The human study was approved by the Ethics Review Committee of Peking University Third Hospital. All study subjects signed an informed consent form. Totally, 24 subjects were recruited and subdivided into a healthy non-smoking (HNS) group, a smoker group and a COPD group. HNS controls were defined as never-smokers having a post-bronchodilator FEV₁/FVC ≥ 0.7. Smokers were defined as having a smoking history of ≥ 10 pack-years with a post-bronchodilator FEV₁/FVC ≥ 0.7. COPD patients were defined as having ≥ 10 pack-years of smoking history, and a post-bronchodilator FEV₁/ FVC < 0.7. All COPD patients recruited were clinically stable. The study subjects had received lung surgery for solitary tumors, and lung tissues with a maximum distance from the tumor were collected by a pathologist.

All the procedures and protocols carried out in the mouse experiments were approved by Animal Care Committee of Peking University Third Hospital. 6–8week old C57BL/6 female mice, weight 22–25 g, were provided by Beijing Vital River Experimental Animal Company (Beijing, China). All the mice were housed in a specific pathogen-free facility with free access to sterilized food and water. These mice were exposed to filtered air or CS (Baisha cigarettes with filter, Hunan, China) with a nose−only smoke exposure system (SG−300; SIBATA, Saitama, Japan), which has been described in our previous study (Zhou et al., 2019; Xiong et al., 2020). Each mouse was exposed to cigarettes or filtered air for two times a day, 50 min each time with 20-min smoke -free intervals, 5 days a week, for 24 weeks consecutively.

Mouse lung tissues were obtained, fixed with 4% paraformaldehyde for 24 h, embedded with paraffin, then cut into sections of 4 μm thickness, followed by H&E staining. The airspace enlargement in mice was quantified by the mean linear intercept (MLI) and destruction of alveolar walls in mice was quantified by destructive index (DI), which has been described in our previous studies (Zhou et al., 2019; Xiong et al., 2020).

Lung tissues were cut into 5 μm sections, incubated in 0.3% H₂O₂-CH₃OH for 15 min for blocking endogenous peroxidase activity, then treated with citrate buffer (pH 6.0) using a microwave oven for 15 min to retrieve antigen, and followed by blocking with 5% bovine serum albumin for 30 min at room temperature (RT). Subsequently, the tissues were incubated overnight with antibodies against TRPV4 (1:100, Alomone labs, Jerusalem, Israel), human gasdermin D (GSDMD) (1:200, Abcam, Cambridge, United Kingdom), mouse GSDMD (1:1,500, Bioss Biotechnology Co., Ltd, Beijing, China), human gasdermin D N-terminal fragment (GSDMD-N) (1:400, Abcam), and mouse gasdermin D C-terminal fragment (GSDMD-C) (1:400, Abcam). The tissues were then incubated with horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG (ZSGB-Bio, Beijing, China) at 37°C for 30 min. Finally, slides were visualized by staining with a 3,3′-Diaminobenzidine (DAB) detection system kit (ZSGB-Bio). Images were photographed by a microscope and analyzed by Image-Pro Plus 6.0 software (Media Cybernetics, MD, United States).

Human bronchial epithelial cells (16HBEs) were purchased from Bai Ye Biotechnology Center (Shanghai, China). Cells were cultured in Roswell Park Memorial Institute medium supplemented with 10% fetal calf serum (Gibco, Grand Island, NY, United States), 1% penicillin/streptomycin (Gibco, Grand Island, NY, United States) at 37°C with 5% CO₂. 16HBEs were pretreated with 10 μM of a TRPV4 antagonist [GSK205 (catalog number HY120691A, MCE, Monmouth Junction, NJ, United States)] for 1 h and then stimulated with medium only or with CSE for another 24 h.

