The poly (ADP-ribose) polymerase inhibitor olaparib attenuates established pulmonary fibrosis in a large animal model

Abstract

Background

Poly (ADP‒ribose) polymerase (PARP) is a constitutive enzyme involved in regulating various biological processes in health and disease. In the present study, we investigated the role of PARP1 in pulmonary fibrosis and assessed the efficacy of a clinically approved PARP inhibitor, olaparib, for the treatment of pulmonary fibrosis in a large animal bleomycin model.

Methods

Sheep (n = 12) received two fortnightly instillations of bleomycin (3 U) and saline into separate lung lobes of the same animal. Two weeks after the second bleomycin/saline exposure, sheep were randomly assigned to two groups and treated twice/week orally with either olaparib (10 mg/kg; treated group) or vehicle solution (control group) for four weeks. Olaparib concentrations in plasma were analysed using ultra-high-performance liquid chromatography. The degree of inflammation and fibrosis was assessed using a semiquantitative histopathology scoring method. Masson’s trichrome staining and hydroxyproline assays were used to evaluate collagen deposition in the lungs. We also determined whether olaparib treatment was targeting the migration of sheep lung fibroblasts in vitro using a 2D wound scratch assay.

Results

Olaparib was rapidly taken up into plasma following oral delivery, returning to baseline levels within 24 h. Therapeutically, olaparib treatment significantly reduced bleomycin-induced PARP1 overexpression and attenuated lung injury and fibrosis. Inflammation and fibrosis scores were significantly lower in bleomycin-injured lobes of sheep treated with olaparib compared to those treated with vehicle only. Additionally, a significant reduction in collagen deposition in the lungs of olaparib-treated sheep compared to vehicle-treated sheep was demonstrated by histopathology and hydroxyproline analyses. In vitro, olaparib significantly inhibited the migration of primary sheep lung fibroblasts.

Conclusions

Olaparib treatment reduced bleomycin-induced PARP1 overexpression and significantly attenuated established pulmonary fibrosis. Our data suggest that the activation of PARP1 plays a key role in the pathology of pulmonary fibrosis, and provides strong support for the potential repurposing of olaparib and similar PARP inhibitors for the treatment of pulmonary fibrosis.

Introduction

Idiopathic pulmonary fibrosis (IPF) is a chronic, progressive and fibrotic lung disease characterized by excessive extracellular matrix (ECM) deposition in the lung parenchyma that replaces healthy tissue. As a result, the alveolar architecture is destroyed, and lung compliance and gas exchange are compromised, ultimately leading to respiratory failure and death [1]. Currently, the two antifibrotic drugs, pirfenidone and nintedanib, are the only drugs available for pharmacological treatment of IPF [2–4]. Although these currently approved drugs can slow down the progression of the disease and the decline in lung function, they do not completely stop the fibrosis process. Moreover, these drugs are associated with unwanted adverse effects [3, 4]. Therefore, there is an unmet need to either develop new therapeutic options for IPF or explore the potential of repurposing existing clinically approved drugs for other indications.

Poly (ADP-ribose) polymerases (PARPs) are constitutive enzymes that play important roles in DNA repair and the maintenance of genome integrity, as well as in the regulation of various metabolic and signal transduction processes in health and disease [5–8], including modulating key pathways involved in the development of fibrosis [9]. The 17 enzymes that make up PARPs share a common domain that catalyses the transfer of ADP-ribose residues derived from NAD+, thus building a poly (ADP-ribose) chain and creating posttranslational changes in targeted proteins, a process known as PARylation [10]. PARP1 is the most abundant member and most widely studied enzyme in the PARP family [11].

Notably, several recent preclinical studies have revealed novel mechanisms that involve PARP1 in fibrogenesis of multiple organs [9], such as in LPS-induced myocardial fibrosis in rats [12], carbon tetrachloride-induced liver fibrosis in mice [13], angiotensin II-induced aortic fibrosis in rats [14] and bleomycin-induced pulmonary fibrosis in mice [15–18], highlighting the potential of PARP1 as a therapeutic target for these conditions.

