Di-functional magnetic nanoflowers: A highly efficient support for immobilizing penicillin G acylase

Article abstract of DOI:10.1002/jccs.201900559

The novel di-functional magnetic nanoflowers (DMNF) which had both epoxy groups and hydrophilic catechol as well as phthaloquinone groups capable of covalently coupling of penicillin G acylase (PGA) were characterized by scanning electron microscopy, transmission electron microscope (TEM), vibrating sample magnetometer, N₂ adsorption, and so on. The studies showed that DMNF possessed “hierarchical petal” structure of nanosheets had specific saturation magnetization of 39.7 emu/g and average pore diameter of 25.4 nm as well as specific surface area of 17.28 m²/g. For hydrolysis of penicillin G potassium catalyzed by the PGA immobilized on DMNF with enzyme loading of 106 mg/g-support, its apparent activity reached 2,667 U/g, which benefited from the “hierarchical petal” and large pore structure of the magnetic DMNF leading to high enzyme loading and fast diffusion of substrate molecules to the immobilized PGA to reaction. The apparent activity of the immobilized PGA could keep 2,408 U/g (above 90% of its initial activity) after repeating use for 10 cycles. The magnetic immobilized PGA exhibited excellent operational stability due to covalently coupling of the enzyme molecules between the support by covalent interaction of the amino groups of PGA and the reactive groups of epoxy, catechol, and phthaloquinone groups on DMNF. Furthermore, the PGA displayed good acid and alkaline resistance as well as thermal stability by immobilization using DMNF.

Full text of DOI:10.1002/jccs.201900559

Received: 29 December 2019
DOI: 10.1002/jccs.201900559
Revised: 22 March 2020
Accepted: 23 March 2020
A R T I C L E
Di-functional magnetic nanoflowers: A highly efficient
support for immobilizing penicillin G acylase
Zheng Zhaoyu | Xue Ping
Li Peng
| Shi Keren | Zhang Weiwei | Hu Chunmiao |
State Key Laboratory of High-efficiency
Utilization of Coal and Green Chemical
Engineering, Ningxia University,
Yinchuan, China
Abstract
The novel di-functional magnetic nanoflowers (DMNF) which had both epoxy
groups and hydrophilic catechol as well as phthaloquinone groups capable of
covalently coupling of penicillin G acylase (PGA) were characterized by scan-
ning electron microscopy, transmission electron microscope (TEM), vibrating
sample magnetometer, N2 adsorption, and so on. The studies showed that
DMNF possessed “hierarchical petal” structure of nanosheets had specific sat-
uration magnetization of 39.7 emu/g and average pore diameter of 25.4 nm as
well as specific surface area of 17.28 m2/g. For hydrolysis of penicillin G potas-
sium catalyzed by the PGA immobilized on DMNF with enzyme loading of
106 mg/g-support, its apparent activity reached 2,667 U/g, which benefited
from the “hierarchical petal” and large pore structure of the magnetic DMNF
leading to high enzyme loading and fast diffusion of substrate molecules to the
immobilized PGA to reaction. The apparent activity of the immobilized PGA
could keep 2,408 U/g (above 90% of its initial activity) after repeating use for
10 cycles. The magnetic immobilized PGA exhibited excellent operational sta-
bility due to covalently coupling of the enzyme molecules between the support
by covalent interaction of the amino groups of PGA and the reactive groups of
epoxy, catechol, and phthaloquinone groups on DMNF. Furthermore, the PGA
displayed good acid and alkaline resistance as well as thermal stability by
immobilization using DMNF.
Correspondence
Xue Ping, State Key Laboratory of High-
efficiency Utilization of Coal and Green
Chemical Engineering, Ningxia
University, Yinchuan, China.
Email: ping@nxu.edu.cn
Funding information
National Natural Science Foundation of
China, Grant/Award Numbers: 21865023,
21961028
K E Y W O R D S
di-functional reactive groups, “hierarchical petal” structure of nanosheets, immobilization,
magnetic nanoflowers, penicillin G acylase
1 | INTRODUCTION
applied catalytic chemical reactions in the synthesis of
pharmaceuticals and fine chemicals as well as degrada-
With the development of modern biotechnologies such as
gene engineering and protein engineering, a variety of
enzymes have been efficiently expressed and produced in
large scale. Enzymes as biocatalysts could have been
tion of organics in wastewater.1 However, free enzyme is
very sensitive to the environment and it is unstable in
acid, alkali, high temperature, and organic solvent so that
its catalytic activity was greatly reduced. Moreover, if free
© 2020 The Chemical Society Located in Taipei & Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
J Chin Chem Soc. 2020;1–11.
http://www.jccs.wiley-vch.de
1

2
ZHAOYU ET AL.
enzyme was directly used in catalytic process, it is diffi-
cult to separate enzyme from reaction system to reuse.
The immobilization technology of enzyme is an impor-
tant subject in the field of industrial application of
enzymes because it not only overcomes the above short-
coming but also maintains the unique catalytic activity of
enzyme. Penicillin G acylase (PGA, EC 3.5.1.11) as a
powerful biocatalyst plays a greatly important role in the
manufacture of semi-synthetic β-lactam antibiotics. It
can hydrolyze penicillin G to produce primary intermedi-
ate of nuclei-6-aminopenicillanic acid (6-APA) as well as
can also catalyze the synthesis of semisynthetic penicil-
lins with a high catalytic performance in mild reaction
conditions.2 It has long been the goal toward exploiting
new-type supports and immobilization techniques to
improve the catalytic efficiency of PGA.
