Epithelium-derived exosomes promote silica nanoparticles-induced pulmonary fibroblast activation and collagen deposition via modulating fibrotic signaling pathways and their epigenetic regulations

Background In the context of increasing exposure to silica nanoparticles (SiNPs) and ensuing respiratory health risks, emerging evidence has suggested that SiNPs can cause a series of pathological lung injuries, including fibrotic lesions. However, the underlying mediators in the lung fibrogenesis caused by SiNPs have not yet been elucidated. Results The in vivo investigation verified that long-term inhalation exposure to SiNPs induced fibroblast activation and collagen deposition in the rat lungs. In vitro, the uptake of exosomes derived from SiNPs-stimulated lung epithelial cells (BEAS-2B) by fibroblasts (MRC-5) enhanced its proliferation, adhesion, and activation. In particular, the mechanistic investigation revealed SiNPs stimulated an increase of epithelium-secreted exosomal miR-494-3p and thereby disrupted the TGF-β/BMPR2/Smad pathway in fibroblasts via targeting bone morphogenetic protein receptor 2 (BMPR2), ultimately resulting in fibroblast activation and collagen deposition. Conversely, the inhibitor of exosomes, GW4869, can abolish the induction of upregulated miR-494-3p and fibroblast activation in MRC-5 cells by the SiNPs-treated supernatants of BEAS-2B. Besides, inhibiting miR-494-3p or overexpression of BMPR2 could ameliorate fibroblast activation by interfering with the TGF-β/BMPR2/Smad pathway. Conclusions Our data suggested pulmonary epithelium-derived exosomes serve an essential role in fibroblast activation and collagen deposition in the lungs upon SiNPs stimuli, in particular, attributing to exosomal miR-494-3p targeting BMPR2 to modulate TGF-β/BMPR2/Smad pathway. Hence, strategies targeting exosomes could be a new avenue in developing therapeutics against lung injury elicited by SiNPs. Graphical Abstract Supplementary Information The online version contains supplementary material available at 10.1186/s12951-024-02609-y.


Introduction
The rapid advancement of nanotechnology has brought about an urgent need to assess the potential toxicological impacts of nanoparticles on both the environment and human health.Silica nanoparticles (SiNPs) are commonly used nanomaterials, finding their way into a wide range of products, food additives, and biomedical applications.Inhalation was recognized as the most significant route of exposure to SiNPs, particularly in the workplace.It has been shown that nanoparticles presented in welding fumes may contribute, at least in part, to pulmonary inflammation [1].Liao et al. found pulmonary dysfunction in SiNPs handling workers, characterized by a decreased 25% forced expiratory flow (FEF25%) [2].Increasing animal studies have shown that SiNPs could cause pulmonary oxidative stress, mitochondrial dysfunction, inflammation, DNA damage, and epigenetic alterations, eventually contributing to lung fibrotic lesions [3][4][5][6].Nevertheless, the pathogenesis and underlying mediators remain incompletely understood, and therapeutic targets and effective treatments are lacking.
The respiratory epithelium serves as a first line of contact with harmful substances.Studies have demonstrated that epithelium injury and dysfunction by repeated and sequential inhaled stimuli promoted mesenchymal fibroblast activation [7], resulting in collagen deposition and lung fibrogenesis [8].In particular, exosomes can serve as mediators for cell-to-cell communication through transferring information packages [9][10][11].Exosomes with a lipid bilayer membrane structure are known as extracellular vesicles (EVs; 30-150 nm), formed by inward depression of the internal membrane during endocytosis and released into the extracellular process after fusion with the plasma membrane [12,13].Exosomes contain biomolecules [14], including lipids, proteins, and nucleic acids (miRNA, mRNA).Notably, most exosomal biologic effects have been ascribed to miRNAs through epigenetic regulation [15].Emerging data have indicated that the miRNAs in exosomes (also named exosomal miRNA) act as a key mediator in regulating cell differentiation and phenotypic changes [16,17].For example, Njock et al. have illustrated that exosomal miRNAs can regulate the alteration of pulmonary function [18].Previous studies have reported aberrant miRNA expression profiles in SiNPs-stimulated lung tissue and in vitro cultured lung cells [19][20][21], where miRNAs may participate in epithelial-fibroblast communication [22,23].Despite these advances, the specific role of exosomes/exosomal miRNA in the epithelial-fibroblast cross-talk and resultant pulmonary fibroblast activation and collagen deposition upon SiNPs exposure remains unclear.
Based on these findings, this work focused on elucidating the role and regulatory mechanism of exosomal miR-NAs from lung epithelial cells in fibroblast activation and subsequential collagen deposition.The differential miR-NAs in exosomes derived from SiNPs-stimulated lung epithelial cells were screened via transcriptome sequencing.Then, the regulatory mechanism for exosomal miR-NAs on fibroblast activation and collagen proliferation was explored in vivo and in vitro.

