Antibiofilm activity of nanosilver coatings against Staphylococcus aureus

Abstract

Implant infections due to bacterial biofilms constitute a major healthcare challenge today. One way to address this clinical need is to modify the implant surface with an antimicrobial nanomaterial. Among such nanomaterials, nanosilver is arguably the most powerful one, due to its strong and broad antimicrobial activity. However, there is still a lack of understanding on how physicochemical characteristics of nanosilver coatings affect their antibiofilm activity. More specifically, the contributions of silver (Ag)⁺ ion-mediated vs. contact-based mechanisms to the observed antimicrobial activity are yet to be elucidated. To address this knowledge gap, we produce here nanosilver coatings on substrates by flame aerosol direct deposition that allows for facile control of the coating composition and Ag particle size. We systematically study the effect of (i) nanosilver content in composite Ag silica (SiO₂) coatings from 0 (pure SiO₂) up to 50 wt%, (ii) the Ag particle size and (iii) the coating thickness on the antibiofilm activity against Staphylococcus aureus (S. aureus), a clinically-relevant pathogen often present on the surface of surgically-installed implants. We show that the Ag⁺ ion concentration in solution largely drives the observed antibiofilm effect independently of Ag size and coating thickness. Furthermore, co-incubation of both pure SiO₂ and nanosilver coatings in the same well also reveals that the antibiofilm effect stems predominantly from the released Ag⁺ ions, which is especially pronounced for coatings featuring the smallest Ag particle sizes, rather than direct bacterial contact inhibition. We also examine the biocompatibility of the developed nanosilver coatings in terms of pre-osteoblastic cell viability and proliferation, comparing it to that of pure SiO₂. This study lays the foundation for the rational design of nanosilver-based antibiofilm implant coatings.

Introduction

Implant infection is a major problem in healthcare today, causing both financial and psychological burden in several thousand patients per year globally.¹ Such infections typically result from exogenous bacteria entering the surgical site or the surgical wound, but also from bloodborne transportation of endogenous bacteria to the implantation site.² Those microbes can form a biofilm on the abiotic implant surface. In biofilms, the bacterial communities are protected and engulfed in an acidic³ extracellular matrix (ECM) consisting of proteins, deoxyribonucleic acid (DNA) and polysaccharides.⁴ This shielding reduces the effect of systemic treatments with antibiotics, or other antimicrobials, and as a result often requires additional revision surgeries, especially when caused by drug resistant strains. Almost 2 million infections occur each year in healthcare of which 50-70% can be attributed to medical devices according to the US Center of Disease Control⁵ with enormous costs. For example, the average cost for an arthroplasty hip revision surgery due to infection may reach up to 100 000 USD per patient.⁶

Scientists have tried to tackle this medical challenge with various antibiofilm approaches, spanning from chemical modification of the implant surface to trigger a biocidal effect, to physical removal of the biofilm with shear stress, pressure, lasers or electric waves, enzymatic treatments, as well as application of nanoparticles/small molecule agents for ECM degradation.^7,8,9 Especially the implant surface chemistry and modifications of it with metals or their respective oxides which are known to have an antimicrobial effect has attracted considerable attention.⁷ Zinc (Zn), copper (Cu), iron oxide (Fe₂O₃) and especially Ag are widely researched as surface coatings due to their inherent antimicrobial properties, particularly in the form of nanoparticles due to their high surface to volume ratio.^10,11,12,13 For example, small nanosilver particles release higher concentrations as Ag⁺ ions in aqueous solution than large ones,¹⁰ which is attributed to the dissolution of the native oxide layer.^14,15 Those ions are the major driving force for the antibacterial activity of nanosilver at small sizes¹⁶ and generate reactive oxygen species (ROS) that may disrupt, among others, the DNA of bacteria.^17–19 Another antibacterial mechanism of nanosilver arises through the direct contact with the bacterial cell and the following distribution of it through the high surface energy.²⁰

