Film and microfibrillar substrate production

Random block co-polymer of poly(ethylene oxide terephthalate) (PEOT) and poly(butylene terephthalate) (PBT), with 300 Da PEG and PEOT/PBT ratio (w/w) of 55/45 (300PEOT55PBT45, acquired from PolyVation) was used to produce films and microfibrillar substrates. 300PEOT55/PBT45 granules were melted at 180 °C under slight pressure (~100 kg) in a circular 23 mm mold between two silicon wafers (Si-mat, Kaufering, Germany) to produce flat films. Films were punched out using a 22 mm punch to fit in a 12-well plate.

The electrospinning polymer solution was prepared by dissolving 20% (w/v) 300PEOT55PBT45 in 3:7 1,1,1,3,3,3,-Hexafluoro-2-propanol (HFIP):chloroform, overnight at room temperature under agitation. For heparin functionalization, 2% (w/v) poly(ethylene glycol) (PEG) with 2 NH2 end-groups (Mw: 3350 kDa) (PEG-NH2), was added to the polymer solution and mixed for 4 h before electrospinning.

ESP scaffolds (microfibrillar substrates) were produced on a slowly rotating (100 RPM) 19 cm diameter mandrel by electrospinning on a polyester mesh (FinishMat 6691 LL (40 gr/m2), generously provided by Lantor B.V.) with 12 mm holes, on top of aluminum foil. The following parameters were maintained: 15 cm working distance between needle and rotating mandrel, 1 ml/h flow rate, 23–25 °C and 40% relative humidity, a needle charge between 10 and 15 kV and collector charge between −2 and −5 kV. Individual ESP scaffolds were punched out with a diameter of 15 mm over the 12 mm holes in the polyester mesh and removed from the aluminum foil. This resulted in 15 mm ESP scaffolds with a 12-mm diameter surface for cell culture and a 1.5-mm polyester ring around it to improve handleability. Using this method, up to 100 microfibrillar substrates were produced under exactly equal parameters.

Before cell culture, microfibrillar substrates and films were sterilized in 70% ethanol for 15 min and dried at room temperature until visually dry. The 1.5-mm polyester ring was covered with a rubber 15 mm outer- and 12 mm inner-diameter O-ring (Eriks) to keep the scaffolds from floating in tissue culture well plates.

Functionalization of microfibrillar substrates with BSA or bFGF

Before coupling of bovine serum albumin (BSA)-FITC conjugate (ThermoFisher Scientific) or basic fibroblast growth factor (bFGF) (Neuromics), ethanol sterilized microfibrillar substrates were incubated in 0.5 M NaOH for 30 min at room temperature to open the ester bond of the 300PEOT55PBT45 polymer. Scaffolds were thoroughly washed five times with water and then incubated with 4 mg/ml N-(3-Dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDC) (Sigma-Aldrich) and 10 mg/ml N-hydroxysuccinimide (NHS) (Sigma-Aldrich) in milliQ water, or in milliQ water only, without EDC-NHS, as negative control, for 30 min at room temperature on a rocking plate. EDC-NHS solution was removed and 500 µl of 1 µg/ml BSA or 20, 200, or 2000 ng/ml bFGF (thus in total 10, 100, or 1000 ng bFGF) in water was added to the scaffolds in a 24-well-plate well and incubated overnight at 4 °C on a rocking plate.

For BSA-FITC functionalization, the following day scaffolds were washed five times with water and scaffold fluorescence was measured in the fluorescein channel on a Clariostar plate reader (BMG Labtech). For sodium dodecyl sulfate (SDS) wash, 1% (w/v) in water was added to the functionalized scaffolds and incubated under agitation at room temperature overnight. The following day, scaffolds were thoroughly washed five times with water and measured on the plate reader as described before.

For bFGF functionalized scaffolds, the bFGF solution with which the scaffolds were incubated was harvested, before any washes, to be analyzed by bFGF ELISA. The scaffolds were then washed five times with water, once with PBS and once with medium. For the bFGF scaffolds, all solutions were sterilized by filtration through at 0.2 µm filter. As we could not directly measure bFGF on the scaffolds like we measured the fluorescent BSA-FITC, we measured the bFGF that did not bind to the scaffolds. The amount of unbound bFGF was measured by quantifying the harvested bFGF solutions with which the scaffolds were incubated, using a bFGF ELISA Kit (Abcam), according to manufacturer’s protocol. In this way, we quantified the amount of bFGF that was left over in solution, and thus did not bind to the scaffolds. This measurement was then used to compare scaffolds with and without EDC/NHS and to estimate the amount of bFGF that bound to the scaffolds.

Heparin functionalization of microfibrillar substrates

In all, 1.5 mg/ml heparin sodium salt from porcine intestinal mucosa (Sigma-Aldrich) was mixed with 4 mg/ml EDC and 10 mg/ml NHS in water (or water only, without EDC-NHS, as negative control) and directly added to the 300PEOT55PBT45 + PEG-NH2 microfibrillar substrates and incubated overnight at 4 °C.

