Assembly of nitroreductase and layered double hydroxides toward functional biohybrid materials

Article abstract of DOI:10.1016/j.jcis.2018.07.126

The development of new multifunctional materials integrating catalytically active and selective biomolecules, such as enzymes, as well as easily removable and robust inorganic supports that allow their use and reuse, is a subject of ongoing attention. In

Full text of DOI:10.1016/j.jcis.2018.07.126

Accepted Manuscript
Assembly of Nitroreductase and Layered Double Hydroxides toward Functional
Biohybrid Materials
Felipe Bruna, Christine Mousty, Pascale Besse-Hoggan, Isabelle Batisson,
Vanessa Prevot
PII:
S0021-9797(18)30889-0
https://doi.org/10.1016/j.jcis.2018.07.126
YJCIS 23917
DOI:
Reference:
Journal of Colloid and Interface Science
To appear in:
Received Date:
Revised Date:
Accepted Date:
24 April 2018
24 July 2018
28 July 2018
Please cite this article as: F. Bruna, C. Mousty, P. Besse-Hoggan, I. Batisson, V. Prevot, Assembly of Nitroreductase
and Layered Double Hydroxides toward Functional Biohybrid Materials, Journal of Colloid and Interface
Science (2018), doi: https://doi.org/10.1016/j.jcis.2018.07.126
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Assembly of Nitroreductase and Layered Double
Hydroxides toward Functional Biohybrid Materials
Felipe Bruna†‡, Christine Mousty†, Pascale Besse-Hoggan†, Isabelle Batisson‡, Vanessa
Prevot*,†
† Université Clermont Auvergne, CNRS, Sigma-Clermont, ICCF, Institut de Chimie de
Clermont-Ferrand, F-63000 CLERMONT-FERRAND, France
‡ Université Clermont Auvergne, CNRS, Laboratoire Microorganismes : Génome et
Environnement, F-63000 CLERMONT-FERRAND, France
KEYWORDS: enzyme immobilization, layered double hydroxides, nitroreductase, biohybrid
materials, biomolecules, mesotrione
ABSTRACT
The development of new multifunctional materials integrating catalytically active and selective
biomolecules, such as enzymes, as well as easily removable and robust inorganic supports that
allow their use and reuse, is a subject of ongoing attention. In this work, the nitroreductase
NfrA2/YncD (NR) from Bacillus megaterium Mes11 strain was successfully immobilized by
adsorption and coprecipitation on layered double hydroxide (LDH) materials with different
compositions (MgAl-LDH and ZnAl-LDH), particle sizes and morphologies, and using different
enzyme/LDH mass ratios (Q). The materials were characterized and the immobilization and
1

catalytic performance of the biohybrids were studied and optimized. The nitroreductase-
immobilized on the nanosized MgAl-LDH displayed the best catalytic performance with 42-46%
of catalytic retention and more than 80% of immobilization yield at saturation values of enzyme
loading Cs ≈ 0.6 g NR/g LDH (Q = 0.8). The adsorption process displayed high enzyme-LDH
affinity interactions yielding to a stable biohybrid material. The increase in the amount of
enzyme loading favoured the catalytic performance of the biohybrid due to the better
preservation of the native conformation. The biohybrid was reused several times with partial
activity retention after 4 cycles. In addition, the biohybrid was successfully dried maintaining the
catalytic activity for several weeks when it was stored in its dry form. Finally, thin films of
NR@LDH biohybrid deposited on glassy carbon electrodes were evaluated as a modified
electrode applied for nitro-compound detection. The results show that these biohybrids can be
used in biotechnology applications to efficiently detect compounds such as dinitrotoluene. The
search for new non-hazardous chemical designs preventing or reducing the use of aggressive
chemical processes for human being and the environment is the common philosophy within
sustainable chemistry.
1. INTRODUCTION
Highly reactive and selective microbial biomolecules, such as enzymes, are recognized as
powerful tools involved in many well-established natural and synthetic processes. In the last
decades, the enzyme immobilization has gained worldwide interest for a more widespread use
ranging from industrial biotransformation, biosensing, drug delivery to food processing [1-4].
Indeed enzyme immobilization on suitable supports produces heterogeneous biohybrids with
potentially new functionalities synergistically offered at molecular scale by the intrinsic
properties of the host material and its interactions with biomolecules. The produced functional
biohybrids usually display interesting advantages such as the possibility of reuse, improved
2

operational stability, ease of separation of the reaction medium, reduction of contamination in
the final product and improvement in the control of biocatalytic reactions.
Obviously, both the immobilization process and the type of support are key factors for an
optimal biohybrid performance. Several techniques for the enzyme immobilization were
described in the literature, such as adsorption, entrapment, covalent binding or cross linking [5,
6]. With the aim of a cost-effectiveness biohybrid, the immobilization technique should be
simple and inexpensive, using chemically soft and reproducible processes and avoiding
hazardous solvents or reagents. Different types of organic and inorganic materials have been
extensively studied and reported as efficient supports for enzyme immobilization, such as
chitosan, cellulose, epoxide-containing polymers, silica, oxides and clays, evidencing the need of
optimizing the conditions for each enzyme/immobilization matrix couple [5, 7-11].
Among the various supports available, layered double hydroxides (LDH) show many
advantages compared to other organic and inorganic host structures, such as improved physico-
chemical stability, biocompatibility and resistance to microbiological attacks. Within the
potential inorganic supports, LDH are promising materials especially due to their large surface
area with abundant basic binding sites, 2D open framework, unique exchange properties,
tuneable particle size and low cost [12]. LDH, also referred as anionic clays, can be described
III
according to a brucite-type metal hydroxide structure with a general formula [M IIMx (OH) ]x+
1−x
2
−
Xx n n ∙mH2O. On one hand, the molar ratio and the type of divalent, MII, and trivalent, MIII, metals
/
in the layers control the physico-chemical properties like the charge density or basicity of the
materials. On the other hand, the type and nature of the exchangeable anion, Xn-, may modify
their adsorption properties. These features provide to the LDH compounds a high structural and
morphological versatility. Such kind of inorganic materials have been extensively studied as
matrices of confinement for various biomolecules ranging from amino acids, DNA, to enzymes,
3

