Therapeutic effects of adipose-tissue-derived mesenchymal stromal cells and their extracellular vesicles in experimental silicosis

This study was approved by the Health Sciences Ethics Committee of the Federal University of Rio de Janeiro (CEUA 188/13). All animals received humane care in compliance with the principles of laboratory animal care formulated by the National Society of Medical Research and the Guide for the Care and Use of Laboratory Animals prepared by the National Academy of Sciences, USA.

A total of 50 female C57BL/6 mice (8–12 weeks old) were divided into two groups: silica (SIL), in which animals received an intratracheal injection of 20 mg of silica in 50 μL of saline, and control (CTRL), in which the animals received 50 μL of saline intratracheally. After 15 days, the SIL group was further randomly divided into four groups: Sal (intratracheal injection of 50 μL of PBS), AD-MSC (dose of 100,000 AD-MSCs in 50 μL of PBS), EV5 (EVs obtained from 100,000 AD-MSCs for 24 h, also in 50 μL of PBS), and EV6 (10× higher EV dose than EV5).

Male C57BL/6 mice (weight 20–25 g, 8 weeks old) were used as donors. MSCs from adipose tissue (epididymal fat pad) were obtained as previously described [25]. A brief description of the isolation procedure and culture conditions is given in Additional file 1.

At the third passage, approximately ten million cells were characterized as MSCs through flow cytometry and induction of differentiation into osteoblasts and chondroblasts as previously described [26]. Cells from the third to fifth passage were used for EV isolation and instillation. For direct instillation, cells were detached with trypsin, washed, and re-suspended in PBS. For in vitro internalization essays, lung fibroblasts were obtained from healthy C57BL/6 mice and cultivated in DMEM containing 1% antibiotic solution, 10% fetal bovine serum (FBS) and 15 mM HEPES for up to 2 weeks [27]. MH-S cells were purchased from ATCC (#CRL2019) and cultivated in RPMI 1640 with 10% FBS, 1% penicillin-streptomycin (10,000 U/mL, Thermo Fisher Scientific, USA) and 5 mM 2-mercaptoethanol according to the vendor’s instructions. RAW264.7 cells were exposed to silica for 2 h (100 μg/mL) followed by treatment with EVs or regular medium for 24 and 48 h.

AD-MSCs were kept in exosome-free medium (with the same proportion of FBS, previously ultracentrifuged without dilution at 100,000×g overnight to remove bovine EVs [28]) at 70–90% confluence. After 24 h, the conditioned medium was removed, and EVs were enriched as follows: centrifugation at 300×g for 10 min, supernatant centrifugation at 3000×g for 20 min, and final supernatant ultracentrifugation at 100,000×g for 2 h (modification from Théry et al. [29]).The pellet was washed and ultracentrifuged with PBS once and then re-suspended in PBS for downstream applications. The methods used for characterization of EVs are given in Additional file 1.

AD-MSCs were stained with Vybrant DiI (Life Technologies, USA) before isolation of EVs. For in vitro internalization experiments, MH-S cells and primary fibroblasts were incubated with 3 × 10⁸ stained vesicles per well in 12-well cell culture plates for 24 h, washed with PBS, fixed in methanol, and mounted with ProLong Gold Antifade Mountant with DAPI DNA dye (Life Technologies, USA). The slides were imaged in a Zeiss LSM 710 confocal microscope at 200× and 400× magnification. For flow cytometry analysis, cells were washed and re-suspended in PBS after incubation and used without fixation.

For in vivo distribution experiments, stained EVs were delivered intratracheally in 50 μL of PBS in 5 C57BL/6 female mice, 15 days after instillation of 20 mg of silica. The lungs were collected 5 min after the instillation, snap frozen in liquid nitrogen, embedded in optimal cutting temperature medium (Tissue-Tek OCT compound, Electron Microscopy Sciences, Fisher Scientific, USA), and cryosectioned into 10 μm slices. Slides were mounted with ProLong Gold/DAPI as above and imaged at 200× magnification with a Zeiss fluorescence microscope.

