Antimicrobial activity of Ti-ZrN/Ag coatings for use in biomaterial applications

Severely broken bones often require external bone fixation pins to provide support but they can become infected. In order to reduce such infections, novel solutions are required. Titanium zirconium nitride (Ti-ZrN) and Ti-ZrN silver (Ti-ZrN/Ag) coatings were deposited onto stainless steel. Surface microtopography demonstrated that on the silver containing surfaces, Sa and Sv values demonstrated similar trends whilst the Ra, average height and RMS value and Sp values increased with increasing silver concentration. On the Ti-ZrN/Ag coatings, surface hydrophobicity followed the same trend as the Sa and Sv values. An increase in dead Staphylococcus aureus and Staphylococcus epidermidis cells was observed on the coatings with a higher silver concentration. Using CTC staining, a significant increase in S. aureus respiration on the silver containing surfaces was observed in comparison to the stainless steel control whilst against S. epidermidis, no significant difference in viable cells was observed across the surfaces. Cytotoxicity testing revealed that the TiZrN coatings, both with and without varying silver concentrations, did not possess a detrimental effect to a human monocyte cell line U937. This work demonstrated that such coatings have the potential to reduce the viability of bacteria that result in pin tract infections.


Results
Chemical composition. Energy Dispersive X-Ray (EDX) analysis was undertaken on the four Ti-ZrN/Ag coatings. The silver content demonstrated averages of 6.0%, 15.6% and 24.7% for magnetron powers of 90 W, 130 W and 160 W respectively (Fig. 1). Backscattered electron images were obtained using the backscattered electron detector, which resulted in visualisation of the chemical elemental distribution on the surface. Due to heavier elements being able to backscatter electrons more freely, they appeared on the image as a brighter phase resulting in the silver (heavier than zirconium) to be visualised as white nanoparticles thus showing their distribution throughout the surfaces (Fig. 2). The Ti-ZrN was examined and characterised as a negative control, which showed a relatively homogenous surface, demonstrated by a similar colour phase ( Fig. 2A). The Ti-ZrN/6.0 at% Ag showed silver particles of various sizes around 8 nm, with large variations being observed in their distribution (Fig. 2B). The Ti-ZrN/15.6 at% Ag displayed a different morphology, with the overall size of the silver nanoparticles being around 5 nm and these were more regularly distributed but more tightly packed (Fig. 2C). The Ti-ZrN/24.7 at% showed mixed sized, silver particles (ranging from 10-20 nm), which were closely spaced (Fig. 2D).
Structural Morphology. The surface microtopography was determined to visualise and quantify surface features that were of microbial dimensions. Results from the linear profiles demonstrated that the Ti-ZrN coatings did not alter surface microtopography, even in the presence of an increased silver concentration. The average width of the features from the linear profiles ranged between 2.9 µm (Ti-ZrN) and 2.0 µm (Ti-ZrN/6.0 at% Ag) ( Fig. 3a-d) and were demonstrated to be larger than the average of a size of a Staphylococcus spp. cell (1 µm). However, the average depth values were in the range of 0.029 µm-0.037 µm and therefore were not significant in relation to the size of a Staphylococcus spp. cell (Figs 3a-d).
The white light profilometer was used to obtain roughness values of the surfaces at a microtopography scale. Of the five surfaces, no significant difference (p > 0.05) was demonstrated in the S a values (arithmetic average height) (0.02-0.03 µm) although for the silver containing surfaces, the S a values demonstrated the same trend as the S v values (surface valley depth) (Fig. 3e). S p values, which represent the surface peak height showed that the lowest value recorded was on the stainless steel control and the pure Ti-ZrN (0.08 µm and 0.10 μm, respectively), whilst the silver coatings decreased from the greatest S p value (0.21 µm) for the Ti-ZrN/6.0 at% Ag coating and decreased with increasing silver content to 0.09 µm for the Ti-ZrN/24.7 at% Ag coating (Fig. 3e). The S v values (valley depths), demonstrated that the stainless steel and Ti-ZrN coatings were similar (0.23 µm and 0.21 µm respectively), whilst although lower the Ti-ZrN/6.0 at% Ag and Ti-ZrN/24.7 at% Ag coatings also demonstrated similar values (0.14 µm and 0.11 µm respectively), whilst the Ti-ZrN/15.6 at% Ag) demonstrated the greatest S v value (0.34 µm) (Fig. 3e).
