conservation scores of most residues in active site were higher
than 6, which indicated that they were conservative (Figure S1).
Furthermore, a higher binding energy (>-4.5 kcal/mol) of pyruvate
was observed from 9 out of these 10 transaminases, suggesting
donor, more than 85% conversion was reached by 2 h reaction
(91.7% for ATA-117-Rd11; 86.0% for ABAT1), yielding 1b with
high e.e. (>99%). They also showed the highest activity toward α-
KG (341 U/g for ATA-117-Rd11, 199 U/g for ABAT1). Importantly,
the ATA117-Rd11 showed the lowest activity toward pyruvate
(only 3 U/g), which indicated that the side-reaction may not be
detectable. Therefore, ATA-117-Rd11 was selected as a desired
biocatalyst for development of the single-TA-catalyzed cascade.
Interestingly, ATA-117-Rd11 was
a
highly engineered
transaminase for the biomanufacture of sitaglipitin (a widely used
therapeutic agent for type-2 diabetes) by 11 rounds of directed
evolution, in which its substrate-binding pocket was enlarged to
extend its substrate specificity.[12]
Table 1. The activities of selected enzymes toward PPO, α-KG and pyruvate
Activity
toward
PPO
Activity
toward α-
KG
Activity
toward
pyruvate
(U/g)
e.e. of
L-PPT
(%) b
Conversion of
PPO (%) a
Entry
Enzyme
Genbank
(U/g)
(U/g)
1
2
3
ATA-117-Rd11
ATA-117
3WWJ_A
3WWI_A
115±4
83±3
75±2
341±12
145±9
143±2
3±0.2
6±0.5
91.7
75.9
36.2
>99
98.4
92.3
R-selective transaminase
WP_010910285.1
20±1.1
4
AT-ωATA
EAQ46075.1
65±2
51±3
13±0.9
36.9
99.2
5
6
Aminotransferase
WP_011462786.1
WP_011349475.1
74±3
83±4
75±4
12±1.2
13±0.2
21.8
20.7
98.1
98.3
Amine transaminase
(fold IV)
114±8
ABAT1
7
8
(4-aminobutyrate
aminotransferase)
ABAT2
(4-aminobutyrate
aminotransferase)
AKK18527.2
108±4
87±5
199±10
80±6
19±0.2
17±0.6
86.0
46.1
96.8
94.2
WP_001095559.1
ABAT3
9
(4-aminobutyrate
aminotransferase)
WP_021700965.1
79±4
201±11
16±0.2
58.2
95.3
]
aConv. (PPO) = [D−PPT +[L−PPT]×100%；b푒. 푒. = [[L−PPT]−[D−PPT]×100%; [D-PPT], [L-PPT] and [PPO] are the concentrations of D-PPT, L-PPT and PPO after reaction,
[
]
[
]
]
L−PPT +[D−PPT]
PPO 0− PPO
respectively, and [PPO]0 was the initial concentration of PPO.
Since the ATA-117-Rd11 exhibited higher activities toward
PPO and α-KG than other selected TAs, and low activity toward
pyruvate, ATA-117-Rd11 was employed as the single biocatalyst
to develop the enzyme cascade. As the principle shown in Figure
1C, glutamate acid was used as the amino donor for supplying
the amino group to PPO and the L-aspartate acid was utilized as
co-substrate to regenerate the amino donor. The substrate PPO
and the amino donor was initially set as 100 mM and 20 mM,
respectively. The concentration of co-substrate L-aspartic acid
was optimized from 100 mM to 300 mM. The highest conversion
was observed when the L-aspartic acid concentration was higher
than 110 mM. Furthermore, when the substrate concentration
was increased to 200 mM, the optimal co-substrate concentration
was 220 mM. Taken together, the optimal ratio between 1d and
1a was 1.1, which is substantially lower than that used in TA-
catalyzed reaction (> 4 equivalent amino donor).
The effects of pH and temperature on the single-TA-catalyzed
cascade were tested in range of 30 to 60 °C (temperature) and
6.0 to 11.0 (pH). The conversion was increased from 15.6%
(30 °C) to 83.9% (40 °C), but decreased when the temperature is
higher than 40 °C. (Figure S2A). While, when the pH was lower
than 7.0 or higher than 8.5, the conversion was lower than 80%.
