One-step Synthesis of α-Keto Acids from Racemic Amino Acids by A Versatile Immobilized Multienzyme Cell-free System

Article abstract of DOI:10.1002/cctc.201800359

The elevated value of α-keto acids has pushed scientists to explore more efficient and less expensive alternatives for their synthesis. In this work, an immobilized tri-enzyme system that produced α-keto acids in “one-pot” from l- or racemic mixtures of diverse amino acids was presented. The system combined a broad-spectrum amino acid racemase (BsrV), a d-amino acid oxidase (DAAO) and catalase (CAT). BsrV racemized l-amino acids into their d-enantiomers, DAAO catalyzed the stereospecific oxidative deamination of the d-amino acids into their corresponding α-keto acids, ammonium ion, and H₂O₂. Finally, CAT converted the inactivating H₂O₂ into H₂O and O₂, which can be reused by the oxidase reaction. BsrV thermal stability was improved 3,300-fold by immobilizing the enzyme on glyoxyl-activated agarose beads. DAAO and CAT were co-immobilized on agarose beads functionalized with glutaraldehyde groups for enhancing their stabilities and eliminating H₂O₂ in a much more effective way. To show the versatility of this system, racemic mixtures of amino acids were converted in their corresponding α-keto acids. The coupling of the three immobilized enzymes permitted conversions of approximately 99 % through a dynamic kinetic resolution process. This system conserved 100 % of its initial effectiveness after 8 reaction cycles. Collectively, our innovative tri-enzyme system for the synthesis of α-keto acids opens the door for a cheapening in the production of many pharmaceutical and cosmetics.

Full text of DOI:10.1002/cctc.201800359

Accepted Article
Title: One-step synthesis of alpha-keto acids from racemic amino acids
by a versatile immobilized multienzyme cell-free system
Authors: Alejandro H. Orrego, Fernando López-Gallego, Akbar
Espaillat, Felipe Cava, José M. Guisan, and Javier Rocha-
Martin
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: ChemCatChem 10.1002/cctc.201800359
Link to VoR: http://dx.doi.org/10.1002/cctc.201800359
A Journal of
www.chemcatchem.org

ChemCatChem
10.1002/cctc.201800359
FULL PAPER
One-step synthesis of alpha-keto acids from racemic amino acids
by a versatile immobilized multienzyme cell-free system
[
a]
[b] [c]
[d]
[d]
Alejandro H. Orrego, Fernando López-Gallego,
Akbar Espaillat, Felipe Cava,* José M.
[
a]
[a]
Guisan* and Javier Rocha-Martin*
Abstract: The elevated value of α-keto acids has pushed scientists
to explore more efficient and less expensive alternatives for their
synthesis. In this work, we present an immobilized tri-enzyme system
that produces α-keto acids in “one-pot” from L- or racemic mixtures of
diverse amino acids. The system combines a broad-spectrum amino
acid racemase (BsrV), a D-amino acid oxidase (DAAO) and catalase
methylthiobutyric acid (α-KMTB). α-KIC is an intermediary
metabolite of leucine degradation and synthesis, used for the
treatment of chronic glomerulonephritis, hepatitis B and colon
cancer.[ Furthermore, α-KIC has been investigated for the de-
repression of silenced tumour suppressor genes,[3] and it can also
increase muscular power, stimulate milk production and secretion
of insulin.[1a, 3-4] While, α-KMTB (keto form of methionine) inhibits
3]
(CAT). BsrV racemizes L-amino acids into their D-enantiomers,
[
5]
DAAO catalyzes the stereospecific oxidative deamination of the D-
the growth of methionine-dependent tumours , and has also
been used in the treatment of uremic patients.[6]
amino acids into their corresponding α-keto acids, ammonium ion, and
H
2
O
2
. Finally, CAT converts the inactivating H
2
O
2
into H
2
O and O
2
,
Currently, most α-keto acids are being generated by multiple-
step chemical synthesis, which results in high production costs
and can also cause environmental pollution.[ Consequently,
there is considerable interest in the development of eco-friendly
alternatives for the production of α-keto acids. One of these
alternatives has been the use of microbial production through a
direct fermentation process and whole-cell biocatalyst.[ In
general, direct fermentation uses low-priced carbon sources, but
it has low productivity. Whole-cell biocatalyst has several
advantages such as high yield and productivity. However, this
method requires the genetic manipulation of metabolic pathways
which can be reused by the oxidase reaction. BsrV thermal stability
was improved 3,300-fold by immobilizing the enzyme on glyoxyl-
activated agarose beads. DAAO and CAT were co-immobilized on
agarose beads functionalized with glutaraldehyde groups for
7]
2 2
enhancing their stabilities and eliminate H O in a much more effective
8]
way. To show the versatility of this system, racemic mixtures of amino
acids were converted in their corresponding α-keto acids. The
coupling of the three immobilized enzymes permitted conversions of
approximately 99% through a dynamic kinetic resolution process. This
system conserves 100% of its initial effectiveness after 8 reaction
cycles. Collectively, our innovative tri-enzyme system for the
synthesis of α-keto acids opens the door for a cheapening in the
production of many pharmaceutical and cosmetics.
of
manipulation can cause the accumulation of α-keto acids that
promotes slower cell growth rate requiring expensive
a
microorganism through metabolic engineering.[8] This
a
[
9]
supplementation . The use of soluble enzymes to produce α-keto
acids has as one of the main disadvantages the previous
purification of the enzyme that increases the cost of the
bioconversion regarding the whole-cell method. However,
enzymatic bioconversion could minimize waste by substrate
utilization of the whole-cell biocatalyst. D-amino acid oxidases
Introduction
α-keto acids are very relevant organic compounds in the
pharmaceutical, food, cosmetic and animal feed industries.
They are precursors of biofuels as n-butanol, HIV antiretroviral
drugs,[1b] antibiotics,[2] sweetener aspartame,[1d] chiral α- and β-
amino acids.[2] Among the α-keto acids with a greater potential
market are α-ketoisocaproate (α-KIC) and α-keto-γ-
[1]
(
DAAO, EC 1.4.3.3) have also used for the enzymatic production
[
1a]
[10]
of α-keto acids from D-amino acids. The main disadvantages
of this method are: (i) a yield of only 50% of the desired
enantiomer is obtained starting from a D,L-amino acid mixture,
2 2
and (ii) H O produced during the reaction can act on the newly
formed α-keto acid forming one-carbon shorter carboxylic acid
and can negatively affect the operational stability of the
enzyme.[ Another important aspect that should be considered
during the production is the separation of α-keto acids,
particularly when they are to be used as building blocks for drug
production. To resolve this problem, an effective approach is the
use of immobilized enzymes that permits an easy regulation of
the reaction process thereby reducing the effects of possible
inhibition by products and reaction conditions.[8, 12]
In this arena, systems biocatalysis has emerged as a new
discipline that permits a more efficient synthesis of chemical
compounds by the spatial organization of enzymes in vitro to
accomplish complex reaction cascades.[13] Heterogeneous
biocatalysis (involving enzymes immobilized on solid supports) is
being successfully applied to design multi-enzyme systems for
synthetic and analytical chemistry by harnessing the exquisite
selectivity of enzymes with smaller ecological footprint compared
[
[
a]
b]
Mr. A. H., Orrego, Prof. J. M., Guisan, Dr. J., Rocha-Martin
Department of Biocatalysis
Institute of Catalysis and Petrochemistry (ICP) CSIC
Campus UAM. Cantoblanco. 28049 Madrid (Spain)
E-mail: javirocha@icp.csic.es; jmguisan@icp.csic.es
Dr. F., López-Gallego
Departamento de Química Orgánica
Instituto de Síntesis Química y Catálisis Homogénea (ISQCH)
CSIC-Universidad de Zaragoza
11]
50009 Zaragoza (Spain)
[
[
c]
d]
ARAID Foundation, Zaragoza, Spain
Dr. A., Espaillat, Dr. F. Cava
Department of Molecular Biology and Laboratory for Molecular
Infection Medicine Sweden
Umea Centre for Microbial Research. Umea University. Umea
(Sweden)
E-mail: felipe.cava@molbiol.umu.se
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201800359
FULL PAPER
to chemical processes.[14] Thus, immobilized multi-enzyme
systems have been developed to carry out chemo-, regio- and
catalase (CAT) from bovine liver (Scheme 1). Finally, we
demonstrated the functionality of this immobilized multienzyme
system by achieving the synthesis of α-ketoisocaproate (α-KC),
α-KIC and α-KMTB via an enzyme-catalyzed DKR.