The hydrophobicity of the coatings were calculated by obtaining contact angles using two polar solvents (water and formamide) and one non-polar solvent (1-bromonaphthalene), in order to calculate the ΔG iwi , (the quantitative measure for hydrophobicity/hydrophilicity). The results demonstrated that all five surfaces tested possessed negative ΔG iwi values indicating a hydrophobic nature. The pure Ti-ZrN was the least hydrophobic at ΔG iwi −40, whilst the Ti-ZrN/15.6% Ag was the most hydrophobic at ΔG iwi −77 ( Fig. 5; Table 1). All the surfaces were significantly different to one another, with the exception being between the 6.0% Ag and 15.6% Ag coatings ( Fig. 5; Table 1).
Ion coupled plasma -atomic emission spectroscopy (ICP-AES) was used to quantify leaching of elements from the coatings (silver and zirconium). The results demonstrated that the silver coatings leached silver (Fig. 6) into the surrounding broth with an increase in silver ions leached with increasing silver concentration. After 24 h, the number of silver ions leached from the silver containing surfaces was 0.15 ppm, 0.69 ppm and 0.25 ppm (Ti-ZrN/6.0 at% Ag, Ti-ZrN/15.6 at% Ag and Ti-ZrN 24.7 at% Ag respectively).
Antimicrobial analysis. In order to determine the viability of the microorganisms retained, LiveDead ™ staining was used (Fig. 7). Overall the numbers of live (Fig. 7a) and dead cells (Fig. 7b) retained were greatest on the stainless steel surfaces. Both the numbers of live and dead S. epidermidis cells were retained in fewer numbers (S. epidermidis percentage coverage range live 0.00-0.07%; dead 0.02-0.07%) than S. aureus (percentage coverage range 0.47-4.69%; dead 0.65-2.75%) (Fig. 7a). The results demonstrated that the Ti-ZrN and Ti-ZrN/ Ag surfaces contained less viable S. aureus and S. epidermidis cells on the surface than the stainless steel control, and a decrease in live cells was observed as the silver content increased (live cells S. aureus 2.78%, 1.30%,   (Fig. 7a). On the silver containing surfaces the percentage coverage of the retained live bacteria followed the same trend as the surface roughness S p values. The Ti-ZrN/15.6at.Ag (1.71%) and Ti-ZrN 24.7 at% (1.77%) surfaces demonstrated a significant increase in dead S. aureus cells compared to the Ti-ZrN (0.65%) and Ti-ZrN/6.0 at% Ag (0.52%) coatings, however there was no significant increase in dead cells between the Ti-ZrN/15.6 at% Ag and the Ti-ZrN/24.7 at% Ag surfaces. The number of dead S. epidermidis cells (Fig. 7b) was found to be greatest on the Ti-ZrNAg containing surfaces at 0.03%, 0.05%, 0.06% and 0.07% on the Ti-ZrN, Ti-ZrN/6.0 at% Ag, Ti-ZrN/15.6 at% Ag and Ti-ZrN 24.7 at% Ag respectively.