For transient transfection, cells were cultured in six-well culture plates overnight prior to transfection. Short interfering (si) RNA or control oligonucleotide (GenePharma, Jiang Su, China) were incubated with 4 μl of JetPRIME^® in 200 μl of jetPRIME buffer (Polyplus-transfection) for 15 min at room temperature. A final concentration of 50 nM of TRPV4 siRNA or control oligonucleotide were added to cell culture plates. After 24 h, CSE was added and stimulated for another 24 h. Finally, cells were harvested for subsequent experiments.

CSE was prepared according to our previous study (Zhou et al., 2019). Briefly, five cigarettes (Baisha, China Tobacco Industry Co., Ltd., Hunan, China) were bubbled through 10 ml medium at a constant velocity (Hyclone, Logan, UT, United States). The solution was filtered through a 0.22 μm filter, which was served as the 100% CSE work solution.

Cell viability was detected by the cell counting Kit-8 (CCK8) assay kit (KeyGEN Biotechnology Co., Ltd., Jiangsu, China). In brief, Cells were treated with medium or different concentrations of CSE or GSK205 for 24 h, then washed with PBS for three times and incubated with CCK8 reagent for another 2 h. Subsequently, the absorbance was assessed by spectrophotometer at 450 nm.

To assess the protein expression levels of TRPV4, NLRP3, Caspase-1, GSDMD, GSDMD-N, GSDMD-C, dynamin-related protein 1 (DRP1), mitochondrial fission factor (MFF), OPA1 mitochondrial dynamin like GTPase (OPA1), and mitofusin 2 (MFN2), lung tissue or 16HBE cells were lysed and equal amount of protein was subjected to 8 or 10% SDS-PAGE, subsequently, protein was transferred to 0.22 μm PVDF membranes (Merck-Millipore, Carrigtwohill, Ireland). The membranes were blocked with 5% non-fat milk powder and then incubated with antibodies overnight at 4°C: GAPDH, β-actin, DRP1, MFF, OPA1, MFN2 (1:1,000, Cell Signaling Technology, Danvers, MA, United States), Caspase-1, NLRP3 (1:1,000, RD system, Minneapolis, MN, United States), GSDMD, GSDMD-N, and GSDMD-C (1:1,000, Abcam, Cambridge, United Kingdom).

Membranes were then washed and incubated with HRP-conjugated donkey anti-mouse IgG or antirabbit IgG antibody for 1 h at RT. Finally, membranes were visualized with enhanced chemiluminescence (MerckMillipore, Carrigtwohill, Ireland). Quantitative images were analyzed using Image J software.

Gene expressions of TRPV4, IL1B, IL8, IL18, NQO1 [NAD(P)H quinone dehydrogenase 1], MNSOD [superoxide dismutase 2 (mitochondrial), and CAT (catalase)] were determined by qRT-PCR. RNA was isolated by TRIzol reagent (Thermo Fisher Scientific, MA, United States) and the concentration was determined by Nano Drop 2000. Then RNA was reverse transcribed into cDNA using HiScript III RT Supermix for qPCR(+gDNA wiper) Kit (Vazyme, Nanjing, China). The qPCR reactions were performed on the Applied Biosystems^® QuantStudio^® 5 in a 20 μl reaction system by using ChamQ Universal SYBR qPCR Master Mix Kit (Vazyme, Nanjing, China). The primer sequences were: human TRPV4: 5′-GGCTTTTCCTCTCTCCTCCC-3′ (forward) and 5′-AGGGTGGACTCCAGCAGAT-3′ (reverse); mouse TRPV4: 5′-TCCTGAGGCCGAGAAGTACA-3′ (forward) and 5′-TCCC CCTCAAACAGATTGGC-3′ (reverse); human GAPDH/Gapdh: 5′-AAATCAAGTGGGGCGATGCTG-3′ (forward) and 5′-GCA GAGATGATGACCCTTTTG-3′(reverse); mouse GAPDH/ Gapdh: 5′-AAATGGTGAAGGTCGGTGTGAAC-3′ (forward) and 5′-CAACAATCTCCACTTTGCCACTG-3′(reverse); IL-1β: 5′-ACTGAGAGTGATTGAGAGTGGAC-3′ (forward) and 5′-AACCCTCTGCACCCAGTTTTC-3′ (reverse); IL-8: 5′-AT GATGGCTTATTACAGTGGCAA-3′ (forward) 5′-GTCGGAGA TTCGTAGCTGGA-3′ (reverse); IL-18: 5′-TCTTCATTGACC AAGGAAATCGG-3′ (forward) and 5′-TCCGGGGTGCATTAT CTCTAC-3′ (reverse); NQO1: 5′-CAGTGGCATGCACCCA GGGAA-3′ (forward) and 5′-GCATGCCCCTTTTAGCCTTG GCA-3′ (reverse); MNSOD: 5′-ACAGGCCTTATTCCACTGCT-3′ (forward) and 5′-CAGCATAACGATCGTGGTTT-3′ (reverse); and CAT: 5′-TAAGACTGACCAGGGCATC-3′ (forward) and 5′-CAAACCTTGGTGAGATCGAA-3′ (reverse). The real-time PCR conditions were: 95°C for 15 min, followed by 40 cycles of 95°C for 10 s and 60°C for 32 s. Relative gene expression to control was calculated with the 2^{–Δ Δ CT} method. Gene expression was calculated relative to that of GAPDH/Gapdh in triplicates.

