You are on page 1of 8

Colloids and Surfaces B: Biointerfaces 208 (2021) 112030

Contents lists available at ScienceDirect

Colloids and Surfaces B: Biointerfaces


journal homepage: www.elsevier.com/locate/colsurfb

Silver nanoparticle-protein interactions and the role of lysozyme as an


antagonistic antibacterial agent
M. Beatriz Espeche Turbay a, b, *, Valentina Rey a, b, Rita D. Dorado a, Marcelo C. Sosa a,
Claudio D. Borsarelli a, b, *
a
Instituto de Bionanotecnología del NOA (INBIONATEC), CONICET, Universidad Nacional de Santiago del Estero (UNSE), RN9, km 1125, G4206XCP, Santiago del
Estero, Argentina
b
ICQ – Facultad de Agronomía y Agroindustrias, UNSE, Av. Belgrano (S) 1912, Santiago del Estero, Argentina

A R T I C L E I N F O A B S T R A C T

Keywords: The photoreductive synthesis and antibacterial activity of silver nanoparticles (AgNP) prepared in the presence
Silver nanoparticles of bovine serum albumin (BSA) and lysozyme (LZ) were evaluated. AgNP@BSA showed similar antibacterial
Antibacterial activity activity to those stabilized with citrate (AgNP@CIT) and to an AgNO3 solution, suggesting the releases of Ag+ as
Antagonistic effect
the mechanism of death. In contrast, AgNP@LZ solutions showed no activity, although LZ behaves as a
Protein corona
moderately antibacterial peptide. Furthermore, the addition of LZ to the AgNP@CIT or AgNP@BSA solutions
Lysozyme-silver interactions
induced their agglomeration and suppressed their original antibacterial efficacy. This antagonistic antibacterial
effect exerted by LZ on AgNPs is associated with electrostatic interactions exerted by LZ. Specific metal-LZ in­
teractions produce a harder protein corona on AgNP@LZ that retains Ag+, while LZ acts as a glue for AgNP@CIT
or AgNP@LZ due to its opposite electrical charge, besides strong binding to Ag+avoiding the bactericide effect.
Therefore, bactericidal effects of AgNP in biological media may be modulated by specific protein interactions.

1. Introduction materials have been used for anti-infective treatments since ancient
times, basically by non-specific toxic effects exerted by the release of
The irresponsible use of antibiotics has contributed to the prevalence silver ions, Ag+, and the overproduction of reactive species oxygen
of super-resistant microorganisms capable of surviving classical drug (ROS), minimizing the possibility that pathogens can develop resistance
treatments, even those of the new generation, causing considerable mechanisms [11].
economic losses and deaths [1,2]. Hence, infectious diseases represent However, in biological environments, the presence of other bio­
one of the most dangerous health threats to overcome due to the molecules can modulate the antibacterial efficacy of AgNPs [12,13]. For
persistence of resistant pathogens [2]. example, proteins are the most abundant components in biological
Therefore, the search and development of new approaches to control fluids, so the understanding of protein-nanoparticle interactions is very
contaminations and infections with mechanisms of action that minimize important, considering that the formation of a protein layer or corona
or avoid the development of antimicrobial resistance are crucial [3]. around the nanoparticle acts as an active biointerface between the
Regarding this, recent advances in nanotechnology point to new and nanocores and cells [12–14]. The effect of the protein corona around
innovative alternative therapies that employ nanoparticles (NP) as AgNPs using various proteins (e.g., ubiquitin, collagen, human or bovine
carriers for antibiotics or self-antimicrobial agents to reduce the toxic serum albumin, blood plasma, yeast extracts, and fetal calf) on the
effect of drugs, as well as to increase antimicrobial drug concentrations cytotoxicity and antimicrobial effects of the bioconjugates have been
at the target site [4,5]. Among the multiple nanotechnological appli­ evaluated [12,15–18]. The data reported indicate that the properties
cations of silver nanoparticles (AgNPs) [6], their use for the control and and functionalities of the protein corona depend on several factors, such
treatment of bacterial infections, whether due to planktonic or biofilm as the size and/or shape of the nanoparticles, the concentration of
proliferation, represents an emerging field [7–10]. Silver-based proteins, the functionalization of the surface, etc., while some

* Corresponding authors at: Instituto de Bionanotecnología del NOA (INBIONATEC), CONICET, Universidad Nacional de Santiago del Estero (UNSE), RN9, km
1125, G4206XCP, Santiago del Estero, Argentina.
E-mail addresses: beaespeche@gmail.com (M.B. Espeche Turbay), cborsa@unse.edu.ar (C.D. Borsarelli).

https://doi.org/10.1016/j.colsurfb.2021.112030
Received 26 April 2021; Received in revised form 30 June 2021; Accepted 7 August 2021
Available online 10 August 2021
0927-7765/© 2021 Elsevier B.V. All rights reserved.
M.B. Espeche Turbay et al. Colloids and Surfaces B: Biointerfaces 208 (2021) 112030