There are several ways in which PARP1 may be involved in the pathogenesis of IPF. It has been demonstrated that the lungs of both IPF patients and rodent models of lung fibrosis exhibit excessive PARP1 activity. Compared with those isolated from control subjects, human fibroblasts isolated from the lungs of IPF patients showed overactivation of PARP1 [16]. In a similar study, PARP1 was shown to regulate myofibroblast differentiation by regulating the expression of the gene encoding α-SMA [16]. Furthermore, PARP1 inhibition reduces the expression of MMP-2 and MMP-9, both of which are involved in the pathogenesis of pulmonary fibrosis [19, 20].

There is growing evidence in the literature that inhibition of PARP1 is protective against lung inflammation and fibrosis in mouse models [15–18]. PARP1 knockout mice exhibited reduced pulmonary fibrosis in response to bleomycin-induced lung injury [16]. Similarly, pharmacological inhibition of PARP1 via the administration of 3-aminobenzamide (3-ABA) and 5-amino isoquinolinone (5-AIQ) suppressed the development of bleomycin-induced lung injury by blocking the attachment of nuclear factor kappa B (NF-kB) to DNA and reducing the levels of tumor necrosis factor-alpha (TNF-a) and interleukin-1b (IL-1b) [17]. Furthermore, another PARP1 inhibitor, HYDAMTIQ, was shown to ameliorate bleomycin-induced lung fibrosis by inhibiting the TGF-β/SMAD signalling pathway [15]. These studies in bleomycin mouse models suggest that PARP1 inhibition could be an attractive approach for the treatment of IPF.

Animal models are useful tool for understanding the pathobiology of complex diseases such as pulmonary fibrosis and for testing novel therapeutics. While most studies for pulmonary fibrosis are conducted in bleomycin-induced murine models, this model has several key limitations that impact its ability to translate to human disease. Indeed, many therapies that have shown benefit in this mouse model have failed to translate successfully into human clinical trials [21], and this is likely due to significant morphological differences between murine and human lungs. Large animals, such as sheep, possess anatomical and physiological features that closely resemble human lungs, making them suitable translational models for pulmonary fibrosis [22]. Importantly, the current study will utilise a sheep model of pulmonary fibrosis that has served previously for translational preclinical testing of new treatments [23–27].

Several PARP inhibitors are currently approved as clinical treatments for various cancer types due to their ability to impede DNA repair processes and the cell division of cancer cells. These include olaparib, talazoparib, rucaparib, and niraparib for the treatment of cancers, including ovarian, pancreatic, prostate and germline BRCA-mutant breast cancers [28, 29].

Olaparib (also known as AZD2281) is a small molecule, orally active PARP inhibitor that has been developed as a monotherapy as well as in combination with radiation or chemotherapy for a variety of solid tumours [28, 29]. Accumulating preclinical data highlighted above suggests the possibility of repurposing clinically approved PARP inhibitor drugs for nononcologic conditions, including idiopathic pulmonary fibrosis. Repurposing already approved drugs is an attractive approach in drug development as it offers the possibility of expedited and cost-effective development of new treatment options for combating disease. The present study aimed to investigate the role of PARP1 in pulmonary fibrosis and assess the therapeutic efficacy of the clinically approved PARP inhibitor olaparib in a large animal model of pulmonary fibrosis.

Materials and methods

Induction of lung fibrosis in sheep

Twelve (n = 12) female merino sheep (12–15 months of age, mean live weight 28.47 ± 0.78) were used in these studies. The sheep were sourced from the Federation University animal ethics committee-approved commercial supplier. The study was conducted according to the ARRIVE guidelines [30] and compiled with the Australian Code for the Care and Use of Animals for Scientific Purposes, with experimental procedures involving sheep approved by the Animal Ethics Committee (AEC) of Federation University Australia. All sheep were treated with anthelminthic to eliminate any possible parasite infections prior to experimentation and were judged free of significant pulmonary disease on the basis of clinical examination. Animals were housed in pens or yards as a group and provided with water and feed ad libitum.