Ag has been explored as an antibiofilm coating material on implants in several studies because of its superior antimicrobial properties (Supporting Information (SI), Table 1). For example, Goderecci and coworkers deposited nanosilver oxide (AgO) films on Ti foil discs and studied their antimicrobial activity against S. aureus and E. coli. They found Ag⁺ ions in the cultures indicating that ions were the major driving force after reduction in zone of inhibition.²¹ Thukkaram et al. were able to reduce the colony forming units (CFUs) of the same two pathogens by 4- to 5-logs with their nanocomposites consisting of 25 nm sized nanosilver and an amorphous hydrocarbon matrix.²² One research group reduced the CFUs of S. epidermidis by 5-logs with Ag NPs having an average diameter of 25 nm.²³ Another report exploited the extract of the plant Arbutus unedo to reduce Ag⁺ ions and triggered a similar antibacterial effect against four strains. The size difference of nanoparticles was achieved by varying the plant concentration.²⁴ It was even possible to anchor Ag NPs to a polyethyleneimine self-assembled monolayer and achieve an antimicrobial response against E. coli and S. aureus.²⁵ However, there is still a bottleneck in the clinical translation of nanosilver as an implant coating primarily because of the lack of understanding on the fundamental physicochemical properties that drive the antibiofilm performance. It is not clear whether there is an optimal nanosilver size, coating thickness or if the antibiofilm activity of the nanosilver coatings is driven mainly by the released Ag⁺ ions upon contact with fluids or by the direct contact of bacteria with the nanoparticles. Hence, there is a need to systematically study these parameters and their impact on biofilm inhibition.

Here, to better understand the fundamental physicochemical parameters of nanosilver coatings that dictate their antibiofilm activity, we produced such coatings by flame aerosol direct deposition followed by in situ flame annealing on Si substrates. With this scalable and reproducible nanomanufacture process, it is possible to fabricate homogeneous nanoparticle coatings on substrates with high control over nanoparticle size, coating thickness and composition. The nanosilver size is finely tuned by supporting it on amorphous nanostructured SiO₂ and by controlling the Ag-content in the composite AgSiO₂ nanoparticles during flame synthesis.¹⁰ We performed a detailed characterization of the physicochemical and morphological properties of the as-deposited nanostructured coatings, and studied their antibiofilm activity in vitro against S. aureus, a major pathogen found on implant infections,⁵ in addition to their in vitro biocompatibility evaluation in terms of cell viability and proliferation using MC3T3-E1 pre-osteoblastic cells. Finally, we utilized a novel in vitro S. aureus antibiofilm co-incubation model by placing both nanosilver and pure SiO₂ coatings in the same well to better understand the role of the released Ag⁺ ions and the nanosilver particles separately.

Material & Methods

Deposition of nanoparticles and characterization

AgSiO₂ nanoparticles were coated on a silicon (Si) substrate using flame spray pyrolysis (FSP). The corresponding nanopowder was collected on a glass fiber filter (AlbetLabScience) with the aid of a vacuum pump (Busch). A liquid precursor consisting of stoichiometric concentrations of silver acetate (Sigma, > 99%) for a final Ag-content in the AgSiO₂ nanoparticles from 0 to 50 wt% and 0.3 M of hexamethyldisiloxane (HMDSO, > 98%, Sigma) were dissolved in 2-ethylhexanoic acid (2-EHA, > 99%, Sigma) together with acetonitrile (Sigma, > 99.8%, ratio 1:1). This solution was then fed through a capillary with a syringe pump at a flow rate of 3, 5 or 10 mL/min (New Era Pump Systems, Inc.) and out of 100 mL syringes (SGE Analytical Science) and was dispersed with 3, 5 or 8 L/min flow rate of oxygen (> 99.5 %, AGA Gas AB) (EL-FLOW Select, Bronkhorst Ruurlo). The fine dispersed mist was ignited with a support flame of methane (1.5 L/min flow rate) and oxygen (3.2 L/min flow rate) (both gases > 99.5%, AGA Gas AB). Silicon substrates (5 mm side length,) were mounted onto a water-cooled holder over the nozzle at a distance of 16 cm (SI, Figure S1). The deposition time varied from 15 to 90 s. After deposition, all coatings were in situ annealed using an ethanol flame at distance 11 cm for 20 s.²⁶ For particle characterization, transmission electron microscopy (TEM) was performed with a FEI Tecnai Spirit BioTwin (120 kV) and 2kx2k Veleta OSiS CCD camera. These images were obtained by dispersion of appropriate amounts of nanopowders in ethanol via sonication. Then a droplet was carefully put on the copper grid and dried overnight.

