2.1 Cell lines and cell culture

The mouse macrophage cell line RAW264.7, mouse embryonic fibroblast cell line NIH-3T3, human monocyte leukaemia cell line THP-1 and human embryonic fibroblast cell line HFL1 were purchased from the China Center for Type Culture Collection (CCTCC). HFL1 cells were cultured in Ham's F-12K (Kaighn's) medium (F12k; Gibco) supplemented with 10% foetal bovine serum (FBS; BI, Kibbutz Beit-Haemek). RAW264.7 and NIH3T3 cells were cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco) supplemented with 10% FBS. THP-1 cells were cultured in Roswell Park Memorial Institute (RPMI)-1640 medium (BI, Kibbutz Beit-Haemek, Israel) supplemented with 10% FBS. All the cells were cultured at 37°C with 5% CO₂. To perform a wound closure assay, cells were cultured in growth medium without FBS. THP-1 cells were treated with phorbol 12-myristate 13-acetate (PMA; 100 ng/mL, Sigma-Aldrich, Merck KGaA,) for 24 hours to induce differentiation into an adherent macrophage phenotype before SiO₂ exposure. Before administration, silicon dioxide (SiO₂; S5631, 1-5 µm, Sigma-Aldrich, Merck KGaA,) was packaged in tinfoil and baked overnight (180°C, 16 hours) for sterilization and inactivation of any endotoxin contamination. In this study, the SiO₂ dosage was 200 μg/mL. The conditioned medium of macrophages was harvested after treatment with SiO₂ for 48 hours. For an ER stress inhibition assay, fibroblasts were pre-treated with 4-phenylbutyric acid (4-PBA) (500 μmol/L, Sigma-Aldrich, Merck KGaA,) for 4 hours in growth medium without FBS. After washing with sterile phosphate-buffered saline (PBS) 3 times, these fibroblasts were treated with SiO₂-Exos +4-PBA (500 μmol/L) for different time points in growth medium supplemented with 10% FBS.

2.2 Exosome isolation, identification and treatment

Exosomes were isolated from conditioned medium of RAW264.7 or THP-1 cells by differential centrifugation (Beckman Coulter). RAW264.7 and THP-1 cells were cultured in growth medium with 10% exosome-depleted FBS, and the cell culture supernatant (SN) was collected after 48 hours of culture. The collected SN was centrifuged at 3 000 × g at 4°C for 10 minutes to remove cells and cellular debris and then at 20 000 × g at 4°C for 25 minutes to remove microvesicles. Vesicles smaller than 200 nm were collected by centrifugation at 110 000 × g at 4°C for 120 minutes after filtration with a 0.22 μm filter (Millipore, MA). The precipitate was resuspended in ice-cold sterile PBS, and the mixture was ultracentrifuged at 110 000 × g at 4°C for 120 minutes to remove contaminated proteins. Finally, the exosome pellets were resuspended in specific resuspension buffers according to different objectives. These exosomes could be used immediately or stored at −80°C. The number of exosomes was determined by measuring the total protein of exosomes by a Micro-BCA assay (Pierce). To verify the exosomes, the expression of the extracellular vesicle-related protein markers CD63 (1:1000; ab134045; Abcam) and HSP70 (heat shock protein 70; 1:1 000; ab2787; Abcam) and the exosome-specific marker TSG101 (tumour susceptibility gene 101; 1:1 000; 14497-1-AP; Proteintech) was detected by Western blot analysis. For transmission electron microscopy (TEM; FEI) analysis, exosomes were prefixed with 1% glutaraldehyde, followed by negative staining with a negative-staining solution (4% uranyl acetate (pH 4) + 2% methylcellulose). Then, the preparation could be viewed and imaged by TEM. The range of size distribution of exosomes was confirmed by nanoparticle tracking analysis (NTA; Zetasizer Nano ZS, Malvern Instruments, Worcestershire). For cell treatment, 1 × 10⁵ recipient cells were treated with 50 μg of exosomes at different time points.