stereoselective
biotransformations.[15]
Stereoselective
biotransformations can be mainly classified into two different
categories: asymmetric synthesis and kinetic resolution of
racemates (KR).[16] A main disadvantage of KR is its limited yield
of 50%, and this step is usually economically inefficient.[16-17] One
solution to overcome this theoretical bottleneck is through a
dynamic kinetic resolution (DKR), which consists in the
introduction of a racemisation step in the process where the
racemic starting material is transformed into one enantiomer and
allows theoretical yields of ca.100% of pure products. In this way,
the addition of a racemisation step to the oxidative deamination
of a D,L-mixture by a DAAO could reach an ideal yield of 100% of
the desired product. The enzymes that catalyze the conversion
between amino acid enantiomers are called amino acid
racemases.[ Amino acid racemases (EC 5.1.1.X) are usually
classified into two major categories, PLP-dependent and PLP-
independent racemases, based on the use of pyridoxal-5-
phosphate (PLP) as a cofactor.[19] Although these enzymes are
DAAO
BsrV
CAT
Scheme 1. Multienzyme biosynthetic pathway proposed to produce α-keto
18]
acids. R = -CH
2 2 3 2 3 2
-CH -S-CH (Methionine); R= -CH -CH-(CH ) (Leucine); R= -
(
CH -CH (Norleucine).
2
)
3
3
mainly monospecific (i.e., racemize one amino acid), the recently Results and Discussion
discovered PLP-dependent broad spectrum racemases (Bsrs)
accept several amino acids as substrates[20], thus having a great
biotechnological potential for industrial applications. In particular,
Vibrio cholerae Bsr (BsrV) is a homodimeric enzyme that
racemizes aliphatic and basic amino acids as well as a few non-
proteinogenic amino acids such as norleucine, homoserine,
ornithine, N-acetyl lysine methyl ester, diaminobutyrate and
amino•butyrate.[20a]
For industrial applications, protein immobilization is a powerful
tool to overcome the limitations of enzymes as catalysts and make
an economically feasible industrial process.[21] Enzyme
immobilization may simplify the downstream processing, can
enhance enzyme stability, and consequently, it allows the reuse
of the catalyst.[21a, 22] Moreover, immobilization protocols can
affect not only stability but also other enzyme properties such as
activity, selectivity, and inhibition.[21a, 23] A challenging step in the
immobilization of PLP-dependent enzymes is to preserve the
A novel tri-enzymatic system catalyzes the bioconversion of
racemic mixtures of amino acids or L-amino acids to α-keto
acids
To produce α-keto acids from a racemic mixture of amino acids or
L-amino acids (Scheme 1) we developed of a novel tri-enzymatic
system composed by: (1) BsrV, which allows the in situ
continuous racemization of L-amino acids into D-amino acids.[
Introducing this racemization step, the theoretical yield of 50%
20a]
[18]
can be overcome and reach 100% through a DKR process. (2)
DAAO, a flavoenzyme that catalyzes [10a, 28] the stereospecific
oxidative deamination of the D-amino acid into the corresponding
α-keto acid, ammonium ion, and H
introduced in the system to remove the H
water and O2. CAT is fundamental in this system because
otherwise, H could convert α-keto acids to one-carbon shorter
carboxylic acids. In addition, H has a strong denaturing effect
2
O
2
;[11] and (3) CAT was
yielding innocuous
2
O
2
2
O
2
internal aldimine bond (Schiff base) between the ε-NH
2
of a lysine
2 2
O
[10b,
of the enzyme and the aldehyde groups of the cofactor (Scheme
S1) during the immobilization process.[24] Only a few PLP-
dependent enzymes have been successfully immobilized, being
the epoxy and glutaraldehyde activated supports the most
common methodologies used.[25] In this regard, strategies such as
those involving reduction steps have been avoided since it could
irreversibly inactivate the enzyme by reducing the Schiff’s base
between the enzyme and the PLP cofactor (Scheme S1).[26]
Conversely, multipoint covalent attachment on glyoxyl agarose
on proteins and influences the operational stability of DAAO.
11]
2
Moreover, CAT recycles half of consumed O by the oxidase
reaction.
The reaction performed by soluble DAAO stopped after a few
minutes (data not shown), in agreement with a previous report by
Fernandez-Lafuente et al., suggesting that the presence of
gas/aqueous interfaces and H
2
O
2
could rapidly inactivate the
removal, and
soluble DAAO.[11] To solve this issue and the H
O
2
2
thus avoiding the non-enzymatic oxidative decarboxylation of the
reaction product as well as enzyme inactivation, DAAO and CAT
were irreversibly co-immobilized in a ratio 1:10 (mg DAAO/mg
CAT) onto the same porous glutaraldehyde-activated agarose
[
21a, 23a,
are the most successful strategies reported in the literature.
2
7]
Here, we provide an efficient and environmentally-friendly
alternative to the current α-keto acids production protocols by the
design of an immobilized tri-enzyme cell-free system. This
innovative pathway synthesizes α-keto acids from their
corresponding racemic amino acid mixtures or L-amino acids by
combining three different enzymes: a BsrV from V. cholerae, a D-
amino acid oxidase from Trigonopsis variabilis (DAAO) and
(Ag-Glutaraldehyde) structure, where the H
eliminated in situ by CAT, as previously described.
2 2
O produced can be
[
11]
Immobilization-stabilization of BsrV on activated agarose
beads
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201800359
FULL PAPER
To address the design and fabrication of a heterogeneous
enzyme system for the production of α-keto acids, BsrV was
immobilized onto differently activated agarose-based matrixes.
The different immobilization procedures were evaluated by
determining the immobilization yield, expressed activity (Table 1)
and enzyme stability (Figure 1). As expected, we found a great
variability in the immobilization yield and expressed activity
depending on the immobilization chemistry used (Table 1).
a non-covalent immobilization method. BsrV was immobilized on
agarose activated with epoxide groups (Ag-Epoxide), cyanogen
bromide (Ag-CB), glutaraldehyde (Ag-Glutaraldehyde) and
primary amino groups (Ag-MANAE). The expressed activity and
immobilization yield are summarized in Table 1. BsrV reversibly
immobilized on Ag-MANAE through rich areas in acid residues
(Asp and Glu), expressed 98% of its initial activity with 90% of
immobilization yield. When BsrV was covalently immobilized on
Ag-Epoxide, the immobilization rate was slow, and after overnight
incubation, BsrV expressed 14% of its initial activity and
immobilization yield was 61%. Meanwhile, BsrV irreversibly
immobilized on Ag-Glutaraldehyde, and Ag-CB expressed 61 and
Table 1. Parameters of BsrV immobilization onto agarose via different
chemistries.
7
8% of its initial activity, respectively.
The different immobilized preparations were incubated at 70
Immobilization yield
Expressed activity
Activated agarose[a]
%)[b]
(%)[c]
(
Ag-CB[d]
ºC and pH 7.5 to evaluate their thermal stabilities (Figure 1). We
did not include BsrV on the Ag-Epoxide given the low
immobilization yield and expressed activity obtained. Our results
show that all immobilized BsrVs were more stable than the soluble
counterpart, which was instantly inactivated (data not shown).
Covalent immobilization promoted higher stabilization than ion
exchange immobilization (Ag-MANAE). BsrV immobilized on Ag-
G (for 1 hour at 25 ºC) at alkaline conditions showed the highest
thermal stability. While BsrV immobilized on Ag-Glutaraldehyde
was 6-fold more stable than BsrV immobilized on Ag-MANAE and
Ag-CB, which presented similar half-life times. Despite the low
expressed activity of the BsrV immobilized Ag-G, the high stability
achieved by this biocatalyst through this immobilization protocol
led us to further investigate the possibility of obtaining a more
active biocatalyst.
