Microplastics exacerbate ferroptosis via mitochondrial reactive oxygen species-mediated autophagy in chronic obstructive pulmonary disease

ABSTRACT

Microplastics (MPs) induce mitochondrial dysfunction and iron accumulation, contributing to mitochondrial macroautophagy/autophagy and ferroptosis, which has increased susceptibility to the exacerbation of chronic obstructive pulmonary disease (COPD); however, the underlying mechanism remains unclear. We demonstrated that MPs intensified inflammation in COPD by enhancing autophagy-dependent ferroptosis (ADF) in vitro and in vivo. In the lung tissues of patients with COPD, the concentrations of MPs, especially polystyrene microplastics (PS-MPs), were significantly higher than that of the control group, as detected by pyrolysis gas chromatography mass spectrometry (Py-GCMS), with increased iron accumulation. The exposure to PS-MPs, 2 μm in size, resulted in their being deposited in the lungs of COPD model mice detected by optical in vivo imaging, and observed in bronchial epithelial cells traced by GFP-labeled PS-MPs. There were mitochondrial impairments accompanied by mitochondrial reactive oxygen species (mito-ROS) overproduction and significantly increased levels of lysosome biogenesis and acidification in pDHBE cells with PS-MP stimulation, triggering occurrence of ferritinophagy and enhancing ADF in COPD, which triggered acute exacerbation of COPD (AECOPD). Reestablishing autophagy-dependent ferroptosis via mitochondria-specific ROS scavenging or ferroptosis inhibition alleviated excessive inflammation and ameliorated AECOPD induced by PS-MPs. Collectively, our data initially revealed that MPs exacerbate ferroptosis via mito-ROS-mediated autophagy in COPD, which sheds light on further hazard assessments of MPs on human respiratory health and potential therapeutic agents for patients with COPD.

Abbreviations: ADF: autophagy-dependent ferroptosis; AECOPD: acute exacerbation of chronic obstructive pulmonary disease; Cchord: static compliance; COPD: chronic obstructive pulmonary disease; CQ: chloroquine; CS: cigarette smoke; DEGs: differentially expressed genes; Fer-1: ferrostatin-1; FEV 0.1: forced expiratory volume in first 100 ms; FVC: forced vital capacity; GSH: glutathione; HE: hematoxylin and eosin; IL1B/IL-1β: interleukin 1 beta; IL6: interleukin 6; MDA: malondialdehyde; Mito-ROS: mitochondrial reactive oxygen species; MMA: methyl methacrylate; MMF: maximal mid-expiratory flow curve; MMP: mitochondrial membrane potential; MOI: multiplicity of infection; MPs: microplastics; MV: minute volume; PA: polyamide; PBS: phosphate-buffered saline; PC: polycarbonate; pDHBE: primary human bronchial epithelial cell from COPD patients; PET: polyethylene terephthalate; PIF: peak inspiratory flow; PLA: polylactic acid; pNHBE: primary normal human bronchial epithelial cell; PS-MPs: polystyrene microplastics; PVA: polyvinyl acetate; PVC: polyvinyl chloride; Py-GCMS: pyrolysis gas chromatography mass spectrometry; SEM: scanning electron microscopy; Te: expiratory times; Ti: inspiratory times; TNF/TNF-α: tumor necrosis factor.

Introduction

Currently, plastic products are used widely, and it is estimated that approximately 11 billion tons of plastic waste will be present in the environment by 2025 [1]. Plastics undergo continuous fragmentation and shred into micrometer-scale particles termed microplastics (MPs), which are often detected in water, soil, and air [2]. The ubiquity of MPs (plastic particles <5 mm) in the global biosphere raises increasing concerns about their implications for human health [3]. In the past, the understanding of the harm of MPs has chiefly focused on their environmental impact, but an increasing number of studies have confirmed the presence of MPs in human organs, greatly endangering human life and health [4]. Recent evidence have indicated that MPs can float and be transported in the atmosphere, and plastic fibers are found in human lung tissue [5–8], suggesting that airborne MPs are easily inhalable. However, it is unclear whether airborne MPs pose a substantial risk to human respiratory health.

Chronic obstructive pulmonary disease (COPD) is the third leading cause of death worldwide [9]. Acute exacerbation of COPD (AECOPD) can be precipitated by several factors, of which long-term exposure to airborne particulate matter, including smoke, dust, and MPs, negatively impacts health status, rates of hospitalization and readmission, and disease progression [10]. MPs affect the inflammation, metabolic response, and death mode of cells through contact and deposit with respiratory mucosa and lung cells [11,12], especially in the pathological conditions of airway inflammatory diseases such as COPD, which involve the destruction of airway and alveolar walls. MPs, of which the predominant type is often polystyrene (PS-MPs), can induce oxidative stress, impair mitochondrial function, and abnormal iron accumulation in human lung epithelial cells to induce cell death [13] that have been linked with increased risk of AECOPD [12,14]. These provide ample incentive to explore more information on the potential risk of PS-MPs; however, there is still a lack of research on the mechanism of PS-MPs in AECOPD.

Autophagy is a highly integrated process that maintains cell homeostasis, whereas dysregulated autophagy in diseased states is related to various cell death modalities [15,16]. Autophagy-dependent cell death is used to functionally define a type of cell death that mechanistically relies on the autophagic machinery or its components [17]. Recent studies have shown that imbalanced oxidative stress induces dysregulated autophagy, accompanied by the accumulation of ions in cells [18,19], leading to lipid peroxidation and rupture of cell membranes, which is a newly discovered mode of cell death termed autophagy-dependent ferroptosis (ADF) [20,21]. As a form of regulated cell death, ADF also leads to overwhelming accumulation and release of inflammatory factors in the lung [21], which plays an important role in the development of acute exacerbation of various respiratory diseases, such as COPD and asthma. A previous study indicated a significant increase in MPs particle deposition in the airway epithelium in patients with asthma [22]. In addition, an increasing number of studies have demonstrated that MPs significantly affect the production of mitochondrial reactive oxygen species (mito-ROS), iron metabolism and transport, and the antioxidant defense system [23–25]. As the initiator and amplifier of ferroptosis, mito-ROS plays an important role in driving the Fenton reaction and lipid peroxidation, and its large accumulation is closely related to the occurrence of ADF [26–28]. Our previous research revealed mitochondrial dysfunction and a significantly increased level of mito-ROS in the progression of COPD [29]. However, whether the accumulation of mito-ROS triggers ADF in acute exacerbation COPD induced by MPs exposure requires further investigation.

Given the lack of crucial data on the hazard of MPs exposure in COPD, we designed this study to explore the differential distribution and content of MPs in the lung tissues of patients with COPD and discern the mechanisms by which the PS-MPs induce mitochondrial dysfunction and autophagy-dependent ferroptosis in vitro and in vivo, to provide new evidence on the potential health risks of environment-related MPs exposure and further therapeutic strategies for poor prognosis caused by inhaled MPs in patients with COPD.

Results

Significantly increased levels of MPs were deposited in the lung tissues of patients with COPD and in model mice

To explore whether airborne MPs pose a substantial risk to human respiratory health, the concentrations of MPs, including polystyrene (PS), fabric, polyvinyl alcohol (PVA), polycarbonate (PC), polyethylene terephthalate (PET), polylactic acid (PLA), methyl methacrylate (MMA), and polyamide (PA), in the lung tissues of patients with COPD and the control group were detected by Py-GCMS. Specific information about the lung tissue samples and the content of MPs detected in the samples is summarized in Table S1. The concentration and percentages of PS, fabric, PVA, PC, PET, PLA, MMA, and PA in each sample are presented in Figure 1A,B. The results showed that the concentration and category of MPs in the lung tissues of patients with COPD were significantly higher than those of the control group calculated based on the weight in Figure 1C, of which the levels of PS-MPs were significantly increased with the severity of COPD (Figure 1D). And shown in Figure 1E, the levels of iron content were significantly increased in lung tissue of COPD patients compared to the control group. The shape and size of PS-MPs used in the study were then assessed by SEM. Both 500 nm and 2-μm PS-MPs aggregated slightly into larger complexes in ddH2O and were characterized as spherical particles with a rough surface, as shown in Figure 1F. To determine the size of PS-MPs that can be inhaled and deposited in the respiratory system, PS-MPs-GFP were constructed (Figure 1G), and their aggregation and distribution were assessed by optical in vivo imaging in COPD mice after intranasal administration of 40 mg/kg PS-MPs per day continuously for 21 days (Figure 1H). The results proved that the exposure of PS-MPs, 2 μm in size was significantly deposited in the lungs of the COPD mouse model, as shown in Figure 1I.

Figure 1.

Significantly increased microplastics were deposited in the lung tissues of patients with COPD and model mice. (A) The components and contents of microplastics in the lung tissues of the control group and patients with COPD were detected by py-gcms. (B) The percentages of detected microplastics in the lung tissues of the control group and patients with COPD. (C) The concentrations of total microplastics in the lung tissues of patients with COPD compared to the control group were detected by py-gcms. (D) The concentrations of PS-MPs in the lung tissues of mild and moderate, severe and very severe COPD groups were compared to the control group. (E) The levels of iron in the lung tissue lysis buffer of control and COPD groups are shown. (F) The features of PS-MPs (500 nm and 2 μm size) were observed by scanning electron microscope. (G) Chemical formula of PS-MPs. (H) Protocol for the establishment of COPD mice with or without 500 nm or 2 μm PS-MPs exposure. (I) The aggregation and distribution of PS-MPs-gfp were assessed by optical in vivo imaging. **p < 0.01 and ***p < 0.001.