Calbryte 520 AM (AAT Bioquest, San Francisco, CA, United States) was used to assess the intracellular Ca²⁺ levels. In brief, cells were loaded with 2 mM Calbryte 520 AM at 37°C in the dark for 1 h, and then washed with PBS. Thereafter, cells were stimulated with CSE for 20 min. The intracellular Ca²⁺ influx was recorded by a confocal laser scanning microscope for 20 consecutive minutes, at 5 s per frame.

2′,7′-dichlorofluorescin diacetate (DCFH-DA) (Beyotime Biotechnology Co., Ltd., Shanghai, China) was used to assess the intracellular ROS with fluorescent microscopy. In brief, cells were pretreated with GSK205 10 μM for 1 h, and then stimulated with medium only or CSE for another 24 h. These cells then underwent incubation with 10 μM DCFH-DA for 20 min at 37°C. Finally, the intracellular ROS was assessed at 488/525 nm using fluorescent microscopy (Leica, Germany) and analyzed by Image-Pro Plus 6.0 software (Media Cybernetics, MD, United States).

Mitochondrial ROS was assessed by MioSox Red (Invitrogen, Life Technologies, Carlsbad, CA, United States). Briefly, after corresponding treatment, cells were stained with 5 μM Mio Sox Red in the dark for 10 min. The mitochondrial ROS was determined at 510/580 nm using fluorescent microscopy (Leica, Germany) and analyzed by Image-Pro Plus 6.0 software (Media Cybernetics, MD, United States).

LDH was assessed using an LDH assay kit (KeyGEN Biotechnology Co., Ltd). Briefly, cells were handled as described above, then the supernatants were harvested and assessed for LDH release with a microplate reader at 490 nm according to the manufacturer’s instruction.

Hoechst 33342/PI staining was measured using apoptosis and necrosis Assay Kit (Beyotime Biotechnology Co., Ltd., Shanghai, China). After corresponding treatments described above, cells were stained with 5 μl Hoechst 33342 and 5 μl PI in 1 ml staining buffer solution on ice in the dark for 20 min. Then, PI positive cells were photographed by confocal microscopy.

MMP was assessed with Mitotracker deep red (Invitrogen, Life Technologies). After corresponding treatment, cells were incubated with 50 nM Mitotracker deep red for 30 min at 37°C, and then detected at 644/655 nm by flow cytometry.

After corresponding treatment, mitochondria were labeled with the MitoTracker Green (Beyotime Biotechnology Co., Ltd., Shanghai, China) for 30 min in the dark. The mitochondrial morphology was photographed by a confocal microscope (Leica SP5, Germany). MitoTracker Green fluorescence intensity, mitochondrial Aspect Ratio and Form Factor was analyzed by Image J software. A minimum of 20 mitochondria were analyzed.