97.4 ± 0.8
60.7 ± 1
91 ± 0.9
90 ± 2
78 ± 2
14 ± 1
Ag-Epoxide
Ag-G
4 ± 0.5
98 ± 0.8
61 ± 1.5
Ag-MANAE
Ag-Glutaraldehyde
98.2 ± 1.6
[
a] Agarose 6% beads cross-linked (BCL) was activated with different
functional groups as described in the methods to immobilize the enzyme
through different chemistries: Ag-CB, agarose support activated with
cyanogen bromide groups: Ag-Epoxide, agarose support activated with
epoxide groups; Ag-G, fully activated agarose glyoxyl support (100
µmol/mL); Ag-MANAE, monoaminoethyl-N-aminoethyl agarose; and Ag-
Glutaraldehyde, agarose support activated with glutarladehyde. [b]
Immobilization yield (%)= (activity of the solution offered for immobilization
–
activity in the supernatant at the end of the immobilization process)/
(activity of the solution offered for immobilization) x 100. [c] Expressed
activity (%)= (activity in immobilized preparation) / (activity of the solution
offered for immobilization – activity in the supernatant at the end of the
immobilization process) x 100. [d] In the case of the matrix Ag-CB the
agarose used was 4BCL. All data are the mean value of three separate
experiments where the error value was never higher than 5%.
In general, we observed that when the agarose beads were
not functionalized with glyoxyl groups (with cyanogen bromide
groups, Ag-CB; epoxide groups, Ag-Epoxide; monoaminoethyl-N-
aminoethyl, Ag-MANAE; and glutaraldehyde, Ag-Glutaraldehyde),
BsrV retained more catalytic activity (Table 1). BsrV immobilized
on Ag-G retained no more than 4% of its initial activity. This is
likely because the multipoint covalent immobilization of enzymes
on Ag-G requires the reduction of the Schiff bases formed.[23a, 29]
BsrV’s PLP is covalently attached to the enzyme via a Schiff base
through an internal aldimine bond between the ɛ-amino group of
a Lys residue of the protein and aldehyde group of the cofactor
(
Scheme S1). Thus, the covalent immobilization methods as Ag-
G that require a reducing step may produce a dramatic reduction
in the catalytic activity likely caused by the reduction of the internal
aldimine bond between Lys and PLP that prevents the exchange
reaction of the Schiff base and product release. The same result
was observed when the soluble BsrV was reduced with
borohydride. The reduction of Schiff´s base in the active center
resulted in the loss of catalytic activity and the disappearance of
the characteristic PLP-absorption peak between 410-420 nm
Figure 1. Thermal inactivation courses of different BsrV preparations at 70 ºC
and pH 7.5. Symbols: Ag-MANAE (diamonds-dash-dot line), Ag-Glutaraldehyde
(triangles-dot line), Ag-CB (cross-dash line) and Ag-G (circles-solid line).
Reconstitution of the enzymatic activity of the PLP-
dependent enzyme BsrV immobilized by multipoint covalent
attachment on activated glyoxyl-agarose supports:
Optimization of the BsrV immobilization
(
Figure S1). Therefore, the preservation of the Schiff base
established between cofactor and enzyme during the
immobilization process is essential to keep the enzymatic activity.
As an alternative to Ag-G, we investigated other covalent
immobilization methods that do not require a reduction step and
In PLP-dependent enzymes, the addition of NaBH
enzymatic activity and absorption spectrum, and it is the same for
4
affects both
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201800359
FULL PAPER
BsrV. Surprisingly, the incubation with PLP after the reduction
step led to the recovery of all BsrV’s expressed activity (Table 2).
The immobilization of BsrV on agarose beads activated with
glyoxyl groups at alkaline pH orients the enzyme through the
surface region containing the highest number of Lys residues on
the support surface.[23a] Under these conditions, glyoxyl groups
are very stable and long incubation times promote a higher
number of enzyme-support bonds. This stronger covalent
attachment may reduce the flexibility of the immobilized enzyme
molecules resulting in a greater stabilization against small
conformational changes caused by heat, organic solvents,
mentioned above, the addition of exogenous PLP before
reduction does not lead to an increase of the catalytic activity of
BsrV, supporting this assumption. However, when the enzyme is
incubated with PLP after reduction, it could be hypothesized that
the empty active center could suffer a conformational change
leading to an open form allowing the binding of an exogenous PLP.
Molecular dynamics simulations should be performed to support
this hypothesis. As a result, the activity of the immobilized
racemase would be quantitatively recovered with only one
operative active center. When BsrV immobilized on Ag-G was
subjected to a second reduction with borohydride and incubation
in the presence of PLP, the enzyme did not have the ability to
recover its catalytic activity again. This fact could reflect that each
active site of BsrV irreversibly bound a PLP molecule, which
results in the blockage of the enzyme active centers. Similar
mechanisms have been described for some oligomeric PLP-
dependent enzymes such as glutamate-1-semialdehyde
aminomutase,[32] aspartate aminotransferase,[26] and glycine
decarboxylase.[33] The binding of a ligand to one subunit in, for
example, P-protein and aminomutase, cause a conformational
change that influences the binding affinity of the second subunit
for the ligand in a positive or negative way (positive or negative
cooperativity, respectively). This asymmetry may provide a
mechanism where the open/close state of the monomeric
catalytic centers alternates within the dimeric enzyme. Therefore,
this mechanism would prevent the enzyme from being presented
in a closed conformation in both monomers at the same time.
extreme pH values or denaturing agents, considerably increasing
enzyme stability.[23a,
27]
In this line, we screened different
immobilization conditions (temperature and time) to improve the
expressed activity and enzyme stability. As shown in Table 2, the
highest expressed activity was obtained with an incubation
enzyme-support at 4 ºC for 6 h. It is noteworthy that there is a
direct correlation between immobilization time and the loss of
(
expressed) activity. In other words, a longer incubation time
increases the number of enzyme-support interactions leading to
a decrease in expressed activity. One possible explanation could
be that BsrV requires a certain grade of flexibility and the
multipoint covalent attachment might rigidify in excess the
enzyme structure hindering the conformational changes that
occur during catalysis.[30]
4
As had been noted previously, the reduction with NaBH of the
BsrV would form a stable covalent bond between the C4′ of the
PLP and the ε-amino group of the K74 residue of the enzyme
active center. This adduct formed could not be released from the
enzyme, even under denaturing conditions,[31] resulting in the loss
of enzymatic activity. To investigate the mechanism by which
4
BsrV inactivated by NaBH recovered its catalytic activity upon
addition of exogenous PLP, we performed fluorescence and
enzymatic assays (Figure 2). A pre-incubation of BsrV (soluble or
immobilized on Ag-CB) with an excess of PLP, (without any
reduction step), did not lead to an increment in the enzyme activity
(
data not shown). In contrast, after the first reduction with
borohydride followed by incubation with PLP, BsrV was able to
recover its initial activity. Consistently, the immobilized biocatalyst
reduced once and subsequently incubated with PLP showed an
increase of the fluorescence at 400 nm. However, upon a second
reduction, it was impossible to recover the enzyme activity (or the
decrease of the fluorescence around 400 nm) after a new
incubation with exogenous PLP (Figure 2).
Figure 2. Fluorescence emission spectra upon excitation at 325 nm and
expressed activity of the BsrV immobilized on Ag-G after different treatments.