CTC-DAPI fluorescent staining demonstrated the metabolic activity of the bacteria and the counterstain of the total bacterial cells (Fig. 8). The results demonstrated that when compared to the stainless steel of Ti-ZrN surfaces, the silver present reduced the coverage of respiring S. aureus cells but an increase in silver content increased the amount of respiring bacteria (0.94%, 0.20%, 0.35% and 0.76% on the Ti-ZrN, Ti-ZrN/6.0 at% Ag, Ti-ZrN/15.6 at% Ag and Ti-ZrN/24.7 at% Ag surfaces respectively. The results obtained from the S. epidermidis showed similar percentage coverage on each surface (range 0.14-0.22%), with an exception being the Ti-ZrN/24.7 at% Ag where no viable bacteria were observed. Cytotoxicity Assay. In order to assess the surface coatings suitability for application in conjunction with biomaterials, such as pin tract infections cytotoxicity assays were carried out (Fig. 9). The Ti-ZrN and Ti-ZrNAg surface coatings were tested against a commercial human monocyte cell line U937 utilising a cell viability (MTS) method. Results demonstrated that all of the surfaces demonstrated little to no cytotoxic effect. The negative control (untreated cells) gave the highest percentage viability (98.95%), followed by TiZrN/15.6 at% Ag (98.54%). The

Discussion
A range of surface coatings were produced on medical grade stainless steel in order to determine if the number of viable bacteria retained could be reduced. EDX analysis showed that the silver contents of the coatings were Ti-ZrN/6.0 at% Ag, Ti-ZrN/15.6 at% Ag and Ti-ZrN/24.7 at% Ag silver for magnetron powers of 90 W, 130 W and 160 W, respectively. All of the coatings had a stoichiometric zirconium and nitrogen ratio. The nitrogen content appeared to be lower than expected, however this could be due to nitrogen's low molecular weight making it difficult to detect using EDX. Backscattered electron detection demonstrated the distribution of silver particles throughout the surface coatings. The silver particles were situated close together, thus allowing antimicrobial activity if bacteria were to make contact with the surface (via contact with the silver particle). Throughout this study, a higher silver concentration correlated with a higher the antimicrobial efficacy (as observed in live percentage coverage). This contact-initiated mode of antimicrobial action is well established in the case of silver nanoparticles, which provide a much greater degree of access due to silver ions due to their large surface area 20 . Upon contact with the bacteria, the silver nanoparticles may leach ions which attach to the cell membrane and penetrate the bacterial cell 20 , interact with sulphur-containing proteins (as well as phosphorous containing compounds such as DNA), and attach to thiol groups 24,25 . Thus, silver has shown to inhibit respiration and growth, by damaging protein and enzymes, ultimately leading to loss of cell membrane integrity and cell death 20,[26][27][28] . Previous studies into nitride/silver nanocomposites have demonstrated that not all microorganisms are susceptible to leaching of silver, whereas composite-coatings have been suggested to have a contact kill efficacy 16,29 .
The microtopography of a substrate is a vital parameter when evaluating antimicrobial efficacy, specifically when surfaces have features around the micron range, which is similar to that of the dimension of a bacterium (in this case Staphylococci), which can result in an increase in bacterial retention 30,31 . The optical profilometry line profiles, demonstrated that the deposition of thin films (1 µm thick) onto stainless steel using magnetron sputtering did not affect the microtopography of the surfaces. This was confirmed via S a values which were similar for all surfaces indicating that the coatings did not affect the micro-roughness of the surfaces. S v demonstrated the same trend as the S a values which also was the same trend observed of the formazan production of S. epidermidis in the CTC-DAPI assays. However, the S p value demonstrated the same decreasing trend seen on the silver surfaces of the number of the live cells of both S. aureus and S. epidermidis suggesting that as the concentration of silver increases, so do the tops of the nanoparticles that protruded from the matrix of the Ti-ZrNAg which results in a greater antimicrobial kill. This was confirmed by the AFM values which demonstrated that with increasing silver content, the nanotopography of the surfaces also increased.
Hydrophobicity varied throughout the surface coatings and with the silver containing coatings, surface hydrophobicity was found to follow the same trend as the S a and S v surface results. All surfaces exhibited hydrophobicity, with Ti-ZrN/15.6 at% Ag the most hydrophobic (∆G iwi −77.22) and Ti-ZrN the least (∆G iwi −39.63) 32 .
Zirconium and silver ion leaching was analysed via ICP-AES. In the case of silver, leaching of the ions into the surrounding broth was observed. The minimum inhibitory concentration (MIC) of silver against Staphylococci spp., has been demonstrated to be around 10 ppm 33 . Sandstrom (2011) found that silver produced an MIC of 8 ppm towards S. aureus and a minimum bactericidal concentration (MBC) of 32 ppm 34 . However, in our work, the leaching of the silver ions into the broth was much lower than the bacteriostatic value obtained for the MIC. This may be a result of the bacteria being in contact with the silver on the surfaces, since work by Vaidya et al., (2017) demonstrated that silver in ionic form did not demonstrate high levels of antimicrobial activity when compared to other metal ions 35 . In the case of zirconium ions, all the surfaces demonstrated extremely low levels of leaching. In a previous study, it had been established that this level of leaching would not be enough to damage human cells -with a toxic amount being around 150 ppm 36 .
