Mechanical Stretch Promotes Sustained Proliferation and Inflammation in Developing Human Airway Smooth Muscle

Abstract

Preterm infants frequently require respiratory support, including continuous positive airway pressure (CPAP), that imposes mechanical stretch on highly compliant perinatal airways. How this excess stress impacts airway development and function is not completely understood. Using human fetal airway smooth muscle (fASM), a key cell type in airway contractility and remodeling, as a model, we investigated the effects of stretch, focusing on the role of mechanosensitive ion channels Piezo1 and Piezo2. We found that CPAP-like static stretch did not alter Piezo1 and Piezo2 protein expression per se and had minimal effect on fASM cell proliferation or IL-6 production during the stretch period. However, CPAP-like stretch produces long-term effects in fASM, leading to increased cell proliferation and IL-6 production during the post-stretch period, though interestingly, it does not enhance extracellular matrix deposition. The role of Piezo channels appears context-dependent in that the Piezo1 antagonist GsMTx4 reduced baseline proliferation in non-stretched cells but slightly increased proliferation in stretched cells. Piezo1 and Piezo2 inhibition did not alter IL-6 production. These results suggest that stretch induces sustained increase in cell proliferation and inflammatory responses, which may contribute to long-term remodeling in former preterm infants initially exposed to CPAP.

Graphical Abstract

Introduction

Infants born preterm often have respiratory insufficiency soon after birth and require respiratory support therapies such as supplemental oxygen, mechanical ventilation, and continuous positive airway pressure (CPAP) (1–4). CPAP is an increasingly used noninvasive therapy that has been shown to reduce neonatal mortality and the incidence of bronchopulmonary dysplasia (BPD) in preterm infants. In addition to promoting alveolar opening and distension, CPAP exerts outward intraluminal pressure on bronchial airways and distends the highly compliant airways of premature infants. Despite its clinical benefits in terms of oxygenation in prematurity, the longer-term impact of the mechanical forces of CPAP particularly on the bronchial airways is less explored. Limited studies in mouse models show that neonatal CPAP increases airway hyperresponsiveness and remodeling (5–7), suggesting that CPAP may have long-term detrimental effects. How such effects are mediated is not yet clear.

The lung is a mechanosensitive organ continuously exposed to dynamic mechanical stimuli, including airflow, pressure fluctuations, and tissue deformation during respiration. Mechanotransduction—the process by which cells convert mechanical stimuli into biochemical signals—plays a critical role in lung physiology and pathophysiology. Among the critical components governing lung mechanics are mechanosensitive ion channels, particularly the Piezo proteins (Piezo1 and Piezo2) that are large transmembrane, nonspecific ion channels expressed by multiple organ systems. Piezos respond to a variety of mechanical stimuli including stretch, touch, shear stress, compression, and pulsatile strain (8–12). Piezo channels play critical roles in the development of multiple organs and in congenital diseases. For example, Piezo1 global knockout mice die in the embryonic stage with defects in vascular remodeling (13, 14) and Piezo2-deficient mice die within 24 h of birth due to respiratory distress (15). In humans, Piezo1 mutations are associated with genetic diseases, including generalized lymphatic dysplasia (16, 17), the rare congenital human disorder Prune Belly Syndrome (18), and autosomal dominant dehydrated hereditary stomatocytosis (19–21). Human Piezo2 mutations are associated with Gordon syndrome, Marden-Walker syndrome, distal arthrogryposis (22, 23), and a broad spectrum of peripheral sensory dysfunctions (24–28). Genetic deletion studies show that Piezo proteins are critical for lung development, with essential roles in pulmonary vascular formation and respiratory adaptation (15, 29–31). Airway smooth muscle (ASM), a key cell type involved in airway tone, contractility and remodeling, expresses both Piezo1 and Piezo2 proteins (32–34), with developing fetal ASM (fASM) showing higher expression compared to adult ASM (32). Recent studies, including our own, show that Piezo channels regulate ASM calcium signaling, extracellular matrix (ECM) production, cell stiffness, and cell migration (32–34).

Perinatal oxygen is known to induce airway inflammation in the context of BPD as well as wheezing disorders (35–37). Such inflammatory responses can have longer term effects on the developing airway. In adult airways, mechanical forces can also induce airway inflammation (38–40). Whether similar effects occur in developing airway is not known.