A laparotomy was done immediately after the determination of lung mechanics (as described in Additional file 1) and heparin (1000 IU) was injected intravenously into the vena cava. The trachea was clamped at end expiration, and the abdominal aorta and vena cava were sectioned, yielding a massive hemorrhage that quickly killed the animals. The left lung was then removed, fixed in 3% buffered formaldehyde and embedded in paraffin. Slices were cut (4-μm thick) and stained with hematoxylin–eosin. Lung histology analysis was performed with an integrating eyepiece with a coherent system consisting of a grid with 100 points and 50 lines (known length) coupled to a conventional light microscope (Olympus BX51, Olympus Latin America-Inc., Brazil). The area fraction of granuloma was measured with ImageJ v1.48 software (NIH, USA). The number of mononuclear and polymorphonuclear cells in pulmonary tissue and inside the granuloma was determined using the point-counting technique across 10 random, non-coincident microscopic fields [30]. For the assessment of collagen fiber deposition, slices were stained using the Picrosirius method [31]. The fraction of area occupied by collagen fibers (stained in red) in relation to total tissue area was quantified using Image Pro Plus version 5.1 (Media Cybernetics, USA).

For protein isolation, the right lobes of the lungs were snap frozen in liquid nitrogen and kept at − 80 °C until analysis. Protein was extracted from tissue homogenates using RIPA buffer. Levels of interleukin (IL)1-β, transforming growth factor (TGF)-β, tumor necrosis factor (TNF)-α, and IL-6 in lung tissue were measured by ELISA using matched antibody pairs from PrepoTech (Rocky Hill, NJ, USA) and R&D Systems (Minneapolis, MN, USA), according to the manufacturers’ instructions. Protein levels of TNF-α and TGF-β were also measured in cell supernatants using the same methods.

The total number of macrophages was quantified separately in alveolar septa and granuloma through reaction with monoclonal antibody F4/80 rat anti-mouse (AbD Serotec). The antibody was detected using a secondary antibody labeled with peroxidase (Histofine mouse MAX PO, Nichirei Biosciences, Japan) followed by the chromogen substrate, diaminobenzidine (liquid DAB, Dakocytomation, USA). Thirty microscopic fields were randomly selected, avoiding vessels and bronchi, using a digital camera (Evolution, Media Cybernetics, USA) coupled to a light microscope (Eclipse 400, Nikon, Japan), and a computer with graphical interface software (Q-Capture 2.95.0, Silicon Graphics, USA). High-quality images were captured and the number of labeled cells was divided by the total number of cells per field.

We observed release of EVs from AD-MSCs using transmission electron microscopy (Fig. 1a, b). After sequential centrifugation and ultracentrifugation of the conditioned media [29], the final fraction obtained (which should be enriched in smaller EVs) was observed by scanning electron microscopy (Fig. 1c). Spherical structures under 150 nm in diameter were prevalent in the sample. Larger amorphous structures that suggest aggregation were observed. When measured by nanoparticle tracking analysis (Fig. 1d), the vesicle size distribution was bimodal with mode diameters of 145 nm and 285 nm. This size distribution also indicates the presence of larger populations of vesicles, which is expected for a sample containing vesicular and non-vesicular moieties. The mean size was around 250 nm, in agreement with what was described for human AD-MSC EVs by Lopatina and colleagues [32] using a similar NTA technique. We did not detect EVs larger than 1000 nm, indicating the absence of relevant amounts of apoptotic bodies in our samples. EVs were positive for common EV surface markers: CD63, CD81, Lamp1, and CD9 (Additional file 2); 10⁶ cells (EV6 dose) produced on average 3.84 × 10⁹ EVs, with a standard deviation of 1.92 × 10⁹. No differences in size distribution or concentration were observed when comparing EVs from cells at different passages (data not shown). The fraction of vesicles obtained did not present 18S or 28S ribosomal RNA fractions, indicating negligible cytosolic contamination, but small RNAs were detected. Total RNA concentration was around 50 ng/10⁹ vesicles. Protein content was around 40 μg/10⁹ vesicles.

