Repeated instillations of bleomycin in male mice induce a perivascular remodeling at distance of fibrotic areas in the context of progressive pulmonary fibrosis

Abstract

Idiopathic pulmonary fibrosis (IPF) is the archetypal form of progressive pulmonary fibrosis (PPF), as well as being the most common and severe type of interstitial lung disease. IPF is characterized by progressive, diffuse interstitial remodeling, stiffening of the lung parenchyma and is also most often associated with pulmonary hypertension (PH), which significantly impacts the clinical course and outcome of the disease. To improve targeted therapeutic options and survival predictions, as well as offering an overall better understanding of the disease, researchers need models that best reflect the characteristics observed in patients and allow the study of comorbidities associated with pulmonary fibrosis. Chronic pulmonary fibrosis was induced in male C57BL/6 mice through intratracheal instillation (ITI) of 3 low‐doses of bleomycin (BLM) every 2 weeks. Mice were sacrificed 15 or 90 days after the last ITI. Clinical follow‐up and histological analysis were performed to study lung and perivascular remodeling. Our results suggest that fractionated bleomycin instillation generates a persistent and progressive pulmonary fibrosis in mice. This is associated with collagen remodeling in the perivascular space outside fibrotic areas. We hypothesize that this remodeling, accompanied by a mild but significant increase in pulmonary artery media thickness, could be a precursor to pulmonary fibrosis‐associated hypertension. These results suggest that this model could improve our understanding of pulmonary remodeling and vascular lesions in the early stages of pulmonary fibrosis.

We established a model of progressive pulmonary fibrosis via repeated intratracheal instillation (ITI) of leomycin (BLM) (A). Perivascular remodeling was observed in areas distant from the fibrotic areas and lesions induced by repeated ITI of BLM simulate those observed in IPF patients (B). Thus, this model has the capacity to track and investigate the pattern and spatial distribution of fibrotic lesions over time and help us to better understand and treat pulmonary remodeling and vascular lesions in their early stages. Experimental design presented in this graphical abstract were created using BioRender.com.

1. INTRODUCTION

For many years, the classification of interstitial lung disease (ILD) has been highly compartmentalized, with the different types being formally and practically distinguished. ¹ However, patients with different fibrotic ILD types can develop a similar phenotype, characterized by self‐sustaining fibrosis, irreversible loss in lung function, and premature death despite appropriate treatment. ¹ , ² The concept of progressive pulmonary fibrosis (PPF) has emerged and represents a devastating class of diseases that could become an important public health issue. Idiopathic pulmonary fibrosis (IPF) is the most common and severe form of ILD and is considered the archetypal form of PPF.

IPF is most often associated with comorbidities present before or appearing concomitantly or subsequently after progression begins that significantly impact the clinical course and outcome of the disease. ³ , ⁴ For example, several microvascular and macrovascular impairments have been identified in IPF. ⁵ These alterations can lead to the development of pulmonary hypertension (PH), a severe comorbidity that arises after the onset of IPF. ³ , ⁴ To our knowledge, there is currently no reliable model for studying the mechanisms involved in the progression of pulmonary fibrosis and associated PH. However, developing this type of model is important in order to consider the development of therapeutic strategies. Indeed, there are limited therapeutic options for these patients who usually have a poor prognosis and a reduced survival rate. The most widely used model of pulmonary fibrosis is the experimental mouse model, in which fibrosis is induced by a single intratracheal instillation (ITI) of bleomycin (BLM). ⁶ , ⁷ However, this model has several drawbacks, particularly with regard to the distribution and characteristics of the lesions, as well as spontaneous resorption. ⁷ , ⁸ In addition, associated comorbidities, especially vascular remodeling potentially leading to pulmonary hypertension (PH), are rarely described in this model. Recent studies have investigated the effect of repeated instillation of BLM on the onset of the disease. ⁸ , ⁹ , ¹⁰ , ¹¹ The distribution of fibrotic lesions across the different lobes of the mouse lungs was found to be more uniform and persistent. ¹⁰ , ¹¹ The present study aimed to propose the use of this model to evaluate comorbidities associated with progressive pulmonary fibrosis, especially those involving perivascular remodeling.

2. METHODS

2.1. Animals and ethical approval

The experimental procedures were conducted in 51 males C57BL/6 mice, aged 8 weeks and weighing 25 ± 2 g at the beginning of experimentation. The animals were acclimatized for at least 7 days to the local animal facility and housed in standard conditions in a 12 h/12 h light/dark cycle, at an ambient temperature of 20–22°C and had ad libitum access to water and food.

