Aim of the study
This study aimed at developing a new non-cytotoxic coating on medical-grade titanium based on rhamnolipid biosurfactant and testing its efficacy in reducing the amount of microbial biofilm formation. In-vitro quantitative tests with two standard biofilm former staphylococcal strains were used to identify the best coating process and generate quantitative data on biofilm inhibition along time. Additional tests were realized on different commercial titanium surfaces in order to provide a proof of principle for the application of the proposed coating to a finished dental implant component.
To address this study aim, we structured the experimental design into three phases: 1) optimization of the BS-coating process, 2) investigation of the anti-biofilm efficacy of BS-coated titanium discs at different time-points, 3) testing the anti-biofilm efficacy of the BS-coating deposited on commercial surface morphologies used for dental implants.
Phase 1 was performed by running controlled experiments comparing the biofilm growth at 24 h on laboratory polished titanium surfaces coated using BS solutions at different concentrations ranging from 0 mg/mL (uncoated controls) to 4 mg/mL. Two representative biofilm former Staphylococcus spp. strains were considered in the study. Experiments were performed in quadruplicate using the crystal violet (CV) method to quantify the total biofilm biomass formed on coated discs compared to uncoated controls. In addition, a cytotoxicity test was performed on titanium discs coated with the highest BS concentration (4 mg/mL) to obtain preliminary data on their biocompatibility.
Phase 2 addressed the anti-biofilm efficacy of laboratory polished and BS-coated discs with the optimal process at longer time-points. The CV method and the 3-(4,5-dimethylthiazolyl-2-yl)-2,5-diphenyltetrazolium bromide reduction assay (MTT) were used to quantify both the total biofilm biomass and the metabolic activity of the biofilm formed after 24, 48 and 72 h of incubation. CV tests were repeated in three experimental sets, having four replicates for each set. MTT tests, were performed in quadruplicate on a single set. Data were analyzed in order to provide the percentage of biofilm inhibition at the different time-points for each of the two experimental strains.
Phase 3 implemented the optimal coating process on three commercial titanium surface morphologies, representative for a variety of dental implants available on the market. Anti-biofilm efficacy of coated commercial surfaces was assessed by quantifying the biofilm biomass by CV method at 24 h and calculating the percentage of biofilm inhibition with respect to equivalent uncoated controls. Tests were carried out in triplicate on a single experimental session.
Details on the BS production, biofilm former strains selection, biofilm formation, and tests for quantifying biofilm amount and coating anti-biofilm efficacy are reported below.
The biosurfactant R89 (R89BS) was obtained from the rhamnolipids-producing strain Pseudomonas aeruginosa 89 . R89BS was produced, extracted and chemically characterized as described by Ceresa et al. . Briefly, a loop of P. aeruginosa 89 overnight culture was inoculated into 40 mL of Nutrient Broth II (Sifin Diagnostics GmbH, Berlin, Germany) and incubated at 37 °C for 4 h at 140 rpm. Afterwards, 24 mL of the seed culture were inoculated in 1.2 L of Siegmund–Wagner medium and incubated at 37 °C for five days at 120 rpm. The bacterial cells were removed by centrifugation (Sorvall RC-5B Plus Superspeed Centrifuge, Fisher Scientific Italia, Milano, Italy) at 7000 rpm for 20 min and the supernatant was acidified with 6 M H2SO4 at pH 2.2, stored overnight at 4 °C and extracted three times with ethyl acetate (Merck KGaA, Darmstadt, Germany). The organic phase was anhydrified and evaporated to dryness under vacuum conditions and the composition of the raw extract confirmed by mass spectrometry analysis as reported previously .
Medical-Grade Titanium Discs and Surface Characteristics
Titanium alloy Ti6Al4V (medical-grade 5) discs (TDs) with four different surface finishing processes and morphologies were considered in this study. TDs 10 mm in diameter and 2 mm in thickness were obtained from computer numerical control machining and were subsequently polished in the laboratory with increasing fine-grained silicon-carbide abrasive paper up to 4000 grit to obtain a flat surface. Laboratory polished discs were used for optimizing anti-biofilm coating (Phase 1) and evaluating the microbial inhibition efficacy along time (Phase 2). To remove impurities and grinding residues, laboratory polished TDs were cleaned by sonication for 15 min each in three consecutive solutions, 100% acetone, 70% v/v ethanol in distilled water and 100% distilled water, as indicated in Ghensi et al. . The discs were then disinfected by immersion for a minimum of 24 h in 70% v/v ethanol in water and stored in these conditions until further use. TDs were dried under a laminar flow immediately before testing.