unexpected interactions of the protein corona with surrounding mole­ Fluorescence spectra were obtained with an Agilent Cary Eclipse
cules could modify the antimicrobial activity [12–14,19]. instrument.
Regarding this, here we report the effect of globular proteins such as Fourier-transform infrared (FTIR) spectra were acquired at room
bovine serum albumin (BSA) and lysozyme (LZ) on the formation and temperature with a Jasco FTIR-4600 spectrometer equipped with the
antibacterial activity of AgNPs prepared by a photoreduction reaction ATR PRO ONE single reflection accessory under dry airflow to eliminate
using Irgacure 2959 as a ketyl radical photoinitiator [20,21]. The for­ water vapor distortions in the spectra. Four spectra for each sample were
mation and specific protein interactions of the bionanocomposites averaged and processed with the Spectra Analysis™ software provided
AgNP@BSA and AgNP@LZ, together with AgNP solutions stabilized by Jasco.
with citrate ions (AgNP@CIT), were monitored by various techniques Raman spectra were recorded at room temperature with a Thermo
(UV–vis, fluorescence, FTIR, Raman, dynamic light scattering, zeta po­ Scientific DXR Raman instrument equipped with a 532 nm CW laser of
tential, and transmission electron microscopy) [21], while the antibac­ 10 mW as the excitation source. The samples were placed on gold-coated
terial activity of the different AgNPs was evaluated in Gram-positive and sample slides and 100 acquisitions were accumulated with an exposure
Gram-negative bacteria using the zone of inhibition and minimum time of 5 s for all samples to improve the signal-to-noise ratio.
inhibitory concentration methods [22].
The present results demonstrate that LZ, a recognized antibacterial 2.4. AgNPs characterization
protein [23], induces the loss of the antibacterial efficacy of AgNPs, by
exerting intense specific interactions, either as a protein corona on the Hydrodynamic diameter (dh) and zeta-potential (ζ) of the different
silver nanocore of AgNP@LZ or by attaching to the coating layer of AgNP solutions were determined with the Horiba SZ-100 particle size
AgNP@BSA or AgNP@CIT producing their agglomeration. This unan­ analyzer using a 532 nm CW laser, focused at 173◦ relative to the
ticipated antagonistic antibacterial effect is associated with the strong detection device. The data were analyzed with the Horiba NextGen
binding of the silver ions Ag+ to LZ, which dramatically reduces the Proyect™ SZ-100 software.
intrinsic toxic effect of AgNPs on the bacteria. To our knowledge, there Nanoparticle images were obtained with a Zeiss Libra 120 TEM at the
are still few studies on understanding the physicochemical aspects of Integral Center for Electron Microscopy (CIME) of CONICET in
protein-metal/ion interfacial interactions that can modulate the bio­ Tucumán, Argentina. A diluted aliquot of the sample solution was
logical activity of AgNPs [13,17]. In this sense, the present results may evaporated on a carbon-coated copper mesh screen. Mean size with
help to design nanomaterials with greater antimicrobial efficacy in standard deviation (SD) was calculated from at least four different re­
biological media. gions in randomized fields of view using the ImageJ™ free software.

2. Experimental section 2.5. Antibacterial activity

2.1. Materials and strains The viability tests of the strains mentioned above were performed
using the techniques recommended by the CLSI (https://clsi.org/s
Silver nitrate (AgNO3), sodium citrate dihydrate (C6H5Na3O7⋅2H2O), tandards/). The agar diffusion method was used to detect the antimi­
sodium borohydride (NaBH4), 2-hydroxy-1-[4-(2-hydroxyethyl) crobial effect by placing a 50 μL aliquot of each sample in the well while
phenyl]-2-methyl-1-propanone (Irgacure®, I-2959), lysozyme (from the minimum inhibitory concentration (MIC) was quantified using the
chicken egg white, approximately 58,100 units/mg, LZ) and bovine macrodilution technique [22].
serum albumin (≥98 %, BSA) of analytical grade were purchased from
Sigma-Aldrich Argentina and used without further treatment. The so­ 2.6. Lysozyme (LZ) lytic activity assay
lutions were prepared with ultra-pure water, provided from a MiliQ
system. Compressed ultrapure nitrogen (99.99 %) was purchased from The lytic activity of LZ towards the cell walls of Micrococcus lyso­
Indura Argentina. Staphylococcus (S.) aureus (ATCC 25923), Escherichia deikticus (ATCC 4698) (Sigma-Aldrich) at different concentrations of
(E.) coli O157H7, Serratia (S.) marcescens SmP9 and Klebsellia (Kb.) AgNO3 was monitored by the changes in the optical density (OD) of a
pneunomiae CR1 strains were kindly supplied by CERELA-CONICET (S. suspension of 0.15 mg.mL− 1 of the microorganism dispersed in ultra­
M. de Tucumán, Argentina). SmP9 and CR1 were isolated from clinical pure water.25
samples and displayed multidrug-resistant (MDR) phenotypes [24].
Bacterial growth medium, Luria-Bertani (LB) medium, and antimicro­ 3. Results and discussion
bial sensitivity test were performed on Mueller-Hinton (MH) broth, and
agar was purchased from Laboratorios Britania, Argentina. 3.1. Photochemical synthesis of AgNPs capped with BSA and LZ

2.2. Synthesis of silver nanoparticles Fig. 1 shows the UV–vis spectra of the surface plasmon band (SPB) of
AgNPs obtained after full reduction of Ag+ ions with ketyl radicals
Silver nanoparticles aqueous solutions stabilized with citrate generated by photo-fragmentation of the non-toxic benzoin I-2959 in the
(AgNP@CIT), lysozyme (AgNP@LZ), or bovine serum albumin presence of 1 mM citrate (AgNP@CIT), 10 μM LZ (AgNP@LZ), and 1 μM
(AgNP@BSA) were prepared by photoreduction of AgNO3 with I-2959 BSA (AgNP@BSA), respectively. The concentration of each protein used
as previously reported [21], and is briefly described in section one of was minimal to obtain stable nano-colloidal solutions of AgNP@LZ [21]
Supplementary Material (SM1). AgNP solutions were stored at refrig­ or AgNP@BSA [26], respectively. In all cases, the position of the
erator temperature and their colloidal stability was confirmed by maximum wavelength (λmax) of SPB was located at <410 nm, Table 1],
monitoring the characteristic surface plasmon band (SPB) of AgNP be­ which suggests the formation of small spherical AgNPs with a core
tween 300 and 700 nm by UV–vis spectroscopy as described below. In all diameter less than 10 nm [27]. Red-shift and broadening of the SPB of
cases, almost no changes in absorbance of the SPB were observed after AgNP@LZ and AgNP@BSA relative to that for AgNP@CIT are typical for
one week of preparation of the solutions. protein-stabilized AgNPs [15,16,28–33], due to dielectric changes pro­
duced by the formation of a protein corona that surrounds the metalcore
2.3. Spectroscopic measurements [12,14].
Table 1 also shows the hydrodynamic diameter (dh) and zeta po­
UV–vis absorption spectra were registered using either an Ocean­ tential (ζ) values for the prepared AgNPs. For AgNP@CIT, dh = 3 ± 1 nm
Optics USB2000 or Hewlett Packard 8453 spectrophotometers. was similar to dAgNP = 4 ± 1 nm as previously measured by TEM [21], as