other proteins and even bacteria [13-16]. Indeed, several enzymes such as alkaline phosphatase,
lipase, α-amylase or transketolase have been successfully immobilized and characterized on
different phases of LDH-type materials through various immobilization strategies, evidencing
the need of optimizing the process for each enzyme to access to efficient biohybrid [17-20].
Likewise, some applications involving enzymes/LDH biohybrids have been described in the
literature, especially as biocatalysts or biosensors [21, 22].
Nitroreductases are interesting enzymes able to reduce nitroaromatic derivatives often highly
toxic, mutagenic, or carcinogenic which makes them of great interest for potential environmental
applications [23]. The reduction of nitroaromatic compounds is catalysed via a two or one-
electron transfer by nitroreductase type I or II, respectively [24, 25]. Recently, we have
identified and produced nitroreductases called NfrA1 and NfrA2/YcnD from Bacillus
megaterium Mes 11 capable to biotransform mesotrione (See supporting information Figure S1),
an herbicide belonging to the triketone family broadly used on maize crops and having a
nitroaromatic moiety [26, 27]. These enzymes belong to the NfrA Frp family of the nitro-flavine
mononucleotide (FMN) reductase superfamily (oxygen insensitive nitroreductase type I) using
nicotinamide adenine dinucleotide phosphate (NADPH) or nicotinamide adenine dinucleotide
(NADH) as cofactors [28]. Its broad activity spectrum for biotransforming several nitroaromatic
compounds including insecticides, fungicides, antibiotics and prodrugs opens the way to various
opportunities for potential applications of NfrA2/YcnD, for instance in biosensor development or
bioremediation as suggested by Carles et al. [26].
It is noteworthy to mention that in the literature, there is little work that has the immobilization
of enzymes with nitroreductase activity (See supporting information Table S1). A nitroreductase
from E. cloacae fused with a maltose binding protein, and immobilized on glassy carbon
electrodes (GCE) by entrapment within redox active organic polymer, was applied as biosensor
4

for the detection of 2,4,6-trinitrotoluene (TNT) [29]. Recently, Komarova et al. [30] reported a
nitroreductase-modified ion-selective field effect transistor for the detection of TNT and other
nitroaromatic explosives. Additionally, the immobilization of the nitrobenzene nitroreductase
from P. pseudoalcaligenes JS45 on polyethyleneimine-mediated silica particles and the
immobilization of a genetically modified nitroreductase from E. coli on gold nanoparticles have
been described for the activation of the CB1954 cancer prodrug [31, 32].
In this work, we were interested to investigate the efficiency of LDH to immobilize
nitroreductases which in our knowledge, has never been reported. Having in mind the elements
exposed above, we report here a rational approach to the development and characterization of
NfrA2/YcnD (NR) immobilized biohybrids using LDH compounds as supports. For that, Mg/Al
and Zn/Al LDH matrices were prepared and characterized. The enzyme immobilization on those
LDH phases was carried out using different soft chemical routes: adsorption or direct
coprecipitation. Thus, the so-prepared biohybrids were characterized and the performance of the
different NR@LDH biohybrids for the mesotrione-biotransformation was assayed. Additionally,
the NR activity was also studied electrochemically to investigate the suitability of NR-based
biohybrids for biotechnology applications, such as biosensors.
2. EXPERIMENTAL SECTION
2.1. Reagents
Bradford reagent was obtained from Bio-Rad and Ni-NTA resin from Qiagen. Bovine serum
albumin (BSA, purity > 98%), mesotrione (Pestanal®, purity 99.9%), 2,3 dinitrotoluene (DNT),
nicotinamide adenine dinucleotide (NADH), 3-(N-morpholino)propanesulfonic acid (MOPS)
were purchased from Sigma-Aldrich. Others chemicals used in this work were of analytical
grade or better and used as received without further purification.
2.2. Instrumentation
5

The X-ray diffraction patterns of powder samples (PXRD) were recorded using Philipps X-
Pert Pro diffractometer with Cu Kα radiation source (λ = 0.15405 nm). Patterns were recorded
over the 5–70° (2θ) range in steps of 0.033° with a counting time of 200 s per step.
Infrared transmission spectra (FTIR) were recorded using the KBr pellet technique over a
wavenumber region of 400–4000 cm−1 with a Nicolet 5700 spectrometer from Thermo Scientific.
The sample was mixed and homogenised with dry potassium bromide in a 1% of weight.
Hydrodynamic diameter (d, nm) and zeta potential (ζ-potential, mV) were obtained for the
pristine LDH samples and for the biohybrids by dynamic light scattering (DLS) and
electrophoretic light scattering (ELS), respectively, using a Zetasizer Nano ZS instrument
(Malvern Instruments) coupled to an automatic titration unit (MPT-2). The measurements were
performed using buffered sample suspensions containing 0.05 M MOPS, pH 7.5.
Fluorescence spectra of the free and immobilized enzymes in aqueous solution (~ 0.1 mg / mL)
were recorded at room temperature at λexc = 280 nm and λem = 300-450 nm. A Varian Cary
Eclipse fluorescence spectrophotometer was used, adopting a 5 nm bandpass on both excitation–
emission. FESEM characteristics of the samples were imaged by a Zeiss supra 55 FEG-VP
operating at 3 keV and 10 kV combined with an energy dispersive X-ray (EDX) analyzer.
Transmission electron microscopy (TEM) images were recorded on a Hitachi 7650 microscope
operated with an acceleration voltage of 80 kV. To perform the characterization, a drop of the
suspension was deposited on a 400 mesh holey carbon-coated copper grid and dried at room
temperature. AFM images were obtained on a Bruker Nanoscope IIIa atomic force microscope.
Imaging was achieved using the repulsive mode in intermittent contact, commonly known as the
‘‘tapping” mode.
2.3. NfrA2/YncD expression and purification
6

