TY - JOUR AU1 - Riegert, Janine AU2 - Töpel, Alexander AU3 - Schieren, Jana AU4 - Coryn, Renee AU5 - Dibenedetto, Stella AU6 - Braunmiller, Dominik AU7 - Zajt, Kamil AU8 - Schalla, Carmen AU9 - Rütten, Stephan AU1 - Zenke, Martin AU1 - Pich, Andrij AU1 - Sechi, Antonio AB - Introduction Biomaterials are often use as guidance structures in a variety of applications. For instance, biomaterials can be used to deliver pharmaceutically active compounds or cells to specific locations and can contribute to the repair of damaged tissues. Furthermore, biomaterials can mimic the physical and chemical features of the extracellular matrix thus supporting wound healing [1–4]. At the cellular level, biomaterial chemistry and topography are often exploited to regulate numerous cellular processes including differentiation, cell adhesion and migration as well as dendritic cell function [5–11]. Microgels play a central role in several aspects of the biomaterials research. Microgels are colloids characterized by distinctive physical and chemical properties, which include a porous structure, swelling in aqueous media, surface activity, and a very flexible chemical functionality. Another fundamental feature of microgels is their high responsiveness to several external stimuli such as temperature, pH, light, redox potential, magnetic fields and enzymes [12–16]. All of these features make microgels crucial building blocks in the context of several applications such as coatings, drug and gene-delivery systems, catalysis, water purification, sensing devices and cosmetic applications [17–25]. In addition, microgels can be readily attached to solid substrates (physically or chemically) to form linear arrays or films [26]. In this context, we have developed a printing technology that allows controlled functionalization of solid substrates with microgels and the variation of several microgel properties, for instance topology, degree of swelling and chemical structure [27]. In the context of biological applications, microgels have been used for drug and nucleic acid delivery as well as tissue regeneration [28–33]. Moreover, temperature-responsive microgels have been specifically targeted to cancer cells to induce their necrosis or apoptosis [34,35]. Inhibition of tumor cell proliferation has also been achieved via the release of doxorubicin or paclitaxel from pH-sensitive microgels [36,37]. Finally, advanced tissue engineering applications have included the support of mouse fibroblasts cell adhesion and proliferation [38] and the regulation of the adhesion of different cell types using temperature-sensitive microgels [39]. Cell adhesion and migration are two fundamental biological processes required for biomaterial-supported tissue regeneration and engineering. Hence, the need for tailored interfaces and guidance systems that mimic the extracellular matrix thus supporting cell adhesion and migration. Several studies have proposed a number of strategies for controlling cell adhesion and migration by biomaterials. To mention a few, linear random cell migration has been promoted using microgrooves [40–42]. Moreover, complex geometries such as asymmetric teardrop islands have been shown to be able to convert random cell migration to directional cell migration [43–46]. Cell adhesion of various cell types has also been regulated by the use of elliptical rings with tunable height and shape or polymer brush nanoarrays [47–49]. In a previous study, we have demonstrated that microgels can be used to control cell adhesion and migration [50]. Substrate-anchored microgel arrays greatly influenced the distribution and orientation of focal adhesions and the actin cytoskeleton resulting in the alignment of these cytoskeletal structures in parallel with the microgel arrays. Remarkably, increasing the spacing of the microgel arrays from 1000 to 2000 nm augmented the motility of B16F1 cells by a factor of 2. In addition, temperature-responsive reduction of microgel stiffness has been shown to effectively regulate cell migration [50]. These findings demonstrate that microgels can be used not only for investigating important aspects of cell migration, but also for supporting and tailoring such behavior. To refine the understanding of the impact of surface-grafted microgel arrays on cell