2
M.B. Espeche Turbay et al. Colloids and Surfaces B: Biointerfaces 208 (2021) 112030

The slower growth kinetics for the photoreduction reaction in the


presence of proteins than with citrate also provides evidence of the
formation of a protein corona around the AgNPs (inset of Fig. 1). The
sigmoidal kinetic curves can be interpreted according to the
Kolmogorov-Johnson-Mehl-Avrami (KJMA) or aggregate growth model,
Eq. (1), where kg is the first-order rate constant of particle growth and
nAv is the Avrami exponent [37].
[ ( )n ]
AUCt = AUC∞ 1 − exp − kg t Av (1)

In general, kg is strongly associated with the nanoparticle growth rate


rather than nucleation rate, while nAv is related to the nucleation
mechanism and shape of the growing nanoparticle [37]. The calculated
nAv values are shown in Table 1 and are in the typical range expected for
radical-mediated growth of spherical metal nanoparticles [21,37]. On
the other hand, the kg value for AgNP@CIT was almost two and four
Fig. 1. UV–vis absorption spectra of silver nanoparticle aqueous solutions times higher than for AgNP@BSA and AgNP@LZ, respectively, indi­
stabilized with 1 mM citrate (AgNP@CIT), 1 μM bovine serum albumin cating the role of the protein corona as a steric barrier for the growth
(AgNP@BSA), and 10 μM lysozyme (AgNP@LZ) prepared by photochemical pathway. Interestingly, the kg value for AgNP@BSA was higher than for
reduction of 0.2 mM AgNO3 in the presence of 0.2 mM Igarcure 2959 (I-2959) AgNP@LZ, although BSA is almost five times as bulky as LZ. Previously,
at room temperature. Inset: kinetic growth curves of the area under the curve we have shown that kg was independent of the initial concentration of LZ
(AUC) for the formation of the surface plasmon band (SPB) of the AgNPs. Solid used for the photochemical synthesis [21], and therefore, the larger kg
black lines represent the non-linear fitting with the aggregation-growth or
value in the formation of AgNP@BSA may suggest that BSA has a more
KJMA model, Eq. (1).
flexible structure than LZ, favoring its adsorption on the metal nanocore.

can be expected for a nanoparticle capped with a small molecule. On the 3.2. Lysozyme effect on the antibacterial activity of the AgNPs
other hand, for AgNP@LZ and AgNP@BSA, an average dh = 13.5 ± 4.0
nm was obtained, in agreement with the values previously reported for Table 2 shows the diameters of the zone of inhibition (mm) observed
these nanoparticles [21,31]. Then by comparison with the TEM di­ after overnight growth at 37 ◦ C of the bacterial strains containing 50 μL
ameters [31,33], the thickness of the protein corona (rcorona) was esti­ aliquot of AgNP or control solutions, as shown in the inset of Fig. 2a.
mated ≈ 3.5 nm and 4.5 nm for AgNP@LZ and AgNP@BSA solutions, Except for Kb. pneunomiae, which is a strain intrinsically resistant to the
respectively. These rcorona values can be expected for a monolayer of LZ toxic activity of Ag+ ions [38], both AgNP@CIT and AgNP@BSA solu­
corresponding to an elongated ellipsoid of 3 × 3 × 4.5 nm [34], and for a tions showed efficient antibacterial activity against the rest of the
monolayer of BSA with an equilateral triangular prism shape of 8.4 × 8.4 strains, and similar to that obtained with a 0.2 mM AgNO3 solution
× 8.4 × 3.2 nm [35], respectively. Therefore, the diameter of the taken as the positive control for the toxic effect of Ag+ ions [7,39].
spherical metal core can be approximated to dAgNP = dh –2×rcorona, that Interestingly, AgNP@LZ solutions prepared with 1.4 μM or 10 μM LZ
is, ≈5 nm and ≈6 nm for AgNP@BSA and AgNP@LZ, respectively. These were completely ineffective in inhibiting the growth of all bacteria, even
dAgNP values were used to calculate the AgNP concentration of each though a 10 μM LZ solution was able to partially prevent the growth of
solution as explained in the supplementary material (section SM2), S. aureus. On the other hand, aliquots of citrate, I-2959, or BSA solutions
assuming the complete conversion of the precursors according to the added at the same concentrations used for the preparation of AgNP so­
plateau observed in the growth kinetics of the photoreduction reaction lutions did not show antibacterial activity, Table 2.
(inset of Fig. 1). The calculated concentrations are reported in Table 1 Taking into account that at neutral pH both Gram-positive and Gram-
and were of the same order of magnitude as those calculated using the negative bacteria show negative ζ values [36,40], the toxic effect of
Lambert-Beer law with a molar extinction coefficient of ≈2 × 108 M− 1 AgNP@CIT and AgNP@BSA, both with negative surface charge, may be
cm− 1 for AgNP@CIT with diameters <10 nm, as reported by Paramelle associated with the release of Ag+ ions from the silver nanocore, which
et al. [27]. in aerobic conditions is favored by the dissolution of Ag2O formed on the
The ζ values of the photochemically synthesized AgNPs were similar surface of the nanoparticle [7,41]. To confirm this mechanism of bac­
to those previously reported for each nanocolloid prepared by different terial death, the inhibition zone produced by the different AgNPs on
methods [31,32]. The large negative value ζ = − 45 mV for AgNP@CIT is S. aureus was analyzed as a function of the NaCl concentration, Fig. 2a.
associated with the complete ionization of the citrate carboxylic groups The results indicated that at ~100 μM NaCl the antibacterial activity of
at neutral pH. The different ζ values observed for AgNP@BSA (− 19 mV) both AgNP@CIT and AgNP@BSA was lost, while for AgNP@LZ the in­
and AgNP@LZ (+60 mV), confirm the formation of a protein corona crease in ionic strength did not modify the initial absence of activity. It is
around the metal core that stabilizes the nanocolloid [12,14]. The sign noteworthy that this threshold value for the complete loss of antibac­
change of the zeta potential between AgNP@LZ and AgNP@BSA is terial activity was coincident with the minimal inhibitory concentration
consistent with the net surface charges of each native protein at neutral (MIC = 100 μM) elicited by AgNO3 on both E. coli and S. aureus (Table S1
pH, due to isoelectric points of 11 and 4.8 for LZ and BSA, respectively of SM).
[36]. It is well-known that chloride ions (Cl–) induce dissolution and