2.2. Experimental design

The mice received 3 ITIs at two‐weekly intervals with either phosphate‐buffered saline (PBS) (n = 23) or bleomycin (BLM, Sigma‐Merck, Saint‐Quentin‐Fallavier, France, 0.8 IU/g) (n = 28) to induce lung fibrosis (Figure S1A). ⁹ The mice were sacrificed 15 days (Figure S1A, PBS + 15d, n = 12; BLM + 15d, n = 12) or 90 days (PBS + 90d, n = 11; BLM + 90d, n = 16) after the third ITI in order to appreciate the course of the disease. At the end of the experiment, mice were sedated by intraperitoneal (IP) injection of a mixture of ketamine 1000 (Virbac, 06510 Carros, France) and xylazine Rompun 2% (Bayer Healthcare, Gaillard, France) (100 and 20 mg/kg, respectively), in order to collect the lungs for histology and SIRCOL evaluation (n = 26) or to evaluate lung compliance (n = 25).

2.3. Collection of lung samples

Lung samples were collected as previously described. ⁹ , ¹² Mice were tracheotomized, exsanguinated, and perfused with PBS via the right ventricle. The right lobes were ligated, and the left lobe was inflated with 4% paraformaldehyde (PFA) at 20 cm H₂O through the trachea, then both lungs and heart were removed. Right lobes were snap‐frozen for RNA/protein analysis; and the left lung was fixed in 4% PFA for 24 h, paraffin‐embedded, and sectioned (5 μm) for histology as previously described. ¹² , ¹³

2.4. Special staining of tissue sections

After deparaffinization and rehydration through graded ethanol solutions, lung slides were stained with Masson's trichrome, Sirius red–fast green, hematoxylin–eosin or Movat's pentachrome methods (see below). After staining, the sections were dehydrated and mounted prior to analysis.

2.4.1. Masson's trichrome staining and quantification of fibrotic areas

The slides were stained using Masson's trichrome method: nuclei with Weigert's iron hematoxylin (Roth, X907.1), cytoplasm and muscle with Biebrich scarlet‐acid fuchsin (Merck, CI 42685), and collagen (types I and III) with aniline blue (VWR, 34015182). The collagen in the lung parenchyma was quantified by measuring aniline blue‐stained areas as a percentage of total tissue surface by using the plugin « ColorDeconvolution2 » in the ImageJ software.

2.4.2. Sirius red–fast green staining and analysis of the periarterial region

The collagen was stained with 0.1% Sirius red and epithelia were counterstained with 0.1% fast green. Stained lung sections were imaged under white and polarized light (× 400) and analyzed using ZEN software (ZEISS). The arterial and periarterial regions were delineated by marking the internal vessel perimeter under polarized light (Figure S1B, step 1, white line). Then, a 200‐pixel‐wide surrounding area was selected (Figure S1B, step 1, red line). Binarisation of both the birefringent pixels and the contour to give the perimeter in this area was obtained using the tool “plugin selection” in the Image J software (Figure S1B, steps 2 to 4).

2.4.3. Movat's modified methods and hematoxylin–eosin for the analysis of changes in arterial structure

To examine the components of the pulmonary arteries in proximity to the airways in details, two consecutives were used. First sections were stained using Movat's method with alcian blue pH 2.5 (Sigma‐Aldrich, AG3157‐25G), Verhoeff's solution (modified Weigert hematoxylin, Roth, 3816.3), alcoholic saffron (Gatinais saffron) and acid fuschin (Sigma, F8129). ¹⁴ , ¹⁵ The second sections were stained using Harris hematoxylin (Sigma‐Aldricht, MHS16, Saint‐Quentin‐Fallavier, France) and an aqueous eosin solution (Merck, Art.1144, Saint‐Quentin‐Fallavier, France) for assessing the media thickness in pulmonary arteries close to the airways. The mean media thickness was obtained by the calculation of the length and segment using ImageJ (https://imagej.net/ij/).

2.5. Lung collagen assessed by Sircol® assay

The amount of soluble pulmonary collagen, comprising the collagen released after enzymatic digestion with 0.1 mg/mL pepsin (516360‐500MG, EMD Millipore, MERK, Saint‐Quentin‐Fallavier, France), was measured on the right superior lobe of different groups using the Sircol test, in accordance with the manufacturer's instructions (Biocolor Ltd.®).

