Male and female C57BL/6 mice (8-12 weeks) weighing 18-22 g were obtained from the Brazilian National Institute of Cancer (INCA, Rio de Janeiro, Brazil) and Multidisciplinary Center for Biological Investigation (University of Campinas, São Paulo, Brazil). Three-month-old male Syrian hamsters, weighing 110 to 120 g, were purchased from Anilab (São Paulo, Brazil) and maintained in our animal facility according to regulations given by the local ethics committee (CEUA UFRJ: 021/16 and 158/18). Animals were housed at constant temperature of 25°C under a 12-h light/dark cycle with free access to food and water. All procedures were performed according to the guidelines of the Committee on Ethical Use of Laboratory Animals of the Federal University of Rio de Janeiro (CEUA UFRJ: 092/14).

The pulmonary fibrosis model was induced by intratracheal instillation (i.t.) of BLM [Sigma Aldrich, Saint Louis, MO, USA; 0.06 U/mouse in a final volume of 30 μl of saline (SAL)] as previously described (22, 23). Mice were divided into three groups: 1) SAL group = mice received i.t. administration of SAL and i.v. administration of ethanol 0.1% and 0.01% (diluted in SAL) on days 7 and 10, respectively; 2) BLM group = mice received BLM i.t. injection on day 0 and i.v. administration of SAL on days 7 and 10, respectively; and 3) BLM + ATRvD1 group = mice received BLM i.t. injection on day 0 and i.v. administration of ATRvD1 (Cayman Chemicals, Ann Arbor, MI, USA) 2.5 μg/kg in 0.1% ethanol and 0.5 μg/kg in 0.02% ethanol on days 7 and 10, respectively. On day 14, bronchoalveolar lavage (BAL) was performed. The lungs were immediately removed postmortem and used for histological analysis, cytokine quantification by ELISA, and detection of primary cell-derived EVs.

The tracheas were clamped at the end-expiration, the lungs were removed en bloc and immersed in buffered formalin (10%) for 48 h before embedding in paraffin, and sections (5 μm thick) were obtained. Histologic sections were stained with hematoxylin and eosin for structural analysis or by Picrosirius red staining to evaluate matrix protein deposition, mainly collagen. The tissue sections stained with Picrosirius red were also analyzed by polarized light microscopy. Areas were photographed with an Olympus BX53 light microscope (Tokyo, Japan).

Lung sections were immunostained with antibodies against collagen type I (Santa Cruz Biotechnology, Santa Cruz, CA, USA) and alpha-smooth muscle actin (α-SMA; Abcam, Cambridge, MA, USA). Immunohistochemistry was performed as previously described (24).

For immunofluorescence, lung sections were immunolabeled with primary antibodies against α-SMA and fibronectin (Abcam, Cambridge, MA, USA) at 1:400 and 1:800 dilutions, respectively. The slides were incubated with fluorescent secondary goat anti-rabbit IgG H&L (Alexa Fluor^® 488; Invitrogen, Carlsbad, CA, USA) diluted 1:400 in blocking solution. The slides were counterstained with VectaShield with DAPI (Vector Laboratories, Burlingame, CA, USA) for nuclei staining.

BAL was performed as described elsewhere (25). Briefly, BAL was performed by the i.t. administration of phosphate-buffered saline (PBS) solution (1 ml), and after a thoracic massage, the instilled fluid was gently retracted to maximize BAL fluid recovery and to minimize shearing forces.

We performed total and differential counts in the BAL fluid. The cells were diluted in Turk’s solution (1:10), and the total leukocyte count was performed in a Neubauer chamber. Mononuclear cells and neutrophils were quantified on Diff-Quik-stained cytocentrifuge slides (Cytospin 3; Shandon, Thermo Fisher Sientific, Waltham, MA). The data were expressed as the number of cells (×10⁵) per ml.

In order to determine fibrosis, the right lung tissue was prepared for hydroxyproline assay as previously described by Woessner (26). Lung hydroxyproline content was determined spectrophotometrically by absorbance at 550 nm, and the results were expressed as nanograms of hydroxyproline per milligram of tissue.

