GLP-1 receptor agonist ameliorates experimental lung fibrosis

Idiopathic pulmonary fibrosis (IPF) is a chronic, progressive, and fatal lung disease. This disease is characterized by an excessive accumulation of extracellular matrix deposition that modify normal lung physiology. Up to date, there are not efficient therapeutic tools to fight IPF. Glucagon-like peptide-1 receptor (GLP-1R) activation plays an essential role in lung functions in normal and in pathological conditions. The aim of the present study was to study the possible beneficial effects of the administration of the GLP-1R agonist, liraglutide, in the pathogenesis of the fibrotic process in an animal model of pulmonary fibrosis induced by bleomycin. We observed that liraglutide decreased mRNA expression of collagen, hydroxyproline and key enzymes for the synthesis of collagen. In addition, GLP-1R activation restored the ACE2 mRNA levels modulating the activities of the RAS components, increased the production of surfactant proteins (SFTPa1, SFTPb, SFTPc) and promoted an improvement in pulmonary and cardiac functionality, including a partial restoration of lung alveolar structure. Liraglutide effects are shown at both the pro-inflammatory and fibrosis phases of the experimental disease. For these reasons, GLP-1 might be regarded as a promising drug for treating pulmonary fibrosis.

Liraglutide treatment in extracellular matrix synthesis. In the inflammatory phase (day 7), BLMadministered rats showed increased levels of the mRNA of collagen type I alpha 1 chain (Col1a1, p = 0.0001) and Fibronectin 1 (Fn1, p ≤ 0.0001, Fig. 2a). In this phase the LIR treatment did not modify the altered mRNA levels of Col1a1 neither Fn1 in BLM-administered rats (Fig. 2a). Bronchoalveolar lavage (BAL) soluble collagen levels were greatly increased after bleomycin administration (p < 0.0001, Fig. 2b) and LIR treatment was ineffective to restore them at that time (Fig. 2b). Total hydroxyproline (OH-Pro) levels were paradoxically decreased in day 7 after BLM instillation (p = 0.0178, Fig. 2c).
Effect of liraglutide treatment during fibrotic phase in extracellular matrix components. As we previously have shown, bleomycin instillation statistically increases the mRNA expression level of col1a1 and fn1 (Fig. 8a). Liraglutide treatment once fibrotic process is established, from day 10 to day 20, normalize mRNA expression of these two extracellular matrix elements (Col1a1  Fig. 8b), but not FN1 protein levels (Fig. 8c).

Discussion
Interstitial Pulmonary Fibrosis is a deleterious and progressive disease of the lung with fatal course and not well-understood pathophysiology. Although the incidence of IPF is modest, its rate is increasing in occidental countries, and about 80,000 in the U.S.A. and more than 300,000 patients in the E.U. are suffering this agonic illness. Recent data on Sars-Cov-2 pneumonia suggests a fibrotic sequel in some of the more severe cases 20 , what may increase the incidence in next years. IPF survival ranges 3.8 years from diagnostics, being shorter than many of common neoplasia. Up to date, there are not efficient treatments to fight this disease; despite few drugs have some palliative effects, as nintedanib and pirfenidone 5 . Therefore, the search of new therapeutic options for IPF is mandatory. Our previous and other authors´ work showed that GLP-1 and the agonists of the GLP-1R, highly expressed in pulmonary tissues, promote very potent effects in the lungs, affecting production and secretion of surfactant lipids 11,12 and proteins 13 , regulation of the components of the pulmonary RAS and vasoconstriction/ www.nature.com/scientificreports/ vasodilatation balance 14,15 ; and inflammatory mediators 16 . Since it becomes a challenging opportunity to test the GLP-1 receptor agonist, liraglutide, in an experimental model of IPF and if it has any potential therapeutic capacities in this dramatic disease.
In the present work, we studied the effect of liraglutide administration during the inflammatory phase of the well-established animal model of IPF induced by intratracheal instillation of BLM, and its consequences at two different moments: day 7 in the middle of the inflammatory phase induced by BLM instillation, and day 21 at beginning of the pro-fibrotic phase that develops after BLM-induced injury. In addition, we have studied the effect of liraglutide administration after finished the inflammatory phase and once the fibrosis process initiated upon fibrosis markers (Fig. 8).This model has been shown to have several advantages, highlighting that it resembles histological and functional changes of human lung fibrosis and fulfils the criteria for a good animal model of IPF 6 , is quite easy to perform, and widely accessible and reproducible. This model, reproduces two phases of the disease: an initial inflammatory phase ranging 8-10 days, and a subsequent second phase where collagen interstitial depot predominates, lasting up to 25-27 days after BLM instillation 7 . This model has been widely used to study experimental IPF and current palliative drugs were tested in it.