This article is protected by copyright. All rights reserved.

10.1002/cbic.202000488
ChemBioChem
COMMUNICATION
The optimal pH value was selected as 8.0 in Tris-HCl buffer
(Figure S2B). Furthermore, the enzyme cascade was performed
in a whole cell format. The cell concentration (DCW: 0.5-5 g/L) is
optimized and the optimal lyophilized cell concentration is 3 g/L.
(Figure S2C). Moreover, since ATA-117-Rd11 is PLP-dependent,
it is necessary to optimize the PLP concentration. When the PLP
concentration is higher than 0.2 mM (toward 3 g/L cells), the
conversion of 1a is higher than 80% (Figure S2D). The further
increase in PLP concentration (0.2 mM) did not enhance the
conversion of PPO in this cascade.
To evaluate the potential of this cascade for synthesis of L-PPT,
different concentrations of PPO were used as substrate and
lyophilized E. coli cells harboring ATA-117-Rd11 used as
biocatalyst. Under optimized conditions (pH 8.0 and 40 °C), the
conversion of 100 mM 1a was almost complete (conversion >
96%, >99% e.e.) within 24 hours using 3 g·L−1 lyophilized E. coli
co-expressing ATA-117-Rd11, and the side-product alanine was
undetectable. To explore the performance of this cascade, the
substrate concentration was gradually increased from 100 mM to
200 mM and 500 mM, and the ammonium and glucose
concentrations were also increased to the same extent. As shown
in Figure 2, the conversion of 200 mM and 500 mM 1a (89.04
g·L−1) reached 95% and 92% after 24 h reaction with the same
amount of biocatalyst (3 g·L−1 lyophilized cells harboring ATA-
117-Rd11), affording an L-PPT with e.e.>99%. Compared to the
reports, the highest conversion (>95%) was achieved by the ATA-
117-Rd11 catalyzed cascade, in which only 3 g·L −1 biocatalyst
was used with 1.1 equivalent aspartate and 0.2 equivalent
glutamate (in total 1.3 equivalent amines). Due to unfavorable
thermodynamic equilibrium, the higher amount amino donor (> 4
equivalent) should be employed in TA catalyzed amination
reaction toward 20 g/L PPO (Table 2). Therefore, the single-TA-
catalyzed cascade displays a great potential in biocatalytic
synthesis of L-PPT.
Table 2. Comparison of the cascade with other reported reactions for production of L-PPT
Substrate
loading (g/L)
Catalyst loading
(g/L)
Time
(h)
Conversion
(%)
Enzyme
Amidase
Substrate
Amino donor
Reference
a PPT
[18]
150
20
0.03a
34b
-
6.8
14
49.7
91.2
derivatived
α-Transaminase
(single enzyme reaction)
[19]
[15]
1a
1a
1a
>4 equiv. glutamate
AspTA and GOT
(Coupled enzyme
cascade)
0.2 equiv.
glutamate,
1.2 equiv. 1d
0.08 for TAa,
89
89
24
24
Up to 85
92.4
0.02 for GOTa
0.2 equiv.
glutamate,
1.1 equiv. 1d
ATA-117-Rd11
(single enzyme cascade)
3c
This study
a purified enzyme used as catalyst.
b The immobilized α-TA used as catalyst.
c The catalyst loading as dry cell weight.
d The rac-4-(hydroxy(methyl)phosphoryl)-2-(2-phenylacetamido) butanoic acid (a PPT derivative) used as substrate.
In a kilogram-scale experiment, a 5-ton fermenter (Figure S3)
was used to produce E. coli cells harboring ATA-117-Rd11, which
yielded approximately 150 kg E. coli wet cells (see supporting
information for the fermenter settings in deail). In a 1000 L
reaction mixture, 80 kg PPO was mixed with 10 kg pre-frozen E.
coli cells (wet cells) harboring ATA-117-Rd11 in Tris-HCl buffer
(pH = 8.0). During the 24 h reaction, the pH was automatically
adjusted to 8.0 by titrating with NH3·H 2O or HCl. The samples
were taken from the bioreactor every 4 h.The reaction process
was monitored by the determination of residual PPO
concentration and generated PPT using HPLC, as previously
described.[11] The formed product was purified by an eight-step
purification process shown in supporting information, [20] and the
total yield was about 81% (from ≈440 mol PPO to ≈ 360 mol L-
PPT). Finally, approximately 70 kg of L-PPT was obtained and
characterized using HPLC (purity = 92.3% with 99% e.e. (Figure
S4), ESI-MS (m/z= 180.0437, calcd. for C5H11NO4P-[M]-:
180.0426, Figure S5) and NMR analysis (Figure S6-S7).