adhesion and migration, it is necessary to expand the range of topographic and mechanical features of microgel arrays and test their effect on different cell types. To this end, we generated a set of microgel arrays in which (i) the spacing between adjacent arrays was varied between 300 and 1600 nm, or (ii) their stiffness was varied by changing their degree of cross-linking (2.3 or 5 mol%). We studied the influence of these microgel arrays features on cell adhesion and migration using two model cell types: melanoma and Sertoli cells. Furthermore, we tested whether Gas2L1 (growth arrest specific 2 like 1), a target of thyroid hormone receptor that is associated with the actin and microtubule cytoskeletons and is also important for focal adhesion dynamics and cell migration [51–53], plays a role in the adhesion and migration of Sertoli cells on microgel arrays. The present findings clearly show that spacing and rigidity of surface-grafted microgel arrays can be manipulated to effectively modulate cell adhesion and motility of diverse cell types. Materials and methods Materials N-Isopropylacrylamide (NIPAm, Acros Organics 99%) was recrystallized from hexane before use. N,N’-Methylenebis(acrylamide) (BIS, Sigma-Aldrich, 99%) and 2,2′-Azobis(2-methylpropionamidine)-dihydro-chloride (AMPA, Sigma-Aldrich, 97%) and hexadecyl(trimethyl)ammonium bromide (CTAB, Sigma-Aldrich 98%) were used as received. Microgels synthesis Microgels were synthesized by precipitation polymerization.[54] NIPAm, BIS and CTAB were dissolved in ultra-pure water (150 mL) in a double wall reactor and heated to 70°C (see S1 Table in S1 Appendix for more details). Nitrogen was purged over the solution for 30 minutes. AMPA was dissolved in a small amount of water and added to initiate the polymerization, which lasted for 4 hours. Microgels were subsequently purified by dialysis against 5 liters of deionized water for seven days with repeated exchange of water (three times each day) using a membrane with a molecular weight cut-off (MWCO) of 12,000–14,000 Da (ZelluTrans, Roth). The concentration of the microgel solution was determined by gravimetric analysis. Microgels were stored in water, which was removed by lyophilization prior to use. Preparation of PDMS wrinkles PDMS stamps were prepared as described earlier [50,55,56]. PDMS was produced from the dual component Sylgard 184 elastomer kit (Dow Chemical) by mixing the monomer (33 g) with the base (3.3 g) for one minute and pouring the solution into a 10x10 cm plate to obtain a 3 mm thick film. PDMS solution was pre-cured and degassed over night at room temperature before final curing at 80°C for two hours. To produce the wrinkled stamps, a custom-made stretching device was used. A 1 x 2.5 cm block of PDMS was clamped into the device and stretched to 130% of its original size by increasing the distance between the clamps from 1.3 cm to 1.7 cm. Oxidation of the PDMS surface was performed in a low-pressure plasma oven (Plasma Activate Flecto 10 USB; Plasma Technology GmbH, Germany) with ambient air plasma at a pressure of 0.2 mbar and a power of 100 W. This process was performed for either 15 sec (300 nm), 120 sec (800 nm), 480 sec (1200 nm) or 900 sec (1600 nm), after which the tension was released and the wrinkled PDMS stamp was placed on a glass surface to maintain its stability. Printing of microgels on glass substrates Microgels were printed on glass coverslips as described earlier [27,50]. Briefly, glass coverslips were cleaned by sequential exposure to acetone, water and isopropanol in an ultrasonic bath (5 minutes each) followed by drying in a stream of nitrogen and activation in a plasma oven (Plasma Activate Flecto 10 USB; Plasma Technology GmbH, Germany) at 0.2 mbar for 300 sec. For the printing process, 15 μL of microgel solution was placed in the middle of a glass cover slip. For printing microgel arrays, wrinkled PDMS stamps were used (see section above), whereas for printing microgel films PDMS stamps, which were not stretched before oxidation, were used. The stamp was placed on the glass coverslip at a tilted angle and gently dropped onto the microgel droplet. Air bubbles and excess microgel solution were removed by gently pushing the stamp with tweezers. The stamp/coverslips combination was allowed to dry overnight (or at least for 12 hours). After removing the PDMS stamp, microgels were grafted to the surface by low pressure argon plasma in a plasma oven (Plasma Activate Flecto 10 USB; Plasma Technology GmbH, Germany). The oven was purged five times, by changing the pressure between 0.5 mbar and 0.1 mbar for cycles of 30 sec. Pressure was equilibrated for 60 sec at 0.2 mbar prior to cross-linking the microgel surface with argon plasma. Surface activation was performed for 23 sec at a pressure of 0.2 mbar and a power of 100 W. From this point, microgels could be used immediately or stored at room temperature. Characterization of microgels and microgel arrays The hydrodynamic diameter (DH,x°C) of microgels was determined by dynamic light scattering (DLS). For this purpose, 5 μL of microgel solution was diluted with 1.2 mL of ultra-pure water and measured with a Zetasizer ZS (Malvern Instruments GmbH) using a 633 nm laser and analyzing its back scatter at 173°. To investigate the thermoresponsive properties of microgels, the hydrodynamic diameter was measured at temperatures between 15°C to 50°C in 1°C steps. The volume phase transition temperature (VPTT) was determined as the inflection point in the plot of the hydrodynamic radius versus the temperature. The degree of microgel swelling (Rx°C, 50°C) was calculated by comparing the hydrodynamic diameter at x°C to the hydrodynamic diameter in the collapsed state (DH, 50°C) according to the Eq (1): 1 To investigate the structure of the microgel surface, atomic force microscopy (AFM) measurements were performed. To this end, 1x1 cm silica wafers were cleaned for 15 minutes in toluene in an ultrasonic bath and dried in a nitrogen stream. The dried wafers were further cleaned with a high-pressure carbon dioxide jet stream. The cleaned silica wafers were activated for 300 sec in a low-pressure plasma oven at 0.2 mbar (Plasma Activate Flecto 10 USB; Plasma Technology GmbH, Germany). A volume of 50 μL of a 1% diluted microgel solution was spin coated onto the activated wafer (WS-650-SZ-6NPP/Lite, Laurell) at an acceleration of 800 rpm/s and a speed of 2000 rpm for 1 minute. For AFM analyses, a NanoScope V (Digital Instruments Veeco Instruments Santa Barbara, CA) equipped with a J-Scanner was used. Uncoated NCH-50 (Nano World Point probe) cantilevers were used as probes with a resonance frequency of 320 kHz and a force constant of 42 N m-1. All measurements were performed in tapping mode and the images were analyzed with Gwyddion (version 2.53). The contact diameter of the microgels (DAFM) and the height of the microgels (hAFM) in dry state could then be determined from the AFM values. The deformation of microgels was calculated using the Eq (2): 2 The stability of microgel arrays in aqueous solutions was determined by placing them in ultra-pure water for up to 48 hours. Samples were then left to dry under ambient conditions and analyzed by AFM as described above. Cell culture B16F1 cells (ATCC CRL 6323) and B16F1 cells stably expressing RFP-zyxin [10,50] were grown in DMEM high glucose supplemented with 10% FCS, 2 mM L-glutamine, 1 mM sodium pyruvate, 100 μg mL-1 streptomycin and 100 U mL-1 penicillin at 37°C, 5% CO2. Control and Gas2L1 knock out Sertoli cells were grown in DMEM/F12 [1:1] supplemented with 10% FCS, 2 mM L-glutamine, 100 μg mL-1 streptomycin and 100 U mL-1 penicillin at 37°C, 5% CO2 [51]. Immunofluorescence and scanning electron microscopy Cells were fixed and permeabilised as described earlier [10,50,57]. Briefly, the actin cytoskeleton was labelled with Alexa 594-conjugated phalloidin (0.3 U mL-1, cat. no. A12381, Thermo Fischer). Nuclei were labelled with the DAPI (5 μg mL-1, cat. no. D1306, Thermo Fischer). Vinculin was labelled using an anti-vinculin antibody (1:400, cat. no. V9131, hVin1, Sigma-Aldrich) followed by Alexa 594-conjugated goat anti-mouse IgG (2 μg mL-1, cat. no. A11005, Thermo Fischer). Cover slips were mounted in Prolong Gold antifade agent (cat. no. P36934, Thermo Fischer). Images were acquired with a cooled, back-illuminated charge-coupled device camera (Cascade 512B; Princeton Instruments, USA) driven by IPLab Spectrum software (Scanalytics, USA) using a Plan-Apochromat 100x/1.30 numerical aperture oil immersion objective. Scanning electron microscopy was performed as