Animal experiment and ethics statement
The chronic inhalation exposure model of SiNPs was established in Wistar rats through intratracheal instillation.In brief, male Wistar rats (6-w old) were acquired from the Experimental Animal Center of Capital Medical University (Beijing, China) and lived in comfortable surroundings with plenty of water and food, maintained a constant humidity of 50 ± 5%, specific temperature of 24 ± 1 °C under 12-h light/dark cycles.After 1-w adaptation, experimental animals were randomly assigned to four groups, including one control group and three SiNPs exposed groups.In the exposed groups, rats were administered with SiNPs suspension via intratracheal instillation at doses of 1.5, 3.0, and 6.0 mg/kg•bw, respectively, once a week for a total of 24 treatments.As previously described [24], the SiNPs used in this experiment were synthesized using the Stöber approach, and their detailed characterization was described in the additional file Fig. S1 and Table S1.Its applied dosage was based on previous inhalation studies of SiNPs [25,26], where the administered dose was evaluated based on occupational exposure scenarios.The rats were intratracheally instilled with 0.9% saline in the control group.Finally, the rats fasted overnight, and tissues were harvested and stored at − 80 °C.The animal experiment was carried out following the Animal Care and Use Committee at Capital Medical University (Ethical number: AEEI-2021-269).

Histopathology
Fresh rat lung tissues were placed in a fixative solution and embedded in paraffin.Subsequently, hematoxylin-eosin (HE; Servicebio, China) and Masson staining (Servicebio, China) were operated on to evaluate histopathological damage following the protocol.The stained slides were scanned with a Pannoramic automated slide scanner (3DHISTECH, Hungary).Masson staining was used to detect the degree of collagen deposition in rat lungs, and its corresponding semi-quantification was conducted using Image J software.The ratio of the bluepositive stained area to all of the lung interstitial area was calculated, known as collagen volume fraction (CVF).The severity of pulmonary injury was estimated using the Ashcroft Score, as previously described [27].

Hydroxyproline content detection
Hydroxyproline (HYP) is known as the sensitive biochemical marker that indicates collagen fiber changes [28] and the severity of fibrosis [29].In animal tissues, HYP is essentially found in collagen [30].The HYP measurement was recommended as the "gold standard" for preclinical evaluation of pulmonary fibrosis by the 2017 American Thoracic Society guidelines [31].According to the product's manual, the hydroxyproline (HYP) content in lungs (50 mg) or culture supernatants was determined through a Hydroxyproline Assay Kit (Jiancheng, China).