The crystal structure of the as deposited coatings was examined using X-ray diffraction (Rigaki MiniFlex 600). The film morphology was evaluated with scanning electron microscopy (SEM) (Phenom Pharos, Thermo Fisher, and Gemini Ultra 55, Zeiss) and the thickness was measured by breaking the substrates before the examination. The Ag⁺ ion release was measured by inductively coupled plasma mass spectrometry (ICP-MS). For ICP-MS measurements, sterilized substrates were immersed in 5 mL tryptic soy broth (TSB) each and incubated at 37 °C. After 24 hrs, supernatants were collected, transferred to Falcon tubes and acidified with ultra-pure HNO₃ prior to analysis using an Agilent 7900 ICP-MS. Spiked samples were included to rule out matrix effects. The pure silicon substrates were weighed before and after deposition to determine the mass of the coatings alone. Then the nominal Ag mass (wt%) was estimated.

For the determination of the actual Ag mass the coatings were incubated for 2h in 2ml of 20% HNO₃ at room temperature. Afterwards, 1ml of this solution was mixed with 9ml ultra-pure water and analyzed using an Agilent 5110 ICP-OES.

Antibiofilm activity against S. aureus: biofilm growth protocol and inhibition evaluation

A laboratory type strain of S. aureus (ATCC 25 923) was cultured in tryptic soy broth (TSB) medium at 37 °C overnight and subsequently diluted to an optical density (OD) of 0.01 at 600 nm. All samples were sterilized at 210 °C (Carbolite, Gero) and placed into a 48 well plate. 300 μl of the bacterial solution were pipetted in each of the substrate-containing wells and the plate was incubated at 37 °C for 24 h. The samples were carefully removed from the wells and washed 3 times in PBS, before being placed in Eppendorf tubes containing 2 mL of PBS. A rigorous washing protocol was employed to remove any attached bacteria on the substrates. The substrates in Eppendorf tubes were vortexed for 30 s, ultrasonicated (bath sonicator, VWR) for 1 min and vortexed for 30 s again.⁸ The remaining PBS was serially-diluted and plated on Luria Agar plates (pH 7.5) for overnight incubation at 37 °C. The colonies were then counted on the next day and the final numbers of the colony forming units/mL (CFUs/mL) were calculated and plotted.

For SEM characterization of the formed biofilms, the substrates were taken out of the Eppendorf tubes. The samples were thoroughly washed and placed in a new well plate. As a next step, they were incubated at 4 °C for 24 h in glutaraldehyde. After that the remaining water was drawn out by placing them in an ethanol series (5, 10, 20, 25, 50 and 80%). All fixed samples were then examined with the SEM for qualitative evaluation. For the coincubation model a 24 well plate was chosen to fit in two coated substrates at the same time. For this reason, the volume of the bacterial solution was also adjusted to 400 μL

In vitro biocompatibility

Pre-osteoblastic cell culture maintenance and cell seeding

Pre-osteoblastic cells MC3T3-E1 derived from the murine calvaria, were cultured in alpha-MEM minimum essential media (PAN Biotech, Germany), supplemented with 10% fetal bovine serum (PAN Biotech, Germany), 100 units/ml of penicillin, 100 mg/ml of streptomycin (PAN Biotech, Germany), 1% amphotericin (Gibco, USA). Cells between passages 7 and 12 were used for the experiments. Prior to cell seeding, the samples were immersed in culture media for 24 h to promote an Ag⁺ ion burst release that is expected to affect the cell viability. The cell-loaded samples were maintained in a humidified atmosphere under 5% CO₂ at 37 °C and the culture media were changed twice a week. A tissue culture treated polystyrene (TCPS) was used as control surface.