described earlier [10,50,51,57]. Imaging and analysis of cell motility and focal adhesion dynamics To analyze cell motility, cells seeded on glass cover slips, microgel films or arrays were imaged for 24 h (at 37°C and 5% CO2) using an Axio Observer Z1 inverted microscope (Carl Zeiss, Germany) equipped with a Plan-Apochromat 10x objective and an AxioCam MRm (Carl Zeiss, Germany) driven by Zen 2 software (Carl Zeiss, Germany). Images were acquired every 5 min at multiple locations using a motorized X-Y stage. To determine the average speed and directionality of cell motility, manual tracking of the cells’ centroid was done using the ImageJ plugin MTrackJ [58]. Directionality of cell movement was calculated by analyzing all angular displacements measured between subsequent frames as described earlier [50]. Imaging of focal adhesion dynamics was performed by total internal reflection fluorescence (TIRF) microscopy using an Axio Observer Z1 inverted microscope equipped with a motorized TIRF slider (Carl Zeiss, Germany). Excitation of RFP-zyxin was carried out using a 561 nm laser (running at 10% of the nominal output power of 100 mW). The depth of the evanescent field was ≈70 nm. Images were acquired every 10–15 sec using an Evolve Delta EM-CCD camera driven by ZEN 2 software (Carl Zeiss, Germany). For all experiments, exposure time, depth of the evanescent field, and electronic gain of the EM-CCD camera were kept constant. The analysis of focal adhesion dynamics was achieved using a segmentation and tracking algorithm [59,60] to determine the following focal adhesion parameters: assembly and disassembly rates, area, life span and speed (i.e., speed of the apparent movement of FAs relative to the substrate). To determine the turnover of zyxin within focal adhesions, Sertoli and B16F1 cells stably expressing RFP-zyxin were used [51,57]. Briefly, focal adhesions were imaged by TIRF and fluorescence recovery after photobleaching (FRAP) microscopy for 15–20 min. One min after the beginning of image acquisition, a portion of a single focal adhesion (∅ 1 μm) was bleached for 1 second using a 405 nm laser at maximum power (100 mW) driven by a UGA-40 control unit (Rapp Opto Electronic GmbH, Germany). The same conditions (area bleached and the duration and intensity of the laser impulse) were applied for all experiments [51,61]. FRAP analysis was performed in two steps. Firstly, ImageJ (developed by Rasband, W.S., National Institute of Health, Bethesda, USA, http://imagej.nih.gov/ij/) was used to measure the average pixel intensity of three distinct regions of interest (ROI): ROI1: bleached area; ROI2: unbleached area within the cell; ROI3: background. Secondly, easyFRAP was used to normalize the FRAP recovery curves and calculate the mobile fractions as described [62]. Statistical analysis 150–200 samples were analyzed (i.e., motile cells or dynamic focal adhesions) from 2–3 independent experiments. For the motility studies, cellular speed and directionality were analyzed, whereas for focal adhesion dynamics studies, assembly and disassembly rates, speed, size and life span of focal adhesions were analyzed. For the analysis of zyxin turnover at focal adhesions, its mobile fraction was analyzed. Prism 8 (GraphPad Software Inc., CA) was used to generate all graphs and statistics. Pairwise statistical analyses were performed using the two-tailed Mann–Whitney nonparametric U-test and the null hypothesis (the two groups have the same median values, i.e., they are not different) was rejected when p > 0.05. Multiple comparisons were performed using the one-way ANOVA test in combination with the Tukey method with a statistically significant difference set at p < 0.05. In all box plots, the line in the middle of the box indicates the median, the top of the box indicates the 75th quartile, whereas the bottom of the box indicates the 25th quartile. Whiskers represent the 10th (lower) and 90th (upper) percentile, respectively. Materials N-Isopropylacrylamide (NIPAm, Acros Organics 99%) was recrystallized from hexane before use. N,N’-Methylenebis(acrylamide) (BIS, Sigma-Aldrich, 99%) and 2,2′-Azobis(2-methylpropionamidine)-dihydro-chloride (AMPA, Sigma-Aldrich, 97%) and hexadecyl(trimethyl)ammonium bromide (CTAB, Sigma-Aldrich 98%) were used as received. Microgels synthesis Microgels were synthesized by precipitation polymerization.