Table 1
Some properties of AgNP solutions obtained by UVA photolysis of aqueous solutions of 0.2 mM AgNO3 and 0.2 mM I-2959 in presence of citrate (CIT), lysozyme (LZ),
and bovine serum albumin (BSA) as capping agents. Check the text for the meaning of abbreviations.
3
Capping agent λmax (nm) Fwhm (nm) dh (nm) dAgNP (nm) [AgNP] (nM) ζ (mV) kg/10− (s− 1) nAv
a
1 mM CIT 390 ± 1 63 ± 1 3±1 4±1 10.2 ± 1.5 –45 ± 5 11.0 ± 3.0 2.1 ± 0.2
1 μM BSA 409 ± 1 76 ± 1 14 ± 4 5±2 5.2 ± 0.9 –19 ± 2 5.0 ± 0.1 1.1 ± 0.1
10 μM LZ 409 ± 1 73 ± 1 13 ± 4 6±2 3.0 ± 0.5 +60 ± 1 2.9 ± 0.1 1.4 ± 0.1
a
From TEM measurements [21].

3
M.B. Espeche Turbay et al. Colloids and Surfaces B: Biointerfaces 208 (2021) 112030

Table 2
Inhibition zone (mm)a for the inactivation of Gram-positive and Gram-negative bacteria produced by the addition of 50 μL of AgNP solutions stabilized with citrate
(CIT), lysozyme (LZ), or bovine serum albumin (BSA), and by different control solutions. [AgNPs] are those indicated in Table 1.
AgNP solution Control solution

Strains AgNP@CIT AgNP@BSA AgNP@LZb AgNP@CIT +10 AgNP@BSA +10 0.2 mM 0.2 mM I- 1 mM 10 МM 1 μM
МM LZ МM LZ AgNO3 2959 CIT LZ BSA

S. aureus 8.5 9.0 n.d. n.d. n.d. 9.0 n.d. n.d. 6.0 n.d.
E. coli 7.0 10.0 “ “ “ 7.0 “ “ n.d. “
S. marcescens 9.0 10.0 “ “ “ 8.0 “ “ “ “
K. pneumoniae n.d.c n.d. “ “ “ n.d. “ “ “ “
a
Relative standard deviation ±6 %.
b
Prepared either with 1.4 μM or 10 μM lysozyme solutions.
c
n.d. = not detected.

aggregation phenomena of AgNPs by stoichiometric reaction with the


released Ag+ to form insoluble AgCl(s) with a Ksp = 1.8 × 10− 11 [42].
This effect can be observed by the spectral changes of the SPB of the
AgNPs as a function of the NaCl concentration, as is shown for
AgNP@CIT in Fig. 2b. Each spectrum was recorded after 10 min of
addition of the NaCl aliquot, ensuring a constant change in the absor­
bance of the SPB. The strong decrease in SPB absorption together with
the increase in the spectral baseline due to the light scattering effect
confirms the dissolution of AgNP@CIT with the formation of insoluble
colloidal particles which is reflected in the color discoloration of the
solution (inset of Fig. 2b).
Comparison of the relative values of the area under the curve (AUC)
of the PBS of the AgNPs as a function of the NaCl concentrations in­
dicates that the saline effect was greater for AgNP@CIT, milder for
AgNP@BSA, while it was almost absent for AgNP@LZ, Fig. 2c. Consid­
ering the longer incubation time used for the antibacterial experiments
than for the spectroscopic ones, to confirm the differential release of
Ag+, the absorbance changes in the SPB of the aqueous solutions of the
different AgNPs were monitored during 44 h of incubation at 37 ◦ C,
observing redshifts of the SPB maximum of 22 nm, 5 nm and 0.8 nm for
AgNP@CIT, AgNP@BSA, and AgNP@LZ, respectively. These spectral
changes are produced by the adsorption of dissolved Ag+ on the surface
of AgNPs, which amount is proportional to the concentration of Ag+
released by the nanoparticle, and the concentration of adsorbed ions
[Ag+]ad can be estimated by Eq. (2) [43].
[( ) ]
2
λ
+
[Ag ]ad = [Ag]0 − 1 (2)
λ0

λ0 and λ are the absorption wavelength of the maximum SPB observed at


the initial and final incubation times, respectively, while [Ag]0 repre­
sents the initial silver concentration in the colloidal solution, which is
assumed to be the same as that of the precursor (0.2 mM AgNO3). Thus,
[Ag+]ad was estimated as 23.2 μM for AgNP@CIT, 4.9 μM for
AgNP@BSA, and 0.8 μM for AgNP@LZ, confirming the trend in Ag+
release capacity as AgNP@CIT > AgNP@BSA>> AgNP@LZ.
These results confirm that the antibacterial mechanism of AgNP@­
CIT and AgNP@BSA is mainly due to the slow release of Ag+ in the
culture media, where both types of AgNP act as reservoirs of silver ions
[7], while the supply of Ag+ in AgNP@LZ is blocked by its protein
corona. A blocking effect for the release of Ag+ ions has been also re­
ported for metallic silver surfaces with adsorbed LZ at a concentration
greater than ≈70 μM, and was explained by the association of the silver
ions in the adsorbed protein layer together with electrostatic repulsion
Fig. 2. a) Effect of NaCl concentration on the inhibition diameter for S. aureus
elicited by adding 50 μL of AgNP@CIT, AgNP@BSA, and AgNP@LZ solutions.
effects on the transport of Ag+ through the positively charged pro­
The arrow indicates the minimal inhibition concentration (MIC) of AgNO3 for tein/solution interface preventing their exit into the bulk solution [44].
S. aureus. Inset: Agar well diffusion test for S. aureus in the absence and presence Table 2 also indicates that the addition of 10 μM of LZ to the
of NaCl as a function of increasing amounts (μL) of AgNP@CIT. b) NaCl effect AgNP@CIT or AgNP@BSA solutions resulted in the total loss of the
on the surface plasmon band (SPB) of AgNP@CIT solution. Inset: color fading original antibacterial activity of both types of nanoparticles. In contrast,
produced in the AgNP@CIT solutions by NaCl. c) Variation of the relative area the addition of 1 μM BSA to AgNP@CIT or AgNP@LZ solutions did not
under the curve (AUC) of the SPB of different AgNPs solutions with the con­ modify the antimicrobial activity of each nanoparticle solution.
centration of NaCl.