Given that preterm infants exposed to CPAP experience abnormal mechanical stretch, it is important to investigate how excess stretch impacts on developing ASM in the short- and long-term, and whether Piezo-mediated mechanotransduction is involved. We used human fASM cells as an in vitro model of the developing airway to address the hypothesis that stretch, via Piezo channels, has prolonged detrimental effect on developing airways.

Materials and Methods

Cell Culture

Canalicular stage human fetal ASM (fASM) cells (18–22 weeks gestation, a fetal age which closely approximates to the age of potential NICU survival during which mechanical support including CPAP would be administered) were enzymatically dissociated from tracheobronchial trees following fetal demise, as previously described (41). The Mayo Clinic Institutional Review Board (IRB) considers deidentified fASM acquisition exempt. Each sample represents cells isolated from an individual lung and correlates to one “n”. All cell lines were used at passage of 3–7. Genetic sex of fASM samples were determined and both female and male cells were utilized in all experiments. Cells are characterized as smooth muscle cells through expression of key markers such as smooth muscle actin, calponin, and acetylcholine (ACh) receptor (41, 42). fASM were cultured in 10% fetal bovine serum (FBS), phenol red-free Dulbecco’s Modified Eagle Medium (DMEM)/F12 growth media and growth-arrested in 0.5% FBS media for 24h before being exposed to experimental conditions. Our lab previously tested for mycobacterium with negative results; no testing was conducted during the period of this study.

Cell Stretch and Treatment

Following an established protocol in our laboratory (32, 43, 44), cells were seeded on collagen-coated 6-well Bioflex plates at 1.5 × 10⁵/well, serum starved the next day for 24h and then put on a Flexcell system. For static stretch experiments, cells were exposed to 5%, 10% or 15% static stretch. Non-stretched cells were served as controls. For CPAP-like stretch experiments, one set of cells was exposed continuously to 5% cyclic strain at 40 cycles/min mimicking breathing and served as a “control”. Another set of cells underwent 5% cyclic strain superimposed with 5% or 10% static stretch thus mimicking CPAP. In some experiments, fASM cells were treated with Piezo1 antagonist GsMTx4 (1 μM, MedChemExpress, HY-P1410) or Piezo2 antagonist D-GsMTx4 (1 μM, MedChemExpress, HY-P1410B) for one hour prior to stretch and the treatments continued thereafter during the stretch protocol. After 24h stretch treatment, fASM cells were either trypsinized and counted or were lysed in Cell Lysis Buffer (Cell Signaling Technology, 9803S) with protease inhibitors. The concentration and treatment time of antagonists were chosen based on our previous study showing the efficiency of Piezo1 and Piezo2 antagonists in inhibiting stretch induced changes in intracellular Ca²⁺ and ECM production in developing ASM (32).

For the post-stretch experiments, recombinant human IL-13 (R&D Systems 213-ILB) 25 ng/ml or TGFβ (R&D Systems 240-B) 2 ng/ml were added into the culture for three days in 96-well plates.

Protein Measurements

Protein quantification was performed using BCA protein assays (Thermo Fisher #89901). Protein expression levels were measured using JESS, a capillary-based electrophoresis system (Protein Simple, San Jose CA, United States). Following the manufacturer’s instructions, 0.3 μg protein was loaded into 12-230-kDa or 66-440-kDa JESS separation modules with appropriate primary and secondary antibodies validated for that system. The primary antibodies used were Piezo1 (Novus Biologicals, NBP1–78537, 1:20 dilution), Piezo2 (Novus Biologicals, NBP2–61130, 1:25 dilution), Ki67 (Cell Signaling Technology, 9449s, 1:20 dilution), Cyclin E1 (Abcam, ab3927, 1:25 dilution), proliferating cell nuclear antigen (PCNA) (Cell Signaling Technology, 13110, 1:50 dilution). The secondary antibodies were purchased from ProteinSimple: Anti-Mouse Secondary HRP Antibody (# 042–205) and Anti-Rabbit Secondary HRP Antibody (# 042–206). Total Protein Detection Module for Chemiluminescence based total protein assays (#DM-TP01) were used to determine the total protein expression. The expression of target proteins was normalized to total protein. Digital representations of the electropherograms were used and then quantified using Compass for Simple Western Software. All antibodies were validated with cells treated with inhibitors or siRNA knockdown.