Reduced (solid line, A), reduced and incubated with PLP (dash line, B), reduced
twice and incubated once with PLP (dot line, C) and reduced twice and
incubated twice with PLP (dash-dot line, D). All conjugates had the same protein
concentration (9 mg BsrV/g Ag-G). Relative fluorescence was calculated:
BsrV is a globular homodimer with one catalytic active site per
monomer preceded by a wide entry channel (Figure S2).[20a] We
hypothesize that the dimeric nature of BsrV plays a major role in
the recovery of the enzymatic activity after the reduction step and
subsequent incubation with PLP. The BsrV dimer may present
alternatively one functional catalytic site filled with a PLP molecule
while the other site is PLP-free and inactive. During the reduction
step, the active center that binds PLP would become irreversibly
inactive because of the formation of a covalent attachment
푅퐹푈푥푥푥
푅퐹푈350
푅퐹푈푥푥푥
푅퐹푈350
퐹푛 (푆) =
퐹푛 (퐵) =
푛푅퐹 = 퐹푛 (푆) − 퐹푛 (퐵)
The relative enzymatic activity (inserted table) is referred to the suspension
activity of each conjugate after the first reduction with borohydride. PLP:
incubation with 8 mM PLP for 24 hours at pH 7.0. Red: reduction step with
4
1mg/mL NaBH for 30 min at pH 10.0.
4
between the PLP and the active site Lys. NaBH do not react in
the second active center since the PLP is not attached to the Lys
residue perhaps reflecting a reduced accessibility of the cofactor
due to the conformation that BsrV presents at that moment. As
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201800359
FULL PAPER
Table 2. Reconstitution of the BsrV activity and optimization of the immobilization of BsrV on Ag-G.
Expressed activity
Expressed
activity after
reduction (%)
6.6 ± 1
3 ± 2
Recovered activity
after incubation with
Incubation
time (h)
Incubation
temperature (ºC)
Immobilization yield
(%)
Biocatalyst
before reduction
%)[
b]
[b]
PLP (%)
[c][d]
(
Ag-CB
Ag-G 1h RT
Ag-G 3h RT
15
1
3
6
16
4
25
25
4
97.2 ± 0.5
95 ± 2
93 ± 2
96 ± 4
95 ± 4
76 ± 1
55 ± 2
50 ± 4
64 ± 5
49 ± 3
74 ± 2
57 ± 2
50 ± 3
65 ± 4
48 ± 2
[
[
a]
a]
[
2 ± 2
4 ± 2
4 ± 1
a]
[
Ag-G 6h 4 ºC
a]
Ag-G 16h 4 ºC
4
[
a] Agarose 6BCL was activated with glyoxyl groups.[b] Immobilization yield (%) = (activity of the solution offered for immobilization – activity in the supernatant at
the end of the immobilization process)/ (activity of the solution offered for immobilization) x 100. [c] Expressed activity (%) = (activity in immobilized preparation) /
activity of the solution offered for immobilization – activity in the supernatant at the end of the immobilization process) x 100. [d] Incubation was performed in the
(
presence of 8 mM of PLP for 24 hours at pH 7.0. After incubation with PLP, the conjugate was vacuum filtered and washed with 25 mM sodium phosphate buffer
pH 7.0. Immobilization times varied from 1 to 6 h and immobilization temperatures were 25 ºC and 4 ºC. The proportion of support/aqueous medium was 1/10 (w/v).
All data are the mean value of three separate experiments where the error value was never higher to 5%.
Thermal stability of the optimal immobilized BsrV
biocatalysts
The thermal stability of the optimum immobilized biocatalysts was
compared at pH 7.0 and 70-80 ºC. Although at 70 ºC all
conjugates were stable for at least for 2 hours (data not shown),
in Figure 3 (80 ºC) we readily appreciated that the incubation
enzyme-support at 4 ºC for 6 hours produces the highest
stabilization (half-life time was 3 hours at 80 ºC). In contrast,
longer incubation times at 4 ºC led to less stable immobilized
biocatalyst (half-life time was 1 hour).
The half-life times of the soluble BsrV and immobilized BsrV
on Ag-CB could not be calculated at 80 ºC and pH 7.0 due to their
extremely rapid inactivation. Once we set the optimal conditions
to obtain the best immobilized BsrV biocatalyst, we compared the
Figure 3. Thermal inactivation courses at 80 ºC and pH 7.5. BsrV immobilized
immobilized BsrV on Ag-G 6h against the BsrV covalently
immobilized biocatalyst (Ag-CB) under mild conditions and the
soluble BsrV to accurately calculate their half-life times and
stabilization factors. For this purpose, we tested the thermal
stability of these preparations at 50 ºC (Table S1). BsrV
immobilized on Ag-G under optimal conditions was approximately
for 1 hour at 25 ºC (circles-solid line), 3 hours at 25 ºC (diamonds-dash line), 6
hours at 4 ºC (squares-dot line) and 24 hours at 4 ºC (triangles-dash-dot line).
All preparations were immobilized on Ag-G support at pH 10. The proportion of
support/aqueous medium was 1/10 (w/v).
α-keto acids production by the heterogeneous multienzyme
system
3,300-fold more stable than the soluble enzyme (half-life time was
nearly 331 h). Even, the covalent immobilization of BsrV under
mild conditions on Ag-CB (immobilization at pH 7.0 by the most
reactive amine groups) was 67-fold more stable than the soluble
form. This stabilization could be due to the dimeric nature of the
enzyme (containing two N-termini). Thus, a two-point covalent
To use BsrV and DAAO for α-keto acids production, the substrate
compatibility between both enzymes was evaluated (Table S3).
On one side, BsrV reversibly racemizes ten of the natural chiral
amino acids knowns, including both non-β-branching aliphatic
amino acids (Ala, Leu, Met, Ser, Cys, Gln, and Asn) and positively
charged amino acids (His, Lys and Arg). BsrV was not able to
utilize aromatic (Tyr, Trp and Phe) and negatively charged (Asp
and Glu) amino acids. Moreover, BsrV enzyme displayed minimal
activity towards β-branched aliphatic amino acids (Val, Thr and
Ile). In addition, BsrV catalyzed the racemization of several non-
proteinogenic amino acids such as norleucine, ornithine,
diaminobutyric acid, aminobutyric acid, N-acetyl lysine methyl
ester and homoserine.[20a] On the other side, DAAO oxidizes a
broad spectrum of amino acids except Cys (Table S3). However,
Asp, Glu, Ser and Pro are poor substrates for DAAO.
attachment between enzyme and support could occur, resulting
in a more stable biocatalyst.[23a]
Otherwise, the preservation of the enzyme quaternary
structure is critical to avoid the dissociation of its enzyme subunits
to retain its catalytic activity. Thus, we analyzed if all the enzyme
subunits were involved in the immobilization process by SDS-
PAGE as described previously.[34] In this way, the supernatants
obtained after boiling the immobilized preparations in the
presence of SDS and mercaptoethanol were analyzed by this
technique. Under these incubation conditions, only the non-
covalently attached subunits would be released from the support
to the supernatant. The multipoint covalent immobilization of BsrV
onto Ag-G support under alkaline conditions achieved to
immobilize the two enzyme subunits to the support (Figure S3).
As first approach, we investigated the production of α-
ketocaproate by the immobilized multienzyme system in the
presence and absence of the BsrV from a racemic mixture of the
non-proteinogenic amino acid D,L-norleucine. Figure 4 shows
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201800359
FULL PAPER
that the maximum consumption of D,L-norleucine by the bi-
enzymatic system in which DAAO and CAT were co-immobilized
on the same carrier (Ag-Glutaraldehyde) without adding BsrV was
completed in just 5 h with a production rate of 12.52 mg of
product/h. These results are in agreement with the differential
specific activity of BsrV towards different amino acid.[20a] The initial
phase of the biotransformations was fast, with a 90% conversion
of the racemic mixtures of methionine, leucine and norleucine in
2, 6 and 7 h of reaction, respectively (Figure 5). Taking these
results into account, it would be necessary to optimize the
proportions of enzymes of the biosystem for each
biotransformation to obtain the highest production of α-keto acids.