4
M.B. Espeche Turbay et al. Colloids and Surfaces B: Biointerfaces 208 (2021) 112030

Furthermore, the loss of the antimicrobial effect induced by LZ to the Instead, the TEM image inserted in Fig. 3c confirms that the addition
AgNP@CIT and AgNP@BSA solutions was parallel to the spectral of LZ to AgNP@CIT solution produces large aggregates of dh ≈1 μm, due
modifications of the SPB observed for each nanoparticle, as shown in to the agglomeration of small individual AgNPs rather than Ostwald’s
Fig. 3a for AgNP@CIT. nucleation-coalescence process to produce larger particles [37]. The
As the LZ concentration increased, a continuous broadening and agglomeration effect of AgNP is driven by the strong reduction of ξ value
redshift occurred along with a reduction in the absorbance of the SPB. for negatively charged nanoparticles by positively charged LZ molecules
These spectral changes are parallel to a strong decrease of the negative ξ (ξ ≈+7 mV at neutral pH) [46], as shown in the inset of Fig. 3a for
value of the colloidal solution (inset of Fig. 3a). In contrast, the addition AgNP@CIT. Therefore, LZ acts as an “electrostatic glue” of negatively
of BSA to the AgNP@CIT solution almost did not modify either the SPB charged nanoparticles.
spectrum or ξ value (Fig. 3b).
Moreover, a different effect on the dh of AgNP@CIT was produced by 3.3. About the antagonistic antibacterial effect exerted by LZ
the addition of LZ or BSA (Fig. 3c). BSA only slightly increased the
particle dispersibility of the AgNP@CIT solution, probably due to the Taking into account the intrinsic antibacterial function of AgNPs [7]
presence of small aggregates of protein-AgNP@CIT and/or protein di­ and lysozyme [23], respectively, the present results suggest a surprising
mers, considering dh ≈ 7 nm for monomeric BSA [45]. antagonistic antibacterial effect exerted by LZ in solutions of negatively
charged AgNPs. An antagonistic effect occurs between two or more
molecules when the functionality of the mixture is weaker or absent than
that given by the sum of the individual contributions [47,48].
In the present case, the antagonistic antibacterial effect may be
associated with strong specific interactions of LZ with silver, which may
affect the release of the toxic species Ag+ into the medium. To explore
this possibility, Fig. 4 compares the normalized infrared absorption of
the Amide I band (1700− 1600 cm− 1) of each native protein with that for
AgNP solution stabilized with the same amount of protein. This band is
typically used to monitor changes in protein secondary structure, and for
typical globular proteins like LZ and BSA, the Amide I band shows a
narrow Gaussian shape with a maximum at 1658 cm-1 that is charac­
teristic of a predominant α-helix protein conformation [49,50]. In
contrast, for the AgNP@BSA and AgNP@LZ solutions, the Amide I band
of LZ was distorted showing an increase in absorbance around 1601
cm-1, 1625 cm-1. and in the region between 1675 and 1725 cm -1, as
shown in each differential spectrum of Fig. 4. These modifications are
evidence of a relative increment in the β-sheet content (1625 cm-1) and
of extended chains (1675− 1725 cm-1) as the result of a partial dena­
turation of the protein after the formation of the protein corona adsor­
bed on the silver nanocore [49–51].
Comparison of the Raman spectral changes of each native protein
and its AgNP@protein solution provides information on the strength of
the protein-metal core interaction, Fig. 5a and b. Native proteins showed
relatively low-intensity Raman spectra with a maximum at 1640 cm− 1

Fig. 3. a,b) UV–vis spectral changes produced in 10.2 nM AgNP@CIT solution


upon addition of increasing amounts of native LZ or BSA, respectively. Insets:
Variation of the zeta potential (ξ) with the protein concentration. c) Hydro­
dynamic diameter (dh) of AgNP@CIT solutions in the absence and presence of 1 Fig. 4. a, b) Normalized ATR-FTIR spectra of the native protein (10 μM LZ and
μM BSA or 10 μM LZ, respectively. Inset: TEM image of AgNP@CIT obtained 20 μM BSA) and of the respective AgNP@Protein colloidal solution, together
after the addition of 10 μM LZ. with the calculated differential absorption spectra (ΔA = AAgNP – AProtein).