To measure bound heparin, scaffolds were washed five times with milliQ water and stained for 30 min with an alcian blue staining solution (0.1% alcian blue, 10% ethanol, 0.1% acetic acid, 0.03 M MgCl2 in water (all Sigma-Aldrich)). Scaffolds were washed once with MQ water and incubated for 30 min at room temperature in destaining solution (10% ethanol, 0.1% acetic acid, 0.03 M MgCl2 in water). Scaffolds were washed again once with water and then incubated for 30 min in 1% SDS to extract the heparin-bound alcian blue from the scaffolds. The absorbance of this solution was measured in a Clariostar plate reader.

For cell culture, the heparin-functionalized scaffolds were washed five times with milliQ water and incubated overnight at 4 °C with 500 µl 2000 ng/ml bFGF. The following day scaffolds were washed five times with water, once with PBS and once with medium. All solutions were sterilized by filtration through at 0.2 µm filter.

Cell culture

Human dermal fibroblasts (Lonza) were expanded at 2000 cells/cm2 in DMEM + Glutamax medium (Thermo Fisher Scientific) supplemented with 10% (V/V) fetal bovine serum (FBS) (Sigma-Aldrich).) Bone marrow derived hMSCs were isolated by Texas A&M Health Science Center62. Briefly, aspirated bone marrow was centrifuged to isolate mononuclear cells. The hMSCs were further expanded and tested for differentiation potential. hMSCs were received at passage 1 and were further expanded at 1000 cells/cm2 in α-MEM + Glutamax medium (Thermo Fisher Scientific) supplemented with 10% FBS. MG63 cells (ATCC) were expanded at 5000 cells/cm2 in DMEM + Glutamax+10% FBS medium. All cells were cultured in 37 °C in 5% CO2 until reaching 70–80% confluency. Cells were trypsinised in 0.05% Trypsin and 0.53 mM EDTA (ThermoFisher Scientific) and hMSCs and fibroblasts were used for experiments at passage 5. MG63 cells were used at passage 90.

Unless otherwise stated, all experiments were harvested at day 7. For scaffold experiments, hMSCs and fibroblasts were cultured at 1000 cells/cm2 in TCP and films, and 30.000 cells/microfibrillar substrate in growth medium with 100 U/ml penicillin-streptomycin. All other experiments were done in medium without penicillin-streptomycin. In bFGF medium conditions, 10 ng/ml bFGF was added to the medium.

For blebbistatin and MRTF/SRF inhibitor experiments, hMSCs and fibroblasts cells were seeded at 1000 cells/cm2 on TCP and cultured for 6 days. MG63 cells were seeded at 5000 cells/cm2 cultured for 2 days, because of a very high proliferation rate. After the initial culture period in growth medium, 100 µM blebbistatin (Sigma-Aldrich) in 0.2% DMSO in growth medium, or 12.2 µM MRTF/SRF inhibitor CCG203971 in 0.1% DMSO in growth medium, or respective DMSO control was added to the cells for 24 h.

To test the responsiveness of hMSCs to bFGF in the presence of MRTF/SRF inhibitor, hMSCs were seeded at 1000 cells/cm2 in TCP and cultured for 7 days in 0.1% DMSO control, 0.1% DMSO + 10 ng/ml bFGF, 24.4 µM MRTF/SRF in 0.1% DMSO or 24.4 µM MRTF/SRF in 0.1% DMSO + 10 ng/ml bFGF, all in hMSC growth medium.

DNA quantification

To lyse cells for DNA quantification, cells were washed 2x with PBS and freeze-thawed dry twice before RLT lysis buffer (Qiagen) was added. Microfibrillar substrates were removed from the polyester ring after the last PBS wash. Samples were then freeze-thawed 3x in lysis buffer. TCP plates and films were scraped with a cell scraper after the first freeze-thaw in lysis buffer. Microfibrillar substrates were left in lysis buffer. Samples were diluted 100–400x, depending on expected number of cells per samples, in Tris-EDTA buffer (10 mM Tris-HCl, 1 mM EDTA, pH 7.5) and λ-DNA standard was made in the same final solution (0.25-1% RLT in Tris-EDTA buffer). Pico green assay (ThermoFisher Scientific) was then used to quantify DNA, according to the manufacturer’s protocol.

Protein isolation and western blot

Protein was isolated in a custom lysis buffer to allow for the detection of membrane proteins with western blot. Other buffers, such as RIPA buffer, were tried for FGFR1 western blot without success. The buffer consisted of 150 mM NaCl, 0.5% sodium deoxycholate, 1% SDS, 1% NP-40, and 50 mM Tris-HCl in water, set to pH 7.4. The buffer was supplemented with cOmplete™ Mini EDTA-free Protease Inhibitor Cocktail (Sigma-Aldrich).

Samples were washed in cold PBS twice before lysis. Microfibrillar substrates were removed from the supporting polyester ring. To get sufficient proteins, 6–12 films or 15–20 microfibrillar substrates were combined in 300–400 µl for a single protein isolate. Experiments were repeated 3 or 4 times to obtain sufficient replicates. In total, 6 or 10 cm dishes were used for TCP samples. TCP and film conditions were scraped in lysis buffer with cell scrapers. Microfibrillar substrates were submerged in lysis buffer and incubated for ~30 min in lysis buffer because the scaffolds were removed from the protein isolate. Samples were not spun down to maintain potentially non-dissolved membrane proteins in solution.