Cytotoxicity results demonstrated that the Ti-ZrN and Ti-ZrNAg surface coatings had no detrimental effect upon the human monocyte cell line U937. Zirconium is already widely used in the construction of prosthetic devices, due to its excellent biocompatibility to human cells and good mechanical and chemical properties [37][38][39] . Further, when exposed to oxygen, zirconium becomes zirconium oxide which is also biocompatible, thus making it an ideal material for application within pin tracts 40 . The oligodynamic characteristics of silver are well documented throughout the literature 19,41,42 . This, coupled with good biocompatibility has led to silver being used in a plethora of biomedical applications including, wound dressings 43-45 , antibacterial cream for burns patients (silver sulfadiazine) 46,47 and surface coatings for equipment/in-dwelling patient devices [48][49][50] . In light of the cytotoxicity results, the application of TiZrN in combination with silver for pin tracts, has promising potential to reduce bacterial colonisation, without possessing a detrimental effect to the host cells.
LiveDead ™ staining results demonstrated a decrease in live S. aureus cells on the silver containing coated samples, and an increase in antimicrobial efficacy (reduction of live cells) as silver concentration increased. An increase in dead percentage coverage was observed against the coatings with a higher silver concentration with TiZr-N/15.6 at% Ag and TiZr-N/24.7 at% Ag. For S. epidermidis, an extremely low percentage coverage of live and dead cells were observed on all the substrata. A decrease in the percentage coverage of live S. epidermidis cells was also observed with increasing silver concentration. The LiveDead ™ staining kit, stains all bacterial cells green using CYTO-9, whilst cells with loss of membrane integrity are the over-stained using propidium iodide (red) 51 . However, cells may still be viable with a compromised membrane, therefore other methods should be used to check cell viability, especially if silver ions have the ability to enter the cell via channels, leading to DNA damage and growth inhibition 52,53 .
The CTC-DAPI staining assay, displayed the respiration of individual microorganisms, due to the reduction of tetrazolium (in the electron transport chain of the microorganism), leading to the production of insoluble, fluorescent formazan [54][55][56] . Results demonstrated that S. aureus displayed more respiring cell coverage than S. epidermidis. It was demonstrated that an increase in silver content did not decrease cell respiration as expected, but there was a significant increase in S. aureus respiration in comparison to the stainless steel control. S. epidermidis SCIEntIfIC REPORTS | (2018) 8:1497 | DOI:10.1038/s41598-018-20013-z resulted in similar respiration on all the surfaces with the exception of the Ti-ZrN/24.7 at% Ag suggesting that the silver ions affect the bacteria in different ways 57,58 . Growth inhibition from bacteriostatic antibiotics have been shown to supress cellular respiration, whilst bactericidal antimicrobial formations are associated with accelerated respiration 50 . Therefore, it might be suggested that more than one assay should be used in order to confirm cell viability.

Conclusion
The Ti-ZrN/Ag coatings demonstrated antimicrobial efficacy. Surface hydrophobicity followed the same trend as the S a and S v values. S p values demonstrated the same trend as the increase in silver concentration in the surfaces since the silver produced nano-peaks. The changes in nanotopography with increasing silver concentration was confirmed with AFM. The increase in the silver concentrations of the surfaces resulted in a decrease in bacterial viability. Cytotoxicity testing revealed the Ti-ZrN/Ag coatings elicited no detrimental effect against a human monocyte cell line. This work demonstrated that such coatings have the potential to reduce the viability of bacteria that result in pin tract infections.