50%. After 6 h of reaction, the amount of D,L-norleucine
consumed did not vary. This result was due to DAAO which can
only transform D-norleucine into its corresponding α-keto acid,
remaining unchanged the L-norleucine irrespective of the reaction
time. BsrV can reversibly racemize norleucine. Thus, by coupling
BsrV to the bi-enzymatic system DAAO:CAT one would in
principle enrich the reaction mixture in D-norleucine, which is a
substrate for the DAAO. Therefore, this tri-enzyme system would
theoretically produce pure α-KC, consuming 100% of the racemic
norleucine. As shown in Figure 4, the addition of the optimal
immobilized BsrV to the reaction media led to the consumption of
the remaining L-norleucine resulting in a conversion of 96% of
conversion after 24 h. Therefore, the introduction of
racemization step in the process may allow yields approaching
00% and leading to pure α-keto acids.
a
1
Figure 4. Bioconversion of D,L-norleucine to α-KC. D,L-Norleucine
consumption in two different reactions: (i) DAAO-CAT, the reaction started
adding 100 mg of co-immobilized DAAO:CAT on Ag-Glutaraldehyde; and (ii)
DAAO-CAT + BsrV, the reaction started adding 100 mg of co-immobilized
DAAO:CAT on Ag-Glutaraldehyde and 100 mg of immobilized BsrV on Ag-G.
All the reactions were carried out in 10 mL of 25 mM sodium phosphate pH 7.0
containing 10 mM of D,L-norleucine and 25 ºC for 24 hours.
To demonstrate the applicability of the immobilized
multienzyme cell-free system for the production of different α-keto
acids, we performed in vitro biotransformations of the racemic
mixtures of leucine, methionine, in addition to norleucine, to their
corresponding α-keto acids. Figure 5 shows the full-time courses
of the α-KC (A), α-KIC (B) and α-KMB (C). The coupling of the
three immobilized enzymes in an enzymatic cascade reaction led
to conversions above 99% in the three studied biotransformations
(
Figure 5 and Figure S4).
The enzymatic conversions of D,L-norleucine and D,L-leucine
to their corresponding α-keto acids were >99% after 24 h of
reaction. In the biotransformation of D,L-leucine to α-KIC, it was
necessary to add twice more immobilized BsrV (Table S2)
whereas the conversion of 10 mM D,L-methionine to α-KMB was
Figure 5. Reaction courses of D,L-norleucine (A), D,L-leucine (B) and D,L-
methionine (C) to their corresponding α-keto acids. The biotransformations were
performed in 10 mL of 25 mM sodium phosphate pH 7.0, 10 mM of the racemic
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201800359
FULL PAPER
mixture of amino acids, 200 mg of co-immobilized DAAO:CAT and 200 mg of
Ag-G-BsrV (400 mg for D,L-Leucine) at pH 7.0, 25 ºC for 24 hours. Symbols:
D,L-amino acid (squares-solid line) and α- keto acid (circles-dash line).
uncommon immobilization carrier for PLP-dependent enzymes),
recovering completely its expressed activity thanks to the
exogenous incubation with PLP after the reduction step. The
immobilized BsrV biocatalyst was 3,300-fold more stable than its
soluble form (at 50 ºC and pH 7.5). Moreover, this immobilized
BsrV biocatalyst was successfully coupled to co-immobilized
DAAO and CAT. The applicability of the immobilized multienzyme
system was demonstrated by the production of α-KC, α-KIC, and
α-KMB starting from racemic mixtures of norleucine, leucine and
methionine, respectively. This biosystem offer two remarkable
advantages: i) the catalytic promiscuity of BsrV and DAAO
permits the use of a single system to transform a wide range of
substrates and, ii) DKR facilitates conversions near to 100%.
Therefore, we have developed an efficient process for the
biotechnological production of α-keto acids that could be an
alternative to the traditional chemical synthesis approach. This
immobilized enzyme system enables the greatly simplification of
the reaction control and reactor design, as well as the final product
separation and its application in continuous flow packed-bed
reactors for an intensification of this bioprocess.
In recent years, metabolic engineering approaches have
shown remarkable potential for α-keto acids production. Although,
strains are only modified by overexpression or knockout of
operons or particular genes, and in many cases without achieving
an efficient regulation of the metabolic pathways.[9, 12, 35] The
production of α-keto acids by whole-cells of Corynebacterium
glutamicum, Rhodococcus opacus DSM 43250 and E. coli
expressing an engineered L-amino acid oxidase (LAAO) from
Proteous vulgaris has been previously reported.[12, 35b, 36] α-KIC
productivity (Yp/s=gproduct/gsubstrate) of our immobilized system was
higher (approximately Yp/s=0.99) than those previously
described for whole-cells of Corynebacterium (Yp/s=0.014 and
0
.17) and Rhodococcus (Yp/s=0.5).[36b, 36c] α-KMB Yp/s was also
higher (approximately Yp/s=0.99) than previously reported for its
production by E. coli cells expressing an engineered LAAO from
P. vulgaris (Yp/s=0.91).[12] The effect of immobilization of E. coli
whole-cells expression LAAO from P. vulgaris with Ca2+ alginate
was also investigated by G.S. Hossain et al., reporting that after Experimental Section
4
reaction cycles the biocatalyst only kept 50% of its catalytic
activity.[ Therefore, microbial production requires the genetic
manipulation of the metabolic routes of a microorganism involved
in the synthesis of the different α-keto acids. Moreover, it requires
12]
Chemicals
Pyridoxal-5-phosphate (PLP), o-phenylenediamine (OPD), amino acids,
peroxidase from horseradish (HRP) and catalase from bovine liver were
purchased from de Sigma-Aldrich Co. (St. Louis, IL USA). D-amino acid
oxidase from Trigonopsis Variabilis was kindly donated by Recordati SRL.
Ni-NTA agarose was acquired from Qiagen (Hilden, Germany). Complete
Protease Inhibitor cocktail tablets were from Roche (Basel, Switzerland).
Agarose beads 6BCL were purchased from Agarose Beads Technology
the use of expensive and complex nutrients, and in many cases,
it can generate by-products.[8] Another important limitation for the
use of LAAOs is the lack of a suitable heterologous expression
system for enabling the production of large amounts of active
37]
recombinant LAAOs with
a
broad substrate spectrum.[
Considering all the above, this method requires more
investigation to be implemented on an industrial scale.
(Madrid, Spain). Cyanogen bromide-activated-Sepharose 4BCL was
purchased from de General Electric Healthcare (Uppsala, Sweden).
Protein concentrations were determined with BCA Protein Assay Kit from
Pierce (Rockford, IL). Solvents were acdquired from Scharlab (Barcelona,
Spain). Buffers and other reactives were from Sigma-Aldrich Co. (St. Louis,
IL USA).
Finally, the operational stability of the multienzyme system
was tested in the production of α-KC from a racemic mixture of
norleucine. As is shown in Figure S5, the catalytic activity of the
immobilized multienzyme system was kept for at least 8 reaction
cycles with a conversion over 96% in each cycle. It is noteworthy
to mention that it was not necessary to add exogenous PLP to the
multienzyme system during or after each reaction cycle since it
was not detected a reduction in the reaction rate for at least eight
reaction cycles. This fact indicates that the cofactor is strongly
bonded to the K74 residue inside the active center of the BsrV as
we demonstrated in this work. In most of the reactions performed
by PLP-dependent enzymes, the addition of exogenous PLP (0.1-
Production of broad-spectrum racemase from Vibrio cholerae (BsrV)
The V. cholerae gene encoding BsrV was cloned in pET28b for expression
in E. coli BL21(DE3) cells.[20a] Expression was induced (at OD600=0.4) with
1
1
mM IPTG for 3 h. Cell pellet was resuspended in 50 mM Tris HCl pH 7.5,
50 mM NaCl, and Complete Protease Inhibitor Cocktail Tablets, and
lysed with 3 passes through a French press. The protein was purified from
cleared lysates (30 min, 50,000 rpm) on Ni-NTA agarose columns, and
eluted with a discontinuous imidazole gradient. Finally, the enzyme was
dialyzed twice during a 2 hours period and once more during a 16 hours
period against a 25 mM sodium phosphate solution at pH 7.5.
1
00 mM) was required to stabilize the enzyme or regenerate the
[25a, 38]
cofactor as described in previous reports.