In recent years, magnetic supports for immobilizing
PGA have attracted extensive attention due to its easy
and rapid separation, which was separated from mixture
reaction system by a magnetic force, avoiding the loss of
biocatalysts during separation and washing mechanical
operation. In general, Fe3O4 nanoparticles are employed
as magnetic sources to design and prepare magnetic
supports.3–6 Chen et al.7 synthesized paramagnetic
polymer microspheres as support immobilizing PGA
through polymerization of glycidyl methacrylate, ally
glycidylether, and methacrylamide on the surface of
silica-coated Fe3O4 nanoparticles. The PGA immobilized
on composite microspheres had initial activity of 430 U/g
and retained 99% of its initial activity after recycling
10 times. In addition, immobilized PGA possessed high
thermal stability and pH stability. Zhang et al.8 synthe-
sized functionalized magnetic mesoporous microspheres
Fe3O4@SiO2@m-SiO2-DHA with specific saturation mag-
netization of 61.25 emu/g, which was prepared by modi-
fying with ɤ-(2,3-epoxypropoxy)propytrimethoxysilane
and dendritic hyperbranched amine (DHA) onto the sur-
face of Fe3O4@SiO2@m-SiO2. The apparent activity of
PGA immobilized on Fe3O4@SiO2@m-SiO2-DHA was
approximately 475 U/g; meanwhile, the enzyme activity
retention rate was 94.5% after cycling of 10 runs. Yang
et al.9 reported the aldehyde-functionalized mesoporous
silica-Fe3O4 nanocomposites with a pore size of 7.4 nm
via co-condensation of tetraethylorthosilicate and trim-
ethoxysilylpropanal in the presence of triblock copolymer
of Pluronic P123, NaCl, and Fe3O4 nanoparticles in the
neutral solution. The magnetic immobilized PGA pos-
sessed considerable operation stability. So that, it retained
91.0% of initial activity after recycled for 10 times. As a
potential biomedical material, magnetic inorganic
nanocomposites not only serve as the supports for
immobilizing enzyme but are also applied for drug and
RNA transmission and multiple imaging agents.10–13
With the rapid development of magnetic nanocomposites
and the improvement of immobilization technology, the
“enzyme-carrier” complex has both the new characteris-
tics of the support and the efficient catalytic performance
of the enzyme molecule. For example, the support pos-
sesses a unique morphology structure and the surface is
rich in functional groups that are easy to covalently bind
to PGA. Li et al.14 reported a flexible dendrimer-grafted
flower-like magnetic microcarriers with glutaraldehyde
as crosslinking agent for PGA immobilization, and the
catalytic activity of immobilized PGA maintained 82.14%
after reused for 9 times. However, over-crosslinking of
glutaraldehyde on enzyme and the reversibility of Schiff
base chain cause many inherent difficulties to decrease
activity recovery of immobilized enzyme.15 The active
epoxy group on the magnetic porous polymer micro-
spheres can be covalently bound to the amino groups of
PGA, so that the PGA can be immobilized on the mag-
netic polymer microsphere under mild conditions. It was
considered that the porous supports with large specific
surface area and pore volume are beneficial to reduce the
contact limitation between the catalyst and the substrate
as well as to the migration and diffusion of substrate mol-
ecules, especially for bulky PGA of 7 nm × 5 nm × 5.5 nm.
The magnetic nanoflowers formed via the self-assembly
of Fe3O4 nanochips possessed large pore size and specific
surface area, and they could provide more opportunities
for enzyme molecules to come into contact with the sup-
port surface in the immobilized.16–19
Herein, the novel magnetic composite nanoflowers
with “hierarchical petal” structure and reactive groups of
epoxy groups and hydrophilic catechol as well as
phthaloquinone groups were obtained for immobilizing
PGA. The catalytic performances of the PGA
immobilized on the magnetic composite nanoflowers
were studied for hydrolyzing penicillin G potassium salt
to prepare 6-APA. In addition, the kinetic constants of
the catalyst were determined. The influence of reaction
temperature, pH value, and dopamine amount on the
catalytic performance of immobilized PGA was also
discussed.
2 | EXPERIMENTAL
2.1 | Materials
Penicillin G acylase (PGA, 862 U/ml) was provided by
Zhejiang Shunfeng Haideer Co., Ltd., China. Met-
hacryloxy propyl trimethoxyl silane (MPS) was supplied
from Shanghai Macklin Biochemical Co., Ltd., China.
FeCl3ꢀ6H2O, potassium persulfate (KPS), urea, and ethyl-
ene glycol (EG) were purchased from sinopharm

ZHAOYU ET AL.
3
Chemical Reagent Co., Ltd., China. Glycidy methacrylate
(GMA) and Ethylene dimethacrylate (EDMA) were pro-
duced from Acros Organics, USA. Dopamine hydrochlo-
ride was provided by Ark Pharm, USA. Penicillin G
potassium was purchased via Shanghai Aladdin Bio-
Chem Technology Co., Ltd., China. The other reagents
were of laboratory grade obtained from different com-
mercial suppliers.
specific operation was as follows: 2 g of aqueous solution
of EDMA (5%), 0.1 g GMA and 5 mg dopamine were
sequentially modified on the surface of spherical Fe3O4
nanoparticles (0.5 g) under the same reaction conditions.
The prepared di-functional spherical magnetic
nanoparticles were named as DSMN.