Cell viability and proliferation assay

The viability and proliferation of MC3T3-E1 pre-osteoblasts cultured on AgSiO₂-coated biofilm was quantitatively assessed using the resazurin-based metabolic assay PrestoBlue® (Invitrogen, USA) as previously described.²⁷ Briefly, the coatings were placed into 48-well plates and each sample was seeded with 10⁴ cells. At each experimental time point of 2, 7 and 12 days of culture, the PrestoBlue® reagent was added directly to the wells at a 1:10 dilution in culture medium and incubated at 37 °C for 60 min, before measuring the absorbance at 570 and 600 nm in a spectrophotometer (Synergy HT, Biotek, USA). The metabolic activity of living cells was correlated with the cell number by means of a calibration curve. Two independent experiments were performed in triplicates.

Statistical analysis

All statistical analysis was performed in GraphPad Prism 9 (La Jolla, USA) using One-way ANOVA comparing the Ag-loaded samples to the control (pure SiO₂). For the in vitro cytotoxicity studies, the experimental data was analysed using two-way ANOVA followed by Dunnett’s multiple comparisons test between groups. The symbols designate as follows unless stated otherwise for an individual figure: * = p<0.05, ** = p<0.01, n.s. = statistically non-significant difference compared to the TCPS control surface.

Results & Discussion

Physicochemical and morphological characterization of nanosilver coatings

Flame synthesis allows for nanoparticle production and deposition on surfaces in a single step. With this process, Si substrates were coated with nanoparticle films consisting of composite AgSiO₂ nanoparticles and in situ annealed to increase their adhesion, cohesion and structural stability.²⁶ By carefully tuning the liquid precursor stoichiometric composition, six different nominal Ag-contents (weight %, wt%) were prepared: 0 (pure SiO₂), 10, 20, 30, 40 and 50 wt% (from now on xAgSiO₂, with x being the Ag-content). For example, the nominal content of 1 g of the 10AgSiO₂ sample was 0.1 g Ag and 0.9 g SiO₂. There is a good agreement between the actual and nominal Ag-content in the xAgSiO₂ samples, in agreement with the literature¹⁵ (SI, Figure S2). Figure 1a,b shows TEM images of the as-prepared pure SiO₂ nanoparticles and the 50AgSiO₂ ones. In both micrographs, the characteristic fractal-like agglomerated structure of the as-prepared nanoparticles can be observed. The presence of the electron-dense high atomic number Ag nanoparticles relative to the lighter SiO₂ nanoparticles is evident from the dark contrast areas in Figure 1b.¹⁴ The presence of Ag in the composite AgSiO₂ nanoparticles is further validated here by the macroscopic examination of the as-deposited coatings on the Si substrates as shown in Figure 1c with two substrates for each Ag-content sample and the uncoated Si substrates for comparison. The coating from the pure SiO₂ (Ag-content x = 0 wt%) appears white in colour, while the colours of all the AgSiO₂ coatings range from bright yellow to dark brown for increasing Ag-content. This colour originates from the plasmonic properties of nanosilver and is in agreement with the literature.^10,16,28,29 Figure 1d shows the XRD patterns of all the as-deposited coatings on the Si substrates for increasing Ag-content and the characteristic (111) peak of Ag metal is indicated with a dotted line (the sharp peaks at 29˚ and 33˚ are attributed to the Si substrate). In agreement with the literature, there is no evidence for oxide formation by XRD analysis, even though such Ag nanoparticles have one or two monolayers of an oxidized surface.^15,30 As the Ag-content increases, the main diffraction peak for Ag metal (111) becomes narrower indicating increasing crystal sizes. These average crystal sizes as estimated by the main diffraction peak (36-40˚) are plotted in Figure 1e as a function of the Ag-content. For an increasing Ag-content the Ag crystal size grows from 4 nm to 11 nm in agreement with the literature.^14,31 Moreover, this also highlights the high control of Ag size with the nanomanufacture process employed here.

