The aim of this study was to establish a new FSS model that is easy to use as well as simple to assemble in order to create reproducible fluid shear forces on cells close to implant material surfaces. Todays’ commonly used commercial flow devices differ in geometry and function, which makes comparisons between experiments difficult [4, 10, 26, 27]. The benefits of this novel testing device are reproducible laminar flows under controlled conditions (regulated temperature as well as steady partial pressure of CO₂). Due to its reproducibility, the stimulation of osteoblast cells by shear forces becomes assessable.

In this FSS chamber, osteoblasts were cultured on the bottom of a rotating round glass panel that moves within a resting liquid. Computerised simulations determined a value of 200 rpm as the optimal system configuration in which a constant laminar flow occurs without pulsatile character. When creating laminar flows, induction of turbulences at boundary surfaces results in flow instability. To reduce this negative effect occurring in frequently used stationary devices, cells were cultured on a carrier plate, which is placed within the lower petri dish. In this context, the direct contact between the carrier plate and another interface was omitted. Laminar flows were chosen to achieve a good reproducibility. This required a flow profile that is characterised by parallel moving liquid layers [26] that are present in the area in between the upper and lower plate. To define the most favourable position of the cell-bearing surface, computerised simulations were performed. Herein, it could be demonstrated that rising shear forces along the plate surfaces’ (0–2 dyn/cm²) are too low for osteoblast test cell stimulation, which occurs at about 10 dyn/cm² [28, 29]. The bottom of the glass plate generated enough shear forces (10 dyn/cm² in the periphery) to meet the requirements of an osteoblast-stimulating laminar flow chamber. Further on, the simulations indicated that the flow profile in between the two plates was not influenced by peripheral turbulences alongside the peripheral regions. To verify a cellular realignment towards the shear direction, cells were microscopically examined prior and after exposure to shear forces for 24 h upon a spinning disc at a speed level of 200 rpm. Even if not sufficiently meaningful alone, observing changes in osteoblast cell morphology are still appropriate methods to verify the good usability of a flow chamber for the generation of reproducible FSS. Although not statistically significant, a tendency of cellular realignment towards the liquid flow direction was demonstrated. Similarly, several studies have revealed characteristic changes of osteoblast morphology triggered by fluid shear stress, which depends on exposure time and strength [22, 30]. Likewise to our findings, these changes are characterised by the formation of actin stress fibres, which in turn align towards the longitudinal cell axis and mainly appear near the nucleus [8, 31]. However, the manual approach of analysing the actin fibres’ orientation has to be stated as a drawback of the present study, since it does not meet the requirements of a valid measurement. Immunofluorescence microscopy is largely a qualitative, or semiquantitative, approach with a limited capability of precise fibre differentiation and/or quantification since standard binary thresholds failed to exhaustively segment all fibres because of their wide variations in intensity and background levels [32]. Hence, more objective measurements could be provided by the use of an automated software-assisted processing. In this context, the FibreScore Algorithm by Lichtenstein et al. presents a potential solution for quantification, since it allows the reliable segmentation of each actin fibre. The procedure itself is based on the acquisition of different pixel intensities, whereby it works through the correlation of pixel adjacent regions with synthetic fibre templates at different orientations and their assignment for the central pixel with the highest correlation coefficient among all orientations [32].