Figure 2. Biosynthesis of L-PPT using ATA-117-Rd11 catalyzed cascade
toward 100, 200, and 500 mM PPO.
In summary, we demonstrated a single-transaminase catalyzed
cascade that enables the high conversion of high concentration
PPO. The e.e. of L-PPT is higher than 99% and side-product
alanine is almost undetectable. In the kilogram-scale reaction,
approximately 70 kg L-PPT was yielded after 24 h with a high e.e.
value (>99%) when 80 kg PPO was loaded as substrate.
This article is protected by copyright. All rights reserved.

10.1002/cbic.202000488
ChemBioChem
COMMUNICATION
[10]
[11]
[12]
a) H. Zhou, L. Meng, X. Yin, Y. Liu, J. Wu, G. Xu, M. Wu, L. Yang,
Appl Catal A-Gen 2020, 589, 117239; b) D. X. Jia, H. P. Xu, C. Y.
Sun, C. Peng, J. L. Li, L. Q. Jin, F. Cheng, Z. Q. Liu, Y. P. Xue, Y. G.
Zheng, Bioproc. Biosyst. Eng. 2020, 43, 1599–1607; c) L.-Q. Jin, F.
Peng, H.-L. Liu, F. Cheng, D.-X. Jia, J.-M. Xu, Z.-Q. Liu, Y.-P. Xue,
Y.-G. Zheng, Process Biochem. 2019, 85, 60-67.
Acknowledgements
This work was financially supported by Natural Science
Foundation of China (21978268, 31700693 and 31970046) and
the National Defense Science and Technology Innovation Zone
Foundation of China.
a) S. A. Kelly, S. Pohle, S. Wharry, S. Mix, C. C. R. Allen, T. S. Moody,
B. F. Gilmore, Chem Rev 2018, 118, 349-367; b) Y. P. Xue, C. H.
Cao, Y. G. Zheng, Chem. Soc. Rev. 2018, 47, 1516-1561; c) D. P.
Gavin, F. J. Reen, J. Rocha-Martin, I. Abreu-Castilla, D. F. Woods,
A. M. Foley, P. A. Sánchez-Murcia, M. Schwarz, P. O’Neill, A. R.
Maguire, F. O’Gara, Sci. Rep. 2019, 9, 20285.
Conflict of interest
a) A. Gomm, W. Lewis, A. P. Green, E. O'Reilly, ChemCatChem
2016, 22, 12692-12695; b) C. K. Savile, J. M. Janey, E. C. Mundorff,
J. C. Moore, S. Tam, W. R. Jarvis, J. C. Colbeck, A. Krebber, F. J.
Fleitz, J. Brands, P. N. Devine, G. W. Huisman, G. J. Hughes,
Science 2010, 329, 305-309; c) L. Skalden, C. Peters, J. Dickerhoff,
A. Nobili, H.-J. Joosten, K. Weisz, M. Höhne, U. T. Bornscheuer,
ChemBioChem 2015, 16, 1041-1045; d) D. Koszelewski, K. Tauber,
K. Faber, W. Kroutil, Trends Biotechnol. 2010, 28, 324-332; e) M. S.
Weiß, I. V. Pavlidis, P. Spurr, S. P. Hanlon, B. Wirz, H. Iding, U. T.
Bornscheuer, ChemBioChem 2017, 18, 1022-1026.
The authors declare no conflicts of interest.