Cell culture and exposure
Human lung epithelial cells (BEAS-2B cells) were obtained from Cell Resource Center (Shanghai, China) and seeded in DMEM medium (Corning, USA) containing 10% fetal bovine serum (FBS; Gibco, Australia) within a cell incubator (Thermo Fisher Scientific, USA).The BEAS-2B cells were incubated until 80% confluency and then treated with SiNPs (0 or 25 μg/mL) for 24 h.To investigate molecular mechanisms, GW4869 was applied to assess the impact of exosomes derived from BEAS-2B cells on fibroblast activation.GW4869, as a potent, specific, non-competitive inhibitor of membrane-neutral sphingomyelinase, is known to reduce exosome release significantly [32] and barely has toxicity to cultured cells [33].BEAS-2B cells were preprocessed with GW4869 (5 μM; Selleck, USA) for 24 h [34] and then subjected to 25 μg/mL SiNPs for another 24 h.Given that GW4869 is insoluble in water but soluble in DMSO, we set a solvent control treated with an equal amount of DMSO.Lastly, the BEAS-2B cell supernatants (SB) were gathered and filtered with filter membranes (0.22 μm; Millipore, Billerica, USA).
Human lung fibroblast cells (MRC-5) were acquired from the Cell Resource of the Chinese Academy of Science (Beijing, China) and seeded in MEM medium (HyClone, USA) containing 10% FBS (Gibco, Australia) and 1% Non-Essential Amino Acids (NEAA; Gibco, USA) in a cell incubator.To investigate the epitheliumfibroblast communication, MRC-5 cells in the logarithmic phase with about 80% confluence were stimulated by the culture supernatants from BEAS-2B cells without or with SiNPs (named as SB-Ctr or SB-SiNPs).

Cell proliferation assay
According to the protocol, fibroblast proliferation was assessed by cell counting-8 (CCK-8; Dojindo Laboratories, Japan).Briefly, the relative cell viability was expressed with 100% control cell viability, based on OD at 450 nm in the microplate reader (BioTek, USA).

Wound healing assay
When the cultured MRC-5 cells in a 6-well plate were grown to 100% confluency, fibroblasts were scratched with a sterile 200 μL pipette tip and recorded as 0 h scratch status under the microscope (Leica, Germany).Afterward, cells were cultured in SB-Ctr or SB-SiNPs for 24 h and again examined microscopically (Leica, Germany).Image J software analyzed the wound healing rate using the mathematical equation: wound healing rate (%) = (0 h scratch area-24 h scratch area)/0 h scratch area × 100%.

Real-time cell analysis (RTCA)
MRC-5 cells were cultured in RTCA E-plates (ACEA Biosciences; Germany) for 24 h and then incubated with the SB-Ctr or SB-SiNPs for another 24 h by changing half of the medium.The cell growth status was monitored in real-time, and the signals were collected every 15 min until 24 h.The analysis was performed using the xCELLigence RTCA instrument (ACEA Biosciences, Germany).

Exosomes extraction and characterization
The exosomes derived from BEAS-2B cells, including SB-Ctr or SB-SiNPs, were extracted following previous publishing [10,35,36].In brief, the supernatants were centrifuged (2,000 g; 10 min; 4 °C) and filtered to discard cellular debris and microvesicles.Then, the supernatants underwent serial centrifugation at 10,000g for 30 min and 100,000g for 70 min using an SW32Ti rotor (Beckman, USA) at 4 °C.The pellets were further purified by being dissolved in phosphate buffered solution (PBS; Servicebio, China) and ultracentrifuged at 100,000g for 70 min at 4 °C.Lastly, the exosomes were resuspended with PBS.

Exosomes uptake assay
The MRC-5 cells were incubated in confocal dishes to 80% confluence and cultured with DIO-labeled exosomes for 24 h.Nuclei were stained with Hoechst 33342 (10 μM; ThermoFisher Scientific, USA) for 30 min.LSCM examined the uptake of exosomes by MRC-5 cells.