Fig. 1.

TEM images of the 0 (a) and 50AgSiO₂ nanoparticles (b) showing the characteristic flame-made fractal-like agglomerate nanostructures. Scale bar is the same for both images. (c) Photograph of the Si substrates with the various AgSiO₂ coatings (Ag content x = 0 to 50 wt%) and a blank Si substrate. As the Ag loading increases the colour changes from yellow to dark brown/black. (d) XRD patterns of all xAgSiO₂ samples highlighting the main diffraction peak of silver metal. For smaller loadings, this peak is broader indicating smaller crystal sizes. (e) Average Ag crystal size d_XRD plotted as a function of nominal Ag-content (wt%). (n=3)

To further examine the morphology of the deposited nanostructured coatings here, we performed SEM analysis as shown in Figure 2 with top-view (a-c) and side view (d-f) images of such coatings for Ag-contents x = 0, 30, 40 and 50 wt%. The coatings exhibit similar morphology for the different Ag-contents and are homogeneous with high porosity, characteristic for their production method.²⁶ The film thickness is 15 μm ± 2 μm for deposition time t_d = 60 s in all samples. The highly porous nanostructure likely facilitates the release of Ag⁺ ions from the nanosilver particles throughout the whole coating.^10,16 Figure 3a shows the Ag⁺ ion release after 24 h at 37 °C from the deposited coatings as a function of Ag-content upon their immersion into bacterial culture medium TSB. The Ag⁺ ion release increases for increasing Ag-content. Similar Ag⁺ ion release was observed in pure water (SI, Figure S3). For increasing Ag-content there is higher Ag-mass present on the substrates that in turn dictates the Ag⁺ ion concentration. The Ag⁺ ion release occurs within the first few hours upon incubation, that are important for the initial bacterial adhesion, and slowly reaches equilibrium after 24 h (SI, Figure S4). Figure 3b shows the fraction of Ag as Ag⁺ ions as a function of average Ag crystal size from the data in Figure 3a. Upon normalizing the Ag⁺ ion release to the total Ag mass present in solution, it becomes evident that small nanosilver particles release higher fraction of their mass as Ag⁺ ions, which is well in line with the literature.^10,29

Fig. 2.

(a-d) SEM images of 4 different Ag-contents (x = 0, 30, 40, 50 wt%) from top- (a-c) and side-view (d-f). The coatings exhibit a porous homogeneous morphology.

Fig. 3.

(a) Ag⁺ ion release as a function of nominal Ag-content x (wt%) in xAgSiO₂ coatings. (b) Fraction of Ag as Ag⁺ ions calculated by panel (a) as a function of average nanosilver crystal size. (n=3)

Antibiofilm activity against S. aureus

The antibiofilm activity of the nanosilver coatings was evaluated against S. aureus. Figure 4 shows the SEM images of nanoparticle coatings on Si substrates for all Ag-contents after their 24 h incubation at 37 °C with S. aureus bacterial suspension (OD₆₀₀ = 0.01) and subsequent gentle washing and drying.⁸ On the pure SiO₂ coating, the complete surface was covered with attached 3D-structured microcolonies of S. aureus, indicative of early biofilm formation, which differs from non-adherent planktonic state bacteria.³² With increasing Ag-content of the coatings, fewer bacterial colonies/biofilm were found on the substrates, and beyond the x = 20 wt% Ag sample, there was no evidence of surviving microcolonies on the substrates.

Fig. 4.

(a-f) SEM images of coatings with different Ag-content (x = 0 – 50 wt%) after 24 h incubation in the presence of S. aureus.