2.6. α‐SMA immunofluorescence (IF)

IF was performed on 5 μm‐thick, deparaffinized tissue sections, using an anti‐mouse alpha‐smooth muscle actin (α‐SMA) antibody (red; A5228, Sigma‐Aldrich; 1/500) and the nuclei were counterstained with DAPI (blue). Briefly, after antigen retrieval in a boiling citrate buffer, sections were incubated with 5% of normal horse serum for 1 h to block nonspecific antibody binding sites. The sections were then incubated with the primary antibodies overnight at 4°C. The following day, the slides were incubated with secondary antibodies conjugated to a fluorochrome for 1 h (Alexa Fluor™ 546; A11003, Invitrogen; 4 μg/mL). The periarterial space was examined for the presence of anti‐α‐SMA‐stained cells.

2.7. Lung compliance measurement

At the end of the experimental procedure (i.e. the last ITI + 15d or + 90d), the mice were sedated via an IP injection of a mixture of 100 mg/kg of ketamine 1000 (Virbac, 06510 Carros, France) and 20 mg/kg of xylazine Rompun 2% (Bayer Healthcare, Gaillard, France), and then tracheotomised and ventilated for compliance measurement. Lung dynamic compliance (C_COV, mL/cmH₂O) was evaluated by plethysmography (Emka technologies, Paris, France) and the covariance as previously described. ⁹ , ¹² Briefly, the anesthetised mice were ventilated and a differential pressure transducer was used to obtain a flow signal reflecting the expansion and contraction of the thorax during each ventilation cycle. Compliance was calculated following acquisition of the flow signal and pressure signals. The volume was computed by integrating flow, proportional to the pressure difference across the pneumotachograph, the sole airflow route.

2.8. Statistical analysis

Graph and statistical analysis were performed with GraphPad Prism software (version 10). Data were presented as the mean ± standard deviation (SD). The normality of the data distribution was verified using Shapiro–Wilk and Kolmogorov–Smirnov tests. Comparisons among groups were assessed through two‐way ANOVA followed by the post‐hoc Tukey test. p < 0.05 differences were considered significant.

3. RESULTS

3.1. Repeated ITI of bleomycin led to progressive pulmonary fibrosis

Parenchymal lung lesions were characterized by an increased in collagen deposition distributed throughout the whole lung (Figure 1A,B). In the BLM + 15d group, remodeling of the lung tissue was observed, with lesions being present throughout the lung, predominantly in subpleural (Figure 1A) and perivascular areas (Figure 1B). In the BLM + 90d group, a worsening of these lesions was observed, with their distribution extending across the entire lung surface. In the BLM + 15d group, an increase in the total collagen fraction was observed (Figure 2A,B), which continued to evolve and have spread 90 days after the last ITI, as assessed by the Sircol assay (Figure 2A) and the aniline blue quantification (Figure 2B) following Masson's trichrome staining. This increase was associated with a decrease in lung compliance, which was observed 15 days after the final ITI of BLM (see Figure 2C). However, no significant additional decrease was demonstrated 3 months later.

FIGURE 1.

(A, B) Representative photographs of 5 μm thick Masson's trichrome‐stained lung sections, taken close to the pleura (A) and centred on the perivascular area (B) at × 400 magnification (scale bar = 100 μm; Zeiss microscope). This allows fine analysis of the collagen deposits stained by aniline blue. Yellow stars indicate areas of dense fibrosis. (C, D) Representative photographs of 5 μm thick lung sections stained with Sirius red–fast green close to the pleura (C) and centred on the perivascular area (D) in polarized light at × 400 magnification (scale bar = 100 μm; Axioscan 7). Yellow stars indicate areas of dense fibrosis. Yellow arrows indicate perivascular collagen deposits.

FIGURE 2.

(A) Total lung soluble collagen quantification by Sircol® in the right lung of mice. (B) Analysis of lung parenchyma collagen. The percentage of collagen labeled with blue aniline was quantified in lung sections stained with Masson's trichrome (panels A, B) compared to the total surface area of the analyzed lung tissue (Image J). (C) Lung dynamic compliance was measured using the covariance method (C_COV, mL/cmH₂O; Emka System®). This was measured in anesthetised and intubated mice after mechanical ventilation. (D) The ratio of the birefringent pixels to pixels defining the perimeter corresponds to the quantification of the analysis described in Figure S1B. Data are expressed as mean ± SD (G, H, I, J). ns, *, **, *** and**** indicate, respectively, not significant, p < 0.5, p < 0.1, p < 0.001 and p < 0.0001.

