Evaluation of different glutaryl acylase mutants to improve the hydolysis of cephalosporin C. in the absence of hydrogen peroxide

Article abstract of DOI:10.1002/adsc.200700320

2-Oxoadipoyl-7-ACA is an intermediate in the conversion of cephalosporin C. (CPC) to 7-aminocephalosporanic acid (7-ACA) when using a new route involving D-amino acid oxidase, catalase and glutaryl acylase. A key point in the reaction design is to avoid the accumulation of hydrogen peroxide in the reaction medium as the yields of 7-ACA decrease in the presence of this compound due to its low stability. Looking for an enzyme with improved activity towards 2-oxoadipoyl-7-ACA, different mutants of glutaryl acylase from Pseudomonas SY-77 with an improved activity towards adipoyl-7-ACA were evaluated. The best results on 2-oxoadipoyl-7-ACA hydrolysis were found with the double mutant Y178F+F375H, which showed a K_cat increase of 6.5-fold and a K _m decrease of 3-fold compared to the wild-type (wt) enzyme. When this enzyme was tested in the tri-enzymatic system to convert CPC into 7-ACA, this mutant permitted us to reach more than an 80% yield of 7-ACA using a 3-fold mass excess compared to DAAO; while the wt enzyme gave only a 40% yield. Therefore, the application of this new mutant to the one-pot conversion of CPC to 7-ACA gives very good result in terms of efficiency, yield and rate of the process.

Full text of DOI:10.1002/adsc.200700320

UPDATE
DOI: 10.1002/adsc.200700320
Evaluation of Different Glutaryl Acylase Mutants to Improve the
Hydolysis of Cephalosporin C in the Absence of Hydrogen
Peroxide
a
a
b
b
Fernando López-Gallego, Lorena Betancor, Charles F. Sio, Carlos R. Reis,
b
a
b,
Pol Nadal Jimenez, Jose M. Guisan, Wim. J. Quax, *
and Roberto Fernandez-Lafuente *
a,
a
Departamento de Biocatalisis, ICP-CSIC, 28049 Madrid, Spain
Fax : (+34)-91-585-4760; E-mail: rfl@icp.csic.es
Department of Pharmaceutical Biology, University of Groningen, Groningen, The Netherlands
b
E-mail: w.j.quax@rug.nl
Received: June 29, 2007; Published online: January 4, 2008
Abstract: 2-Oxoadipoyl-7-ACA is an intermediate
in the conversion of cephalosporin C (CPC) to 7-
aminocephalosporanic acid (7-ACA) when using a
new route involving d-amino acid oxidase, catalase
and glutaryl acylase. A key point in the reaction
design is to avoid the accumulation of hydrogen
peroxide in the reaction medium as the yields of 7-
ACA decrease in the presence of this compound
due to its low stability. Looking for an enzyme with
improved activity towards 2-oxoadipoyl-7-ACA, dif-
ferent mutants of glutaryl acylase from Pseudomo-
nas SY-77 with an improved activity towards adipo-
yl-7-ACA were evaluated. The best results on 2-ox-
oadipoyl-7-ACA hydrolysis were found with the
double mutant Y178F+F375H, which showed a
cephaloporanic acid (7-ACA) or 7-aminodesacetoxy-
cephalosporanic acid (7-ADCA) are among the most
used antibiotics. It was estimated that approximately
1.95 million kg of 7-ADCA and 2.14 million kg of 7-
[
1]
ACA were produced in 2000.
Chemical deacylation of cephalosporin C (CPC),
[2]
a fermentation product, is the primary method used
industrially to produce 7-ACA. This method is eco-
nomically feasible using currently available chemical
processes. However, due to the large amounts of haz-
ardous chemicals used, there is a concern for environ-
mental and safety reasons. To overcome these prob-
lems, efforts are being made to develop an efficient,
entirely enzymatic process for the conversion of CPC
to 7-ACA.