PS-MPs induced both ferroptosis and autophagy in a concentration-dependent manner

Considering the potential toxic impacts of PS-MPs on the respiratory system, primary bronchial epithelial (pHBE) cells were used to evaluate the potential toxicity after exposure to PS-MPs. The cytotoxicity of PS-MPs (0–300 μg/ml, 0–48 h) for pHBE cells was assessed by the Cell Counting Kit-8 (Figure S1). The results showed that there was significant cell death from the concentration of 200 μg/ml and time of 24 h with PS-MPs stimulation in pDHBE cells. As detected by flow cytometry, the levels of mito-ROS continued to rise from 6 to 24 h shown in Figure 2A,B. As depicted in Figure 2C,D, the significantly increased levels of mito-ROS in pDHBE cells were early from 50 μg/ml and highest in 200 μg/ml concentration of PS-MPs stimulation after 24 h. Mito-ROS was closely associated with ferroptosis and autophagy, and related key proteins were detected. The results of western blot analysis showed that the expression level of ACSL4 was significantly increased and levels of GPX4 and ferritin were significantly decreased, while the expression level of LC3-II:I was significantly increased and the level of SQSTM1/p62 was significantly decreased with increased concentration of PS-MPs stimulations, especially in 200 μg/ml (Figure 2E–J), suggesting that PS-MPs induced both ferroptosis and autophagy in a dose-dependent manner. The increased number of LC3B dots, representing the accumulation of autophagosomes, induced by PS-MPs in pDHBE with AdPlus-mCherry-GFP-LC3B infection is shown in Figure 2K.

Figure 2.

Microplastics induced ferroptosis and autophagy in a dose-dependent manner. The pNHBE and pDHBE cells were cultured and treated with 0, 50, 100, 150, or 200 μg/ml PS-MPs for 0–48 h. (A-B) The levels of mitochondrial ROS induced by PS-MPs with time course were detected using Mito-sox Red staining via flow cytometry. (C-D) The levels of mitochondrial ROS induced by PS-MPs with concentration gradient were detected using Mito-sox Red staining via flow cytometry. The protein expression levels of ACSL4, GPX4, SQSTM1, LC3-II:I, and ferritin in pDHBE cells were measured by western blotting (J) and normalized to ACTB/β-actin (E – I). (K) Cells were transfected with mCherry-GFP-LC3B adenovirus for 24 h and then cultured with fresh medium for 24 h, and the autophagy levels with LC3B dots were detected. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

PS-MPs induce autophagy and ferroptosis accompanied by mitochondrial dysfunction and enhanced lysosome function in pDHBE cells

To explore the potential mechanism of cell toxicity after PS-MPs exposure, we determined the gene‐expression profiles of untreated pDHBE cells compared with profiles of cells treated with PS-MPs using high-throughput sequencing of the transcriptome. There are 694 differentially expressed genes (DEGs) shown in the heat map, of which 261 were significantly upregulated and 433 were downregulated in pDHBE cells with PS-MPs stimulation, including PTGS2, SQSTM1, NCOA4, GPX4, MAP1LC3B and ACLS4 genes (Figure 3A–B). And shown in Figure 3C, the DEGs are enriched in the ferroptosis pathways between pDHBE cells with and without PS-MPs stimulation by KEGG enrichment analysis. Consistently, exacerbated ferroptosis alterations with increased RNA and protein expression levels of ACSL4 and PTGS2 and decreased level of GPX4 were significantly enhanced. And the level of Ferritin was significantly decreased in protein expression but not RNA level in pDHBE cells induced by PS-MPs (Figure 3D–H,K,L). In the meantime, the expression level of LC3B-II:I was significantly increased and that of SQSTM1 was significantly decreased, indicating enhanced autophagy occurrence (Figure 3D,I,J). Next, we estimated the levels of MDA (a lipid peroxidation marker) and reduced GSH (a biomarker of predictable cellular redox homeostasis) using an MDA and GSH assay kit, respectively. As shown in Figure 3P,Q, there were significantly decreased levels of GSH and increased levels of MDA in pDHBE cells induced by PS-MPs.

Figure 3.

PS-MPs induced autophagy and ferroptosis in pDHBE cells. The gene‐expression profiles of pDHBE cells treated with and without PS-MPs using highthroughput sequencing of the transcriptome. (A) The DEGs between pDHBE cells with and without PS-MPs stimulation are shown in the heatmap. (B) The DEGs between pDHBE cells with and without PS-MPs stimulation are shown in the volcano plot. (C) The DEGs were analyzed using KEGG pathway‐enrichment analysis. The protein expression levels of ACSL4, PTGS2, GPX4, ferritin, LC3-II:I and SQSTM1 in pDHBE cells after exposure to 200 μg/ml of PS-MPs were measured by western blotting (D) and normalized to ACTB/β-actin (E – H, J). (I) The protein expressions of LC3-II:I. The mRNA expression levels of PTGS2 (K), GPX4 (L), ACSL4 (M), FTH1 (N) and FTL (O) in pDHBE cells after PS-MPs exposure were measured by qRT-pcr and normalized to ACTB/β-actin. The levels of GSH (P) and MDA (Q) in the cell lysis buffer are shown. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

To explore further mechanisms, 200 μg/ml PS-MPs in 2-μm size with GFP label were adopted to stimulate pDHBE cells stained with MitoTracker were used to track mitochondria. The results revealed that abundant PS-MPs entered the cytoplasm of pDHBE cells after 24 h, accompanied by a decreased mass of mitochondria, as shown in Figure 4A. Electron microscopy revealed that the morphology of both autophagosome and ferroptotic mitochondria characterized by the shrinkage of mitochondria and an increase in mitochondrial intimal density were more significant after PS-MPs exposure (Figure 4B). Moreover, regulatory changes in key molecules involved in mitochondrial dynamics and functions were further verified. The results showed that the expression of DNM1L/DRP1 regulating mitochondria fission was significantly increased, whereas the expression of MFN2 promoting mitochondria fusion was significantly decreased in pDHBE with PS-MPs. That is, PS-MPs induced fragmentation of mitochondria in pDHBE. And the expressions of PINK1 and PRKN were significantly increased, suggesting there was mitochondria damage induced by PS-MPs (Figure 4C-G). The mitochondrial dysfunction feedback loops enhanced activation of mito-ROS, and there was an increase in mito-ROS levels in pDHBE cells stimulated by PS-MPs compared to the control group (Figure 4H). Then the lysosome function was further assessed. There was enhanced lysosome acidification detected by LysoTracker and LysoSensor probes (Figure 4I), and significantly elevated lysosome biogenesis with increased expression levels of LAMP1, LAMP2 and TFEB in pDHBE cells with PS-MP stimulation (Figure 4J-M). Collectively, PS-MPs induce autophagy and ferroptosis along with mitochondrial impairment and enhanced lysosome function in pDHBE cells.

Figure 4.

PS-MPs induced mitochondrial dysfunction and enhanced lysosome function in pDHBE cells (A) pDHBE cells were treated with 200 μg/ml PS-MPs-GFP for 12 or 24 h and then incubated with Mito-Tracker Red. The distribution and content of PS-MPs-GFP, as well as the mitochondrial mass, were detected using fluorescence microscopy. (B) Morphological changes in the mitochondria of pDHBE cells after exposure to 200 μg/ml of PS-MPs were observed by transmission electron microscopy. (C) The protein expressions of PRKN, PINK1, MFN2 and DNM1L for molecular regulation of mitochondrial dynamics and functions of pDHBE cells treated with PS-MPs were measured by western blotting. (D-G) The protein expressions of PRKN and PINK1 normalized to GAPGH, and DNM1L and MFN2 normalized to ACTB/β-actin. (H) The levels of mitochondrial ROS were detected with Mito-sox using flow cytometry. (I) the LysoTracker staining and lysosomal pH measurement. The protein expression levels of LAMP1, LAMP2 and TFEB in pDHBE cells after exposure to 200 μg/ml of PS-MPs were measured by western blotting (J) and normalized to ACTB/β-actin (K-M). *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Mito-ROS is involved in the autophagy and ferroptosis induced by PS-MPs

To confirm the correlation between mito-ROS and ferroptosis induced by PS-MPs, we pretreated pDHBE with the mito-ROS scavenger mito-TEMPO. As shown in Figure 5A, the overproduction of mito-ROS induced by PS-MPs was attenuated by mito-TEMPO. As shown in Figure 5B, the JC-1 fluorescence intensity detected by flow cytometry shifted from red-emitting (PE) aggregates to green-fluorescent (FITC) monomers after PS-MP treatment and was rescued by mito-TEMPO. This indicated that the MMP was conspicuously declined by PS-MPs in pDHBE cells, and mito-TEMPO salvaged the drop in MMP, because the red/green JC-1 fluorescence ratio can be used as a sensitive measure of MMP. The crosstalk between mito-ROS and autophagy has been widely confirmed; however, it is unclear whether the mito-ROS caused by PS-MPs are associated with autophagy. As shown in Figure 5C–E, the increased autophagy biomarker LC3-II:I levels and decreased autophagy substrate SQSTM1 levels caused by PS-MPs were reversed by mito-TEMPO. Simultaneously, the increased accumulation of autophagosome with increased LC3B dots induced by PS-MPs in pDHBE was reduced after mito-TEMPO treatments, as shown in Figure 5F. These findings indicate that the PS-MPs-induced upregulation of autophagy is mediated by mito-ROS. In addition, mito-TEMPO inhibited the increased MDA content and reduced GSH levels induced by PS-MPs (Figure 5G–H). Moreover, the increased expression of ACSL4 and PTGS2 and the decreased level of GPX4 induced by PS-MPs were all reversed by mito-TEMPO (Figure 5I–L). As chemical agonist of mito-ROS, H₂O₂ was applied to cell model to verify the changes of autophagy and ferroptosis, and the effects of H₂O₂ on autophagy and ferroptosis were similar to those induced by PS-MPs (Figure S2). Together, these data indicate that PS-MPs-induced autophagy and cellular ferroptosis are mediated by mito-ROS.