5
M.B. Espeche Turbay et al. Colloids and Surfaces B: Biointerfaces 208 (2021) 112030

due to the mixing of the characteristic Amide I band (1660 cm− 1) and therefore, LZ can produce a harder protein corona on AgNPs than BSA.
the vibratory stretching of aromatic amino acids (Trp/Tyr) around 1615 Ultracentrifugation experiments for LZ and AgNP@LZ solutions using a
cm− 1, together with a shoulder at 1400 cm− 1 associated with the sym­ 20 kDa cut-off filter combined with absorption and fluorescence detec­
metric stretching of carboxylic residues (COO–). Moreover, the weakest tion confirmed that only the solution of native LZ passed completely
band at 815 cm− 1 indicates the hydrogen bonding interactions of aro­ through the cut-off filter, but no free protein was collected for the
matic residues with water [52,53]. AgNP@LZ solution (Fig. S2 of SM). Taken together, these results
On the other hand, in both AgNP@protein solutions, the typical ef­ strongly suggest that the LZ molecules are tightly attached to the AgNPs.
fect of surface-enhanced Raman scattering (SERS) was observed in the The accepted model for the formation of protein corona on NPs in­
spectrum of each protein induced by contact with the silver nanocore, volves a dynamic process in which free protein molecules in solution are
increasing the Raman intensity by several orders of magnitude [53]. constantly exchanged with those adsorbed on the surface of the NPs
Additionally, a new band at 235 cm− 1 due to the Ag-N(Protein) vibra­ [19]. Therefore, proteins with a high attraction for the nanoparticle
tory mode was observed (insets of Fig. 5a and b), together with the surface will form a hard corona with tightly bound proteins in the
relative decrease of the band at 815 cm− 1 due to the breakdown of the innermost layer, and with a slow protein exchange lifetime of several
hydrogen-bonding network between aromatic residues and water mol­ hours. On the other hand, proteins with lower affinity will form a softer
ecules, suggesting that these residues interact with the metal surface corona, with faster protein exchange rates occurring in a few seconds to
[53,54]. minutes [12,14,19,51].
The SERS effect in the 1700− 1100 cm− 1 region was almost 30-times Therefore, different protein corona dynamics can be expected for
more intense in AgNP@LZ than for AgNP@BSA solutions, confirming a AgNP@LZ and AgNP@BSA solutions. In the first case, the release of Ag+
much stronger interaction of LZ than BSA with the silver nanocore and, from the inner core of AgNP is blocked by the strongly adsorbed LZ
corona, with very slow molecular exchange, preventing the toxic effect
during the incubation of the strains. In the case of AgNP@BSA, the
formation of a softer protein corona with faster molecular exchange rate
is anticipated, allowing the oxidation of silver on the AgNP surface and
subsequent release of Ag+ in the medium, which eventually induce
bacterial death.
The Raman spectrum of Fig. 5c indicates that also native LZ strongly
interacts in the presence of a molar excess of dissoved Ag+, since a
similar spectral fingerprint to that observed for AgNP@LZ was observed
[54], although the intensity enhancement is almost 10-times lower.
Nevertheless, the intense band at 240 cm− 1 attributed to the Ag+
interaction with the N-atom of the imidazole ring of residue His15 of LZ
was observed [55], with an association constant Ka = 1.9 μM− 1 deter­
mined by isothermal titration calorimetry [56]. Instead, the binding of
Ag+ to BSA is much weaker, since a Ka = 0.004 μM− 1 was determined by
BSA fluorescence quenching by AgNO3 measurements (Fig. S3 of SM).
As a consequence of the strong Ag+–LZ interaction, the lytic activity
of native LZ decreased with increasing AgNO3 concentration (Fig. S4 of
SM), in a similar way to that observed by Wu et al. [56]. This result
confirms that the binding of Ag+ ions modifies the conformational
structure at the specific site for LZ lytic activity, in particular, the Glu35
and Asp52 residues which are known to be directly responsible for
enzyme action cleavage of glycosidic bond in the alternating β-linked
(1-4) copolymer of the N-acetyl-D-muramic acid. [25,55].

4. Conclusions

In this work, we have demonstrated the suitability of the photore­


duction method of Ag+ ions using I-2959 as the photoiniator for pre­
paring thermodynamically stable AgNP solutions in the presence of
citrate or proteins (LZ and BSA) as stabilizers, as shown in Scheme 1.
The negatively charged AgNP@CIT and AgNP@BSA showed efficient
antibacterial activity, mainly due to their capability to act as an nano-
reservoir for the release of the toxic Ag+ ions. However, the prepared
AgNP@LZ was completely ineffective for bacterial inactivation, despite
the intrinsic antimicrobial function of LZ [23]. In addition, the presence
of native LZ in both AgNP@CIT and AgNP@BSA resulted in the com­
plete loss of the original antibacterial activity of these nanoparticles.
This antagonistic antibacterial effect exerted by positively charged
LZ is modulated by the strength of specific silver-lysozyme interactions,
resulting in the formation of a harder protein corona in AgNP@LZ, while
inducing electrostatic-mediated agglomeration of both AgNP@CIT and
Fig. 5. a, b) Raman spectra of the native protein (10 μM LZ and 20 μM BSA) AgNP@BSA, along with the strong association of free Ag+. These effects
and of the respective AgNP@Protein colloidal solution; and c) of 10 μM LZ in decrease the viability of Ag+ ions to kill bacteria, Scheme 1.
the absence and presence of 200 μM AgNO3, respectively. Note the difference of These results were also in opposition to previous reports in which an
intensity scale due to the SERS effect. Insets: SERS band of the Ag–N(Protein) antibacterial effect of AgNP@LZ was observed [32,57]. However, in
vibrational mode around 235 cm− 1. these cases, the bionanocomposite was chemically synthesized under

6
M.B. Espeche Turbay et al. Colloids and Surfaces B: Biointerfaces 208 (2021) 112030

Scheme 1. Pathways related to the antago­


nistic effect of lysozyme (LZ) on photochemi­
cally prepared AgNPs. I-2959 is the
photoiniciator that after absorption of a UVA
photon generates ketyl radicals to reduce Ag+
to Ag◦ , with acetone and HEBA (4-(2-hydrox­
yethoxy)benzoic acid) as by-products [20]. Red
arrows indicate steps that do not release Ag+
ions as toxic species to bacteria, due to the
formation of a hard protein corona (AgNP@LZ),
along with the high binding of Ag+ to LZ.
Furthermore, LZ also interacts electrostatically
with negatively charged AgNPs, e.g. capped
with citrate ions (AgNP@CIT) or bovine serum
albumin (AgNP@BSA), to induce AgNPs
agglomeration and binding of free Ag+,
reducing the initial ability of these nano­
particles to release Ag+ ions and cause bacterial
death (as shown by the green arrow steps).

