This article is mentioned in:

Introduction

A total of 375,000 hip and knee endoprostheses implants are performed in Germany each year (1). In 1.5% of cases, implant-associated infections (IAI) occur following total joint arthroplasty (2). Despite the relatively low prevalence of IAI, it may have severe consequences for patients, physicians and the health care system (3). In the majority of cases, infections are initiated by Staphylococcus aureus, a biofilm-forming coagulase-negative Staphylococcus epidermidis (S. epidermidis), or gram-negative species (4). Early infections occur within weeks following surgery and are usually acquired intraoperatively. In addition, IAI may manifest as a late infection, occurring between 3 and 24 months following surgery, resulting from hematogenous spreading of the bacteria from other foci (5,6). IAI requires local and systemic antimicrobial therapy, which necessitates the removal of the contaminated endoprosthetic implant and insertion of a temporary cement implant (spacer) loaded with antibiotics (7–9).

There is currently a requirement for novel endoprosthetic materials with optimal surface and mechanical characteristics that prevent bacterial biofilm formation avoid bacterial resistance mechanisms and mediate biocompatibility and osseointegration (10). In recent years, magnesium has emerged as a promising implant material solution. Magnesium has various favourable properties, as it is a freely accessible metal and an essential cation serves an important role in the metabolism of human cells (11,12). The biocompatibility of magnesium as a coating is influenced by the corrosion rate, corrosion products and the varying pH-value of the solution (13). A previous in vitro study revealed that rapidly corroding magnesium in a thin layer on titanium (Ti6Al4V) was a suitable candidate for implant coatings with antimicrobial and biocompatible properties (14). Monocultures of S. epidermidis and human osteoblasts (hOB) revealed that magnesium coatings may prevent initial bacterial adhesion while remaining biocompatible according to the cytotoxicity test standards of the International Organisation for Standardisation (15). In contrast to monocultures, co-cultures provide a more comprehensive model of the development of the infection, as demonstrated in a previous study (16).

The aim of the present study was to investigate the antibacterial potential and cytotoxic effects of a thin, rapidly corroding magnesium coating on titanium samples in a co-culture model of S. epidermidis and hOBs.

Materials and methods

Titanium samples coated with pure magnesium

The present study used Ti6Al4V discs (DOT GmbH, Rostock, Germany) coated with pure magnesium for the experiments. Uncoated Ti6Al4V discs (a common implant material) served as a control. The discs were 11 mm in diameter and 2.5 mm in thickness. Magnesium-sputtering targets (diameter, 20 cm; 3.5 mm thickness; purity, 3N5) were produced by Fhr Anlagenbau GmbH (Ottendorf-Okrilla, Germany). The magnetron sputtering process was conducted with a Von Ardenne CS730 cluster machine (Von Ardenne GmbH, Dresden, Germany) with a 200 W sputtering power and 2.3×103 mbar pressure. The samples were coated with layers of magnesium (5-µm thick) by breaking the vacuum and turning the samples. The temperature of the substrate holder did not surpass 76°C, and all process parameters were performed in accordance with those described by Schlüter et al (17). The roughness (Rz) of the surfaces was measured via 3D laser scanning microscopy (VK-X260, Keyence Corporation, Osaka, Japan). The Ti6Al4V disc roughness was between 8 and 12.7 µm for the magnesium-coated discs as determined in our previous study (14). The samples were Υ-sterilized at 25 kGy following the sputtering process.