Proliferation assay

Proliferation of fASM cells based on DNA content was assayed using the CyQUANT^™ NF Cell Proliferation Assay kit (Invitrogen, C35006). Cells in Bioflex plates were treated for 24h as described above and then were trypsinized, counted, and seeded in black 96-well plates (3,000 cells/well). Three days later, cell proliferation was measured by using the CyQUANT^™ NF Cell Proliferation Assay kit following the manufacturer’s protocol. Dye binding to DNA (fluorescence) was measured on a FlexStation microplate reader. Cell numbers were calculated based on fluorescence signal calibrations.

ECM Deposition Assay

Cells were seeded at 3,000/well in 96-well plates as described in the proliferation assay. After CyQUANT NF assay, the cells in the wells were lysed with 0.016N NH₄OH. The wells were washed, and then were blocked in Li-Cor Odyssey blocking buffer before overnight incubation in primary antibodies, including anti-collagen I + collagen III (Abcam, ab34710, 1:200) and anti-fibronectin (Abcam, ab2413, 1:500). After washing, plates were incubated with infrared dye-conjugated goat anti-rabbit IgG secondary antibody (Li-Cor Biosciences 926–32211), washed, and imaged using a Li-Cor Odyssey XL system with densitometry quantification. Wells incubated with the secondary antibody were used as background controls. ECM expression levels were normalized to cell numbers derived from CyQUANT NF assay.

ELISA

Cells in Bioflex plates were treated for 24h as described above and then were trypsinized, counted, and seeded in black 96-well plates (3,000 cells/well). Three days later, culture supernatants were harvested. IL-6 levels in culture supernatants were measured using Human IL-6 DuoSet ELISA kit (R&D Systems, DY206) following the manufacturer’s protocol. Absorbance was read at 450 and 530 nm using a FlexStation microplate reader.

Statistical Analysis

GraphPad Prism Software (La Jolla,CA, USA) was used for all statistical analyses. Based on data distribution and experimental designs, specific statistical analyses were applied. We highlighted the statistical test used in the legend of every figure. Experimental data were analyzed by Wilcoxon test, Two-way ANOVA, or paired t-test based on experimental designs. Statistical significance is indicated by exact P value. Values are presented as means α SD.

Results

Regulation of Piezo protein expression by stretch

We have shown previously that 5% static stretch (mimicking CPAP) superimposed on 5% oscillatory stretch (mimicking breathing) increases both Piezo1 and Piezo2 mRNA expression in fASM (32). To further investigate whether the extent of stretch matters, we exposed fASM cells to different levels of stretch (5%, 10% and 15%) for 24h. First, we examined the effects of static stretch alone on Piezo protein expression. We found that 5% static stretch significantly increased Piezo1 but not Piezo2 expression, whereas 10% and 15% static stretch did not significantly alter either Piezo1 or Piezo2 expression (Fig. 1A–C). Next, we analyzed the effects of static stretch superimposed on 5% oscillatory stretch. We found that 5% and 10% CPAP did not have significant effects on Piezo1 or Piezo2 expression (Fig. 1D–F). To avoid the detrimental effects associated with respiratory support therapies, lower levels of oxygen or stretch are often used clinically for preterm infants (1–4). We aimed to understand whether the low-level stretch produces detrimental effects in fASM. Thus, we chose to use 5% CPAP in subsequent experiments.

Figure. 1.

The effect of stretch in Piezo protein expression in fASM cells. (A, B): fASM cells grown on Flexcell plates were exposed to 5%, 10% or 15% of static stretch for 24h. Piezo1 and Piezo2 protein expression was determined by JESS assays. The expression levels in stretched cells were normalized to that of control cells without stretch treatment. Representative JESS images are shown in (C). SS: static stretch. Total protein signal values (x 10⁵) are shown at the bottom of each lane. (D, E): fASM cells grown on Flexcell plates were exposed to 5% or 10% of static stretch superimposed on 5% oscillatory stretch (CPAP) for 24h. The Piezo1 and Piezo2 protein expression levels in these cells were normalized to that of control cells with 5% oscillatory stretch only. Representative JESS images are shown in (F). Data are expressed as means ± SD (n = 6–7 cell lines, each dot represents one cell line). Open symbol: male. Solid symbol: female. All data pass Shapiro-Wilk Normality tests. For statistical analysis, paired t tests are applied to control and treatment groups. Exact P value is above the data in Fig. 1A.