RNA isolation and RT-qPCR analysis
The total RNA in the exosomes, cells, and lung tissues was isolated using a Trizol reagent (Life Technologies, USA) following the product manual.Based on RNA quantification spectrophotometrically at 260 nm and 280 nm, the RNA (1 μg) was reverse transcribed into cDNA for mRNA and miRNA detection through Prime-Script ™ RT reagent Kit with gDNA Eraser (Takara, Japan) and Mir-X ™ miRNA First-Strand Synthesis Kit (Takara, Japan), respectively.The level of miR-494-3p and mRNA expressions of collagen type I alpha 1 chain (COL1A1), collagen type III alpha 1 chain (COL3A1), fibronectin 1 (FN1), α-SMA and bone morphogenetic protein receptor 2 (BMPR2) were detected by TB Green ® qRT-PCR kit (Takara, Japan).Data were normalized to ACTB (mRNA) or U6 (miRNA).The primer sequences involved are presented in Table S2.

miRNA array analysis and luciferase reporter assay
Total RNA was extracted from exosomes, and miRNA microarray analysis was conducted using Agilent Arrays (Biotechnology, China).The differentially expressed genes were served as down-or upregulated with p < 0.05 and the absolute fold change > 1.5.Then, the target gene prediction was applied using TargetScan, miRDB, miRWalk, and StarBase databases.According to the bioinformatics analysis, the binding site between bone morphogenetic protein receptor 2 (BMPR2) and miR-494-3p was affirmed by the Dual-Luciferase Reporter Assay System (Promega, USA).In brief, the wild and mutant reporter plasmids of BMPR2-3 'UTR were constructed with pmir-GLO vector and transfected into 293T cells.Lastly, bioluminescence was read at 48 h post-transfection.

Overexpression of BMPR2 by lentiviral transfection
BMPR2 overexpression lentivirus (OE-BMPR2) and negative control lentivirus (OE-NC) were purchased from Oligobio.MRC-5 cells were routinely seeded in sixwell plates for 24 h.Subsequently, the culture medium was substituted with serum-free medium (1 mL), and the MRC-5 cells were transfected with OE-BMPR2 or OE-NC lentivirus (Oligobio, China) at a multiplicity of infection (MOI) of 10 along with the 8 μg/mL polybrene (Oligobio, China) for 24 h.Afterward, the culture medium was changed to a complete medium for another 48 h.To note, GFP expression of the lentivirus reporter gene was viewed, and cells exhibiting a fluorescence rate of at least 80% were applied in the following experiments.

Western blot
The protein expression was tested by Western blot analysis and normalized by GAPDH.In brief, the total proteins from cultured cells, lung tissue, or exosomes were abstracted using a Protein Extraction kit (KeyGEN, China) and then quantified with BCA assay (Dingguo, China).The samples (20 μg) were subjected to sodium dodecyl sulfate-polyacrylamide (SDS-PAGE) gels and then blotted on polyvinylidene difluoride (PVDF) membranes, subsequently blocked with 5% milk (room temperature, 2 h).After three times washing with Trisbuffered saline with 0.05% Tween 20, the membrane was incubated with primary (4 °C, overnight) and secondary antibodies (room temperature, 2 h) diluents.The primary antibodies for CD9, CD63, BMPR2, Smad1, transforming growth factor beta (TGF-β), COL3A1 and FN1 were obtained from Santa Cruz, USA, and that for TSG101, calnexin, α-SMA from Abcam, UK, and that for p-Smad1/5/9, Smad3, p-Smad3, COL1A1 and GAPDH from CST, USA.The anti-mouse or -rabbit IgG was acquired from CST, USA.At last, the images were captured using the enhanced chemiluminescence (ECL) method and analyzed using Image J software.

Statistical analysis
All statistical differences were conducted through SPSS 22.0 software.Data were expressed as mean ± standard deviation (SD).Student's t-test was used to compare two groups and ANOVA to multiple comparisons.A twotailed Pearson correlation test was applied to describe the correlation between miR-494-3p and BMPR2 levels.p < 0.05 manifests statistical significance.