For further quantification of the biofilm growth inhibition of the developed nanostructured coatings, the retrieved bacteria were cultured and plated to count their CFUs/mL (Figure 5a) after rigorous washing (SI, Figure S5).³³ On pure SiO₂ there is a high bacterial growth that gradually decreased with increasing Ag-content, reaching a 5-log CFUs/mL reduction at the highest Ag-content used (50AgSiO₂). Reductions greater than a log unit of 3 is considered “bactericidal”.³⁴ In the present study and experimental conditions, this threshold was reached with 30AgSiO₂ and similar reductions have been observed in the literature,²² even though the log-reduction often depends on the experimental conditions. The 10AgSiO₂ substrate resulted in a slightly higher number of CFUs/mL forming on the surface, compared to the control. This might be explained by a potential metabolic response of the biofilm community to sublethal metal concentrations. Indeed, most bacteria react to environmental stresses by producing more biomass as a defence mechanism.³⁵ However, this difference was not statistically significant in the present experimental model. Figure 5b shows a high magnification SEM image of the pure SiO₂ coating showing the 3D structured microcolonies and the typical roundly/globally formed structure of S. aureus. In contrast, Figure 5c-e show the structure of bacteria grown on coatings from 40AgSiO₂ and 50AgSiO₂, in which the envelope and the structure of the cells appears compromised. In particular, in the middle image of Figure 5d,e, the bacterial cell appears disrupted (yellow arrows), an indication of being partially or completely lysed. The bright spots (yellow arrows) on the bacterial cell might be the result of accumulated charge on non-continuous areas on the membrane or indicate any Ag metal accumulation (with higher Z-contrast). However, EDX analysis (not shown) did not reveal any Ag metal accumulation, suggesting that in these areas the bacterial membrane may indeed be damaged. Similar observations have been reported by Taglietti et al. where Ag nanoparticles accumulated outside of the cell envelop and later penetrated and disrupted it.³⁶

Fig. 5.

(a) CFUs/mL-counts of retrieved bacteria after 24 h incubation with coatings from all Ag-contents. Each symbol represents the data obtained from the same technical replicate (n > 9). (b) High magnification SEM images of S. aureus bacteria grown on the pure SiO₂ with an intact and globally shaped envelope. **** p < 0.0001 (c-e) Damaged and compromised cell structures of bacteria attached to substrates coated with 40 and 50AgSiO₂. Arrows indicate damage sites on the bacterial membrane.

In order to study the effect of the thickness of the nanostructured coatings on their antibiofilm activity, we performed flame aerosol deposition for different durations of the 30AgSiO₂ sample. By keeping the Ag-content the same (30AgSiO₂), the nanosilver size is identical in all different thicknesses. Due to the high porosity of the deposited coatings³⁷ (approx. 88% for the 30AgSiO₂, please see SI, Figure S6 for more SEM images and calculation), the Ag⁺ ion release occurs from the whole film affecting the antibiofilm activity. The coating thickness is finely tuned by controlling the deposition duration during the nanoparticle aerosol self-assembly on the substrates. Figure 6a-c shows three cross-sectional SEM images of the different 30AgSiO₂ film for three deposition durations (t_d = 15, 30 and 45 s), in addition to the t_d = 60 s as shown in Figure 2d. Figure 6d shows the measured thickness of the coatings as a function of deposition time t_d. Now upon incubating all four samples with the different coating thicknesses with S. aureus bacterial suspension, a clear thickness-dependency on the resulting CFUs/mL was observed (Figure 6e). While the 8 μm thick coatings (after t_d = 30 s) induced a slight, but not statistically significant, decrease in CFUs/mL, the 12.5 μm thick coatings (t_d = 45 s) showed a reduction of more than 2-log units. Following a trend, the 15 μm thick deposited substrate of 30AgSiO₂ reached 3-log units, which is the same reduction as in the first biofilm inhibition experiment (Figure 5a). This thickness-dependent antibiofilm activity highlights that there is a minimum nanosilver content, and thus released Ag⁺ ion concentration, required to achieve biofilm growth inhibition. In fact, upon plotting the biofilm growth CFUs/mL % inhibition as a function of the Ag⁺ ion concentration present in solution from all tested coatings here (for all Ag-contents and all thicknesses, Figure 7), there is a clear trend showing that almost all biofilm growth is inhibited for Ag⁺ ion concentrations >1mg/L. Ag⁺ ion concentrations greater than 1 mg/L lead to a bactericidal activity (3 logs reduced) which reaches a plateau at around 1.5 mg/L (5 logs).