Keywords: L-phosphinothricin • transaminase • enzyme cascade
• virtual screening • biocatalysis
[1]
a) G. P. Horsman, D. L. Zechel, Chem. Rev. 2017, 117, 5704-5783;
b) E. Bayer, K. H. Gugel, K. Hägele, H. Hagenmaier, S. Jessipow,
W. A. König, H. Zähner, Helv. Chim. Acta. 1972, 55, 224-239; c) A.
R. Fraser, S. M. Ridley, Planta 1984, 161, 470-474.
H.J. Zeiss, J. Pestic. Sci. 1994, 41, 269-277.
[13]
[14]
[15]
[16]
a) A. Schulz, P. Taggeselle, D. Tripier, K. Bartsch, Appl. Environ.
Microbiol. 1990, 56, 1-6.
J. David Rozzell, in Method Enzymol, Vol. 136, Academic Press,
1987, pp. 479-497.
[2]
[3]
Y. M. Li, X. H. Du, Q. H. Zhou, S. D. Chen, Org. Prep. Proced. Int.
2014, 46, 565-568.
K. Bartsch, R. Schneider, A. Schulz, Appl. Environ. Microbiol. 1996,
62, 3794-3799.
[4]
[5]
N. Minowa, M. Hirayama, S. Fukatsu, B. Chem. Soc. Jpn. 1987, 60,
1761-1766.
a) I. G. Fotheringham, N. Grinter, D. P. Pantaleone, R. F. Senkpeil,
P. P. Taylor, Bioorg. Med. Chem. 1999, 7, 2209-2213; b) D. J. Ager,
T. Li, D. P. Pantaleone, R. F. Senkpeil, P. P. Taylor, I. G.
Fotheringham, J. Mol. Catal. B-Enzym. 2001, 11, 199-205.
H. Ashkenazy, E. Erez, E. Martz, T. Pupko, N. Ben-Tal, Nucl. Acids
Res. 2010, 38, W529–W533.
A. L. Ahmad, P. C. Oh, S. R. Abd Shukor, Biotechnol. Adv. 2009, 27,
286-296.
[6]
[7]
J. Xu, F. Li, Y. Xue, Y. Zheng, Chirality 2019, 31, 513-521.
Kim F. Albizati, S. Kambourakis, A. Grubbs, B. C. Borer,
US9683001B2 2017.
[17]
[18]
[19]
[20]
X. Kang, X. Cai, Z. Liu, Y. Zheng, Bioresource Technol. 2019, 289,
121658.
[8]
[9]
a) K. Bartsch, US6686181B1 2004; b) C. C. Cao, F. Cheng, Y. P.
Xue, Y. G. Zheng, Enzyme Microb. Technol. 2020, 135, 109493.
a) X. J. Yin, J. P. Wu, L. R. Yang, Appl. Microbiol. Biot. 2018, 102,
4425-4433; b) X. J. Yin, Y. Liu, L. Meng, H. Zhou, J. Wu, L. R. Yang,
Adv. Synth. Catal. 2019, 361, 803-812; c) F. Cheng, H. Li, K. Zhang,
Q. H. Li, D. Xie, Y. P. Xue, Y. G. Zheng, J. Biotechnol. 2020, 312,
35-43.
D. Jia, Z. Liu, H. Xu, J. Li, J. Li, L. Jin, F. Cheng, Z. Liu, Y. Xue, Y.
Zheng, J. Biotechol. 2019, 302, 10-17.
S. Lv, Y. Guo, Y. Xue, J. Xu, Y. Zheng, Sep. Sci. Technol. 2020, 55,
779–787.
This article is protected by copyright. All rights reserved.

10.1002/cbic.202000488
ChemBioChem
COMMUNICATION
Entry for the Table of Contents
Insert graphic for Table of Contents here. ((Please ensure your graphic is in one of following formats))
Asymmetric synthesis of L-phosphinothricin: A single-transaminase-catalyzed biocatalytic cascade for L-phosphinothricin
synthesis was developed by employing a desired biocatalyst ATA-117-Rd11. High conversion, high e.e. of product and low side reaction
was achieved by employing only 1.3 equivalent amine donors. In a kilogram-scale experiment, ATA-117-Rd11, about 70 kg L-
phosphinothricin (e.e. >99%) was yielded from 80 kg PPO within 24 hours.
This article is protected by copyright. All rights reserved.