Pulmonary fibrosis was induced using an established protocol [24, 27], where each sheep received two separate doses of bleomycin (BLM; 3 U in 5 ml of saline), with a two-week interval between administrations (Fig. 1). Briefly, the method of BLM administration involved direct bolus instillation into the left or right caudal lung lobe (random selection based on animal ID number) via the biopsy port of a bronchoscope. One lung lobe (right or left caudal lung lobe) received BLM, while the contralateral lobe in the same animal received sterile saline as an internal control (Fig. 1A). Administration of bleomycin/saline and BAL sampling was performed in fully awake sheep using a fibre-optic endoscope, with Lignocaine gel (2% w/v) applied as a local anesthetic to the nasal passage and outside of the endoscope [23–27].

Fig. 1.

Induction of pulmonary fibrosis and therapeutic intervention in sheep. (A) Schematic diagram showing localised delivery of bleomycin (BLM) and saline to individual caudal lung lobes in sheep. While one lung lobe received 3 U BLM, the contralateral lobe in the same animal received saline as an internal control. (B) Timeline (weeks) showing bleomycin injury (blue arrows), olaparib and vehicle treatments, bronchoalveolar lavage fluid (BAL) sampling (green arrows) and necropsy (red arrows)

Administration of olaparib

Two weeks after the final BLM or saline exposure, sheep were randomly assigned to two groups (n = 6/group): treated and untreated. Group size selection was based on our previous therapeutic studies and in consultation with a biostatistician for identifying statistically significant changes in disease-related histopathology [24, 27]. The treated group received an oral dose of 10 mg/kg (olaparib, MedKoo Biosciences; prepared on day of use in a vehicle solution of 10% DMSO/10% 2-hydroxy-propyl-β-cyclodextrin in PBS) olaparib twice/week for 4 weeks, whereas the control group received vehicle alone. Oral dosing of olaparib or vehicle solution involved inserting a feeding tube (7 mm internal diameter) via the nasal passage (guided by a fibre-optic endoscope) into the upper region of the oesophagus. Immediately prior to this, a solution of copper sulfate (10% CuSO₄ w/v, 20 mL) was administered into the esophagus to ensure effective closure of the reticular groove and enable oral drug delivery directly into the abomasum [31].

Sheep were monitored for clinical drug safety (in-life measures, including heart rate, body temperature, feed intake and changes in body weight) throughout the experimental period. All animals were housed under the same conditions in group pens or yards during experimentation. The experimental timeline, including BLM/saline exposures, sampling and treatment dosing, is shown in Fig. 1B. No subjects were excluded from providing data for analysis (exclusion criteria was based on intolerance to experimental procedures or ill-health during experimentation).

Quantification of olaparib in plasma

Pharmacokinetic (PK) evaluation was performed once during the first week of treatment in all sheep receiving olaparib (n = 6). Peripheral blood samples were collected predose and then at 0.25, 0.5, 0.75, 1, 3, 5 and 24 h following olaparib administration by drawing blood from the jugular vein into heparinized vacutainers (BD Biosciences). Plasma was prepared and stored frozen (−20 °C) prior to determination of olaparib concentrations using ultra-high-performance liquid chromatography (UHPLC) with tandem mass spectrometric detection, as previously described [32]. The PK parameters of maximum plasma concentration (C_max) and time to C_max (T_max) were read directly from the data.

Bronchoalveolar lavage sampling

Bronchoalveolar lavage (BAL) sampling was performed via bronchoscopy for all sheep at baseline and at two weeks and six weeks after the second BLM/saline exposure (Fig. 1B). For BAL collection, a bronchoscope was directed into the left and right caudal lung lobes, typically traversing 3–4 airway branches, followed by gentle infusion and withdrawal of 20 mL of sterile saline through the bronchoscope biopsy port [27]. BAL samples were centrifuged at 300 g to separate BAL cells and BAL fluid. BAL cell cytospots were then prepared on glass microscope slides and stained with Hem-Quik™ for differential cell leukocyte counts conducted in a blinded manner [27].