[54] NIPAm, BIS and CTAB were dissolved in ultra-pure water (150 mL) in a double wall reactor and heated to 70°C (see S1 Table in S1 Appendix for more details). Nitrogen was purged over the solution for 30 minutes. AMPA was dissolved in a small amount of water and added to initiate the polymerization, which lasted for 4 hours. Microgels were subsequently purified by dialysis against 5 liters of deionized water for seven days with repeated exchange of water (three times each day) using a membrane with a molecular weight cut-off (MWCO) of 12,000–14,000 Da (ZelluTrans, Roth). The concentration of the microgel solution was determined by gravimetric analysis. Microgels were stored in water, which was removed by lyophilization prior to use. Preparation of PDMS wrinkles PDMS stamps were prepared as described earlier [50,55,56]. PDMS was produced from the dual component Sylgard 184 elastomer kit (Dow Chemical) by mixing the monomer (33 g) with the base (3.3 g) for one minute and pouring the solution into a 10x10 cm plate to obtain a 3 mm thick film. PDMS solution was pre-cured and degassed over night at room temperature before final curing at 80°C for two hours. To produce the wrinkled stamps, a custom-made stretching device was used. A 1 x 2.5 cm block of PDMS was clamped into the device and stretched to 130% of its original size by increasing the distance between the clamps from 1.3 cm to 1.7 cm. Oxidation of the PDMS surface was performed in a low-pressure plasma oven (Plasma Activate Flecto 10 USB; Plasma Technology GmbH, Germany) with ambient air plasma at a pressure of 0.2 mbar and a power of 100 W. This process was performed for either 15 sec (300 nm), 120 sec (800 nm), 480 sec (1200 nm) or 900 sec (1600 nm), after which the tension was released and the wrinkled PDMS stamp was placed on a glass surface to maintain its stability. Printing of microgels on glass substrates Microgels were printed on glass coverslips as described earlier [27,50]. Briefly, glass coverslips were cleaned by sequential exposure to acetone, water and isopropanol in an ultrasonic bath (5 minutes each) followed by drying in a stream of nitrogen and activation in a plasma oven (Plasma Activate Flecto 10 USB; Plasma Technology GmbH, Germany) at 0.2 mbar for 300 sec. For the printing process, 15 μL of microgel solution was placed in the middle of a glass cover slip. For printing microgel arrays, wrinkled PDMS stamps were used (see section above), whereas for printing microgel films PDMS stamps, which were not stretched before oxidation, were used. The stamp was placed on the glass coverslip at a tilted angle and gently dropped onto the microgel droplet. Air bubbles and excess microgel solution were removed by gently pushing the stamp with tweezers. The stamp/coverslips combination was allowed to dry overnight (or at least for 12 hours). After removing the PDMS stamp, microgels were grafted to the surface by low pressure argon plasma in a plasma oven (Plasma Activate Flecto 10 USB; Plasma Technology GmbH, Germany). The oven was purged five times, by changing the pressure between 0.5 mbar and 0.1 mbar for cycles of 30 sec. Pressure was equilibrated for 60 sec at 0.2 mbar prior to cross-linking the microgel surface with argon plasma. Surface activation was performed for 23 sec at a pressure of 0.2 mbar and a power of 100 W. From this point, microgels could be used immediately or stored at room temperature. Characterization of microgels and microgel arrays The hydrodynamic diameter (DH,x°C) of microgels was determined by dynamic light scattering (DLS). For this purpose, 5 μL of microgel solution was diluted with 1.2 mL of ultra-pure water and measured with a Zetasizer ZS (Malvern Instruments GmbH) using a 633 nm laser and analyzing its back scatter at 173°. To investigate the thermoresponsive properties of microgels, the hydrodynamic diameter was measured at temperatures between 15°C to 50°C in 1°C steps. The volume phase transition temperature (VPTT) was determined as the inflection point in the plot of the hydrodynamic radius versus the temperature. The degree of microgel swelling (Rx°C, 50°C) was calculated by comparing the hydrodynamic diameter at x°C to the hydrodynamic diameter in the collapsed state (DH, 