The production and purification of the recombinant His-tagged NfrA2/YncD enzyme was
obtained as previously described [26], with some modifications. Typically, E. coli BL21 (DE3)
pLys cells carrying the recombinant plasmid were grown overnight at 30 ºC in Luria-Bertani
(LB) medium containing 100 µg/mL of ampicillin and 34 µg/mL of chloramphenicol, under
agitation (140 rpm) on an orbital shaker. The expression of the enzyme was realized by induction
at OD600 nm = 0.6 with 1 mM of isopropyl β-D-1-thiogalactopyranoside (IPTG) and incubation of
the culture at 30 ºC overnight. Crude extracts were obtained by lysis of cells in a One Shot Cell
Disrupter (Constants Systems), under 2.6 kbar pressure (3 cycles), in 50 mM sodium phosphate
buffer solution (PBS), pH 8, containing 300 mM NaCl and 10 mM imidazole (7 mL/g of cells).
◦
Cell debris were eliminated by centrifugation (20000 g for 30 min at 4 C) and the supernatant
was loaded on a Ni2+-nitrilotriacetate agarose (Ni-NTA) chelating affinity column (Qiagen)
previously conditioned with the same PBS. After washing the column with 50 mM PBS pH 8,
containing 300 mM NaCl and 20 mM of imidazole at 4 ºC, the His-tagged NfrA2/YncD was
finally eluted with 50 mM PBS, pH 8, containing 300 mM NaCl and 250 mM imidazole.
Fractions containing the enzyme were pooled and dialyzed overnight in 2 mM PBS pH 8. A
second dialysis was carried out in deionized water for 5 h.
The size of the purified protein was monitored with SDS-PAGE molecular mass standards,
between 10 and 250 kDa (BioRad). SDS-PAGE was performed in 12 % (w/v) polyacrylamide
gel as described by Laemmli (1970) [33]. The SDS-PAGE showed a good expression of the
protein with a major band in the purified extract corresponding to the monomer size of
NfrA2/YcnD as decribed by Carles et al. [26] (See supporting information Figure S2). The
purified NR enzyme was lyophilized and the powder was stored at -18 ºC until needed.
2.4. Enzyme activity assay
7

The protein content of free and immobilized enzyme samples was estimated by the Bradford
assay [34], using a Coomassie blue-based protein dye reagent (BioRad) and BSA as the standard
protein. Briefly, the Bradford reagent was added to a solution containing free (1-6 μg of protein)
or immobilized NR (0.5-3 μg of protein), to a final volume of 20 % (w/w), and shook manually
in a 96 - well microplate. In the case of the immobilized NR, the content of LDH was ≤ 2 %
(w/v) in all the samples to ensure the complete dissolution of the material and release of the
enzyme, maintaining the linearity of the microassay. After 10 min, the absorbance at 595 nm of
the mixture was measured using a microplate reader (TECAN®) and compared to the BSA
standard curve made in the same buffered conditions.
Enzyme activity of the NR for degrading 0.1 mM mesotrione was evaluated at 28 ºC as
previously described [26], in 50 mM MOPS, at optimum pH 6.5, in the presence of 0.7 mM
NADH and 12.5 or 25 µg/mL of free or immobilized enzyme, respectively. The reaction was
stirred and stopped after 1 h by the addition of concentrated hydrochloric acid with final
concentration of 7 % (v/v). The catalytic reaction was linear in the whole time interval. The
disappearance of mesotrione was then monitored by HPLC as previously described [26]. One
unit of nitroreductase activity (U) was defined as the amount of enzyme which transforms 1
μmol of mesotrione equivalent per minute under the above experimental conditions. Specific
activity, expressed as units of enzyme per mg of protein (U/mg or µmol/min.mg), was monitored
during the different purification stages, until obtaining the lyophilized protein. The enzymatic
activity was conserved during the whole process, with a purification index of 10.9 compared
with the crude extract. Indeed, the specific activity of the crude extract was 2.4 mU/mgof protein
and that of the purified NR was 25.4 ± 6.3 mU/mg protein.
When DNT was tested as a substrate, the enzymatic activity of free NR was determined via the
consumption of NADH by spectrophotometry at 340 nm (= 6220 M-1cm-1) using a Cary 300
8

UV-Vis (Agilent). The reaction was carried out in 0.1 M PBS (pH = 6.5) containing 300 µM
DNT, 350 µM NADH and 20 µg of NR. The specific activity was 2.99 U/mg protein (1 U = 1
µmol of NADH consumed/min).
2.5. Preparation of LDH samples
Two samples of nitrate-intercalated LDH compounds, namely MgAl-LDH and ZnAl-LDH,
were prepared with 2:1 M2+/M3+ molar ratio by classical coprecipitation (see supporting
information) [35]. MgAl-NO3 nanoparticles (noted MgAl-LDHnano) were prepared by flash
coprecipitation, followed by a hydrothermal treatment as previously reported [36, 37]. A volume
of 100 mL of metal nitrate solution (Mg/Al = 2, 0.4 M) was rapidly added to 300 mL of a 0.25 M
NaOH solution under vigorous stirring. The dispersion was aged for 30 min at pH 10.0 in ice
bath. The whole procedure was carried out under N2 atmosphere. The dispersion obtained was
hydrothermally treated at 150 ºC and after 4 h, it was air-cooled. Then, the sample was
centrifuged at 18500 g for 30 min, washed three times with deionized water and dispersed in 50
mL of deionised water. The resulting nanoparticles were re-dispersed in deionized water and
stored as a colloidal dispersion (26 %, w/w) at 4 ºC.
2.6. Immobilization of NR on LDH materials
In this study, two different immobilization strategies were investigated to prepare NR@LDH
biohybrids, adsorption and direct coprecipitation, since it is well-known that immobilization step
can strongly modify the enzyme activity [38]. NR was immobilized on MgAl-LDH, ZnAl-LDH
and MgAl-LDHnano materials by adsorption process, mixing enzyme and LDH suspensions.
Suspensions of 2 mL of 50 mM MOPS pH 7.5, containing 2 mg of LDH and NR with initial
concentrations (Ci) between 0.1 and 2 g/L were gently shaken at 30 rpm. After shaking overnight
at 4 ºC, the suspensions were centrifuged at 11000 g for 15 min, washed with deionized water
several times and stored in 50 mM MOPS pH 7.5 at the same temperature. The NR adsorption
9