Necropsy and lung tissue sampling

Sheep were euthanized by intravenous barbiturate overdose at six weeks after the last BLM dose (Fig. 1B). Lung tissues were dissected, and the targeted lung lobes were identified and carefully dissected free from the surrounding tissue. Individual lobes were inflated with a 1:1 mixture of optimal cutting temperature compound (OCT; Tissue Tek, Miles Inc., PA, USA) and sterile PBS solution. Several transverse lung tissue blocks (0.5 cm thick) of the inflated lobe were cut, fixed in 10% neutral-buffered formalin and processed in paraffin for histopathological assessment as previously described [24, 27].

Immunohistochemistry and PARP1 expression

Immunohistochemistry for PARP1 expression was performed on 5 μm sections of formalin-fixed lung tissue. Briefly, tissue sections were microwave-treated for antigen retrieval prior to incubation with rabbit anti-PARP1 (EPR18461, 1:200; Abcam), followed by incubation with horseradish peroxidase (HRP)-conjugated goat anti-rabbit immunoglobulin (IgG) (P0448, 1:200; Dako). Positive staining was visualized following incubation with 3,3’-diaminobenzidine tetrahydrochloride (Pierce™ DAB Substrate Kit; Thermo Scientific), and the sections were counterstained with haematoxylin. PARP1 immunopositive (PARP1⁺) cells were counted in a blinded manner in 20 randomly selected digital images captured under light microscopy (captured at ×20 objective magnification).

Histopathology

Formalin-fixed lung tissue section (5 μm) were stained with haematoxylin and eosin Y (H&E), and histopathology of the lung parenchyma was evaluated using a semiquantitative scoring system as described previously [24, 27, 33]. Briefly, the slides were examined in a blinded manner, ensuring that the assessor remained unaware of the specific treatment groups. For each differentially treated lung lobes in each animal, 10 random, nonoverlapping fields (x20 objective magnification) lacking large bronchi and major pulmonary vessels were assessed from 1 H&E-stained section based on the scoring criteria for inflammation and fibrosis. The scores from all ten fields were averaged to obtain representative scores for the parameters assessed for each differentially treated lung lobe.

Quantitation of lung collagen levels

Masson’s trichrome staining and hydroxyproline assays were used to extrapolate the collagen content and concentration in the lung as an indication of connective tissue deposition and fibrosis. Briefly, for the histological assessment of collagen, 10 randomly selected fields (x20 objective magnification) excluding large blood vessels and bronchi from Masson’s trichrome-stained sections were captured using a camera attached to a microscope and computer. Image capture and analysis were performed in a blinded manner on coded slides. The photomicrographs were analysed using Image-Pro^® Plus software with the ‘colour selector’ tool to measure the percentage of area of collagen (blue stained tissue), and the results were averaged. Additionally, a biochemical hydroxyproline assay was used to extrapolate the collagen content (Hydroxyproline Assay kit, ab222941; Abcam), and the assay was performed according to the manufacturer’s instructions.

Primary lung fibroblast isolation and culture

Primary sheep lung fibroblast cultures were established from lung explants obtained from fibrotic lung lobes of sheep with bleomycin-induced pulmonary fibrosis (n = 3 sheep), as previously described [26]. In brief, tissue from the distal parenchyma was rinsed in 80% ethanol for sterilisation, then minced into 1–2 mm³ pieces. These pieces were placed into a medium containing 10% (vol/vol) fetal bovine serum (FBS) and 2% penicillin-streptomycin (Invitrogen) in Dulbecco’s Modified Eagle Medium (DMEM) (Invitrogen) and centrifuged for 5 min at 1000 rpm. The supernatant was removed, and the tissue pellet was resuspended in the same medium. The disrupted explants were then plated onto T75 tissue culture flasks (BD Biosciences, North Ryde, Australia) and cultured in a humidified atmosphere with 5% CO2 at 37 °C. After overnight incubation, non-adherent cells were removed, and the adherent cells continued to grow in DMEM under the same conditions. Cells between passages 3 and 5 were used for the in vitro wound healing assay.