Materials and Methods
Bacterial Preparation. The microorganisms Staphylococcus aureus NCTC 8532 and Staphylococcus epidermidis NCTC 11047 were used throughout this study. Stock cultures were stored at −80 °C. When required the cultures were thawed and inoculated onto nutrient agar media (Oxoid, UK) and incubated at 37 °C for 24 h. The inoculated agar plates were kept refrigerated at 4 °C and replaced every four weeks to maintain the genotype. Sterile brain heart infusion (BHI) broths (Oxoid, UK) (10 mL) were inoculated with S. aureus or S. epidermidis and incubated overnight in an orbital incubator (610 g) at 37 °C for 24 h. Cultures were removed from incubation and the cells were washed in sterile membrane filtered water (MilliporeElix, USA) (10 mL) by centrifuging at 600 g for 8 min. The supernatant was removed and the cells were re-suspended in sterile distilled water. Cells were diluted to an optical density (OD) of 1.0 ± 0.05 at 540 nm using a spectrophotometer (Jenway 6305, UK), calibrated against distilled water. Cell numbers were determined in colony forming units/mL (CFU/mL) using serial dilutions and were determined to be 0.33 × 10 8 CFU/mL for S. aureus and 0.76 × 10 8 CFU/mL for S. epidermidis. The diluted cell suspension (100 µL) was spread and repeated in duplicates on brain heart infusion agar (Oxoid, UK) and incubated at 37 °C for 24 h. Surface coatings. The surface coatings were deposited onto medical grade 316 L stainless steel (Aalco, UK), at with a range of different powers used on the silver target in order to produce a range of silver concentrations in the coatings (90 W, 130 W and 160 W). The Ti in the surface coating title denotes that a titanium interlayer was used to improve the adhesion of the coating to the surface of the stainless steel, however the titanium was not evident as part of the chemistry of the surface coatings.

Scanning Electron Microscopy (SEM), Energy-Dispersive X-ray (EDX) Spectroscopy and
Electron Backscattering analysis. Scanning electron microscopy was used to display the coatings morphology and structure using a Zeiss Supra VP40 field emission gun scanning electron microscope. EDX was performed on the samples to determine the chemical composition of the coatings (Edax Trident) using an EDAX Sapphire Si (Li) detector, and quantified using a standardless ZAF algorithm. The chemical composition was calculated as an atomic percentage (at%), giving the percentage of the said atom relative to the total number of atoms in the scan. Electron backscattered imaging spectroscopy was undertaken on the samples to map the silver particle distribution on the substrate (n = 3).
Atomic Force Microscopy. Roughness parameters were obtained using an explorer AFM (Veeco, USA) operated in contact mode using a force constant of 0.12 Nm −1 and a silicon nitride tip. Scans with the Ti-ZrN coating and Ti-ZrN/Ag coatings were performed. Three replicate scans were performed on three separate coupons (n = 3). Hydrophobicity Measurements. The hydrophobicity was calculated by obtaining the surface contact angles of two polar liquids: HPLC grade water (Fisher Scientific; Loughborough, UK) and Formamide (Sigma Aldrich, UK) and one non-polar solvent: 1-bromonaphthelene (Sigma Aldrich, UK). The contact angles were obtained using the sessile drop technique (Kruss MobileDrop II, Kruss, Germany). This instrument deposited 5 μL volume drops from a standard dropper. The drop contact angle on the surface is then measured using a camera and prism inside the device. The image was analysed using Kruss SW23 (DSA2) (Kruss, Germany) analysis software, using the Young-Laplace bubble fit method to obtain the angle between the 'bubble' and the surface interface. Different surfaces were used for each solvent. Surface tension parameters for polar and apolar liquids were used to calculate physicochemical parameters (Table 1)

LIVE/DEAD ™ Staining.
To monitor cell viability, cells retained on surfaces were treated using a live/dead stain (LIVE/DEAD ™ Baclight ™ bacteria viability kit, Invitrogen, Scotland). After diluting the stains in dimethyl sulfoxide (Sigma, UK) according to the manufacturer's instructions, the propidium iodide and Syto 9 dye components were individually diluted in a 1:10 solution in sterile distilled water in eppendorf tubes. Five microliters of each dilution was then mixed together, and 10 μL was spread across the sample and allowed to air dry in the dark in a microbiological Class II safety hood. Following drying samples were stored in the dark at 4 °C. Potentially viable and damaged cells were distinguished under the fluorescence microscope (Nikon Eclipse E600, Surrey, UK) since it is assumed that viable cells appeared green, whilst non-viable or membrane compromised cells appeared red. The microscope was mounted with an F-View II black and white digital camera (Soft Imaging System Ltd., Helperby, UK, supplied by Olympus, Hertfordshire, UK). This system used a Cell F Image Analysis package (Olympus, Hertfordshire, UK). The percentage coverage of the live and dead cells retained on the surface was measured using separate selective UV filters across the same field of view. Twenty fields of view using each UV filter were taken per surface.