2.4 | Preparation of immobilized PGA
and testing of enzyme loading
2.2 | Preparation of Fe3O4 nanoflowers
Di-functional magnetic nanoflowers (DMNF) (0.2 g) were
dispersed into the mixed solution containing 3.75 ml
phosphate buffer solution (0.01 mol/L, pH 7.8) and
1.25 ml PGA original solution. After mechanical shaking
at 30ꢁC for 12 hr, PGA immobilized on DMNF were col-
lected by a magnet and then washed with phosphate
buffer solution three times to remove free PGA, and the
filtrate was recovered for the determination of enzyme
loading. The immobilized PGA on DMNF was denoted as
PGA/DMNF.
As a comparison, PGA was immobilized onto differ-
ent supports of Fe3O4 nanoflower, PGMN, and DSMN.
Three kinds of carriers of 0.2 g were dispersed into the
mixture containing 3.75 ml phosphate buffer solution
(0.01 mol/L, pH 7.8) and 1.25 ml PGA original solution
under the same immobilization conditions, correspond-
ingly. The obtained immobilized PGA were labeled as
PGA/Fe3O4 nanoflowers, PGA/PGMN, and PGA/DSMN,
respectively.
The concentration of protein in PGA solution before
and after immobilization was determined via Bradford's
dye binding assay. A little diluted immobilized PGA fil-
trate and original enzyme solution were mixed with
coomassie brilliant blue, respectively. Afterward, the
absorbance of solution was determined at 595 nm by UV-
3300 spectrophotometer, and the protein concentration
of filtrates was calculated according to the standard
curve. The enzyme loading of immobilized PGA was
obtained from the following formulas:
FeCl3ꢀ6 H2O (0.1 g) and 3 g urea were dissolved in 40 ml
EG and then stirred for 30 min at the temperature of
50ꢁC. Afterward, the homogeneous solution was trans-
ferred to Teflon-lined autoclave and reacted at 200ꢁC for
4 hr. After the reaction, the sediment was washed several
times with water and ethanol to remove unreacted ions
and organic solvent. Ultimately, the product was calcined
at 400ꢁC for 1 hr under the N2 atmosphere at 5ꢁC/min in
tube furnace to obtain the Fe3O4 nanoflowers. Spherical
Fe3O4 nanoparticles were prepared via a versatile
solvothermal reaction.20
2.3 | Preparation of di-functional Fe3O4
nanoflowers
Fe3O4 nanoflowers (0.5 g) were dispersed into 320 ml eth-
anol and 100 ml deionized water, which was heated to
40ꢁC in a water bath, and adding 5 ml MPS into the solu-
tion under the mechanical stirring. After being stirred for
12 hr, the precipitate was separated by a magnet and
washed several times with ethanol and water. Then, 0.2 g
microspheres abovementioned were dispersed in the mix-
ture consisting of 0.1 g GMA and 2 g EDMA (5%) aqueous
solution, and 10 g KPS (0.5%) aqueous solution was added
into the mixture at 70ꢁC for 6 hr with stirring. After the
completion of polymerization, the polymeric EDMA-
GMA magnetic nanoflowers (PGMN) were separated by a
magnet and washed several times with ethanol and
water, and then the products were dried in vacuum dry-
ing oven. Subsequently, 0.2 g PGMN was dispersed in Tris
buffer solution (0.05 mol/L, pH 8.5) containing 5 mg of
dopamine hydrochloride. The thin film of polydopamine
on the surface of PGMN was formed after shaking at the
temperature of 30ꢁC for 12 hr. After that, the Fe3O4 nan-
oflowers modified by polymeric EDMA-GMA and
polydopamine were separated in magnetic field and
washed three times with water and ethanol. The di-
functional magnetic nanoflowers were named as DMNF.
As a comparison, spherical Fe3O4 was used as support
matrix and di-functionalized via the same method. The
ðm1 −m2Þ
Enzyme loading Q ðmg=gÞ =
ð1Þ
ms
In Equation (1), where m1 is the mass of added
enzyme (mg); m2 is the protein quality in total filtrate
(mg); ms is the mass of support (g).
2.5 | Characterization
The size and morphology of samples were obtained via
scanning electron microscopy (SEM, JEOL JSM-6360LV)

4
ZHAOYU ET AL.
and transmission electron microscopy (TEM, Hitachi
HT7700). The magnetization curve was measured at
room temperature under a varying magnetic field using a
vibrating sample magnetometer (VSM, Lake Shore 7304).
X-ray diffraction (XRD, K/MaX-3c) was used to investi-
gate the crystal structure of the magnetic supports. The
IR spectra were measured in the range from 4,000 to
400 cm−1 via Fourier transform infrared spectrometer
with KBr pellets. Low-temperature N2 adsorption was
performed to determine the pore size and specific surface
area of the magnetic microspheres using a Micromeritics
Model 2010 ASAP apparatus.
images and shown in Figure 1. Early in the reaction
(0.5 hr), the sample was composed of a large number of
Fe3O4 nanoparticles. When the reaction time was
extended to 1.5 hr (Figure 1b), a large number of clus-
tered “petals” could be observed at the expense of the
nanoparticles. After reaction for 4 hr, it can be clearly
observed that the three-dimensional nanostructures
expressed flower-like morphology with no Fe3O4
nanoparticles, and the sample consisted entirely of 3D
nanoflowers. Increasing the reaction time to 5.5 hr, the
morphology of Fe3O4 nanoflowers did not change signifi-
cantly. At the same time, a few Fe3O4 nanoparticles
appeared around the 3D “layered petals.” It was observed
that the Fe3O4 nanoflowers formed at earlier stages were
decorated with Fe3O4 nanoparticles again (Figure 1e),
and no special “layered petals” reserved as shown in
Figure 1f (8 hr).
