Piezo1 Activation by Shear Stress Induces Corneal Epithelial Homeostasis Disruption

Abstract

Purpose

The shear stress induced by abnormal blinking compromises tear film stability. This study aimed to investigate the effects of shear stress on corneal epithelial homeostasis and the role of the mechanosensor Piezo1 in this process.

Methods

In vivo, abnormal ocular surface shear stress was induced in C57BL/6J mice using a controlled blinking model. Corneal damage was assessed via fluorescein staining and tear secretion. In vitro, human SV40 immortalized corneal epithelial cells were exposed to low-level shear stress (L-SS, 2.6 dyn/cm²) or high-level shear stress (H-SS, 7 dyn/cm²), calibrated by computational fluid dynamics simulation. Piezo1 was modulated using Yoda1 and lentivirus-mediated knockdown. Proliferation, migration, pyroptosis, and intercellular junctions were evaluated via EdU, scratch assays, transmission electron microscopy, and immunofluorescence, respectively.

Results

In vivo, increased blinking frequency caused corneal damage, reduced tear secretion, and Piezo1 upregulation. In vitro, both L-SS and H-SS activated Piezo1, modulating proliferation and migration via Ca²⁺ influx. L-SS upregulated zonula occludens 1 (ZO-1), promoting tight junction formation. In contrast, H-SS induced Piezo1 overactivation, which downregulated ZO-1, compromising barrier function, and upregulated pyroptosis markers (NLRP3, caspase-1, GSDMD, and IL-1β). Notably, these shear stress–induced biological effects were significantly reversed after Piezo1 knockdown.

Conclusions

This study established Piezo1 as a critical mechanosensor that mediates intensity-dependent responses caused by ocular surface shear stress in the corneal epithelial homeostasis. Specifically, Piezo1-mediated mechanotransduction under high-intensity shear stress triggers barrier failure and pyroptosis, highlighting Piezo1 as a promising therapeutic target for managing shear stress-related ocular surface disorders.

Dry eye disease (DED), a prevalent multifactorial chronic ocular disorder, is characterized by the disruption of corneal epithelial homeostasis. The corneal epithelium, serving as the primary physical barrier of the ocular surface, is constantly exposed to diverse mechanical stimuli, including blink-induced shear stress, alterations in extracellular matrix stiffness, and external physical forces such as eye rubbing and contact lens wear.^1,2 These stimuli significantly influence corneal epithelial homeostasis, especially shear stress induced by blinking. Physiological blinking distributes the tear film evenly across the entire eye surface, creating shear stress with uniform direction.³ However, altered blink patterns—such as increased frequency and incomplete blinks—disrupt tear film stability, accelerate tear evaporation, and elevate the risk of DED.^4–7 Accelerated tear film breakup, caused by tear film instability, is a key driver of dry eye.^8,9 Recently, the prevalence of dry eye syndrome associated with a short tear film breakup time has risen markedly^10,11; blink training can improve both tear film stability and dry eye symptoms.¹²

Piezo1, as a mechanosensitive cation channel, can detect diverse extracellular mechanical forces, including shear stress, osmotic pressure, and tensile forces, and transduces them into intracellular signals.¹³ It regulates a wide array of physiological functions, including cardiovascular, respiratory, urinary, and immune processes.¹⁴ Upon mechanical stimulation, activated Piezo1 channels convert mechanical signals into intracellular inflammatory cascades,^15–17 thereby triggering inflammation in various tissues.

Recent studies have highlighted the pivotal role of Piezo channels in ocular mechanobiology. Although both Piezo1 and Piezo2 are expressed across various ocular tissues,¹⁸ they govern distinct physiological domains. Piezo2 is primarily localized to corneal sensory neurons as a mechanotransducer involved in nociception and chronic pain signaling,^19,20 with its expression potentially altered in dry eye states.²¹ However, whereas Piezo2 is restricted to peripheral sensory processing, Piezo1 exhibits a more dominant distribution within structural and homeostatic ocular tissues, particularly within the trabecular meshwork, ciliary body, and ganglion cell layer.²² Functionally, Piezo1 regulates aqueous humor outflow and IOP homeostasis within the iridocorneal angle,²³ maintains lens fluid dynamics by modulating intracellular hydrostatic pressure,²⁴ and negatively regulates force-induced neurite elongation in retinal ganglion cells via calcium influx.²⁵ Spanning the anterior segment to the posterior pole, these diverse roles underscore the ubiquitous importance of Piezo1 as a key mechanotransducer throughout the ocular tissues.

Specifically regarding the ocular surface, mechanical stress has been shown to stimulate Piezo1 in conjunctival epithelial cells, leading to upregulated IL-6 expression and enhanced neutrophil infiltration.²⁶ Notably, Piezo1 has been implicated in the development and progression of DED. In a hyperosmolar stress-induced DED model, Piezo1 promoted chronic inflammatory responses in the corneal epithelium via activation of the NLRP3 inflammasome.²⁷ Moreover, the expression of Piezo1 is significantly higher in the corneal epithelial layer than in other corneal layers,¹⁸ suggesting its potential role in sensing and responding to mechanical stimuli at the ocular surface.