CTC DAPI Assays. 5-cyano-2,3-ditolyl tetrazolium chloride (CTC) is a redox dye which reduces into formazan and fluoresces red in the presence of UV radiation. Coupons were subjected to a standardised retention assay before staining 17 . CTC (Sigma Aldrich, UK) was prepared from powder and dissolved to a 5 mM concentration in sterile membrane filtered water (Millipore Elix, Germany), the CTC was also filter sterilised (0.2 μm Acrodisk, UK) before use. Direct staining of the coupons was performed by pipetting 500 μL of CTC onto the substrata/retained bacteria and incubated in a sealed container, without agitation, for 30 min at 37 °C. Following incubation, the surfaces were rinsed with sterile distilled water with a 3 mm nozzle at a 45° angle for five seconds and counter stained by pipetting 500 μL of 1 μg/mL 4′,6-diamidino-2-phenylindole (DAPI) (Sigma Aldrich, UK), prepared from the powder dissolved in sterile membrane filtered water. This was then pipetted onto the surface and left to incubate at room temperature in the dark, in a class II microbiological flow cabinet for 15 min. The coupons were rinsed again with sterile distilled water and then left to dry in the class II cabinet for 45 min until the coupon surface was completely dry. Stained samples were stored at 4 °C, in the dark and the substrata plus retained cells were visualised using epifluorescent microscopy (Nikon Eclipse E600 epifluorescence microscope, Tokyo, Japan), and Cell-F SCIEntIfIC REPORTS | (2018) 8:1497 | DOI:10.1038/s41598-018-20013-z image visualisation software (Olympus, UK) was used for image capture and analysis. The two stains were differentiated by using filters of 590/650 nm for red excitation and emission spectra (CTC/Formazan) and 350/470 nm for the blue (DAPI) excitation and emission spectra, respectively. Cytotoxicity Assays. Cell Culture. The commercial human monocyte cell line U937 (Health Protection Agency Culture Collections, Salisbury, UK) was cultured under aseptic conditions at 37 °C and 5% CO 2 in RPMI-1640 media (Lonza, Belgium) supplemented with 10% foetal bovine serum (Sigma-Aldrich, Dorset, UK) and 2% penicillin-streptomycin (Lonza, Slough, UK). The U937 cell suspension was maintained at 5 × 10 5 million cells/ mL by resuspension in fresh media every alternate day. Sterile filtered 0.4% trypan blue dye (Sigma-Aldrich, UK) was used to check cell viability using a 1:1 ratio of cell suspension to trypan blue and then counting the number of non-viable cells that take up blue dye using a TC10 automated cell counter (BioRad, Singapore). The viability of cells was above 90% for experimental purposes.
Cell Viability (MTS). S. steel, Ti-ZrN, Ti-ZrN/6.0 at% Ag, Ti-ZrN/15.6 at% Ag and Ti-ZrN/24.7 at% were sterilized with IMS before being placed in a 24-well plate. U937 monocytes were seeded in the 24-well plate (5 × 10 5 cell per well) containing the metals, and incubated at 37 °C and 5% CO 2 for 24 h. Cells (2.5 × 10 5 cell) were then spun down at 500 g for 5 minutes and re-suspended in 100 µL RPMI-1640 media. A 20 µL volume of Cell Titer 96 AQueous One Solution Reagent (Promega, UK) was added to the cells and incubated at 37 °C and 5% CO 2 for 1-4 h. Formazan absorbance was measured at 490 nm and results are expressed as percentage of the untreated control cells. U937 cells seeded in blank wells were used as a negative control. Cells treated with pure ethanol were used as a positive control.
Statistical analysis. The standard error of the mean was shown on the graphs using error bars. p values were calculated at the 95% confidence level using student t-tests. For the cytotoxicity assays, IBM SPSS Statistics (Version 22) was used to perform all statistics. The normality test and one way ANOVA with Tukey pairwise comparisons were conducted on all the data. Graphs were drawn using Microsoft Excel 2013. In all cases, P < 0.05 was considered statistically significant.
Data availability. The datasets generated during and/or analysed during the current study are available from the corresponding author upon reasonable request.