2.6 | Testing of apparent activity for
immobilized PGA
The apparent activity of the immobilized PGA for cata-
lyzing hydrolysis of penicillin G potassium to produce
6-APA was determined by the titration method as fol-
lows21: The immobilized PGA were added into 10 ml
phosphate buffer solution (0.01 mol/L, pH 7.8) and incu-
bated under the temperature of 30ꢁC for a period of time.
Subsequently, 10 ml 8% (w/v) of penicillin G potassium
solution (which has been preheated at 30ꢁC) was added,
and the mixture was titrated with 0.1 mol/L standard
NaOH solution. Finally, the volume of NaOH aqueous
consumed within 3 min was recorded. The apparent
activity of immobilized PGA was calculated from the fol-
lowing Equation (2).
3.2 | Characterization of DMNF
The SEM images were shown in Figure 2. As seen in the
inset of Figure 2a, Fe3O4 nanoflowers with diameter of
3 μm were made of uniform sheet, which exhibited a
unique appearance of 3D “layered petal.” At the same
time, it can be observed that the surface of DMNF
became rough, but it still maintained unique 3D “layered
petal” (Figure 2b), indicating that the introduction of
polymeric GMA and polydopamine would not destroy
the morphology of composite supports, which was in
accordance with the synthesis of DSMN (Figure 2d). In
addition, the pores of DMNF were mainly in the range of
mesoporous size, which were significantly superior to tra-
ditional spherical microspheres (as shown in Figure 2c,
d); thus, providing a larger contact area between PGA
and supports and contributing to the reduction of mass
transfer resistance in the process of catalysis.
Further observing the microstructure of Fe3O4 nan-
oflowers and DMNF, the high magnification of TEM
images in Figure 3b were clearly exhibited “petal” mor-
phology of DMNF. It can be seen that Fe3O4 nanoflowers
were uniformly surrounded by thin coating, further dem-
onstrating that the modification of polymeric GMA and
polydopamine would not destroy the primary flower-like
architectures of Fe3O4 nanoflowers. The “layered petal”
support with larger specific surface area and mesoporous
size is beneficial for the immobilization of PGA because
it can increase the density of valid functional groups and
provide a friendly biological micro-environment.
C × V × 1000
AðU=gÞ =
ð2Þ
m × t
In Equation (2), where C represents the concentration
of NaOH standard solution (mol/L); V is the volume of
consumed NaOH solution (ml); m is the dry weight of
support (g); and t is the reaction time (min).
To test its reusability, at the end of each reaction, the
PGA/DMNF was separated from the reaction system by
magnet and washed with phosphate buffer solution for
three times. The PGA/D2MNF was continuously used for
10 times of the hydrolysis reaction.
3 | RESULTS AND DISCUSSION
3.1 | Morphology of flower-like Fe3O4
nanostructures
The crystal phases and compositions of corresponding
samples of Fe3O4 nanoflower, PGMN, and DMNF were
examined by X-ray diffraction, as shown in Figure 4. It
can be seen that Fe3O4 nanoflowers were in an inverse
The morphologic evolution of iron oxide precursor over
time was investigated under keeping other reaction con-
ditions the same as the typical experiment by SEM

ZHAOYU ET AL.
5
FIGURE 1 SEM images of Fe3O4 nanoflowers prepared for different reaction times of (a) 0.5 hr, (b) 1.5 hr, (c) 4 hr, (d) 5.5 hr, (e) 6.5 hr,
(f) 8 hr
FIGURE 2 SEM images of Fe3O4
nanoflowers (a), DMNF (b), Fe3O4
spherical nanoparticles (c), and DSMN (d)
cubic spinel structure and the diffraction peaks at 2θ of
30.3ꢁ, 35.7ꢁ, 43.2ꢁ, 53.4ꢁ, 57.4ꢁ, and 62.9ꢁ were accurately
corresponding to the indices of (220), (311), (400), (422),
(511), and (440) crystal planes of Fe3O4 without other
impurity peaks, which proved that good crystallinity of
Fe3O4 can be obtained after calcined at 400ꢁC.22 The
XRD patterns showed that Fe3O4 particles in PGMN and
DMNF possessed favorable crystallinity owing to the
peak position of samples basically the same as that of the
standard Fe3O4 in Figure 4a, which indicated that the
crystal structure of Fe3O4 did not change in the process
of polymerization and coating. The broadening peak at
2θ in 10ꢁ–18ꢁ in Figure 4c belonged to the amorphous
polymeric components in DMNF.
The FT-IR spectra of Fe3O4 nanoflowers, PGMN, and
DMNF were presented in Figure 5. The characteristic

6
ZHAOYU ET AL.