Stretch does not have immediate effects on fASM cell proliferation

To determine whether CPAP impacts fASM cell proliferation, we subjected cells to 5% oscillatory stretch (control), or 5% static stretch superimposed on 5% oscillatory stretch (CPAP). Some cell groups were treated with Piezo antagonists GsMTx4 (Piezo1) or D-GsMTx4 (Piezo2). After 24h of stretch, cells were harvested and counted. Alternatively, cells were lysed immediately for protein expression of proliferation markers Ki67, PCNA, and cyclin E1. We found that stretch did not have immediate significant effects in fASM cell proliferation based on cell counts (Fig. 2A) and expression of proliferation markers (Fig. 2B–D). Moreover, the presence of Piezo antagonists did not significantly affect fASM cell proliferation under either control or stretch conditions during the stretch period.

Figure. 2.

The short-term effect of stretch in fASM cell proliferation. fASM cells on Flexcell plates were pretreated with piezo inhibitors for 1h and then were exposed to 5% static stretch with 5% oscillatory stretch (5% CPAP) with continuous presence of the inhibitors. Control cells were exposed to 5% oscillatory stretch only. Cells were harvested after 24h treatment and counted (A). The protein expressions of Ki67, PCNA, and Cyclin E1 in these cells were determined by JESS (B-D). Representative JESS images are shown beneath each figure with the same sample order. The first lane shows the protein ladder. Open symbol: male. Solid symbol: female. All data pass Shapiro-Wilk Normality tests. For statistical analysis, Two-Way ANOVA with Tukey’s multiple comparison test are applied. Data are expressed as means ± SD (n = 6–7 cell lines, each dot represents one cell line). The interaction p-values are below the corresponding panels.

Stretch has long-term effect on fASM cell proliferation

fASM cells were exposed to stretch on Bioflex plates for 24h with or without Piezo antagonists as described above. Cells were then harvested and cultured further for 3 days without stretch to mimic the post-stretch period. Cell proliferation was measured by CyQUANT NF assays. Our results show that fASM cells previously exposed to stretch/CPAP have significantly higher levels of proliferation in the absence of any further stimulation (Fig. 3 A, B). The presence of Piezo1 antagonist GsMTx4 during the first 24h stretch treatment significantly inhibited proliferation in control (non-stretched) cells but did not inhibit the proliferation of cells previously exposed to stretch/CPAP (Fig. 3A). Piezo2 antagonist D-GsMTx4 did not have significant effects in either control or CPAP-exposed cells (Fig. 3B).

Figure. 3.

The long-term effect of stretch in fASM cell proliferation. fASM cells on Flexcell plates were pretreated with inhibitors to Piezo1 (A) or Piezo2 (B) for 1h and then were exposed to 5% static stretch with 5% oscillatory stretch (5% CPAP) with or without the inhibitors. Control cells were exposed to 5% oscillatory stretch only. Cells were harvested after 24h treatment and seeded into 96-well plates. After 3-day culture without stretch, cell proliferation was measured by CyQUANT assay. Open symbol: male. Solid symbol: female. All data pass Shapiro-Wilk Normality tests. For statistical analysis, Two-Way ANOVA with uncorrected Fisher’s LSD test are applied. Data are expressed as means ± SD (n = 6–7 cell lines, each dot represents one cell line). The interaction p-values are below the corresponding panels.

Stretch has long-term effect on IL-6 production

We investigated whether stretch regulates inflammatory responses in fASM cells by examining IL-6 production during stretch and after stretch. IL-6 is a predictor of perinatal sepsis (45, 46), and increased IL-6 is a marker of BPD (47–54). Furthermore, oxygen is known to increase cytokines such as IL-6 in developing lung (55–58). Thus, potential stretch effects on IL-6 can have longer term effects. We found that IL-6 levels in the culture supernatants were comparable between control and CPAP-exposed fASM cells at the end of 24h stretch treatment (Fig. 4A). Compared to controls, fASM cells previously exposed to CPAP produced significantly higher levels of IL-6 when these cells were cultured for additional three days (Fig. 4B). Because IL-6 levels were highly variable among experiments done on different days using different cell lines, we normalized the IL-6 values of CPAP-exposed cells to the values of control cells with each cell line to better understand stretch effect (Fig. 4C). Piezo1 and Piezo2 antagonists did not have significant effects in IL-6 production in either control or CPAP-exposed cells (Fig. 4D, E).