Pierce BCA protein assay kit (Thermo Fisher Scientific) was used to quantify total protein concentration. 20 µg protein was incubated in 10% 2-Mercaptoethanol (Sigma-Aldrich) in laemmli loading buffer (Bio-Rad) for 37 °C for 20 min for FGFR1 western blots and at 95 °C for 5 min for all other western blots. Samples were loaded into 4–15% polyacrylamide gels (Bio-Rad) and blotted to 0.45 µm PVDF membranes (Bio-Rad) using semi-dry transfer. Membranes were blocked in 5% (w/v) fat free milk (Bio-Rad) in TBS + 0.05% (v/v) tween-20 (Sigma-Aldrich) for 1 h, except for SRF western blots, which had to be blocked in 2% (w/v) BSA (VWR) + 0.05% tween-20 in PBS to work. Primary antibodies were incubated in their respective blocking buffer overnight at 4 °C. All antibodies were ordered from Abcam: FGFR1: ab76464 1/500; Paxillin: ab32084 1/1000; Zyxin: ab58210 1/1000; YAP1: ab52771 1/1000; SRF: ab53147 1/250; and TBP: ab51841 1/1000. Blots were incubated the following day with 0.33 µg/ml Goat-anti-rabbit or -mouse HRP (Bio-Rad) in blocking buffer for 1 h at room temperature. To visualize the protein bands, blots were incubated with Clarity Western ECL (Bio-Rad) for 1–5 min right before imaging.

Immunofluorescence and imaging

Cells were fixed with 3.6% (v/v) paraformaldehyde (Sigma-Aldrich) in PBS for 20 min at room temperature. To block and permeabilize, fixed cells were incubated in 2% (w/v) BSA + 0.1% (v/v) triton X (VWR) in PBS. Zyxin or paxillin (Abcam, ab58210 and ab32084, respectively, both 1/1000) were incubated in 2% (w/v) BSA + 0.05% (v/v) tween-20 in PBS overnight at 4 °C. The following day, 1/1000 Goat-anti-mouse Alexa Fluor 568 or Goat-anti-rabbit Alexa Fluor 488 was incubated overnight at 4 °C in 2% (w/v) BSA + 0.05% (v/v) tween-20 in PBS. The next day, samples were stained with DAPI (Sigma-Aldrich, 0.14 µg/ml in PBS + 0.05% (v/v) tween-20) to stain nuclei. Images were taken on a confocal microscope.

Focal adhesions were quantified manually by counting the number of focal adhesions per cell using Fiji. Between 17 and 27 cells were counted per condition, from 5 to 10 different images from biological triplicates. Cell area was measured by manually outlining the cells and measuring surface area using Fiji. The number of focal adhesions was normalized to the cell area.

Lentiviral production and transduction

To produce lentiviral particles, human embryonic kidney 293FT (HEK) cells were seeded at 60,000 cells/cm2 in DMEM + Glutamax+10% FBS. Cells were transfected with pMDLg pRRE, pMD2.G, pRSV Rev (Addgene) and one of the pLKO.1 shRNA plasmids using 5:1 lipofectamine 2000 (Thermo Fisher Scientific):DNA (v/w) 24 h after seeding. The following TRC pLKO.1 constructs (Dharmacon) were used: ZYX: TRCN0000074204 and TRCN0000074205; PXN: TRCN0000123134 and TRCN0000123136; YAP1: TRCN0000107265 and TRCN0000107266; and non-targeting shRNA control (RHS6848). Medium was changed 16 h post-transfection to hMSC growth medium. Lentivirus was harvested and filtered through a 0.45-µm filter 24 h and 48 h after the change to hMSC growth medium.

In all, 24 h after thawing at 1000 cells/cm2, hMSCs were transduced with the lentiviral medium for 16 h. Medium was replaced with growth medium the following day. 48–72 h post transduction, medium was replaced with growth medium + 2 µg/ml puromycin for 72 h. A total of 9–10 days after thawing, hMSCs were passaged and seeded at 1000 cells/cm2 on TCP for 7 days in growth medium before protein harvest.

Statistics and reproducibility

The statistical tests and number of biological replicates and/or experiments are stated in the figure subtexts. Each experiment used at least three biological replicas. Cells selected for quantification of focal adhesions were selected randomly. Films and electrospun scaffolds were also randomly assigned to different experimental groups. Shapiro–Wilk test was used to test for normal distribution of each experimental group before further statistical analysis. To test for significance of absolute differences in experiments with multiple comparisons between groups, a One-way ANOVA with Tukey’s post hoc was performed. For relative differences between multiple experimental groups, log values were used for repeated measures ANOVA, with Tukey’s post-hoc test. For experiments with a single comparison, two-tailed Student’s t test was used for absolute differences, and ratio–paired t-test for relative differences. Significance was set at p < 0.05. Statistical analysis was done using Graphpad Prism 8.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.