Wound healing assay

Cell migration with respect to wound healing was assessed using a 2-dimentional (2D) scratch method on sheep lung fibroblasts (n = 3 sheep), as previously described [26]. Briefly, primary lung fibroblasts generated from explant cultures were seeded at 5 × 10³ cells per well in a 24 well plate and grown to confluence. After 24 h in serum-free medium, an artificial wound was scratched in each monolayer using a 100 µl pipette tip, following a grid line drawn on the bottom of the plate creating a linear cell-free area. Cells were washed, followed by the addition of complete media (DMEM with 10% FBS) containing olaparib at 50µM, 100µM and 200µM. Control wells contained DMEM with 10% FBS alone. Cells were imaged under an inverted microscope to assess growth at time 0 and 24 h following wound generation. Photomicrographs were uploaded to ImageJ and a line was drawn around the perimeter to provide a measure of the area of the wound, which was used to determine the wound closure percentage.

Statistical analysis

Statistical analysis was performed using GraphPad Prism for Windows (GraphPad Software Inc., La Jolla, CA, United States). Inflammation and fibrosis scores were analysed using the Mann‒Whitney test. Two-way ANOVA was used to compare the BAL leukocyte counts. Other parameters were assessed for a Gaussian distribution using the D’Agostino normality and Pearson omnibus test. For parametric data, paired two-tailed t tests were performed to compare lung segments that received bleomycin with saline-infused control segments, whereas unpaired t tests were used to compare data between olaparib-treated and vehicle-treated animals. For data that did not meet the assumptions of a Gaussian distribution, the Wilcoxon signed-rank test was used. A p value of < 0.05 was considered to indicate a significant difference. All values are reported as the means (± SEM).

Results

Olaparib treatment and pharmacokinetics

In n = 6 sheep, olaparib was administered orally, based on an established protocol allowing the drug to be delivered directly to the true stomach [31]. No adverse health events were recorded throughout the experiment or in response to olaparib treatment (no treatment-specific changes in heart rate, body temperature, feed intake, or body weight).

The PK profile for olaparib was assessed following oral olaparib treatment: olaparib was rapidly taken up into plasma following oral delivery, with a mean C_max of 2632 ± 575 ng/mL and a T_max of 1 h, returning to baseline levels within 24 h (Fig. 2).

Fig. 2.

Plasma PK profile (over 24 h) following oral olaparib administration in sheep. Mass spectrometric assessment of plasma olaparib levels; the data are expressed as the means ± SEM (n = 6 sheep)

Olaparib treatment reduces PARP1 overexpression in lung fibrosis

Immunohistochemistry revealed that an increase in PARP1 expression was associated with the development of lung fibrosis. Following the induction of lung fibrosis and in the absence of any treatment, PARP1 expression was elevated in fibrotic (BLM) lung lobes compared to saline lung lobes (Fig. 3). There was a significant reduction in PARP1 expression in BLM-exposed lungs from olaparib-treated animals as determined by PARP1⁺ cell counts (Figs. 3A and B).

Fig. 3.