Figure. 4.

The effect of stretch in IL-6 production by fASM cell. Cells on Flexcell plates were exposed to 5% CPAP or 5% oscillatory stretch (control) with or without Piezo inhibitors for 24h. Culture supernatants were harvested and were assayed for IL-6 production by ELISA (A). CPAP-exposed or control cells were harvested and then cultured in 96-well plates for 3 days without stretch. IL-6 levels in the 3-day culture supernatants were determined by ELISA. IL-6 production fold changes were calculated by normalizing IL-6 produced by CPAP-exposed cells to IL-6 produced by control cells within each fASM cell line (B-D). Open symbol: male. Solid symbol: female. All Data pass Shapiro-Wilk Normality tests except Data in (D) which are then logarithmic transformed. Data are analyzed with paired t-test (A), Wilcoxon test (B, C), or Two-Way ANOVA with uncorrected Fisher’s LSD test (D, E). Data are expressed as means ± SD (n = 5–7 cell lines, each dot represents one cell line). The interaction p-values are below the corresponding panels.

Stretch blunts long-term IL-13 response

IL-13 is a proinflammatory cytokine that has important roles in tissue remodeling (59, 60). We examined whether stretch has long-term effect in IL-13 induced responses in fASM. After 24h stretch treatment, CPAP-exposed fASM or control cells were cultured with or without IL-13 for three additional days. Subsequently, cell proliferation and IL-6 production in the supernatants were measured by CyQUANT assay and ELISA, respectively. As expected, IL-13 stimulated fASM cell proliferation in control fASM cells (Fig. 5A). CPAP exposure alone increased fASM cell proliferation. However, additional IL-13 stimulation failed to further enhance proliferation of CPAP-exposed fASM cells (A, B). Like proliferation, IL-13 stimulated IL-6 production in control fASM cells (Fig. 5C). But unlike proliferation, IL-13 treatment induced higher levels of IL-6 production in both control and CPAP-exposed fASM cells. However, when fold changes were calculated, IL-13 induced IL-6 response was lower in CPAP-exposed cells compared with control cells (Fig. 5D).

Figure. 5.

The long-term stretch effect in IL-13 response. Control or 5%CPAP-exposed fASM cells were cultured in 96-well plates for 3 days with or without IL-13 (25 ng/ml). Cell proliferation was measured by CyQUANT assay (A). IL-6 levels in the culture supernatants were determined by ELISA (B). Open symbol: male. Solid symbol: female. Data in (A) but not (B) pass Shapiro-Wilk Normality tests. Data in (B) are then logarithmic transformed. Data are expressed as means ± SD (n = 5–7 cell lines, each dot represents one cell line) and analyzed with Two-Way ANOVA with Tukey’s multiple comparisons test. The interaction p-values are below the corresponding panels.

Stretch does not have long-term effects on ECM deposition

We examined the long-term effects of stretch in fASM remodeling by culturing control and CPAP-exposed fASM cells with or without TGFβ stimulation for 3 days. Subsequently, we performed ECM deposition assays using antibodies recognizing collagen I + collagen III, or fibronectin. Control and CPAP-exposed fASM cells showed similar basal deposition levels for collagen I + collagen III, and fibronectin. Moreover, TGFβ induced comparable levels of ECM deposition in control and CPAP-exposed fASM cells (Fig. 6).

Figure. 6.

The long-term effect of stretch in ECM deposition. Control or CPAP-exposed fASM cells were cultured in 96-well plates for 3 days with or without TGFβ (2 ng/ml). The amount of Col1/Col3 or fibronectin deposited on the plates were measured by ECM deposition assays. The presented values were normalized to cell numbers in the plates. Open symbol: male. Solid symbol: female. All data pass Shapiro-Wilk Normality tests. For statistical analysis, Two-Way ANOVA with Tukey’s multiple comparisons test were applied. Data are expressed as means ± SD (n = 5–6 cell lines, each dot represents one cell line). The interaction p-values are indicated below the corresponding panels.

