Contents lists available atScienceDirect
NanoImpact
journal homepage:www.elsevier.com/locate/nanoimpact
Research paper
Nanosilver: An innovative paradigm to promote its safe and active use
D. Gardini
a, M. Blosi
a,⁎, S. Ortelli
a, C. Delpivo
a,1, O. Bussolati
a,b, M.G. Bianchi
b, M. Allegri
b,
E. Bergamaschi
a,c, A.L. Costa
aaNational Research Council of Italy (CNR) - Institute of Science and Technology for Ceramics (ISTEC), Via Granarolo 64, I-48018 Faenza, Italy bLaboratory of General Pathology, Department of Medicine and Surgery, University of Parma, Via Volturno 39, I-43125 Parma, Italy cDepartment of Public Health and Pediatrics Public Health, Toxicology, University of Torino, P.zza Polonia, 94, 10126 Torino, Italy
A R T I C L E I N F O Keywords: Nanosilver Toxicity Antimicrobial activity Safe application Risk Quality Index A B S T R A C T
A general model useful to assess the applicability, in terms of safety and efficiency, of nanosilver sols in real applications was developed. This led to the definition of an operative parameter, named Quality Index (QI), able to rank the capacity of the samples to be applied effectively in safe concentration ranges on the basis of the mere knowledge of the toxicological and antibacterial limits of the samples and of the fraction of silver ions (Ag+)
with respect the total silver in the samples and in the samples after application. The new approach was applied to three nanosilver formulations to be used as antibacterial agents on tile surfaces, differing, essentially, for the relative amount of silver ions and silver nanoparticles (Ag NPs): i) a (diluted) commercial sample (Ag1); ii) a sample arising from the latter after an ultrafiltration treatment applied as remediation strategy (Ag31); and iii) a sample synthesized at lab scale in controlled conditions (AgL). Such sols were characterized with in vitro toxicity and antibacterial tests to collect the inputs required by the model. A general rationale to classify partially soluble nanomaterials, exploitable in antibacterial applications, was thus introduced, opening a new perspective toward the promotion of a safer by-design management of nanomaterials.
1. Introduction
Silver-based materials have been broadly used from long times in different applications as photography and wound treatments, as well as pigments, conductive/antistatic composites, catalysts, and biocides (Lara, 2011; Nowack et al., 2011). Colloidal silver sols (suspensions of silver nanoparticles, Ag NPs) have been commercialized for health and medical reasons since the early twentieth century and at present they are marketed as a dietary supplement and alternative medicine cure-all (Seltenrich, 2013). Nevertheless, improved capabilities in nanoscience and nanoparticle synthesis and engineering justify the novel/revived attention on Ag NPs for antimicrobial applications (Woodrow Wilson International Center for Scholars, 2013) with an estimated amount of about 320 tons/year produced and used worldwide.
The mechanisms by which Ag NPs exert toxicity and, consequently, antimicrobial effects, are not fully understood (Behra et al., 2013;Huk et al., 2014;McShan et al., 2014;Reidy et al., 2013) but it is commonly accepted that the release of silver ions (Ag+) represents the primary mechanism of antibacterial action, with a negligible particle-specific activity (Ivask et al., 2014a;Ma et al., 2012;Wijnhoven et al., 2009;Xiu et al., 2012). As extensively reviewed by Liu and coworkers (Liu et al.,
2010), the majority of silver ions comes from oxidation of the zer-ovalent metallic particle, typically by reaction with dissolved oxygen (O2), mediated by protons and other components in the surrounding fluid phase. Nevertheless, some unreacted silver ions can remain after sol-gel synthesis, increasing the concentration of dissolved ions, and, as a consequence, the antibacterial activity of the sample.
At the base of the ion-based toxicity pathway there is a strong in-teraction with respiratory and transport proteins, due to the high a ffi-nity of silver ions with thiol groups present in the cysteine residues of those proteins (Marambio-Jones and Hoek, 2010). The hypothesized role of the Ag NPs is both to generate a sustainedflux of Ag+ions from their surfaces and to transport and let available active Ag+ ions to sensitive biological targets on cell membranes or within cells, following particle attachment or endocytosis, respectively. This could justify many discrepancies found in literature between “silver ions” and “particle-specific” toxicity/antibacterial mechanisms (Lubick, 2008). Studies supporting a particle-specific antibacterial activity (Gunawan et al., 2009; Ivask et al., 2012; Jin et al., 2010; Lok et al., 2006; McQuillan et al., 2012;Navarro et al., 2008;Pal et al., 2007) are, in fact, based on the lower toxicant reactivity of Ag+ ion solutions in comparison to colloidal silver samples containing an equivalent amount
https://doi.org/10.1016/j.impact.2018.06.003
Received 17 January 2018; Received in revised form 21 June 2018; Accepted 22 June 2018
⁎Corresponding author at: Institute of Science and Technology for Ceramics (ISTEC), Via Granarolo 64, I-48018 Faenza, Italy.
1Present address: Leitat Technological Center, Carrer de la Innovació, 2, 08225 Terrassa (Spain).
E-mail address:magda.blosi@istec.cnr.it(M. Blosi).
Available online 24 June 2018
2452-0748/ © 2018 Published by Elsevier B.V.
of dissolved Ag+ions. However the effect could be explained by con-sidering not different toxicity pathways, but an increased bioavail-ability of Ag+ions to bacterial cells, depending on their release and transport in aqueous phases from the Ag NPs surfaces (Ivask et al., 2014a, 2014b).