Figure 5.

Mito-ros was involved in mitochondrial dysfunction and ferroptosis induced by microplastics in pDHBE cells. The pDHBE cells were treated with 200 μg/ml PS-MPs and/or 40 μM mitoTEMPO for 24 h. (A) The levels of mitochondrial ROS were detected using MitoSOX red dye via flow cytometry. (B) MMP levels were determined using the JC-1 kit via flow cytometry. (C) The protein expression levels of LC3-II:I and SQSTM1 for molecular regulation of autophagy of pDHBE cells treated with PS-MPs and/or mitoTEMPO were measured by western blotting. (D) The protein expressions of SQSTM1 normalized to ACTB/β-actin. (E) The protein expressions of LC3-II:I. (F) Cells were transfected with mCherry-GFP-LC3B adenovirus for 24 h and then cultured with fresh medium for 24 h, and the autophagy levels with LC3B dots were detected in pDHBE cells treated with PS-MPs and/or mitoTEMPO. The levels of MDA (G) and GSH (H) in the cell lysis buffer are shown. The protein expression levels of GPX4, ACSL4, and PTGS2 in pDHBE cells treated with PS-MPs and/or mitoTEMPO were measured by western blotting (I) and normalized to ACTB/β-actin (J – L). *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Ferroptosis induced by PS-MPs is triggered by mito-ROS-dependent autophagy

To explore the link between mito-ROS-dependent autophagy and ferroptosis, chloroquine (CQ), a pharmacological inhibitor of autophagy, was used in pDHBE to inhibit the degradation of LC3B during autophagy. The cytotoxicity experiment of CQ on pDHBE cells was performed and it found that there was nontoxic at concentrations of 10 μM for 24 h used in the later verification (Figure S3). Indeed, the PS-MPs-induced mito-ROS accumulation was apparently inhibited by CQ (Figure 6A). In PS-MPs-exposed pDHBE cells, upregulation of LC3B with accumulation of autophagosomes was reinforced, and the downregulation of SQSTM1 was reversed by CQ, resulting from CQ restrained the fusion of autophagosomes and autolysosomes (Figure 6B–D,H). Also, the results indicated that lysosome biogenesis and acidification in pDHBE cells induced by PS-MP stimulation were not impaired by CQ (Figure 6E–H). After treatment with CQ, overloaded MDA and impaired GSH levels under PS-MPs exposure were alleviated (Figure 7A–B). Additionally, the CQ inhibitor significantly inhibited PS-MPs-induced ferroptosis in pDHBE. The results demonstrated that the downregulated expression of GPX4 and upregulated expression of ACSL4 and PTGS2 induced by PS-MPs were reversed by CQ (Figure 7C–F). These data demonstrate that mito-ROS-dependent autophagy is an active regulator of PS-MPs-induced ferroptosis in pDHBE.

Figure 6.

Ferroptosis was triggered by mito-ros-dependent autophagy in PS-MPs-treated pDHBE cells. (A) The levels of mitochondrial ROS in pDHBE cells treated with 200 μg/ml PS-MPs and/or 10 μM CQ were detected with MitoSOX red dye using flow cytometry. (B) Cells were transfected with mCherry-GFP-LC3B adenovirus for 24 h and then cultured with fresh medium for 24 h, and the autophagy levels with LC3B dots were detected in pDHBE cells treated with PS-MPs and/or CQ. (C) the protein expressions of LC3-II:I. (D) The protein expressions of SQSTM1 normalized to ACTB/β-actin. (E) the protein expressions of LAMP1 normalized to ACTB/β-actin. (F) the protein expressions of LAMP2 normalized to ACTB/β-actin. (G) the protein expressions of TFEB normalized to ACTB/β-actin. (H) the protein expression levels of LC3-II:I, SQSTM1, LAMP1, LAMP2 and TFEB in pDHBE cells treated with PS-MPs and/or CQ were measured by western blotting. (I) the LysoTracker staining and lysosomal pH measurement in pDHBE cells treated with PS-MPs and/or CQ. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Figure 7.

Ferritinophagy triggered ferroptosis by regulating iron homeostasis in PS-MPs-treated pDHBE cells. The levels of MDA (A) and GSH (B) in the cell lysis buffer were shown. The protein expression levels of GPX4, ACSL4, and PTGS2 in pDHBE cells treated with PS-MPs and/or CQ were measured by western blotting (F) and normalized to ACTB/β-actin (C – E). (G) The levels of iron in the cell lysis buffer were shown. The protein expression levels of ferritin and NCOA4 in pDHBE cells treated with PS-MPs and/or CQ were measured by western blotting and normalized to ACTB/β-actin (H – J). (K) The protein expressions of TOMM20 and LC3B were shown using immunofluorescence staining. (L) The protein expressions of ferritin and LC3B were shown using immunofluorescence staining. (M) The protein expressions of NCOA4 and SQSTM1 were shown using immunofluorescence staining. (N) The changes of protein-protein interactions between NCOA4 and ferritin, LC3B in pDHBE with PS-MP stimulation were detected by co-ip. (O-P) The change of half-life of ferritin protein induced by PS-MP in pDHBE cells was measured. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Ferritinophagy triggers ferroptosis by regulating iron homeostasis in PS-MPs-exposed pDHBE cells

To investigate the mechanism by which autophagy regulates ferroptosis under PS-MPs stimulation, the role of PS-MPs-induced autophagy in the expression of cellular iron content was explored. The augmented iron content induced by PS-MPs was obviously decreased by CQ treatment (Figure 7G). The cellular free iron concentration is closely involved in the regulation of ferroptosis, and it is mainly controlled by ferritin. NCOA4 (nuclear receptor coactivator 4) is an autophagy cargo receptor that mediates the autophagic degradation of ferritin. Therefore, the expressions of ferritin and NCOA4 were measured. As shown in Figure 7H–J, the inhibition of ferritin protein and the augmentation of NCOA4 protein by PS-MPs were reversed by CQ. The above results indicate that iron homeostasis is regulated by autophagy to trigger ferroptosis in PS-MPs-exposed pDHBE. To distinguish macroautophagy, mitophagy and ferritinophagy, the presences of protein colocalization were detected by immunofluorescence assay, and the results indicated that NCOA4, but not mitochondrial marker TOMM20 or ferritin, were co-localized with the autophagy marker in pDHBE with PS-MP stimulation (Figure 7K-M). Then the results of co-immunoprecipitation proved that there were enhanced protein-protein interactions between NCOA4 and ferritin, LC3B in pDHBE with PS-MP stimulation (Figure 7N). And there was reduction of the half-life of ferritin and increased levels of iron release induced by PS-MP in pDHBE (Figure 7O-P), suggesting that the degradation efficiency of ferritin increases. Further the stable knockdown of NCOA4 or overexpression of FTH1 were successfully constructed in pDHBE cells (Figure S4). And when pDHBE cells with NCOA4 knockdown are established, ferroptosis induced by MPs was alleviated to a certain extent including changes of key ferroptosis proteins and indicators for lipid peroxidation (Figure 8A,C-J). While FTH1 is overexpressed in pDHBE cells, the protein expression of NCOA4 was increased, and level of ferroptosis increased correspondingly resulting from significantly increased levels of ferritin degradation and iron release (Figure 8B,K-R). Together, these results concluded that ferritinophagy occupies a dominant role in the mechanism of PS-MP-induced- ferroptosis in pDHBE cells, which is mediated by the receptor NCOA4 that specifically bound and marked ferritin as autophagic cargo, and leads increased level of degradation of ferritin and iron release to induce ferroptosis.

Figure 8.