3.2. Progressive pulmonary fibrosis was associated with periarterial remodeling

An increase in perivascular collagen was observed away from the areas of fibrosis, particularly around the pulmonary arteries (Figure 1C,D). Periarterial collagen quantification was performed after Sirius red staining and examination under polarized light. An increase in collagen fibers wrapping around the vasculature was observed 15 days after the last ITI (BLM + 15d, Figure 2D). Furthermore, a population of fluorescent, anti‐α‐SMA‐stained cells was observed within the periarterial space of non‐fibrotic lung areas at 15 and 90 days after the last BLM ITI (Figure 3). This was not observed in the PBS control groups.

FIGURE 3.

Immunofluorescence staining of 5 μm‐thick lung tissue sections labeled with an anti‐mouse alpha‐smooth muscle Actin (α‐SMA) antibody (red) and nuclei counterstained with DAPI (blue) at × 400 magnification (scale bar = 100 μm; Zeiss microscope). The yellow arrow indicates α‐SMA‐positive cells around arteries in mice that received ITI of bleomycin, but not in the PBS control groups.

3.3. Mild, yet significant remodeling of the pulmonary arteries is observed in the same areas

Histologically, no arterial structural lesions were observed in BLM conditions when compared with the control group. Notably absent were endothelial cell anomalies such as swelling, nuclear hyperchromasia, cuboidal modification and intimal thickening. We examined the elastic fibers of the pulmonary arteries in proximity to the airways using Movat's method (Figure 4A, black arrows show elastic fibers and stars indicates pulmonary artery closed to airway). In BLM group (Figure 4A2–3), there were no apparent differences as compared to the controls (Figure 4A1): the two elastic laminae were clearly visible with no duplication or fragmentation of the elastin. However, when we evaluated the media thickness in HE sections (Figure 4B1–4) using a high‐magnification microscopic examination (method explained in the Figure 4 legend), a trend towards increased media thickness was observed in the BLM + 15d group, and a significant increase was observed in the BLM + 90d group (Figure 4C).

FIGURE 4.

Pulmonary artery sections in the vicinity of the airway were analyzed using either the Movat pentachrome staining method (A) or the hematoxylin and eosin method (B). (A1, 2) Pulmonary arteries close to airways in control PBS + 90d (A1), and BLM + 90d (A2) mice, flagged with asterisks (*) (bar = 10 μm). (A3) A higher magnification of BLM + 90d (A2), revealing no changes, that is no duplication or fragmentation of elastic fibers (black arrows), or collagen intimal deposition (the yellow square) (bar = 5 μm). (B1–3) The steps for assessing the media thickness. The selected artery is shown in B1 (bar = 30 μm), and a segment of the vascular wall with a clear transverse section is observed and photographed at a higher magnification (B2, bar = 15 μm). The media contained in this segment was manually outlined (B3). The length and segment were then calculated using ImageJ (https://imagej.net/ij/) (see B4), which provided the mean media thickness. The results are reported in the graph in C (the X‐axis shows the groups, and the Y‐axis shows the media thickness, expressed in pixels; 100 pixels = 2.25 μm). The asterisks indicates that the difference between the two sets of values is significant (p < 0.05).

4. DISCUSSION

The concept of PPF classification emerged from observations of the progressive nature of the lung fibrosis. As recently reported by others, ¹⁶ our BLM mouse model recapitulates some histological features of progressive lung fibrosis including an increase in collagen content without new ITI of BLM. However, as with some of the discrepancies reported in IPF between imaging and functional physiology tests, ¹⁷ we observed no significant added decrease in lung compliance thereafter at day 90, although a wide dispersion was observed. As 3 out of 8 mice exhibited a dramatically low compliance level, a larger sample size and consecutive evaluations would be required for the results to be considered statistically significant.