In order to definitively overcome the confusion arising when solu-bility of Ag NPs is discussed and, at the same time, to take into con-sideration that the biological reactivity is related to the coexistence of silver nanoparticles and silver ions, it is necessary to understand that Ag NPs in their metallic form (noble metal) are totally insoluble and che-mically inert, but when they come in touch with water, under aerobic conditions, their surface is oxidized with the creation of a silver oxide (Ag2O)/silver hydroxide (AgOH) layer and the consequent release of silver ions. The solubility of the Ag2O/AgOH layer depends on ther-modynamic parameters (pH, temperature), but also on kinetic factors due to the non-stoichiometry solubility, which is surface-dependent property, being affected by surface area, presence of defects, particle size and shape, presence of capping agents (Misra et al., 2012). If only the release mechanism was considered, no differences in the anti-bacterial activity should be expected once samples were normalized for the content of Ag+ions in solution. As this does not occur, it is ne-cessary to take into account also the contribution of the NPs. In order to explain the combined toxic effects of silver ions and particles, it is ne-cessary to change paradigm and to think in terms of different bioa-vailability of Ag+ions with respect to the cells (Navarro et al., 2015). In fact, they will act differently depending if they are free in the electro-lyte solution outside the cells, if they are adsorbed on Ag NP surfaces, or if they are released from the Ag NP surfaces into the intracellular en-vironment once penetrated through a Trojan-horse pathway ( Bermejo-Nogales et al., 2016;Gliga et al., 2014; Ivask et al., 2014b) as sche-matized inFig. 1.
Considering the extremely complex equilibrium speciation and chemical pathways that lead Ag+ions to react with biological targets, the control of the content of Ag+ions and of their distribution/avail-ability within the biological matrix is, for sure, the main challenging objective to control Ag NP hazardous properties. In this context, some design strategies, acting on surface chemistry of the nanoparticles or on their nano-structuring, allow to control the reactivity of Ag+ions (i.e., their adverse effects against biological targets: both cytotoxicity and antibacteride) and, at the same time, to preserve the toxicant action at the base of antibacterial activity.
In this work three nanosilver sols, produced starting from the same sol-gel synthesis route and differing only for the content of Ag+ions
with respect to the total silver present in the sols in whatever form (NP or ion) have been characterized in terms of decrease in cell viability (IC50) and antibacterial performance (MIC99) after an industrially re-levant application (spray deposition as antibacterial coating on ceramic tile surfaces). The half maximal inhibitory concentration (IC50) and the lowest concentration of Ag NPs that inhibits the visible growth of a microorganism after overnight incubation – the minimum inhibitory concentration (MIC)– have been measured and expressed in terms of Ag+dose, in order to allow a direct comparison with AgNO
3solutions. It was thus possible to verify if silver nanosol reactivity was only Ag+ dependent or if there was a specific Ag NP activity. Such data have been used to illustrate a general rationale to classify silver colloidal samples as potential antibacterial agents, in relation with the conditions for their safe use. In particular from the Ag+ion doses and dilutions ne-cessary to keep the Ag+ion concentration lower than IC50but higher than MIC, it was possible define and calculate the Quality Index (QI), that gives indications about the capacity of the sample to be applied as antibacterial agent in a safe concentration range developing a general approach for the safe design and management of partially soluble na-nomaterials. The suggested new approach is illustrated, and the Quality Index evaluated, on the basis of in vitro toxicity data, but in principle it could be also applied by using in vivo toxicity or ecotoxicological data and no (or low) observed adverse effect concentration (NOAEC or LOAEC) or predicted no effect concentration (PNEC) as reference values (Borm et al., 2006). The information generated in these cases would be more realistic and also closer to the expectations of the European reg-ulation for biocides that currently imposes important limitations to the approaches based on in vitro data.
2. Materials and methods 2.1. Materials
A commercial silver nanosol, provided by Colorobbia Italia SpA (4 wt% silver concentration), investigated also by NanoValid con-sortium as reported by Jemec and co-workers (Jemec et al., 2016) and patented in 2010, was diluted with Milli-Q water down to 1 wt%, in order to avoid possible gelations during sample treatment, resulting in a sol called Ag1.
Starting from such a diluted sol, an ultrafiltrated sample (Ag31) was obtained by applying a series of ultrafiltrations by using the con-servative Stirred Cell (SC) apparatus, (Merck Millipore) using a 100 kDa polyethersulfone membrane disc and recovering at each step both Ag
NPs-rich supernatant (ultrafiltered sol) and Ag+
ions solution permeate (filtered solution). All the volume of the samples was filtered in the stirred cell. Thefinal total concentration of the purified Ag31 sample is 0.44 wt%.
Synthesized silver nanosol (AgL) at 0.5 wt% was obtained by sol-gel synthesis using at lab scale level the same reducing (glucose) and complexing (polyvinylpyrrolidone) agents, with the same stoichiometry ratio, as used at industrial level for the synthesis of the commercial patented sample (http://google.com/patents/WO2010100107A2, n.d.).
AgNO3solutions were used as benchmark material for toxicity ex-periments. Dulbecco's modified Eagle's medium (DMEM) and Fetal Bovine Serum (FBS) were from Euroclone SpA (Pero, Milan, Italy). Bovine Serum Albumin (BSA) was from Sigma-Aldrich, Milan, Italy. 2.2. Size and zeta potential analysis
Hydrodynamic diameter and zeta potential of the Ag NPs in the three nanosilver sols after dilution in Milli-Q water and DMEM medium at 128μg/mL of total silver concentration, were evaluated with a Zetasizer Nano ZSP (model ZEN5600, Malvern Instruments, UK). Hydrodynamic diameters were obtained from Dynamic Light Scattering (DLS) data, setting the measurements in back-scattering mode with a detection angle of 173°. After a 2 min temperature equilibration step, 1 mL of sample volume was subjected to three consecutive measure-ments performed at 25 °C, and particle size distributions by intensity were obtained by averaging these measurements.