Although static stresses such as osmotic pressure have been shown to induce inflammation through Piezo1, it remains unknown whether the dynamic ocular surface shear stress generated during blinking is involved in the pathogenesis of DED. We propose that Piezo1 may function as a mechanosensor for aberrant shear stress, converting these physical signals into pathological responses, such as inflammasome activation and barrier disruption. Therefore, this study aimed to investigate the effects of ocular surface shear stress on corneal epithelial homeostasis from a biomechanical perspective, elucidate its role in the disruption of corneal epithelial homeostasis, and determine whether Piezo1 acts as a key mediator in this mechanotransduction pathway.

Methods

Animal Models

To simulate abnormal shear stress–induced dry eye, we developed a controlled blink mouse model. During the procedure, mice were anesthetized by intraperitoneal injection of Delivector Avertin (Dowobio, Shanghai, China) at a dose of 16 µL/g. Using a 5-0 surgical suture to apply traction to the eyelid, which was connected to a horizontal reciprocating device, we controlled the blink frequency by adjusting the device's frequency. Mice in the control group were similarly anesthetized and sutured, but their eyelids were allowed to remain in a natural closed state. Three experimental groups were thus established: a control group (0 blinks/min), a low-frequency blink group (10 blinks/min), and a high-frequency blink group (40 blinks/min). Each group underwent a 20-minute daily intervention for 3 consecutive days. Observations were conducted on days 0, 7, and 14 post intervention. To prevent ocular surface desiccation, saline (0.9% NaCl) was administered dropwise during the controlled blink procedure.

Eight-week-old female C57BL/6J mice (n = 60) were housed in the animal breeding unit of Xi'an Jiaotong University School of Medicine. All experimental procedures were reviewed and approved by the Institutional Animal Care and Ethics Committee of Xi'an Jiaotong University (No. XJTUAE2025-3763).

Tear Production and Corneal Staining Assessment

Corneal epithelial damage was assessed by fluorescein staining using 2.5% sodium fluorescein and observation under cobalt blue light. The total area of fluorescein staining of the cornea was scored in a blinded manner as follows²⁸: 0 (none), 1 (<5%), 2 (5%–25%), 3 (25%–50%), and 4 (>50%). Tear production was measured by placing a standardized phenol red-impregnated cotton thread (Tianjin Jingming New Technological Development, Tianjin. China) at the lateral canthus for 60 seconds, with the moistened length recorded in millimeters.

Cell Culture, Shear Stress Application, and Lentiviral Transfection

The human SV40 immortalized corneal epithelial cell (HCEC) line, which was kindly provided by Prof. Wei Chen (School of Ophthalmology and Optometry and Eye Hospital, Wenzhou Medical University, Wenzhou, China), was cultured in DMEM-F12 (Gibco; Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% fetal bovine serum in an incubator at 37°C and 5% CO₂. HCECs were seeded in six-well plates at a density of 5 × 10⁴ cells/cm². The plates were then placed on an orbital shaker (SLK-O3000-S; SCILOGEX, Rocky Hill, CT, USA) and maintained under standard culture conditions at 37°C. With an orbital radius of 10 mm, two rotation speeds—50 rpm and 200 rpm—were applied to generate low-level shear stress (L-SS) and high-level shear stress (H-SS), respectively. Consistent with the fluid dynamic characteristics observed in actual culture processes, the orbital motion induces rotation of the culture medium about the central axis, subjecting the liquid to centrifugal force and generating specific flow patterns. Shear stress exposure began 24 hours after seeding and lasted for 8 hours. Control cells from the same passage were cultured under static conditions.

Calcium-free medium was obtained by adding 0.2g of EGTA (Macklin, Shanghai, China) powder to 50 mL of complete medium and dissolving it thoroughly. Yoda1 (GC18233, GLPBIO, Montclair, CA, USA), a Piezo1 ion channel agonist, was dissolved in DMSO, and the final working concentration of Yoda1 was 0.2 µg/mL.

For the generation of stable Piezo1-knockdown cells, HCECs were seeded into 24-well plates and transduced with the Piezo1 knockdown lentiviral vector LV-Piezo1 (Genechem Co., Ltd, Shanghai, China) at an multiplicity of infection of 50, using HiTrans A (Genechem Co., Ltd) as the transduction reagent. Forty-eight hours post transduction, cells were selected with 1 µg/mL puromycin to establish stable cell lines.