In addition to laboratory polished TDs, three commercial micro-morphologies were considered in this study to test efficacy of the R89BS-coating on representative commercial titanium surfaces used in real manufacturing processes for dental implants. For these tests, discs 10 mm in diameter and 1 mm in thickness were provided as sterile coupons from dental implants manufacturer companies with the following surface morphologies on a single face of the disc: computer numerical control machined and polished (M&P) flat surface for transmucosal implant components (CLC Scientific, Vicenza Italy), Laser-Lok® (L-L) micro-treaded medium-roughness surface for trans-mucosal implant components (BioHorizons, Birmingham AL, USA) and Resorbable-Blast Texturing (RBT) blasted high-roughness surface for bone contacting dental components (BioHorizons, Birmingham AL, USA). Discs with commercial surfaces were stored at environmental temperature within the original packaging until experimental use.
Surface Coating Process
Coating of TDs surface was performed in 24-well polystyrene plates, fitting one disc for each well. A solution of R89BS in sterile phosphate buffer saline (PBS) was freshly prepared before use according to the desired concentration of BS, ranging from 2 to 4 mg/mL. One milliliter of the desired R89BS solution was added to each well and R89BS-coating was obtained by physical adsorption of the biosurfactant at the titanium surface for 24 h at 37 °C. During coating process, the polystyrene plates containing the samples were agitated at 70 rpm on an orbital shaker. Uncoated control discs having the same surface morphology did not undergo any coating treatment.
Phase 2 and Phase 3 experiments were performed respectively on laboratory polished and commercial titanium discs coated using a 4 mg/mL R89BS solution in sterile PBS at 37 °C for 24 h at 70 rpm. Untreated discs having the same surface morphology were used as controls.
At the end of the immersion period, test discs were aseptically transferred to new 24-well polystyrene plates and dried under a laminar flow to set the BS-coating at the surface.
Cytotoxicity of R89BS-Coated Titanium
The potential cytotoxicity of R89BS-coated TDs was evaluated using a previously reported method . Briefly, a lactate dehydrogenase (LDH) assay (ISO 10993) (TOX7 In Vitro Toxicology Assay Kit, Sigma-Aldrich, Darmstadt, Germany) was performed using normal lung fibroblasts (MRC5), according to TOX7 operative procedures. The cell line was maintained in modified Eagle’s medium (MEM) supplemented with 10% fetal calf serum (FCS), l-glutamine (2 mM), sodium pyruvate (1 mM), 1% non-essential amino acids, and 1% antibiotics at 37 °C, 5% CO2, and 95% relative humidity.
Titanium discs coated with 4 mg/mL R89BS were immerged in fresh cell culture medium at 37 °C for 24 h in dynamic conditions, obtained by orbital shaking at 1 Hz, to favor biosurfactant removal from the surface. At the same time, cells were seeded in 96-well tissue culture plates and cultured in standard medium until about 70% confluence (24 h). The growth medium was then removed and replaced with the conditioned surface-contacting medium (200 µL/well).
The cytotoxic effect was measured on the basis of the amount of LDH released by cells after 48 h of exposure to the surface contacting medium. The positive control for cytotoxicity was constituted by fully lysate cells after exposure to 0.5% Triton X. Negative control was obtained from cells in reduced medium without surfactant. LDH level was evaluated by light absorbance at 490 nm (Tecan Spark 10 M). Assays were carried out in quintuplicate per each test condition.
Biofilm Growth on Titanium Discs
Two reference strains, Staphylococcus aureus ATCC 6538 and Staphylococcus epidermidis ATCC 35984, were used in this study, given their ability of producing high amount of slime according to the methods and criteria proposed by Christensen et al.  and subsequently detailed by Stepanovic and co-workers . Strains were stored at –80 °C in Tryptic Soy Broth (TSB) (Scharlab Italia, Milano, Italy) supplemented with 25% glycerol and grown on Tryptic Soy Agar (TSA) plates at 37 °C for 20 h before experimental assay.
S. aureus ATCC 6538 and S. epidermidis ATCC 35984 suspensions at the concentration of 1×107 Colony Forming Unit per mL (CFU/mL) were prepared in TSB supplemented with 1% w/v glucose (Scharlab Italia) to induce slime production .
Biofilm formation on the surface of the titanium discs (both coated and uncoated control discs) was obtained in 24-well polystyrene plates fitting one disc for each well. One milliliter of bacterial suspension was added to each well, thus guaranteeing submersion of the disc into the medium with bacteria. Plates were then immediately incubated at 37 °C for 24 h at 70 rpm in air.
In case the planned incubation period lasted 48 or 72 h, discs were aseptically transferred every 24 h into a new plate containing 1 mL of sterile TSB supplemented with 1% w/v glucose to provide fresh nutrients for the sessile bacterial cells. At the end of the incubation period, the suspension was removed using a micro-pipette, and the discs were gently washed twice with sterile PBS to remove non-adherent cells.