Fig. 6.

(a-c) SEM images indicating film thicknesses of different deposition times (side view) of the 30AgSiO₂ sample. (d) Thicknesses plotted as a function of the deposition time (n=3). (e) CFUs/mL as a function of different deposition times and the positive control (n = 3). *** p < 0.001

Fig. 7. A summary of the different silver containing coatings used ins this study and the inhibition of CFUs as a function of the measured and calculated Ag⁺ ion release.

For a further quantification the role of the released Ag⁺ ions in comparison to that of the direct bacterial contact with the nanosilver coatings, we designed a co-incubation experiment (Figure 8a). In this approach, a 50AgSiO₂ coating, exhibiting the highest Ag⁺ ion release (Figure 3c) and antibiofilm activity (Figure 5a), is placed in the same well with a pure SiO₂ coating and both are covered with a S. aureus suspension. Furthermore, to examine the effect of nanosilver size on coatings with the same composition and thickness, three different 50AgSiO₂ samples were produced featuring different Ag particle sizes (d_XRD = 6, 11 and 18 nm) and deposited for adjusted durations to obtain similar coating thicknesses (SI, Figure S7 also shows their Ag⁺ ion release). After 24 hours, the CFUs/mL on both the 50AgSiO₂ and the pure SiO₂ substrates retrieved from the same well are counted and compared to the control conditions (i.e. no AgSiO₂ coating present but only pure SiO₂), allowing for the investigation of the effect from the released Ag⁺ ions only. Any biofilm growth inhibition on the pure SiO₂ coating would originate from the activity of the Ag⁺ ions released from the nanosilver coating in the same well. Figure 8b shows the CFUs/mL (the higher CFU-counts in these experiments than in Figure 5 are a result from the different experimental conditions, please see Materials and Methods section) in relation to the Ag crystal sizes of the 50AgSiO₂ (d_XRD = 6, 11 and 18 nm) and the pure SiO₂ of that same well. In the case of the AgSiO₂ substrate, there was an Ag size-dependent increase of CFUs/mL. In particular, the 6 nm nanosilver displayed the lowest CFUs/mL counts for the 50AgSiO₂, as well as for the pure SiO₂ coating. Interestingly, the 11 nm 50AgSiO₂ substrate showed a significant inhibition, whereas the pure SiO₂ coating in that same well did not show a significant reduction, when compared to the control (pure SiO₂, no AgSiO₂). Finally, the largest size of nanosilver particles used in this study did not exhibit any significant biofilm inhibition when compared to its corresponding control. Barzan et al. showed a similar size dependency of the antibacterial response of nanosilver against E. coli.³⁹ It should be noted that antibacterial and antibiofilm experiments are highly dependent on the experimental conditions. For example, even though here with our largest nanosilver we could not detect a significant antibiofilm effect, other studies that utilize even larger nanosilver structures exhibit some antibacterial effect. However, to be able to compare different studies³⁸, similar, if not identical, conditions have to be present including initial bacterial load and volume, growth medium incubation period. Nonetheless, these results indicate that the released Ag⁺ ions are the major driving force behind the antibiofilm activity of nanosilver, especially for the nanosilver with the smallest size (d_XRD = 6 nm). However, with increasing nanosilver size (d_XRD = 11 nm), the released Ag⁺ ions did not inhibit the biofilm growth on the pure SiO₂ coatings to the same extent as on the AgSiO₂ coatings. This might be attributed to additional antibiofilm activity from direct bacterial contact with the nanosilver, or perhaps also to an increased local Ag⁺ ion concentration at the proximity of the AgSiO₂ coatings.

Fig. 8.