PARP1 lung expression. (A) Representative micrographs showing PARP1⁺ immunostaining (brown stained cells) in saline- and BLM-exposed lung lobes of vehicle control and olaparib-treated sheep. (B) PARP1⁺ cell counts in saline- and BLM-exposed lungs; the data are expressed as the means ± SEM (n = 6/group); *p < 0.05, **p < 0.01. Scale bars = 50 μm

Olaparib treatment attenuates lung histopathology and collagen deposition

Olaparib treatment significantly reduced BLM-induced lung injury and fibrosis in sheep. In vehicle-control sheep, prominent fibrosis was observed in BLM-exposed lung lobes, while saline-exposed lung lobes in the same animal exhibited normal lung architecture with no signs of lung injury/fibrosis (Fig. 4A). In olaparib-treated sheep, lung injury and fibrosis were markedly attenuated in the BLM-exposed lung lobes, which had a similar appearance to that of the saline-exposed lobes (Fig. 4A). This finding was supported by semiquantitative analyses of inflammation and fibrosis scores, which were significantly lower in the BLM-exposed lobes of olaparib-treated sheep than in those of vehicle controls (Fig. 4B, C).

Fig. 4.

Lung histopathology. (A) Representative H&E staining of saline- and BLM-exposed lung lobes from vehicle only treated (control) and olaparib treated sheep. (B-C) Histopathology scoring data for (B) inflammation and (C) fibrosis as assessed on H&E-stained sections from the differentially treated lung lobes. The data are expressed as the means ± SEM (n = 6/group); *p < 0.05, **p < 0.01

Masson’s trichrome staining revealed a reduction in the degree of tissue fibrosis and connective tissue deposition in olaparib-treated compared to vehicle-only treated sheep (Fig. 5A). The area of collagen staining (blue staining in Masson’s trichrome stain) was significantly lower in the BLM-exposed lung lobes of olaparib-treated sheep than in those of vehicle treated controls (Fig. 5B). In support of the histopathology data on collagen deposition, biochemical hydroxyproline assay data revealed a significant reduction in the total hydroxyproline content of BLM-exposed lung lobes following olaparib treatment (Fig. 5B).

Fig. 5.

Collagen deposition in the lungs. (A) Masson’s trichrome staining of BLM-exposed lung lobes in vehicle-only treated (control) and olaparib-treated sheep. (B) Graphs showing the evaluation of collagen content; Masson’s trichrome blue-stained collagen area percentage (upper panel) and total hydroxyproline content (lower panel); data are expressed as the means ± SEM (n = 6/group), *p < 0.05, **p < 0.01

Olaparib treatment reduces BAL inflammatory cells

The BAL cellular profiles at baseline, 2 weeks and 6 weeks following the 2nd BLM dose are provided in Fig. 6. Total leukocyte, lymphocyte, and neutrophil numbers and percentages in BAL fluid were increased following BLM administration (Fig. 6A-F). Olaparib treatment significantly reduced the number of neutrophils as well as the number and percentage of lymphocytes in BAL fluid at 6 weeks after the 2nd BLM injury (Fig. 6A, B, F). BAL fluid macrophages were the predominant cell population (Fig. 6C, G).

Fig. 6.

BAL inflammatory cells in differentially treated lung lobes at baseline (0), week two (2) and week six (6) after 2nd BLM/saline exposure. (A) neutrophil numbers, (B) lymphocyte numbers, (C) macrophage numbers, (D), total leukocyte numbers, (E) neutrophil percentage, (F) lymphocyte percentage, and (G) macrophage percentage. data expressed as means ± SEM (n = 6/group), *p < 0.05, **p < 0.01, ***p < 0.001.The cell numbers are shown as ×10⁴ cells/mL BAL fluid. The data are expressed as the means ± SEM (n = 6/group), *p < 0.05, **p < 0.01, ***p < 0.001

Olaparib inhibits the migration of primary sheep lung fibroblasts

During a 24-hour incubation period following the creation of the wound, primary pulmonary fibroblast cells migrated to almost completely cover the wound area in the control (no olaparib) group. In contrast, migration appeared to be less pronounced where olaparib was present at concentrations of 50 µM, 100 µM and 200 µM (Fig. 7A). Analysis of the percentage of wound closure using Image J software showed an apparent dose-dependent effect of olaparib on fibroblast cell migration, with the 200 µM olaparib providing significant inhibition of fibroblast cell migration (Fig. 7B).

Fig. 7.