The stable knockdown of NCOA4 or overexpression of FTH1 regulated cellular ferroptosis and iron release in pDHBE cells with PS-MPs stimulation. The pDHBE cells with NCOA4 knockdown or FTH1 overexpression were established. (A) The protein expression levels of ACSL4, NCOA4, ferritin, GPX4 and LC3-II:I in shRNA-nc- or shNCOA4-pDHBE cells after 200 μg/ml PS-MPs exposure were measured by western blotting. (B) The protein expression levels of ACSL4, NCOA4, ferritin, GPX4 and LC3-II:I in OE-NC- or FTH1-pDHBE cells after 200 μg/ml PS-MPs exposure were measured by western blotting. (C) The protein expressions of GPX4 normalized to ACTB/β-actin in shRNA-nc- or NCOA4-pDHBE cells. (D) The protein expressions of NCOA4 normalized to ACTB/β-actin in shRNA-nc- or NCOA4-pDHBE cells. (E) The protein expressions of LC3-II:I in shRNA-nc- or NCOA4-pDHBE cells. (F) The protein expressions of ferritin normalized to ACTB/β-actin in shRNA-nc- or NCOA4-pDHBE cells. (G) The protein expressions of ACLS4 normalized to ACTB/β-actin in shRNA-nc- or NCOA4-pDHBE cells. The levels of MDA (H), GSH (I) and iron content (J) in the cell lysis buffer of shRNA-nc- or NCOA4-pDHBE cells after 200 μg/ml PS-MPs exposure were shown. (K) Tthe protein expressions of GPX4 normalized to ACTB/β-actin in OE-NC- or FTH1-pDHBE cells. (L) The protein expressions of NCOA4 normalized to ACTB/β-actin in OE-NC- or FTH1-pDHBE cells. (M) The protein expressions of LC3-II:I in OE-NC- or FTH1-pDHBE cells. (N) the protein expressions of ferritin normalized to ACTB/β-actin in OE-NC- or FTH1-pDHBE cells. (O) the protein expressions of ACLS4 normalized to ACTB/β-actin in OE-NC- or FTH1-pDHBE cells. The levels of MDA (P), GSH (Q) and iron content (R) in the cell lysis buffer of OE-NC- or FTH1-pDHBE cells after 200 μg/ml PS-MPs exposure were shown. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Fer-1 inhibited MPs-induced cellular ferroptosis in pDHBE cells

To confirm the effect of PS-MPs on ferroptosis regulation, ferrostatin-1 (Fer-1), a ferroptosis inhibitor, was used in the present study. Compared to the PS-MPs group, the level of mito-ROS was significantly reduced in the PS-MPs+Fer-1 group, indicating the rescue effect of Fer-1 on PS-MPs-induced cell mitochondria damage (Figure 9A). In addition, the content of MDA was much lower and GSH levels were enhanced in the PS-MPs+Fer-1 group than in the PS-MPs group of pDHBE cells (Figure 9B–C). The increased mRNA and protein levels of ACSL4 and PTGS2 along with decreased levels of GPX4 in the PS-MPs group were also improved by Fer-1 treatment in vitro (Figure 9D–J). Consistent with these changes, the decreased protein expression of GPX4 induced by PS-MPs was identified in pDHBE cells, and Fer-1 treatment attenuated its effects via immunofluorescence analysis (Figure 9K). Overall, these results suggested the key role of ferroptosis in PS-MPs-induced pDHBE injury from another perspective.

Figure 9.

Fer-1 inhibited cellular ferroptosis induced by microplastics in pDHBE cells. The pDHBE cells were treated with 200 μg/ml PS-MPs and/or 10 μM fer-1 for 24 h. (A) The levels of mitochondrial ROS were detected using MitoSOX red dye via flow cytometry. The levels of MDA (B) and GSH (C) in the cell lysis buffer are shown. The protein expression levels of GPX4, ACSL4, and PTGS2 in pDHBE cells were measured by western blotting (D) and normalized to ACTB/β-actin (E – G). The mRNA expression levels of GPX4 (H), ACSL4 (I), and PTGS2 (J) in pDHBE cells treated with PS-MPs and/or fer-1 were measured by qRT-pcr and normalized to ACTB/β-actin. (K) the protein expression of GPX4 was shown using immunofluorescence staining. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Fer-1 inhibited MPs-induced lung inflammation and ferroptosis in COPD mice

To explore weather PS-MPs induced ferroptosis in vivo, a COPD mouse model with PS-MPs intranasal administration was established, as described in Figure 10A, and the iron content was measured in lung tissues. Compared to the control and PS-MPs groups, the iron content in lung tissue was apparently increased in the COPD+PS-MPs group (Figure 10B). Further, the protein expressions of MFN2, GPX4 and SQSTM1 were significantly downregulated, whereas DNM1L, PTGS2, ACSL4, and LC3-II:I expression were significantly upregulated in PS-MPs-exposed COPD mice compared to the control and PS-MPs groups (Figure 10C–J). Compared to the control and PS-MPs groups, the levels of mito-ROS content was apparently increased and mitochondrial membrane potential was significantly dampened in mouse lung tissue of the COPD+PS-MPs group (Figure 10K-L). These data indicate that PS-MPs induce mitochondrial impairment, autophagy and ferroptosis in COPD mice.

Figure 10.

Microplastics induced both mitochondrial autophagy and ferroptosis in COPD mice. (A) Protocol for the establishment of COPD mice with or without 2 μm PS-MPs challenge and/or synchronous fer-1 or CQ treatment. (B) The levels of iron in the lung tissue lysis buffer of mock and COPD mice with or without PS-MPs challenge are shown. (C) The protein expression levels of PTGS2, GPX4, and ACSL4 for the molecular regulation of ferroptosis, LC3-II:I and SQSTM1 for the molecular regulation of autophagy, and DNM1L and MFN2 for the molecular regulation of mitochondrial dynamics in the lung tissues of mock and COPD mice with or without PS-MPs challenge were measured by western blotting. (D) The protein expression of PTGS2 normalized to ACTB/β-actin. (E) The protein expression of GPX4 normalized to ACTB/β-actin. (F) The protein expression of ACSL4 normalized to ACTB/β-actin. (G) The protein expression of LC3-II:I. (H) The protein expression of SQSTM1 normalized to GAPDH. (I) the protein expression of DNM1L normalized to ACTB/β-actin. (J) The protein expression of MFN2 normalized to ACTB/β-actin. (K) The levels of mitochondrial ROS were detected using MitoSOX red dye by multimode reader. (L) MMP levels were determined using the JC-1 kit by multimode reader. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

The therapeutic effect of Fer-1 against PS-MPs-induced lung inflammation and ferroptosis was further evaluated in vivo using a COPD mouse model. First, the lung functions of mice were detected, and there were significant decreases in respiratory airflow during expiration (volume expired in first 100 ms of fast expiration, FEV 0.1:FVC), maximal mid-expiratory flow curve (MMF), peak inspiratory flow (PIF), and minute volume (MV), and increases in static compliance (Cchord) in the COPD+PS-MPs mice compared to the COPD mice. PS-MPs exposure also increased expiratory times (Te) and decreased inspiratory times (Ti) (Figure 11A–G), indicating the exacerbation of respiratory airflow limitation induced by PS-MPs in COPD mice. However, Fer-1 treatment decreased Cchord, calibrated respiratory times (Ti and Te), and improved airflow limitation reflected by elevated FEV 0.1:FVC, MMF, PIF, and MV (Figure 11A–G). The results suggest that Fer-1 treatment has beneficial effects on lung function in AECOPD mice induced by PS-MPs. Then, the levels of the inflammatory cytokines IL6, IL1B/IL-1β, and TNF/TNF-α in bronchoalveolar lavage fluid (BALF) were measured, and the results indicated that PS-MPs exposure significantly elevated inflammatory cytokines in COPD mice and the relief of the inflammatory response in the PS-MPs+Fer-1 group compared to the PS-MPs group (Figure 11H–J). These results indicate that the ferroptosis inhibitor Fer-1 exerts therapeutic action against PS-MP-induced lung inflammation.

Figure 11.