Enzymatic assays
Conclusions
Enzymatic assays were carried out on UVmini-1240 spectrophotometer
from Shimadzu (Kyoto, Japan). Racemase activity was checked
spectrophotometrically using L-arginine as substrate. We followed the
increase in the absorbance at 450 nm promoted by coupling the
racemization and the oxidative deamination of the amino acid by DAAO
with the reaction between the hydrogen peroxide and o-phenylenediamine
This work aimed to design and develop a biocatalytic process for
the production of α-keto acids from racemic amino acids or L-
amino acids by a supported cell-free platform based on a tri-
enzyme system composed of BsrV, DAAO, and CAT. Here, we
have been able to efficiently immobilize the BsrV on Ag-G (an
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201800359
FULL PAPER
catalyzed by peroxidase that produces a colored compound (Scheme S2).
The reaction mixture was 1.5 mL of 10 mM L-Arginine, 0.5 mL of 0.1
mg/mL OPD, 100 µL of 0.05 mg/mL of HRP and 50 µL of 0.042 mg/mL
DAAO all of them in 25 mM phosphate buffer pH 7.5. This reaction solution
was preincubated at 25 ºC. The reaction was initialized by adding a
maximum of 0.1 U of BsrV. The assays were performed at 25 ºC with
magnetic stirring in a thermostatized cuvette. One racemase unit is defined
as the amount of enzyme that consumed 1 µmol of L-arginine per minute
under the previously described conditions.
The activities of the other enzymes were measured according to modified
literature procedures for DAAO and CAT.[11, 39] DAAO activity was followed
by the increase in the absorbance at 420 nm promoted by coupling the
oxidative deamination of D-amino acids catalyzed by DAAO with the
reaction between the hydrogen peroxide and o-phenylendiamine
catalyzed by HRP that produces a coloured compound. The reaction
mixture was: 1.5 mL of 0.1 M sodium phosphate buffer, 10 mM D-amino
acid pH 7.0, 0.5 mL of 1.85 mM o-phenylendiamine in distilled water and
by vacuum filtration, washed with distilled water and added to 10 mL of
phosphate pH 7.5 buffer containing 20 U of BsrV. After 15 min at 4 ºC, the
conjugate was vacuum filtered and washed with phosphate buffer. The
remaining CNBr groups were blocked with 1 M Ethanolamine pH 8.5
during 2 h.
Immobilization on glutaraldehyde support (Ag-Glutaraldehyde)
One g of support was added to 10 mL of 10 mM sodium phosphate pH 7.5
containing 20 U of BsrV. Immobilization time was 1 h at 25 ºC. The
conjugate was vacuum filtered and washed with 10 mM sodium phosphate
buffer pH 7.5.
Immobilization on monoaminoethyl-N-aminoethyl support (Ag-
MANAE)
One g of support was added to 10 mL of 10 mM sodium phosphate pH 7.5
containing 20 U of BsrV. Immobilization time was 1 h at 25 ºC. The
conjugate was vacuum filtered and washed with 10 mM sodium phosphate
buffer pH 7.5.
0
.1 mL of 2 mg/mL HRP in 0.1 m sodium phosphate pH 7.0. This reaction
solution was preincubated at 25°C. The reaction was initialized by adding
a maximum of 0.1 DAAO units. One DAAO unit is defined as the amount
of enzyme that produced 1 μmol of H O (or oxidize 1 μmol of D-amino
2 2
acid) per minute under the previously described conditions. CAT activity
was determined spectrophotometrically by monitoring the decomposition
Immobilization on epoxide support (Ag-Epoxide)
2 2
of H O , via measurement of the change in the absorbance at 240 nm. A
2
.9 mL portion of H O
2 2
, 0.14% (w/w) in 50 mM sodium phosphate buffer
One g of support was added to 10 mL of 10 mM sodium phosphate pH 7.5
containing 20 U of BsrV. Immobilization time was 16 h at 25 ºC. The
conjugate was vacuum filtered and washed with 10 mM sodium phosphate
buffer pH 7.5.
pH 7.0, was incubated with 0.2 mL of enzyme solution. All measurements
were carried out at 25 °C. One CAT unit was defined as the amount of
enzyme that decomposes 1 μmol of hydrogen peroxide per minute under
the previously described conditions.
Co-immobilization of D-amino acid oxidase (DAAO) and catalase
Preparation of different agarose supports
(CAT)
Glyoxyl agarose (Ag-G) was prepared as described previously.[29]
Monoaminoethyl-N-aminoethyl agarose (Ag-MANAE) was prepared as
described previously.[40] Agarose beads activated with Glutaraldehyde
A co-immobilized preparation of DAAO and CAT were prepared. The best
conjugate was prepared as previously described,[43] containing 10 mg of
CAT and 1 mg of DAAO per gram of Ag-Glutaraldehyde support.
(
Ag-Glutaraldehyde) was prepared as described previously.[41] Epoxide
agarose was prepared as described previously.[42]
SDS-PAGE analysis
Enzyme immobilization
Sodium dodecyl sulphate polyacrylamide gel electrophoresis was
performed as previously described. The BsrV conjugates were boiled in
[
44]
All the immobilizations were carried out in a Stuart Scientific roller mixer
SRT6 at 33 rpm. The catalytic activity of the blank, supernatant and
suspension were determined during the process. Also, the final catalytic
activity was measured after immobilization. All glyoxyl preparations were
washed with 25 mM of sodium phosphate buffer pH 7.5 after the reduction
with sodium borohydride and incubated for 16 h in 25 mM phosphate buffer
pH 7.5 containing 8 mM of PLP. The excess of cofactor was removed with
several washes with 25 mM of sodium phosphate buffer pH 7.5.
disruption buffer containing mercaptoethanol and SDS, thus releasing
non-covalently bound protein from the support.[23b] The different
immobilized preparations contains the same amount of protein (2.1 mg
protein/mL agarose) and the soluble protein concentration was 0.3 mg
protein/mL buffer. Samples were analyzed using 12% polyacrylamide gels,
and proteins were detected by Coomassie blue staining.
Study of the thermal stability
Immobilization on glyoxyl support (Ag-G)
The thermal inactivation assays were performed at 50 ºC, 70 ºC and 80 ºC
on 25 mM sodium phosphate buffer at pH 7.5. The proportion of
support/aqueous medium was 1/10 (w/v). Samples were periodically
removed and the activity measured. Half-life times were calculated as
previously described.[45]
One g of Ag-G support was added to 10 mL of 0.1 M Bicarbonate buffer
pH 10.0 containing 20 U of BsrV. Immobilization times varied from 1 to 6
h and, also, the two different temperatures were tested, 4 ºC and 25 ºC.
The conjugates were reduced during 30 minutes with 10 mg of sodium
borohydride after the incubations at pH 10.0. Finally, the conjugates were
vacuum filtered and washed to remove the excess of borohydride with 25
mM sodium phosphate buffer pH 7.0.
Fluorescence of PLP on different immobilized BsrV biocatalysts
Fluorescence spectrums were done to different immobilized BsrV
preparations (specially prepared for the occasion with 9 mg of protein per
gram of support) as previously described;[46] the measures were done with
Immobilization on cyanogen bromide support (Ag-CB)
2
00 µL of a suspension of 1/5 (w/v) of different preparations. Fluorometric
First, 0.3 g of dry Ag-CB support (approximately 1 g of wet support) was
hydrated on 10 mL of distilled water pH 2.0-3.0. Wet support was collected
measurements were carried out on a Varioskan Flash Multimode Reader
fluorimeter from Thermo-Fisher Scientific (Waltham, MA USA). The λ
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201800359
FULL PAPER
excitation was 325 nm and the emission spectra ranged from 350 nm to
587; c) G. Kuhlman, J. A. Roth, P. J. Flakoll, M. J. Vandehaar,
S. Nissen, J. Nutrit. 1988, 118, 1564-1569.
a) G. Quash, G. Fournet, Curr. Med. Chem. 2009, 16, 3686-
500 nm.
[
5]
3
700; b) A. L. Subhi, P. Diegelman, C. W. Porter, B. Tang, Z. J.
Lu, G. D. Markham, W. D. Kruger, J. Biol. Chem. 2003, 278,
9868-49873.