Computational Fluid Dynamics (CFD) Simulation and Model Validation

To characterize the shear stress parameters generated by the orbital shaker and validate the in vitro model, a transient-state CFD simulation was performed using ANSYS Fluent (Ansys, Canonsburg, PA, USA). The culture system was modeled as a gas–liquid two-phase flow within a cylindrical dish (37-mm diameter, 2.08-mm liquid height) at rotational speeds of 50 rpm and 200 rpm (Fig. 1A). Given that the gas and liquid domains share identical stress and velocity boundary conditions at the interface, the system was simplified using the volume of fluid model to track the gas–liquid interface, with the flow assumed to be incompressible. This approach allows the calculation of the dynamic flow behavior and the resulting wall shear stress on the cell layer. Detailed mathematical derivations used in the simulation are provided in the Supplementary Material.

Figure 1.

Comparison of shear stress and velocity distribution in culture dishes. (A) Single-well culture dish model. (B, C) Shear stress distribution at 50 rpm (B) and 200 rpm (C). At 50 rpm, the shear stress intensity reached a maximum of 0.26 Pa (2.6 dyn/cm²). At 200 rpm, the shear stress intensity increased significantly, with the maximum shear stress reaching 0.7 Pa (7 dyn/cm²), but the overall distribution characteristics showed no significant change. (D, E) Velocity distribution at 50 rpm (D) and 200 rpm (E). At 200 rpm, the liquid surface velocity increased significantly, but the overall distribution characteristics showed no significant change.

CFD simulations confirmed that, at 50 rpm, the shear stress remained relatively low (≤ 2.6 dyn/cm²) and exhibited radial inhomogeneity, with higher intensity concentrated in an annular region near the sidewall (Fig. 1B). At 200 rpm, the overall shear stress increased significantly, reaching a maximum of 7 dyn/cm², with the high-stress zone expanding across the entire bottom surface and peaking at the center (Fig. 1C). The liquid–surface velocity fields at 50 rpm (Fig. 1D) and 200 rpm (Fig. 1E) both displayed a decay pattern from the sidewall and center toward the intermediate radial positions. These results validate that increasing the rotational speed substantially elevates the magnitude of shear stress and flow velocity without altering their fundamental spatial distribution within the culture dish.

Cell Viability Assay (CCK-8)

HCECs were treated with varying Yoda1 concentrations. CCK-8 reagent (BS350B, Biosharp, Shanghai, China) was diluted according to the manufacturer's instructions. Absorbance was measured at 450 nm using a microplate reader (Bio-Tek Instruments, Winooski, VT, USA).

Immunofluorescence

For corneal immunofluorescence staining, mouse eyeballs were collected after euthanasia, fixed, paraffin embedded, and sectioned. Following dewaxing and antigen retrieval in a 70°C water bath, the sections were processed for staining. For cell immunofluorescence, HCECs were grown in six-well plates and fixed with 4% paraformaldehyde for 30 minutes. All samples were then blocked with 5% bovine serum albumin for 30 minutes at room temperature, incubated with primary antibodies overnight at 4°C, and then with Cy3-conjugated secondary antibodies (1:200) for 1 hour in the dark. Before mounting, nuclei were counterstained with DAPI for 10 minutes. Antibody details are provided in Table 1.

Table 1.

Antibody Specifications and Dilution Ratios for Immunofluorescence

Antibody	Catalogue No.	Dilutions	Company
Piezo1	15939-1-AP	1:200	Proteintech
ZO-1	21773-1-AP	1:200	Proteintech
Phalloidin	CD4677	1:100	Coolaber

To visualize the cytoskeleton, samples were incubated with Alexa Fluor 488–conjugated phalloidin (1:100) for 1 hour at room temperature. After three washes with PBS, nuclei were counterstained with DAPI for 10 minutes.

Intracellular Free Ca²⁺ Assay and ROS Assay

For Ca²⁺ measurement, HCECs were seeded in six-well plates. After intervention, cells were washed thrice with HBSS (BL559A, Biosharp) and incubated in HBSS containing Rhod2-AM (MedChemExpress, Shanghai, China) for 30 minutes at 37°C.

ROS generation was detected using the CM-H2DCFDA probe (Molecular Probes, Invitrogen, Waltham, MA, USA) following the kit instructions. After incubation with 10 µM CM-H2DCFDA dye for 20 minutes at 37°C, cells were washed twice with PBS and the fluorescence was quantified using a fluorescence microscope (Axio Observer 7, Zeiss, Jena, Germany).

Cell Proliferation and Migration Assays

Proliferation was assessed using the BeyoClick EdU-488 Kit (C0071, Beyotime Biotech, Shanghai, China). Cells were incubated with 2 × EdU working solution for 2 hours, followed by Click reaction and Hoechst 33342 nuclear staining.

Migration was evaluated via scratch wound assay. A uniform wound was created in the monolayer with a sterile 200-µL pipette tip. Cells were pretreated with shear stress before wounding to investigate its sustained impact on epithelial homeostasis. Previous studies confirm that shear pre-exposure is essential for modulating HCEC migration, whereas postscratch intervention shows no significant effect.²⁹ Wound healing was monitored using an inverted microscope at 0 (immediately after wounding), 12, and 24 hours after wounding. The remaining wound area at each time point was quantified using ImageJ software. All data were normalized to their respective 0-hour areas.