(a) Illustration of the co-incubation experiment with two substrates in one well. The pure SiO₂ coating shares a well with the 50AgSiO₂ coating. The Ag⁺ ions are released into the environment upon contact with the growth medium and affect the growth on both samples. (b) the CFUs/mL plotted for three different Ag crystal sizes of the 50AgSiO₂ coatings and the CFUs/mL of the corresponding pure SiO₂ coatings sharing a well with those 50AgSiO₂ coatings (in comparison to the control well in which there is only a pure SiO₂ coating) (n = 3). *** p < 0.001, ** p < 0.01

In vitro biocompatibility

The in vitro biocompatibility of all AgSiO₂ coatings was evaluated using the PrestoBlue® cell viability assay with MC3T3-E1 pre-osteoblastic cells directly grown on substrates (preconditioned for 24 h in cell-free alpha-MEM), to simulate the wound healing phase that starts days/weeks after the initial surgery,⁴⁰ and by when the released Ag⁺ ions would have eliminated any bacteria and be depleted. This is especially important for implants susceptible to surface colonization where a quick local response is crucial.⁴¹ Figure 9 shows the cell number per well as a function of the Ag-content of the six different samples (Ag-content from 0 to 50 wt%) after 2, 7 and 12 days. All substrates show an increasing cell number over time indicating that the cells can grow and proliferate on all coatings. When compared to the pure SiO₂ coating, only the highest Ag-containing coating of 50 wt% show a significantly lower difference in cell number. The other Ag-containing coatings exhibit slightly lower cell number values than the pure SiO₂ one (not statistically significant), especially after 12 days, probably attributed to any residual Ag⁺ ion release. However, all coatings (Ag-contents x: 0-50 wt%) exhibit lower cell numbers after 12 days than the tissue culture treated polystyrene surface (SI, Figure S8). Hence even though the cells are alive and proliferate on the AgSiO₂ coatings, improvements on the coating composition and/or morphology might enhance their biocompatibility even further.

Fig. 9.

The cytotoxicity of the coatings is evaluated by the PrestoBlue® cell viability assay. Pre-osteoblastic cells MC3T3-E1 are grown on the pre-conditioned coatings (Ag-content x = 0-50 wt%) and the number of living cells per well is determined photometrically after 2, 7 and 12 days in culture. (n=3) ** p < 0.005

Conclusions & Outlook

To summarize, in this study we produced composite nanostructured AgSiO2 coatings by flame aerosol direct deposition that allows for a facile control over the nanoparticle composition and size. Flame aerosol nanoparticle deposition offers several advantages for nanostructured coatings fabrication as it is a rapid (coatings are produced within seconds) nanomanufacture process that combines particle formation and coating fabrication in a single step. Here, we systematically varied the Ag-content in the AgSiO₂ nanoparticles and the coating thickness tuning the average Ag size and Ag⁺ ion release in aqueous solutions. We then studied the antibiofilm activity of the developed coatings against S. aureus, a clinically relevant pathogen often present in biofilms associated with implant infections. The coatings with the highest Ag⁺ ion release exhibited the strongest antibiofilm activity indicating that Ag⁺ ions dominate the antibiofilm activity of nanosilver. This is especially important for solving the question which mechanism of nanosilver (nanoparticles vs ions) is responsible for its antibiofilm properties²⁰. Our co-incubation experiments illustrate that nanosilver-coated implants featuring Ag particle sizes of 6 nm exhibit long-ranging antimicrobial effects due pronounced Ag⁺ ion release. For larger Ag particle sizes, antibiofilm activity is attenuated and contained to the local environment. The results here further strengthen the observed size-dependent antimicrobial activity of nanosilver against planktonic bacteria^{10,14–16,39} and extend this mechanism to biofilm-forming bacteria. It should be noted that antibacterial and antibiofilm experiments are highly dependent on the experimental conditions. For example, even though here with our largest nanosilver we could not detect a significant antibiofilm effect, other studies that utilize even larger nanosilver structures exhibit some antibacterial effect. However, to be able to compare different studies,³⁸ similar, if not identical, conditions have to be present including initial bacterial load and volume, growth medium incubation period. These insights may prove valuable in the design of potent nanosilver-based antibiofilm implant coatings with high cytocompatibility where accurate control over Ag⁺ activity is imperative. Future studies could focus not only on the in vivo demonstration of the antibiofilm activity of such coatings but also on their successful osseointegration.