Production of α-keto acids
4
[
[
6]
7]
R. N. Dilger, C. Kobler, C. Weckbecker, D. Hoehler, D. H. Baker,
J. Nutrit. 2007, 137, 1868-1873.
A. J. L. Cooper, J. Z. Ginos, A. Meister, Chem. Rev. 1983, 83,
Reaction mixtures were composed by 10 mM of D,L-amino acid, 200 mg
of co-immobilized DAAO:CAT (1 mg DAAO/g support and 10 mg CAT/g
support), 200 (for D,L-norleucine and DL-methionine) or 400 mg (for D,L-
leucine conversion) mg of Ag-G-BsrV (loaded with 19 mg BsrV/g support),
in a 25 mM phosphate buffered solution at pH 7.0. Ten milliliters of reaction
mixture were placed in a 20 mL cap bottle (1:1 relation of air:reaction
mixture). The reaction was started by the addition of immobilized
biocatalysts (total volume of biocatalysts was always maintained lower
than 8.5% of the reaction volume) and maintained under gentle agitation
at 25 ºC. The reusability of the immobilized multi-enzyme system were
performed with 10 mM of D,L-norleucine, 250 mg of co-immobilized
DAAO:CAT (1 mg DAAO/g and 10 mg CAT/g) and 250 mg of Ag-G-BsrV
321-358.
[8]
Y. Song, J. Li, H.-d. Shin, L. Liu, G. Du, J. Chen, Bioresour.
Technol. 2016, 219, 716-724.
V. Bückle-Vallant, F. S. Krause, S. Messerschmidt, B. J.
Eikmanns, Appl. Microbiol. Biotechnol. 2014, 98, 297-311.
a) L. Pollegioni, G. Molla, Trends Biotechnol. 2011, 29, 276-
[
[
9]
10]
283; b) M. García-García, I. Martínez-Martínez, Á. Sánchez-
Ferrer, F. García-Carmona, Biotechnol. Prog. 2008, 24, 187-
191.
R. Fernández-Lafuente, V. Rodriguez, J. M. Guisán, Enzyme
Microb. Technol. 1998, 23, 28-33.
G. S. Hossain, J. Li, H.-d. Shin, G. Du, M. Wang, L. Liu, J. Chen,
PLoS One 2014, 9, e114291.
a) W.-D. Fessner, New Biotechnology 2015, 32, 658-664; b) D.
Tessaro, L. Pollegioni, L. Piubelli, P. D’Arrigo, S. Servi, ACS
Catal. 2015, 5, 1604-1608.
a) C. Schmidt-Dannert, F. Lopez-Gallego, Microb. Biotechnol.
2016, 9, 601-609; b) Q. Ji, B. Wang, J. Tan, L. Zhu, L. Li,
Process Biochem. 2016, 51, 1193-1203; c) F. López-Gallego,
E. Jackson, L. Betancor, Chemistry – A European Journal, n/a-
n/a.
[11]
[
[
12]
13]
(19 mg/g), In the same conditions previously described for 3 h of reaction
time. The conjugates were washed with sodium phosphate buffer between
cycles. Reaction course was monitored by periodically withdrawing and
filtering samples which were adequately diluted to be further analyzed by
High-Performance Liquid Chromatography (HPLC) on a Spectra system
with a UV100 detector from Thermo-Fisher Scientific (Waltham, MA USA)
using an Ultrabase C18 5µm 150x4.6mm column from Análisis Vínicos
[14]
(
Tomelloso, Spain). Elution of compounds was carried out at an isocratic
flow of 0.75 mL/min of a mobile phase of 2.5 mM phosphate at pH 2.5 in
5:5 water/methanol (95:5). Detection of compounds was done at 210 nm.
[
15]
a) Z. Liu, J. Zhang, X. Chen, P. G. Wang, ChemBioChem 2002,
9
3, 348-355; b) J. Rocha-Martín, B. d. l. Rivas, R. Muñoz, J. M.
Retention times were 4 min for D,L-leucine, 26 min for α-KIC, 10 min for
D,L-norleucine, 58 min for α-KC, 4.1 min for D,L-methionine and 19.7 min
for α-KMBA. The conversion degree was estimated after a relationship of
the peak’s area and a calibration curve in the same elution conditions.
Initial velocities (production rates) were determined plotting the α-keto acid
amount as a function of time.
Guisán, F. López-Gallego, ChemCatChem 2012, 4, 1279-1288;
c) W. Kang, J. Liu, J. Wang, Y. Nie, Z. Guo, J. Xia, Bioconj.
Chem. 2014, 25, 1387-1394; d) F. Zhao, H. Li, Y. Jiang, X.
Wang, X. Mu, Green Chem. 2014, 16, 2558-2565; e) J. Rocha-
Martin, S. Velasco-Lozano, J. M. Guisan, F. Lopez-Gallego,
Green Chem. 2014, 16, 303-311; f) J. Rocha-Martin, A. Acosta,
J. M. Guisan, F. López-Gallego, ChemCatChem 2015, 7, 1939-
1947; g) S. Velasco-Lozano, J. Rocha-Martin, E. Favela-Torres,
J. Calvo, J. Berenguer, J. M. Guisán, F. López-Gallego,
Biochem. Eng. J. 2016, 112, 136-142; h) S. Velasco-Lozano, E.
S. da Silva, J. Llop, F. López-Gallego, ChemBioChem, n/a-n/a.
E. García-Urdiales, I. Alfonso, V. Gotor, Chem. Rev. 2005, 105,
Acknowledgements
[
[
[
[
16]
17]
18]
19]
313-354.
Javier Rocha-Martin is grateful to the Juan de la Cierva fellowship
M. Ahmed, T. Kelly, A. Ghanem, Tetrahedron 2012, 68, 6781-
6802.
C. Femmer, M. Bechtold, T. M. Roberts, S. Panke, Appl.
Microbiol. Biotechnol. 2016, 100, 7423-7436.
(
IJCI-2014-19260) funded by the Spanish Ministry of Economy,
Industry and Competitiveness. Research in the Cava lab is funded
by The Knut and Alice Wallenberg Foundation (KAW), The
Laboratory of Molecular Infection Medicine Sweden (MIMS), the
Swedish Research Council and the Kempe Foundation.
T. Yoshimura, N. Esaki, J. Biosci. Bioeng. 2003, 96, 103-109.
a) A. Espaillat, C. Carrasco-Lopez, N. Bernardo-Garcia, N.
Pietrosemoli, L. H. Otero, L. Alvarez, M. A. de Pedro, F. Pazos,
B. M. Davis, M. K. Waldor, J. A. Hermoso, F. Cava, Acta
Crystallogr. D Biol. Crystallogr. 2014, 70, 79-90; b) H. Lam, D.
C. Oh, F. Cava, C. N. Takacs, J. Clardy, M. A. De Pedro, M. K.
Waldor, Science 2009, 325, 1552-1555; c) D. Matsui, T. Oikawa,
N. Arakawa, S. Osumi, F. Lausberg, N. Stäbler, R. Freudl, L.
Eggeling, Appl. Microbiol. Biotechnol. 2009, 83, 1045-1054; d)
A. D. Radkov, L. A. Moe, J. Bacteriol. 2013, 195, 5016-5024.
a) C. Mateo, J. M. Palomo, G. Fernandez-Lorente, J. M. Guisan,
R. Fernandez-Lafuente, Enzyme Microb. Technol. 2007, 40,
1451-1463; b) B. Brena, P. González-Pombo, F. Batista-Viera,
in Immobilization of Enzymes and Cells: Third Edition (Ed.: J.
M. Guisan), Humana Press, Totowa, NJ, 2013, pp. 15-31.
a) C. A. Godoy, B. d. l. Rivas, V. Grazú, T. Montes, J. M. Guisán,
F. López-Gallego, Biomacromolecules 2011, 12, 1800-1809; b)
K. Hernandez, R. Fernandez-Lafuente, Enzyme Microb.
Technol. 2011, 48, 107-122.