[
3]
Kcat increase of 6.5-fold and a K decrease of 3-fold
After many efforts, the direct conversion of CPC
to 7-ACA by cephalosporin C acylase has been suc-
m
compared to the wild-type (wt) enzyme. When this
enzyme was tested in the tri-enzymatic system to
convert CPC into 7-ACA, this mutant permitted us
to reach more than an 80% yield of 7-ACA using a
[
4]
cessfully reported. However, nowadays, the most
widely used enzymatic route at the industrial level is
the conversion of CPC to 7-ACA in a two-pot system
with two different enzymes. Firstly, d-amino acid ox-
idase (DAAO) catalyzes the oxidative deamination of
CPC yielding 7-b-(5-carboxy-5-oxopentanamido)-
cephalosporanic acid (2-oxoadipoyl-7-ACA) and hy-
drogen peroxide. This latter product is able to cata-
lyse the decarboxylation of 2-oxoadipoyl-7-ACA pro-
ducing glutaryl-7-aminocephalosporanic acid (GL-7-
ACA). Secondly, GL-7-ACA is hydrolysed to 7-ACA
3-fold mass excess compared to DAAO; while the
wt enzyme gave only a 40% yield. Therefore, the
application of this new mutant to the one-pot con-
version of CPC to 7-ACA gives very good result in
terms of efficiency, yield and rate of the process.
Keywords: 7-ACA; d-amino acid oxidase; gutaryl
acylase; hydrogen peroxide; semi-synthetic cephalo-
sporins
[5]
and glutaric acid by glutaryl acylase (GAC). The
main drawback of this system is the presence of hy-
drogen peroxide during the reaction, which leads to
the inactivation of both enzymes, and prevents the
use of just one reactor for the process, to avoid the
[6]
Introduction
exposition of GAC to this inactivating reagent. Re-
cently, a new route for the enzymatic conversion of
[
7]
The semi-synthetic cephalosporins, which are pro- CPC to 7-ACA was reported. This is a tri-enzymatic,
duced from a cephalosporin nucleus, either 7-amino- one-pot system where hydrogen peroxide is eliminat-
Adv. Synth. Catal. 2008, 350, 343 – 348
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
343

Fernando López-Gallego et al.
UPDATE
ed in situ due to the action of catalase (CAT) which is
[
8]
co-immobilised with DAAO. The main advantage
over the conventional system is the absence of hydro-
gen peroxide, avoiding enzyme inactivation by this
compound and permitting the use of just one reactor.
However, now the product of the DAAO oxidation is
2-oxoadipoyl-7-ACA, a product that is unstable and
needs to be rapidly hydrolysed to 7-ACA, in order to
obtain sufficiently high 7-ACA yields.
The main problem of this system is the relatively
low specific activity of wild-type GAC towards
Scheme 1. Adipoyl-7-ADCA
and
2-oxoadipoyl-7-ACA
structures. The difference between the two substrates is
mainly in their aliphatic side-chains. In the case of 2-oxoadi-
poyl-7-ACA, the side chain has a carbonyl group in the a-
carbon, whereas the adipoyl-7-ADCA does not have this
carbonyl group in its side-chain. The differences in the C-3
modifications between the two substrates do not translate in
2
-oxoadipoyl-7-ACA. As a consequence, a large
excess of GAC over DAAO is needed to prevent ac-
cumulation of 2-oxoadipoyl-7-ACA. Thus, any im-
provement on the GAC activity-affinity against this
[
7]
new compound may permit us to reduce the GAC/ a different way to the interaction of both substrates with the
DAAO ratio and increase the global volumetric activ- binding pocket of the enzyme.
ity of the biocatalysts employed.
During the last years, several mutants of different
glutaryl acylases from different microorganisms, Result and Discussion
mainly from Pseudomonas species, have been devel-
oped in order to improve the affinity and activity of Kinetic Characterisation of GAC Activity towards 2-
[9]
the enzyme towards adipoyl 7-ADCA. This com- Oxoadipoyl-7-ACA
pound is produced by feeding adipic acid to a Penicili-
um chrysogenum, which had been genetically modi- The kinetic parameters were determined for four mu-
[
10]
fied to express an expandase enzyme, and the hy- tants, two single mutants (Y178H and N266H) and
drolysis of adipoyl-7-ADCA would yield 7-ADCA in two double mutants (Y178H+N266H and Y178F+
an easy and environmental friendly way, opening a F375H), and the wild-type (wt) enzyme towards glu-
new route to produce the cephalosporin nucleus.
