Male Sprague Dawley rats (Charles River Italia, Lecco, Italy), weighing 200–250 g, were housed 3 or 2 per cage in standard conditions (light 7 a.m.–7 p.m., dark 7 p.m.–7 a.m., temperature 10 °C–24 °C, relative humidity 40%–70%) at arrival. Before any use, animals followed an acclimatization period of 5 days and no prophylactic or therapeutic treatment were administered to any animals during this time. Animals had free access to tap water and were fed with a standard certified laboratory mouse diet ad libitum (4RF21, Mucedola srl, Milan, Italy for Charles River Italia). They were checked for abnormalities or sign of health problem after arrival in the animal facility. All the procedures involving animals were reviewed, approved and authorized by the Italian Public Health System for Animal Health and Food Safety within the Italian Ministry of Health (MoH) (authorization number 246/2021-PR). All the procedures were performed within a certified animal facility: AAALAC (American Association for Accreditation of Laboratory Animal Care, https://www.aaalac.org/). Experiments were performed in full compliance with the European ethics standards in conformity with directive 2010/63/EU, Italian D. Lgs 26/2014, the revised ‘Guide for the Care and Use of Laboratory Animals’ (Guide for the Care and Use of Laboratory Animals, 1996), and the ARRIVE guidelines (Animal Research: Reporting of In Vivo Experiments).

Rats were randomized after weighing and body weight were checked every day before treatment to obtain an indication about health conditions. According to published protocol for this model [15], a total of 20 animals received intratracheal BLM (batch 100012570, Baxter Oncology GmbH, Halle, Germany) administrations at the starting point of the experiment and in the fourth day, using a dose of 1.5 U/kg/0.6 mL, as shown in Fig. 1. Especially, rats were anesthetized in cages through a sevoflurane gas (4% in oxygen) and then attached by incisor teeth to a support inclined about 45° respect to the work surface. Holding the mouth open, tongue was moved with tweezers and the trachea was observed using a small otoscope (000A3754, Welch-Allyn Li-Ion, Hallowell EMC, Pittsfield, MA, USA) accessorized of a rat intubation speculum (200A3588, Hallowell EMC, Pittsfield, MA, USA). While the vocal cords were opened, bleomycin was injected through a metal canula. The same procedure was carried out for the control group composed of 14 rats which received a 0.9% saline solution (batch 19TAAHC0, Sodium Chloride 0.9%, Eurospital S.p.A, Trieste, Italy), the same used for the BLM dilution. Animals were sacrificed at two different time points to investigate alveolar epithelial dysfunction at different stages of the disease: 6–8 rats for the control group at day 21 and day 28, respectively, and 9–11 BLM-treated rats at day 21 and 28, respectively. They were anesthetized with pentothal sodium solution (batch 23PT001A/1, MSD Animal Health, Rahway, NJ, USA) injected intraperitoneally and sacrificed by bleeding from the abdominal aorta. Lungs were washed through transcardiac perfusion with a saline solution and explanted. Left lobes were separated and used for histological analysis, while right lobes were allocated for markers quantification.

Fig. 1.

Experimental protocol. BLM, bleomycin; SAL, saline.