Transmission Electron Microscopy and Live Cell Imaging

After the shear stress intervention, HCECs were harvested by scraping at a 45° angle. Cell precipitates were fixed in a mixture of 2.5% glutaraldehyde and 4% paraformaldehyde for at least 2 hours, followed by post-fixation with 1% OsO₄ for 1 hour. After dehydration through a graded acetone series (50%, 70%, 90%, and 100%), the samples were embedded in Epon 812 resin and polymerized at 60°C for 48 hours. Ultrathin sections (70–90 nm) were stained with uranyl acetate and lead citrate for 15 minutes each and observed under a transmission electron microscope (H-7650, Hitachi, Tokyo, Japan). For live cell imaging, HCECs cultured in 3.5-cm dishes were placed on a live cell imaging system (DMI6000B, Leica, Wetzlar, Germany) after shear stress application, and images were acquired every 15 minutes over a continuous 24-hour observation period.

RT-qPCR and Western Blot Analysis

Total RNA was isolated using the RNAfast200 kit (220011, Fastagen Biotech, Shanghai, China) and reverse transcribed with a Reverse Transcription Kit (11142ES10, Yeasen, Shanghai, China). Quantitative PCR was performed on a Bio-Rad (Hercules, CA, USA) instrument using primers listed in Table 2. For Western blotting, cells were lysed in RIPA buffer (PL001-2A, Zhhcbio, Airport New City, China) with protease and phosphatase inhibitors. Proteins were separated by SDS-PAGE, transferred to PVDF membranes, blocked with rapid-blocking solution (PS108P, Epizyme, Shanghai, China), and incubated with primary antibodies (Table 3) overnight at 4°C. After incubation with horseradish peroxidase–conjugated secondary antibodies, bands were visualized using a Bio-Rad ChemiDoc XRS+ system and analyzed with ImageJ (NIH, Bethesda, MD, USA).

Table 2.

Primer Sequences Used for PCR

Gene	Forward Primer	Reverse Primer
Piezo1	GCTCAAGGCCTTCTGGTG	GTGTAGACGGCGATGAGGAC
ZO-1	GAAATACCTGACGGTGCTGC	GAGGATGGCGTTACCCACAG
NLRP3	GCTGGCATCTGGGGAAACC	TTCAGGGTCCTTAGGCTTCG
IL-1β	CCTGAGCTCGCCAGTGAAAT	AAGCCCTTGCTGTAGTGGTG
GSDMD	TAAACTGGGCGGAGGGATGA	CCAGAAGTGTTGAGGGCAGAA
Caspase-1	AGAAAAGCCATGGCCGACAA	CTCTTCCTTGTTCAGCACCCT
Actin	CTTCGCGGGCGACGAT	CCACATAGGAATCCTTCTGACC

Table 3.

Antibody Specifications and Dilution Ratios for WB Analysis

Antibody	Catalogue No.	Dilutions	Company
Piezo1	15939-1-AP	1:800	Proteintech
ZO-1	21773-1-AP	1:1000	Proteintech
NLRP3	R23319	1:1000	ZENBIO
IL-1β	16806-1-AP	1:4000	Proteintech
Tubulin	T2200	1:3000	Sigma
Gapdh	GB15004-100	1:4000	Servicebio

Statistical Analysis

Statistical analyses were performed using GraphPad Prism software (version 10.1.2; GraphPad Software, La Jolla, CA, USA). Data normality was confirmed using the Shapiro–Wilk test, and all datasets passed the normality test (P > 0.05). For comparisons among three or more groups, one-way ANOVA followed by Tukey's post hoc test was used for most experimental data. Specifically, two-way ANOVA followed by Tukey's post hoc test was applied to analyze the wound healing rates over time in the scratch assay. Statistical significance was represented as: *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Results

Controlled Blink Induces Piezo1 Upregulation and Disrupts Corneal Epithelial Homeostasis in Mice

To investigate the impact of shear stress on corneal epithelial homeostasis, we established a controlled blink mouse model by threading surgical sutures through the upper eyelid of the mice and connecting them to a frequency-controlled lever device (Figs. 2A, 2B). Mice were divided into three experimental groups: control (0 blinks/min), low-frequency blink (10 blinks/min), and high-frequency blink (40 blinks/min). Corneal status was evaluated on days 1, 7, and 14 post intervention. Each group contained at least six mice. Corneal fluorescein staining indicated that corneal epithelial damage in mice deteriorated with increasing shear stress (Figs. 2C, 2D). On day 7 post intervention, tear secretion decreased in both blinking groups, with a more pronounced reduction in the high-frequency blink group (Fig. 2E). Immunofluorescence analysis revealed that Piezo1 expression in the mice corneal epithelium increased significantly by day 7 in the high-frequency blink group, whereas upregulation in the low-frequency group was delayed, becoming significant only by day 14 (Figs. 2F, 2G). These findings suggest that abnormal shear stress induces Piezo1 activation, which may contribute to ocular surface damage.