Thus, there are available several GAC mutants that
taryl-7-ACA and 2-oxoadipoyl-7-ACA (Table 1).
The wt enzyme showed Km and Kcat values on
present an important increase in the adipoyl-7-ADCA 2-oxoadipoyl-7-ACA that were 11-fold higher and 17-
[
9,11]
hydrolase activity.
In Pseudomonas SY-77 GAC, it fold lower, respectively, as compared to glutaryl-7-ACA.
[11b]
was possible to identify some residues – Y178,
The N266H mutant showed a very high K for the
m
[
11a]
N266, and P375
– that are important for the spatial 2-oxoadipoyl-7-ACA, even higher than the Km of wt
arrangement of adipoyl-7-ADCA in the substrate enzyme towards this substrate. On the other hand, the
binding pocket of the enzyme. Adipoyl-7-ADCA has Kcat of the mutant N266H for 2-oxoadipoyl-7-ACA
the same side-chain length as 2-oxoadipoyl-7-ACA was slightly increased with respect to the same param-
and the only difference between both compounds is eter for the wt enzyme. Due to these values of Km
the carbonyl group in the a-carbon of 2-oxoadipoyl-7- and Kcat the catalytic efficiencies were very similar for
ACA (Scheme 1).
Here, we have studied the catalytic properties of
different SY-77 GAC mutants for their activity to-
both wt and N266H mutant.
Mutant Y178H showed
a catalytic efficiency
A
H
R
U
G
wards 2-oxoadipoyl-7-ACA and their impact on yield comparing to the catalytic efficiency of wt enzyme to-
and conversion rate in the one-pot conversion of CPC wards the same substrate. This fact is mainly due to
to 7-ACA.
the increment in the Kcat (K slightly decreased).
m
Table 1. Kinetics parameters of wt and different mutant GAC towards different substrates. The reactions were carried out as
described in the Experimental Section at 378C and pH 7.5. The catalytic efficiency was defined as K /K .
cat
m
Enzyme
Glutaryl-7-ACA
2-Oxoadipoyl-7-ACA
À1
À1
K [mM]
m
Kcat [s ]
K /K
cat
K [mM]
m
Kcat [s ]
K /K
m
m
cat
wt
Y178H
N266H
Y178H+N266H
Y178F+F375H
0.34Æ0.02
0.8Æ0.05
0.4Æ0.03
1.4Æ0.1
2.0Æ0.01
1.4Æ0.1
5.8
1.7
6.1
0.35
0.12
3.9Æ0.1
2.2Æ0.1
5.0Æ0.3
19.4Æ0.3
1.27Æ0.1
0.12Æ0.01
0.4Æ0.02
0.03
0.18
0.047
0.003
0.63
2.33Æ0.1
0.5Æ0.02
0.33Æ0.1
0.24Æ0.01
0.05Æ0.002
0.77Æ0.04
2.81Æ0.1
344
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 343 – 348

Evaluation of Different Glutaryl Acylase Mutants
UPDATE
When double mutant Y178H+N266H was ana-
lysed, the catalytic efficiency towards 2-oxoadipoyl-7-
ACA was the lowest of all studied enzymes, even
more, the catalytic efficiency decreased (10-fold) to-
wards 2-oxoadipoyl-7-ACA comparing with the wt
enzyme. This very low catalytic efficiency was due to
both a K increase and a K decrease (Table 1), al-
m
cat
though it has been reported to improve the enzyme
[
12]
performance against adipoyl-7-ACA.