(A)
(B)
1053
(C)
1384
1629
1730
930
2920
2972
3406
1156
FIGURE 3 TEM images of Fe3O4 nanoflowers (a) and
573
DMNF (b)
3500
3000
2500
2000
1500
-1)
1000
500
Wavenumbers
(cm
FIGURE 5 FT-IR spectra of Fe3O4 nanoflowers (a), PGMN
(b), and DMNF (c)
(A)
45
(B)
(C)
(a)
(b)
30
15
0
10
20
30
40
50
60
70
80
-15
-30
-45
2
θ
/ degree
FIGURE 4 XRD patterns of Fe3O4 nanoflowers (a), PGMN
(b), and DMNF (c)
-20000
-10000
0
10000
20000
Magnetic Field (Oe)
absorption band located at 573 cm−1 in Figure 5a was
assigned to the Fe-O bond, indicating that magnetic nan-
oflowers were Fe3O4. It can be seen from spectra in
Figure 5b,c that the peak at 573 cm−1 of Fe3O4 could still
be observed, illustrating that the introduction of
polydopamine and polymeric GMA has no effect on the
structure of Fe3O4, which is consistent with the discus-
sion of SEM and XRD. The appearance of new peak in
Figure 5b at 1,053 cm−1 may be attributed to the linkage
FIGURE 6 Magnetization curve of DMNF (a) and image of
DMNF (b) in a magnetic field
Magnetic property characterization for DMNF is
shown in Figure 6a in which there is no hysteresis in the
magnetic curve with remanence and coercivity being
zero, indicating that it possessed superparamagnetism. It
was seen that the saturation magnetization of DMNF
reached 39.7 emu/g, lower than that of Fe3O4 nanoflower
matrix (61 emu/g).16 However, as shown in Figure 6b,
the DMNF particles dispersed in water are effectively sep-
arated by magnetic force. In our experiment, it was also
observed that the DMNF particles did not appear to
agglomerate during the immobilization process. This
indicates that the magnetic self-agglomeration between
the magnetic particles is weakened after the surface mod-
ification of Fe3O4 nanoflower by polydopamine and
polymeric GMA.
between
OH of Fe3O4 and
O Si O
of MPS.
Besides, the absorption peaks at 1,597 cm−1 and
1,730 cm−1 were related to the stretching vibration of
C C and carbonyl group, respectively. The absorption
peak located at 930 cm−1 was corresponded to the epoxy
group of polymeric glycidyl methacrylate,23 which proved
that glycidyl methacrylate was successfully modified on
the surface of supports. Besides, the absorption bands at
1629 cm−1 and 1,384 cm−1 were associated with the
stretching vibration of the carbonyl group and hydroxyl
group on polydopamine.24 The analytical results of FT-IR
spectra confirmed that polymeric glycidyl methacrylate
and polydopamine were successfully modified in mag-
netic nanoflowers.
The nitrogen adsorption–desorption isotherms
(Figure 7) of Fe3O4 nanoflowers and DMNF showed the
typical H3 hysteresis loop, indicating that pore structure
of Fe3O4 nanoflowers and DMNF was slit pores piled up

ZHAOYU ET AL.
7
3000
2500
2000
1500
1000
500
180
150
120
90
120
100
80
60
40
20
0
Apparent activity (U/g)
Enzyme loading (mg/g)
2667
(a)
Fe O nanoflower
3
4
0.16
0.12
0.08
0.04
0.00
106
1466
1067
933
67
0
20 40 60 80 100 120 140
Pore Diameter (nm)
60
52
45
30
0
0.0
0.2
0.4
0.6
0.8
1.0
Immobilized PGA
Relative Pressure (P/Po)
FIGURE 8 The apparent activity and enzyme loading of
immobilized PGA
80
60
40
20
0
(b)
DMNF
0.25
0.20
0.15
0.10
0.05
0.00
G potassium produce 6-APA are investigated and shown
in Figure 8. It was obviously observed that enzyme load-
ing (106 mg/g) of PGA/DMNF was significantly higher
than protein content of PGA/Fe3O4 nanoflowers (52 mg/
g) and PGA/PGMN (67 mg/g), which was mainly due to
polydopamine film with more active functional groups
for covalently coupled PGA formed on the surface of
PGMN, illustrating that it could provide an excellent
microenvironment for the immobilization of enzyme.
The apparent activity of PGA/DMNF can be up to
2,667 U/g at optimal reaction conditions, which was 2.5
times as much as that of PGA/DSMN (1,067 U/g and
45 mg/g). There are some reports on the immobilization
of PGA with functional magnetic carriers, which focus
mainly on the functional groups of amino, aldehyde, car-
boxyl, and epoxy groups. Ling et al.25 immobilized PGA
on magnetic Fe3O4@chitosan nanoparticles by the Schiff
base reaction and the immobilized PGA with protein
loading of 6.2 mg/g-support was obtained under the opti-
mal conditions. Shi et al.26 used glutaraldehyde as a
cross-linking agent to immobilize PGA on aminopropyl-
functionalized silica-coated magnetic microspheres via
the condensation of tetraethylorthosilicate and
γ-aminopropyltriethoxysilane onto the surface of Fe3O4
microspheres. The result showed that the apparent activ-
ity of immobilized PGA was 384 U/g. Xue et al.27 pre-
pared magnetic hydrophilic polymer microspheres
containing oxyalkyl groups for immobilizing PGA to pro-
duce 6-APA. Under the optimal reaction conditions, the
apparent activity and enzyme loading of immobilized
PGA were only 821 U/g and 65.3 mg/g, respectively. At
present, PGA is mostly immobilized on spherical mate-
rials, and there is a lack of research on PGA immobiliza-
tion by carriers with different morphologies. However, it
is possible to reduce the resistance and improve the
0
20 40 60 80 100 120 140
Pore Diameter (nm)
0.0
0.2
0.4
0.6
0.8
1.0
Relative Pressure (P/Po)
FIGURE 7 N2 adsorption—desorption isotherm curves and
corresponding pore size distribution of Fe3O4 nanoflowers (a) and
DMNF (b)
by plate-like particles,17–19 which was consistent with the
results of SEM images and TEM images. Moreover, it
showed that Fe3O4 nanoflowers (35.3 nm) and DMNF
(25.4 nm) both had large pore size; meanwhile, their spe-
cific surface area were 17.29 m2/g and 17.28 m2/g, respec-
tively. Supports with “layered cross petal” architecture
can be employed to immobilize PGA, which facilitates
the contact with enzyme molecules and be in favor of the
entrance of substrates to the internal surface of
immobilized PGA thus producing higher diffusion effi-
ciency of substrate and promoting the catalytic perfor-
mance of PGA.