strong protein denaturation conditions, such as using methanol as a [2] D.I. Andersson, D. Hughes, Antibiotic resistance and its cost: is it possible to reverse
resistance? Nat. Rev. Microbiol. 8 (2010) 260–271.
solvent [32], or highly alkaline media [57]. Therefore, it can be spec­
[3] R. Vivas, A.A.T. Barbosa, S.S. Dolabela, S. Jain, Multidrug-resistant bacteria and
ulated that the synthesis conditions and the presence of other chemical alternative methods to control them: an overview, Microb. Drug Resist. 25 (2019)
agents in the preparation of AgNPs can modify the strength of the 890–908.
silver-protein interactions and, later, the antibacterial efficacy of the [4] S. Zaidi, L. Misba, A.U. Khan, Nano-therapeutics: a revolution in infection control
in post antibiotic era, Nanomed. Nanotechnol. Biol. Med. 13 (2017) 2281–2301.
biocomposite. [5] L. Wang, C. Hu, L. Shao, The antimicrobial activity of nanoparticles : present
In conclusion, the present results point to the relevant role of the situation and prospects for the future, Int. J. Nanomed. 12 (2017) 1227–1249.
specific interactions of AgNPs with proteins for the fate of their bio­ [6] B. Calderón-Jiménez, et al., Silver nanoparticles: technological advances, societal
impacts, and metrological challenges, Front. Chem. 5 (2017) 1–26.
logical function, which may result in unexpected behaviors, such as the [7] B. Le Ouay, F. Stellacci, Antibacterial activity of silver nanoparticles: a surface
antagonistic antibacterial effect that is reported here. science insight, Nano Today 10 (2015) 339–354.
[8] N. Durán, et al., Silver nanoparticles: a new view on mechanistic aspects on
antimicrobial activity, Nanomed. Nanotechnol. Biol. Med. 12 (2016) 789–799.
CRediT authorship contribution statement [9] A.V. Domínguez, R.A. Algaba, A.M. Canturri, Á.R. Villodres, Y. Smani,
Antibacterial activity of colloidal silver against gram-negative and gram-positive
M. Beatriz Espeche Turbay: Validation, Investigation, Writing - bacteria, Antibiotics 9 (2020) 1–10.
[10] S. Tang, J. Zheng, Antibacterial activity of silver nanoparticles: structural effects,
Review & Editing. Valentina Rey: Validation, Investigation. Rita D. Adv. Healthc. Mater. 7 (2018), 1701503.
Dorado: Validation, Investigation. Marcelo C. Sosa: Validation, [11] S. Chernousova, M. Epple, Silver as antibacterial agent: ion, nanoparticle, and
Investigation. Claudio D. Borsarelli: Conceptualization, Resources, metal, Angew. Chem. - Int. Ed. 52 (2013) 1636–1653.
[12] N. Durán, C.P. Silveira, M. Durán, D.S.T. Martinez, Silver nanoparticle protein
Writing - Review & Editing, Visualization, Supervision, Project admin­ corona and toxicity: a mini-review, J. Nanobiotechnol. 13 (2015) 1–17.
istration, Funding acquisition. [13] S. Kittler, et al., The influence of proteins on the dispersability and cell-biological
activity of silver nanoparticles, J. Mater. Chem. 20 (2010) 512–518.
[14] C. Gunawan, M. Lim, C.P. Marquis, R. Amal, Nanoparticle-protein corona
Declaration of Competing Interest complexes govern the biological fates and functions of nanoparticles, J. Mater.
Chem. B 2 (2014) 2060–2083.
[15] P.S. Nayak, et al., Lactoferrin adsorption onto silver nanoparticle interface:
The authors report no declarations of interest. Implications of corona on protein conformation, nanoparticle cytotoxicity and the
formulation adjuvanticity, Chem. Eng. J. 361 (2019) 470–484.
Acknowledgments [16] E.I. Alarcon, et al., Human serum albumin as protecting agent of silver
nanoparticles: role of the protein conformation and amine groups in the
nanoparticle stabilization, J. Nanopart. Res. 15 (2013) 1374.
The authors thanks for the financial support to the following [17] J.H. Shannahan, et al., Silver nanoparticle protein corona composition in cell
Argentinean Institutions: Universidad Nacional de Santiago del Estero culture media, PLoS One 8 (2013).
[18] U. Hansen, A.F. Thünemann, Characterization of silver nanoparticles in cell culture
(UNSE, grant #23A/254), Consejo Nacional de Investigaciones Científ­ medium containing fetal bovine serum, Langmuir 31 (2015) 6842–6852.
icas y Tecnológicas (CONICET, grant #PUE-2018-035) and Fondo para [19] P. Del Pino, et al., Protein corona formation around nanoparticles - from the past to
la Investigación Científica y Tecnológica (FONCyT grant #PICT-2019- the future, Mater. Horiz. 1 (2014) 301–313.
[20] M.L. Marin, K.L. McGilvray, J.C. Scaiano, Photochemical strategies for the
02052).
synthesis of gold nanoparticles from Au(III) and Au(I) using photoinduced free
radical generation, J. Am. Chem. Soc. 130 (2008) 16572–16584.
Appendix A. Supplementary data [21] V. Rey, et al., Kinetics and growth mechanism of the photoinduced synthesis of
silver nanoparticles stabilized with lysozyme, Colloids Surf. B Biointerfaces 172
(2018) 10–16.
Supplementary material related to this article can be found, in the [22] M. Balouiri, M. Sadiki, S.K. Ibnsouda, Methods for in vitro evaluating antimicrobial
online version, at doi:https://doi.org/10.1016/j.colsurfb.2021.112030. activity: a review, J. Pharm. Anal. 6 (2016) 71–79.
[23] B. Masschalck, C.W. Michiels, Antimicrobial properties of lysozyme in relation to
foodborne vegetative bacteria, Crit. Rev. Microbiol. 29 (2003) 191–214.
References [24] C. Rodríguez, et al., Successful management with fosfomycin + ceftazidime of an
infection caused by multiple highly-related subtypes of multidrug-resistant and
[1] W.C. Reygaert, An overview of the antimicrobial resistance mechanisms of extensively drug-resistant KPC-producing Serratia marcescens, Int. J. Antimicrob.
bacteria, AIMS Microbiol. 4 (2018) 482–501. Agents 52 (2018) 737–739.