Left lobes were insufflated through a tracheal cannula with a 10% formalin buffer, necessary to expand and fix the lungs and stored in test tubes filled with 10% formalin buffer (Sigma-Aldrich, Saint Louis, MO, USA). Then, the formalin fixed lungs, were paraffin embedded (FFPE). Each longitudinally oriented lung was cut to obtain 5 µm thick sections representing the long axis of the entire parenchyma on slides. After that, they were stained with the Masson Trichrome staining (activities performed by Histolab Verona, Verona, Italy) that allowed to highlight collagen deposition and fibrotic areas into the lung parenchyma. The slide images are acquired using the NanoZoomer S-60 Digital scanner (Hamamatsu Photonics K.K., Shizuoka, Japan) at 20$\times$ magnification. In addition, fibrotic lesions were assessed by Ashcroft score evaluation, using a method based on the Ashcroft scale, as described previously by Ashcroft et al. [16] and modified by Hübner et al. [17]. According to this approach, lung sections approximately 2 $\times$ 1 cm were examined by the observer under 10$\times$ magnification. Each field (40 to 50 per section) was individually evaluated for the fibrosis severity and assigned a score from 0 (normal lung) to 8 (total fibrotic oblation of the field) according to the Ashcroft scale. The fibrosis score for the entire lung section was calculated by summing all grades and dividing by the number of fields analyzed. Fibrosis quantification was furthermore confirmed by the automated software VIS analysis Protocol Package (VIS; V.2017.2.4.3387) which used the deep learning intelligence. The automated analysis offers an observer-independent measurement and was conducted using an optimized and internally validated Image Analysis Protocol Package (APP). The APP was designed to identify and quantify fibrotic lung lesions by detecting connective tissue accumulation, excessive collagen deposition, and cellular hyperproliferation, showing a significant correlation with the Ashcroft score. Before any quantification, the VIS APP was calibrated on control animals from each experiment to exclude normal collagen tissue and large bronchi, blood vessels, and emphysema were excluded from the lung area. This software returns a measure of the percentage of fibrotic area to the total lung tissue.

Immunofluorescence was carried out on lung sections of BLM and vehicle treated animals using the Leica BOND RX automated stainer (BOND Research Detection System DS9455, Leica Biosystems, Nussloch, Germany). Slides underwent an unmasking step using EDTA buffer pH 8.5 and consequently they were incubated for 15 minutes with the blocking solution composed of 10% of the donkey serum (D9663, Sigma-Aldrich, Saint Louis, MO, USA) in PBS 1$\times$ (10010023, Thermo Fisher Scientific, Waltham, MA, USA) and Triton 10% (Triton X-100, 9036-19-5, Sigma-Aldrich, Saint Louis, MO, USA). Thereafter they were incubated for 45 minutes with primary antibodies, such as mouse anti-alpha smooth muscle actin (1:200, ab7817, Abcam, Cambridge, UK), rabbit anti-KRT8 (1:1000, ab217173, Abcam, Cambridge, UK) or mouse anti-KRT8 (1:200, M1603-2, Thermo Fisher Scientific, Waltham, MA, USA), rabbit anti-Vimentin (1:200, ab92547, Abcam, Cambridge, UK), rabbit anti-Prosurfactant Protein C (1:200, ab312850, Abcam, Cambridge, UK), followed by the secondary antibodies incubation of 60 minutes in goat anti-mouse AlexaFluor555 (A-32727, Thermo Fisher Scientific, Waltham, MA, USA), donkey anti-rabbit AlexaFluor488 (A-21206, Thermo Fisher Scientific, Waltham, MA, USA), donkey anti-mouse AlexaFluor488 (A-21202, Thermo Fisher Scientific, Waltham, MA, USA), donkey anti-rabbit AlexaFluor647 (A-31573, Thermo Fisher Scientific, Waltham, MA, USA), donkey anti-rabbit AlexaFluor555 (A31572, Thermo Fisher Scientific, Waltham, MA, USA). Cell nuclei were stained with 4${}^{\prime }$,6-diamidino-2-phenylindole (DAPI) (P36931, Life Technologies, Thermo Fisher Scientific, Waltham, MA, USA). Immunofluorescence images were acquired with the Zeiss LSM 710 confocal microscope (Carl Zeiss, Oberkochen, Germany).

The mean fluorescence intensity of KRT8, alpha-smooth muscle actin ($\alpha$-SMA) and Pro-SPC was measured on the 20$\times$ immunofluorescence microscopy images of five animals per group and selecting four regions each slide. Fibrotic regions of BLM rats at both 21 days and 28 days were analyzed excluding bronchi, which were not even considered in the controls. FIJI (ImageJ v.1.8.0_345, can be downloaded at fiji.sc.) were used for the quantification analysis following the steps described by Shihan et al. [18]. KRT8+ Pro-SPC+ cells were counted through the QuPath software (v.0.5.1, https://qupath.github.io/), as used by Wang et al. [19], on the 60$\times$ immunofluorescence microscopy images and their abundance were reported as percentage of the total number of cells.