isotherm on MgAl-LDHnano was performed and fitted to the linear form of Freundlich’s equation,
f
lf
lf
Cs = K CeN , and Langmuir−Freundlich’s equation, Cs = (Cm Cen )/(K + Cen ), as previously
f
d
described [19] .
Two biohybrids were prepared by the direct coprecipitation route following the method
described above and performed in this case in presence of the enzyme NR in an adapted reduced
scale material (see supporting information) [39]. In both cases, the amount of immobilized
protein on the support (Cs) was quantified indirectly from the difference between the initial (Ci)
and the equilibrium (Ce) solution concentrations of protein, respectively, through the following
equation:
(1)
where Cs (mg/g) is the amount of NR adsorbed, or adsorption capacity, at the equilibrium
concentration Ce (mg/L); Ci (mg/L) is the initial protein concentration in solution; V is the
solution volume (L) and m is the adsorbent mass (g). In addition, the amount of immobilized
enzyme was also determined directly from the NR@LDH biohybrids using the Bradford assay as
described above.
2.7. Storage stability and recyclability experiments
For storage stability, the biohybrids were stored at 4 ºC in two forms: in solution (50 mM
MOPS, pH 7.5) or as a lyophilized powder. The biocatalytic activity was compared with that of
the free enzyme stored under the same conditions. The specific activity of the free or
immobilized enzyme was checked during at least one month, using 0.1 mM mesotrione under the
same assay conditions as those mentioned before. The pristine LDH was used as a negative
control for catalytic activity toward mesotrione adsorption and degradation.
To study the effect of recovering and recycling the biocatalysts, a NR@LDH biohybrid sample
was repeatedly exposed to mesotrione. The NR@LDH biohybrid (0.15 mg), with a protein
loading of Cs = 0.6 mg/mg MgAl-LDHnano, was added to the reaction medium (1 mL) containing
10

100 µM mesotrione and 700 µM NADH in 50 mM MOPS pH 6.5 at room temperature. The
NR@LDH biohybrid was recovered by centrifugation in the same reaction vials (1.5 mL
eppendorf tubes) and used immediately in a new reaction cycle. Then, a volume of 0.1 mL of
concentrated hydrochloric acid was added to 0.5 mL of supernatant and the NR activity was
determined as described in the previous section. The reaction was then repeated 3 additional
times resulting in 4 sequential cycles of the same NR@LDH sample.
2.8. Electrochemical experiments
Cyclic voltammetry (CV) was recorded out with a three-electrode system using a PGSTAT
204 controlled by the NOVA software (Metrohm). A saturated calomel electrode (SCE) was
used as the reference electrode and a Pt wire as the counter electrode. The working electrode was
a modified glassy carbon electrode (GCE, A = 0.196 cm2). Before the deposition of the
NR@MgAl-LDHnano biohybrid, the GCE was polished with a diamond paste (1 µm), washed with
acetone, then polished with alumina slurry (0.04 µm), and finally rinsed with ethanol and water.
A drop (50 µL) of NR@MgAl-LDHnano mixture, corresponding to 100 µg of biohybrid material
(Enzyme/LDH 1:1) was deposited on the surface of GCE and the bioactive layer was allowed to
dry under vacuum for 1 h. Before use, the NR@MgAl-LDHnano/GCE modified electrodes were
rehydrated in 0.1 M PBS (pH = 6.5) for 15 min and then transferred into the electrochemical cell
containing 0.2 mM mesotrione or DNT and 0.4 mM NADH in 0.1 M PBS to carry out CV
experiments at a scan rate of 10 mV/s under argon atmosphere.
3. RESULTS AND DISCUSSION
3.1. Synthesis and biological activity of NR@LDH biohybrids
In this study, purified NfrA2/YncD from Bacillus megaterium Mes11, noted NR, was
immobilized on nitrate-intercalated MgAl-LDH and ZnAl-LDH biocompatible matrices by two
different approaches: adsorption and coprecipitation leading to biohybrids named NR@LDH-ads
11

and NR@LDH-cop, respectively. According to the literature and our previous studies concerning
enzyme immobilization on LDH-type materials [17, 20], the initial M2+/M3+ ratio in the LDH
layer was fixed at 2 in order to maximize the charge density of the layers of these materials.
Nevertheless, other key parameters were optimized: a) the type of the immobilization procedure,
b) the chemical composition of the LDH and c) the enzyme/LDH mass ratio ranging from 0.1 to
2.
For the adsorption, three different LDH samples were prepared either by the classical method
of coprecipitation [35], ZnAl-LDH and MgAl-LDH, or flash coprecipitation MgAl-LDHnano [36,
37] to tune the particle size and aggregation state. All the samples prepared exhibited the typical
XRD patterns of hydrotalcite-like materials (Figure 1) with a basal spacing d of 0.86 nm in
agreement with the intercalation of nitrate anions between the LDH layers (See supporting
information Table S2). The chemical analyses confirmed the LDH coprecipitated with a M2+/M3+
ratio near the expected one, with values of 2.2, 2.0 and 1.9 for ZnAl-LDH, MgAl-LDH and
MgAl-LDHnano, respectively.
12