Fer-1 inhibited lung inflammation and ferroptosis induced by microplastics in COPD mice. Lung function, including volume expired in first 100 ms of fast expiration (FEV 0.1:FVC) (A), maximal mid-expiratory flow curve (MMF) (B), peak inspiratory flow (PIF) (C), static lung compliance (cchord) (D), inspiratory time (ti) (E), expiratory time (te) (F), and minute volume (MV) (G), of COPD mice with or without 2 μm PS-MP (40 mg/kg) challenge and/or synchronous fer-1 (10 mg/kg) treatment. (H – J) levels of cytokines (IL6, IL1B/IL-1β, and TNF/TNF-α) in the BALF of mice. The protein expression levels of PTGS2, GPX4, and ACSL4 in lung tissue were measured by western blotting (M) and normalized to ACTB/β-actin (K – L, N). The levels of MDA (O) and GSH (P) in the lung tissue lysis buffer are shown. (Q) HE staining of lung sections was imaged by a microscope using 10× and 40× objectives. (R) Morphological changes in the mitochondria of the mouse airway epithelium were observed using transmission electron microscopy. (S) Protein expression of GPX4 and ACSL4 in the lung tissues of mice were shown using immunofluorescence staining. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

The ferroptosis level in lung tissue was evaluated to analyze the effects of Fer-1. There was a significantly increased expression of PTGS2 and ACSL4 along with a reduced level of GPX4 in the COPD+PS-MPs mice compared to the COPD mice, suggesting that PS-MPs treatment promoted ferroptosis in lung tissues. Furthermore, these effects were ameliorated by co-treatment with Fer-1 (Figure 11K–N). Similarly, the results showed that overloaded MDA and decreased GSH levels were significant under PS-MPs exposure in COPD mice, which were alleviated by Fer-1 treatment (Figure 11O–P). As shown in HE staining, lung morphology showed local pulmonary emphysema and inflammatory infiltration in COPD mice, whereas PS-MPs exposure resulted in significantly elevated pulmonary leukocyte infiltration and airway wall thickening, which were rescued after Fer-1 treatment (Figure 11Q). As shown in the mitochondrial ultrastructural micrographs, membrane rupture and reduction or disappearance of cristae were observed in the PS-MPs-exposed COPD mice, and these events improved with Fer-1 intervention (Figure 11R). Coincident with the results of western blot analysis, simultaneous upregulation of ACSL4 and downregulation of GPX4 in the airway epithelium of PS-MPs-exposed COPD mice by immunofluorescence staining were reversed by Fer-1 treatment (Figure 11S). Collectively, these results indicated that Fer-1 alleviated PS-MPs-induced excessive lung inflammation in COPD by inhibiting ferroptosis in vivo.

Microplastics induce mito-ROS-autophagy-dependent ferroptosis in acute exacerbation of COPD mice

The therapeutic action of CQ against PS-MPs-induced lung inflammation and ferroptosis was further evaluated in vivo using the COPD mouse model. Importantly, CQ treatment decreased Cchord, calibrated respiratory times (Ti and Te), and improved airflow limitation reflected by elevated FEV 0.1:FVC, MMF, PIF, and MV (Figure 12A–G). The results suggest that CQ treatment has beneficial effects on lung function in AECOPD mice induced by PS-MPs. The autophagy and ferroptosis levels in lung tissue were evaluated to analyze the effect of CQ. There were significantly increased expressions of LC3-II:I, PTGS2, and ACSL4 along with reduced levels of SQSTM1 and GPX4 in the COPD+PS-MPs group than in the COPD group, suggesting that PS-MPs treatment promoted autophagy and ferroptosis in lung tissues. Furthermore, these effects on changes in PTGS2, ACSL4, SQSTM1, and GPX4 were alleviated, and the obstruction of autophagy flux with increased LC3B accumulation was intensified by co-treatment with CQ (Figure 12H–J). Then, the levels of the inflammatory cytokines IL6, IL1B/IL-1β, and TNF/TNF-α in BALF were measured, and the results indicated that PS-MP exposure significantly elevated inflammatory cytokine levels in COPD mice and the relief of the inflammatory response in the PS-MPs+CQ group compared to the PS-MPs group (Figure 12K–M). These results indicate that the ferroptosis inhibitor CQ exerts therapeutic action against PS-MPs-induced lung inflammation. Similarly, the results showed that overloaded MDA and decreased GSH levels were significant under PS-MPs exposure in COPD mice, which were alleviated by CQ treatment (Figure 12N–O). As shown by HE staining, the lung morphology showed local pulmonary emphysema and inflammatory infiltration in COPD mice, whereas PS-MPs exposure resulted in significantly elevated pulmonary leukocyte infiltration and airway wall thickening, which were rescued after CQ treatment (Figure 12P). As shown in the mitochondrial ultrastructural micrographs, membrane rupture and reduction or disappearance of cristae were observed in the PS-MPs-exposed COPD group, and these events improved with CQ intervention (Figure 12Q). Coincident with the results of western blot analysis, simultaneous upregulation of LC3B and downregulation of GPX4 in the airway epithelium of PS-MPs-exposed COPD mice was observed by immunofluorescence staining, and upregulation of LC3B was reinforced and the downregulation of GPX4 was reversed by CQ treatment (Figure 12R). To further explore changes in iron homeostasis, the expression levels of ferritin and NCOA4 were measured. As shown in Figure 12S–T, the inhibition of ferritin protein and the augmentation of NCOA4 protein by PS-MPs were reversed by CQ. The above results indicate that iron homeostasis is regulated by autophagy to trigger ferroptosis in PS-MPs-exposed COPD mice. We concluded that MPs induce mito-ROS-autophagy-dependent ferroptosis in acute exacerbation of COPD (Figure 13).

Figure 12.

Microplastics induce mito-ros-autophagy-dependent ferroptosis in acute exacerbation of COPD mice. Lung function detection, including volume expired in first 100 ms of fast expiration (FEV 0.1:FVC) (A), maximal mid-expiratory flow curve (MMF) (B), inspiratory time (ti) (C), expiratory time (te) (D), peak inspiratory flow (PIF) (E), static lung compliance (cchord) (F) and minute volume (MV) (G), in COPD mice with or without 2 μm PS-MPs (40 mg/kg) challenge and/or synchronous CQ (20 mg/kg) treatment. (H) Protein expression of GPX4 normalized to ACTB/β-actin. (I) the protein expression levels of PTGS2, GPX4, and ACSL4 for molecular regulation of ferroptosis, and LC3-II:I and SQSTM1 for molecular regulation of autophagy in the lung tissues were measured by western blotting. (J) The protein expressions of PTGS2, ACSL4 and SQSTM1 normalized to ACTB/β-actin, and LC3-II:I. (K – M) Levels of cytokines (IL6, IL1B/IL-1β, and TNF/TNF-α) in the BALF of mice. The levels of MDA (N) and GSH (O) in the lung tissue lysis buffer are shown. (P) HE staining of lung sections was imaged by a microscope using 10× and 40× objectives. (Q) The morphological changes in the mitochondria of the mouse airway epithelium were observed by transmission electron microscopy. (R) The protein expression of GPX4 and LC3B in the lung tissues of mice are shown using immunofluorescence staining. The protein expression levels of ferritin and NCOA4 in lung tissues were measured by western blotting (S) and normalized to ACTB/β-actin (T). *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Figure 13.

Microplastics intensify ferroptosis via mitochondria-mediated autophagy in chronic obstructive pulmonary disease.

Discussion

Microplastic pollution has been identified as an escalating global threat to human health [6], with exposure to MPs being intricately linked to the occurrence and mechanisms of ADF [12,28], a novel regulatory pathway for non-apoptotic cell death in various diseases of the respiratory system [30]. Prolonged exposure to airborne environmental pollutants is among the most significant predisposing factors for COPD, contributing to disease progression and poor prognosis [31]; however, crucial data on the hazards of inhaled MPs exposure in the progression of COPD are scarce. This comprehensive study ascertained the differential distribution and content of MPs in the lung tissue of patients with COPD and the mechanisms by which PS-MPs induce mitochondrial impairment and ADF in vitro and in vivo in patients with COPD to provide new evidence regarding the potential health risks of environment-related MPs exposure and further alternative therapeutic strategies to the clinical challenge of inhaled-MPs‐induced the exacerbation of COPD.

Particles with an aerodynamic diameter between 1 and 3 μm are mostly retained in the bronchi and bronchioles, whereas particles less than 1 μm can adhere to alveolar segments or be expelled from the lungs by breathing [32–34]. Furthermore, it reported that fine particulate matter (PM 2.5) is strongly associated with AECOPD by inducing oxidative stress and inflammation reaction [35]. Using Py-GCMS analysis, we illustrated that the concentration of MPs, including polystyrene (PS-MPs), in lung tissue of patients with COPD was significantly higher than that of the control group. Furthermore, the exposure of PS-MPs, 2-μm in size rather than 500-nm PS-MPs was deposited in the lungs of the COPD mouse model and detected in primary COPD bronchial epithelial (pDHBE) cells. These findings indicate that long-term exposure to airborne MPs at the micron scale can enter the body through breathing, accumulating in the lungs and respiratory tract and being involved in the pathogenesis and progression of COPD.

Ferroptosis was primarily found as a form of non-apoptotic cell death dependent on iron, which was grouped into a novel type of regulated cell death [30,36]. It has been reported that ferroptosis is involved in the development of numerous diseases, including acute lung injury, asthma, and COPD [37–39]. So far, few studies have linked ferroptosis to the process of PS-MP-induced acute exacerbation of COPD. In this study, primary DHBE cells were isolated and an animal model of COPD was established after exposure to cigarette smoke for six months was established to emulate the authentic physiological state. The expression changes of ferroptosis-related proteins, including GPX4, ACSL4, and PTGS2, MDA, GSH and iron content, and mitochondrial morphology induced by PS-MPs were measured in both the pDHBE and lung tissues of COPD mice. In general, we found that excessive airway inflammation and the factors that promote ferroptosis were significantly positively induced by PS-MPs in vitro and in vivo in patients with COPD. Thus, PS-MPs exacerbate excessive inflammation in COPD by enhancing ferroptosis, and ferroptosis has great potential as a therapeutic target for MPs-induced AECOPD.