50°C) according to the Eq (1): 1 To investigate the structure of the microgel surface, atomic force microscopy (AFM) measurements were performed. To this end, 1x1 cm silica wafers were cleaned for 15 minutes in toluene in an ultrasonic bath and dried in a nitrogen stream. The dried wafers were further cleaned with a high-pressure carbon dioxide jet stream. The cleaned silica wafers were activated for 300 sec in a low-pressure plasma oven at 0.2 mbar (Plasma Activate Flecto 10 USB; Plasma Technology GmbH, Germany). A volume of 50 μL of a 1% diluted microgel solution was spin coated onto the activated wafer (WS-650-SZ-6NPP/Lite, Laurell) at an acceleration of 800 rpm/s and a speed of 2000 rpm for 1 minute. For AFM analyses, a NanoScope V (Digital Instruments Veeco Instruments Santa Barbara, CA) equipped with a J-Scanner was used. Uncoated NCH-50 (Nano World Point probe) cantilevers were used as probes with a resonance frequency of 320 kHz and a force constant of 42 N m-1. All measurements were performed in tapping mode and the images were analyzed with Gwyddion (version 2.53). The contact diameter of the microgels (DAFM) and the height of the microgels (hAFM) in dry state could then be determined from the AFM values. The deformation of microgels was calculated using the Eq (2): 2 The stability of microgel arrays in aqueous solutions was determined by placing them in ultra-pure water for up to 48 hours. Samples were then left to dry under ambient conditions and analyzed by AFM as described above. Cell culture B16F1 cells (ATCC CRL 6323) and B16F1 cells stably expressing RFP-zyxin [10,50] were grown in DMEM high glucose supplemented with 10% FCS, 2 mM L-glutamine, 1 mM sodium pyruvate, 100 μg mL-1 streptomycin and 100 U mL-1 penicillin at 37°C, 5% CO2. Control and Gas2L1 knock out Sertoli cells were grown in DMEM/F12 [1:1] supplemented with 10% FCS, 2 mM L-glutamine, 100 μg mL-1 streptomycin and 100 U mL-1 penicillin at 37°C, 5% CO2 [51]. Immunofluorescence and scanning electron microscopy Cells were fixed and permeabilised as described earlier [10,50,57]. Briefly, the actin cytoskeleton was labelled with Alexa 594-conjugated phalloidin (0.3 U mL-1, cat. no. A12381, Thermo Fischer). Nuclei were labelled with the DAPI (5 μg mL-1, cat. no. D1306, Thermo Fischer). Vinculin was labelled using an anti-vinculin antibody (1:400, cat. no. V9131, hVin1, Sigma-Aldrich) followed by Alexa 594-conjugated goat anti-mouse IgG (2 μg mL-1, cat. no. A11005, Thermo Fischer). Cover slips were mounted in Prolong Gold antifade agent (cat. no. P36934, Thermo Fischer). Images were acquired with a cooled, back-illuminated charge-coupled device camera (Cascade 512B; Princeton Instruments, USA) driven by IPLab Spectrum software (Scanalytics, USA) using a Plan-Apochromat 100x/1.30 numerical aperture oil immersion objective. Scanning electron microscopy was performed as described earlier [10,50,51,57]. Imaging and analysis of cell motility and focal adhesion dynamics To analyze cell motility, cells seeded on glass cover slips, microgel films or arrays were imaged for 24 h (at 37°C and 5% CO2) using an Axio Observer Z1 inverted microscope (Carl Zeiss, Germany) equipped with a Plan-Apochromat 10x objective and an AxioCam MRm (Carl Zeiss, Germany) driven by Zen 2 software (Carl Zeiss, Germany). Images were acquired every 5 min at multiple locations using a motorized X-Y stage. To determine the average speed and directionality of cell motility, manual tracking of the cells’ centroid was done using the ImageJ plugin MTrackJ [58]. Directionality of cell movement was calculated by analyzing all angular displacements measured between subsequent frames as described earlier [50]. Imaging of focal adhesion dynamics was performed by total internal reflection fluorescence (TIRF) microscopy using an Axio Observer Z1 inverted microscope equipped with a motorized TIRF slider (Carl Zeiss, Germany). Excitation of RFP-zyxin was carried out using a 561 nm laser (running at 10% of the nominal output power of 100 mW). The depth of the evanescent field was ≈70 nm. Images were acquired every 10–15 sec using an Evolve Delta EM-CCD camera driven by ZEN 2 software (Carl Zeiss, Germany). For all experiments, exposure time, depth of the evanescent field, and electronic gain of the EM-CCD camera were