Magnesium ion release and pH of the supernatants

The magnesium-coated Ti6Al4V samples were pre-incubated for 24 h in Modified Eagle's osteogenic cell culture medium (MEM; EMD Millipore, Billerica, MA, USA) without calcium or antibiotics, containing 10% foetal calf serum (EMD Millipore), 1% HEPES buffer (Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) and the following osteogenic additives: 100 nM dexamethasone, 50 µg/ml L-ascorbic acid and 10 mM β-glycerophosphate, which were all obtained from Sigma-Aldrich; Merck Millipore (Darmstadt, Germany). Following 24 h pre-incubation and at 2, 4 and 7 days following infection, supernatants were transferred to 1.5-ml Eppendorf reaction tubes (Eppendorf, Hamburg, Germany). Digestion was achieved using 65% concentrated nitric acid (1.0% v/v; Sigma-Aldrich, Merck Millipore). The concentration of magnesium ions released from magnesium coatings was quantified using atomic absorption spectrometry ZEEnit 650P in combination with a graphite tube atomizer (all from Analytik Jena AG, Jena, Germany). The analysis process started with evaporation steps at 80, 90 and 110°C. This was followed by a pyrolysis phase at 900°C in the platform tube, heated to 2,000°C to evaporate and transform compact fragments into atoms. This step involved elemental analysis using a resonating cathode lamp and a magnesium cathode at a wavelength of 202.6 nm. Following 0, 2 and 7 days, the pH value of the supernatants was determined using an InoLab® pH 720 pH meter (WTW GmbH, Weilheim, Germany).

Isolation and cultivation of hOBs

The isolation of hOBs was conducted according to the protocol previously described by Jonitz et al (18). The osteoblasts were obtained from the cancellous femoral head of 6 patients that underwent total hip replacement surgery (3 male donors, mean age, 72±8.6 years; 3 female donors, mean age, 78±3.8 years) at the Department of Orthopaedics University Medicine Rostock (Rostock, Germany). The present study was approved by The Local Ethics Committee of Rostock (Rostock, Germany; registration number, A2010-10) and written informed consent was obtained from each patient.

Co-culture of hOBs and S. epidermidis

The procedures used to co-culture hOBs and S. epidermidis were the same as those described by Zaatreh et al (16). hOBs (25,000 cells/ml) at passage three were transferred to a 24-well-plate (Sarstedt, Nümbrecht, Germany) and cultured on the Ti6Al4V sample discs in MEM with adjunctions for 24 h. Similarly, hOBs and S. epidermidis were cultured on Ti6Al4V sample discs coated with magnesium for 24 h. A prepared overnight culture of S. epidermidis (RP62A; ref. no. 35984; American Type Culture Collection, Manassas, VA, USA) was subsequently used for mono-species infections with a multiplicity of infection of 0.04 (25,000 osteoblasts/ml, 1,000 colony forming units/ml). Co-cultures were incubated for a period of 7 days under standard aerobic conditions (37°C and 5% CO2). At 2 and 4 days following infection the medium was replenished.

Determination of hOB viability

hOBs that adhered to the discs were washed with 1X phosphate-buffered saline (PBS; EMD Millipore), followed by the addition of 200 µl trypsin/EDTA (1X; Gibco; Thermo Fisher Scientific, Inc.) for 3 min under aerobic conditions (37°C and 5% CO2). The hOBs were then mechanically relocated from the discs with a pipette tip (Eppendorf). The solution with the osteoblasts was transferred to a 1.5 ml Eppendorf reaction tube (Eppendorf) centrifuged at 169 × g for 4 min at 4°C, and washed with 1X PBS. Quantification of the number of viable primary osteoblasts on the test samples was determined by trypan blue staining (Sigma-Aldrich; Merck Millipore). Trypan blue enters the damaged membranes of dead cells, leaving viable cells unstained. The viable cells were counted using a Thoma cell counting chamber, obtained from Paul Marienfeld GmbH & Co. KG (Lauda-Königshofen, Germany) according to manufacturer's protocol (19) and using an Olympus CKX41SF optical light microscope, (Olympus Soft Imaging Solutions GmbH, Hamburg, Germany). Measurements were performed at 2 and 7 days of co-culture.

Quantification of biofilm-bound S. epidermidis

The test samples were transferred to glass test tubes (Greiner Bio-One International GmbH, Kremsmünster, Austria) containing 1 ml PBS (1X). S. epidermidis were removed by ultrasonic treatment with device settings at 100% for 4 min (BANDELIN BactoSonic, GmbH & Co. KG, Berlin, Germany) at days 2 and 7 of co-culture. The solution in the glass test tube was diluted in 1X PBS and plated onto tryptic soy broth (TSB)-agar plates (Thermo Fisher Scientific, Inc.). Following 24 h incubation at 37°C and 5% CO2, the number of colony forming units were quantified.