Lung samples were prepared from the whole lung homogenized in 1 ml of PBS containing a mixture of protease inhibitors (Complete; Sigma-Aldrich, Switzerland). The levels of interleukin (IL)-1β, MMP-9, monocyte chemoattractant protein-1 (MCP-1/CCL-2), tissue inhibitor of metalloproteinases 1 (TIMP-1), TGF-β, and tumor necrosis factor-alpha (TNF-α) in the lung samples were measured by ELISA (R&D Systems, Minneapolis, MN, USA), according to the manufacturer’s instructions. The results were expressed as ng/mg of protein (27).

Lung tissues were cut into small pieces of 1 mm³ and then incubated in collagenase I 0.2% (Sigma-Aldrich) for 1.5 h at 37°C. After the incubation, the samples were centrifuged, and the pellet was incubated with ammonium chloride potassium bicarbonate (ACK) lysing buffer for blood cell depletion. After centrifugation, cells were resuspended in DMEM F12 medium, supplemented with 10% fetal calf serum (FCS) and 100 U/ml of penicillin/streptomycin. The cell suspension was subsequently transferred to a flask and placed at 37°C in a 5% CO₂ incubator, and the experiments were performed from the third to the eighth passages. Primary fibroblasts (1 × 10⁵ cells/ml) were plated with DMEM F12 medium (unstimulated) and incubated with TGF-β (10 ng/ml) or with TGF-β + ATRvD1 (100 ng/ml) for 24 h.

Alveolar macrophages were collected via BAL. After cell count, the cells were resuspended in DMEM F12 medium with 10% FCS and adhered for 2 h. Thus, cultures were washed and incubated for further 18 h to reduce the presence of B1 cells. Macrophages were cultured in medium alone (unstimulated) or with TGF-β (10 ng/ml) or with TGF-β + ATRvD1 (100 ng/ml) for 24 h. Culture purity was assessed by flow cytometry with 95% of macrophage cells (F4/80⁺ cells). The supernatants obtained were subjected to EV analysis by flow cytometry.

Plasma or conditioned media from macrophages were collected and centrifuged at 250×g for 10 min, and the supernatants were stored at −20°C. To obtain the EVs, the supernatant from plasma or conditioned media were (ultra)centrifuged at 100,000×g for 4 h. The pellets were resuspended in annexin V buffer or FACS buffer and were stored at −20°C until use. The EVs were assessed using an Accuri™ C6 Flow Cytometer (Becton Dickinson, Franklin Lakes, NJ, USA) through the evaluation of annexin V (Thermo Fisher Scientific, Waltham, MA, USA), F4/80-PE (Biolegend, San Diego, CA, USA), SinglecF-alexa488, LY6G-APC, and CD3-alexa488 (BD Pharmingen, San Diego, CA, USA)-positive events. The gated region containing the EVs was previously delimited by using known concentrations of 1-µm beads to estimate the precise amount of EVs, as previously described (28).

The lungs were collected on day 14 after SAL or BLM installation. Briefly, the lungs obtained from each mouse were digested by collagenase A (1 mg/ml; Sigma-Aldrich) for at least 40 min, with shaking at 37°C. Phosphate-buffered saline containing 10% fetal bovine serum was added to inactivate the enzyme, and the solution was passed through a 40-µm cell strainer. Cells (10⁶ cells/ml) obtained from the lungs or BAL were incubated with LIVE/DEAD BV510 according to the manufacturer’s instruction (Life Technologies, Carlsbad, CA, USA) followed by anti-mouse CD16/32 Fc block (eBioscience, San Diego, CA) for 15 min at 4°C. Different cell populations were identified using the following corresponding antibodies: a) Ly6C-FITC, Ly6G-Alexa647, CD11b-Pecy7, CD11c-PECy5, and SinglecF-PE (monocytes, neutrophils, eosinophils, and alveolar macrophages) or b) CD11b-APC, CD11c-PECy5, SinglecF-FITC, CD86 PECy7, and CD206-PE (alveolar macrophage activation). Samples were acquired by BD LSRFortessa X-20 Cell Analyzer (BD Biosciences, San Rose, CA) and then analyzed by FlowJo software (Ashland, OR, USA).

