Supplemental Material & Methods
S 1 Pillar micropattern device fabrication
The pillar micropattern devices have been generated by photolithography with the help of newly fabricated moulds with holes of defined depth and diameter, which served to achieve a homogeneous thick layer of photo-reactive polymers. To this end, the epoxy-based SU-8 10 resin (Microresist Technologies, Germany) was spin-coated under clean room conditions onto a clean silicon wafer with a defined speed. The solvent was evaporated by a two step baking process before the wafer was exposed to UV light. The photo-initiated ring opening reaction was accomplished at 95 °C using a hot plate and developing was performed in a propylenglycol-monoethylether-acetate (PGMEA)-containing developer for 1 min. For spin-coating, 2 ml of photoresist was spun on a silicon wafer (+5.08 cm) in a two step process. To reach a 15 µm thickness, a SU-8 10 resist was spun on with 2000 rpm for 40 s and the wafer was softbaken on a hot plate at 65 °C for 2 min, followed by baking at 95 °C for 5 min. After cooling for 10 min to room temperature (RT), UV-exposure was performed on a mask aligner (MJB3, Karl Suess, Germany), equipped with a 400 W mercury lamp. Exposure time depended on resist thickness and varied from 2.5 to 4 s. To complete cross-linking in the exposed areas, the wafer chips were heated up to 65 °C for 1 min, followed by heating up to 95 °C for 2 min, and cooled to RT for 10 min. The unexposed parts were redissolved in PGMEA for up to 1 min while shaking and the structure was blown dry with a stream of nitrogen. The resulting moulds were covered with the silicone derivate polydimethylsiloxane (PDMS, DowCorning, Midland, Michigan, USA) that had previously been mixed vigorously with a short hydrosilane cross-linker in the ratio 10:1. To remove trapped air, the mix was degassed for 30 min under vacuum at 7 x 10–2 mbar. The pillar arrays were generated by pressing the moulds upside down onto a drop of PDMS on a 24 x 24 mm coverslip, followed by curing for 4 h at 65 °C, a process which leads to get a defined E-modulus of 0.6 MPa. After cooling to RT, the substrates were peeled with a razor blade and overlapping PDMS residues were removed under sterile conditions. Each pillar array presented a dimension of 8 x 8 mm with a cell growth area of 64 mm2. The pillars had a height of 15 µm, a diameter of 5 µm and were arranged in squares with pillar micropatterns of 5, 7, 9 and 11 µm. Since PDMS is a hydrophobic material and the PDMS pillar microstructure leads to water repellency, the pillar tops were coated with fibronectin (FN, 10 µg/ml) by carefully pipetting FN drops to the pillar surfaces (10 sec at RT). Immediately after withdrawing the drop, the surface was rinsed with PBS, kept moistened with PBS containing 0.05 % Tween-20 and directly used for cultivating cells on the pillar tops. To control successful biofunctionalization, indirect immunofluorescence stainings (iIF) were performed with a mouse anti-human fibronectin antibody (Abcam, Cambridge, UK) visualized by a secondary antibody labelled with Alexa Fluor 594 (goat anti-mouse IgG, Invitrogen, Darmstadt, Germany). Prior to cell seeding, the pillar devices were accurately checked and rejected if the pillars stuck together.
S 2 Cell isolation, propagation and characterization
Human mesenchymal stem cells (hMSC) were isolated using the Bone Marrow Procedure Pack system (Bone Marrow Aspiration Concentrate, Harvest Technologies Corp., Plymouth, MA, USA) according to manufacturer’s instructions, followed by a Ficoll gradient centrifugation step (Invitrogen, Darmstadt, Germany), and a plastic adherence purification step. The cells were cultivated in hMSC proliferation medium supplemented with 10 % fetal calf serum, 50 ng/ml amphotericin and 50 µg/ml gentamicin (all Provitro, Berlin, Germany) and kept at 37 °C, 5 % CO2 in a humidified atmosphere. The formation of colonies in culture was monitored and documented with a Leica D-Lux3 CCD camera (Leica Camera, Solms, Germany) connected to a Leica DMIL inverted microscope (Leica Microsystems, Wetzlar, Germany). In order to identify the nature of the hMSC that were consecutively seeded onto pillar micropatterns, the standard hMSC characterization criteria of the International Society for Cellular Therapy (ISCT) were employed: MSC must be (i) adherent to plastic in vitro, (ii) positive for CD105, CD73 and CD90, while lacking CD45 and CD34, and (iii) able to differentiate at least to osteoblasts, adipocytes and chondroblasts in vitro.