SiNPs caused fibroblast activation and collagen deposition in vivo
Firstly, the impact of SiNPs on lungs was assessed by establishing an in vivo model, and the specific experimental details were shown in Fig. 1A.HE staining of lung sections (Fig. 1B) indicated the alveolar structure was basically normal, without obvious destruction, and inflammatory infiltration in the control group.On the contrary, significant lung injuries were presented in SiNPs-exposed rats, as featured by widened alveolar septa, consolidation of alveolar cavities, and even cell nodules.Masson staining and its semi-quantitative (Fig. 1C, D) revealed that SiNPs promoted collagen accumulation in the rat lungs.Concertedly, the HYP content in the rat lungs (Fig. 1E) and Ashcroft score analysis (Fig. 1F) revealed that SiNPs induced pulmonary impairments with dose-dependent exacerbation, leading to lung fibrotic phenotypic changes.Simultaneously, up-regulated protein expressions, including COL1A1, COL3A1, FN1, and α-SMA, were detected in SiNPs-exposed rat lungs (Fig. 1G, H), indicating fibroblast activation and excessive accumulation of extracellular matrix (ECM) to contribute to pulmonary fibroplasia.Taken together, these data revealed that respiratory exposure to SiNPs via intratracheal instillation caused fibroblast activation, collagen, and ECM deposition, and the resultant lung fibrotic lesion.

SiNPs-stimulated supernatants from pulmonary epithelial cells triggered fibroblast proliferation, migration, activation, and collagen deposition in vitro
Given the pathogenesis of pulmonary fibroplasia as the result of repeated and sequential inhaled stimuli-induced injury to lung epithelium [38,39], we next evaluated the intercommunication between epithelial cells and fibroblasts in response to SiNPs stimulation by establishing an indirect co-culture system in vitro.The specific experimental details are shown in Fig. 2A.Based on the cell viability determination in our previous publishing [40], here, the administered dosage of SiNPs to BEAS-2B cells was set to be 25 μg/mL, where cell viability remained up to 75%.The effects of supernatants secreted by BEAS-2B cells on MRC-5 cell proliferation were investigated via CCK-8 assay and RTCA.In comparison to SB-Ctr, SB-SiNPs exposure for 24 h resulted in a prominent rise in cell viability of MRC-5 (Fig. 2B).The enhanced fibroblast proliferation upon SB-SiNPs stimuli was also confirmed by the RTCA system (Fig. 2C).In addition, the wound healing assay demonstrated a significantly enhanced cell migration ability of MRC-5 cells when treated with SB-SiNPs (Fig. 2D).To evaluate the status of fibroblast activation and the deposition of collagen and ECM, we measured the mRNA and protein levels of COL1A1, COL3A1, α-SMA and FN1 (Fig. 2E-G).In the SB-SiNPs group, higher expressions of COL1A1, COL3A1, α-SMA, and FN1 in MRC-5 cells were noticed compared to that in the SB-Ctr group.These data confirmed the induction of fibroblast proliferation, migration, and activation in the context of lung epithelial injury elicited by SiNPs.

Exosomes from lung epithelial cells as a key mediator for fibroblast activation upon SiNPs exposure
Based on available evidence, exosomes have been considered novel communicators of cell-to-cell communication and regulators of health and disease [41].To further explore the potential communication between lung epithelial and lung fibroblasts, the exosomes in BEAS-2B supernatants were extracted through conventional ultracentrifugation (Fig. 3A).As manifested in TEM images (Fig. 3B), the extracted exosomes were cup-shaped with the double concave disc.The average particle size of exosomes was 118.4 nm in the Ctr group and 122.5 nm in the SiNPs group (Fig. 3C and Table S3).Also, exosomes were identified by Western blot (Fig. 3D), which were positive for CD9, CD63, and TSG101 and negative for calnexin.As demonstrated in Fig. 3E, the DIO-labeled Ctr-or SiNPs-exosomes could be effectively transported into MRC-5 cells.Further, to confirm whether exosomes mediated lung epithelium-fibroblast communication upon SiNPs stimulation, GW4869, an inhibitor of exosome production and release, was applied (Fig. 4A).As manifested in Fig. 4B, GW4869 markedly reversed SiNPs-induced HYP release from MRC-5 cells, indicating that inhibition of exosomes alleviated SiNPs-induced collagen deposition.Consistently, SB-SiNPs exposure triggered activation of MRC-5 cells, while GW4869 could inhibit it, as illustrated by α-SMA immunofluorescence staining (Fig. 4C) and Western blot (Fig. 4D, E).Accordingly, inhibiting exosomes released by GW4869 could significantly alleviate SiNPs-induced fibroblast activation, collagen, and ECM deposition.That is to say, exosomes derived from BEAS-2B served as key mediators involved in the lung epithelium-fibroblast communication upon SiNPs exposure, contributing to fibroblast activation and the aforementioned fibrogenesis in the lungs.