Figure 1. XRD patterns of a) ZnAl-LDH b), MgAl-LDH c) MgAl-LDHnano and d) NR@MgAl-
LDHnano at Cs = 0.6 mg/mg (* Mg(OH)2).
The TEM images (Figure 2) of the precursors further reveal the differences between the two
synthetic routes used, in relation to their morphology and particle size. The samples prepared by
the classical method of coprecipitation, MgAl-LDH and ZnAl-LDH, present the usual densely
packed particles, configured as aggregates (Figures 2A-2B). In contrast, small polydisperse
particles with an almost spherical shape are observed for the MgAl-LDHnano sample (Figure 2C).
The particle sizes deduced from the TEM images range from 70 to 130 nm, with an average
around 90 nm. This value is close to the hydrodynamic diameter of 130 ± 54 nm (polydispersity
index: 0.17) obtained by DLS. In the case of samples prepared by classical coprecipitation,
greater particle size about 600-700 nm was inferred.
13

Figure 2. TEM images of A) ZnAl-LDH B), MgAl-LDH C) MgAl-LDHnano and D) NR@MgAl-
LDHnano at Cs = 0.6 mg/g.
A key factor leading to the immobilization of proteins on LDH and other inorganic materials is
the electric charge parameters [40]. As expected, the zeta potential values for all the LDH
matrices are positive ranging from 21 mV to 37 mV at a pH of 7.5 (see supporting information
Table S.2). For NR, the isoelectric point (IEP) was determined at pH 5.7 (See supporting
information Figure S3) close to the theoretical IEP = 6.1 calculated on the basis of the amino
acid sequence of NfrA2/YncD from B. megaterium [41, 42]. Considering this IEP, the protein
will be negatively charged at basic pH (50 mM MOPS, pH 7.5) used in the immobilization
process. A priori, these opposite values should favour electrostatic interactions between both
components and promote the biohybrid formation [12, 19, 20].
Figure 3A shows the influence of the chemical composition of the LDH (MgAl vs ZnAl),
prepared by classical or flash coprecipitation, on the immobilization yields and the enzyme
activity of the NR immobilized by adsorption or coprecipitation at different NR loadings.
14

Concerning the immobilization yield, the enzyme was completely immobilized for the lower
enzyme amount (Enzyme/LDH initial mass ratio: Q = 0.5), proving an excellent immobilization
efficiency in all the cases. At Q ≥ 1, modulations were observed according to the matrices and
conditions used. The amount of immobilized enzyme in the NR@ZnAl-LDH-ads biohybrid was
close to 100 % for the whole Q range studied, while in the NR@MgAl-LDH-ads it significantly
decreased with the increase of the Q ratio, evidencing the material saturation. Such influence of
the LDH chemical composition on the enzyme immobilization rate was previously observed for
transketolase and alkaline phosphatase using MgAl-LDH and ZnAl-LDH as supports [17, 38].
However, the immobilization capacity of the inorganic materials was not only influenced by its
chemical composition but also by the immobilization method: for instance, the amount of
immobilized enzyme on the MgAl-LDH matrix was significantly higher when the
coprecipitation route was used as an alternative immobilization strategy (NR@MgAl-LDH-cop
biohybrid). Obviously, the coprecipitation process overcomes the limitation of the adsorption
surface area, promoting a high immobilization yield and favouring the dispersion of the
biomolecule within the solid phase [43, 44]. Charradi et al. [45] also observed the increase effect
of the immobilization yield of the heme protein hemoglobin on Mg2Al-LDH compounds
prepared by coprecipitation. However, a different result was observed after the immobilization of
transketolase on Mg2Al and Zn2Al-LDH inorganic supports, where the adsorption process
yielded to a high enzyme immobilization compared with the in situ-coprecipitation [38]. As
mentioned above, electrostatic or coulombic interactions due the opposite surface charge values
promoted the stable assembly of LDH enzymes and a high performance of NR immobilization.
The higher NR immobilization by adsorption displayed by the ZnAl-LDH sample may be related
to the different specific buffering properties due to a lower surface basicity of the sample ZnAl-
LDH compared with the MgAl-LDH matrix [17, 38, 46].
15

In parallel, the biohybrid performance toward mesotrione biodegration as a function of Q was
also investigated (Figure 3B). The free enzyme exhibited a specific activity for mesotrione
biotransformation of ≈ 25 mU/mg protein. Systematically, the enzyme immobilization induced a
decrease of the enzymatic activity. An activity loss compared with the free enzyme has been
commonly observed in non-specific adsorption processes of enzymes randomly bound to solid
surfaces, especially in the so-called “soft” proteins [17, 19, 45, 47-49]. However, a better
catalytic performance for the samples based on NR immobilized on MgAl-LDH matrices by
adsorption was systematically observed, showing an increase of the specific NR activity higher
than 4-fold compared with NR immobilized on the ZnAl-LDH phase at an enzyme/LDH initial
mass ratio Q = 1. As mentioned above, this trend could be due to the difference of surface
properties for these two LDH chemical compositions as already observed in the literature [17]. It
is noteworthy that the strategy for the NR immobilization had a greater influence in the final
performance of the biohybrids. The NR immobilized by coprecipitation, NR@MgAl-LDH-cop,
significantly decreased its activity performance compared with the NR@MgAl-LDH-ads,
probably due to the pH instability of this enzyme at pH = 9 used for the biohybrid synthesis.
Indeed, at this basic pH, the free enzyme activity dropped by 90 % after 3 h, meanwhile the free
enzyme fully retained its activity at pH 7.5 in 50 mM MOPS for several days. Conversely, the
LDH particle size had not a noticeable influence in the final performance of the biohybrids.
However, the data obtained using the MgAl-LDH sample, prepared by classical coprecipitation,
were poorly reproducible and rather heterogeneous, probably due to the greater particle size
distribution of this sample.
16