3.3 | The activity of PGA immobilized on
nanoflowers
The results of apparent activity and enzyme loading of
immobilized PGA for catalyzing hydrolysis of penicillin

8
ZHAOYU ET AL.
FIGURE 9 The schematic diagram
of covalent interaction between PGA
and DMNF
OH
O
O
O
N
N
O
O
O
O
O
O
O
O
HN
NH
OH
OH
HO
HO
OH
HO
phosphate-buffered solution
OH
HO
O
O
O
O
N
O
N
O
O
O
O
O
OH
O
DMNF
PGA/DMNF
Penicillin G acylase
2800
2600
2400
2200
2000
1800
1600
160
140
120
100
80
Apparent activity (U/g)
Enzyme loading (mg/g)
100
90
80
70
60
50
(A)
(B)
60
2
3
4
5
6
7
8
25
30
35
40
45
50
55
Added amount of dopamine hydrochloride (mg)
(o
Temperature
C)
FIGURE 10 Effect of dosage of dopamine hydrochloride on
catalytic performance of PGA/DMNF
FIGURE 11 Effect of temperature on activity of PGA/DMNF
(a) and free PGA (b)
diffuseness in immobilization of bulky PGA molecules of
by controlling the morphologies of carriers. It was
observed that the PGA-immobilized DMNF with 3D
“hierarchical petal” displayed high apparent activity of
2,667 U/g for hydrolyzing penicillin G potassium to pro-
duce 6-APA.
Using the PGA immobilized on the di-functional
magnetic nanoflowers with mesoporous-size having high
activity of 2,667 U/g and 106 mg/g-support protein load-
ing, where higher than other supports for immobilized
PGA, illuminating that there were two kinds crucial
design, that is, larger pore structure and effective func-
tional groups. The schematic diagram of covalent interac-
tion between PGA and DMNF is shown in Figure 9. The
reactive epoxide functional groups of glycidyl methacry-
late can be directly covalent coupling the amino groups
of PGA via ring opening reactions and polydopamine
film could also offer an easy way to combine the amino
groups of PGA by means of aldolic condensation or Schiff
base reactions to enhance the structural rigidity and oper-
ational stability of PGA.28–29
The effect of dosage of dopamine hydrochloride on
catalytic performance of PGA/DMNF was performed. As
shown in Figure 10, with the dosage of dopamine rising,
the apparent activity and enzyme loading of PGA/DMNF
first increased and the maximum activity and enzyme
loading were both obtained at 5 mg. The apparent activity
and enzyme loading of PGA/DMNF reached 2,667 U/g
and 106 mg/g-support. As the dosage of dopamine hydro-
chloride increased to 8 mg, the apparent activity and
enzyme loading of PGA/DMNF decreased to 1863 U/g
and 83.6 mg/g-support, respectively.
The optimal temperature of free PGA and
PGA/DMNF was determined in the temperature range of
25–55ꢁC (Figure 11). It can be observed that the optimum
temperature of free PGA was obtained at 35ꢁC, while

ZHAOYU ET AL.
9
Running title (a short but clear subtitle, less than 50 characters)
100
90
80
70
60
50
40
3000
2500
2000
1500
1000
500
(A)
(B)
11
0
1
2
3
4
5
6
7
8
9
10
Cycle number
FIGURE 13 Reusability of PGA/DMNF
6
7
8
9
10
pH
7
6
5
4
3
2
1
0
FIGURE 12 Effect of pH on activity of PGA/DMNF (a) and
free PGA (b)
(A)
PGA/DMNF which showed maximum activity, was
shifted to 45ꢁC. When the temperature of free PGA
higher than 35ꢁC with the relative activity falls sharply,
the activity of PGA/DMNF still increased with the raise
of temperature until 45ꢁC, demonstrating that the heat
resistance of PGA/DMNF is better than that of free PGA.
Also, it was obvious that the temperature was as high as
55ꢁC; PGA/DMNF maintained quite a high relative activ-
ity (89%), which was higher than its counterparts, free
PGA (50%). This reaction trend may be explained such
that the enhanced intermolecular interaction between
PGA and the support, making PGA structure more rigid
and protecting the active site from damage.30
(B)
-240 -200 -160 -120 -80 -40
0
40
80 120
1/[S] (L/mol)
FIGURE 14 Lineweaver-Burk plots of PGA/DMNF (a) and
free PGA (b)
In this study, the optimal pH of free PGA and
PGA/DMNF was also investigated with the pH values
varied from 6 to 11 and was seen in Figure 12. With the
enhancement of pH values, it was obviously that free
PGA and PGA/DMNF showed similar trends and their
maximum relative activity were both obtained at pH of
8. It was worthy to note that PGA/DMNF could tolerate
harsher pH condition than free PGA, especially at pH of
11. The PGA/DMNF retained 75% of its initial activity,
while that of its counterpart using free PGA as bio-
catalyst was only 50%. The main reason for this phenom-
enon could be attributed to the introduction of polymeric
GMA and dopamine, whose “synergistic effect” made the
connection between PGA and DMNF more stable, thus
exhibiting the excellent acid and alkaline resistance.
