Background
Idiopathic pulmonary fibrosis (IPF) is a severe type of lung fibrosis with a median survival of 2–3 years [
1]. The pathogenesis of IPF is still unclear, although marked progress has been made recently both in clarifying disease mechanisms and in developing new therapeutic agents. At present, no pharmacological therapy is able to cure the disease but two drugs, pirfenidone and nintedanib i.e. BIBF1120, have been shown to slow the progression of the disease [
2‐
4] whereas the previously used N-acetylcysteine had no effect on the outcome [
5,
6].
Changes in epithelial and mesenchymal cells as well as the interaction between these cells are the main characteristics of IPF whereas it is currently believed that inflammatory processes play only a minor role. One widely accepted hypothesis to explain the mechanisms in IPF pathogenesis postulates that an injury of the alveolar epithelium results in excessive production of extracellular matrix (ECM) proteins, growth and transcription factors and cytokines by fibroblasts [
7]. The fibroblast focus, a typical histological feature of IPF, is a specific aggregate of cells, especially fibroblasts and myofibroblasts covered by injured and hyperplastic epithelium, and ECM produced by myofibroblasts [
8]. Studies have revealed that IPF patients with a high number of fibroblast foci have a shortened survival [
9]. In addition, the extent of expression of alpha smooth muscle actin (α-SMA), as a marker of myofibroblasts, in the lungs of IPF-patients, has been shown to be negatively associated with patient survival [
10].
In our previous studies, we have observed that it is possible to isolate and culture fibroblast and myofibroblast containing cell lines both from the bronchoalveolar lavage (BAL) fluid and lung tissue samples of patients with different types of lung diseases including IPF. Furthermore, we characterized these cells by a variety of methods including electron and immunoelectron microscopy [
11,
12]. We have noted that myofibroblasts from different lung diseases display different ultrastructural and functional properties [
11,
12]. In particular, we and other investigators have observed that fibroblasts and myofibroblasts containing cells lines cultured from IPF patients are more invasive than the cells obtained from other lung diseases [
11,
13]. It has also been reported that fibroblastic cells from IPF patients have a higher amount of α-SMA, a lower growth rate and a higher number of apoptotic cells than found in controls [
14].
Most of the previous preclinical studies investigating the effect of potential anti-fibrosis drugs have been conducted by using animal models [
15]. For example, bleomycin-induced fibrosis in mice, rats or hamsters has been the most commonly used study protocol. Although testing in animal models is rational, it is often difficult to extrapolate the results to the human diseases. For instance, bleomycin-induced fibrosis in rodents resembles rather poorly the IPF in humans [
16,
17], and further, the pulmonary anatomy and cellular components of rodents and other experimental animals are very different from their human counterparts. There are very few studies which have utilized human lung cells to investigate novel therapeutic agents for pulmonary fibrosis, and even fewer that have used cells originating from IPF patients. Moreover, most of the previous studies have focused on only one pharmacological agent and not compared two or more drugs using the same study protocol. Surprisingly, the mechanisms of action of many promising anti-fibrosis drugs are still not fully understood. A better understanding of mechanism could help us selecting the most suitable therapy for each individual patient as well as in developing improved combination treatment modalities in the future [
18].
The aim of the present study was to evaluate the effects of pirfenidone and nintedanib on the ultrastructural and functional properties of stromal cells such as fibroblasts and myofibroblasts collected from BAL and lung tissue samples both from patients with IPF and from control lung.
Methods
Study subjects
The study material comprised lung tissue from 7 patients with IPF and from 4 control patients having normal peripheral lung. The patients underwent diagnostic BAL, diagnostic surgical lung biopsy or surgery for lung cancer during 2008–2012 in Oulu University Hospital (Table
1). All control patients were nonsmokers with normal lung function and normal lung histology outside the lung tumor. Pieces of lung tissues were collected from non-involved areas outside the tumor as previously described [
19]. As the cells of IPF patients were derived from diagnostic samples before the year 2012, none of the study subjects was treated with pirfenidone or nintedanib before the cells were derived.
Table 1
Sample information
Total number | 4 | 7 |
Smoking status | Non-smoker | 4 | 3 |
Ex-smoker | | 3 |
Smoker | | 1 |
Number of samples derived from | BAL | 0 | 4 |
Biopsy | 0 | 2 |
Lobectomy for lung cancer | 4 | 1 |
Ethics, consent and permissions
The donors were informed and interviewed before the operation. Each patient provided written informed consent. The study protocol was approved by the Ethical Committee of Northern Ostrobothnia Hospital District in Oulu (64/2001, amendment 2005, 2/2008).
Cell culture
Cell samples were collected and stromal cells were cultured as described previously [
11,
12]. Briefly, an aliquot of BAL-sample or collagenase-digested lung biopsy specimen was centrifuged (300 g, 10 min) and plated at a density of approximately 40,000 cells/cm
2 in a medium consisting of Minimun essential medium Eagle α modification (Sigma-Aldrich, Inc, St Louis, MO, USA) supplemented with 13 % heat-inactivated fetal bovine serum (PromoCell, Heidelberg, Germany), 2 mM L-glutamine, 100 U/ml penicillin, 0.1 g/l streptomycin, 2.5 mg/l amphotericin B and 10 mM HEPES (all from Sigma-Aldrich). The cells were passaged at near-confluence and used for experiments in passages 2–5. The cells were exposed to 0.1–0.5 mM pirfenidone (Santa Cruz Biotechnology) or 0.1–0.5 μM nintedanib by adding the drug into the cell culture medium with or without serum. Pilot studies were used to select drug concentrations that were low enough not to harm the cells but high enough to cause responses. The effects of the drugs were tested also in the presence of 2–5 ng/ml transforming growth factor β1 (TGFβ1) (Sigma-Alrich) in serum-free conditions.