Frozen right lobes were weighted and homogenized in 10 mL of ice-cold 1$\times$ PBS (10010023, Thermo Fisher Scientific, Waltham, MA, USA) per g of tissue using the gentleMACS Dissociator (Miltenyi Biotec, Bergisch Gladbach, Germany) with the protease and phosphatase inhibitor cocktail (PPC2020, Sigma-Aldrich, Saint Louis, MO, USA). Cell Lysis Buffer (895347, R&D, Minneapolis, MN, USA) was added to a rate of each sample and then centrifuged (5000 g for 10 minutes) to obtain a supernatant cell lysate. The total amount of proteins was quantified using the DC protein assay (500-0116, Bio-Rad, Hercules, CA, USA) after diluting samples 1:10. Pro-collagen 1 protein was quantified in the supernatant through the Mouse Pro-Collagen I alpha 1 Enzyme-linked Immunosorbent Assay (ELISA) Kit (ab210579, Abcam, Cambridge, UK), already used in rats by Huang et al. [20]. Results were expressed as pg/mg of protein.

The statistical analysis was performed using GraphPad Prism 10 software version 10.0.2 for Windows (GraphPad Software, San Diego, CA, USA). All data were reported as mean $\pm$ SEM. The comparisons between BLM-treated animals and corresponding vehicles were carried out using the one-way ANOVA Dunnett’s test and the Student’s T test for parametric analysis. Especially the first test was used to compare more than two group and the second one to compare just two of them. The p-value $\le$ 0.05 was considered statistically significant. *p 0.05, **p 0.01, ***p 0.001, ****p 0.0001.

To assess the lung fibrosis induced by BLM, we analyzed representative Masson’s Trichrome-stained lung tissue sections comparing BLM-treated animal, at two different time points, and the relative controls (Fig. 2A). The staining highlighted severe fibrotic areas starting branching in rats receiving BLM at both day 21 and day 28. Moreover, large areas of normal lung parenchyma were replaced with dense collagen deposition as fibrotic single masses or confluent conglomerates corresponding to an Ashcroft score around 4 (colored in blue by Masson’s staining), while these features were absent in SAL rats. It is also possible to appreciate that BLM injury resulted in a patchy distribution of fibrotic lesions consisting of normal lung architecture in alternance to affected areas.

Fig. 2.

Quantification of fibrosis in rats treated with bleomycin. (A) Masson’s Trichrome-stained lung tissue of the control group and the BLM-treated animals at 21 and 28 days. Scale bar 5 mm or 1 mm. (B) The bar graph shows the Ashcroft Score quantification of lung damage in different experimental groups treated with BLM. ****p 0.0001 vs. Saline of corresponding time point (Student’s T test). (C) Graphic representation of the four classes of severity in different experimental groups. Physiological (from Ashcroft score 0 to 1) is represented in white, mild (2–3) in grey, moderate (from 4 to 5) in dark grey and finally black corresponds to the most severe score (from 6 to 8). ****p 0.0001, **p 0.01 vs. Saline, (One way ANOVA followed by Dunnett’s test). (D) Percentage of fibrotic area calculated in different experimental groups using the Visiopharm platform. *p 0.05, **p 0.01 vs. Saline of corresponding time point (Student’s T test). (E) Pro-collagen I protein quantification carried out through an ELISA assay. ****p 0.0001 vs. Saline of corresponding time point (Student’s T test). SAL, Saline; BLM, Bleomycin; ELISA, Enzyme-linked Immunosorbent Assay.