The double mutant Y178F+F375H showed the
best results in all parameters studied towards 2-oxoa-
dipoyl-7-ACA; an improvement of 3- and 6.5-fold in
K , and K , respectively, the values for K and K
cat
m
cat
m
being 1.27 and 0.77, respectively. The Kcat of this
mutant towards 2-oxoadipoyl-7-ACA was around
40% of the Kcat of wt GAC towards glutaryl-7-ACA,
while the K was only around double than of the Km
m
of the wt GAC towards glutaryl-7-ACA. This double
mutant was obtained through rational randomisation
of the most important amino acids implied in the in-
teraction between the enzyme and the substrate, ad-
dressed to improve the efficiency against adipoyl-7-
ADCA. Thus, this mutant also has a highly improved
activity on 2-oxoadipoyl-7-ACA allowing us to go
from a catalytic efficiency around 200-fold lower to-
wards 2-oxoadipoyl-7-ACA than glutaryl-7-ACA
(
using wt enzyme), to only a factor of 9 (using double
mutant Y178F+F375H). Considering that all these
enzymes presented an improved performance against
adipoyl-7-ACA, results suggest that the 2-oxo group
in 2-oxoadipoyl-7-ACA may be playing an important
role in the recognition of this substrate by the
enzyme. Thus, a detailed modelling of the enzyme in-
teractions was performed.
Figure 1. 3D structure of the binding of the side chain of
glutaryl-7-ACA by glutaryl acylase. Residues Asn 266, Tyr
1
78 and Phe 375 were mutated yielding different single and
double mutants. The residue Arg 255 interacts with the car-
boxy group of the GL-7-ACA side chain.
Models of the Active Site of Native and Mutated
Glutaryl Acylase with Bound Substrates
The best results for both adipoyl-7-ADCA and 2-
oxoadipoyl-7-ACA hydrolysis were obtained with
double mutant Y178F+F375H. This mutant was ob-
[12]
Position 266 was demonstrated to be indirectly in- tained through a rational randomisation strategy.
volved in binding the carboxyl group of the substrate Substitution of F375 by various residues was shown to
[
11a]
[13].
side-chain via Arg255
(Figure 1). Replacement of improve adipoyl-7-ADCA hydrolysis activity.
Asn266 by His was very favourable to the binding be-
Both mutations would have a synergistic effect for
[
11a]
tween the enzyme and the adipoyl-7-ADCA,
but the spatial arrangement of both adipoyl and 2-oxoadi-
less favourable for binding to 2-oxoadipoyl-7-ACA poyl side-chains, explaining the improvements on the
Table 1). The model does not give a straightforward hydrolysis activity towards longer side-chain sub-
explanation for this. strates (adipoyl-7-ADCA and 2-oxoadipoyl-7-ACA).
(
In contrast, certain mutations in position 178 im- Remarkably, the double mutant was able to hydrolyse
prove the 2-oxoadipoyl-7-ACA hydrolase activity 2-oxoadipoyl with a Kcat value which is 40% of the wt
with regard to the wt enzyme. The change of Y178 by enzyme hydrolysis activity towards glutaryl-7-ACA.
His was earlier described to increase the hydrolysis of This is an interesting value to use this mutant in the
[
11b]
adipoyl-7-ADCA,
and was explained by assuming one-pot tri-enzymatic synthesis to convert CPC to 7-
that histidine allows better accommodation of the ACA.
[
11b]
longer side-chain of the new substrates (Figure 1).
When looking in more detail at the 3D model of
This also seems to apply to the hydrolysis activity to- positions 178 and 375 of cephalosporin acylase it
Y178F+F375H
wards 2-oxoadipoyl-7-ACA, which is improved 6 fold seems that in SY-77
the scissile bond of 2-
(
Table 1). oxoadipoyl-7-ACA is placed at a more favourable po-
Adv. Synth. Catal. 2008, 350, 343 – 348
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
345

Fernando López-Gallego et al.
UPDATE
Figure 3. One pot 7-ACA production with native (open sym-
bols and dotted lines) and mutant Y178F/F375H (full sym-
bols and solid lines).The reactions were carried with 3-fold
mass excess of GAC (for mutant and native enzyme) with
respect to DAAO. Samples were withdrawn at different
times and analysed by HPLC: 7-ACA (squares). 2-oxoadipo-
yl-7-ACA (triangles).