Zeta potential measurements were performed at 25 °C by Electrophoretic Light Scattering (ELS) technique. Smoluchowski equa-tion was applied to convert the electrophoretic mobility to zeta po-tential. After a 2 min temperature equilibration step, samples under-went three measurements delayed of 120 s, and zeta potential was obtained by averaging these measurements. Before and after zeta po-tential analyses, a size measurement was performed to check that the hydrodynamic diameter had not changed, i.e. that neither aggregation/ dissolution phenomena had occurred.
Morphological analysis was carried out using a High Resolution (HR) Transmission Electron Microscopy (TEM) in JEOL JEM-2100F multipurpose with afield emission source operating between 80 and 200 kV. Agglomeration factor was calculated by dividing DLS mean diameter for TEM derived diameter.
2.3. Determination of the silver ionic ratios (Ag+/Ag)
The as prepared samples closed and wrapped in aluminum foil al-lowed to equilibrate with ambient air for at least 3 days after pre-paration, in order to achieve the stationary equilibrium. The amount of Ag+ ions with respect the total amount of silver, Ω
Ag+(0), was
de-termined measuring the concentration of total silver in the sols by X-ray Fluorescence (XRF) (Panalytical Axios Advanced, NL) and the con-centration of silver ions by Atomic Absorption Spectroscopy (AAS) (Perkin Elmer, USA) in the permeates, once separating the ions from solid particles, through an ultrafiltration membrane, as reported also by Jemec et al. 2016(Jemec et al., 2016). We used Ultra Centrifugal Filter (UCF) units (Amicon Ultra-15, 10 kDa, Millipore) for the samples Ag31 and AgL and Stirred Cell apparatus for the sample Ag1, assuming that the ultrafiltration allows to separate completely the ionic fraction from the solid one (see Electronic Supplementary Information– Section B). Optimal ultrafiltration conditions for UCF were obtained with starting volume sample of 13 mL, a relative centrifugal field (RCF) of 3520g (corresponding to 5000 rpm for the centrifuge used) and spin time of 30 min.
In order to remove the excess of ions and establishing the amount of ions leached at equilibrium the samples were subjected to repeated ultrafiltration steps in the UCF units (see Fig. S2 – Electronic Supplementary Materials); the filtered liquid was replaced with the
same volume of fresh Milli-Q water and the resulting sol left to equi-librate for 24 h. This treatment was repeated several times till the concentration of silver ions in the permeates reached an asymptotic value. Assuming that the washing step occurring during the application of the sols on the tile surfaces can be approximated by thefirst ultra-filtration step, the silver ion fraction in the sols after application, ΩAg+(3)has been approximated with the concentration of silver
mea-sured in permeates.
2.4. Dispersion of nanosilver sols within biological media
Nanosilver sols werefirst diluted in a 0.05 wt% solution of Bovine Serum Albumin (BSA) in Phosphate Buffered Saline (PBS) without calcium and magnesium, in order to obtain the more concentrated 100× stock suspension with Ag dose of 12.8 mg/mL (corresponding to 8 mg/cm2for a well of 96-well trays considering afilling volume of 200μL). This stock was then diluted in 0.05 wt% BSA in PBS to obtain the other 100× stock suspensions. All the stocks were then vortexed for 30 s, sonicated for 15 min, re-vortexed for 30 s andfinally dispersed in complete growth medium (Dulbecco's modified Eagle's medium (DMEM) + 10 wt% of Fetal Bovine Serum (FBS)).
2.5. Cells and experimental treatments
Mouse peritoneal monocyte-macrophage cells (Raw 264.7 line) were obtained from the Istituto Zooprofilattico della Lombardia e dell'Emilia (Brescia, Italy). Raw 264.7 cells were cultured in DMEM supplemented with 10 wt% FBS, 4 mM glutamine, streptomycin (100μg/mL) and penicillin (100 U/mL). Cells were routinely cultured in 10-cm diameter dishes (Falcon™ 353,003, Corning, New York, NY (USA)), maintained in a humidified atmosphere of 5% CO2in air. For cytotoxicity tests, cells were seeded in complete growth medium (DMEM + 10% FBS) in 96-well multiwell plates (Falcon™ 353072), at a density of 30 × 103 cells/well. 24 h after cell seeding, the complete growth medium was replaced with fresh medium supplemented with Ag NPs prepared at different doses from the 100× stocks at 128 mg/L. Doses of NPs were adjusted in order to obtain silver concentrations per unit of surface ranging from 0.3125μg/cm2 to 80μg/cm2 (corre-sponding to a range from 0.5 to 128μg/mL). In all the experiments, vehicle (0.05 wt% BSA in PBS) was added to the control.
2.6. Cytotoxicity tests to determine the in vitro toxicity limit (IC50) Cell viability was assessed with the indicator resazurin, a non-fluorescent molecule which is converted by intracellular enzymes in the fluorescent compound resorufin (λem= 572 nm). After 24 h of in-cubation in presence of Ag NPs, cell viability was tested replacing medium with a solution of resazurin (44μM) in serum-free medium. After 20 min,fluorescence was measured at 572 nm with a multimode plate reader Perkin Elmer Enspire (Waltham, Massachussets, USA). Since nanomaterials could interfere with cytotoxicity tests, a pre-liminary experiment was performed Ag NP preparations (80μg/cm2of Ag1 and Ag31) with resazurin (44 mM) in the absence of cells or with cells incubated for 20 min in a medium containing resazurin (44 mM) and then measuringfluorescence. In either case, there was no difference in thefluorescence signal compared with samples incubated without NPs. Dose-response curves have been calculated with a GraphPad™ software, version 5.0 and the half maximal inhibitory concentration IC50 parameter evaluated.