It has been previously described that the fibrotic progression is promoted by different cytokines. TGF-β1 is the profibrotic cytokine considered the most important molecule to stimulate extracellular matrix synthesis 21 . CTGF is also an extracellular matrix cytokine involved in profibrotic action of TGF-β1 22 . These two cytokines together drive the myofibroblast differentiation, the starring cell type involved in synthesis and secretion of extracellular matrix in fibrotic responses 4 . The presence of myofibroblast in the tissue might be revealed by identifying specific local markers. ACTA2, is the most specific, reliable, and used marker for the identification and quantification of myofibroblast during lung fibrosis 23 . BLM administration induced a differential expression of those cytokines and myofibroblast marker Acta2. Thus, in the initial inflammatory phase just Ctgf mRNA expression is increased without modifying Tgfb1 and Acta2 mRNA expression. In such phase, the inflammatory events predominate but Ctgf elevation reveals the start of underlying mechanisms which promote fibrosis. During active fibrotic phase, in day 21, the expression of these two profibrotic cytokines Ctgf and Tgfb1 and the myofibroblast marker Acta2 are expected to be markedly elevated, showing a complete pro-fibrotic profile. In our study, liraglutide treatment reverts the elevation of these three molecules, both pro-fibrotic cytokines and the myofibroblast marker Acta2. Our results confirmed the initial hypothesis that the GLP-1R activation has positive effects in fibrosis-promoting factors. Previously, it was shown that GLP-1R activation can decrease in a dose-dependent manner, the mRNA expression of Ctgf in primary culture of human mesangial cells 24 , which agrees with our results.
However, what is relevant and determines the course of the fibrotic process is the collagen depot in the interstitial space of lungs. This is part of the reparative response of the tissues to severe injury accompanied by inflammation. In many tissues and conditions, this reparative response is just partially regenerative substituting some of the original cells by collagen fibres. Collagen type I alpha 1 chain (COL1A1) is the best representative We have shown that Col1a1 expression levels and soluble collagen in BAL are increased from the beginning of the process including both the inflammatory and fibrotic phase in this model. Meanwhile OH-Pro tissue levels were just elevated in the fibrotic phase, revealing an accelerated production rate of collagen at that time. These findings indicate that the transcriptional mechanisms involved in collagen synthesis were activated previously to the beginning of the fibrotic phase. An increase in BAL soluble collagen levels already detected in the inflammatory phase is likely due to alveolar epithelial cells necrosis 28 before collagen depot. Liragutide treatment was able to block the synthesis and deposition of collagen in the fibrotic phase. Similarly, in other models of cardiac and renal fibrosis, the activation of GLP-1R was able to reduce the fibrotic process 18,29 . For first time, we have described this effect in a lung fibrosis model.
The enzyme prolyl 4-hydroxylase (P4H) plays a critical role to hydroxylate proline residues to 4-Hydroxyproline, that is essential for the stability of the collagen quaternary structure 30 . We have shown that the P4ha3 mRNA expression levels are increased in the two phases of the BLM model. There is a positive correlation between an increased collagen synthesis and P4H activity in animal models of pulmonary fibrosis and in other human diseases with a fibrotic process 31 . Therapeutic inhibition of this enzyme is today regarded as a promising strategy to prevent fibrosis. Some P4H inhibitors have been tested and they appear to reduce collagen deposition 32 . Up to date, these inhibitors are not clinically relevant due to their low potency, secondary toxicity and poor selectivity 30 . We here describe for first time that GLP-1R activation can inhibit P4ha3 markedly at transcriptional level and since it may constitute a new and promising therapeutic option for the treatment of fibrotic diseases.