kept constant. The analysis of focal adhesion dynamics was achieved using a segmentation and tracking algorithm [59,60] to determine the following focal adhesion parameters: assembly and disassembly rates, area, life span and speed (i.e., speed of the apparent movement of FAs relative to the substrate). To determine the turnover of zyxin within focal adhesions, Sertoli and B16F1 cells stably expressing RFP-zyxin were used [51,57]. Briefly, focal adhesions were imaged by TIRF and fluorescence recovery after photobleaching (FRAP) microscopy for 15–20 min. One min after the beginning of image acquisition, a portion of a single focal adhesion (∅ 1 μm) was bleached for 1 second using a 405 nm laser at maximum power (100 mW) driven by a UGA-40 control unit (Rapp Opto Electronic GmbH, Germany). The same conditions (area bleached and the duration and intensity of the laser impulse) were applied for all experiments [51,61]. FRAP analysis was performed in two steps. Firstly, ImageJ (developed by Rasband, W.S., National Institute of Health, Bethesda, USA, http://imagej.nih.gov/ij/) was used to measure the average pixel intensity of three distinct regions of interest (ROI): ROI1: bleached area; ROI2: unbleached area within the cell; ROI3: background. Secondly, easyFRAP was used to normalize the FRAP recovery curves and calculate the mobile fractions as described [62]. Statistical analysis 150–200 samples were analyzed (i.e., motile cells or dynamic focal adhesions) from 2–3 independent experiments. For the motility studies, cellular speed and directionality were analyzed, whereas for focal adhesion dynamics studies, assembly and disassembly rates, speed, size and life span of focal adhesions were analyzed. For the analysis of zyxin turnover at focal adhesions, its mobile fraction was analyzed. Prism 8 (GraphPad Software Inc., CA) was used to generate all graphs and statistics. Pairwise statistical analyses were performed using the two-tailed Mann–Whitney nonparametric U-test and the null hypothesis (the two groups have the same median values, i.e., they are not different) was rejected when p > 0.05. Multiple comparisons were performed using the one-way ANOVA test in combination with the Tukey method with a statistically significant difference set at p < 0.05. In all box plots, the line in the middle of the box indicates the median, the top of the box indicates the 75th quartile, whereas the bottom of the box indicates the 25th quartile. Whiskers represent the 10th (lower) and 90th (upper) percentile, respectively. Results Preparation and characterization of microgels We have previously demonstrated that cell migration can be effectively modulated by changing the spacing and the degree of microgel array swelling [50]. Hence, we decided to analyze these aspects in more detail by generating microgel arrays using a higher amount of cross-linker or by varying their spacing from 300 to 1600 nm. Since the generation of microgel arrays with smaller spacing requires microgels with a small hydrodynamic diameter, we initially concentrated our efforts on setting up a method that would readily allow the control of this parameter. To this end, we took advantage of surfactants, which are known to stabilize precursor microgel particles during the polymerization process, resulting in smaller microgel particles [63–65]. Specifically, we synthesized N-Isopropylacrylamide (NIPAm) microgels, cross-linked by N,N-Methylenebis(acrylamide) (BIS), in the presence of the surfactant hexadecyl(trimethyl)ammonium bromide (CTAB) at the concentration varying between 0wt% and 2.5wt% of the total mass of all products, keeping the concentrations of monomer, cross-linker and initiator constant (S1 Table in S1 Appendix). For simplicity, we will refer to the four microgel preparations as: MG small (generated in the presence of 2.5mol% CTAB), MG medium (0.5mol% CTAB), MG large (0mol% CTAB) and MG large-stiff (0mol% CTAB, 5mol% BIS). The hydrodynamic diameters of microgels were determined by dynamic light scattering (DLS) (Table 1). Download: PPT PowerPoint slide PNG larger image TIFF original image Table 1. Physical properties of microgels. https://doi.org/10.1371/journal.pone.0257495.t001 The hydrodynamic diameter of the microgels in the swollen state (20°C), at temperatures below the