Figure 3. Immobilization yield (A) and activity performance (B) of the NR@LDH biohybrids:
NR@ZnAl-LDH-ads (black), NR@MgAl-LDH-ads (red horizontal lines), NR@MgAl-LDHnano
(blue vertical lines) and NR@MgAl-LDH-cop (grey) at different Q values (experiments run in
triplicate).
In summary, the results highlight the complexity of the immobilization process of enzymes on
LDH inorganic materials which strongly depends on the inherent properties of the enzyme-LDH
biosystem selected and the immobilization strategy which determine the interactions between the
17

two components. The NR immobilization on the MgAl-LDHnano matrix by adsorption at NR/LDH
initial mass ratio Q = 1 exhibited a good immobilization yield (80 %) and the best catalytic
activity (44 %) compared to free enzyme with a good reproducibility of the immobilization
process (RSD < 1.5 %). Consequently, the NR@MgAl-LDHnano system was further investigated
for a better understanding of the immobilization process and biohybrid performance.
3.2. Physico-chemical and structural characterization of NR@MgAl-LDHnano
Figure 4A displays the adsorption isotherm of the NR on the MgAl-LDHnano which corresponds
to an H-type shape, according to the classification reported by Giles et al. [50]. Similarly, the
enzymatic activity increased with the amount of immobilized NR, following a similar pattern. As
proposed in a previous study [19], an adequate description of the protein adsorption process on
LDH compounds can be obtained by the non-linear fitting of the adsorption isotherm to the
Freundlich model and the composite Langmuir-Freundlich model. Both models offered a good
fit to the adsorption isotherm data (Table 1). However, the Langmuir-Freundlich equation was
somewhat more adequate to describe the general process of NR adsorption, showing Cm values
close to those obtained experimentally. The Kf and Kd parameters are characteristic of systems
with strong interactions between both components involved in the adsorption process, i.e. the
enzyme and the support. The nfl values higher than 1 and the Nf values close to 0, clearly
suggested a positive cooperativity in binding and a heterogeneous nature of the enzyme
adsorption [51]. This may indicate that the increase of NR adsorbed on the LDH is favoured by
attractive forces due to lateral enzyme-enzyme interactions, as observed for another
enzyme/LDH system [19].
The “high affinity” isotherm could be attributed to the existence of electrostatic interactions,
considering the opposite charge values of both components at slightly basic experimental pH 7.5.
Indeed at this pH, the amount of NR adsorbed led to a marked negative progression on the
18

surface charge of the nanomaterial, from a positive zeta potential (ζ) value of the pristine LDH
(+35 mV) to ζ values close to that of the free enzyme (Figure 4B). Additionally, this charge
inversion, together with the increase of the biocatalytic activity with the enzyme loading,
demonstrated the successful immobilization of the NR on the MgAl-LDHnano sample.
Figure 4. A) Non-linear fit of the adsorption isotherm of NR on MgAl-LDHnano using the
Langmuir-Freundlich model (red curve; a), and NR specific activity (blue curve; b) and B)
evolution of the ζ with the enzyme loading (experiments run in triplicate). Red dot line: ζ value
of the free enzyme (50 mM MOPS, pH 7.5 and 25 °C).
Table 1. Freundlich and Langmuir-Freundlich parameters for NR adsorption on MgAl-LDHnano
Freundlich model
Kf Nf
0.69 ± 0.08 0.20 ± 0.08 0.92
Langmuir-Freundlich model
Cm Kd
0.59 ± 0.15 5316 ± 540 1.74 ± 0.55 0.99
R²
R²
nlf
The XRD pattern of the NR@MgAl-LDHnano biohybrid (Figure 1) confirms the preservation of
the well-crystallized LDH structure after the enzyme adsorption. The d spacing remains the same
19

compared to the pristine MgAl-LDHnano material, suggesting that nitrate is the main interlayer
anion and no intercalation of the NR occurred. Nevertheless, after the adsorption, the reflections
of the side Mg(OH)2 phase already observed in the precursor MgAl-LDHnano are still present
conjointly with two additional reflections at 14.4° and 24.0° which could correspond to the
partial intercalation of an anionic species from the buffer adsorption medium. Vibration bands
typical of both LDH-type and the enzyme structure are systematically observed by FTIR
spectroscopy in the NR@MgAlLDHnano biohybrid sample (Figure 5A), confirming the presence
of adsorbed enzyme on the LDH material. The characteristic broad stretching vibration band at
ca. 3500 cm-1 (νO–H) is attributed to OH groups of interlayer water molecules and the hydroxylated
brucite-layers. The vibration bands ranging between 400 and 800 cm-1 were assigned to the νM-O
and δO-M-O lattice vibrations. Additional bands characteristic of the νC-H and amide vibration bands
of the enzyme are present at 2750 cm-1 and 800-1700 cm-1. Noticeably, the intensity of the band
attributed to ν(NO3) at 1387 cm-1 in the biohybrid and LDH material slightly decreased but was
still present indicating that nitrate-intercalated LDH remains as the main LDH phase after the
immobilization of the enzyme which adsorbed at the particle surface rather than by intercalation
in the interlayer space. This result is in agreement with those observed previously for enzyme
adsorption using LDH matrices [20, 52].
We further investigated aqueous suspension of NR@MgAl-LDHnano biohybrids by fluorescence
spectroscopy and compared the spectra with that of the native enzyme in order to obtain
information about the immobilization effect on the structural compactness and unfolded state of
NR (Figure 5B). In fact, fluorescence characteristics of the aromatic amino acid residues of
enzymes, such as tryptophan (Trp) or tyrosine (Tyr), are valuable parameters to evaluate
conformational changes upon enzyme immobilization [53, 54]. When excited at λexc = 280 nm,
free NR displayed a maximum fluorescence emission at 339 nm, close to the maximum emission
20

produced by the amino acid residue Trp (350 nm). Once immobilized, the NR fluorescence
spectra changed with the enzyme loading into the NR@MgAl-LDHnano biohybrids. At low
enzyme loading values, the fluorescence yield decreased and a new peak at red wavelength
values was clearly observed. These results suggest a change in the local environment polarity of
the fluorophore residues of the immobilized enzyme which became more accessible to the
solvent molecules compared to the native NR, as a consequence of the compactness relaxation of
the enzyme structure. Similar results have been reported in the literature for the immobilization
of amylase for instance using LDH as support [15, 19, 55]. However at higher saturation loading
(Cs = 0.62 mg/mg), the emission spectrum was similar to that of the free enzyme, highlighting
that in the NR case, this effect was reduced at saturated enzyme loading values.
21

