The human primary epidermal keratinocytes (HEKs) were purchased from ScienCell Research Laboratories. The HEKs were cultured with keratinocyte medium (ScienCell Research Laboratories, Carlsbad, CA, United States) at 37°C with 5% CO₂. The medium was changed every 3 days. We used HEKs from passages three to five.

HEKs were seeded on six-well flexible silicone rubber BioFlex plates (Flexcell International, Burlington, NC, United States) at a density of 5 × 10⁵ cells/well in 2 ml of medium. Cells were cultured for 24 h to reach 60–80% confluence before mechanical stretch was applied. Cyclic mechanical stretch was applied with 10% amplitude at 0.5 Hz for 24 h by using an FX-5000T Flexcell Tension Plus device (Flexcell International, Burlington, NC, United States) as previously reported (Zhou et al., 2015). HEKs cultured in the same plates but left non‐stretched served as controls. The diagram of cyclic mechanical stretch device is shown in Figure 1D.

Total proteins were extracted from cells by using Radio immune precipitation assay (RIPA) lysis buffer. Concentrations of protein were detected by the bicinchoninic acid (BCA) assay (Thermo Fisher Scientific). 10 μg of total protein were separated by 10% SDS-PAGE, followed by transfer to PVDF membranes (Millipore, United States). The membranes were blocked with 5% bovine serum albumin at room temperature for 1 h and then probed with primary antibodies against Piezo1 (1:1000; SAB), fibronectin, Vimentin (1:1000, all from Abcam, Cambridge, United Kingdom), MMP9, E-cadherin, N-cadherin (1:5000, all from Abcam), MMP2, α-SMA, GAPDH (1:1000, all from Cell Signaling Technology, Danvers, MA, United States). Next day, after washing with TBST 10 min for three times, bands were then incubated with secondary antibodies and visualized using an ECL detection system (Millipore, Bedford, MA, United States). ImageJ software (National Institutes of Health, Bethesda, MD, United States) was used for quantitative analysis of immunoreactive bands.

For Piezo1 silencing, HEKs were transfected in six-well plates with 100 nM Piezo1 siRNA by using Lipofectamine RNAiMAX reagent (Invitrogen, Carlsbad, CA, United States) according to the manufacturer’s protocol. The sequences were as follows: Piezo1-siRNA, 5′-AGA​AGA​AGA​UCG​UCA​AGU​ATT-3′ (sense) and 5′-UAC​UUG​ACG​AUC​UUC​UUC​UTT-3′ (antisense), negative control (NC) siRNA, 5′-GUG​AGC​GUC​UAU​AUA​CCA​UTT-3′ (sense) and 5′-AUG​GUA​UAU​AGA​CGC​UCA​CTT-3′ (antisense). The sequences used were self-selected.

The Piezo1 inhibitor GsMTx4 (Alomone Labs, Jerusalem, Israel) was purchased and dissolved in PBS solution. 5 µM GsMTx4 was used for all experiments according to the manufacturer’s protocol. The incubation time with GsMTx4 is 24 h, accompanying by stretch process.

To analyze calcium concentration, the calcium assay kit (Abcam, Cambridge, MA, United States) was purchased. Before calcium detection, cell numbers in every group exhibit no statistical difference. The cell lysates collected were used for analyzing cytosolic calcium level. Measurement was performed in a 96-well plate, at 575 nm by using an Infinite M200 Pro microplate reader (Tecan, Männedorf, Switzerland).

Migration assays were performed using Transwell chambers (Corning, Tewksbury, MA, United States) as described previously (Fang et al., 2019). HEKs were seeded in keratinocyte medium without keratinocyte growth supplement (KGS) in the upper chambers. The lower chambers were filled with keratinocyte medium. After 24 h, the migrated HEKs were fixed and stained for 20 min in a 0.1% crystal violet solution. Images of migrated HEKs on the lower filters within three random fields were captured with a microscope. Migrated HEKs numbers were calculated by the ImageJ software.