10 mMof substrate. The specific activity of DAAO to
carry out the oxidative deamination of CPC was
much higher than the specific activity of wt GAC to
carry out the hydrolysis of 2-oxoadipoyl-7-ACA
(
around 20-fold higher). Moreover, DAAO will be
acting against a high concentration of substrate while
GAC will be acting against very low concentrations of
2-oxoadipoyl-7-ACA (because we intend to prevent
Figure 2. Structural impression of the mutational effects. accumulation of this product).
Overlay of 2-oxoadipoyl-7-ACA and active site residues in
When the one-pot system was carried out with the
the structural models of wild-type glutaryl acylase proteins
immobilised mutant GAC Y178F+F375H as biocata-
lyst, the reaction time with mutant GAC was 3.5-fold
shorter than the one with wt GAC to get the 50%
yield of 7-ACA (Figure 3). Moreover, the maximum
Y178F F375H
SY-77
grey).
(dark grey) and refined wild-type (light
sition with respect to the Ser199 when compared with accumulation of 2-oxoadipoyl-7-ACA in the reaction
wt (Figure 2). (using a ratio of GAC/DAO of 3) was of 80% with
Also the presence of polar residues allowing addi- the wt GAC and 40% with the mutant one
tional hydrogen bonds may stabilise the position of (Figure 3). Even more interestingly, the mutant
the side-chain and thereby also of the scissile bond to- enzyme yielded a 95% conversion of 7-ACA while
wards the catalytic serine.
the native enzyme yielded only around 60%
The best result was therefore obtained when simul- (Figure 3).
taneously sufficient space for the longer side-chain
However, the difference between the 2-oxoadipoyl-
was created and formation of new hydrogen bonds is 7-ACA hydrolase activity of wt and mutant (Y178F+
allowed.
F375H) GAC in the one-pot system is lower than one
could expect regarding the catalytic efficiency im-
provement. This fact could be due to different causes.
For example, the wt and mutant GAC acts over a
very low concentration of 2-oxoadipoyl-7-ACA, thus,
the most important kinetic parameter would be the
Use of the Mutant Enzyme in the Tri-Enzymatic
Conversion of CPC to 7-ACA
The conversion of CPC to 7-ACA was carried out Km (3-fold higher for the mutant than for the wt
using a tri-enzymatic system in only one reactor com- enzyme). Moreover, the system with wt enzyme accu-
paring wt GAC and the mutant enzyme Y178F+ mulated more 2-oxoadipoyl-7-ACA than the system
F375H (Figure 3). The specific activities towards with mutant enzyme. This fact means that the 2-oxoa-
2-oxoadipoyl-7-ACA of both immobilised enzymes dipoyl-7-ACA concentration using wt GAC is much
were 0.058 and 0.58 IU/mg, respectively, hydrolysing higher than the concentration over which the mutant
346
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 343 – 348

Evaluation of Different Glutaryl Acylase Mutants
UPDATE
enzyme is working. In this way, although the mutant Purification of Enzymes
enzymes was 18-fold more efficient than the wt one
Mutant and wt enzymes were purified by three chromatog-
raphy steps on a Duoflow system (Bio-Rad) using columns
from http://www.jbc.org/cgi/inline?Amersham Biosciences.
E.coli DH10B containing the plasmids encoding the desired
when the kinetics parameters were calculated, in the
one-pot system, the mutant enzyme was 3.5-fold more
efficient than the wt one.
enzymes were grown for 24 h at 308C in 300 mL of 2*YT
À1
medium supplemented with 50 mgmL
chloramphenicol
Conclusions
and 0.1% glycerol. Cell-free extract was made by sonication
10 min, output 4, 50% duty cycle on a Sonifier 250; Bran-
(
son) and centrifugation (30 min at 14,000 rpm).After this
treatment, the cell-free extract was loaded onto a HiTrapQ
column. The proteins were eluted with a linear gradient of
Some important GAC residues implied in the recogni-
tion and hydrolysis of 2-oxoadipoyl-7-ACA have been
detected , permitting us to produce a mutant glutaryl
acylase having much better properties for the hydroly-
sis of 2-oxoadipoyl-7-ACA than the wild-type
enzyme. However, the mutant enzymes were selected
0
–1MNaCl in 50 mMTris-HCl, pH 8.8. Afterwards, the
fractions containing the enzyme were loaded onto a HiTrap
phenyl-Sepharose HP column. The protein was eluted with
a gradient of 0.7–0M(NH ) SO . The pooled fractions con-
4
2
4
for the hydrolysis of adipoyl-7-ADCA, a substrate taining the enzyme were desalted on a HiPrep 26/10 desalt-
somehow similar but not identical to the new one. ing column. The pooled fractions in 50 mMTris-HCl, pH 8.8
Thus, saturation mutations in these positions could were stored at À208C. Typical yields were 10 mg of more
than 90% pure enzyme per liter of culture.