The fibrotic condition was confirmed by Ashcroft score evaluation, using a method based on the Ashcroft scale [16, 17]. BLM rats at 21 days reached a score of 3.75 (p 0.0001) and at 28 days a score of 4.04 (p 0.0001) compared to controls (Fig. 2B). As shown in Fig. 2C, BLM animals sacrificed on day 21 and 28 had a high frequency of the highest scores compared to controls (consider physiological from 0 to 1 of Ashcroft score, mild from 2 to 3, moderate from 4 to 5, severe score from 6 to 8). Especially, rats at 28 days showed more severe score than those on the 21st day. With an automated software of fibrosis analysis (VIS analysis Protocol Package), we confirmed a significant increase of fibrosis in BLM treated animals compared to SAL (Fig. 2D), with a rate of increase of 75% (p 0.01) and 65% (p 0.05) at day 21 and 28, respectively. In agreement with data obtained by the semiquantitative measurements of fibrosis, we found significant increase levels of Pro-Collagen I in right lung homogenates of rats receiving BLM and sacrificed on day 21 (p 0.01) and 28 (p 0.05), as demonstrated in Fig. 2E.

Further confirmation of the fibrosis development in the rat BLM model was the activation of fibroblasts into myofibroblasts. Notably, the immunofluorescence staining of alpha-smooth muscle actin ($\alpha$-SMA) was flanked to the corresponding area of the Masson Trichrome stained slide, which emphasized collagen deposition and fibrotic regions. The $\alpha$-SMA fluorescence revealed the presence of myofibroblasts, especially SAL and BLM-animals images showed physiological $\alpha$-SMA positive cells in vessel and bronchial walls, but they were found in the parenchyma exclusively of BLM treated rats at both 21 and 28 days (Fig. 3A). This evidence was an additional proof of the fibrotic lesions in BLM treated animals.

Fig. 3.

Histological and immunofluorescence evaluation of rat lungs at 21 and 28 days after SAL or BLM treatment. (A) Representative images of the parenchyma lung sections of rats receiving BLM or saline at 21 and 28 days. On the left tissue lungs stained with Masson’s Trichrome, scale bar 500 µm. On the right, correspondent immunofluorescence staining of myofibroblasts colored in red using the $\alpha$-SMA antibody and nuclei (DAPI) in blue. Magnification 10$\times$, scale bar 100 µm, and zoom at magnification 20$\times$, scale bar 50 µm. (B) Quantification of $\alpha$-SMA mean fluorescence intensity in the parenchyma (n = 5 per group). **p 0.01, ****p 0.0001 vs. Saline, ^#⁢#⁢#p 0.001 vs. BLM 28 days (Student’s T test). $\alpha$-SMA, alpha-smooth muscle actin; SAL, Saline; BLM, Bleomycin; DAPI, 4${}^{\prime }$,6-diamidino-2-phenylindole.

The quantification of the $\alpha$-SMA staining was reported as mean fluorescence intensity (Fig. 3B) and highlighted the highest myofibroblasts activation in the 21 days BLM rats (p 0.0001) compared to the control group. In addition, BLM animals at 21 days had a major $\alpha$-SMA fluorescence intensity compared to the BLM animals at 28 days (p 0.001). Finally, these latter showed a higher level than the SAL (p 0.01). This evidence was a clue to the ability of rodents in resolving fibrosis and fibroblasts activation [19] that in the rat model occurs from day 21 onward.

To further investigate the presence of aberrant basaloid cells in BLM-induced fibrotic lesions, lung sections of BLM and SAL treated animals were stained with the KRT8 and $\alpha$-SMA marker. Immunofluorescence images acquired with the confocal microscope, clearly showed the presence of KRT8+ cells both in rats receiving saline or bleomycin. In control rats KRT8+ cells were located physiologically in the bronchus as expected and previously demonstrated [12] while in BLM rats they were, further into the bronchi, moreover in the parenchyma [13]. They were found especially nearby myofibroblasts, whose presence is highlighted by $\alpha$-SMA, in BLM rats sacrificed at day 21. BLM rats sacrificed at day 28, instead, had less aberrant basaloid cells of those sacrificed on day 21, suggesting that the disappearance of KRT8+ cells agreed to the failure of the animal model in the disease progression [19] (Fig. 4).

Fig. 4.