Discussion
Over the last two decades, significant development has been made in evaluating the health effects caused by SiNPs exposure.Following the respiratory inhalation, SiNPs could directly contact and accumulate in the lung tissues [45], resulting in lung injuries and collagen hyperplasia [46].However, limited data are available in terms of pathogenesis and molecular mediators accounting for observed pathological lesions.In the present work, pulmonary injury and ECM deposition have successfully occurred upon chronic exposure to SiNPs, and a dosedependent fibrogenic capability was manifested.Consistent with Yu et al. research [47], we observed that SiNPs induced widened alveolar septa, formation of various cell nodules, and collagen deposition in the rat lungs (Fig. 1B-D), especially at a higher level (6.0 mg/kg•bw).In accordance with our previous publishing [4], the lung impairments elicited by SiNPs were also dependent on exposure of duration, as characterized by alveolar structure destruction and inflammatory infiltration at an early stage, and gradually progressed into fibrotic lesions upon persistent particle stimulation.Moreover, our experimental results showed that in the lung tissues of SiNPsexposed rats, COL1A1, FN1, COL3A1, and α-SMA were dose-dependent upregulated (Fig. 1G, H), indicating fibroblast activation and collagen deposition to promote the pulmonary fibrogenesis.The dose-response effect of SiNPs on the lung tissues hinted great concern should be paid for exposure control, especially to SiNPs-handling workers.Pitifully, there is no established occupational exposure limit for SiNPs hitherto.In light of quite a slight damage upon SiNPs exposure at the dose of 1.5 mg/ kg•bw, our data may provide a valuable reference to policymakers to ensure nano-safety.
The insult to epithelial cell integrity by inhaled toxicants may initiate lung impairments or diseases via the cross-talk between epithelial cells and fibroblasts [48].When the lung epithelial cells are injured, a process of interstitial repair is triggered through the activation of fibroblasts.However, if this activation is abnormal, leading to excessive proliferation and ECM deposition, that may result in the progression of fibroplasia [49,50].Owing to the ample evidence of lung epithelial cell injury caused by SiNPs [51][52][53], we next investigated the lung epithelium-fibroblast communication in response to SiNPs stimulation.Indeed, the supernatants from SiNPsstimulated BEAS-2B cells facilitated fibroblast proliferation (Fig. 2B, C), migration (Fig. 2D), and activation (Fig. 2E-G).Nevertheless, the specific molecular mechanisms have not yet been elucidated.
Exosomes are regulators of intercellular communication by transforming their contents, which can be affected by external stimuli [54].In recent years, exosome-based mechanisms in tissue and organ damage have attracted more and more attention.Enhanced EVs secretion and function were reported in fibrotic patients and animal models [55,56], where EVs act as carriers for signaling mediators and thus contribute to fibrogenesis.In this study, the inhibition of exosome secretion from lung epithelial cells by GW4869 can greatly alleviate the activation of fibroblasts and ECM deposition by SiNPs (Fig. 4), hinting strategies targeting exosomes may be a potential way against the transformation of pulmonary fibroblasts into myofibroblasts.Consistent with our results (Fig. 4), Yu et al. [10] ever presented primary human bronchial epithelial cells-derived EVs mediated myofibroblast differentiation in lung fibroblasts.Likewise, exosomes from hypoxia-caused epithelial cells were verified to promote lung fibroblast proliferation and trans-differentiation, probably by perturbing miR-30d-3p-regulated HSF1 inhibition in a HOTAIRM1-dependent mechanism [57].