Total RNA was extracted from lung tissue or 10⁶ primary mouse lung fibroblasts with the TRIzol reagent (Sigma) following the manufacturer’s procedures. Two micrograms of RNA was used for cDNA synthesis using the High-Capacity cDNA Reverse Transcription Kit (Life Technologies, Carlsbad, CA, USA). Quantitative real-time RT-qPCR was performed with the SYBR-green fluorescence quantification system. The PCR cycling parameters were 95°C (10 min) and then 40 cycles of 95°C (30 s) and 60°C (1 min), followed by the standard denaturation curve. The primer sets were as follows: GAPDH forward (F): 5′AGGTCGGTGTGAACGGATTTG; GAPDH reverse (R): 5′TGTAGACCATGTAGTTGAGGTCA; ALX (F): 5′GCCTTTTGGCTGGTTCCTGTGT; ALX (R): 5′CAAATGCAGCGGTCCAAGGCAA. ALX to GAPDH relative expression was calculated using the comparative Ct method.

Animals were anesthetized with an i.p. injection of ketamine (200 mg/kg) and xylazine (10 mg/kg) (IBCCF, protocol: 021/16 and 158/18). Then, using a cotton swab, the hamster cheek pouch (HCP; left) was inverted and mounted for intravital microscopy observation (without dissection) in an appropriate plate and topically treated with BLM (1 U/100 μl/hamster) with a sterile syringe (26½ G needle). The contralateral HCP received the same volume of SAL. Alternatively, 3 h post-BLM injection, some animals were treated with ATRvD1 (25 µg/kg - 100 µl i.v.) or SAL. After 24 h of BLM application, the animals were once again anesthetized and supplemented with i.v. α-chloralose (2.5% solution in SAL) through a femoral vein catheter. A tracheal cannula (PE 190) was inserted to facilitate spontaneous breathing, and the body temperature was maintained at 37°C by a heating pad monitored with a rectal thermometer. The HCP was prepared and used for intravital microscopy as described (29, 30). Then, the microcirculation of the HCP was assessed after i.v. injection of macromolecular tracer FITC-dextran 150 kDa (100 mg/kg; TdB Consultancy, Uppsala, Sweden) using an Axioskop 40 microscope, objective ×4, and oculars ×10, equipped with appropriate filters (490/520 nm) and a Colibri 2 LED-light source (Carl Zeiss, Germany). A digital camera, AxioCamHRc, and a computer with the AxioVision 4.4 software program (Carl Zeiss) were used for image recording and the subsequent measurement of fluorescence (relative fluorescent units, RFU). The methods for automatic characterization of the HCP microvasculature from intravital microscopy images were published by Bulant and colleagues (31). The image resolution is 1,388 × 1,040 for an area of approximately 5 mm², which yields a pixel spacing of 1.862 µm and, therefore, a pixel area of 3.467 µm². At the end of each experiment, the animals were euthanized by an overdose of ketamine and xylazine or an i.v. injection of KCl 3 M.

Differences between groups were analyzed by one-way ANOVA, followed by the Bonferroni post-hoc test. For all analyses, data were expressed as mean ± SEM. The GraphPad Prism 5 statistical software package (GraphPad Software, La Jolla, CA, USA) was used. A P-value <0.05 was considered significant.

In order to evaluate the effect of ATRvD1 treatment on BLM-established inflammation/fibrosis, mice were treated on days 7 and 10 after BLM (0.06 U/mouse) with ATRvD1 (25 and 5 μg/kg, respectively; i.v.) and sacrificed on the 14^th day, and the lung architecture was analyzed by histology. As observed in Figure 1A , an intense reaction occurred after the BLM challenge (middle panel), characterized by a high degree of inflammation, with cellular accumulation mainly in the interstitium, thickening and edema of the alveolar septa, and a marked reduction in open alveoli numbers compared with SAL-injected animals (left panel). Remarkably, ATRvD1 administrations after BLM resulted in lung structure re-establishment. Despite the tissue restructuration caused by ATRvD1 treatment, it is important to note that under the treatment approach adopted, we still can observe moderate signals of injury and inflammation, with the presence of inflammatory cells in the parenchyma and septa edema when compared with the SAL group (right panel).