permit us to obtain new mutants with even better
properties than the Y178F+F375.
Characterisation of Wild-Type and Mutant Enzymes
Enzyme activities were determined towards glutaryl-7-ACA
and 2-oxoadipoyl-7-ACA using the fluorescamine assay, in
Experimental Section
[15]
a 96-well format. All pipetting steps were performed by a
Multiprobe II (Camberra Packard). The kinetic parameters
of these mutants with the two different substrates were de-
termined by the measuring of the initial hydrolysis rate on a
range of substrate concentrations with a fixed amount of
enzyme. 140 mL of 20 mMphosphate buffer, pH 7.5, with
different concentrations of different substrates (0.005–
Chemicals
The b-lactam substrates like glutaryl-7-ACA, adipoyl-7-
ADCA and CPC (as sodium salt) were kindly donated by
DSM(Netherlands). Fluorescamine was from Sigma–Al-
drich (Lousina, USA). Agarose 4 BCL was supplied by
Pharmacia
0
.5 mMglutaryl-7-ACA and 0.02–6.25 mM2-oxoadipoyl-7-
d-Amino acid oxidase from Trigonopsis variabilis
ACA) were preheated at 378C. 40 mL of phosphate buffer
containing an appropriate amount of purified enzyme were
added, starting the reaction. After 30 and 60 min of incuba-
tion at 378C, 40 mL of the reaction mixture were withdrawn
and mixed with 140 mL of 2.5Msodium acetate pH 4.5 to
(
DAAO) was obtained from Recordati (Milan, Italy). Cata-
lase from bovine liver (CATb) was purchased from Fluka
Buch, Switzerland). Glyoxyl agarose 4-BCL (40 mmoles of
(
glyoxyl groups/g support) and polyethyleneimine 600 KDa-
agarose 4BCL supports were prepared according to previous
À1
[14]
stop the reaction and 20 mL of 1 mgmL fluorescamine in
reports.
The Kromasil C-8, (5 mm, 250”4.6 mm) column
acetone. The absorbance was measured at 380 nm after
was purchased from Analisis vínicos (Tomelloso, Spain).
Amonium acetate (HPLC grade) was purchased from
Merck & Co. Acetonitrile was purchased from Biosolve
6
0 min of incubation at room temperature with the fluoresc-
amine. Kinetics parameters were obtained by fitting the ex-
perimental data from Lineweaver–Burk plots, and the mean
values of at least four independent measured were calculat-
ed. The Kcat was calculated using the theoretical molecular
mass of the mature enzyme, 75.9 kDa.
(
Netherlands). All other reagents were of analytical grade.
Preparation of 2-Oxoadipoyl-7-ACA
5
mL of 50 mMof CPC pH 7 were incubated with 0.1 g of
co-immobilized derivative DAAO/CATb prepared as de-
Models of the Active Site of Native and Mutated
Glutaryl Acylase with Bound Substrates
[7]
scribed in ref. The conversion of CPC to 2-oxoadipoyl-7-
ACA was checked by HPLC. When 100% of CPC had been
consumed, the suspension was vacuum filtered and the su-
pernatant was immediately used in the kinetics characterisa-
tion experiments.
Modelling was performed using Discovery Studio 1.7 (Ac-
celrys, San Diego, CA, USA). To get a structural view of the
mutational effects we made an overlay of selected active
site residues using the structural models of the glutaryl acy-
lase proteins containing the mutations Y178F and F375H.