2.7. Antibacterial tests to determine the antibacterial limit (MIC99) Antibacterial activity against Escherichia coli was selected as mate-rial functional property to be assessed. Escherichia coli strain ATCC8739 was chosen as model microorganism for inactivation experiments. For the antibacterial activity assay, each sample (Ag1, Ag31, AgL and
AgNO3as comparison) was subsequently diluted in KH2PO4buffered solution (achieved diluting 5 mL of buffer at pH 7 in 800 mL of distilled water), producing 10 test tubes with concentrations ranging from 50μg/mL to 0.005μg/mL. 1 mL of inoculum containing 1.5–3.0 × 105
CFU/mL of bacterial was added to 9 mL of the previously prepared Ag diluted sols. The resulting systems were incubated at room temperature and kept under stirring for 1 h. Then 1 mL from each in-cubated tube was withdrawn and added to test tubes containing 9 mL of a phosphate buffer solution. Finally 1 mL from each prepared test tube was inoculated in a plate culture agar, incubated at 37 °C and after 24 h the viable cell count was performed from which the antibacterial limit MIC99was obtained. Three independent experiments were repeated for each sample.
3. Theory and calculation
3.1. Quality Index
In order tofind conditions for a fast evaluation of the biological reactivity of nanosilver sols and to obtain useful information for their safe and effective application as antibacterial coating agents, both toxicity and antibacterial properties should be considered and related to the sample physicochemical properties. Let us consider an aqueous nanosilver sol to be applied as antibacterial agent on a ceramic tile surface. In order to make the process safe and effective, such a sol has to be non-toxic and antibacterial, respectively. The first condition (non-toxicity) is satisfied if the following inequality is verified:
< + + ω ω Ag Ag Tox 0 ( ) (1) whereωAg+(0)is the mass fraction of silver ions in the sol andωAg+Tox
is, in the present model, the in vitro toxicity limit (e.g., the IC50). If such condition is not satisfied, it is necessary to treat the starting sol in some way in order to reduce the silver ion concentration below the safety limit. The easiest way to do this is to dilute the sol with deionized water till the mass fraction of silver ions in the resulting sol,ωAg+(1), satisfies
the inequality: < + + ω ω Ag Ag Tox 1 ( ) (2) Fig. 2shows the dilution step, where m0andωAg(0)are the total mass and the mass fraction of silver in the starting sol, m1andωAg(1)are the total mass and the mass fraction of silver in the resulting sol after di-lution and mWDis the mass of water of dilution to add.
Let us suppose that a known mass m2(< m1) of the diluted sol at concentration ωAg(1) is then applied by spray-gun technique on a ceramic tile surface. After spraying, the process foresees a washing treatment of the surface with a known amount of water (mw). This further treatment will remove part of the silver from the surface (cur-rent 4) leaving on the surface a mass of sprayed sol m3with a mass fraction of silverωAg(3)(Fig.3).
The second condition (antibacterial activity) is satisfied if the mass fraction of silver ions in the sol after the applicationωAg+(3)is higher
than the antibacterial limitωAg+AntiB(e.g., the MIC99):
> + + ω ω Ag Ag AntiB 3 ( ) (3)
So, the upper and lower limits of the amount of diluting water to add (mwD) to a given starting sol at concentrationωAg(0), to satisfy both the condition were established (Eq.(4)):
⎧ ⎨ ⎩ ⎡ ⎣ ⎢ ⎢ − − ⎛ ⎝ ⎜ − ⎞ ⎠ ⎟ + ⎤ ⎦ ⎥ ⎥ − ⎫ ⎬ ⎭ > > ⎛ ⎝ ⎜ ⎜ − ⎞ ⎠ ⎟ ⎟ + + + + + + m ω ω m m ω ω 1 Ω 1 1 Ω 1 1 1 1 1 1 / / Ag Ag Ag Ag AntiB wD Ag Ag Tox 0 0 0 3 0 0 ( ) ( ) ( ) ( ) (4) whereΩAg+(0)andΩAg+(3)are the mass fraction of silver ions with
re-spect to the total amount of silver in the starting sol and in the sol remained on the surface after application (spraying) and washing, re-spectively (see Electronic Supplementary Information– Section D). In order tofind a general descriptor able to easily establish if a starting sol could be made non-toxic remaining antibacterial and, more in general, to assess the applicability of soluble nanomaterials whose biological reactivity is driven by the availability of ionic form, Eq.(4)was re-arranged as in Eq.(5): − − ⎛ ⎝ ⎜ − ⎞ ⎠ ⎟+ > + > + + ω + + + m m ω ω 1 Ω 1 1 Ω 1 1 1 1 1 1 / / / Ag Ag Ag AntiB wD Ag Ag Tox 0 3 0 0 ( ) ( ) ( ) (5) and a dimensionless parameter, named Quality Index (QI), defined as the ratio between the upper and lower limits established by Eq.(5), was introduced: = ∙⎡ ⎣ ⎢ ⎢ − − ⎛ ⎝ ⎜ − ⎞ ⎠ ⎟+ ⎤ ⎦ ⎥ ⎥ + + + + QI ω ω 1 Ω 1 1 Ω 1 1 1 1 / / Ag Tox Ag Ag Ag AntiB 0 3 ( ) ( ) (6) Thus, it was possible tofind a general criterion to establish if a starting sol could be made non-toxic remaining antibacterial. If QI is higher than 1 (QI≥ 1), a safe and effective sol can be obtained, while this is not possible when (QI <1). When QI = 1, the range collapses in a point. The larger is the amplitude of the range and easier is to satisfy both conditions. The quality of the starting sol is considered better when this parameter assumes high values, and, for this reason, the parameter was named Quality Index. The more interesting aspect is that the QI depends only on four parameters, all experimentally measurable and sample-dependent, exceptΩAg+(3)that is also process-dependent:
•
In vitro toxicity limit,ωAg+Tox;•
Antibacterial limit,ωAg+AntiB;•
Mass fraction of silver ions with respect the total amount of silver in the starting sol,ΩAg+(0);•
Mass fraction of silver ions with respect to the total amount of silver in the sol remained on the surface after washing,ΩAg+(3),but it does not depend on the starting concentration of silver (ωAg(0)). This means that is not important the concentration of total silver in the starting sol, but rather which is the distribution of silver between ions and nanoparticles.