Effect of olaparib on the migration of primary sheep lung fibroblasts. (A) Photomicrographs of wound scratch taken at 0 and 24 h (4x magnification). (B) Graph showing wound closure percentage in cells at 24 h following incubation in the presence of increasing concentrations of olaparib (50-200mM). The data are expressed as means ± SEM (n = 3 sheep), ***p < 0.001

Discussion

In this study, we investigated the role of PARP1 in experimental pulmonary fibrosis by assessing the therapeutic effect of the PARP1 inhibitor olaparib. We found that bleomycin injury induces PARP1 overexpression, which is associated with focal regions of lung damage. Inhibition of PARP1 via oral olaparib treatment significantly reduces lung injury and fibrosis in the sheep model. In vitro, olaparib significantly inhibited the migration of primary lung fibroblasts taken from ‘fibrotic’ lungs. Although olaparib has previously been shown to reduce tissue fibrosis [34], to our knowledge, this is the first study to demonstrate the repurposing potential of olaparib for the treatment of pulmonary fibrosis.

Olaparib, a clinically approved anticancer medication for the treatment of several forms of human cancer, did not appear to have any adverse effects following oral administration at the dose used in the present study in sheep. Furthermore, olaparib was rapidly absorbed following oral dosing, achieving peak plasma concentrations at 1 h, consistent with previously reported maximal plasma concentrations after dosing (oral tablet formulation) in human patients [35, 36]. In cancer patients, olaparib has been shown to have a manageable toxicity profile at doses up to 400 mg b.i.d., the previously determined maximum tolerated dose (MTD) [35, 37]. The major adverse effects observed in human patients are nausea, fatigue, vomiting, taste alteration and anorexia [35, 38]. We used a dosage of olaparib (10 mg/kg or ~ 300 mg per animal, with dosing twice per week), which is much lower than the MTD recognized for cancer patients; on a drug delivery equivalence basis, the absence of any obvious adverse effect in our model is thus not unexpected. It is worth noting that in previous studies that used lower doses of olaparib, inhibition of PARP by more than 90% was observed in peripheral blood mononuclear cells from patients treated with 60 mg or more of olaparib twice daily [35], and the extent of PARP1 inhibitory activity did not increase proportionally with increasing doses of olaparib [35, 38].

Although we were able to demonstrate efficacy in this animal model of pulmonary fibrosis, olaparib was cleared from plasma within 24 h with dosing every 3–4 days, suggesting the need for further work to determine the optimal therapeutic dosing in this model of pulmonary fibrosis. Future studies will also need to explore whether continuous dosing is required and to evaluate potential long-term systemic effects, including impacts on systemic immune function.

There is considerable interest in exploring the potential of repurposing clinically approved PARP inhibitors for the treatment of various fibrotic diseases affecting the lungs, heart, liver, and kidneys [9]. Intriguingly, the in vivo doses that have shown efficacy in several preclinical models for other disease indications are typically an order of magnitude lower than those used in preclinical oncology models. This discrepancy may be explained by the necessity for comprehensive or nearly complete inhibition of all PARP1 activity to inhibit DNA repair (in cancer), while partial inhibition of PARP1 might suffice for cytoprotective effects such as restoration of cellular NAD + levels, reduction of free PAR formation or activation of cytoprotective AKT pathways. This suggests that the effective dose of PARP inhibitors for repurposing is likely to be lower than that used for the treatment of cancer [28, 39].

The present study showed an increase in PARP1 expression in bleomycin-exposed lung lobes of control animals, and oral treatment with olaparib (twice weekly over a period of four weeks) resulted in PARP1 levels in bleomycin-exposed lungs returning to near baseline levels. Bleomycin causes DNA damage via direct DNA strand scission and the generation of free radicals [21], events that lead to increased expression and activation of PARP1 [9]. Indeed, the main function of PARP enzymes is to sense DNA damage and activate subsequent repair mechanisms [5–7]. Previous studies measuring PARylated proteins using western blotting have shown that bleomycin damage activates PARP in mouse bleomycin models [15, 18]. Moreover, in vitro exposure of mouse lung slices to bleomycin [40] as well as fibroblasts isolated from IPF patients exhibited significantly higher levels of PARP expression [16]. Our study provides further confirmation of elevated PARP expression/activation following bleomycin damage.