Figure 2.

Establishment of an in vivo controlled blink model in mice. (A, B) Schematic diagram (A) and experimental setup (B) of the murine dry eye model induced by controlled blink. The eyelid blink traction wire was connected to a lever device, the mice were forced to blink through mechanical reciprocating motion, and the ocular surface was continuously perfused with saline via a microinjection pump. n = 6 animals/group. (C, D) Fluorescein staining revealed progressive corneal epithelial erosion after the controlled blink intervention. (E) Tear secretion measured by phenol red cotton thread was significantly reduced after abnormal blinking intervention. (F, G) Representative immunofluorescence images illustrate the expression patterns of Piezo1 in the corneal epithelium of mice under low-frequency versus high-frequency blink conditions, with corresponding quantification of mean Piezo1 fluorescence intensity (n ≥ 3; scale bar, 100 µm). *P < 0.05, **P < 0.01, ***P < 0.001. ns, not significant.

Shear Stress Promotes the Proliferation and Migration of HCECs

To model corneal epithelial wound healing under blinking‑like shear stress, a uniform scratch was created in HCEC monolayers. After wounding, the monolayer was exposed to 8 hours of shear stress and then returned to standard culture conditions. Shear stress–treated HCECs exhibited enhanced wound healing ability (Figs. 3A, 3B). Furthermore, compared with static conditions, both L-SS and H-SS induced a marked increase in HCEC proliferation (Figs. 3C, 3D).

Figure 3.

Shear stress promotes the proliferation and migration of HCECs. (A, B) Immediately after scratching, HCECs were subjected to static, laminar, or turbulent shear stress for 8 hours. The wound healing area was measured (with the 0-hour wound area as the standard), and the percentage of unhealed area was assessed. (C, D) EdU labeling was used to detect the effect of shear stress on the proliferation of HCECs. Red, EdU; blue, Hoechst. The percentages of EdU-positive cells are shown. The data were collected from at least four independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001. ns, not significant.

Shear Stress Induced Cytoskeletal Reorganization and Intercellular Junctions Change of HCECs

Using phalloidin staining, we investigated the effects of different shear stress intensities on the F-actin cytoskeleton in HCECs. Under static culture (0 dyn/cm²), HCECs showed a cobblestone-like morphology (Fig. 4A). Under L-SS, HCECs exhibited significant and organized cytoskeletal reorganization, adopting an elongated morphology aligned parallel to the flow direction, consistent with the findings reported by Molladavoodi et al.²⁹ In contrast, H-SS induced disorganized cellular alignment and an irregular morphology. (Fig. 4A). Transmission electron microscopy indicated that compared with static culture, 8 hours of L-SS intervention narrowed the extracellular space of HCECs and increased the number of bridging granule junctions. Although H-SS also decreased extracellular space, no stable bridging granule cell–cell junctions formed (Fig. 4B). After 8 hours of treatment, L-SS (vs. static) significantly upregulated zonula occludens 1 (ZO-1) at cell junctions, promoting tight intercellular junctions. Conversely, H-SS significantly downregulated ZO-1 expression and induced its discontinuous localization, leading to impaired cellular barrier function (Fig. 4C). Western blot and qPCR confirmed that ZO-1 levels increased under L-SS and decreased under H-SS (Figs. 4D, 4E; Supplementary Fig. S1A).

Figure 4.

In vitro shear stress influences cell alignment and promotes cell–cell junctions formation. (A) Cellular alignment was analyzed by staining cells with Phalloidin (green) and DAPI (blue). (B) Transmission electron microscopy revealed that under L-SS, extracellular gaps narrowed, and bridging granule cell–cell junction formation was enhanced (arrowheads). (C) Cell–cell junction formation was assessed by analyzing the localization and expression of tight junction protein ZO-1 (green). (D, E) The expression of ZO-1 (tight junction protein 1) was evaluated by Western blot (D) and qPCR (E) analysis, showing upregulation in the L-SS group and downregulation in the H-SS group. Quantified protein data are shown in Supplementary Figure S1. The data were collected from at least three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001. ns, not significant.