GPX4 and ACSL4 were considered to be important indicators of intensity of ferroptosis. GPX4 is a selenoprotein glutathione peroxidase that can limit the increase of lipid peroxidation [40,41] and its knockout has been proved to associate with severe pathological phenotypes of ferroptosis [42,43]. ACSL4 can make cells prone to ferroptosis by enriching specific oxidation-sensitive fatty acids in the membrane [44,45]. When earthworms were exposed to polyethylene (PE), the intracellular storage of iron and the expression of GPX4 were inhibited, leading to ferroptosis [46]. In our study, the expression of GPX4 was significantly decreased, while the expression of ACSL4 was significantly increased in pDHBE and COPD mice treated with PS-MPs. Hence, GPX4 and ACSL4 are recognized as sensitive regulators and biomarkers of ferroptosis involvement in COPD pathogenesis. Several reports have demonstrated that ferroptosis is induced by the loss of GPX4 activity due to the depletion of intracellular GSH in CSE-treated Beas2b and some cancer cells [47,48]. In line with our findings, we observed a clear reduction in GSH and an increase in MDA content in response to 24-h PS-MPs exposure in pDHBE cells. Furthermore, ferrostatin-1 treatment increased GSH levels and clearly reduced lipid peroxidation with decreased MDA levels, further supporting the notion that not only GPX4 and ACSL4 levels but also loss of GSH activity was a critical determinant of lipid peroxidation and subsequent ferroptosis in COPD.

A great number of studies demonstrated that ferroptosis was regulated by multiple factors, of which mitochondria played a dominant role in this process [49–51]. Morphologically, ferroptosis was characterized by intact nuclear and mitochondrial abnormalities [52,53], which played a prominent part in the cytopathic effect of PS-MPs-induced pDHBE in our results. Our results verified that PS-MPs cause mitochondrial damage and oxidative stress with excessive production of mitochondrial ROS, diminished MMP production, and mitochondrial fragmentation in pDHBE, which is a prerequisite for the occurrence of ferroptosis [54]. In addition, high levels of mito-ROS contributing to mitochondrial lipid peroxidation were along with MDA accumulation and GSH depletion in PS-MPs-induced pDHBE and COPD mice, which were also pivotal for ferroptosis [55]. Using mito-TEMPO to restrain mito-ROS, we found that damaged mitochondria and augmented ferroptosis induced by PS-MPs were alleviative in pDHBE. Therefore, excessive mitochondrial ROS accumulation intimately mediates ferroptosis activation. Meanwhile, our results discovered that PS-MPs exposure destroyed the integrity of mitochondria, with changed expressions of mitochondrial membrane fission and fusion protein, suggesting that detriments of PS-MPs involved mitochondrial respiration function and metabolic pathways in COPD. Contrastively, it had been demonstrated that nano-plastic exposure also caused mitochondrial damage with increased mitochondrial ROS, while they give rise to unobvious cell death [56,57]. Therefore, we should take it into consideration that attenuating mitochondrial damage is crucial for reducing the occurrence of ferroptosis in patients with COPD at risk of MPs exposure.

The interrelationship between autophagy and ferroptosis has been explored and provided a novel concept regarding regulatory cell death at the molecular level [58,59]. Increasing evidence reveals that autophagy contributes to ferroptotic cell death under certain conditions [20,21,60]. In the present study, our results showed that PS-MPs upregulated the levels of LC3B and reduced the levels of SQSTM1 in pDHBE and COPD mice, suggesting that autophagy was activated by PS-MPs in COPD. More importantly, after the application of the autophagy-specific inhibitor CQ, there were not only diminished level of mito-ROS and expression of ferritin, but also the levels of free iron as well as ferroptosis elevated by PS-MPs were significantly reduced. These results indicate that ferroptosis is a form of autophagic cell death that is ferritin dependent. In addition, the level of autophagy was elevated by mito-ROS, which was derived from the mitochondrial damage induced by PS-MPs. Under normal conditions, mito-ROS-induced autophagy can alleviate oxidative stress to protect cells from damage. Autophagy protects cells by removing ROS to conserve mitochondrial integrity, avoid apoptosis, and promote antigen presentation [18,61,62]. However, excessive mito-ROS can cause autophagic cell death [63,64] and amplify ferroptosis [26] resulting in persistent inflammatory response under COPD. Collectively, MPs exacerbate autophagy-dependent ferroptosis by increasing mito-ROS levels, thereby increasing susceptibility to AECOPD.

In the study, CQ was applied to confirm that the enhanced autophagy-lysosomal function lead to ferritinophagy, but it was reported that there may be some negative impacts of CQ for its inhibition effects on autophagy processes lead to the excessive aggregation of proteins and dysfunction in organelles ultimately promoted cell death [65]. So, nonsignificant cytotoxicity of 10 μM CQ on pDHBE cells was verified and the further results indicated that lysosome biogenesis and acidification in pDHBE cells induced by PS-MP stimulation were not impaired by CQ in the concentration of 10 μM. The effects of CQ on cells was dose dependent. Low doses are mainly used for autophagy inhibition and immune regulation, while high doses may lead to lysosomal dysfunction, mitochondrial damage and cell death [66–68]. So, in practical applications, the appropriate dose and time of treatment should be considered according to the specific disease and treatment target to avoid negative impacts.

The mechanism of autophagy-dependent ferroptosis induced by PS-MPs in COPD was further investigated. We investigated the role of autophagy protein and demonstrated that SQSTM1 was significantly downregulated and LC3B was upregulated by mito-ROS generation during ferroptosis induced by PS-MPs in COPD. When changes in autophagy protein were inhibited by CQ, the increased expression of NCOA4, degradation of ferritin, and iron overload were significantly improved. Several researchers have verified that ferritinophagy serves to ferroptosis by inducing iron-dependent ROS overproduction by mediating the degradation of ferritin [69,70]. Intracellular iron is mostly stored in a non-bioavailable composition by binding to ferritin, so ferritin degradation eventually results in the cytosolic release of chelated iron, and the disruption of iron homeostasis elicits damage to DNA, proteins and lipids through the generation of free radicals and oxidative stress, subsequently leading to tissue damage in COPD [38]. Additionally, it was reported that SQSTM1 could bind to and stabilize NCOA4, which in turn promoted the transfer of iron from ferritin to the lipid peroxidation system, generating ROS and promoting ferroptosis [71]. These suggested that SQSTM1-NCOA4 axis that regulated ROS generation was a positive feedback loop that amplified ferroptosis induced by PS-MPs in COPD.

The results in the study that treatments including mito-TEMPO, chloroquine, and ferrostatin-1 could reverse the detrimental effects of MPs suggested a complex interplay within this axis. The primary effect of MPs is likely the induction of mito-ROS, as mitochondrial dysfunction and oxidative stress are early and central events in many toxicity pathways. MPs, due to their physicochemical properties, can disrupt mitochondrial membranes or electron transport chain/ETC components, leading to mito-ROS production [72]. This is a common early response to cellular stress. Mito-ROS can then trigger ferritinophagy by activating autophagy-related pathways via NCOA4. The release of free iron from ferritinophagy further promotes lipid peroxidation and ferroptosis [73]. The results that mito-TEMPO can reverse the detrimental effects of MPs supports that mito-ROS is upstream and a primary driver of the pathway. However, the effectiveness of CQ and Fer-1 in reversing effects of MPs suggests that ferritinophagy and ferroptosis are also critical nodes in the pathway, but they are likely secondary to mito-ROS induction. And when pDHBE cells with NCOA4 knockdown are established, ferroptosis induced by MPs was alleviated to a certain extent including changes of key ferroptosis proteins and indicators for lipid peroxidation. While FTH1 is overexpressed in pDHBE cells, the protein expression of NCOA4 was increased, and level of ferroptosis increased correspondingly resulting from significantly increased levels of ferritin degradation and iron release. These results provide a more comprehensive understanding of the axis.

In conclusion, significantly increased MPs were deposited and triggered excessive inflammatory reactions in lung tissue to induce acute exacerbation of COPD. The mechanism is summarized as PS-MPs-mitoROS-ferritinophagy-ferroptosis, which involves a pathway where MPs induce increased levels of mito-ROS, and then trigger enhanced lysosome biogenesis and acidification to promote ferritinophagy, leading to elevated levels of ferritin degradation and iron release to amplify ferroptosis in COPD. This feedback loop contributed to the acceleration of excessive ferroptosis, leading to rapid progression and poor prognosis in patients with COPD who were exposed to PS-MPs exposure. Meanwhile, mitochondria-specific ROS scavenging or ferroptosis inhibition ameliorates excessive inflammation via ADF in COPD. This study shed new light on the effects of MPs on human respiratory health and potential therapeutic strategies against the clinical challenge of disease progression in patients with COPD.

Materials and methods

Ethics approval and participation

The collection of all biological samples was conducted with the approval of the Biomedical Ethics Committee of Anhui Medical University (No. 2023805) and in strict accordance with ethical principles. All participants were informed of the study purpose and provided written informed consent. All experimental mice were treated according to the protocols approved by the Animal Care and Ethics Committee of Anhui Medical University (No. LLSC20232126).