Figure 5. A) FT-IR spectra of the NR@MgAl-LDHnano biohybrid (Cs = 0.6 mg/mg) (a), free NR
(b), and pristine LDH (c). B) Fluorescence emission spectra of the colloidal suspension of
NR@MgAl-LDHnano biohybrids with different enzyme loading: Cs= 0.2 mg/mg (a), Cs = 0.4
mg/mg (b), Cs = 0.57 mg/mg (c), Cs = 0.62 mg/mg (d), and the free enzyme (e). Excitation at λexc
= 280 nm.
This likely modification in enzyme configuration due to its immobilization can have a
consequence on its catalytic activity. Therefore, we analysed the evolution of the enzymatic
activity of the biohybrids as a function of the enzyme loading (Cs) (Figure 6). As mentioned
above, the activity of the biohybrids increased with the increase of immobilized enzyme. This
increase reaches a maximum value of ca. 5 mU/mg biohybrid. As shown in the Figure 6 (blue
curve b), higher activity performance of the NR@MgAl-LDHnano biohybrids, defined as NR
specific activity (mU/mg NR) per mg of enzyme loaded on the support (Cs), was observed when
the saturation in terms of enzyme adsorption was achieved (Cs = 0.55 – 0.6 mg/mg). These
values correspond to initial concentrations of NR in solution higher than Ci > 0.8 mg/mL of NR.
22

It is well-known that the coverage of proteins non-specifically adsorbed on solid surfaces
frequently results in conformational and/or orientational changes within the adsorbed surfaces
affecting its bio-activity [51, 56]. After the attachment step, the biomolecule tends to optimize its
interactions with the surface of the LDH which usually leads to some degree of spreading
specially for the so-called “soft” proteins, inducing the conformational modification previously
observed by fluorescence spectroscopy on the NR@MgAl-LDHnano biohybrids (Figure 5B). This
energy minimization due to the enzyme spread across the surface may negatively affect the
biocatalytic performance, explaining the decrease of its efficiency at low load values. On the
other hand, the structure relaxation of the biomolecule strongly depends on the available
adsorption sites [57]. The number of adsorption sites is reduced with the incorporation of more
biomolecules to the surface and new enzyme-enzyme interactions arise in the system. Therefore,
the configuration set at high load values better preserved the native NR conformation and then
enhanced the performance of the biocatalyst [58].
For the α-amylase/LDH system [19] an optimal configuration was observed at average load
values but an enzyme aggregation occurred at high load values, decreasing the specific activity
of the biocatalyst. However, this multilayer adsorption effect is unlikely to occur with the
NR@MgAl-LDHnano system since a single plateau was observed on the adsorption isotherm
(Figure 4A).
23

Figure 6. Effect of NR loading on the specific enzyme activity (a) and performance (b) of the
NR@MgAl-LDHnano biohybrids (experiments run in triplicate).
3.3 Recyclability and storage stability of the NR@MgAl-LDHnano biohybrid
Usually in biohybrid materials, the immobilization of enzymes may allow their reuse and may
increase their catalytic productivity reducing the overall cost associated with the enzyme
production in industrial bioprocesses or biotechnology development. Thus, we explored here the
recovering and recycling effect on the catalytic performance of NR@MgAl-LDHnano biohybrid
(Cs = 0.6 mg/mg) toward the biodegradation of mesotrione. Figure 7A shows the evolution of the
relative activity of the NR@MgAl-LDHnano biohybrid during 4 cycles of reuse. Although a partial
loss of biocatalytic activity was observed at each step, from 80 % to 60 % of the catalytic activity
was retained with respect to the immediately previous cycle, achieving ca. 30 % of activity
retained after 4 cycles of reuse. Since no appreciable release of enzyme was detected in the
supernatant between each cycle, such a loss of enzymatic activity could be related to a NR
denaturation during the recycling process.
24

Figure 7. A) Recyclability of the NR@MgAl-LDHnano biohybrid (Cs = 0.6 mg/mg) using 100 M
of mesotrione in 50 mM MOPS pH 6.5 + 0.7 mM NADH, and 40 min of reaction time by cycle
(experiments run in triplicate). B) Relative NR activity with time of the lyophilized extract of the
free native NR (black squares), a suspension of NR@MgAl-LDHnano biohybrid in 50 mM MOPS
(red circles) and the lyophilized NR@MgAl-LDHnano biohybrid (blue triangles). Samples were
stored at 4 °C.
Another important point concerns the storage of biohybrid materials in the dry state. It
provides several advantages for enzyme-based materials such as preventing enzyme leakage in
long-term storage, avoiding the degradation of the enzyme by microbiological activity or
25