2.3 Trafficking analysis of exosomes

To dynamically trace exosomes, the exosomes were labelled with a PKH26 fluorescent kit (Sigma-Aldrich, Merck KGaA). First, according to the manufacturer's directions, the purified exosomes were incubated with PKH26 dye. Next, the mixture was diluted in ice-cold PBS and then centrifuged at 110 000 × g at 4°C for 90 minutes to collect the PKH26-labelled exosome pellets, which were resuspended in 100 μL of sterile PBS and then added to coculture with HFL1 cells in complete growth medium for 24 hours. After washing 3 times with PBS, an inverted fluorescence microscope (Olympus) was used to detect the appearance of red fluorescence in HFL1 cells.

2.4 Exosome secretion inhibition assay

To inhibit the secretion of exosomes, RAW264.7 and THP-1 cells were pre-treated with the exosome secretion inhibitor GW4869 (10 μmol/L, Cayman Chemical)²⁷ 24 hours prior to SiO₂ exposure. RAW264.7 and THP-1 cells were then treated with SiO₂ (200 μg/mL) + GW4869 (10 μmol/L) for 48 hours and the cell culture SN was harvested.

2.5 Western blot analysis

Cell lysates and exosomal lysates were subjected to SDS-PAGE. Western blot analyses were performed by using anti-collagen I (1:1 000; 14695-1-AP; Proteintech), anti-α-smooth muscle actin (SMA) (1:1 000; 14395-1-AP; Proteintech), anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (1:1 000; 14497-1-AP; Proteintech), anti-immunoglobulin heavy chain binding protein (BIP) (1:1 000; 11587-1-AP; Proteintech), anti-spliced X-box binding protein 1 (XBP1s) (1:1 000; #12782; Cell Signaling Technology), and anti-phospho eukaryotic initiation factor 2 (eIF2) α (P-eIF2α) (1:1 000; #3398; Cell Signaling Technology) antibodies and corresponding HRP-conjugated secondary antibodies. A chemiluminescent system (Amersham ECL Plus) was used for detection.

2.6 Quantitative RT-PCR analysis of mRNA

Following the manufacturer's protocols, Trizol reagent (Thermo Scientific) was used to extract total RNA, and then, a NanoDrop 2000 spectrophotometer (Thermo Scientific) was used to evaluate RNA quality and quantity. Total RNA was reverse transcribed using a cDNA synthesis kit (Genecopoeia), and then, an ABI-7500 instrument (Applied Biosystems; Thermo Fisher Scientific Inc Waltham) was used to perform real-time PCR. The expression levels of collagen I and α-SMA were normalized to GAPDH. The primers used in the reactions are listed below: mouse collagen I 5ʹGCTCCTCTTAGGGGCCACT3ʹ, 3ʹCCACGTCTCACCATTGGGG5ʹ; mouse α-SMA 5ʹGTCCCAGACATCAGGGAGTAA3ʹ, 3ʹTCGGATACTTCAGCGTCAGGA5ʹ; mouse GAPDH 5ʹAGGTCGGTGTGAACGGATTTGʹ, 3ʹTGTAGACCATGTAGTTGAGGTCAʹ; human collagen I 5ʹCTCCGGCTCCTGCTCCTCTTAG3ʹ, 3ʹGGCAGTTCTTGGTCTCGTCACAG5ʹ, human GAPDH 5ʹGGCACCGTCAAGGCTGAGAACʹ, 3ʹTGCAGGAGGCATTGCTGATGATCʹ; and human α-SMA 5ʹCAACGTGGAGCGCAGTGGTC3ʹ, 3ʹCAAGGCAGTGCTGTCCTCTTCTTC5ʹ.

2.7 Proliferation and a wound closure assay

Following the manufacturer's instructions, cell proliferation was detected by Cell Counting Kit-8 (CCK-8 Kit; Dojindo Molecular Technologies). Cell migration was assessed by a wound closure assay in a 2D culture system. 1 × 10⁵ HFL1 or NIH3T3 cells were inoculated into 24-well tissue culture plates. When the cells were approximately 70%-80% confluent, a cross-shaped scratch was carefully drawn in each well using a sterile 200-μL pipette tip. Each well was then washed 3 times to remove all the detached cells. Digital images were captured at different time points (0, 12 or 24 hours), and ImageJ software (National Institutes of Health) was used to quantitatively evaluate the gap width.