A notable finding for this model, which is closely aligned to the human disease, is the observation of an early onset of perivascular remodeling in areas distant from the fibrotic zone, as assessed by α‐SMA immunostaining of perivascular cells that are consistent with activated fibroblasts and myofibroblasts. This remodeling, which occurred distant from the fibrotic areas, has not yet been reported in any other bleomycin mouse model. Vascular changes play a key role in the natural history of IPF, from the early phases of the disease until the development of pulmonary hypertension (PH). ⁵ In PAH as in IPF, pulmonary arterial remodeling is characterized by various patterns of obliteration, as well as medial hypertrophy and intimal and adventitial thickening. ¹⁸ This results in a reduction in the luminal size of medium‐sized arteries due to the thickening of the three arterial tunicae, particularly the intimal layer. ¹⁵ Additionally, arterial elastin content increases and becomes disintegrated and scattered across and within the various layers. As reported in the Results section, we did not observe such lesions in mice that received ITI of BLM. Lesions observed in mice in various PAH models are always described as minimal, in contrast to those observed in other rodents, such as rats. ¹⁹ In keeping with this, even when an elevation in pulmonary pressure is observed in mice under hypoxic conditions, vascular remodeling takes the form of minimal media thickening. ²⁰ Our findings are consistent with these previous observations and indicate that we may detect an early phase of pulmonary hypertension. A limitation of this study was that heart changes associated with pulmonary hypertension could not be experimentally detected. However, as the results suggest vascular remodeling, further studies are necessary to confirm the development of pulmonary hypertension and signs of cardiac repercussions in our model. PH is a prominent complication resulting from multifactorial pathogenic mechanisms, and the critical role of the extracellular matrix in the perivascular space has been reported in experimental PH models. ²¹ , ²² It is generally accepted that PH becomes apparent after IPF onset and is secondary to fibrosis. Therefore, in line with experimental data, it can be hypothesized, that the two conditions may arise independently or together due to potential shared risk factors and mechanistic pathways. As lung transplantation is the only effective long‐term treatment for IPF, particularly when accompanied by PH, ²³ this model provides a valuable basis for further research. Indeed, it would enable us to improve targeted therapeutic options and survival predictions, as well as to offer a better understanding of the disease. ⁴ This is particularly significant given that PH has a prevalence ranging from 8% to 15% in IPF patients in the early stages, rising to over 60% in the later stages. ⁴

5. CONCLUSION

In conclusion, the results obtained support the hypothesis that progressive pulmonary fibrosis, associated with perivascular remodeling, can be induced in this mouse model when challenged with fractionated ITI of bleomycin. We propose that this model could be a valuable tool for investigating comorbidities associated with the development of progressive pulmonary fibrosis, demonstrating the potential initial step of PH. Indeed, it will enable us to track the evolution and spatial distribution of lesions over time, and improve our understanding of, and ability to treat lung remodeling and vascular lesions in their early stages. It may be worthwhile in the future to explore therapeutic approaches targeting vascular remodeling in IPF, for example by developing a reprogramming method in order to limit the remodeling. ²⁴

AUTHOR CONTRIBUTIONS

Céline‐Hivda Yegen: Formal analysis; investigation; methodology; writing – review and editing. Zakaria Maakoul: Formal analysis; investigation; methodology; writing – review and editing. Dominique Marchant: Formal analysis; investigation; methodology; writing – review and editing. Patrice Callard: Formal analysis; investigation. Julia Bourse: Formal analysis; investigation; methodology. Jean‐François Bernaudin: Formal analysis; investigation; methodology; writing – review and editing. Hilario Nunes: Writing – review and editing. Carole Planès: Writing – review and editing. Emilie Boncoeur: Conceptualization; data curation; formal analysis; funding acquisition; investigation; methodology; project administration; supervision; validation; writing – original draft; writing – review and editing. Nicolas Voituron: Conceptualization; data curation; formal analysis; funding acquisition; investigation; methodology; project administration; supervision; validation; writing – original draft; writing – review and editing.

FUNDING INFORMATION

This work was supported by the “Institut Fédératif de Recherche Biomédicale” programs of the Université Sorbonne Paris Nord (NV‐2020) and by a Chancellerie des Universités de Paris Legs Poix grant (Legs Poix 2018).

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

ETHICS STATEMENT

The experiments were approved by the Charles Darwin Ethics Committee (CEEA‐005) and authorized by the French Ministry of Higher Education, Research and Innovation based on the favorable opinion of the Ethics Committee under APAFIS number #18309 and performed in accordance with the European Community Coucil Directive 2010/63/EU on the animal's care.

Supporting information

Figure S1. (A) Experimental protocol. Eight‐week‐old C57BL6/J mice received three intra‐tracheal instillations (ITIs) of PBS or BLM at two‐weekly intervals. The mice were sacrificed 15 days (PBS+15d and BLM+15d groups) or 90 days (PBS+90d and BLM+90d groups) after the last ITI. The lungs were then removed. (B) Steps involved in assessing the content of periarterial collagen on Sirius red‐stained sections observed under polarized light microscopy.

Yegen C‐H, Maakoul Z, Marchant D, et al. Repeated instillations of bleomycin in male mice induce a perivascular remodeling at distance of fibrotic areas in the context of progressive pulmonary fibrosis. Anim Models Exp Med. 2026;00:1‐8. doi: 10.1002/ame2.70247

DATA AVAILABILITY STATEMENT

The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.