4. Results and discussion
4.1. Size and zeta potential analysis
Some colloidal properties of the commercial (diluted) pristine sol (Ag1), of the modified through ultrafiltration sol (Ag31), and of the synthesized at lab scale sol (AgL), as measured in Milli-Q water and in the culture medium used for toxicological tests, are summarized in Table 1.
The samples dispersed in Milli-Q water showed quite similar values of pH, hydrodynamic diameter (dDLS), zeta potential (ζ) and agglom-eration factor. The pH of samples diluted (1:100) in cell media is in agreement with the pH of medium alone that is around 8. The hydro-dynamic diameter of samples and so their agglomeration state in pro-tein rich medium did not change significantly, because the expected Fig. 2. Dilution step of the starting nanosilver sol.
increase of diameter, due to the surface adsorption of aminoacids and proteins (bio-corona theory) (Lynch and Dawson, 2008) is masked by the large amount of PVP surrounding nanoparticles that ensure an high stability of the nanosols over time (all the considered samples did not show in water any aggregation/sedimentation phenomena in one year of storage). The particle size in bacterial test medium (KH2PO4buffered solution) was not reported because did not significantly change if compared to water suspension. The negative zeta potential showed by Ag NPs in Milli-Q water is likely due to the presence of gluconate an-ions, deriving by the oxidation of glucose (used as reducing agent) during the synthesis, which remain adsorbed onto the surface. The transfer to biological media and the slight decrease of the zeta potential value for all the samples confirms the formation of protein corona that levels off values around −8 mV in DMEM, in agreement with the values expected by serum protein at the corresponding medium pH (Casals et al., 2010).
4.2. Silver ionic ratios (Ag+/Ag) in sols,ΩAg+(0)and sols after application
ΩAg+(3)
Most of the confusion arising when data on toxicity of different nanosilver sols are compared is due to the absence of data on the content of Ag+ions in the sols or by the absence of a shared and clear analytical method to get this information and to interpret data in a sound way. By applying mass balances (see Electronic Supplementary Information– Section B), it was possible to determine the fraction of permeated Ag+ions with respect the silver concentration in the starting sol, thus establishing the amount of Ag+ions available to exert toxicant action (see Electronic Supplementary Information– Section B - Table S1).
The mass fraction of silver ions is reported inTable 2. In order to establish the content of Ag+ions present at equilibrium, we performed repeated ultrafiltration steps on sol Ag31 with the destructive UCF
method. The fraction of Ag+ ions decreases after each ultrafiltration step, until it reached an asymptotic value of 0.08% (Table S1– Mass fractions of silver ions in the starting and ultrafiltrated sols), after 3 UFC steps and a total duration of 72 h. This value is consistent with the equilibrium data reported in literature for the PVP-stabilized 40 nm Ag nanoparticles (0.14%) (Reidy et al., 2013). It should be noted that commercial sols contain an excess of ions (Ag+/Ag content higher than that at equilibrium), reflecting the capacity of these commercial sam-ples to storage higher amount of Ag+ions, likely un-reacted ions dis-tributed between solution and solid surface. Otherwise the AgL sample, being synthetized at lab-scale in more controlled conditions, showed a concentration of Ag+ions (about 0.10%) similar to the one found at the equilibrium.
The Ag+/Ag total values (Ω
Ag+(0)) reported inTable 2have been
measured at the original sample concentration. In order to investigate the dependence of dissolution on concentration, we measured in pre-vious works the Ag+/Ag total of all Ag samples at lower concentration in a range consistent with biological tests and we did not find sig-nificant deviations from the values here reported.
4.3. Effect of silver ionic ratio (Ag+/Ag) on cell viability
The effects of the different silver nanosols on cell viability were assessed in cultures of Raw 264.7 murine macrophages with the re-sazurin method (see Electronic Supplementary Information– Fig. S3). Fig. 3. Application step: spraying and washing.
Table 1
pH, size, and zeta potential (ζ) of nanosilver sols dispersed in Milli-Q water and culture medium DMEM. Agglomeration factor for samples dispersed in Milli-Q water.
Sample dTEM(nm) Milli-Q water DMEM
pH dDLS(nm) ζ (mV) Aggl. factora pH dDLS(nm) ζ (mV) Aggl. factora
Ag1 24 ± 11 4.0 130 ± 2 −14.0 ± 0.2 5.4 7.9 131 ± 1 −7.5 ± 0.3 5.5 Ag31 15 ± 11 4.5 103b± 1 −16.3 ± 0.4 6.9 8.0 114 ± 2 −8.6 ± 1.2 7.6 AgL 18 ± 7 4.5 114 ± 1 −10.4 ± 0.2 6.3 n.d.
n.d. = not determined.
a Agglomeration factor is calculated by dividing DLS mean diameter by TEM average diameter (calculated for 100 particles) (see Electronic Supplementary
Information– Section A).
b Bimodal size distribution; size of each peak and relative volume of the two populations are: Peak 1 at 140 nm (49 vol%) and Peak 2 at 20 nm (51 vol%).
Table 2
pH and Ag+/Ag mass ratio,ΩAg+(0), for the target sols.
pH ΩAg+(0)(wt%)
Ag1 2.4 55.00
Ag31 3.5 2.69
Solutions of AgNO3, equivalent in terms of total silver with the nanosols tested, were used for comparison. When expressed as total silver, the IC50values were essentially inversely correlated with the silver ionic ratio (Ag+/Ag), with samples having the lower ratios exhibiting very low cytotoxicity (Table 3). Conversely, when the IC50is expressed on the basis of the Ag+ion concentration, the IC50values for Ag1, Ag31 and AgNO3 were roughly comparable, while the IC50 of AgL was markedly lower (see Electronic Supplementary Information– Section C).