Quantification of planktonic S. epidermidis

The co-culture supernatants containing planktonic S. epidermidis were collected in 15 ml centrifuge tubes (Greiner Bio-One International AG, Kremsmünster, Austria) with 1 ml PBS (1X), and centrifuged at 3,345 × g for 10 min at 4°C on day 2 and 7 of co-culture. The number of planktonic bacteria were quantified by serial dilution in 1X PBS and then counting the number of colony forming units on TSB-agar plates (Thermo Fisher Scientific, Inc.) following 24 h incubation at 37°C and 5% CO2.

Scanning electron microscopy (SEM)

Test samples were fixed in a 2.5% glutaraldehyde solution at 4°C for 24 h. Specimens were subsequently washed with 0.1 M sodium acetate buffer, and dried using a graded ethanol series (5 min in 30% ethanol, 5 min in 50% ethanol, 10 min in 70% ethanol, 15 min in 90% ethanol and twice for 10 min in 100% ethanol). The samples were then dried by critical point drying with CO2 (Critical Point Dryer, Emitech Ltd., Ashford, UK), sputter-coated with gold, and examined with a scanning electron microscope (Zeiss DSM 960A; Zeiss GmbH, Jena, Germany).

Statistical analysis

Data are expressed as the mean ± standard error for three independent experiments. Statistical significance was evaluated with an unpaired two-tailed t-test using SPSS software (version, 20.0; IBM SPSS, Armonk, NY, USA). P<0.05 was considered to indicate a statistically significant difference.

Results

Magnesium ion release of the test samples and pH-values of the media

Ion release data were examined in order to characterize the strength of the corrosion process as it progresses over the 7 days. Fig. 1 shows the release of magnesium ions in the medium following one day of pre-incubation and at 2, 4 and 7 days following infection of the co-culture and the test sample alone. The magnesium-coated titanium alloy released 24 mM/day magnesium ions during pre-incubation and 11 mM/day over the first 2 days, decreasing to 3 mM/day during days 4–7. The magnesium-coated surface in the co-culture with osteoblasts and bacteria released 24 mM/day during pre-incubation and 8 mM/day over the first 2 days, which decreased to 2 mM/day during days 4 to 7. Media were replenished following 1 day of pre-incubation and after 2 and 4 days of co-culture. Magnesium ion release levels in the magnesium-coated samples maintained in medium without bacteria and cells were not significantly greater when compared with the hOB and S. epidermis co-culture samples. This indicates that hOB adhesion and S. epidermidis infection did not significantly alter the release of magnesium.

Figure 1. Magnesium ion release in the co-culture and test sample-only systems. Mg ion release was measured in the medium following 1 day of pre-incubation and at 2, 4 and 7 days for test sample in medium alone and test sample co-cultured with hOBs and S. epidermidis, as determined by atomic absorption spectrometry. The release declines from ~24 mM/day on day 1 of exposure to the medium, to 2–3 mM/day on day 7 in the co-culture and test sample-only systems. Data are expressed as the mean ± standard error (n=3). S. epidermidis, Staphylococcus epidermidis; hOBs, human osteoblasts; Mg, magnesium.

Bacterial growth is influenced by pH, therefore the pH-values of the media were monitored. Table I indicates the pH-values in the co-culture at 0, 2, 4 and 7 days. The magnesium-coated Ti6Al4V altered the pH-value to an alkaline environment whereas the uncoated Ti6Al4V control demonstrated a slightly acidic pH value.

Table I. pH-values for Ti6Al4V and Mg-coated Ti6A14V test samples in co-culture with hOB and S. epidermidis for 0, 2, 4 and 7 days.

Table I.

pH-values for Ti6Al4V and Mg-coated Ti6A14V test samples in co-culture with hOB and S. epidermidis for 0, 2, 4 and 7 days.

Co-culture of hOB and S. epidermidis	Day 0	Day 2	Day 4	Day 7
MEM+10% FCS	7.94±0.07	8.03±0.15	8.01±0.07	8.05±0.03
Ti6Al4V	7.88±0.06	7.83±0.08	7.81±0.03	7.64±0.04
Mg-coating	8.82±0.08	8.56±0.08	8.19±0.01	8.12±0.04

[i] Data are expressed as the mean ± standard deviation (n=3). MEM, Modified Eagle's osteogenic cell culture medium; FCS, foetal calf serum; hOBs, human osteoblasts; S. epidermidis, Staphylococcus epidermidis; Mg, magnesium.