The wild-type structure was also refined using the same pa-
rameters to account for differences in the modelled struc-
tures.
The model for the enzyme was constructed on the basis of
the free P.diminuta KAC-1 (PDB ID 1M2). Hydrogens were
added automatically and the environment of the acylase was
modelled as vacuum. Models of the substrates were con-
Production of Different Glutaryl Acylase Mutants
from Pseudomonas SY-77
Different mutants were constructed and produced as previ-
[11b]
ously detailed. The mutant Y178H was described in ref.
,
the mutant Y178H+N266H (Thesis Linda Otten) and the
[12]
mutant Y178F+F375H were described in ref.
mutant N266H was described in ref.
and the
[11a]
Adv. Synth. Catal. 2008, 350, 343 – 348
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
347

Fernando López-Gallego et al.
UPDATE
structed and energy minimised using the CHARMm force References
[16]
field. Minimisation was done using a dielectric constant of
1
and a non-bonded cut-off distance of 12 Šngstroms. The
[1] R. P. Elander, Appl. Microbiol. Biotechnol. 2003, 61,
385–392.
[2] H. P. B. Fechtig, H. Bickel, E. Vischer, Helv. Chim.
Acta 1968, 51, 1109–1120.
substrate was docked using the grid based approach
[17]
CDOCKER using as coordinates the binding site report-
ed for glutaryl-7-ACA (PDB ID 1JVZ). Initially the enzyme
was fixed and the atoms of the substrate were allowed to
move. In subsequent rounds of minimisation the constraints
on the amino acids forming the active site were removed
and replaced by distance constraints based on the reported
distances observed in the complex with the glutaryl-7-
ACA. The structures were refined by energy minimisation
consisting of 150 steps of steepest descent followed by 2500
iterations of the adopted basis-set Newton–Raphson algo-
rithm.
[3] a) A. Matsuda, K. Toma, K. Komatsu, J. Bacteriol.
1
987, 169, 5821–5826; b) B. B. Oh, Biochem. Biophys.
Res.Commun. 2003, 310, 19–27.
[
[
4] L. Pollegioni, S. Lorenzi, E. Rosini, G. L. Marcone, G.
Molla, R. Verga, W. Cabri, M. S. Pilone, Protein Sci.
[18]
2
005, 14, 3064–3076.
5] a) H. D. Conlon, J. Baqai, K. Baker, Y. Q. Shen, L. B.
Wong, R. Noiles, W. Raush, Biotechnol. Bioeng. 1995,
4
6, 510–513; b) A. A. Nikolov, Enzyme Microb. Tech-
nol. 1994, 16, 1037–1041; c) A. Parmar, H. Kumar, S. S.
Marwaha, J. F. Kennedy, Crit. Rev. Biotechnol. 1998, 18,
Immobilisation of the Enzymes
1
–12.
The co-immobilisation of DAAO and CATb, 1 mg of
DAAO and 10 mg of CATb per gram of support, and the
immobilisation of glutaryl acylase (1 mg of GAC/g of sup-
port) (wt and Y178F+F375H mutant) were carried out as
[
[
6] a) I. de La Mata, F. Ramon, V. Obregon, M. P. Castil-
lon, C. Acebal, Enzyme Microb. Technol. 2000, 27,
2
Mosbach, Appl. Biochem. Biotech. 1991, 27, 239–250.
7] F. Lopez-Gallego, L. Betancor, A. Hidalgo, C. Mateo,
R. Fernandez-Lafuente, J. M. Guisan, Adv. Synth.
Catal. 2005, 347, 1804–1810.
34–239; b) E. S. Dey, J. R. Miller, S. Kovacevi, K.
[7,14]
previously described.
More than 90% of the offered pro-
tein was immobilised in both cases, and we could recover
more than 90% of the offered activity and the kinetics con-
stants did not change after the immobilisation process in
both wt and mutant glutaryl acylase.
[8] R. Fernandez-Lafuente, V. Rodriguez, J. M. Guisan,
Enzyme Microb. Technol. 1998, 23, 28–33.
