Three-dimensional images of rough and smooth implant surfaces were obtained by means of atomic force microscopy (AFM) using the tapping mode scan of an AFM VEECO machine (Plainview, USA); this method was also used to determine the surface topography. We scanned several randomly selected areas measuring either 3 μm × 3 μm or 30 μm × 30 μm for each of the test groups and sterilized all titanium specimens with UV light for 1 h before use.

We isolated a S. epidermidis strain culture (AC-Acession: AF270147) from the skin of one of the authors; the sample was identified and confirmed by 16S rDNA—nucleotide comparison (IDNS® version v3.1.63r14 © SmartGene 2005 Molecular Mycobacteriology). After isolation, S. epidermidis was proliferated in BHI—culture medium (Bacto™ Brain Heart Infusion, BD Becton, Dickinson and Company Sparks, MD, USA). Glycerine was added, and bacterial cultures were stored at −80 °C. Prior to testing, cultures were defrosted and incubated at 37 °C overnight. We cultivated S. sanguinis (strain 20068; DSMZ) in sterile trypticase soy broth (Tryptic Soy Broth; BD Diagnostics, Sparks, MD, USA) supplemented with yeast extract (Sigma-Aldrich, St. Louis, Mo, USA). For both types of bacteria, cells were harvested by centrifugation, washed twice in phosphate-buffered saline (PBS) (Sigma-Aldrich, St. Louis, Mo, USA), and resuspended in normal saline. After that, we adjusted the cells by densitometry (Genesys 10S; Thermo Spectronic, Rochester, NY, USA) at 600 nm to a MacFarland 0.4 standard optical density that equalled the bacterial concentration of approximately 5 × 10 ⁹ cfu (colony forming units)/ml.

We determined the quantity of bacterial adhesion with a fluorescence dye, i.e., the CytoX-Violet Cell Proliferation Kit (Epigentek Group Inc., New York, USA), and recorded fluorescence intensities with an automated multi-detection reader (Fluostar optima; BMG labtech, Offenburg, Germany) at wavelengths of 560 nm excitation and 590 nm emission. High relative fluorescence intensities indicate high numbers of viable adhering bacteria. For simulating the influence of a salivary pellicle, we incubated specimens in 48-well plates with 1 ml of artificial saliva for 2 h prior to adhesion testing [2]. We then removed the saliva, added 1 ml of bacterial suspension to each well, and incubated the well plates at 37 °C for 120 min on an orbital shaker. After biofilm formation, we extracted the bacterial solution by suction and washed the specimens once with PBS to remove non-adherent bacteria. All specimens were transferred to a new 48-well plate. For each well, we added 200 μl PBS and 20 μl CytoX-Violet (indicator solution) and incubated the well plates at 37 °C for 120 min in darkness; 190 μl of the indicator solution from each well was transferred to sterile black 96-well plates, and fluorescence intensities were recorded.