In addition, we studied the local production of proline (Pro), an essential amino acid of collagen. It constitutes, together with his derivative OH-Pro, at least about 25% of the amino acid chain of collagen fibres 33 . In the healing reparative process of wounds, the local Pro levels become increase by 30-50% due to local Pro biosynthesis 34 . The main biosynthetic route involves the urea cycle. Supraphysiological supplementation with l-Arginine has been shown to increase collagen deposition in wounds 35 . There are three key enzymes involved in the local production of Pro: Arginase 1 (ARG1), ornithine aminotransferase (OAT) and pyrroline-5-carboxylate reductase (PYCR); of which we have studied the first and the last of this synthetic route. Bleomycin administration increases mRNA expression level of Arg1 only in fibrotic phase, but its enzymatic activity is increased in the two phases, what corresponds to the previously reported by Endo et al. 36 . For first time, we described that GLP-1R activation lowers www.nature.com/scientificreports/ this enzyme at the transcriptional and activity levels. On the other hand, PYCR has three different isoforms: type 1, 2 and 3. Only the Pycr1 isoform mRNA expression increased after bleomycin administration in the two phases and the others remained unchanged (data not shown). This suggest that PYCR1 is the critical isoform in Pro biosynthesis during pulmonary fibrosis. Liraglutide completely reverts the increase in the mRNA expression of Pycr1 in the fibrotic phase. Liraglutide was also able to modify the expression of the enzymes of Pro biosynthesis at transcriptional level, in a moment of the fibrotic process when its demand was elevated. In addition, fibronectin (Fn1) contributes to the stability of the collagen fibre network by stablishing crossingbridges between them, but it is involved in migration and cell adhesion to the ECM 37 , too. We observed that in BLM-model, Fn1 mRNA expression was increased in both phases studied, and during fibrotic phase (21D) liraglutide reversed this effect. This result suggests that LIR prevents not just the excessive depot of collagen in the tissue, but also the stabilization of the fibres in the ECM.
Complementary and very relevant the tissue structural changes reveal the pathophysiology of the fibrotic process. A specific collagen stain provides valuable information about the fibrosis development and allows an efficient evaluation of the tissue responses 38 . The BLM-model shows histopathological alterations that closely resemble human IPF 39 . Liraglutide was able to reverse at least partially the fibrosis, even at the structural level. It was observed a reduction in collagen fibres deposit, what corresponds with the reduction described in OH-Pro levels and Col1a1 expression. Liraglutide produces a remarkable structural improvement in pulmonary tissue in this model, which correlates with a functional amelioration also reflected by the increase in body weight observed in LIR-treated rats. Bleomycin administration decreases body weight gain due to alveolar epithelial cells necrosis and alveolitis 40 , and since reduced pulmonary function. Liraglutide administration was solidly proven to reduce food intake 41 , and this effect can affect negatively in the body weight of animals. However, we found an increment in the body weight gain in those ill rats by BLM-instillation, induced by LIR administration, which indicates a functional amelioration in the physiopathological process and a reparative response of the tissue.
Lung surfactant system is critical to maintain alveoli open and functional. Surfactant proteins (SFTPs) plays a fundamental role in lung surfactant stability. Bleomycin reduces mRNA expression of all SFTPs studied in inflammatory and fibrotic phase, as previously shown in this animal model 42 . Liraglutide administration during the inflammatory phase restores SFTPs mRNA expression to the level of control group in fibrotic phase. The capacity of liraglutide to stimulate the secretion and the production of lung surfactant improves alveolar stability and gas exchange, which contributes to a better respiratory function. Previously, we have shown that GLP-1 receptor agonists are able to restore the normal levels of mRNA of these proteins in other lung pathological states, such as foetal lung hypoplasia 13 and type 1 diabetes 14 .
IPF is accompanied by pulmonary arterial hypertension (PAH), which is an important predicting fatal prognosis factor in humans 43 . PAH affects right cardiac ventricle promoting hypertrophy, which itself is a major marker of functional altered status and prognosis in PAH, and consequently in IPF 44 . In the BLM-IPF model excessive fibres accumulation disrupts interalveolar interstitium and increase resistance to blood flow in the pulmonary vascular bed promoting PAH 45 and subsequent right ventricle hypertrophy. Liraglutide treatment prevented the development of right ventricle hypertrophy, as we and others described previously in models of lung pathology 14,17 .
In this context, the Renin Angiotensin System (RAS) is crucial in the regulation of cardiovascular system, and it has a special representation in the pulmonary vascular bed playing a fundamental role in the regulation of vessels´ tone. RAS is a complex set of molecules, enzymes, and receptors, organized in two different branches with eventual antagonist functions. Angiotensinogen is the precursor peptide of the RAS active molecules and expressed locally in lung tissue 46 . BLM markedly reduces Agt mRNA expression levels in inflammatory phase (day 7). These changes reveal different vascular events occurring in each time phase of the BLM model: in the first inflammatory phase, the cytokines produced by the affected tissue induce vasodilatation; meanwhile in the last pro-fibrotic phase the interstitial depot of collagen favours increased vascular resistance.