Discussion

CPAP has been increasingly used in the neonatal intensive care unit to treat preterm infants with respiratory issues (1–4). While CPAP is clearly beneficial towards increasing oxygenation and perinatal outcomes, the longer-term impact of CPAP particularly on the more compliant developing airways remains unclear. In mouse models, neonatal CPAP can lead to airway hyperreactivity and remodeling (5–7). To address this question, using human fetal ASM cells as an in vitro model, we investigated the effects of mechanical stretch on fASM proliferation, IL-6 production, and ECM deposition. Recognizing the importance of mechanotransduction, we examined the role of mechanosensitive Piezo channels in stretch-induced responses in fASM. We show here that low level (5%) of static mechanical stretch that mimics CPAP has very mild effects on Piezo protein expression in fASM and no immediate effects in fASM proliferation and IL-6 production. However, CPAP-like stretch produces long-term effects in fASM with increased cell proliferation and IL-6 production following the period of stretch, although the role of Piezo channels in this effect is less clear. These data suggest that the effects of stretch can be sustained and may be responsible for longer-term detrimental effects associated with CPAP.

Mechanical stretch that is a necessary intervention in the NICU for premature infants and imposes beneficial forces on the underdeveloped peripheral lung, but also imposes static stretch on the compliant bronchial airways resulting in greater distention when exposed to externally imposed strain (61–64). The forces of stretch are inherent to the lung given the cyclic stretch the lung undergoes during normal breathing, and even in utero. While stretching is a normal process, excessive stretch may have damaging effects, leading to inflammation, alveolar simplification, and airway hyperresponsiveness (7, 65). This is better demonstrated in the context of mechanical ventilation that provides critical respiratory support for preterm survival (66, 67) but can initiate inflammatory cascades and cytokine production (65), and airway epithelial disruption (68). While CPAP has been less studied, data in animal models suggest CPAP is not completely benign and may also result in long-term changes in airway structure and function similar to reactive airway disease. For example, neonatal mice exposed to 3h cycles of CPAP for the first 7 days showed increased airway reactivity (assessed by lung slice physiology) that was sustained 2 weeks thereafter, suggesting a persistent impact of neonatal CPAP (5).

The consequences of stretch have been explored in vitro using ASM and fibroblast cells. In adult human ASM cells, cyclical stretch induces proliferation and increases migration (69–71). Interestingly, in human embryonic lung fibroblasts lower levels of stretch increase proliferation (72) while higher amplitude stretch decreases proliferation but increases ECM production suggesting a potential dose–response (72). Other studies in adult ASM have also shown stretch impacts on ECM deposition and remodeling (69, 73). However, there are currently no studies exploring the impact of stretch in developing ASM, where we provide insights into both a potential dose response effect, and importantly differences between immediate vs. postponed effects of stretch.

The potential mechanisms by which stretch effects are mediated have been explored in adult but not developing ASM. TGF-β1 has been shown to contribute to ASM hypertrophy and proliferation (74, 75), and in this regard, adult ASM exposed to stretch show increased TGF-β1 expression (75). In the neonatal mouse model of CPAP-induced airway hyperreactivity, the extracellular calcium-sensing receptor has been found to be a potential mediator of stretch effects (7). Interactions between ASM and ECM have also been shown to be important (76, 77). However, the upstream mechanosensitive pathways underlying stretch effects have yet to be elucidated. In this regard, Piezo channels have garnered substantial attention (8, 78–81), especially given their high expression in the lung (29, 82–85) and the potential role of Piezo2 in lung innervation.

We previously showed that human fASM express both Piezo1 and Piezo2 (32) which mediate changes in Ca²⁺ and ECM production. In the current study, we find that stretch differentially regulates Piezo1 and Piezo2 expression in fASM. Specifically, 5% static stretch (representing mild CPAP) significantly upregulates Piezo1 expression, but higher levels of static stretch (10% or 15%) had no significant effects on either Piezo1 or Piezo2 expression. Surprisingly, when static stretch was combined with 5% oscillatory stretch to mimic breathing, neither Piezo1 nor Piezo2 expression was significantly altered at the 24h time point. Using a similar experimental protocol, our previous study showed that 5% CPAP significantly increased Piezo1 and Piezo2 mRNA expression at 12h time point (32). This discrepancy likely reflects the regulation differences between mRNA and protein expression. Moreover, these data contrast with our previous findings in adult ASM where 10% stretch significantly increases Piezo1 and Piezo2 (43), suggesting that regulation of Piezo expression may be age dependent. Here, the upstream regulators of Piezo are not well known, and therefore it is difficult to speculate whether fetal vs. adult ASM differ in the expression or function of such regulators. Nonetheless, given that Piezo proteins are important to lung development, maintaining homeostatic expression of Piezo proteins in fASM upon stretch may be a protective mechanism to ensure normal airway development.