Piezo1 Mediates Shear Stress–Induced Proliferation and Migration of HCECs

To investigate whether Piezo1 mediates shear stress–induced changes, cells were treated with Piezo1 shRNA (to inhibit Piezo1 expression) and Yoda1 (a selective agonist of Piezo1), respectively, after the shear stress intervention. Among the tested sequences, Sh-1 exhibited the greatest knockdown efficiency and was used for subsequent experiments (Figs. 5A, 5B; Supplementary Fig. S1B). The working concentration of Yoda1 (0.2 µg/mL) was selected based on prior literature and a CCK-8 assay to exclude concentrations causing obvious cytotoxicity (Fig. 5C).²⁷ Compared with the static culture, both the protein level and the RNA level of Piezo1 were significantly upregulated in HCECs after the application of shear stress. Treatment with Yoda1 further upregulated Piezo1 expression, which was inhibited by Piezo1 knockdown (Figs. 5D, 5E; Supplementary Fig. S1C). Subsequent assays showed that shear stress significantly enhanced HCEC migration and proliferation compared with static controls. These effects were potentiated by Yoda1 and attenuated by Piezo1 knockdown (Figs. 5F–I). These findings indicate that Piezo1 may function as a mechanosensitive ion channel, mediating the response to the corresponding shear stress in corneal epithelial cells.

Figure 5.

Piezo1 mediates shear stress-induced proliferation and migration of HCECs. (A, B) The inhibitory effect of lentiviral short hairpin RNAs (Sh-1, Sh-2, and Sh-3) on Piezo1 expression in HCECs was evaluated by Western blot (A) and qPCR (B) analysis. (C) The effect of different concentrations of Yoda1 treatment for 24, 48, and 72 hours on cell viability. Based on the results, 0.2 µg/mL Yoda1 treatment for 24 hours was selected as the experimental condition for subsequent studies. (D, E) HCECs were transfected with Piezo1 lentivirus for 48 hours or treated with 0.2 µg/mL Yoda1 for 24 hours. All cells were then exposed to shear stress. Piezo1 protein and mRNA expression analyzed by Western blot (D) and qPCR (E). (F, H) Scratch assay: wound healing area was measured (with the 0-hour wound area as the standard), and the percentage of unhealed area was assessed. (G, I) Cell proliferation assessed by EdU assay. Red, EdU; Blue, Hoechst. The percentages of EdU-positive cells are shown. Quantified protein data are shown in Supplementary Figure S1. The data were collected from at least three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. ns, not significant.

Piezo1-Evoked Ca²⁺ Signaling Is Required for the Proliferation and Migration Enhancement of HCECs Caused by Shear Stress

Piezo1 senses mechanical cues and converts them into downstream physiological signals by facilitating Ca²⁺ influx.^30,31 Thus, intracellular Ca²⁺ levels were evaluated by Rhod2-AM immunofluorescence to investigate the functional changes of Piezo1. The intracellular Ca²⁺ concentration was significantly elevated in HCECs exposed to shear stress compared with the static control group, with a more pronounced increase under H-SS than L-SS. Yoda1 treatment also exhibited increased Ca²⁺ fluorescence intensity, whereas Piezo1 knockdown markedly reduced fluorescence intensity (Figs. 6A, 6B). These results indicate that shear stress promotes extracellular Ca²⁺ influx, which is closely associated with Piezo1 activation. After extracellular Ca²⁺ chelation with EGTA, HCECs were re-exposed to shear stress to assess calcium-dependent effects. Subsequently, the migratory ability of HCECs was significantly decreased in calcium-free medium compared with the complete medium (Figs. 6C, 6D). Similarly, EdU proliferation assays revealed inhibited proliferative capacity under calcium-free conditions (Figs. 6E, 6F). Together, these data suggest that the Piezo1‑mediated regulation of HCEC proliferation and migration under shear stress is achieved through the modulation of calcium influx.

Figure 6.

Piezo1-evoked Ca²⁺ signaling is required for the proliferation and migration enhancement of HCECs caused by shear stress. (A, B) After modulating Piezo1 activity, shear stress was applied. The Rhod2-AM fluorescence intensity was measured, and the intracellular calcium levels were analyzed. (C, D) HCECs were cultured in calcium-free medium. Scratch assay: the wound healing area was measured (with the 0-hour wound area as the standard), and the percentage of unhealed area was assessed. (E, F) HCECs cultured in calcium-free medium. Cell proliferation assessed by EdU assay. Red, EdU; blue, Hoechst. The percentages of EdU-positive cells are shown. The data were collected from at least five independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Activation of Piezo1 by Shear Stress Upregulated the Expression of NLRP3 Inflammatory Vesicles and Triggered Cellular Pyroptosis

The inflammasome sensor NLRP3 is implicated in DED progression^27,32 and promotes pyroptosis via caspase-1/caspase-4–mediated cleavage of GSDMD.^33,34 Consistent with this mechanism, our results showed that, after shear stress intervention, cells showed morphological changes consistent with pyroptosis. Transmission electron microscopy revealed cell swelling and loss of membrane integrity and endoplasmic reticulum expansion (Fig. 7A), and live cell imaging provided supportive evidence of altered cellular status after shear stress (Fig. 7B). Shear stress upregulated Piezo1 and increased protein/mRNA levels of NLRP3, pyroptosis-related proteins (caspase-1, caspase-4, and GSDMD), and inflammatory cytokines (IL-1β, IL-6, and IL-18), with more pronounced changes under H-SS (Figs. 7C–K; Supplementary Fig. S1D). Activating Piezo1 with Yoda1 significantly increased NLRP3, caspase-1, and IL-1β while decreasing ZO-1 (Figs. 7L–Q; Supplementary Figs. S1E, S1F). Conversely, Piezo1 knockdown reduced inflammatory factor expression but elevated ZO-1. These results indicate that Piezo1 acts as a key mechanosensor, converting abnormal shear stress into pathological signals that simultaneously activate the NLRP3 inflammasome and compromise epithelial barrier integrity under excessive mechanical stimulation.