To explore whether different cell culture conditions could increase the yield of EVs, we stressed the cells by serum deprivation. FBS deprivation has advantages from a translational standpoint for the elimination of a xenogeneic component from production protocols. On the other hand, serum deprivation can stress cells and change their physiological state, consequently affecting EV composition and production, which could then affect the therapeutic outcome (positively or negatively) [33]. When cells were deprived of FBS for 24 h, the yield of EVs was significantly decreased (Additional file 3).

To test whether our target lung cells were able to internalize EVs in culture, we stained EVs with a lipophilic dye and observed the uptake of vesicles in vitro in primary lung fibroblasts and a lineage of alveolar macrophages. Macrophages displayed efficient uptake of vesicles (Fig. 2c, e), as expected considering the optimal size of the vesicles for phagocytosis by these cells. Uptake was neither halted nor increased by previous activation of macrophages with silica particles (data not shown). When exposed to silica in the presence of EVs, macrophages of the RAW lineage produced lower amounts of TNF-α and TGF-β (Fig. 2g, h), suggesting inhibition of activation. Uptake of vesicles by fibroblasts was observed in foci for the same concentration of vesicles in vitro, with a lower percentage of cells containing the dye (Fig. 2d, f).

In addition, we investigated the distribution of vesicles in vivo after intratracheal instillation of EVs (Fig. 2a, b). The vesicles reached the alveoli and were especially concentrated closer to the airways. Both healthier areas and areas with intense inflammation presented fluorescence. We followed this uptake from 5 min to 1 h, observing the same fluorescence intensity throughout the parenchyma.

A single dose of AD-MSCs or EVs in the dose equivalent to the production of one million cells for 24 h (EV6) significantly reduced the fraction area of granuloma in the lungs of SIL animals (Fig. 3). Total tissue cellularity was significantly reduced inside the granuloma with the EV treatments (Table 1) but not in the alveolar septa (data not shown). Moreover, the number of macrophages was reduced by the treatment with AD-MSCs or EV6, both in the alveolar septa and inside the granuloma (Fig. 4a and b), thus showing modulation of the inflammatory response by both the cells and EVs. IL1-β and TGF-β expression in the lung tissue was decreased in the same groups (Fig. 5a, b) but not in the group that received the lower dose of EVs, whereas TNF-α was reduced with both concentrations of EVs but not by the cells (Fig. 5c). IL-6 was not significantly affected by silica or by the treatments (Fig. 5d).

Lung fibrosis is the most important consequence of silica inhalation. SIL groups presented increased collagen deposition compared with CTRL groups. Both interstitial fibrosis and the progressive accumulation of fibers around granulomatous tissue were reduced by the treatments with either AD-MSCs or EVs (Fig. 6a, b). Despite the decrease in collagen fibers with both AD-MSCs and EVs, only a mild effect on lung static elastance was observed. Treatment with the higher dose of EV6 was able to reduce this parameter significantly but without complete reversion to control values (Fig. 6c). No significant differences were observed in viscoelastic or resistive pressure coefficients with any of the treatments compared with untreated silicotic animals (Fig. 6d).

When considering a translational approach to the use of EVs, knowledge of EV stability under different storage conditions is needed. We measured the concentration and size distribution of AD-MSC EVs after 1 week of storage in different conditions: (1) room temperature in saline solution; (2) in bronchoalveolar lavage fluid (BALF) collected in PBS; (3) saline solution at 4 °C; (4) BALF at 37 °C; (5) saline solution at − 80 °C. No changes in the mean interquartile range or in the median diameter were observed after 1 week under the different conditions. All storage conditions caused a decrease in concentration of about 50%, but without significant changes in size (Fig. 7a). However, when the conditioned medium was frozen up to 3 weeks before ultracentrifugation, no changes in size distribution or concentration were observed (Fig. 7b). We then tested the ability of EVs to endure aerosolization, a property essential for the proposed local delivery of EVs to treat respiratory illnesses. We used a microsprayer syringe system to generate an aerosol from suspended EVs. There were no changes in the size parameters or concentration of EVs (Fig. 7c). This suggests that no major disruption or aggregation of vesicles occurs during aerosolization.