volume phase transition temperature (VPTT), could be decreased by increasing the amount of CTAB from 753 nm to 162 nm. At the typical temperature of a cell culture (37°C), clearly above the VPTT, the water was released from the polymer network and the hydrodynamic diameter decreased to values between 348 nm and 67 nm. In addition, the degree of swelling and chemical structure of the microgels was almost unaffected by CTAB (S1 Fig in S1 Appendix). It is important to note that the temperature responsive properties are essential for the synthesis of the microgels but were not used as a trigger for modulating cell behavior in this study. The Raman (S2A Fig in S1 Appendix) and FTIR (S2B Fig in S1 Appendix) spectra indicated that the chemical structure of microgels was also unaffected by the addition of CTAB during the synthesis and that the samples were free of surfactant after purification. Importantly, since CTAB is known to be cytotoxic [66], microgel preparations were extensively dialyzed to completely remove CTAB, as indicated by the Raman spectra (S2A Fig in S1 Appendix). Atomic force microscopy (AFM) images were taken to investigate microgel morphology. All microgels had a rounded shape (Fig 1A–1D) and the diameter in dry state (DAFM) decreased with increasing amount of surfactant that was used. Moreover, the contact area of the microgel with the surface and their height above the surface decreased with increasing amount of CTAB, whereas the deformation of the microgel was rather unaffected (S2 Table in S1 Appendix, S2 Fig in S1 Appendix). It must be noted that the size of the smallest microgels was in the range of the limits of the measurement method, thus causing a large variation in the measurements. Furthermore, smaller microgels have a more homogeneous structure of the polymer network, which lead to higher spreading [63]. Taken together, these observations show that the use of CTAB during the reaction does not grossly alter the final microgel properties and is an easy way to control the size of the microgel. Download: PPT PowerPoint slide PNG larger image TIFF original image Fig 1. Analysis of microgel physical and chemical properties. (A-D) Atomic force microscopy images showing microgels generated using 5mol% cross-linker (A) or 0% (B), 0.5% (C) or 2.5% (D) CTAB. The legend on the side of each atomic force microscopy image indicates the height of the arrays. Scale bar: 2 μm. https://doi.org/10.1371/journal.pone.0257495.g001 To increase microgel stiffness, cross-linker concentration was changed from 2.35mol% to 5mol%, without altering the concentration of any other component (S1 Table in S1 Appendix). The physico-chemical characterization showed that large stiff microgels were comparable in many aspects to the large microgels used previously by our group [50] including temperature responsiveness, hydrodynamic diameter in swollen state and chemical composition (Fig 1, S1 Fig in S1 Appendix, S1 Table in S1 Appendix). Moreover, the higher cross-linker concentration resulted in microgels having a greater height and smaller contact area in the dry state (S2 Fig in S1 Appendix), thus leading to lower deformability (S2 Table in S1 Appendix). Printing and characterization of surface-bound microgel arrays To print microgel arrays with different spacing, we adopted a previously published approach [27,50]. However, since the present goal was to fabricate arrays with smaller spacing, it was necessary to consider two fundamental parameters: (i) the size of the microgel and (ii) the wavelength of the PDMS stamp, defined by the thickness of its oxidized surface (i.e., the time of plasma activation) (S4 Fig in S1 Appendix). These two parameters have to complement each other, since any mismatch (e.g., large microgels used to generate smaller spacing), would results in “crippled” microgel arrays. As described earlier, all arrays were cross-linked by low pressure argon plasma to enhance their stability in aqueous media [50]. The argon plasma creates radicals in the polymer chains, which recombine forming a covalent bond [67,68]. Hiltl and colleagues [27] found that optimal alignment of microgels with different chemical composition could be achieved with a ratio of the wavelength of the PDMS stamp to the hydrodynamic diameter of the microgels of DHx1.2< λ