Having determined that ATRvD1 treatment protects lung structures, we evaluated its effect on the cellularity, cytokine production, and MMP-9/TIMP-1 release in lung tissue evoked by BLM administration. The administration of BLM induced an increase in total cell recovery from the BAL fluid ( Figure 1B , left graph), most of them being mononuclear cells (∼7.5 × 10⁵/ml, approximately 88% of recovered cells; middle graph) and minor extension neutrophils (∼1.1 × 10⁵/ml, approximately 11%, right graph). Interestingly, ATRvD1 treatment significantly decreased cell recruitment induced by BLM, as observed for total cells, mononuclear cells, and neutrophils ( Figure 1B ). Next, we observed that BLM-induced TNF-α, MCP-1/CCL-2, IL-1β, and TGF-β protein release was inhibited by ATRvD1 administration ( Figure 1C ). Additionally, ATRvD1 treatment was also able to impair BLM-induced MMP-9 expression, without affecting TIMP-1 expression ( Figure 1C ).

Using flow cytometry analysis, we confirmed the results obtained by the cell counter, in which BLM administration induced an increase in total cells, and ATRvD1 treatment significantly decreased BLM-induced cell recruitment into the lungs and BAL fluids ( Figure 2A ). Moreover, we observed an increase in Ly6C^highLy6G⁻/CD11b⁺ monocytes ( Figure 2B ), SiglecF^highCD11c⁻ eosinophils, and SiglecF^highCD11c^high alveolar macrophages ( Figure 2C ) in the lungs and BAL fluids and an increase in Ly6G^highLy6C^int/CD11b⁺ neutrophils ( Figure 2B ) in the lungs of BLM-challenged mice. ATRvD1 treatment significantly decreased BLM-induced neutrophil, eosinophil, and alveolar macrophage accumulation in the lungs, while monocyte and alveolar macrophage recruitment was reduced in BAL fluids ( Figures 2B, C ). In order to investigate if ATRvD1 could interfere with macrophage polarization, we analyzed CD86 (M1 marker) and CD206 (M2 marker) expressions. We observed that the BLM challenge caused an increase in the number and MFI of CD86⁺ and CD206⁺ cells in the lungs and BAL fluids. ATRvD1 treatment decreased the number and MFI of CD86⁺ and the number of CD206⁺ macrophages in the lungs, without affecting their expression on alveolar macrophages (BAL) ( Figure 2D ).

The impact of ATRvD1 administration on BLM-induced extracellular protein deposition was evaluated by different strategies. In Figure 3 , lung sections were stained by Picrosirius red, a technique commonly used to evaluate the presence of collagen fibers. As observed by ordinary light microscopy (panel A, upper line) and with polarized light (panel A, bottom line), BLM injection induced matrix protein deposition compared with the control group (SAL-injected mice), and ATRvD1 treatment reduced BLM-induced lung matrix protein deposition to a similar condition observed in the vehicle-injected mice. Substantially, we observed that ATRvD1 treatment also inhibited BLM-induced collagen deposition in lung tissue homogenates, indirectly evaluated by hydroxyproline contents ( Figure 3B ). Once again, the hydroxyproline amount observed in the ATRvD1 group was similar to those observed in the control mice. Furthermore, the protective effect of ATRvD1 in impairing BLM-induced matrix protein deposition was also confirmed by specific immune staining for collagen I. As observed in Figure 3C , collagen I staining in the lung of SAL-injected mice was limited to peribronchiolar areas; meanwhile, in BLM-treated mice, collagen I staining was spread over large areas of the parenchyma. Interestingly, ATRvD1 treatment was able to reduce collagen I deposition induced by the BLM challenge ( Figure 3C ). Regarding the effect of ATRvD1 on BLM-induced matrix protein deposition, we expanded our investigation to evaluate fibronectin contents in in-vivo experiments. As illustrated in Figure 3D , the BLM challenge induced the expression of fibronectin in the lung tissue compared with the control group, and similar to what occurred to the collagen contents, fibronectin expression was inhibited by ATRvD1 treatment, showing a similar intensity to those observed in the control group. In addition to our analyses on the 14th day, interestingly, on day 7 after BLM, when mice received the first administration of ATRvD1, it was already possible to detect cell recruitment, collagen deposit, and structural tissue damage in the lungs ( Supplementary Figure 1 ).