7
M.B. Espeche Turbay et al. Colloids and Surfaces B: Biointerfaces 208 (2021) 112030

[25] A.L.N. Prasad, G. Litwack, Measurement of the lytic activity of lysozymes knowledge and recommendations for future studies and applications, Materials
(muramidases), Anal. Biochem. 6 (1963) 328–334. (Basel) 6 (2013) 2295–2350.
[26] G. Wang, Y. Lu, H. Hou, Y. Liu, Probing the binding behavior and kinetics of silver [42] X. Li, J.J. Lenhart, H.W. Walker, Dissolution-accompanied aggregation kinetics of
nanoparticles with bovine serum albumin, RSC Adv. 7 (2017) 9393–9401. silver nanoparticles, Langmuir 26 (2010) 16690–16698.
[27] D. Paramelle, et al., A rapid method to estimate the concentration of citrate capped [43] A. Henglein, Colloidal silver nanoparticles: photochemical preparation and
silver nanoparticles from UV-visible light spectra, Analyst 139 (2014) 4855–4861. interaction with O2, CCl4, and some metal ions, Chem. Mater. 10 (1998) 444–450.
[28] E.I. Alarcon, et al., The biocompatibility and antibacterial properties of collagen- [44] X. Wang, G. Herting, I.O. Wallinder, E. Blomberg, Adsorption of lysozyme on silver
stabilized, photochemically prepared silver nanoparticles, Biomaterials 33 (2012) and its influence on silver release, Langmuir 30 (2014) 13877–13889.
4947–4956. [45] Y. Li, G. Yang, Z. Mei, Spectroscopic and dynamic light scattering studies of the
[29] M. Voicescu, S. Ionescu, D.G. Angelescu, Spectroscopic and coarse-grained interaction between pterodontic acid and bovine serum albumin, Acta Pharm. Sin.
simulation studies of the BSA and HSA protein adsorption on silver nanoparticles, B 2 (2012) 53–59.
J. Nanopart. Res. 14 (2012). [46] S. Ashrafpour, T.T. Moghadam, Interaction of silver nanoparticles with Lysozyme:
[30] V. Banerjee, K.P. Das, Interaction of silver nanoparticles with proteins: a functional and structural investigations, Surf. Interfaces 10 (2018) 216–221.
characteristic protein concentration dependent profile of SPR signal, Colloids Surf. [47] U.H. Abo-Shama, et al., Synergistic and antagonistic effects of metal nanoparticles
B Biointerfaces 111 (2013) 71–79. in combination with antibiotics against some reference strains of pathogenic
[31] A. Gebregeorgis, C. Bhan, O. Wilson, D. Raghavan, Characterization of silver/ microorganisms, Infect. Drug Resist. 13 (2020) 351–362.
bovine serum albumin (Ag/BSA) nanoparticles structure: morphological, [48] P.S. Ocampo, et al., Antagonism between bacteriostatic and bactericidal antibiotics
compositional, and interaction studies, J. Colloid Interface Sci. 389 (2013) 31–41. is prevalent, Antimicrob. Agents Chemother. 58 (2014) 4573–4582.
[32] D.M. Eby, N.M. Schaeublin, Ќ.K.E. Farrington, S.M. Hussain, G.R. Johnson, [49] K.V. Abrosimova, O.V. Shulenina, S.V. Paston, FTIR study of secondary structure of
Lysozyme catalyzes the formation of antimicrobial silver nanoparticles, ACS Nano bovine serum albumin and ovalbumin, J. Phys. Conf. Ser. 769 (2016).
3 (2009) 984–994. [50] P. Sassi, A. Giugliarelli, M. Paolantoni, A. Morresi, G. Onori, Unfolding and
[33] A.V. Yakovlev, O.Y. Golubeva, Synthesis optimisation of lysozyme monolayer- aggregation of lysozyme: a thermodynamic and kinetic study by FTIR
coated silver nanoparticles in aqueous solution, J. Nanomater. 2014 (2014). spectroscopy, Biophys. Chem. 158 (2011) 46–53.
[34] M. Radmacher, M. Fritz, H.G. Hansma, P.K. Hansma, Direct observation of enzyme [51] J.S. Gebauer, et al., Impact of the nanoparticle-protein corona on colloidal stability
activity with the atomic force microscope, Science (80-.) 265 (1994) 1577–1579. and protein structure, Langmuir 28 (2012) 9673–9679.
[35] M.L. Ferrer, R. Duchowicz, B. Carrasco, J.G. De La Torre, A.U. Acuña, The [52] S. Stewart, P.M. Fredericks, Surface-enhanced Raman spectroscopy of amino acids
conformation of serum albumin in solution: a combined phosphorescence adsorbed on an electrochemically prepared silver surface, Spectrochim. Acta Part A
depolarization-hydrodynamic modeling study, Biophys. J. 80 (2001) 2422–2430. 55 (1999) 1641–1660.
[36] K. Rezwan, A.R. Studart, J. Vörös, L.J. Gauckler, Change of ζ potential of [53] J. Hu, R.S. Sheng, Z.S. Xu, Y. Zeng, Surface enhanced Raman spectroscopy of
biocompatible colloidal oxide particles upon adsorption of bovine serum albumin lysozyme, Spectrochim. Acta Part A Mol. Spectrosc. 51 (1995) 1087–1096.
and lysozyme, J. Phys. Chem. B 109 (2005) 14469–14474. [54] G. Chandra, K.S. Ghosh, S. Dasgupta, A. Roy, Evidence of conformational changes
[37] S.P. Shields, V.N. Richards, W.E. Buhro, Nucleation control of size and dispersity in in adsorbed lysozyme molecule on silver colloids, Int. J. Biol. Macromol. 47 (2010)
aggregative nanoparticle growth. A study of the coarsening kinetics of thiolate- 361–365.
capped gold nanocrystals, Chem. Mater. 22 (2010) 3212–3225. [55] M.J. Panzner, S.M. Bilinovich, W.J. Youngs, T.C. Leeper, Silver metallation of hen
[38] P. Kaur, D.V. Vadehra, Mechanism of resistance to silver ions in Klebsiella egg white lysozyme: X-ray crystal structure and NMR studies, Chem. Commun. 47
pneumoniae, Antimicrob. Agents Chemother. 29 (1986) 165–167. (2011) 12479.
[39] Z.M. Xiu, Q.B. Zhang, H.L. Puppala, V.L. Colvin, P.J.J. Alvarez, Negligible particle- [56] Q. Wu, H. Zhang, T. Sun, B. Zhang, R. Liu, Probing the toxic mechanism of Ag+
specific antibacterial activity of silver nanoparticles, Nano Lett. 12 (2012) with lysozyme, Spectrochim. Acta Part A Mol. Biomol. Spectrosc. 151 (2015)
4271–4275. 124–130.
[40] S. Halder, et al., Alteration of Zeta potential and membrane permeability in [57] S. Ashraf, M.A. Chatha, W. Ejaz, H.A. Janjua, I. Hussain, Lysozyme-coated silver
bacteria: a study with cationic agents, Springerplus 4 (2015) 1–14. nanoparticles for differentiating bacterial strains on the basis of antibacterial
[41] B. Reidy, A. Haase, A. Luch, K.A. Dawson, I. Lynch, Mechanisms of silver activity, Nanosc. Res. Lett. 9 (2014) 1–10.
nanoparticle release, transformation and toxicity: a critical review of current

You might also like