Figure 7.

Piezo1 mediates shear stress–induced activation of NLRP3 inflammatory vesicles and cellular pyroptosis in HCECs. (A) Transmission electron microscopy revealed cell swelling and disruption of cell membrane integrity (arrowheads), and endoplasmic reticulum expansion (asterisks). Enlarged views are shown to highlight the ultrastructural changes. (B) Representative bright-field live cell images of HCECs under shear stress. Live cell imaging provided supportive evidence of blurred outlines of pyroptotic cells after shear stress (arrowheads). (C) Western blot analysis indicated the relative protein expression levels of Piezo1, NLRP3, GSDMD, caspase-1, IL-1β, IL-6, IL-18, and caspase-4 in HCECs after the shear stress intervention. (D–K) HCECs were stimulated with low and high shear stress, and qPCR was performed to evaluate the mRNA expression levels of Piezo1, NLRP3, GSDMD, caspase-1, caspase-4, IL-1β, IL-6, and IL-18. (L) Western blot was used to analyze the expression levels of NLRP3 in HCECs treated with Yoda1 or the lentiviral vector LV-Piezo1 and then stimulated with shear stress. (M–P) qPCR was performed to evaluate the mRNA expression levels of NLRP3, caspase-1, GSDMD, and IL-1β in HCECs treated with Yoda1 or the lentiviral vector LV-Piezo1 and then stimulated with shear stress. (Q) Western blot analysis was used to analyze the expression levels of ZO-1 in HCECs treated with Yoda1 or the lentiviral vector LV-Piezo1. Quantified protein data are shown in Supplementary Figure S1. The data were collected from at least four independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. ns, not significant.

Discussion

In this study, we established a controlled blink mouse model to investigate the impact of shear stress on the corneal epithelium. Under physiological conditions, normal mice blink at a frequency of one to two times per minute.³⁵ In our experimental design, we used a frequency of 0 blinks/min as a baseline control to isolate the effects of blinking-induced mechanical shear from general procedural artifacts and environmental exposure. Although a decreased blinking rate primarily causes damage via increased evaporation and exposure,³⁶ elevated blinking in DED—often a compensatory response to refresh the tear film^37,38—simultaneously subjects the ocular surface to excessive mechanical shear and friction,^39,40 thereby exacerbating the symptoms of DED.

In our model, both 10 and 40 blinks/min controlled blinking elicited abnormal ocular surface shear stress. After the intervention, mice exhibited increased corneal fluorescein staining and decreased tear secretion. We propose that sustained high-frequency shear stress elevates friction at the lid wiper–ocular surface interface, concurrently reducing mucin secretion mucin secretion^40,41 and increasing surface roughness.⁴² As the lubricating layer is compromised and surface roughness increases, tear film stability declines and tear evaporation accelerates, ultimately leading to the disruption of epithelial homeostasis.

Shear stress sensing is recognized as a fundamental mechanobiological process observed across diverse cell types, including vascular endothelial cells, osteocytes, chondrocytes, and epithelial cells.^43–45 Distinct flow regimes—laminar, transitional, and turbulent flow—activate distinct cellular signaling pathways that drive varied physiological responses.⁴⁶ In the ocular context, the corneal epithelium is continuously exposed to tear film dynamics, where blinking generates fluid shear stress across the ocular surface.^1,47 Physiological blinking facilitates the uniform distribution of tear film from the inferior lacrimal meniscus, thereby maintaining tear film stability. Conversely, abnormal blinking patterns have been strongly associated with various ocular surface pathologies.^48,49

To investigate the effects of shear stress under controlled conditions, fluid flow was applied at low (2.6 dyn/cm²) and high (7 dyn/cm²) levels by adjusting the rotational speed of an orbital platform. While in vivo tear film dynamics are governed by a complex, multiparametric interplay of tear viscosity, film thickness, and lid velocity, this orbital shaker model provides a validated, controlled mechanical gradient to systematically study cellular responses. Three-dimensional CFD simulations indicated that the Reynolds number reached 526 at 50 rpm and 3685 at 200 rpm (see the formula in the Supplementary Material). Because the Reynolds number is a fundamental dimensionless quantity used to predict fluid flow patterns, we used it as a primary criterion to characterize the mechanical environment in our system. Based on criteria established by Feng et al.⁵⁰ for similar orbital rotating systems, flow at 200 rpm corresponds with a turbulent regime, whereas flow at 50 rpm represents a transitional state between laminar and turbulent regimes.