Furthermore, we tested the effect of ATRvD1 treatment on BLM-induced α-SMA expression, an index of fibroblast differentiation into myofibroblasts. Immunohistochemistry analysis of the lung sections revealed that BLM caused an increase in α-SMA expression compared with the SAL-injected mice, which was reversed by ATRvD1 administration ( Figure 3E ). Next, to confirm the ATRvD1 effects on fibroblasts, we cultured primary isolated lung fibroblasts with and without TGF-β and evaluated the in-vitro differentiation via α-SMA expression. As demonstrated in Figure 3F , TGF-β-stimulated fibroblasts presented an increase in α-SMA expression compared with non-stimulated cells, whereas ATRvD1 co-incubation impaired myofibroblast differentiation.

Next, we investigated the ALX receptor mRNA expression in the lungs harvested from the control, BLM, and BLM-injected mice treated with ATRvD1. We detected the message for ALX in all groups; however, the mRNA expression was reduced in the lungs obtained from the BLM and ATRvD1-treated BLM samples ( Figure 3G ). Moreover, the incubation of fibroblasts with different TGF-β concentrations (10, 50, and 100 ng/ml) did not modify ALX mRNA expression ( Figure 3H ).

We evaluated the influence of ATRvD1 treatment on the production and release of EVs. BLM injection duplicated the EV contents in BAL compared with control mice. Interestingly, ATRvD1 administration impaired BLM-induced EV release, restoring the EV numbers to those observed in the BAL of SAL-treated mice ( Figure 4A , left panel). In addition, using an in-vitro strategy, in which macrophages were cultured with or without the presence of TGF-β, we also tested the ATRvD1 effect. As noted in Figure 4A (right panel), macrophage incubation with TGF-β caused an increase in the number of EVs released to the conditioned medium compared with the supernatants from untreated cells, while ATRvD1 incubation impaired TGF-β-induced EV production.

Additionally, in an attempt to determine the cellular source of the EVs, we first used 1-μm standard beads to define the gate region strategy as shown in panel B of Figure 4 . We observed that BLM induced an increase in F4/80⁺, SiglecF⁺, Ly6G⁺, and CD3⁺ EVs recovered from BAL fluids. Remarkably, ATRvD1 treatment decreased just F4/80⁺ and SiglecF⁺ EV release but did not modify the event number of Ly6G⁺ and CD3⁺ EVs ( Figure 4C ). Our results uncovered a new mechanism exerted by ATRvD1 in macrophages, as an important modulator of EV formation/release.

Evidence in the literature demonstrates that angiogenesis is a crucial step in fibrosis development (32). Moreover, it is also demonstrated that RvD1 presents an anti-angiogenesis effect in a model of corneal neovascularization (33). Using the intravital fluorescence microscopy technique in the HCP model, we observed that BLM administration in loco promoted an intense angiogenesis evaluated at 24 h by FITC-dextran (fluorescent tracer), compared with SAL administration ( Figure 5A ). ATRvD1 treatment prevented BLM-induced neovascularization ( Figure 5A ). These data were quantified by the total fluorescence measurement as shown in Figure 5B . The contralateral cheek pouch in the BLM-induced hamster did not present angiogenesis ( Figure 5A ). Our data demonstrated that ATRvD1 seems to prevent angiogenesis, contributing to the antifibrotic effect of BLM.

Several parameters can be obtained from the intravital microscopy technique (31, 34). Among them, we analyzed the vascular area and the total vascular length. BLM induced a considerable increase in the vascular area compared with the values observed for the SAL-treated HCP or in the contralateral BLM-treated HCP ( Figure 5C ). Meanwhile, ATRvD1 treatment recovered the BLM effect ( Figure 5C ). Similarly, when we analyzed the total vascular lenght, we also observed that BLM induced an increase in total vascular length compared with SAL-treated animals, which was inhibited by ATRvD1 treatment, however it did not reach statistical difference ( Figure 5D ).