Multilineage differentiation assay
Medium supplements were used to provoke the differentiation of hMSC into adipocytes (0.5 mM 3-isobutyl-1-methylxanthine, 1 µg/ml insulin, 1 µM dexamethasone and 100 µM indomethacine), osteoblasts (50 µg/ml ascorbate-2-phosphate, 10 mM beta-glycerophosphate and 10-7 M dexamethasone) and chondrocytes [cell pellets cultivated with 1 ng/ml transforming growth factor-beta (TGF-beta; R&D systems, Minneapolis, MN, USA), 50 µg/ml ascorbate-2-phosphate, 1x insulin transferring selenium supplement-x (Invitrogen, Darmstadt, Germany) and 10-7 M dexamethasone] for 21 d each [1, 2]. For chondrogenic differentiation, 1x106 cells were pelleted at 150 g for 5 min and cultivated in chondrogenic differentiation medium. The adipocytes’ lipid vesicles were visualized with the oil red-o staining technique, the amount of extracellular matrix calcium deposits induced during the osteogenic differentiation was assessed by von Kossa staining, and the chondrogenic differentiation was evaluated by an alcian blue staining of the paraffin-embedded cell pellets, which served to visualize acid mucopolysaccharides and glycosaminoglycanes. For oil red-o staining, the cells were fixed with 4 % paraformaldehyde, washed with 60 % isopropanol, stained with 0.5 % oil red-o and hematoxylin solution, photographed and air-dried. For von Kossa staining, the cells were fixed with 10 % formaldehyde, incubated for 60 min with 5 % AgNO3, washed, developed with 1 % pyrogallol, washed and fixed with 5 % sodium thiosulfate. For Alcian blue staining, the specimens were re-hydrated in descending alcohol series, stained with 1 % alcian blue solution (3 % acetic acid), washed and counter-stained with nuclear fast red solution, dehydrated and mounted with Roti Histokitt (Carl Roth, Karlsruhe, Germany). Cells were photographed with a Leica D-Lux3 CCD camera (Leica Camera, Solms, Germany) connected to a Leica DMIL inverted microscope (Leica Microsystems, Wetzlar, Germany).
S 3 Cell culture on pillar micropatterns
Silicone-made ring fittings that were sealed to the pillar device-mounted glass cover slip served to encircle the individual pillar micropattern cell culture devices. Herein, hMSC were seeded at 3 x 104 cells per pillar micropattern and cultured in hMSC proliferation medium supplemented with 10 % fetal calf serum, 50 ng/ml amphotericin and 50 µg/ml gentamicin (all Provitro, Berlin, Germany) at 37 °C, 5 % CO2 in a humidified atmosphere. Both planar PDMS surfaces and pillar micropatterns were biofunctionalized with fibronectin as described in Supplemental Materials and Methods Section S1 prior to cell seeding.
S 4 Indirect immunofluorescence stainings
For indirect immunofluorescence stainings (iIF), cells were washed with phosphate-buffered saline (PBS, PAA, Pasching, Austria), fixed with 100 % ice-cold ethanol for 20 min at RT and kept at -40 °C until usage. Focal adhesion molecules were detected by rehydrating the cells in PBS-Tween 0.05 % containing 5 % bovine serum albumin and incubation with 1:50 mouse anti-human vinculin, 1:100 mouse anti-human integrin alpha V, 1:50 mouse anti-human histone H3, mouse anti-human beta-actin (all Abcam, Cambridge, UK), or mouse anti-human 1:50 paxillin (R&D Systems, Minneapolis, MN, USA), respectively, for 1 h at RT. After intense washing, cells were incubated with a goat anti-mouse IgG1 Alexa Fluor 488 antibody (Invitrogen, Darmstadt, Germany) for 40 min in the dark. After another wash step, actin fibres were stained with Texas-Red labelled phalloidin (1:40, 20 min, RT) in combination with vinculin, integrin, paxillin and histone H3 stains and washed. The cells were counterstained with DAPI (300 nM, 5 min, RT) in the dark, mounted with Fluoromount-G (Southern Biotech, Birmingham, AL, USA) and photographed with a Biozero BZ-8000 fluorescence microscop equipped with a CCD camera (Keyence Corp., Neu-Isenburg, Germany). Data were collected with the BZ Image Viewer software v2.5 and analyzed with the BZ Image Analyzer Software v2.5 (both Keyence Corp., Neu-Isenburg, Germany).
S 5 Quantitative real time PCR arrays
For assessing the RNA integrity and quantity with the Experion RNA StdSens Chip technology (Bio-Rad, München, Germany), heat-denatured RNA samples and RNA ladder were loaded onto chips primed with filtered gel stain, vortexed and measured immediately in the Experion electrophoresis station. Electropherograms were checked using the Experion software version 3.0 (Bio-Rad, München, Germany) for RNA integrity and quantity. A C1000 Thermal Cycler served to conduct the transcription of RNA to cDNA and a CFX96 Cycler was used to perform qPCR (both Bio-Rad, Munich, Germany). The qPCR array conditions were as follows: 95 °C 10 min followed by 40 cycles of 95 °C 15 sec, 55 °C 40 sec, 72 °C 30 sec. The products’ specificity of each amplicon was checked by examining the melting temperatures (heating at 0.05 °C/s to 95 °C). Negative reverse transcription and negative template controls were included in all PCR runs.
For data analysis, the Ct-values of the investigated genes were normalized to the Ct-values of a housekeeping gene panel including beta-2 microglobulin, ribosomal protein 13a, glyceraldehyde-3-phosphate dehydrogenase and actin beta that were all checked for Ct-value consistency (inclusion criterion: maximal Ct-value difference £ 0.5 irrespective of pillar micropattern). The fold change (2^(-Delta Delta Ct)) was calculated as the normalized gene expression (2^(-Delta Ct)) in the test sample divided the normalized gene expression (2^(-Delta Ct) in the control sample. The fold regulation was used to present the fold-change results in a biological meaningful way: fold-change values greater than one indicated a positive- or an up-regulation and the fold-regulation was described equal to the fold-change. Fold-change values less than one indicated a negative or down-regulation, and the fold-regulation was described as the negative inverse of the fold-change. The fold-regulation data were visualized as 1D-clustergrams with the magnitude of gene expression illustrated in different colours (green: minimal gen expression, black: medium gene expression, red: maximal gene expression).