[20]
Keywords: biocatalysis • immobilization • α-keto acids •
dynamic kinetic resolution • PLP-dependent enzymes
[
1]
a) S. Atsumi, T. Hanai, J. C. Liao, Nature 2008, 451, 86-89; b)
J. V. N. Vara Prasad, J. A. Loo, F. E. Boyer, M. A. Stier, R. D.
Gogliotti, W. J. Turner, P. J. Harvey, M. R. Kramer, D. P. Mack,
J. D. Scholten, S. J. Gracheck, J. M. Domagala, Biorg. Med.
Chem. 1998, 6, 1707-1730; c) D. P. Pantaleone, A. M. Geller,
P. P. Taylor, J. Mol. Catal. B: Enzym. 2001, 11, 795-803; d) S.
Spaepen, W. Versées, D. Gocke, M. Pohl, J. Steyaert, J.
Vanderleyden, J. Bacteriol. 2007, 189, 7626-7633.
[21]
[22]
[23]
[
[
2]
3]
E. A. Porter, X. Wang, H.-S. Lee, B. Weisblum, S. H. Gellman,
Nature 2000, 404, 565.
a) S. Mou, J. Li, Z. Yu, Q. Wang, Z. Ni, J. Int. Med. Res. 2013,
a) G. Fernández-Lorente, F. Lopez-Gallego, J. M. Bolivar, J.
Rocha-Martin, S. Moreno-Perez, J. M. Guisán, Curr. Org. Chem.
2015, 19, 1719-1731; b) J. Rocha-Martin, A. Acosta, J.
Berenguer, J. M. Guisan, F. Lopez-Gallego, Bioresour. Technol.
2014, 170, 445-453.
4
1, 129-137; b) H. Nian, W. H. Bisson, W.-M. Dashwood, J. T.
Pinto, R. H. Dashwood, Carcinogenesis 2009, 30, 1416-1423.
a) W. J. Malaisse, A. Sener, M. Welsh, F. Malaisse-Lagae, C.
Hellerström, J. Christophe, Biochem. J. 1983, 210, 921-927; b)
B. N. Buford, B. J. Koch, Med. Sci. Sports Exerc. 2004, 36, 583-
[
4]
[24]
A. Mozzarelli, S. Bettati, Chemical Record 2006, 6, 275-287.
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201800359
FULL PAPER
[
25]
a) C. López, S. D. Ríos, J. López-Santín, G. Caminal, G. Álvaro,
Biochem. Eng. J. 2010, 49, 414-421; b) S. S. Yi, C. w. Lee, J.
Kim, D. Kyung, B. G. Kim, Y. S. Lee, Process Biochem. 2007,
[37]
[38]
G. S. Hossain, J. Li, H.-d. Shin, G. Du, L. Liu, J. Chen, Appl.
Microbiol. Biotechnol. 2014, 98, 1507-1515.
a) H. Mallin, M. Höhne, U. T. Bornscheuer, J. Biotechnol. 2014,
191, 32-37; b) H. Mallin, U. Menyes, T. Vorhaben, M. Höhne, U.
T. Bornscheuer, ChemCatChem 2013, 5, 588-593; c) I. Slabu,
J. L. Galman, C. Iglesias, N. J. Weise, R. C. Lloyd, N. J. Turner,
Catal. Today 2017; d) I. Slabu, J. L. Galman, N. J. Weise, R. C.
Lloyd, N. J. Turner, ChemCatChem 2016, 8, 1038-1042; e) T.
M. Lammens, D. De Biase, M. C. R. Franssen, E. L. Scott, J. P.
M. Sanders, Green Chem. 2009, 11, 1562-1567.
L. Betancor, A. Hidalgo, G. Fernández-Lorente, C. Mateo, R.
Fernández-Lafuente, J. M. Guisan, Biotechnol. Prog. 2003, 19,
763-767.
R. Fernandez-Lafuente, C. M. Rosell, V. Rodriguez, C. Santana,
G. Soler, A. Bastida, J. M. Guisán, Enzyme Microb. Technol.
1993, 15, 546-550.
L. Betancor, F. López-Gallego, A. Hidalgo, N. Alonso-Morales,
G. D.-O. C. Mateo, R. Fernández-Lafuente, J. M. Guisán,
Enzyme Microb. Technol. 2006, 39, 877-882.
C. Mateo, R. Torres, G. Fernández-Lorente, C. Ortiz, M.
Fuentes, A. Hidalgo, F. López-Gallego, O. Abian, J. M. Palomo,
L. Betancor, B. C. C. Pessela, J. M. Guisan, R. Fernández-
Lafuente, Biomacromolecules 2003, 4, 772-777.
42, 895-898; c) M.-C. Yen, W.-H. Hsu, S.-C. Lin, Process
Biochem. 2010, 45, 667-674.
S. W. Zito, M. Martinez-Carrion, J. Biol. Chem. 1980, 255,
[
[
26]
27]
8645-8649.
C. Mateo, J. M. Palomo, M. Fuentes, L. Betancor, V. Grazu, F.
López-Gallego, B. C. C. Pessela, A. Hidalgo, G. Fernández-
Lorente, R. Fernández-Lafuente, J. M. Guisán, Enzyme Microb.
Technol. 2006, 39, 274-280.
[39]
[40]
[41]
[42]
[
28]
V. I. Tishkov, S. V. Khoronenkova, Biochemistry (Moscow)
2005, 70, 40-54.
[
[
29]
30]
J. M. Guisán, Enzyme Microb. Technol. 1988, 10, 375-382.
a) F. Secundo, Chem. Soc. Rev. 2013, 42, 6250-6261; b) J.
Mukherjee, M. N. Gupta, Biotechnol. Rep. 2015, 6, 119-123; c)
B. Pioselli, S. Bettati, A. Mozzarelli, FEBS Lett. 2005, 579,
2197-2202.
[
[
31]
32]
D. Schiroli, L. Ronda, A. Peracchi, FEBS J. 2015, 282, 183-199.
a) M. Hennig, B. Grimm, R. Contestabile, R. A. John, J. N.
Jansonius, Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 4866-4871;
b) B. Campanini, S. Bettati, M. L. di Salvo, A. Mozzarelli, R.
Contestabile, BioMed Res. Int. 2013, 2013, 10.
[
[
[
33]
34]
35]
T. Nakai, N. Nakagawa, N. Maoka, R. Masui, S. Kuramitsu, N.
Kamiya, EMBO J. 2005, 24, 1523-1536.
A. H. Orrego, L. Trobo-Maseda, J. Rocha-Martin, J. M. Guisan,
Enzyme Microb. Technol. 2017, 105, 51-58.
a) G. S. Hossain, J. Li, H.-d. Shin, L. Liu, M. Wang, G. Du, J.
Chen, J. Biotechnol. 2014, 187, 71-77; b) Y. Song, J. Li, H.-d.
Shin, G. Du, L. Liu, J. Chen, Sci. Rep. 2015, 5, 12614.
a) P. Niu, X. Dong, Y. Wang, L. Liu, J. Biotechnol. 2014, 179,
[43]
[44]
[45]
[46]
R. Fernández-Lafuente, V. Rodriguez, J. M. Guisán, Enzyme
Microb. Technol. 1998, 23, 28-33.
A. Bastida, P. Sabuquillo, P. Armisen, R. Fernández-Lafuente,
J. Huguet, J. M. Guisán, Biotechnol. Bioeng. 1998, 58, 486-493.
O. Romero, J. M. Guisán, A. Illanes, L. Wilson, J. Mol. Catal. B:
Enzym. 2012, 74, 224-229.
E. Adams, Anal. Biochem. 1969, 31, 118-122.
[
36]
56-62; b) M. Vogt, S. Haas, T. Polen, J. van Ooyen, M. Bott,
Microb. Biotechnol. 2015, 8, 351-360; c) Y. Zhu, J. Li, L. Liu, G.
Du, J. Chen, Enzyme Microb. Technol. 2011, 49, 321-325.
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201800359
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
Layout 1:
FULL PAPER
Heterogeneous multienzyme system
to produce α-keto acids
Alejandro H. Orrego, Fernando López-
Gallego, Akbar Espaillat, Felipe Cava*
José M. Guisan* and Javier Rocha-
Martin*
Page No. – Page No.
One-step synthesis of alpha-keto
acids from racemic amino acids by a
versatile immobilized multienzyme
cell-free system
This article is protected by copyright. All rights reserved.