After corresponding treatment, cells were collected and washed with PBS for subsequent transmission electron microscopic (TEM) detection. Firstly, cells were added with Glutaraldehyde (2.5%) carefully and fixed at 4°C for 1.5 h. Then, the mixture was washed three times with sucrose buffer, fixed, dehydrated, soaked, and embedded, ultrathin sectioned and heavy metal stained. The mitochondrial morphology was assessed by JEM1400 PLUS. Mitochondrial Aspect Ratio and Form Factor in mitochondrial were calculated using Image J software. A minimum of 30 mitochondria were analyzed.

All data were shown as mean ± standard deviation (SD). Statistical analysis was processed with GraphPad Prism 7. Students’ t-tests or one-way ANOVA with Bonferroni post hoc test (equal variances assumed) or Dunnett’s T3 post hoc test (equal variances not assumed) were used when applicable. P-values less than 0.05 were considered statistically significant.

Eight COPD patients, 8 smokers, and 8 HNS controls were enrolled. Table 1 shows the clinical data of the subjects.

A previous in vitro study has implicated pyroptosis in the pathogenesis of COPD (Zhang et al., 2021), however, the expression of pyroptosis in human and mouse COPD were still unknown. Therefore, we investigated the status of pyroptosis of AECs in patients with COPD, smokers and HNS, by performing immunohistochemical staining of lung sections. Patients with COPD showed reduced protein level of GSDMD (Figures 1A–D), but increased protein level of GSDMD-N (Figures 1E–H) in AECs, as compared to smokers and HNS. Moreover, Western blotting analysis revealed decreased GSDMD but increased GSDMD-N protein levels in lung tissue homogenates from COPD patients as compared to smokers and HNS (Figures 1I–K), indicating enhanced pyroptosis in patients with COPD.

We also performed immunohistochemistry for the detection of pyroptosis of AECs in lung sections from mice with CS-induced COPD. Mice in the 6-month CS-exposed group showed enlargement of airway spaces (Figures 2A,B), with significant increase in MLI and DI (Figures 2C,D), which is in line with typical changes of COPD. Consistent with the observations in human subjects, CS-exposed mice showed reduced protein level of GSDMD (Figures 2E–G), but increased protein level of GSDMD-C (a pyroptosis marker in mice) in AECs, as compared to air-exposed mice (Figures 2H–J). Control antibody staining was illustrated in Supplementary Figures 1, 2. Western blotting analysis also demonstrated reduced GSDMD but increased GSDMD-C protein levels in lung tissue homogenates from CS-exposed mice as compared to air-exposed mice (Figures 2K–M). Collectively, these data demonstrated that there was a significant increase in pyroptosis of AECs in human and mouse COPD.

A previous study observed increased TRPV4 mRNA in lung homogenates from patients with COPD in relative to HNS and smokers (Baxter et al., 2014). To further evaluate the protein expression of TRPV4 in COPD, we carried out immunohistochemistry on lung sections from COPD patients. Immunohistochemistry analysis detected increased TRPV4 protein expression in AECs from COPD patients and smokers compared to HNS (Figures 3A–C,F). Similarly, TRPV4 expression was upregulated in AECs from CS-exposed mice as compared to the air-exposed ones (Figures 3D,E,G). CS-exposed mice also showed higher TRPV4 mRNA levels in lung tissues, as compared to the air-exposed controls (Figure 3H). Furthermore, Western blotting analysis observed increased TRPV4 protein level in lung homogenates from CS-exposed mice (Figures 3I,J). Taken together, these results demonstrated that TRPV4 was upregulated in AECs from COPD patients and long-term CS-exposed mice.