2.8 Animal model

The design and methods of the research are in accordance with the requirements of related regulations and procedures (such as the National Institutes of Health Guide for the Care and Use of Laboratory Animals) as well as ethical principles. Six-week-old male C57BL/6 mice were randomly divided into three groups (control, SiO₂ and SiO₂ + GW4869). The silicosis model was induced by intratracheal injection of silica (100 mg/kg bodyweight) that was suspended in 20 μL of sterile PBS. The control group received the same volume of sterile PBS. The animals were killed 28 days after silica exposure, and the lungs and BALF were collected for further analysis. For administration of an exosome secretion inhibitor, GW4869 (2.5 μg/g of bodyweight per day) was dissolved in PBS and pre-administered intraperitoneally to each animal for 7 days before silica exposure. GW4869 was continuously administered to the mice until they were killed at 35 days for further study.

2.9 Haematoxylin and eosin, Masson's trichrome and immunohistochemical staining

Lungs were fixed overnight with 4% paraformaldehyde at a constant pressure to ensure that the lungs were totally submerged, and then, the lungs were dehydrated and embedded in paraffin. The sections were deparaffinized and stained with Masson's trichrome or haematoxylin and eosin (H&E) (Sigma-Aldrich, Merck KGaA). Immunohistochemical staining was performed with an anti-α-SMA antibody (1:200; 14395-1-AP; Proteintech) as described previously.²⁸ Digital images of the area of positive staining were captured by using a light microscope attached to an image-analysis system (K-Viewer, KFBIO).

2.10 Statistics

GraphPad Prism 5 (GraphPad Software) was used for statistical analysis. Numerical data were compared using Student's t test (unpaired, two-tailed) or two-way ANOVA, and the statistical significance was set at P < .05.

3.1 Exosome secretion is increased in the cell culture supernatant of RAW264.7 cells treated with SiO₂

A recent study indicated that the secretion of exosomes into BALF is up-regulated in bleomycin-induced lung fibrosis or idiopathic pulmonary fibrosis,¹⁷ and circulating endothelial microparticle levels are increased in chronic obstructive pulmonary disorder (COPD) patients.²⁹ Hence, we first asked whether the secretion level of exosomes is altered in SiO₂-induced pulmonary fibrosis. For this purpose, we isolated and characterized exosomes from the cell culture supernatant of RAW264.7 cells treated with SiO₂ or a blank control (Figure 1A). The supernatant was collected from RAW264.7 cells after SiO₂ or vehicle administration for 48 hours. Morphological assessment of exosomes by TEM revealed the presence of exosomes (cup-shaped vesicles with a diameter of 30-150 nm; white arrows) (Figure 1B). We further found enriched expression of the exosome-specific marker TSG101 and the specific endosomal markers CD63 and HSP70 in purified exosomes (Figure 1C). The expression levels of HSP70, TSG101 and CD63 were increased in SiO₂-Exos compared to exosomes from control-treated macrophages (NC-Exos), suggesting a potential increase in exosome secretion with SiO₂-exposure. Next, we detected the size distribution of the purified exosomes by NTA (Figure 1D), which showed that the sizes of exosomes were mostly between 30 and 200 nm (peak: l44.5 nm). In addition, we found that the total protein content of SiO₂-Exos increased significantly compared with NC-Exos (Figure 1E, exosome total μg of protein/mL: NC 129.8 ± 13.63, SiO₂ 253.7 ± 17.92, P = .0006).