Bleomycin instillation decreases Ace (Angiotensin-Converting-Enzime) and Ace2 mRNA expression, and this decrease is even more marked on day 21. Previously, other studies had demonstrated that these enzymes are downregulated in animal models of IPF 47,48 , hypothetically due to the endothelial/epithelial injury. LIR treatment additionally decreases the expression of these two enzymes on day 7, but LIR remarkably reverted the expression levels of Ace and Ace2 on day 21. Previous studies of our group in different animal models, showed that LIR restores Ace and Ace2 mRNA expression levels in type 1 diabetes animals, and it consistently induced an overexpression of these two enzymes when administered to normal controls 14 .
Moreover, the angiotensin II receptor, type 1A (Agtr1a) mRNA expression decrease in the two phases of the BLM model, what agrees with the results of immunohistochemistry studies carried out in FPI patients 49 . In contrast, angiotensin II receptor, type 2 (Agtr2) mRNA expression was markedly increased during fibrotic phase, in agreement with other studies [49][50][51] . Liraglutide treatment also prevented this effect. Interestingly, GLP-1R agonists modulate AGTR2 in other tissues as shown in a model of cardiac fibrosis, where the expression of AGTR2 is reduced and GLP-1R activation increases its expression 19 . Even further, MAS1 is the receptor for Ang (1-7), the product of AngII cleavage by ACE2, and it plays a very important role in the vasodilatory, anti-apoptotic and anti-proliferative effects of Ang (1-7) 52 . We have previously reported that ACE2/Ang (1-7)/MAS1 axis was activated and enhanced in lung tissue of weaned pups after GLP-1R activation in an animal model of intrauterine growth restriction 15 . This branch of the pulmonary RAS axis is critical to understand the balance of vascular tone in the lung vascular bed. It fact, the increased activation after liraglutide treatment likely underlying the anti-hypertensive effects of GLP-1R agonists in pulmonary circulation and explains the reversion of the right ventricle hypertrophy we found in the BLM-model, as we previously describe in other models too 14,15 .
As a whole our results show, by the first time, that the administration of liraglutide in the precocious initial phase of lung fibrosis, reduces collagen interstitial deposition and production of precursor materials as proline and hydroxyproline, reduces presence of myofibroblasts in the tissues and the expression of pro-fibrotic cytokines Scientific Reports | (2020) 10:18091 | https://doi.org/10.1038/s41598-020-74912-1 www.nature.com/scientificreports/ as Tgfb1 and Ctgf. Moreover, liraglutide modulates de activity of the pulmonary angiotensins and their receptors improving lung vascular conditions and preventing right ventricle hypertrophy. In addition, liraglutide increase the expression and protein levels of key surfactant proteins, and all together improve the alveolar histological structure, finally ameliorating clinical condition of the animals. Since liraglutide arises as a new promising molecule with potential therapeutic applicability in lung fibrosis, awaiting to reproduce its effects in clinical studies.

Methods
Animals. All experimental procedures were performed in young adult (12 weeks) male Sprague-Dawley rats (200 ± 20 g), obtained from Animal Housing Facility of University of Santiago de Compostela (Santiago de Compostela, Spain). The animals were maintained in the Bio-experimentation Service of the University of Vigo (SB-UVI), on a 12/12 light cycle, at controlled room temperature (22 ± 2 °C) and humidity (50 ± 5%); with ad libitum access to water and standard food (A04; Panlab, Barcelona, Spain). Animals were randomly assigned for each experimental group. Experimental procedure. Rats were treated with liraglutide (LIR, 100 µg/kg per 12 h dissolved on 0.4% acetic acid, subcutaneously) or vehicle (VEH, 0.9% NaCl solution plus 0.4% acetic acid). They were treated using two different patterns. In the first pattern, animals were treated between day-1 and day 6 after endotracheal instillation of BLM. In the second pattern, they were treated between day 10 and day 20. On day 0, rats were intra-tracheal injected with BLM (2.5 mg/kg body weight, dissolved in 200 µL of sterile 0.9% NaCl) or 200 µL of 0.9% NaCl (CT), as previously described 53 . To ensure correct BLM administration, we used only the animals that showed a sustained weight loss at least until day 4. This model characteristically develops in two consecutive phases: a first phase predominantly inflammatory from day 1 to about day 9, and a second phase predominantly showing deposit of collagen in the lung interstitial from day 9 onwards. Groups of eight rats were killed on day 7 and 21 after BLM or NaCl instillation by an overdose of sodium pentobarbital (200 mg/kg, intraperitoneally, Sigma-Aldrich). Trunk blood was obtained directly from abdominal artery trough a puncture trough exposed artery. Blood was collected in a BD Vacutainer tube containing K3EDTA (15%) and aprotinin (250 KIU) (Beckton Dickinson, catalog number 361017), and stored in ice prior to further processing. Animal death was confirmed by abdominal artery exsanguination. Caudal lobe of right lung was removed and stored at − 80 °C for RNA and protein extraction. The medial lobe of right lung was also removed and stored at − 80 °C for OH-Pro quantification and arginase activity assay. Left lung was weighted and then immediately cannulated by the principal bronchia with a 22-gauge plastic cannula (Braun, Melsungen, Germany) and rinsed with 2 mL of cold sterile 0.9% NaCl solution to collect BAL. BAL was stored on ice prior to further processing.