Besides, when the pH was as high as 9–11, the activity of
free PGA and PGA/DMNF fell sharply. The possible rea-
son is that the pH directly affects the surface charge dis-
tribution of PGA, which leads to the change of enzyme
microstructure and affects the binding of PGA and
support.31–32
3.4 | Operational stability of PGA/DMNF
As illustrated in Figure 13, the PGA/DMNF were used for
consecutive cycles to catalyze hydrolysis of penicillin G
potassium salt, and its apparent activity was 2,408 U/g
retaining above 90% of its initial activity after 10 repeti-
tions, which was higher than the activity retention rate
obtained by Jiang's group after repeated use for 9 times.33
The results exhibited the relatively excellent stability and
catalytic activity owing to the co-polymerization of GMA
and dopamine, which benefited from the covalent cou-
pling between their functional groups of epoxy groups,
hydrophilic catechol and phthaloquinone and the amino
group of PGA, meanwhile they provide a more suitable
microenvironment for the for immobilized enzyme. Over-
all, PGA immobilized on DMNF using di-functional cou-
pling method exhibited better adaptability to operational
and economical.

10
ZHAOYU ET AL.
TABLE 1
Apparent kinetics parameters of PGA/DMNF and
ACKNOWLEDGMENTS
free PGA
This work was financially supported by the National Nat-
ural Science Foundation of China (No. 21865023 and
21961028).
Enzyme
preparation
Km
(mol/L)
Vmax
(molꢀL−1ꢀmin−1
)
Free PGA
3.8 × 10−3
9.86 × 10−3
9.34 × 10−3
3.39 × 10−3
ORCID
PGA/DMNF
Xue Ping https://orcid.org/0000-0002-6699-0815
REFERENCES
3.5 | Kinetic constant Km and Vm of
PGA/DMNF and free PGA
[1] X. Pan, L. Wang, J. J. Ye, S. Qin, B. F. He, Appl. Microbiol. Biot.
2018, 102, 1749.
[2] K. Li, Z. B. Chen, D. L. Liu, L. Zhang, Z. H. Tang, Z. Wang,
Y. Zhao, Z. Liu, Polym. Advan. Technol. 2018, 29, 1902.
[3] P. Xue, Y. H. Gu, W. G. Su, H. H. Shuai, J. L. Wang, Appl. Surf.
Sci. 2016, 362, 427.
[4] Y. H. Gu, P. Xue, J. Chin. Chem. Soc. 2018, 65, 696.
[5] A. Morosan, D. E. Mihaiescu, D. Istrati, G. Voicu, M. Radu,
A. Hanganu, R. Stan, Micropor. Mesopor. Mat. 2019, 186, 45.
[6] Z. Y. Zheng, C. M. Hu, C. H. Du, P. Xue, W. W. Zhang,
J. Chin. Chem. Soc. 2019, 66, 1563.
[7] X. Chen, L. Yang, W. C. Zhan, L. Wang, Y. Guo, Y. S. Wang,
G. Z. Lu, Y. L. Guo, Chinese. J. Catal. 2018, 39, 48.
[8] B. L. Zhang, J. Q. Wang, J. J. Chen, H. P. Zhang, D. Z. Yin,
Q. Y. Zhang, Biochem. Eng. J. 2017, 127, 43.
[9] L. Yang, Y. L. Guo, W. C. Zhan, Y. Guo, Y. S. Wang, G. Z. Lu,
Micropor. Mesopor. Mat. 2014, 197, 1.
[10] Y. P. Wang, Y. T. Liao, C. H. Liu, J. Yu, Y. Yamauchi,
M. Shahriar, A. Hossain, K. C. W. Wu, ACS. Biomater. Sci.
Eng. 2017, 3, 2366.
[11] N. D. Thorat, R. A. Bohara, S. A. M. Tofail, Z. A. Alothman,
M. J. A. Shiddiky, M. S. A. Hossair, Y. Yamauchi,
K. C. W. Wu, Eur. J. Inorg. Chem. 2016, 2016, 4586.
[12] N. D. Thorat, R. A. Bohara, V. Malgras, S. A. M. Tofail,
T. Ahamad, S. M. Alshehri, K. C. W. Wu, Y. Yamauchi, ACS.
Appl. Mater. Inter. 2016, 8, 14656.
The hydrolysis rates of penicillin G potassium catalyzed
by free PGA and PGA/DMNF were investigated at differ-
ent substrate concentrations in Figure 14. The apparent
Km constant and the Vmax are calculated via Lineweaver-
Burk plots. As seen in Table 1, the Km value for PGA/DMNF
(9.86 × 10−3 mol/L) was much higher than that for free PGA
(3.8 × 10−3 mol/L), indicating the affinity of immobilized
PGA to the substrate is lower than that of free PGA, while
the Vmax value for PGA/DMNF (3.39 × 10−3 molꢀL−1ꢀmin−1)
was lower than the corresponding value for free PGA
(9.34 × 10−3 molꢀL−1ꢀmin−1). The result manifested that the
diffusion of substrate in DMNF carriers was restricted after
immobilized PGA, and the main reason in the decrease of
the Vmax for immobilized PGA might be attributed to the ste-
ric hindrance effect between enzyme molecules and between
carriers.34
4 | CONCLUSIONS
The magnetic composite nanoflowers with active epoxy
groups and hydrophilic catechol as well as
phthaloquinone groups were successfully designed and
prepared via the polymerization of GMA and coating of
dopamine, which possessed a mesopore size of 25.4 nm
and unique appearance of layered 3D flowers. The results
showed that the dosage of dopamine played an important
role in the catalytic performance on preparation of
6-APA with penicillin G potassium salt catalyzed by
PGA/DMNF. Moreover, the apparent activity and
enzyme loading were obtained of PGA immobilized on
different kinds of supports, in which PGA/DMNF pos-
sessed the apparent activity of 2,667 U/g and enzyme
loading of 106 mg/g-support as well as exhibited super-
paramagnetic properties, that is, 39.7 emu/g. Meanwhile,
the immobilized PGA exhibited superior acid, alkaline
resistance, and thermal stability to the free, and it dis-
played considerable reusability after cycling of 10 runs
with above 90% of its initial activity under magnetic con-
dition. It is obvious that PGA/DMNF is a promising and
efficient biocatalyst in industrial application.