The last result suggests for AgL a key contribution of the nano-particles to cytotoxicity, which in this sample are in large excess with
respect to the ions. Likely, the nanoparticles enhance the bioavailability of Ag+toward the target by means of a Trojan horse mode of action, as schematized inFig. 1.
4.4. Antibacterial activity
The antibacterial activity of the samples Ag1, Ag31, AgL and AgNO3 (as comparison) has been evaluated by assessing the percentage re-duction of the bacterial cell number (Escherichia coli) for different Ag doses (Table 4). Data confirm the high reactivity of nanosilver as an-tibacterial agent at very low concentrations; in fact, with reference to the total silver, the minimum inhibitory concentration (MIC99), corre-sponding to the lowest concentration at which the samples show the highest bacterial reduction (99%), ranges from few mg/L for AgL to someμg/L for Ag1; in particular Ag1 showed 99% of bacterial inhibi-tion for a concentrainhibi-tion of 50μg/L, while Ag31 and AgL much higher values (440 and 4000μg/L, respectively). However, if MIC99is referred to the silver ions concentration all the samples have MIC99of the same order of magnitude, as observed from the diagrams ofFig. 4and from data listed inTable 4.
Based on these results, AgL showed the same reactivity of AgNO3, showing an antibacterial activity which seems dependent only by the Ag + concentration. Conversely, Ag31 and particularly Ag1 samples Table 3
Effect of different silver preparations on cell viabilitya.
Sample Ag+/Ag (wt%) IC50Total silver (mg/L) Confidence Interval (95%) IC50Ag+ ions (mg/L) Confidence Interval (95%) Ag1 55.00 4.66 4.11–5.28 2.56 2.26–2.90 Ag31 2.69 95.7 68.4–134 2.58 1.84–3.61 AgL 0.10 59.6 52.4–67.8 0.060 0.052–0.068 AgNO3 100.00 2.61 2.24–3.05 2.615 2.24–3.050
a For each sample, the IC
50values are obtained from data of two separate
experiments with four independent determinations each. Table 4
MIC99for different Ag nanosols and comparison with AgNO3solution.
Sample MIC99Total silver (μg/L) MIC99Ag+ions (μg/L)
Ag1 50 27.5
Ag31 440 11.9
AgL 4000 4.0
AgNO3 2.5 2.5
Fig. 4. Dose-response histograms of the bacterial depletion (Escherichia coli) as function of Ag+concentration.
Table 5
Data used to identify the dilution ranges and calculate the QI.
ωAg+Tox ωAg+AntiB ΩAg+(0) ΩAg+(3) Upper limit Lower limit QI
(mg/L) (mg/L) (%) (%) for mwD/m0 for mwD/m0
Ag1 2.56 0.0275 55.00 2.69 4.60E + 03 2.19E + 03 2.11 Ag31 2.58 0.0119 2.69 0.34 1.28E + 03 4.71E + 01 26.76 AgL 0.06 0.0040 0.10 0.10 1.25E + 03 8.23E + 01 15.00
seem to be less reactive in terms of Ag+activity, so we can hypothesize that in these cases some the Ag+ions were not bioavailable for the antibacterial action.
4.5. Quality Index
Using in vitro toxicity and antibacterial limits, calculated in the previous paragraphs and reported in Table 5, it was possible to de-termine for each sample a useful descriptor for its“safe” and “effective” use identifying the dilution ranges, schematized also inFig. 5. Thus, it can easily viewed that sample Ag31, derived by pristine commercial sample after ultrafiltration treatment, has the highest range of dilution ratio and, therefore, the highest chance to be applied in a“safe” zone, despite the low concentration of toxicant Ag+ions that was expected to decrease its antibacterial performances as well. Furthermore, AgL, ob-tained at lab scale under more controlled conditions, improved the quality of the commercial sample, despite its higher in vitro toxicity.
On the basis of the previous experimental data, the QI for the three target nanosols was calculated, resulting to be about 2, 27 and 15 for Ag1, Ag31 and AgL, respectively. Furthermore it should be stressed that this model is valid also if in vitro toxicity limit is replaced by other threshold value used to predict toxicity through different adverse out-come pathway (Arts et al., 2016). Within this definition, all the samples tested showed a Quality Index > 1, and, therefore, it was possible to identify a‘safe and active’ concentration range where samples resulted below the toxicity limit but still antibacterial (Fig. 5).
Sample Ag31, simply derived from Ag1, removing after one ultra-filtration treatment part of the excess of free ions is the most promising one, while sample AgL, synthesized in laboratory under controlled conditions, resulted better than the commercial one, suggesting stra-tegies for a control of toxicity by preserving material functionalities for antibacterial coating nanomaterials (safe-by-design approach).
5. Conclusions
Three different nanosilver sols, produced with the same sol-gel synthesis pathway and, ultimately, differing only for the fraction of silver ions with respect the total silver, were compared in terms of decrease in cell viability (IC50), bacterial cell number reduction per-centage and antibacterial performance with reference to an industrially relevant application (deposition of antibacterial nanosilver coatings on ceramic tile surfaces). Ag+/Ag ratio was calculated in water and it was found that a diluted commercial sample (Ag1) presented a high and unexpected content of Ag+ions, around 50%. Ultrafiltration treatment applied to Ag1 or a new synthesis, performed at lab scale following the same procedure as industrial synthesis (AgL), strongly reduced the amount of Ag+ions and, thus, the toxic and antibacterial activities, which were in all cases dependent on available Ag+ion fraction. For each sample dilution limits were calculated, which define a range of
concentrations where the sample can be applied under the IC50 con-centration, but above MIC99, in order to be not toxic but still anti-bacterial. A model was developed for assessing biological sample re-activity in relation to specific applications. A general rationale to classify silver colloidal samples and identify conditions for a safe use, preserving the expected antibacterial activity, was presented. This led to the definition of a Quality Index (QI), representative of the capacity of the sample to be applied and exploited in a safe concentration range. Thus, it was possible, for thefirst time, to develop a general approach for the safe design and management of partially soluble nanomaterials, exploited in antibacterial applications.