Drugs.
The heart was excised, left and right atrium removed, and RV was dissected from LV + S and weighted separately. Fulton´s Index was calculated as the ratio of RV mass/LV + S mass, and it was used as marker of RV hypertrophy.
Broncho-alveolar lavage and blood processing. Blood tubes were centrifuged at 2000 g for 10 min at 4 °C. Supernatant was aliquoted and stored at − 20 °C for further analysis. 500 µL of the BAL collected from the animals was centrifuged at 300 g for 10 min at 4 °C. Supernatant was aliquoted and stored at − 20 °C for further analysis.
Analysis of soluble collagen. Total BAL soluble lung collagen content was quantified using Sirius Red stain absorbance method as previously described 54 . 100 µL of 6 standard-curve samples of Collagen from calf skin, ranging a concentration between 50-1.56 mg/dL (Sigma cat. C9791) and 100 µL of each sample were loaded in duplicate in a 96-well flat bottom plate. Then, 150 µL of collagen-binding dye (0.1% Sirius red stain [Direct red 80, Sigma-Aldrich, cat. 365548] in saturated picric acid solution) was loaded in each well and plate was incubated 60 min at 37 °C. The plate was centrifuged at 3000 g for 10 min, supernatant was discharged, pellet was washed by absolute ethanol for 2 min, and plate was centrifuged at 3000 g for 10 min. Supernatant was discarded and collagen-dye pellet was resuspended using 200 µL of 0.5 M NaOH and incubated at 37 °C for 30 min in obscurity. Finally, 96-well plate dissolved collagen dye pellets were measured spectrophotometrically at 540 nm using a plate reader EnVision (Perkin Elmer, MA, USA). Results were represented as µg of collagen per mL of BAL.
Quantification of TGF-β1 levels by ELISA. Plasma and BAL TGF-B1 levels were quantified by ELISA using a commercial kit (Human TGF beta-1 DuoSet ELISA, cat. DY240-05; and DuoSet ELISA Ancillary Reagent Kit 1, cat. DY007. R&D systems, MN, USA). 175 µL of plasma per sample was activated by acidification with 35 µL of 1 N HCl. Samples were incubated 10 min at room temperature and them they were neutralised using 35 µL of 1.2 N NaOH/ 0.5 M HEPES. 40 µL of plasma per sample was activated by acidification with 20 µL of 1 N HCl. Samples were incubated 10 min at room temperature and them they were neutralised using 20 µL of 1.2 N NaOH/ 0.5 M HEPES. Activated BAL Samples were used undiluted. Activated plasma samples were diluted to a final 20-fold dilution using reagent diluent provided by the kit. 100 µL of each standard and sample were loaded in a previously precoated 96-well plate following manufacturer´s instructions and the rest of the protocol has advanced following them. Results were represented as pg/mL for BAL samples and as ng/mL for plasma samples. Protein expression levels using Western-Blot. Lung homogenate´s total protein level was quantified espectrophotometrically using a commercial Bradford assay (cat. 5000006, Bio-Rad Laboratories, München, Germany). 30 µg of proteins were boiled at 95° with loading buffer; and then they were separated in a SDS-PAGE, using 7.5% precast gels (cat. 4561026, Bio-Rad Laboratories, München, Germany). Proteins separated in the gel were transferred to a PVDF membrane (cat. 1704156, Bio-Rad Laboratories, München, Germany) using a TransBlot Turbo apparatus (Bio-Rad Laboratories, München, Germany).