Determination of hOB viability in co-culture with S. epidermidis

The number of viable hOBs was examined by trypan blue staining (Fig. 2). A total of 25,000 cells were seeded on the two samples. The number of osteoblasts on the magnesium-coated sample was 50,000 following 2 days (P=0.01) and 40,000 at 7 days (P=0.01). For the control, a value of 40,000 cells at day 2 and 20,000 at day 7 were observed. The number of viable osteoblasts was greater on the magnesium-coated sample when compared with the control by ~10,000 at 2 days and ~20,000 at 7 days.

Figure 2. Viability of human hOBs in the co-culture on the test samples for 2 and 7 days. The number of osteoblasts on the titanium alloy sample was decreased by ~20,000 hOB/ml at day 7 compared to day 2, whereas a reduction of 10,000 hOB/ml was observed on the Mg-coated samples. Data are expressed as the mean ± standard error of the mean (n=3). *P<0.05 vs. Ti6Al4V + hOB + S. epidermidis (same day). +P<0.05 vs. day 2 (same material). S. epidermidis, Staphylococcus epidermidis; hOBs, human osteoblasts; Mg, magnesium.

Evaluation of the viability of S. epidermidis in co-culture with hOBs

Measurements of bacterial colonization in the test samples revealed that a greater number of S. epidermidis colonies survived on the control TiAI6V4 surface when compared with the magnesium-coated Ti6Al4V discs (Fig. 3). The magnesium-coated sample demonstrated a significant antibacterial effect, as indicated by the significant decrease in biofilm population by three orders of magnitude following 2 days (P=0.02), and 4 orders of magnitude following 7 days (P=0.03; Fig. 3). The effect was approximately 1 order of magnitude lower for the planktonic population. Complete results of the measurements for S. epidermidis are presented in Fig. 3.

Figure 3. Quantification of viable biofilm-bound and planktonic S. epidermidis in co-culture with hOB cells on the test samples following 2 and 7 days. The number of viable biofilm-bound S. epidermidis on the Mg-coated specimen was significantly decreased when compared with the titanium control over a period of 7 days. Data are expressed as the mean ± standard error of the mean (n=3). *P<0.05 Mg vs. titanium alloy control, day 2, planktonic. +P<0.05 biofilm vs. planktonic, day 2, on Mg. #P<0.05 Mg vs. titanium alloy control, day 7, planktonic. §P<0.05 Mg vs. titanium alloy control, day 7, biofilm. S. epidermidis, Staphylococcus epidermidis; hOBs, human osteoblasts; Mg, magnesium; CFU, colony forming units.

SEM imaging

SEM was performed to qualitatively assess the layer structure and morphology of osteoblasts, as well as the growth patterns and distribution of bacteria on the test specimen. The SEM images shown in Fig. 4 confirm the results of the quantitative measurements of the biofilm-bound bacteria. S. epidermidis colonization was observed to be lower on the magnesium-coated surface when compared with the control on both days 2 and 7 (Fig. 4). The bacteria form biofilm-like structures on the human primary osteoblasts on the two surfaces.

Figure 4. Scanning electron microscopy images of S. epidermidis in co-culture with hOB on the titanium-only test sample and the Mg-coated test sample for 2 and 7 days. A decreased density of S. epidermidis colonization was observed on the Mg-coated surface when compared with the titanium-only test sample (magnification, ×1,000 or ×2,000). Control images are test samples in medium only. S. epidermidis, Staphylococcus epidermidis; hOBs, human osteoblasts; Mg, magnesium.

Discussion

The aim of the present study was to investigate the antimicrobial properties of a rapidly corroding magnesium coating on titanium against S. epidermidis, as well as its effects on hOBs, in a co-culture model. Measurements of hOB and S. epidermidis populations revealed that hOB growth was improved and S. epidermidis growth was inhibited; these alterations differed by several orders of magnitude when compared with the uncoated titanium control.