[13] F. K. Shieh, S. C. Wang, C. I. Yen, C. C. Wu, S. Dutta,
L. Y. Chou, J. V. Morabito, P. Hu, M. H. Hsu, K. C. W. Wu,
C. K. Tsung, J. Am. Chem. Soc. 2015, 137, 4276.
[14] X. Li, L. Tian, Z. Ali, W. Y. Wang, Q. Y. Zhang, J. Mater. Sci.
2017, 127, 43.
[15] Z. X. Huang, S. L. Cao, P. Xu, H. Wu, M. H. Zong, W. Y. Lou,
Chem. Eng. J. 2018, 346, 361.
[16] Z. Ali, L. Tian, P. P. Zhao, B. L. Zhang, A. Nisar, X. J. Li,
H. P. Zhang, Q. Y. Zhang, RSC. Adv. 2015, 5, 92449.
[17] X. L. Tian, K. He, B. X. Wang, S. S. Yu, C. C. Hao, K. Z. Chen,
Q. Q. Lei, Colloid. Surface. A. 2016, 498, 185.
[18] Q. Gao, A. W. Zhao, Z. B. Gan, W. Y. Tao, D. Li, M. F. Zhang,
H. Y. Guo, D. P. Wang, H. H. Sun, R. R. Mao, E. H. Liu,
CrystEngComm. 2012, 14, 4834.
[19] M. H. He, Y. Y. Sun, Y. Tian, D. L. Li, G. Z. Zhao, Micro. Nano.
Lett. 2012, 7, 149.
[20] X. J. Li, J. J. Zhou, L. Tian, Y. F. Wang, B. L. Zhang,
H. P. Zhang, Q. Y. Zhang, Sensor. Actuat. B-Chem. 2017,
241, 413.
[21] K. Henzler, B. Haupt, M. Ballauff, Anal. Biochem. 2008,
378, 184.
[22] S. K. S. Patel, S. H. Choi, Y. C. Kang, J. K. Lee, ACS. Appl.
Mater. Inter. 2017, 9, 2213.

ZHAOYU ET AL.
11
[23] M. Barsbay, Y. Kodama, O. Giiven, Cellulose 2014, 21, 4067.
[24] Y. L. Xiao, J. T. Zai, X. M. Li, Y. Gong, B. Li, Q. Y. Han,
X. F. Qian, Nanoenergy 2014, 6, 51.
[25] X. M. Ling, X. Y. Wang, P. Ma, Y. Yang, J. M. Qin, X. J. Zhang,
Y. W. Zhang, J. Microbiol. Biotechn. 2016, 26, 829.
[26] B. F. Shi, Y. Q. Wang, J. W. Ren, X. H. Liu, Y. Zhang,
Y. L. Guo, Y. Guo, G. Z. Lu, J. Mol. Catal. B-Enzym. 2010,
63, 50.
[32] L. Mosafa, M. Shahedi, M. Moghadam, J. Chin. Chem. Soc.
2014, 61, 329.
[33] Y. J. Jiang, W. Y. Sun, Y. P. Wang, L. H. Wang, L. Y. Zhou,
J. Gao, Y. He, L. Ma, X. Zhang, Enzyme. Microb. Tech. 2017,
96, 42.
[34] A. I. Kallenberg, F. van Rantwijk, R. A. Sheldon, Adv. Synth.
Catal. 2005, 347, 905.
[27] P. Xue, W. G. Su, Y. H. Gu, H. F. Liu, J. L. Wang, J. Magn.
Magn. Mater. 2015, 378, 306.
[28] X. X. Li, D. M. Li, W. F. Wang, R. Durrani, B. Yang,
Y. H. Wang, J. Mol. Catal. B-Enzym. 2016, 133, 154.
[29] M. F. C. Andrade, A. L. A. Parussulo, C. G. C. M. Netto,
L. H. Andrade, H. E. Toma, Biofuel. Res. J. 2016, 3, 403.
[30] S. Matsuura, M. Chiba, T. Tsunoda, A. Yamaguchi, J. Nanosci.
Nanotechno. 2018, 18, 104.
How to cite this article: Zhaoyu Z, Ping X,
Keren S, Weiwei Z, Chunmiao H, Peng L. Di-
functional magnetic nanoflowers: A highly
efficient support for immobilizing penicillin G
acylase. J Chin Chem Soc. 2020;1–11. https://doi.
org/10.1002/jccs.201900559
[31] A. A. Kadam, J. Jang, S. C. Jee, J. S. Sung, D. S. Lee, Carbohyd.
Polym. 2018, 194, 208.