Cell samples were fixed in 4% paraformaldehyde for 20 min at room temperature. Cell samples were then washed, permeabilized and blocked. Antibodies used for immunofluorescence staining were anti-Piezo1 (1:100; SAB), anti-F-actin (1:200, Abcam, Cambridge, United Kingdom), anti-N-cadherin (1:200, Abcam), anti-vimentin (1:200, Abcam), anti-αSMA (1:200, Cell Signaling Technology) an Alexa Fluor 488-conjugated goat anti-mouse secondary antibody (1:200; Jackson ImmunoResearch) and an Alexa Fluor 594-conjugated goat anti-rabbit secondary antibody (1:200; Jackson ImmunoResearch). For phalloidin staining, cells were incubated with Alexa Fluor 647 Phalloidin (Cytoskeleton, Inc., Denver, CO, United States, 1:200) for 30 min at room temperature. Subsequently, samples were stained with DAPI (Solarbio, Beijing, China). Images were captured using a Nikon Eclipse E800 microscope (Nikon, Melville, NY, United States) and a Zeiss 710 laserscanning microscope (Zeiss, Thornwood, NY, United States).

Data are presented as the mean ± SD. Statistical differences among groups were assessed using a two-tailed Student’s t-test or one-way ANOVA. p < 0.05 was considered statistically significant.

First, we investigated whether HEKs express the Piezo1 protein. Immunofluorescence analysis illustrated the presence of the Piezo1 protein in HEKs (Figure 1A). Furthermore, because Piezo1 is a transmembrane cation channel that facilitates Ca²⁺ influx in response to mechanical force (Coste et al., 2010), we tested whether Piezo1 affected Ca²⁺ entry in HEKs by applying the Piezo1 inhibitor GsMTx4 (Bae et al., 2011) and siRNA-mediated Piezo1 knockdown. Western blot and immunofluorescence analyses confirmed the efficiency of siRNA transfection in HEKs (Figures 1B,C). Importantly, in HEKs, the increase in Ca²⁺ concentration induced by mechanical stretch was inhibited by GsMTx4 treatment or Piezo1 knockdown (Figures 1E,F). Collectively, these data indicate that HEKs express Piezo1 and sense mechanical stretch through Piezo1.

To analyze whether Piezo1 is involved in changes in EMT phenotypes, we first tested the effect of Piezo1 activity on cell morphology in HEKs. As expected, some stretched HEKs showed spindle shapes. Meanwhile, cells treated with GsMTx4 or Piezo1 knockdown exhibited polygonal shapes similar to static cells when subjected to mechanical stretch (Figures 2A–D). These findings demonstrate the significant role of Piezo1 in modulating the morphology of HEKs.

Increased migration has been heralded as a key event of EMT (Mittal, 2018). In our research, mechanical stretch-induced migration of HEKs was decreased by GsMTx4 application or Piezo1 knockdown (Figures 3A–D). These results indicate that mechanical stretch-induced HEKs migration might be regulated by Piezo1 activity.

It has been reported that a series of biomarkers are associated with the EMT process (Zeisberg and Neilson, 2009). In our study, we observed that the expression of E-cadherin, a marker of cell-cell adhesion, was downregulated after mechanical stretch, whereas the expression of N-cadherin was upregulated. Significantly, the changes in E-cadherin and N-cadherin expression in response to mechanical stretch were inhibited after blockade and knockdown of Piezo1 (Figures 4A,B). Matrix metalloproteinase 2 (MMP2) and MMP9 are hallmarks of EMT and promote cell migration (Nisticò et al., 2012). Our data showed that the mechanical stretch-induced upregulation of MMP2 and MMP9 was decreased by GsMTx4 and Piezo1 knockdown (Figures 4C,D). Increased expression of α‐smooth muscle actin (α‐SMA) and vimentin are also mesenchymal features that develop during the EMT process (Huang et al., 2012). In our study, mechanical stretch-induced α‐SMA and vimentin upregulation was alleviated in HEKs by the inhibition or knockdown of Piezo1 (Figures 4E,F). Another hallmark of EMT is the upregulation of extracellular matrix (ECM) proteins to reinforce ECM remodeling (Gonzalez and Medici, 2014). Piezo1 inhibition or knockdown inhibited the mechanical stretch-induced increase in fibronectin (Figures 4E,F). Additionally, immunofluorescence staining was conducted to further substantiate the changes in N-cadherin, vimentin and α‐SMA, and the result was consistent with that obtained by western blotting (Figures 4G–I). Therefore, we concluded from these results that Piezo1 was involved in mechanical stretch-induced changes in EMT biomarkers.