In previously conducted monoculture experiments, the number of hOBs that colonized the magnesium-coated surface increased to ~80% when compared with the uncoated titanium control, due to the slightly less favourable growth conditions on the corroding surface (14). However, in the present co-culture experiments, hOB growth on the magnesium-coated samples increased to ~125% following 2 days and 200% following 7 days. The antibacterial surface significantly reduced the negative influence of S. epidermidis on the hOB. This result was achieved despite the presence of the magnesium coating, which demonstrates a negative effect on the growth potential of the hOBs. The negative influence of S. epidermidis on hOB may be due to competition for nutrients and pathogenicity (20). The uninhibited negative effects was observed on the control sample, where hOB numbers were markedly reduced following 7 days when compared with 2 days (a decrease of 20,000 from 40,000).

The increased growth of the hOBs on magnesium-coated samples, and the observed antibacterial effect, may be explained by the following factors: Prevention of bacterial adherence due to the corrosive dissolution process, osmotic stress in the strongest phase of initial corrosion, the microstructure of the surface, an unfavourable increase in pH-value for the bacteria or the inhibitory effects of magnesium ions on the bacteria (21–28). These effects are most prominent in the local environment close to the surface. The effects of dissolution and osmotic pressure decline over 2 days, however the pH shift may also have a negative effect on S. epidermidis growth over the 7-day period. The impact on the planktonic population may be less significant due to reduced exposure to these factors (13,29).

The antibacterial effects observed in the monoculture experiments were reproduced in the co-culture setup, however, they were observed to be one order of magnitude less prominent (14). These observations are in agreement with a prior comparison of a monoculture to co-culture experiment conducted by Zaatreh et al (16). Putative reasons for the diminished effect may include the favourable environment that the hOBs present for S. epidermidis and the protection that they may provide from antibacterial effects of the surface. The planktonic population of S. epidermidis depends on the biofilm population, as this is where most of the bacterial reproduction occurs (30).

In previous studies, the authors demonstrated that monocultures may capture only a limited number of interactions among human cells, bacteria and the tested surface (14,16). Unpredicted behaviour may arise from the complex interaction among human cells, bacteria and implant surfaces. Therefore, the results of co-culture experiments are regarded as more robust (31,32). In the present study, the co-culture model captured the apparent protective effects of the hOB cells, whereby this layer of cells protects bacteria from the adverse effects of the surface. The antibacterial effects were diminished in the co-culture system, however a significant antibacterial effect persisted (three orders of magnitude following 7 days).

The investigation of different experimental parameters must be limited. Future work will aim to test different growth media, a greater number of bacterial strains and species, as well as different osteoblast cell lines. In addition, the length of time for infection to occur, and the ratio of initial hOB and bacterial numbers will be explored. The results of the present study provide significant observations with promising implications for implant material research.

In conclusion, novel implant materials are in high demand due to the severity of IAI, which may be life-threatening for patients with lengthy and demanding treatment procedures and may result in a marked decrease in quality of life. It has been previously demonstrated that titanium coated with a thin, rapidly corroding layer of magnesium may be a suitable candidate material, based on monoculture studies involving primary hOBs and S. epidermidis. The present study applied a more rigorous test system with hOBs and S. epidermidis in co-culture on the magnesium-coated titanium surface. The favourable qualities of this surface as an antibacterial and biocompatible material were verified. These results provide evidence for progression towards future animal model testing.

Acknowledgements

The authors would like to thank Professor Andreas Podbielski (Institute of Medical Microbiology, Virology and Hygiene, University Medicine Rostock) for his help in performing the tests. In addition, the authors would like to thank Dr Marcus Frank, Mr. Wolfgang Labs, Mr. Gerhard Fulda (Electron Microscopy Centre, University Medicine Rostock) Mr. Philipp Pisowocki (Department of Orthopaedics, University Medicine Rostock) and Mrs. Regina Lange (Institute for Electronic Appliances and Circuits, University Medicine Rostock) for their help with SEM. Furthermore, a special thanks to Professor Regine Willumeit-Römer (Institute of Materials Research, Division of Metallic Biomaterials, Helmholtz-Zentrum Geesthacht) for her support and expertise in the field of magnesium as a biomaterial. This study was financially supported by the Helmholtz Virtual Institute MetBioMat (grant no. VH-VI-523) titled ‘In vivo studies of biodegradable magnesium-based implant materials’.

References