Reagents and materials

Non-fluorescent PS-MPs and GFP fluorescent-labeled PS-MPs (PS-MPs-GFP) suspensions (2.5% w:v, 10 ml and 1.0% w:v, 10 ml) were purchased from Tianjin Baseline ChromTech Research Center (6-1-0200 and 7-3-0200) and stored at 4°C. The ferroptosis-specific inhibitor Ferrostatin-1 (SML0583), mito-ROS scavenger mito-TEMPO (SML0737), lysosomal inhibitor chloroquine (CQ; C6628), and DMSO (D2650) were obtained from Sigma-Aldrich. Cycloheximide (CHX) was purchased from MedChemExpress Biotechnology (HY-12320). The malondialdehyde (MDA), glutathione (GSH), and iron assay kits were purchased from the Nanjing Jiancheng Bioengineering Institute (A003-2-2, A006-2-1 and A039-2-1). The Pierce Classic Magnetic IP/Co-IP Kit (88804), MitoSOX Red (M36008) probe to indicate mito-ROS, LysoTracker Red (A66439) probe to track lysosomes and LysoSensor Green (A66436) probe to indicate lysosome acidification were purchased from Thermo Fisher Scientific Inc. The BCA Protein Assay Kit (P0010), MitoTracker Red (C1035), and Mitochondrial Membrane Potential Assay Kit with JC-1 (C2006) were purchased from Beyotime Biotechnology; bronchial epithelial cell growth medium (BEGM; CC-3170) was purchased from Lonza; mouse antibody to PRKN/Parkin (ab77924) and rabbit antibodies to TFEB (ab267351), PINK1 (ab216144), GPX4 (ab125066), ACSL4 (ab155282), SQSTM1/p62 (ab109012), ferritin (ab75973), LC3B (ab192890), DNM1L/DRP1 (ab184248), MFN2 (ab56889), and ACTB/β-actin (ab8226) were purchased from Abcam; rabbit antibody to PTGS2 (12282) and mouse antibody to LC3B (83506) were obtained from Cell Signaling Technology; rabbit antibody to LAMP1 (21997–1) and mouse antibody to LAMP2 (66301–1) were purchased from Proteintech Biotechnology; and rabbit antibody to NCOA4 (PA5–115626) and mouse antibody to ACSL4 (sc -365230) were obtained from Invitrogen and Santa Cruz Biotechnology, respectively.

Pyrolysis gas chromatography mass spectrometry (Py-GCMS) detection

The lung tissues used in Py-GCMS detection were obtained from 13 patients with lobectomy or segmentectomy for lung cancer in situ (as judged by senior pathologists), of which 7 patients were diagnosed with COPD and 6 were without COPD at the First Affiliated Hospital of Anhui Medical University. And samples were at a site over 2 cm away from the edge of the lung cancer. The definitive diagnosis of COPD was based on spirometry, signs, and symptoms according to the 2020 GOLD guidelines [74]. All participants provided written consent and were informed of the purpose of the study in accordance with ethical requirements. 9 types of MPs in the samples were quantified by pyrolysis (Frontier, Japan, Py-3030D) coupled to gas chromatography and mass spectrometry (Shimadzu, Japan, GCMS-QP2020), including PS, fabrics, polyamide (PA), polycarbonate (PC), polylactic acid (PLA), polyethylene terephthalate (PET), methyl methacrylate (MMA), polyvinyl acetate (PVA), and polyvinyl chloride (PVC). The lysis temperature was set to 550°C during the detection. Helium was chosen as the carrier gas with a vent flow of 1 ml/min, and the chromatographic column used was an Rtx-5 MS (Shimadzu, Japan, R221 -75855-30). To achieve chromatographic separation, the following temperature control program was set: hold at 40°C for 2 min, followed by a temperature increase to 320°C at a rate of 20°C per min and hold for 14 min. The selected ion monitoring (SIM) mode was used to identify and quantify the polymers of the target plastic particles with a mass-to-charge ratio scan range of 29–600. The standards for each target MPs were first detected, and a standard curve for MPs quantification was constructed to determine the content of MPs in the lung samples.

Cell isolation and culture

The pNHBE and pDHBE cells were isolated from human bronchial tissues of patients with lobectomy or segmentectomy for lung carcinoma in situ with and without COPD, as described previously [29,75,76]. The isolated human bronchial tissues were located over 2 cm from the edge of the lung cancer. We then applied enzymatic digestion to obtain pNHBE and pDHBE cells. Cells were placed and expanded on a collagen-coated dish in BEGM medium at 37°C in 5% CO₂ in compressed air with high humidity until 70–80% confluency was attained. The cells were identified using immunofluorescence.

Treatment with cells

The pHBE cells were expanded on six-well dishes (3.0 × 10⁵ cells/well) in BEGM medium. When 70–80% confluency was reached, the medium was replaced with basal epithelial cell medium (Lonza, CC-3171) containing 3% fetal bovine serum and without hydrocortisone. The pHBE cells were pretreated with or without PS-MPs (200 μg/ml) and/or ferrostatin-1 (Fer-1, 10 μM), mito-TEMPO (50 μM), chloroquine (CQ, 10 μM) for 24 h. RNA was extracted with TRIzol (Thermo Fisher Scientific Inc, 15596026CN), protein lysates were collected with RIPA lysis buffer (Beyotime Biotechnology, P0013B), and immunofluorescence was analyzed after fixing cells in 4% paraformaldehyde.

Lentivirus infection

The lentivirus infection was conducted to construct a stable knockdown of NCOA4 or overexpression of FTH1 in pDHBE cells. Briefly, the plasmid was cloned into the lentivirus vector, and then the recombinant lentivirus was further transfected into the HEK-293T cells (ATCC, CRL-3216) for amplification and concentration. The captured lentivirus was used to infect the pDHBE cells and puromycin was used to select the stable cells. The infection efficiency was determined using qPCR and western blot analysis.

Western blotting and co-immunoprecipitation (co-IP)

Total proteins were collected after cells were treated with RIPA lysis buffer, as described in our previous study [77]. The cells were dislodged using cell scrapers, and cellular clumps were collected by centrifugation. After stationary positioning on ice for 10 min, the mixture was centrifuged at 12,000 ×g for 10 min at 4°C, and the protein was obtained. All proteins were quantified with a BCA protein quantification kit (Beyotime Biotechnology, P0010) and subjected to western blotting as described previously [78]. The membranes were analyzed using enhanced chemo-luminescence reagent (ECL Advance; Amersham, RPN2236), and the blots were quantified using densitometric analysis with ImageJ (NIH, USA).

To detect protein interactions between NCOA4 and ferritin, LC3B by co-IP, the lysates of pDHBE cells were combine with 4 µg of NCOA4 IP-antibody per sample in a microcentrifuge tube. According to the manufacturer’s instructions for the Pierce Classic Magnetic IP/Co-IP assay kit, dilute the antibody/lysate solution to 500 µL with IP Lysis/Wash Buffer and incubate overnight at 4°C to form the immune complex. Then add the antigen sample/antibody mixture to the tube containing pre-washed magnetic beads and incubate at room temperature for 1 h with mixing. Next, wash beads twice with wash buffer and once with purified water, and elute the antigen/antibody complex. The complexes were later applied to western blot. The labeled protein membranes were observed and quantified using the Tanon system. And IgG was used as a negative control.

Quantitative real-time PCR

According to our previous study [77], RNA was extracted and reverse-transcribed to cDNA according to kit procedures (Hifair III 1st Strand cDNA Synthesis SuperMix for qPCR; Yeasen, 11141ES60), and real-time PCR was performed with Hieff qPCR SYBR Green Master Mix (Yeasen Biotechnology, 11201ES03). The expression rates of target genes were calculated using 2^∆∆Ct analysis relative to the reference gene (ACTB/β-actin), and we then calculated fold-change relative to the medium control. The following primers were used: PTGS2 (human) forward, 5′-AATCTGGCTGCGGGAACACAAC-3′, reverse, 5′-TGTCTGGAACAACTGCTCATCACC-3′; ACSL4 (human) forward, 5′-GCTCTGTCACACACTTCGACTCAC-3′, reverse, 5′-TTCCCTGGTCCCAAGGCTGTC-3′; GPX4 (human) forward, 5′-CCCGATACGCTGAGTGTGGTTTG-3′, reverse, 5′-TCTTCGTTACTCCCTGGCTCCTG-3′; FTH1 (human) forward, 5′- GCCTCCTACGTTTACCTGTCCA-3′, reverse, 5′-GAAGATTCGGCCACCTCGT-3′; FTL (human) forward, 5′-CCGTCAACAGCCTGGTCAATTTG-3′, reverse, 5′- CACGCCTTCCAGAGCCACATC-3′; and ACTB (human) forward, 5′-CACCATTGGCAATGAGCGGTTC-3′, reverse, 5′-AGGTCTTTGCGGATGTCCACGT-3′.

RNA-sequencing analysis

Briefly, total RNA was extracted. RNA-seq was performed using Illumina Novaseq 6000 platform. The retrieved data was compared to the reference genome using HISAT2 aligner, and FPKM of each gene was calculated and the read counts of each gene were obtained by HTSeq-count. Differential expression analysis was performed using the DESeq2. Q value < 0.05 and fold change > 2 was set as the threshold for significant DEGs. The transcriptome sequencing and analysis were conducted by OE Biotech Co., Ltd. (Shanghai, China).