As previously demonstrated in our model, the damage induced by bleomycin exposure significantly increases fibrosis pathology in lung tissue [24, 25, 27, 41]. In the present study, olaparib treatment commencing two weeks post-bleomycin lung injury led to improvements in pathological changes: inflammation and fibrosis scores were significantly reduced and approached the levels observed in the lobes not exposed to bleomycin. Moreover, Masson’s trichrome staining of lung tissue and collagen assessment using the hydroxyproline assay provided further evidence that sheep treated with olaparib had reduced levels of fibrosis in the lobes exposed to bleomycin.

PARP1 has previously been implicated in the pathological mechanisms of pulmonary fibrosis, and the inhibition of PARP1 could prevent fibrosis via multiple pathways [9, 15, 16, 42]. It has been demonstrated that the overexpression of PARP1 can activate the expression of the α-SMA gene in human lung fibroblasts, leading to the differentiation of myofibroblasts, which are key cells that produce ECM [16]. In addition, modulation of the TGFβ-SMAD pathway has been shown to be one of the molecular pathways by which PARP1 contributes to fibrosis. PARP1 is required for TGF-β1-induced Smad3 phosphorylation and nuclear accumulation, enhancing DNA binding to transcription factors and transcription of the CoIα1 and CoIIIα1 genes [42]. In a mouse model of bleomycin-induced pulmonary fibrosis, it has been suggested that the antifibrotic effects of inhibiting PARP1 may be attributed to suppression of the TGFβ-SMAD signalling pathway [15]. Although we have not performed mechanistic analyses in the current proof of concept study, the reduced fibrosis pathology and collagen deposition through PARP1 inhibition shown in our study may involve these mechanisms.

In addition to the reduction in fibrosis observed in the current study, olaparib also had a marked effect on inflammation in the lungs. This effect included reduced lung tissue inflammation in postmortem tissues and a significant reduction in neutrophil and lymphocyte levels in BAL fluid following olaparib treatment in bleomycin-exposed lungs. It has previously been reported that PARP1 is involved in the activation of innate (neutrophils, macrophages, and dendritic cells) and adaptive (T and B lymphocytes) immune cells [43–45]. Furthermore, PARP1 can drive inflammation via the activation of transcription factors such as NF-kappaB and AP-1 to promote the release of proinflammatory mediators [46]. The effects of PARP1 inhibition on inflammation have been demonstrated in vivo in several preclinical models of lung inflammation [47–49].

Conclusions

The present study demonstrated that the overexpression and likely activation of PARP1 play key roles in the pathology of bleomycin-induced pulmonary fibrosis in sheep. The inhibition of PARP1 via oral olaparib treatment provided significant resolution of established pulmonary fibrosis, which provides strong support for the potential repurposing of olaparib and similar PARP inhibitors for the treatment of pulmonary fibrosis.

Authors’ contributions

H.B.D: Study conception, model development, study design and conduct, data collection and analyses, and manuscript preparation and review; A.N.D, and S.B: Data collection and manuscript review; A.Y: study conduct and manuscript review; J.P, D.R and M.M: Data collection and analyses and manuscript review; D.P: Study design, data collection and manuscript review; R.J.B: Study conception, model development, study design and conduct, data collection and analyses, manuscript preparation and review.

Funding

This work was supported by seed funding from Federation University and Allergenix Pty Ltd.

Data availability

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

Declarations

Ethics approval and consent to participate

The study was conducted in accordance with the Australian Code for the Care and Use of Animals for Scientific Purposes, with experimental procedures involving sheep approved by the Animal Ethics Committee (AEC) of Federation University, Australia.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.