We recognize that these defined flow regimes are mechanistic inferences based on our CFD model rather than an exact physical replication of the tear film; however, they offer a rigorous framework for correlating mechanical intensity with biological outcomes. Although transitional flow is often associated with pathological conditions such as aneurysms or vascular stenosis,^51,52 it is indeed physiologically present in the human body. Under smooth surface conditions, fluid generally flows in a steady laminar manner; surface irregularities can induce local vortices and transitional flow. Sustained turbulence typically occurs only when surface roughness exceeds one-half of the channel width.⁵³ Under physiological conditions, tear film spreading during blinking potentially resembles transitional flow, attributable to the inherent roughness of the corneal and conjunctival surfaces. In DED, factors such as corneal ulcers, scarring, chronic inflammation, and tear film instability can further increase surface roughness and alter blink dynamics, promoting turbulence-like flow behavior. Thus, we propose that low shear stress mimics physiological tear film dynamics, and high shear stress represents a pathological mechanical stimulus. Investigating the impact of ocular surface shear stress—particularly under high or pathological conditions—on corneal epithelial cells may offer novel insights for dry eye therapy.

Our study suggests that Piezo1 plays a significant role in mediating the response of the corneal epithelium to abnormal ocular surface shear stress, potentially contributing to the inflammation and epithelial dysfunction associated with ocular surface disorders. Emerging evidence highlights the role of Piezo1 in immune responses and mechanically induced inflammation.^54–57 Specifically, we demonstrated that extracellular mechanical stimuli, such as shear stress, trigger calcium ion influx through Piezo1 channel gating,⁵⁸ which in turn activates the NLRP3 inflammasome—a key driver of inflammatory cascades.^59,60 Consistent with prior work by Ran et al.,⁵⁷ we found that mechanical stimuli promote NLRP3-dependent inflammation via Piezo1. In macrophages, shear stress upregulates Piezo1 and NLRP3, enhancing the secretion of GSDMD, IL-1β, and CCL2 to exacerbate inflammation.^61,62 Using an in vitro hydrodynamic model, we further showed that H-SS upregulates NLRP3 through Piezo1, amplifying inflammatory responses and triggering pyroptosis in human corneal epithelial cells.

Beyond inflammation, Piezo1 influences corneal epithelial barrier integrity, which is critical in DED. Tight junctions, exemplified by ZO-1, safeguard ocular surface integrity, and reduced ZO-1 expression has been linked to epithelial dysfunction in dry eye.⁶³ Our results reveal that L-SS promotes intercellular junction formation and upregulates ZO-1, whereas H-SS impairs junctional assembly, leading to barrier breakdown. Notably, Western blot analysis demonstrated a negative correlation between Piezo1 and ZO-1 expression, aligning with observations in other tissues.⁶⁴ However, the concurrent upregulation of both Piezo1 and ZO-1 under L-SS suggests that additional mechanisms beyond Piezo1-mediated pathways contribute to junctional regulation, underscoring the complexity of mechanical signaling in epithelial homeostasis.

Piezo1 also mediates cellular responses to shear stress in HCECs, promoting proliferation and migration—a finding consistent with its role in endothelial cells.^65–67 Paradoxically, while H-SS disrupts corneal epithelial homeostasis by inducing inflammation and pyroptosis, it simultaneously stimulates proliferation and migration, potentially reflecting a compensatory adaptation to mechanical injury. This dual role emphasizes Piezo1's multifaceted involvement in balancing repair and damage under varying stress conditions.

Several limitations of this study warrant acknowledgment. First, although we modeled ocular surface shear stress effects, the precise shear stress patterns in vivo remain undetermined. Our experimental design applied relatively low shear stress (≤7 dyn/cm²) over a prolonged period to simulate chronic pathological conditions, because we consider shear stress-induced ocular surface injury to be a long-term, cumulative process. Second, for practical reasons related to throughput and cost, we did not use the ibidi flow system (ibidi, Fitchburg, WI, USA) for a more refined fluid simulation. Finally, although the SV40-immortalized HCEC model offered stable and reproducible conditions for mechanobiological investigation, validation with primary corneal epithelial cultures is necessary to confirm clinical relevance and substantiate the role of Piezo1 in human disease.

Conclusions

Piezo1 acts as a critical mechanosensor that transduces ocular surface shear stress into intracellular signaling pathways, regulating epithelial barrier integrity and inflammatory responses. Our findings demonstrate that, although low-intensity shear stress supports corneal epithelial homeostasis, aberrant high-intensity shear stress triggers barrier failure and pyroptosis through Piezo1-mediated mechanotransduction. These insights highlight the significance of Piezo1 in the pathophysiology of altered ocular surface dynamics and suggest its potential as a therapeutic target for corneal injuries induced by pathological mechanical stimulation.