Acknowledgements
The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007–2013) under Grant Agreement n. 280716.
Syed Tofail and Abbasi Ghandi from University of Limerick are acknowledged for providing TEM images.
Claudia Vineis and Alessio Varesano from ISMAC-CNR of Biella (Italy) are acknowledged for providing antibacterial analyses. Appendix A. Supplementary data
Supplementary data to this article can be found online athttps:// doi.org/10.1016/j.impact.2018.06.003.
References
Arts, J.H.E., Irfan, M.-A., Keene, A.M., Kreiling, R., Lyon, D., Maier, M., Michel, K., Neubauer, N., Petry, T., Sauer, U.G., Warheit, D., Wiench, K., Wohlleben, W., Landsiedel, R., 2016. Case studies putting the decision-making framework for the grouping and testing of nanomaterials (DF4nanoGrouping) into practice. Regul. Toxicol. Pharmacol. 76, 234–261.http://dx.doi.org/10.1016/j.yrtph.2015.11.020. Behra, R., Sigg, L., Clift, M.J.D., Herzog, F., Minghetti, M., Johnston, B., Petri-Fink, A., Rothen-Rutishauser, B., 2013. Bioavailability of silver nanoparticles and ions: from a chemical and biochemical perspective. J. R. Soc. Interface 10, 1–5.http://dx.doi.org/ 10.1098/rsif.2013.0396.
Bermejo-Nogales, A., Fernández, M., Fernández-Cruz, M.L., Navas, J.M., 2016. Effects of a silver nanomaterial on cellular organelles and time course of oxidative stress in afish cell line (PLHC-1). Comp. Biochem. Physiol., Part C: Toxicol. Pharmacol. 190, 54–65.
http://dx.doi.org/10.1016/j.cbpc.2016.08.004.
Borm, P.J.A., Robbins, D., Haubold, S., Kuhlbusch, T., Fissan, H., Donaldson, K., Schins, R., Stone, V., Kreyling, W., Lademann, J., Krutmann, J., Warheit, D., Oberdorster, E., 2006. The potential risks of nanomaterials: a review carried out for ECETOC. Part. Fibre Toxicol. 3, 11.http://dx.doi.org/10.1186/1743-8977-3-11.
Casals, E., Pfaller, T., Duschl, A., Oostingh, G.J., Puntes, V., 2010. Time evolution of the nanoparticle protein corona. ACS Nano 4, 3623–3632.http://dx.doi.org/10.1021/ nn901372t.
Gliga, A.R., Skoglund, S., Odnevall Wallinder, I., Fadeel, B., Karlsson, H.L., 2014. Size-dependent cytotoxicity of silver nanoparticles in human lung cells: the role of cellular uptake, agglomeration and Ag release. Part. Fibre Toxicol. 11, 11.http://dx.doi.org/ 10.1186/1743-8977-11-11.
Gunawan, C., Teoh, W.Y., Marquis, C.P., Lifia, J., Amal, R., 2009. Reversible anti-microbial photoswitching in nanosilver. Small 5, 341–344.http://dx.doi.org/10. 1002/smll.200801202.
http://google.com/patents/WO2010100107A2, n.d. Process for preparing stable sus-pensions of metal nanoparticles and the stable colloidal sussus-pensions obtained thereby.
Huk, A., Izak-Nau, E., Reidy, B., Boyles, M., Duschl, A., Lynch, I., Dušinska, M., 2014. Is the toxic potential of nanosilver dependent on its size? Part. Fibre Toxicol. 11 (65).
http://dx.doi.org/10.1186/s12989-014-0065-1.
Ivask, A., George, S., Bondarenko, O., Kahru, A., 2012. Metal-containing nano-anti-microbials: differentiating the impact of solubilized metals and particles. In: Cioffi, N., Rai, M. (Eds.), Nano-antimicrobials. Springer Berlin Heidelberg, Berlin, Heidelberg, pp. 253–290.http://dx.doi.org/10.1007/978-3-642-24428-5_9. Ivask, A., ElBadawy, A., Kaweeteerawat, C., Boren, D., Fischer, H., Ji, Z., Chang, C.H., Liu,
R., Tolaymat, T., Telesca, D., Zink, J.I., Cohen, Y., Holden, P.A., Godwin, H.A., 2014a. Toxicity mechanisms in Escherichia coli vary for silver nanoparticles and differ from ionic silver. ACS Nano 8, 374–386.http://dx.doi.org/10.1021/nn4044047. Ivask, A., Kurvet, I., Kasemets, K., Blinova, I., Aruoja, V., Suppi, S., Vija, H., Käkinen, A.,
Titma, T., Heinlaan, M., Visnapuu, M., Koller, D., Kisand, V., Kahru, A., 2014b. Size-dependent toxicity of silver nanoparticles to bacteria, yeast, algae, crustaceans and mammalian cells in vitro. PLoS One 9, e102108.http://dx.doi.org/10.1371/journal. pone.0102108.