Representative immunofluorescence images of aberrant basaloid cells in rat lungs. On the left tissue lungs stained with Masson’s Trichrome, scale bar 250 µm. On the right, corresponding immunofluorescence staining of aberrant basaloid cells colored in green using KRT8 antibody, $\alpha$-SMA in red and nuclei (DAPI) in blue. Magnification 20$\times$, scale bar 50 µm, and zoom at magnification 60$\times$, scale bar 20 µm. White arrows indicate KRT8+ cells into the parenchyma.

It was hypothesized that aberrant basaloid cells originate from AT2 cells, acting as an intermediate state [12, 13, 18, 21, 22, 23], thereby preventing differentiation into AT1 cells. Immunofluorescence images obtained on our rat BLM model strengthened this assumption. Indeed, KRT8+ of rats on the twenty-first and twenty-eighth day showed a merged staining with the Pro-SPC, in contrast to controls, as shown in Fig. 5A (indicated with pink arrows). Furthermore, it was interesting to note how some aberrant cells expressed exclusively KRT8 (indicated with white arrows), definitely losing the marking of alveolar cells.

Fig. 5.

Immunofluorescence staining and quantification of alveolar type II (AT2) and aberrant basaloid cells in rat lungs. (A) Immunofluorescence staining of aberrant basaloid cells. KRT8 colored in green, pro-SPC in red and nuclei (DAPI) in blue. Magnification 20$\times$, scale bar 50 µm, and zoom at magnification 60$\times$, scale bar 20 µm. The rightmost images represent the split channels of the corresponding adjacent image at 60$\times$, they have the same scale bar, 20 µm. White arrows indicate KRT8+ cells into the parenchyma while pink arrows indicate AT2 cells. (B) Quantification of KRT8 mean fluorescence intensity in the parenchyma (n = 5 per group). (C) Quantification of Pro-SPC mean fluorescence intensity in the parenchyma (n = 5 per group). (D) Percentage of KRT8+ Pro-SPC+ cells on the total number of nuclei labeled with DAPI. **p 0.01, ***p 0.001 vs. Saline (Student’s T test). SAL, Saline; BLM, Bleomycin; Pro-SPC, Prosurfactant protein C.

The highest presence of KRT8+ cells in the BLM-treated rats at 21 (p 0.01) and 28 days (p 0.001), compared to SAL, were demonstrated by the mean fluorescence intensity measure (Fig. 5B), and confirmed the qualitative considerations done in the previous paragraph. A quantitative measure of Pro-SPC (Fig. 5C), instead, reported a similar level in all groups. The apparent higher intensity in the BLM-treated animals is probably due to the higher cell density in the fibrotic regions compared to the physiological parenchyma of control animals [24].

The total number of transient AT2 cells were counted considering their co-expression of the KRT8 and Pro-SPC markers: BLM rats at day 21 and day 28 exhibited a significantly higher proportion of KRT+ Pro-SPC+ cells compared to controls (p 0.01). Specifically, these cells constituted 10.10% and 8.67% of the total cell population in the examined area, respectively (Fig. 5D).

In addition, we evaluated the expression of Vimentin, a mesenchymal marker upregulated in EMT. As showed in Fig. 6 Vimentin was upregulated and localized in the lung parenchyma of BLM rats both at 21 and 28 days. According to other studies [4, 8, 9, 12, 21, 23, 25], where mesenchymal features of aberrant basaloid cells were described, we found colocalization of the immunofluorescence signal of KRT8 and Vimentin in both BLM rats at 21 and 28 days, depictured in Fig. 6 (indicated with white arrows). These data suggested the involvement of KRT8+ in the epithelial-mesenchymal transition process.

Fig. 6.

Immunofluorescence co-staining of KRT8+ cells and Vimentin in rat lungs. KRT8 colored in green, Vimentin in purple and nuclei (DAPI) in blue. Magnification 20$\times$, scale bar 50 µm, and zoom at magnification 60$\times$, scale bar 20 µm. The rightmost images represent the split channels of the corresponding adjacent image at 60$\times$, they have the same scale bar, 20 µm. White arrows indicate cells expressing Vimentin. SAL, Saline; BLM, Bleomycin.