undesirable side-reactions, and improving the handling and transportation. For that, we evaluated
the activity of NR@MgAl-LDHnano biohybrid after lyophilisation (Figure 7B). The lyophilized
form of this biohybrid retained full activity after its rehydration in comparison with its un-dried
form. This result suggests that the different steps involved in the lyophilization treatment did not
affect significantly the performance of the biohybrid material. Furthermore, the storage stability
of the immobilized enzyme with time was investigated in comparison with that of the lyophilized
free native enzyme. NR@MgAl-LDHnano biohybrids were studied in both stored forms at 4 ºC,
that is, as a suspension in MOPS (pH 6.5) and as a lyophilized powder. According to the results,
the biohybrid in suspension retained its activity during 15 days, but a gradual activity loss was
detected after. Nevertheless, the activity was fully maintained with the lyophilized biohybrid, at
least for more than one month of storage. Surprisingly, an increase in the enzymatic activity was
systematically observed after 10 days of storage for both lyophilized NR and lyophilized
NR@/MgAl-LDHnano biohybrid and even for the biohybrid maintained in suspension. Such
behaviour could be link to a slight structural change of the NR in solution with time, further
experiments will be necessary to confirm it.
3.4. Electrochemical response of NR@MgAl-LDHnano modified electrode
In order to test the suitability of NR-based biohybrids for biotechnology applications, such as
biosensors, thin films were prepared using the NR@MgAl-LDHnano biohydrid through a one-step
deposition on a polished glassy carbon electrode. The surface morphology of the biohybrid film
was explored by SEM and AFM measurements (Figure 8). Clearly, the surface of the film is
homogeneous and continuous. The SEM high magnification image and AFM top view image
allow us to distinguish the individual LDH nanoparticles involved into the dense biohybrid film
with particle size in good agreement with the results obtained by DLS.
26

A
1
100
Figure 8. A) FE-SEM and B) AFM images of the NR@MgAl-LDHnano film.
Enzymatic activity of the NR@MgAl-LDHnano films was analysed by cyclic voltammetry (CV).
In fact, the first step of the enzymatic mesotrione transformation by NR, in the presence of its
cofactor NADH, leads to the formation of a hydroxylamino intermediate [26], which can be
detected in electrochemistry at a potential of 0.3 V/SCE [59]. AMBA, the final metabolite of the
enzymatic mesotrione degradation [26] can also be oxidized at a GCE (See supporting
information Figure S4). Unfortunately, its oxidation potential (1.1 V/SCE) is more anodic than
the oxidation potential of the NADH cofactor (0.7 V/SCE) [60]. Therefore, the electrochemical
detection of AMBA cannot be taken under consideration in NR-based amperometric biosensors.
The potential window for CV was therefore fixed between -0.3 and +0.5 V/SCE to prevent, with
the lower potential limit, the direct electrochemical reduction of mesotrione and, with the higher
limit, the electrochemical oxidation of NADH. Figure 9 shows the evolution of the CV curves
when 0.4 mM NADH was added in a 0.2 mM mesotrione solution. The appearance of a small
oxidation peak at +0.30 V/SCE, corresponding to the formation of hydroxylamino intermediate,
confirms that the enzymatic reaction takes place at the modified electrode surface. However, the
peak intensity was very low (0.89 µA cm-2) for a 0.2 mM mesotrione concentration, evidencing a
NR@MgAl-LDHnano modified electrode not very sensitive to detect mesotrione, mainly due to the
27

low specific activity of the enzyme for this substrate (25 mU/mg and KM = 68 µM) [26]. In the
literature, it was reported that nitroreductases are effectively reducing various nitro-compounds
[61]. For instance, we found that NR has a specific activity of 2.99 U/mg and a KM value of 21
µM for dinitrotoluene (DNT). Therefore, DNT was a better substrate to test the electrochemical
activity of the NR@MgAl-LDHnano modified electrode. Figure 7B clearly shows a catalytic
oxidation peak of 11 µA cm-2 at +0.15 V/SCE when 0.2 mM DNT was reduced using this
electrode in the presence of NADH. Similarly, Caygill et al. [62], described that the presence of
nitroreductase NfsA from E.coli into the electrolyte improves the electrochemical detection of
TNT and DNT at a screen–printed electrode, due to an increase of current magnitude of the
oxidation peak of hydroxylamine.
Figure 9. Cyclic voltammograms of (A) 0.2 mM mesotrione and (B) 0.2 mM DNT at
NR@MgAl-LDHnano/GCE without (a) and in presence of 0.4 mM NADH (b) (0.1 M PBS pH 6.5,
v = 10 mV/s).
4. CONCLUSIONS
28

Biohybrids containing nitroreductase (NR) and layered double hydroxide (LDH) particles were
obtained by adsorption process and direct coprecipitation. Two distinct LDH chemical
compositions were tested by modifying the nature of the divalent metal (Mg2+ or Zn2+) involved
within the layer while the LDH particle size and aggregation were varied according to the LDH
synthesis process used. Analysis of the NR immobilization percentage and the retained activity
revealed that the NR adsorbed on LDH nanoparticles (noted NR@MgAl-LDHnano) in a ratio
NR/LDH of Q = 1 gave the best results with 80 % of immobilized NR and 44 % of retained
activity in comparison with the free native enzyme. The adsorption isotherm evidenced a strong
affinity between the enzyme and the LDH nanoparticles further supported by the TEM images
and zeta potential evolution. Remarkably, the highest activity performance of NR@MgAl-
LDHnano biohybrids was observed at the saturation plateau corresponding to Cs ~0.6 mg/mg, due
to a better conservation of the native conformation of the immobilized enzyme. After 4 cycles of
reuse, the NR@MgAl-LDHnano biohybrid still retained 30 % of its initial biological activity,
meanwhile the biological activity was fully maintained using the lyophilized biohybrid, at least
after one month of storage. This work represents the first example of using LDH to immobilize
NR leading to smart functional biohybrid with promising applications in diverse fields.[63]
Our preliminary results using CV evidenced that the enzymatic reaction takes place at an
electrode surface modified by NR@MgAl-LDHnano. While NR@MgAl-LDHnano modified
electrode was efficient to detect DNT, it appears not very sensitive for mesotrione due to the low
specific activity of NR for this substrate. Further studies are currently under investigation to find
another transduction strategy to develop a biosensor for mesotrione, based on NR@MgAl-
LDHnano more sensitive than sensors using clay based materials.[59] Since NR@MgAl-LDHnano
modified electrode is efficient to detect DNT, it should also be of great interest for detection of
29