Mechanical stretch modulates proliferation in several cell types, including fibroblasts, macrophages, and epithelial cells, and Piezo1 is thought to play a role (86–89). How stretch affects ASM cell proliferation has not been explored. Our study shows that 24h of stretch did not affect fASM cell proliferation during stretch but significantly increased proliferation thereafter. Consistent with the absence of changes in cell numbers, cell proliferation-associated proteins, including Ki67, PCNA and cyclin E1, did not show significant increases with stretch treatment. Although the expression levels of these proteins were not assessed following the stretch period, the observed increase in cell numbers suggests that their expression may be elevated post-stretch. This hypothesis will be explored in future studies. Notably, Piezo1 inhibition during the stretch period significantly reduced baseline proliferation in control cells after stretch, indicating a critical role of Piezo1 in homeostatic fASM proliferation. Paradoxically, Piezo1 inhibition during the stretch period tended to increase proliferation in CPAP-exposed cells. Future studies are needed to understand the complex roles of Piezo1 in fASM proliferation with or without mechanical stretch.

Mechanical stretch induces proinflammatory cytokine production, such as IL-6, in various cell types, including neurons (90), smooth muscle cells (91, 92), lung epithelial cells (93, 94), macrophages (95), osteoblasts (96), and fibroblasts (97, 98). Although evidence suggests that Piezo1 is involved in proinflammatory cytokine production in response to hydrostatic pressure or mechanical compression (90, 99), the functional role of Piezo proteins in mechanical stretch-induced cytokine production has not been investigated. In this study, we found that stretch did not affect IL-6 production by fASM during stretch but significantly increased IL-6 levels thereafter. This suggests mechanical stretch induces a sustained inflammatory response in fASM. Inhibition of Piezo1 or Piezo2 did not significantly alter IL-6 production, indicating that alternative mechanotransduction pathways may mediate this prolonged inflammatory response.

Our previous study shows increased secretion of collagen I, collagen III, and fibronectin by fASM during 5% stretch (32). The current study shows that stretch-exposed fASM cells displayed no significant differences in basal ECM deposition compared to controls and exhibited no altered responses to TGF-β stimulation during post-stretch period. Although these findings indicate that CPAP may not have long-term effects in ECM remodeling in fASM cells, ECM deposition in this study was normalized to cell numbers and presented as ECM per cell. Given the increased proliferation of CPAP-exposed fASM cells, the total ECM levels in this population would likely be elevated, potentially contributing to CPAP-induced airway remodeling in long-term.

Our data show that mechanical stretch impairs post-stretch IL-13 responses in fASM cells. CPAP-exposed cells failed to proliferate in response to IL-13 stimulation. Although these cells produce more IL-6 upon IL-13 stimulation, the response was attenuated compared to control cells. Given IL-13 plays critical roles in airway inflammation and remodeling (100–102), desensitize IL-13 response in stretched fASM cells may be an intrinsic mechanism that protects the developing ASM from further exacerbated inflammation.

In summary, this study demonstrates that CPAP-like mechanical stretch produces long-term effects in developing ASM. While acute low-level stretch does not immediately impact cell proliferation or IL-6 production, it increases fASM proliferation and IL-6 production later in post-stretch period. Our findings suggest that even brief exposure to low-level stretch can have detrimental and persistent consequences on ASM functions that may contribute to long-term airway remodeling and hyperresponsiveness in preterm infants receiving CPAP therapy. Our study suggests that the expression of Piezo proteins in fASM is relatively stable under mechanical stretch and Piezo1 activity is not required for stretch-induced IL-6 production. However, the finding that Piezo1 inhibition dysregulates fASM cell proliferation with or without stretch warrants further investigation to understand the role of Piezo proteins in these processes.