[
9] C. F. Sio, W. Quax, Curr. Opin.Biotechnol. 2004, 15,
49–355.
Control of the Reaction
3
The products of the reactions were determined through
HPLC with Kromasil C8 [5 mm, 250”4.6 mm) column,
mobile phase 20 mMm ammonium acetate pH 5.2; acetoni-
trile (95:5 v/v)]. The retention times of the different prod-
ucts were: 4.1 min for CPC, 4.9 min for 7-ACA, 6.7 min for
[
[
10] L. Crawford, A. M. Stepan, P. C. McAda, J. A. Rambo-
sek, M. J. Conder, V. A. Vinci, C. D. Reeves, Biotech-
nol. (NY) 1995, 13, 58–62.
11] a) L. G. Otten, C. F. Sio, J. Vrielink, R. H. Cool, W. J.
Quax, J. Biol. Chem. 2002, 277, 42121–42127; b) C. F.
Sio, A. M. Riemens, J. M. van der Laan, R. M. Ver-
haert, W. J. Quax, Eur. J. Biochem. 2002, 269, 4495–
2-oxoadipoyl-7-ACA and 12.3 min for GL-7-ACA.
Conversion of CPC to 7-ACA by the Tri-Enzymatic
System
4
504.
[
[
[
12] C. F. Sio, Mutans and homologs of cephalosporin acy-
lase, PhD Thesis, University of Groningen, 2004.
Different ratios (2, 3, 5, 7,5 and 15) between the mass of
GAC and DAAO were added to 2 mL of 5 mMCPC in
13] C. F. Sio, L. G. Otten, R. H. Cool, W. J. Quax, Biochem.
Biophys. Res. Comm. 2003, 312,755–760.
14] a) J. M. Guisan, Enzyme Microb Technol. 1988, 10,
0.1Mpotassium phosphate buffer at pH 8 and 25 8C. We used
one DAO biocatalyst co-immobilised with CATb on glyoxyl
agarose (1 mg DAAO:10 mg CATb/g biocatalyst) and other
glutaryl acylase biocatalyst (wt or mutant) (1 mg GAC/g bio-
3
75–382; b) C. Mateo, O. Abian, R. Fernandez-La-
fuente, J. M. Guisan, Biotechnol. Bioeng. 2000, 68, 98–
05; c) N. Alonso, F. Lopez-Gallego, L. Betancor, A.
1
catalyst) immobilised on PEI(25 kDa) and treated with gluta-
A
H
E
N
[14]
Hidalgo, C. Mateo, J. M. Guisan, R. Fernandez-La-
fuente, J. Mol. Catal. B: Enzym. 2005, 35, 57–61.
raldehyde. The suspension was gently stirred and the pH
was maintained at 8.0 by adding 4MNaOH during all the
process. Periodically, samples were withdrawn and analysed
via HPLC as described above. Glutaryl-7-ACA did not
appear through the reaction indicating that the co-immobili-
sation of DAAO and CAT was working perfectly.
[15] F. Reyes, M. J. Martinez, J. Soliveri, J. Pharm. Pharma-
col. 1989, 41, 136–137.
[16] A. D. J. MacKerell, B. R. Brooks, C. L. I. Brooks, L.
Nilsson, B. Roux, Y. Won, M. Karplus, CHARMM:
The Energy Function and Its Parameterization with an
Overview of the Program, in: The Encyclopedia of
Computational Chemistry, (Ed.: P. v. R. Schleyer), John
Wiley, Chichester, 1998, pp 271–277.
Acknowledgements
[
17] G. Wu, D. H. Robertson, C. L. Brooks, M. J. Vieth,
This project has been funded by the Spanish CICYT (PPQ
Comput. Chem. 2003, 24, 1549–1562.
2002–01231A) and by STW, with the grant GBI4704 and part
[
18] Y. Kim, W. G. J. S. Hol, Chem. Biol. 2001, 8, 1253–
was also funded by EU grant EST-020278–2. We gratefully
recognise a predoctoral fellowship and a short-term fellow-
ship for Fernando Lopez-Gallego from Spanish MCYT.
1
264.
348
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 343 – 348