In the pulmonary fibrogenic process, epithelial injury triggers the generation of damage-associated molecular patterns (DAMPs), initiating inflammatory pathways and ultimately contributing to the pro-fibrotic response mediated by mesenchymal cells [58].It is worth noting that exosomes derived from various cell types, including macrophages, epithelial, and fibroblasts, serve as potent mediators of intracellular communication and contribute to the pathological development of lung fibrosis [59,60].For instance, exosomes derived from epithelium in lung tissues demonstrated pro-fibrotic effects by inducing pulmonary fibroblast activation [61] and modulating alveolar macrophages by targeting RGS1-mediated calcium signaling, thereby regulating the immune response [62].Exosomes derived from lung fibroblasts promoted the generation of mitochondrial reactive oxygen species, contributing to mitochondrial damage and resultant senescence in lung epithelial cells [63].Also, exosomes from macrophages were shown to increase the proliferation of pulmonary interstitial fibroblasts and promote the progression of pulmonary fibrosis [64].Thereby, exosomes have been considered to be potential biomarkers for diverse lung diseases.Recent studies have shown that the inhibition of exosome secretion by GW4869 has great potential in the treatment of lung injury [65] (See figure on next page.)Fig. 5 Exosomal miR-494-3p secreted from BEAS-2B cells could target BMPR2 to regulate the TGF-β/BMPR2/Smad pathway in MRC-5 cells.A Heatmap of the differential expressions of miRNA in exosomes derived from BEAS-2B.Differential miRNAs with the absolute fold change (FC) > 1.5 and p < 0.05 were shown.B Potential target genes prediction from 4 databases that could interact with miR-494-3p.The miR-494-3p level was elevated in exosomes (C) and MRC-5 (D, left).In contrast, the level of BMPR2 was decreased in MRC-5 (D, right).Consistently, the levels of miR-494-3p and Bmpr2 were detected in lung tissues (F).N = 4; *p < 0.05 vs control.Meanwhile, the correlation between miR-494-3p and BMPR2 was conducted in vitro (E) and in vivo (G).(H) Dual luciferase assay verified the targeting relationship between miR-494-3p and BMPR2.I, J The molecular expressions of the TGF-β/BMPR2/Smad signaling pathway in the lung tissues, including BMPR2, Smad1, p-Smad1/5/9, Smad3, p-Smad3, and TGF-β.N = 3; *p < 0.05 vs control.The levels of miR-494-3p and BMPR2 (K), and protein expressions of BMPR2, Smad3, p-Smad3, Smad1, p-Smad1/5/9 and TGF-β (L, M) were measured in MRC-5 cells treated with supernatants of BEAS-2B (Ctr, SiNPs, GW4869, GW4869 + SiNPs, or DMSO).N = 3; *p < 0.05 vs SB-Ctr; # p < 0.05 vs SB-SiNPs and silica-induced pulmonary fibrosis [66].Intriguingly, exosomes also manifested superior therapeutic benefits in lung diseases.For instance, Dinh et al. [67] indicated that inhaled lung spheroid cell exosomes could promote pulmonary repair.Normal human bronchial epithelial cells-derived EVs were reported to inhibit TGF-βmediated lung fibroblast activation [35].
miRNAs are known to regulate gene silencing through the post-transcriptional mRNA degradation or repression of the expressed proteins [79].Here, we illustrated that miR-494-3p was a regulator to target BMPR2 (Fig. 5H) to promote profibrotic phenotypes.Referring to previous studies, BMPR2 was declined in the fibrotic lung tissues of patients and animal models [42,44,80,81] and negatively related to the TGF-β/Smad pathway activation [81,82].In our work, BMPR2 and the phosphorylation of Smad1/5/9 were dramatically reduced in SiNPs-exposed rat lungs and fibroblasts, accompanied by elevated TGF-β and ensuing phosphorylated Smad3 (Fig. 5I-M).These data indicated that SiNPs exposure led the TGFβ-BMPR2 balance towards a TGFβ/Smad3 dominant activation direction, resulting in fibroblast activation and accelerating the fibrotic disease development.There is a large body of studies that demonstrated that the imbalanced TGF-β/BMPR2/Smad signaling could lead to pathogenic processes in pulmonary fibroblast activation and collagen deposition [81].In particular, BMPR2 overexpression could greatly alleviate SiNPs-induced fibroblast activation, which was mediated by the recovery of impaired TGF-β/BMPR2/Smad signaling (Fig. 7).Concertedly, Fukihara et al. [83] indicated that transduced BMPR2 into the lung fibroblasts could suppress phosphorylation of Smad2/3.Switch-independent 3a promoted BMPR2 expression through methylation regulation to inhibit pulmonary arterial smooth muscle cell proliferation [84].Overall, these findings distinctly demonstrated that BMPR2 depletion is a crucial hallmark of pulmonary fibrotic phenotype.
Admittedly, this work has some limitations.Apart from epithelial cells, exosomes are released from other lung cells and various immune cells in the lungs.The  work should be carried out to validate the blockade of exosome secretion (e.g., GW4869) or inhibition of exosomal miR-494-3p level, which may be a feasible strategy against lung injury and fibrosis caused by SiNPs.