3.2 SiO₂-Exos promote myofibroblast differentiation in vitro

Previous studies have shown that the activation of macrophage induced by SiO₂ initiates pulmonary fibrosis in silicosis, followed by myofibroblast activation and aberrant collagen deposition.^{2, 3, 30} One of the main mechanisms of myofibroblast activation is the phenotypic transdifferentiation of fibroblasts into myofibroblasts by which myofibroblasts arise and accumulate in lung tissue.³¹ Myofibroblasts are characterized by high expression of the myocyte-specific isoform α-SMA, and increased synthesis and secretion of ECM components (such as collagen I, III and V). Inhibition of myofibroblast differentiation and proliferation can effectively suppress the development of pulmonary fibrosis.³² Additionally, recent studies have indicated that exosomes play a crucial role in fibrotic diseases as carriers of intercellular communication.^{15, 16} The number of exosomes in the BALF of IPF patients is increased, and these exosomes promote lung fibroblast proliferation.¹⁷ Given that high secretion of exosomes is related to SiO₂-induced pulmonary fibrosis, we explored whether exosomes derived from macrophages mediate myofibroblast activation. First, we investigated whether macrophage-secreted exosomes can be ingested by fibroblasts. SiO₂-Exos were labelled with the red fluorescent dye PKH26 (Figure 2A), and then, these exosomes were incubated with HFL1 cells for 24 hours. The red fluorescence in HFL1 cells showed that these exosomes had been effectively phagocytized by the cells (Figure 2A). These results indicate that exosomes derived from macrophages can be ingested by fibroblasts.

Next, we evaluated whether exosomes derived from macrophages can induce myofibroblast differentiation. HFL1 cells and NIH-3T3 cells were treated with PBS, NC-Exos, SiO₂-Exos or exosomes from macrophages treated with SiO₂+GW4869 (SiO₂+GW4869-Exos). Notably, treatment with SiO₂-Exos resulted in significantly increased expression of collagen I and α-SMA in HFL1 cells (Figure 2B) and NIH-3T3 cells (Figure 2C), and these effects were attenuated by pre-treatment of macrophages with GW4869 (Figure 2B,C), which inhibits exosome generation.²⁷ Consistent with these observations, the mRNA expression levels of collagen I and α-SMA were also increased in HFL1 cells (Figure 2D,E) and NIH-3T3 cells (Figure 2F,G) treated with SiO₂-Exos, and these effects could be reversed by GW4869 pre-treatment.

3.3 SiO₂-Exos promote myofibroblast proliferation and migration

Activated myofibroblasts proliferate and migrate to form fibrotic foci and produce excessive ECM components, which are characteristics of fibrotic diseases.^{33, 34} In addition, McBride et al ³⁵ found that mesenchymal stem cell (MSC)-derived exosomes could promote dermal fibroblast proliferation. Therefore, we next explored whether SiO₂-Exos modulate myofibroblast proliferation and migration. SiO₂-Exos or SiO₂ + GW4869-Exos were used to treat fibroblasts, and then, fibroblast proliferation and migration were measured. The results showed that SiO₂-Exos promoted HFL1 (Figure 3A) and NIH-3T3 (Figure 3B) cell proliferation. Consistent with these observations, treatment with SiO₂-Exos significantly promoted the migration of HFL1 cells (Figure 3C) and NIH-3T3 cells (Figure 3D), and these effects were attenuated by pre-treatment of macrophages with GW4869.

3.4 Inhibition of exosome secretion blocks SiO₂-induced proinflammatory cytokine production and pulmonary fibrosis in mice