Emerging evidence indicates that CS exposure can induce pyroptosis of AECs, suggesting a critical role of pyroptosis in the pathogenesis of COPD (Zhang et al., 2021), but the molecules mediating CS-induced pyroptosis was not clear. To determine whether TRPV4 mediates pyroptosis in CS-exposed 16HBEs, we performed pharmacological inhibition or knockdown of TRPV4 function in 16HBEs before stimulation with CSE. TRPV4 siRNA reduced TRPV4 mRNA level by 90% and protein level by 70% (Supplementary Figures 3A–C). To determine the optimal dose of CSE for in vitro experiments, we firstly assessed the cellular toxicity of CSE at concentrations ranging from 2 to 10%. After stimulation with 4% CSE for 24 h, about 80% cells survived (Supplementary Figure 4). Accordingly, 4% CSE exposure for 24 h was applied in following experiments. In addition, TRPV4 inhibitor showed no cellular toxicity at a range from 100 nM to 100 μM (Supplementary Figure 5). After incubation with CSE for 24 h, 16HBEs were harvested for Western blotting and qRT-PCR. We found that CSE induced significant increase in TRPV4 expression at the mRNA level and the protein level (Figures 4A–C), indicating that CS promoted TRPV4 expression, which was consistent with our observations in humans and mice.

As a previous study indicated that blocking Ca²⁺ influx could effectively inhibit pyroptosis (Wang et al., 2020), we thus measured Ca²⁺ influx in 16HBEs after corresponding treatments. We noted a time-dependent Ca²⁺ influx in 16HBEs induced by CSE, while pharmacological inhibition with GSK205 significantly attenuated this Ca²⁺ influx (Figure 4D). To further confirm this observation, we carried out in vitro TRPV4 siRNA experiments, which demonstrated that TRPV4 gene knockdown significantly blocked Ca²⁺ influx induced by CSE (Supplementary Figure 3D). To further explore whether TRPV4 mediates pyroptosis in CS-exposed 16HBEs, we studied the Ca²⁺/NLRP3/caspase-1/GSDMD pathway. We found that CSE induced significant NLRP3 inflammasome activation and pyroptosis in 16HBE cells, as demonstrated by elevated protein levels of NLRP3 (Figures 4F,H), Cleavage Caspase-1 (Figures 4F,J), and GSDMD-N (Figures 4G,L), and decreased levels of Pro-caspase-1 (Figures 4F,I) and total GSDMD (Figures 4G,K). Interestingly, upon using pharmacological inhibition with GSK205, we observed attenuated NLRP3 inflammasome activation and pyroptosis induced by CSE, as demonstrated by decreased protein levels of NLRP3 (Figures 4F,H), Cleavage Caspase-1 (Figures 4F,J), and GSDMD-N (Figures 4G,L), and increased levels of Pro-caspase-1 (Figures 4F,I) and total GSDMD (Figures 4G,K). These results were confirmed by in vitro TRPV4 siRNA experiments. CSE-exposed TRPV4 siRNA 16HBEs showed decreased protein levels of NLRP3, Cleavage Caspase-1 and GSDMD-N, but increased levels of Pro-caspase-1 and total GSDMD compared with CSE-exposed wild type 16HBEs (Supplementary Figures 3F–L). As LDH release was also a marker of pyroptosis (Wu et al., 2018), we detected LDH release in supernants of 16HBEs. Our results showed that, while CSE promoted the release of LDH, pharmacological inhibition with GSK205 or TRPV4 gene knockdown attenuated CSE-induced LDH production (Figure 4E and Supplementary Figure 3E). Moreover, we found that CSE could markedly increased PI-positive cells, which could be rescued by pharmacological inhibition with GSK205 (Figures 4M–P) or TRPV4 gene knockdown (Supplementary Figures 3M–Q). Taken together, these results indicate that TRPV4 mediates CSE-induced pyroptosis via the Ca²⁺NLRP3/Caspase1/GSDMD pathway.