Quantitative Tests for Biofilm Formation
CV test was used in this study to measure biofilm biomass formed on coated or uncoated samples at the desired time-points. All study phases made use of CV test as a first-line quantitative assay. CV test was performed after drying the biofilm at the discs surface in a laminar flow cabinet. Each disc was dipped in 1 mL of 0.2% w/v crystal violet (CV) solution for 10 mins. An additional set of coated discs that did not underwent incubation for biofilm formation was used as blank in each testing session. After removing the CV solution, the discs were washed with distilled water to remove dye excess and air-dried again. The CV bound to the biofilms was then released from the matrix by adding 1 mL of 33% v/v acetic acid (Scharlab Italia) in water.
In Phase 2 a second quantitative test, the MTT reduction assay, was implemented to obtain complementary information about the sessile bacteria organized at the titanium surface. This test was realized by immersing each disc and the biofilm in the hydrated state in 1 mL of a 0.075% w/v MTT solution (Fisher Scientific Italia, Milano, Italy). Five microliters of glucose solution (20% w/v in distilled water) and 10 µL of 1 mM menadione solution (Sigma-Aldrich, Milan, Italy) were then added, and samples were incubated for 30 min at 37 °C. Coated discs without biofilm were also included in each MTT session as blanks. Finally, the formazan crystals were dissolved in 1 mL of a lysing solution composed by 7 parts of dimethyl sulfoxide (Scharlab Italia) and 1 part of 0.1 M glycine buffer (pH 10.2) (Sigma-Aldrich).
CV and MTT resulting solutions were spectrophotometrically read at 570 nm (Victor3VTM, Perkin Elmer, Milano, Italy).
Scanning Electron Microscopy of Titanium Surfaces and Bacterial Biofilms
A qualitative micro-morphological analysis of the three commercial titanium surfaces investigated in Phase 3, and of the Staphylococcus spp. biofilm formed at 24 h with and without the R89BS-coating was carried out by scanning electron microscopy (SEM), as described in Ceresa et al.  with minor modifications. Original untreated commercial surfaces were directly mounted on aluminum stubs using double-sided carbon-conducting tape and imaged without further preparation. Discs with biofilm were dipped in 1 mL of 2.5% w/v glutaraldehyde solution in 0.1 M phosphate buffer for 24 h at 4 °C to preserve the microstructural architecture of the biofilm on the titanium surface. Then, each disc was washed twice with Milli-Q® water, dehydrated by immersion in 70%, 90% and 100% v/v ethanol/water solutions for 10 min each and finally dried overnight under a laminar flow cabinet. Dried samples were then coated by a 10-nm layer of gold using a sputter coater (Emitech K500X, Quorum Technologies, Laughton, UK) to improve their electrical conductivity and thermal stability.
SEM observation was performed using a XL30 (FEI-Philips, Eindhoven, The Netherlands) scanning electron microscope in the high-vacuum mode. A set of four images for each disc were obtained by collecting the secondary electron signal at a magnification of 500×, 1000×, 2000×, and 4000× in order to detect both the titanium surface morphology and the fine structural detail of the microbial cells and of the extracellular matrix on the biofilm. The primary beam energy was set to 10 kV for the titanium surfaces and was lowered to 5 keV when biofilm was present to minimize damage to the organic structures. Possible artefacts due to the sample preparation process  were considered according to indications provided by Hrubanova et al.  and previous experience performed in imaging microbial biofilm formed in-vitro on medical devices [37–39] and in-vivo on titanium abutments [40, 41].
Data Analysis and Statistics
The single titanium disk was considered as statistical unit. Considering that biological data from CV assays or MTT assays are usually not normally distributed, quantitative data obtained from replicated CV and MTT were expressed as medians and interquartile ranges.
Mann-Whitney U test followed by Bonferroni post-hoc test were used to evaluate the effect of the different concentrations of R89BS on Staphylococcus spp. biofilm formation and to study the significance of data in the LDH cytotoxicity assay in comparison to positive and negative controls.
The percentage of cytotoxicity was calculated as follows:
where AR89BS is the absorbance value of samples treated with R89BS and Apos.Ctrl is the absorbance value of positive control (0.5% Triton X).
The effect of R89BS-coated TDs on Staphylococcus spp. biofilm formation was investigated with Mann-Whitney U test, comparing coated discs with uncoated controls at each time-point.
Results were considered to be statistically significant when p < 0.05. Statistical analysis was elaborated by means of the statistical program R, 3.5.3 (R Development Core Team, http://www.R-project.org).
Further, CV and MTT data were normalized with respect to the value of the corresponding blank and the inhibition percentages of biofilm formation was determined using the following formula:
where AR89BS is the absorbance value of BS-coated samples and ACtrl is the absorbance value of the untreated control. Inhibition percentages for biofilm biomass and metabolic activity of sessile microbial cells were provided in tabular form for ease of comparison between different time-points and different commercial surfaces.