Membranes were blocked using a blocking solution (Tris Buffered Saline (TBS)-0.1% Tween 20 with 5% skimmed milk powder) for 1 h and at room temperature. Membranes were incubated overnight at 4 °C with the primary antibodies diluted in the blocking solution against BACT (1:5000, cat. A5441, Sigma-Aldrich); FN-1 (1:1000, cat. SAB4500974, Sigma-Aldrich) and GLP-1R (1:400, cat. sc-66911 Santa Cruz Biotechnology). Membranes were washed 3 times with TBS-0.1% Tween 20 and incubated1 hour at room temperature with a secondary antibody diluted in blocking solution. For BACT an anti-mouse IgG horseradish peroxidase conjugated (1:3000, cat. 170-6516, Bio-Rad Laboratories, München, Germany) and for FN-1 and GLP-1R an anti-rabbit IgG horseradish peroxidase conjugated (1:10,000, cat. 111-035-144, Jackson ImmunoResearch Europe, Ely, United Kingdom). Finally, membranes were washed 3 times with TBS-0.1% Tween 20 and membranes were incubated with the chemiluminescent substrate (Clarity Western ECL substrate, cat. 1705060, Bio-Rad Laboratories, München, Germany). Luminescence signal was visualized using Bio-Rad ChemiDoc XRS system and, in case of BACT and FN1, quantified with Bio-Rad ChemiDoc XRS Image Lab 3.0 software. The capture time was selected at the best point of the slope for signals (every 10 s), avoiding band saturation and allowing quantifications with a minimal variation. Band intensities are normalized using BACT as housekeeping protein, and data represents as percentage of variation versus CT/VEH group.

Hydroxyproline quantification.
A piece of right lung medial lobe was homogenized in deionized water (10 μL per mg of tissue) using a disperser. 600 μL of tissue homogenate was transferred to a Pirex tube with screw top, and same volume of 12 N HCl was added. Tissue homogenate was digested in an oven for 3 h at 120 °C. Afterwards, 1 mL of digested sample was translated to a glass tube that contains 15 mg of activated charcoal and 1 mL of deionized water. This tube was centrifuged at 400 G for 15 min, and 1 mL of the supernatant was disposed in a new glass tube. Samples were neutralized in presence of 10 μL of 1% phenolphthalein with 290 μL of NaOH 10 N. Hydroxyproline quantification was determined using a previously described method 54 . Results were represented as micrograms of Hydroxyproline per milligram of wet tissue.
Arginase activity assay. This assay was previously described by Wynn 54 with minor variations. A piece of lung medial lobe was homogenized in lysis buffer, and activated solution was incubated by arginase substrate solution for 24 h. After that, the urea generated was quantified using a commercial assay (UREA-37, Spinreact, Girona, Spain). Total protein levels were determined in lung homogenate using a commercial Bradford assay (cat. 5000006, Bio-Rad Laboratories, München, Germany). Data were represented as micrograms of urea generated in 24-h period, per mg of protein in lung homogenate.

Scientific Reports
| (2020) 10:18091 | https://doi.org/10.1038/s41598-020-74912-1 www.nature.com/scientificreports/ Histological analysis. In day 21, after BAL sampling, left lung lobe was excised in 3 mm-thick chunks and fixed by immersion in 10% neutral-buffered formalin for 24 h. After fixation, chunks were included in paraffin and 5 µm slices were sectioned using a microtome. The slices were stained using Masson's trichrome staining kit (Sigma-Aldrich cat. HT15) following the manufacturer´s instructions. Images were acquired using Nikon NiE microscope. To evaluate the severity of fibrosis 16 micrographs per sample were taken randomly at 200 ×. Images were analysed in double blind study by 3 different technicians. Each micrograph was scored using modified Ashcroft score that is described previously 58 . The average for the scoring for each sample of the 3 different technicians were represented.
Statistical analysis. Data representation and statistical analysis was performed using Graph Pad Prism 8.0.2 (GraphPad Software, LaJoya, CA, United States). The data were represented as mean ± standard error of the mean (SEM). Comparisons between experimental groups were analysed by two-way ANOVA. In the study was defined two different independent variables: fibrosis and LIR treatment. When the interaction between them where statistically significant a Bonferroni's multiple comparison test was performed. Statistical differences respect to CT/VEH and BLM/VEH groups were represented or two-way ANOVA result unless indicated. By convention, a value of p ≤ 0.05 was accepted as statistically significant.

Data availability
The datasets generated and analysed during the current study are available from the corresponding author on reasonable request. Reprints and permissions information is available at www.nature.com/reprints.
Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.