As CS-induced injury to AECs triggers production and releasing of cytokines, which lead to recruitment of macrophages and neutrophils, participating in the development and progression of COPD (Brusselle et al., 2011; Tuder and Petrache, 2012; Barnes, 2013), we explored whether TRPV4 was possibly involved in this process. Our results showed that pharmacological inhibition of TRPV4 with GSK205 abrogated CSE-induced gene expressions of IL-8, IL-1β, and IL-18 (Figures 5A–C) while attenuated the downregulation of anti-oxidant genes NQO1, MNSOD, and CAT induced by CSE (Figures 5D–F). These findings were confirmed in CSE-exposed TRPV4 siRNA 16HBEs, by showing decreased IL-8, IL-1β, and IL-18 gene expressions but increased NQO1, MNSOD, and CAT gene expressions, compared with CSE-exposed wild type 16HBEs (Supplementary Figures 6A–F). Taken together, these data demonstrate that TRPV4 mediates CSE-induced inflammatory gene upregulation and anti-oxidant gene downregulation.

As ROS and mitochondrial ROS are involved in the assembly of NLRP3 (Abderrazak et al., 2015; Zhang et al., 2021), we thus measured the intracellular and mitochondrial ROS. Exposure to CSE for 24 h significantly elevated intracellular ROS in 16HBEs, while pharmacological inhibition with GSK205 attenuated the effect of CSE (Figures 6A–D,I). Similar results were observed for the mitochondrial ROS level in these cells (Figures 6E–H,J). Additionally, this observation was further confirmed by TRPV4 siRNA experiments with 16HBEs (Supplementary Figures 7A–L).

Mitochondrial dysfunction is known to drive the inflammation associated with COPD (Wiegman et al., 2015). We detected the mitochondrial morphology and activity of 16HBEs after stimulation with CSE. For the detection of mitochondrial morphology, mitochondria were labeled with MitoTracker Green. The mitochondria in CSE-stimulated cells showed decreased relative fluorescence intensity, which could be rescued by pharmacological inhibition with GSK205 (Figures 7A–E) or by TRPV4 siRNA (Supplementary Figures 8A–F). Moreover, mitochondrial morphology was quantified using Aspect Ratio (length, parameter of mitochondrial fusion) and Form Factor (degree of branching, parameter of mitochondrial networking). The result demonstrated that CSE exposure also caused significant reduction of Aspect Ratio and Form Factor, which could be rescued by pharmacological inhibition with GSK205 (Figures 7F,G), or similarly by TRPV4 siRNA (Supplementary Figures 8G,H). Furthermore, ultrastructural analyses by transmission electron microscopy (TEM) observed a significant accumulation of damaged mitochondria with swollen and disrupted cristae in CSE-stimulated cells, which was mitigated by pharmacological inhibition with GSK205 (Figures 7H–K), as quantified by Aspect Ratio and Form Factor, parameters that reflect the complexity of mitochondrial morphology (Figures 7L,M). Likewise, TRPV4 siRNA 16HBEs showed alleviated mitochondrial damage compared to wild type 16HBEs (Supplementary Figures 8I–O). For the detection of mitochondrial activity, Western blotting analysis was carried out, which showed that CSE exposure for 24 h suppressed the expression of mitochondrial fusion protein OPA1 and MFN2, while pharmacological inhibition with GSK205 attenuated this suppression (Figures 7N–P). Moreover, exposure to CSE for 24 h enhanced the expression of mitochondrial fission proteins DRP1 and MFF, which could be abrogated by pharmacological inhibition with GSK205 (Figures 7Q–S). Compared to CSE-exposed wild type 16HBEs, TRPV4 siRNA 16HBEs showed increased protein levels of OPA1 and MFN2 while decreased DRP1 and MFF (Supplementary Figures 8P–U). Additionally, flow cytometry analysis showed decreased MMP after CSE stimulation, while pharmacological inhibition with GSK205 (Figures 7T,U) or TRPV4 siRNA (Supplementary Figures 8V,W) attenuated this decline. In summary, these data indicate that TRPV4 mediates CSE-induced damage of mitochondrial fitness.