Jemec, A., Kahru, A., Potthoff, A., Drobne, D., Heinlaan, M., Böhme, S., Geppert, M., Novak, S., Schirmer, K., Rekulapally, R., Singh, S., Aruoja, V., Sihtmäe, M., Juganson,
1 10 100 1000 10000 (Diluon rao) AgL Ag31 Ag1 Below is no antibacterial Above is toxic
K., Käkinen, A., Kühnel, D., 2016. An interlaboratory comparison of nanosilver characterisation and hazard identification: Harmonising techniques for high quality data. Environ. Int. 87, 20–32.http://dx.doi.org/10.1016/j.envint.2015.10.014. Jin, X., Li, M., Wang, J., Marambio-Jones, C., Peng, F., Huang, X., Damoiseaux, R., Hoek,
E.M.V., 2010. High-throughput screening of silver nanoparticle stability and bacterial inactivation in aquatic media: influence of specific ions. Environ. Sci. Technol. 44, 7321–7328.http://dx.doi.org/10.1021/es100854g.
Lara, H.H., 2011. Silver nanoparticles are broad-spectrum bactericidal and virucidal compounds. J. Nanobiotechnol. 1–8.
Liu, J., Sonshine, D.A., Shervani, S., Hurt, R.H., 2010. Controlled release of biologically active silver from nanosilver surfaces. ACS Nano 4, 6903–6913.http://dx.doi.org/10. 1021/nn102272n.
Lok, C.-N., Ho, C.-M., Chen, R., He, Q.-Y., Yu, W.-Y., Sun, H., Tam, P.K.-H., Chiu, J.-F., Che, C.-M., 2006. Proteomic analysis of the mode of antibacterial action of silver nanoparticles. J. Proteome Res. 5, 916–924.http://dx.doi.org/10.1021/pr0504079. Lubick, N., 2008. Nanosilver toxicity: ions, nanoparticles or both? Environ. Sci. Technol.
42, 8617.http://dx.doi.org/10.1021/es8026314.
Lynch, I., Dawson, K.A., 2008. Protein-nanoparticle interactions. Nano Today 3, 40–47.
http://dx.doi.org/10.1016/S1748-0132(08)70014-8.
Ma, R., Levard, C., Marinakos, S.M., Cheng, Y., Liu, J., Michel, F.M., Brown, G.E., Lowry, G.V., 2012. Size-controlled dissolution of organic-coated silver nanoparticles. Environ. Sci. Technol. 46, 752–759.http://dx.doi.org/10.1021/es201686j. Marambio-Jones, C., Hoek, E.M.V., 2010. A review of the antibacterial effects of silver
nanomaterials and potential implications for human health and the environment. J. Nanopart. Res. 12, 1531–1551.http://dx.doi.org/10.1007/s11051-010-9900-y. McQuillan, J.S., Groenaga Infante, H., Stokes, E., Shaw, A.M., 2012. Silver nanoparticle
enhanced silver ion stress response in Escherichia coli K12. Nanotoxicology 6, 857–866.http://dx.doi.org/10.3109/17435390.2011.626532.
McShan, D., Ray, P.C., Yu, H., 2014. Molecular toxicity mechanism of nanosilver. J. Food Drug Anal. 22, 116–127.http://dx.doi.org/10.1016/j.jfda.2014.01.010.
Misra, S.K., Dybowska, A., Berhanu, D., Luoma, S.N., Valsami-Jones, E., 2012. The complexity of nanoparticle dissolution and its importance in nanotoxicological
studies. Sci. Total Environ. 438, 225–232.http://dx.doi.org/10.1016/j.scitotenv. 2012.08.066.
Navarro, E., Piccapietra, F., Wagner, B., Marconi, F., Kaegi, R., Odzak, N., Sigg, L., Behra, R., 2008. Toxicity of silver nanoparticles to Chlamydomonas reinhardtii. Environ. Sci. Technol. 42, 8959–8964.http://dx.doi.org/10.1021/es801785m.
Navarro, E., Wagner, B., Odzak, N., Sigg, L., Behra, R., 2015. Effects of differently coated silver nanoparticles on the photosynthesis of Chlamydomonas reinhardtii. Environ. Sci. Technol. 49, 8041–8047.http://dx.doi.org/10.1021/acs.est.5b01089.
Nowack, B., Krug, H.F., Height, M., 2011. 120 years of nanosilver history: implications for policy makers. Environ. Sci. Technol. 45, 1177–1183.http://dx.doi.org/10.1021/ es103316q.
Pal, S., Tak, Y.K., Song, J.M., 2007. Does the antibacterial activity of silver nanoparticles depend on the shape of the nanoparticle? A study of the gram-negative bacterium Escherichia coli. Appl. Environ. Microbiol. 73, 1712–1720.http://dx.doi.org/10. 1128/AEM.02218-06.
Reidy, B., Haase, A., Luch, A., Dawson, K., Lynch, I., 2013. Mechanisms of silver nano-particle release, transformation and toxicity: a critical review of current knowledge and recommendations for future studies and applications. Materials 6, 2295–2350.
http://dx.doi.org/10.3390/ma6062295.
Seltenrich, N., 2013. Nanosilver: weighing the risks and benefits. Environ. Health Perspect. 121, a220–a225.http://dx.doi.org/10.1289/ehp.121-a220.
Wijnhoven, S.W.P., Peijnenburg, W.J.G.M., Herberts, C.A., Hagens, W.I., Oomen, A.G., Heugens, E.H.W., Roszek, B., Bisschops, J., Gosens, I., Van De Meent, D., Dekkers, S., De Jong, W.H., van Zijverden, M., Sips, A.J.A.M., Geertsma, R.E., 2009. Nano-silver– a review of available data and knowledge gaps in human and environmental risk assessment. Nanotoxicology 3, 109–138.http://dx.doi.org/10.1080/
17435390902725914.
Woodrow Wilson International Center for Scholars, 2013. Project on Emerging Nanotechnologies. Silver Database.
Xiu, Z., Zhang, Q., Puppala, H.L., Colvin, V.L., Alvarez, P.J.J., 2012. Negligible particle-specific antibacterial activity of silver nanoparticles. Nano Lett. 12, 4271–4275.