Conclusions
In the light of mega-production, extensive application, and increased respiratory exposure of SiNPs, this study aimed to discover the mechanism underlying SiNPselicited lung impairment from the perspective of crosstalk between respiratory epithelial injury and fibroblast activation, with a focus on exosomes and its packaged miRNAs.These data firstly expounded that epitheliumderived exosomes promoted pulmonary fibroblast activation and collagen deposition induced by SiNPs via modulating fibrotic signaling pathways and their epigenetic regulations.Specifically, SiNPs exposure could perturb the exosome secretion from lung epithelial cells; in particular, increased exosomal miR-494-3p led to fibroblast activation and collagen deposition through TGF-β/ BMPR2/Smad signaling via targeting BMPR2.Overall, our study provided a new perspective on the role of exosomal miRNAs in SiNPs-elicited lung toxicity, as well as a scientific foundation for identifying new targets in the mitigation or treatment for SiNPs-induced lung injury.

(
See figure on next page.)Fig. 1 SiNPs promoted fibroblast activation and collagen deposition in vivo.A Schematic representation of animal experimental procedure.B HE staining.Alveolar structural destruction (green arrow), thickened alveolar walls (yellow arrow), and cell nodules (blue arrow) were observed in the lungs with SiNPs administration.N = 4. Representative images of Masson staining C and corresponding semi-quantitative result D illustrated a dose-dependent increase in collagen deposition.N = 4. Scale bar, 100 or 200 μm.E HYP measurement in the lungs.N = 5. (F) Ashcroft score was used to assess the severity of pulmonary fibrosis.N = 4. G, H Western blot assay to detect expressions of fibroblast activation and ECM deposition-associated markers, such as COL1A1, COL3A1, FN1, and α-SMA.N = 3. *p < 0.05 vs control

Fig. 4
Fig. 4 GW4869 attenuated SiNPs-induced fibroblast activation and collagen deposition.A The schematic diagram of cell treatment and grouping.BEAS-2B cells were preprocessed with GW4869 for 24 h and then treated with 25 μg/mL SiNPs for another 24 h.After treatment, the supernatants from BEAS-2B were collected and stimulated to MRC-5 cells for the evaluation of fibroblast activation.B The content of HYP was measured in supernatants of MRC-5.C Immunofluorescence staining for α-SMA in MRC-5 cells.Green: α-SMA; Blue: DAPI.Scale bar, 25 or 100 μm.Meanwhile, protein levels of COL1A1, COL3A1, FN1, and α-SMA were quantified via Western blot in MRC-5 cells (D, E).N = 3; *p < 0.05 vs SB-Ctr; # p < 0.05 vs SB-SiNPs