Transmission electron microscopy (TEM)

According to our previous study [29,77], the cells were fixed at 4°C overnight in a freshly prepared fixative. The samples were rinsed in sodium phosphate-buffered saline (PBS; servicebio, G4202) and then soaked in 2% osmium tetroxide solution for 2 h at 4°C. After washing, samples were stained with 1% uranyl acetate, washed in water, dehydrated in a graded series of alcohol, and infiltrated with epoxy resin (Sigma-Aldrich, 45345). Following polymerization overnight at 65°C, 70-nm sections were cut using an ultramicrotome (Leica EM, UC7, Germany) and picked up on copper grids. The grids were post-stained with uranyl acetate and bismuth subnitrate. The sections were observed using TEM (JEM1400, Japan), and micrographs were recorded using morada G3 (EMSIS GmbH, Germany).

Observation of PS-MPs using scanning electron microscopy (SEM)

The PS-MPs samples suspended in ddH2O were heat-dried by placing the samples in a 50 ml beaker on top of a hotplate, and heating the samples for approximately 30 min until less than 1 ml of water remained. A silicon wafer (Electron Microscopy Sciences, 71893) about 1 cm² in area was used as a substrate to enhance the contrast in the SEM images. A 10-μl wet sample was pipetted onto a cleaned silicon wafer and left to air dry. The substrate was treated with gold sputtering to enhance its conductivity and resolution. Images of the PS-MPs surfaces were taken on substrates and captured using a scanning electron microscope (S-4800, Hitachi, Japan).

mCherry-GFP-LC3B adenovirus transfection

The pHBE cells were cultured and collected, and the cells were seeded in 12-well plates containing slides for cell climbing. After cell adhesion, the mCherry-GFP-LC3B adenovirus was diluted to 20 multiplicity of infection (MOI) and added to the cells according to the manufacturer’s instructions. Cells were transfected with adenovirus for 24 h, and the supernatant was then discarded and replaced with 1 ml fresh BEGM medium containing PS-MPs and/or mitoTEMPO/CQ in each well according to the intervention group, and 1 ml BEGM medium was added to the control group. Cells were then cultured at 37°C in 5% CO₂ in compressed air for another 24 h. The cells were washed with PBS, and an anti-fluorescence quenching coverslip solution containing DAPI (Beyotime Biotechnology, P0131) was added to the slides, and thereafter, the coverslips with the cells were covered. Fluorescent images of the cells were captured using a laser scanning confocal microscope (Zeiss LSM 980, Carl Zeiss AG, Germany).

Glutathione, malondialdehyde, and iron assays

The relative contents of reduced GSH, MDA, and iron in cell lysates and tissue homogenates were detected using a microplate reader and were calculated based on the protein concentration of samples according to the manufacturer’s instructions for the assay kit (Nanjing Jiancheng Bioengineering Institute, A006-2-1, A003–1 and A039).

ELISA

The concentrations of TNF, IL1B, and IL6 were quantified using ELISA kits (Dakewe Biological Technology, 1217202; Absin Bioscience, abs520001; and Dakewe Biological Technology, 1210602) according to the manufacturer’s instructions.

Mitochondrial ROS by mito-sox red

The mito-SOX Red probe was used to determine the levels of mitochondrial ROS, according to the manufacturer’s protocol (Invitrogen, M36008). Cells were seeded in six-well plates, and after stimulation with PS-MPs and/or Fer-1, mitoTEMPO or CQ, cells were incubated with 1 μM mito-SOX Red at 37°C for 30 min. Fluorescence signals were detected using flow cytometry (Beckman, CytoFlex, USA) or multimode reader (Thermo Fisher Scientific Inc., Varioskan LUX, USA).

Mitochondrial membrane potential

Tetrachloro-tetraethylbenzimidazol carbocyanine iodide (JC-1) staining was used to assess Mitochondrial Membrane Potential (MMP) according to the manufacturer’s protocol (Beyotime, C2006; Mitochondrial Membrane Potential Assay Kit with JC-1). Pretreated cells were washed three times with PBS and then incubated with a 1× JC-1 working solution at 37°C for 20 min. After incubation, the cells were washed twice with JC-1 staining buffer and were then coverslipped. Fluorescence signals were detected using flow cytometry (Beckman, CytoFlex, USA) or multimode reader (Thermo Fisher Scientific Inc., Varioskan LUX, USA).

LysoTracker staining and lysosomal pH measurement

LysoTracker Red DND-99 (Invitrogen, A66439) and LysoSensor Green DND-189 (Invitrogen, A66436) were added into the culture medium to a final concentration of 50 nM and 2 μM, respectively. Then cells were incubated at 37°C for 30 min and washed three times with culture medium. The images were acquired using a fluorescence microscope (Zeiss LSM 980).

Measurement of the half-life of ferritin protein

After cells stimulation with PS-MP, protein synthesis was inhibited using 100 μg/ml cycloheximide and the degradation of ferritin protein over time is determined by western blot analysis.

Experimental animals

Male C57BL/6 mice (8 weeks old) were housed in the Laboratory Animal Research Center of Anhui Medical University under standard specific pathogen-free conditions with 12-h light/12-h dark cycles at 22 ± 2°C and free access to the standard laboratory rodent diets and water during modeling experiments. All experimental mice were treated according to the protocols approved by the Animal Care and Ethics Committee of Anhui Medical University (no. LLSC20232126). Mice were randomly divided into seven groups (six mice in each group): 1) air; 2) cigarette smoke (CS) exposure induced COPD; 3) PS-MPs-treated COPD; 4) Fer-1-treated COPD; 5) CQ-treated COPD; 6) PS-MPs and Fer-1-treated COPD; and 7) PS-MPs and CQ-treated COPD. To establish the COPD model, mice were exposed whole bodies to CS in a passive smoking chamber (70 × 40 × 60 cm) with a house-directing flow inhalation and CS exposure system containing in a laminar flow and CS extraction units. Regular CS exposure was performed with 10 cigarettes per run, twice/day, 6 days/week for up to 24 weeks. The control mice were exposed to normal air. To establish PS-MPs-treated models, mice were intranasally administered a 40 mg/kg PS-MPs suspension for 21 consecutive days, whereas air mice were intranasally administered a 40 mg/kg saline. For Fer-1 or CQ treatment, COPD mice were intraperitoneally injected with saline, Fer-1 (10 mg/kg), or CQ (20 mg/kg) in 5% DMSO for 21 consecutive days.

Optical in vivo imaging

After COPD models were established, mice in the intervention group were intranasally administered with 40 mg/kg PS-MPs-GFP suspension in 500 nm or 2 μm size for 21 consecutive days, whereas control mice were intranasally administered with 40 mg/kg saline. The mice were anesthetized with isoflurane, and the aggregation and distribution of PS-MPs-GFP were imaged using a Caliper IVIS Lumina LT imaging system.

Lung function detection

After anesthesia with an intraperitoneal injection of 1% pentobarbital (Injection dose (μl) = mouse body weight (g) × 4), mice were tracheotomized and placed in a whole-body plethysmograph of PFT Pulmonary Maneuvers (DSI Buxco, MN, USA). Forced vital capacity (FVC), volume expired in first 100 ms of fast expiration (FEV 0.1:FVC), and maximal mid-expiratory flow curve (MMF) were measured from the fast-flow volume (FV) maneuver. The inspiratory time (Ti), expiratory time (Te), peak inspiratory flow (PIF), and minute volume (MV) were recorded during resistance and compliance (RC) maneuver. Static lung compliance (chord compliance, Cchord) was measured using the quasi-static pressure-volume (PV) maneuver.

Histological analysis

Mice were euthanatized using high-dose 1% sodium pentobarbital (100 mg/kg, ip), and the left lung lobes were dissected, fixed with 4% paraformaldehyde, and embedded in paraffin. Four-micrometer sections were stained with hematoxylin and eosin (HE; Solarbio, G1120) to evaluate the morphology and inflammatory cell infiltration of lung tissue.

Immunofluorescence staining

The cells in slides or frozen sections were stained with the corresponding anti-rabbit/mouse primary antibodies after permeabilization and blocking. Next, the cells and sections were stained with fluorescently labeled secondary antibodies (Abcam, ab150077, ab150113, ab150080, and ab150116), followed by nuclei staining with DAPI. Fluorescent images of the cells and sections were captured using a laser scanning confocal microscope (Carl Zeiss AG, Germany, Zeiss LSM 980).

Statistical analysis

Continuous variables with normal distribution are presented as mean ± standard error. The t-test or one-way ANOVA was used to determine significance when comparing two or more groups using SPSS23.0 and GraphPad 10. Tukey’s test was employed for post hoc multiple comparisons. Statistical significance is indicated at p < 0.05.

Funding Statement

This research was supported by grants from the National Natural Science Foundation of China [grant number 82170050], clinical medical research and transformation project of Anhui Province [grant number 202304295107020038], and research and practice innovation project of graduate students of Anhui Medical University [grant number YJS20230121]. Thanks for the support of Center for Scientific Research, Anhui Medical University.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data available on request from the authors.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/15548627.2025.2481126