Next, we investigated whether exosomes contributed to the progression of fibrosis in a silicosis model. To determine the exosome-mediated promotive effects on pulmonary fibrosis in silicosis, GW4869 was used to block the generation of exosomes. We constructed a silicosis pulmonary fibrosis model given various treatments (Figure 4A). Either PBS or GW4869 (2.5 μg/g per day) was used to pre-treat mice for 7 days prior to silica suspension exposure; the silica suspension (100 mg/kg) was then administered by intrabronchial injection. GW4869 was continuously administered to the mice until they were killed on day 35 (28 days after exposure to the silica suspension), and the BALF and lung tissues of the mice were harvested and examined. We found that, in BALF, the expression levels of HSP70, TSG101 and CD63 were significantly up-regulated in SiO₂-Exos (the SiO₂ group) compared with control-Exos (the control group) and down-regulated in SiO₂ + GW4869-Exos (the SiO₂ + GW4869 group) compared with SiO₂-Exos (Figure S1A). Consistent with these observations, the total protein content of exosomes also increased significantly in SiO₂-Exos compared with control-Exos and decreased in SiO₂+GW4869-Exos compared with SiO2-Exos (Figure S1B). These results suggested that treatment with GW4869 could inhibit exosome generation in mice. H&E staining showed that in the control group, the lung tissue had clear and intact alveolar structures without obvious inflammatory cell infiltration or fibrosis. In contrast, in the SiO₂ group, the alveolar structures were significantly damaged, along with extensive inflammatory cell aggregation and silicon nodule formation (Figure 4B). Masson's trichrome staining demonstrated that compared with the control group, the sizes of dark blue staining regions in the SiO₂ group were increased significantly (Figure 4C). However, the regions of inflammatory cell aggregation and collagen fibre deposition were reduced accompanied by weaker staining in the SiO₂ + GW4869 group (a group including mice pre-treated with GW4869) than in the SiO₂ group (Figure 4B-C). Additionally, the protein expression of α-SMA was markedly up-regulated in the SiO₂ group compared with the control group and down-regulated in the SiO₂+GW4869 group compared with the SiO₂ group (Figure 4D). Furthermore, the results of an analysis of proinflammatory cytokines in the BALF revealed significant increases in the levels of TNF-α (Figure 4E), IL-1β (Figure 4F) and IL-6 (Figure 4G) in the SiO₂ group compared with the control group. Pre-treatment with GW4869 resulted in a significant decrease in the production of TNF-α, IL-1β and IL-6 in BALF induced by SiO₂ (Figure ). In addition, we observed that the average bodyweight was decreased in the SiO₂ group compared with the control group and increased in the SiO₂+GW4869 group compared with the SiO₂ group (Figure 4H).

3.5 SiO₂-Exos regulate myofibroblast activation through ER stress

ER stress and the UPR are known to contribute to the development of pulmonary fibrosis by regulating myofibroblast differentiation.^{20-22, 36} Additionally, recent studies have shown that exosomes modulate apoptosis through ER stress.^37-39 Consequently, we hypothesized that SiO₂-Exos may regulate myofibroblast activation through ER stress. First, to investigate whether ER stress is involved in regulating myofibroblast differentiation induced by SiO₂-Exos, we examined the expression levels of several protein markers related to these processes. SiO₂-Exos (± 4-PBA) were used to treat fibroblasts, and the expression levels of UPR-related genes (XBP1s and P-eIF2α), ER stress marker (BIP) and myofibroblast differentiation markers (collagen I and α-SMA) were measured. These results showed that after treatment with SiO₂-Exos, the expression levels of BIP, XBP1s and P-eIF2α were up-regulated over time (Figure 5A,B). Similarly, the expression of collagen I and α-SMA in HFL1 cells (Figure 5A) and NIH-3T3 cells (Figure 5B) was also increased over time. When SiO₂-Exos and the ER stress inhibitor 4-PBA were cocultured with fibroblasts, the expression levels of BIP, XBP1s and P-eIF2α were down-regulated (Figure 5C,E). Consistent with these observations, the expression of collagen I and α-SMA in HFL1 cells (Figure 5C,D) and NIH-3T3 cells (Figure 5E,F) was decreased in the combination group (SiO₂-Exos + 4-PBA) compared with the SiO₂-Exos group.

3.6 Inhibition of ER stress attenuates fibroblast proliferation and migration induced by SiO₂-Exos

We next examined whether ER stress is involved in fibroblast proliferation and migration induced by SiO₂-Exos. The fibroblasts were treated with SiO₂-Exos (± 4-PBA), and then, fibroblast proliferation and migration were evaluated. The results showed that SiO₂-Exos promoted HFL1 and NIH-3T3 cell proliferation (Figure 6A,B) and migration (Figure 6C,D), and these facilitating effects were suppressed by 4-PBA.