Proliferation
In the proliferation assay, the cells were plated on 96-well plates with 500 cells per well, 6 parallel wells for each condition. On the next day, the medium was replaced with new medium (control medium with serum, medium with 0.1–0.5 mM pirfenidone and/or 0.1–0.5 mM nintedanib with serum, medium without serum but with 5 ng/ml TGFβ1 or serum-free medium with TGFβ1 and nintedanib or pirfenidone). The number of cells was measured after 1, 3 and 7 days of drug exposure with the MTT-assay (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, Sigma-Adrich). The MTT reagent was added to the wells at a final concentration 0.5 g/l. The cells were allowed to reduce MTT into formazan (2 h at 37 °C) the amount of which was measured spectrophotometrically at a wavelength of 550 nm against background (650 nm) after lysing the cells in DMSO.
Western analysis
Western analysis of α-SMA was performed as described earlier [
11]. Briefly, the cells were lysed in 50 mM Tris, 0.1 % Triton X-100, 0.9 % NaCl supplemented with a protease inhibitor cocktail tablet (Roche, Mannhaim, Germany) and 20 μg aliquots of samples were loaded and run on 12 % SDS–PAGE. The proteins were transferred onto nitrocellulose membrane (Protran, Schleicer and Schuell, Bioscience, Dassel, Germany). After blocking with milk, the membranes were incubated with a 1:1000 dilution of α-SMA antibody followed by 1:1000 diluted secondary antibody (IRDye 800 conjugated anti-mouse IgG, Rockland Immunochemicals, Gilbertsville, PA, USA). Protein intensities were detected and analyzed with an Odyssey infrared imager (Li-Cor Biosciences).
Transmission electron microscopy
All the cultured cells have been previously characterized by transmission electron microscopy (TEM) which showed that morphologically the cell populations consisted of fibroblasts and myofibroblasts [
11,
12]. In order to evaluate the effect of 0.5 mM pirfenidone and 0.5 μM nintedanib on cellular ultrastructure, the cells from control and IPF patients were exposed to the drugs for 4 days in the presence of 5 ng/ml TGFβ1. Samples treated with TGFβ1 alone were used as controls. The samples were fixed and prepared for TEM as described earlier [
11]. Briefly, the cells were fixed in 1 % glutaraldehyde-4 % paraformaldehyde for 10 min. The cells were detached mechanically, pelleted and further fixed for 1 h. The cells were immersed in agarose and fixed in 1 % osmium tetroxide for 0.5 h. The pellet was immersed in Epon LX112 after dehydration in acetone. Uranyl acetate and lead citrate were used for staining ultrathin sections and the cells were examined in a Tecnai G2 Spirit transmission electron microscope.
Immunoelectron microscopy
The cultured cells from two controls and two IPF patients were exposed to 0.5 mM pirfenidone or 0.5 mM nintedanib for 4 days in the presence of 5 ng/ml TGFβ1. Samples treated with TGFβ1 alone were used as controls. The samples were prepared as described earlier [
11]. The cells were fixed in 4 % paraformaldehyde-2.5 % sucrose, immersed first in 12 % gelatin and then in 2.3 M sucrose. Ultrathin sections were incubated with monoclonal anti-human α-SMA antibody (1:1000 dilution, clone 1A4, Dako, Glostrup, Denmark) or monoclonal anti-human fibronectin antibody (1:7000 dilution, clone IST-4, Sigma-Aldrich) followed by secondary antibody and protein A-gold conjugates. Sections were embedded in methylcellulose and visualized as TEM samples.
Collagen gel contraction assay
The contraction assay was essentially performed as described earlier [
20]. Briefly, a total of 300,000 cells and 0.75 mg collagen isolated from rat tail tendon were used for each ml of gel and then 550 μl gels were cast on 24-well plates for 15 min at 37 °C after which the gel was detached and 1 ml of serum-free medium with or without drugs was added to each well. The sizes of the gels were measured daily. The solvent of each drug was used as a control (vehicle).
Invasion
In the invasion assay, the 8 μm pore-sized Transwell inserts for 96-well plates (Corning Incorporated, Lowell, MA, USA) were coated with 50 μl of 1 g/l Matrigel (BD Biosciences) and the plates were incubated at 37 °C O/N as described previously [
11]. A total of 50,000 cells per well were plated on the top of Matrigel and cell culture medium was placed into the lower chamber. 0.5 mM pirfenidone or 0.5 μM nintedanib was added to the medium and the cells were allowed to invade the Matrigel in eight parallel wells. Distinct plates for the preparation of standard curves were prepared for each sample type. After 3 days, Matrigel and cells inside the insert were removed and the inserts and standard plates were incubated in a cell culture medium containing 0.5 g/l MTT-reagent for 2 h. The MTT-reagent was removed and the cells were lysed in DMSO and the absorbance was read as in the cell proliferation assay. The number of invading cells was evaluated from standard curves.
Statistical analysis
Statistical analysis and data visualization was performed by OriginPro 9.1 or by Statistical Package for the Social Sciences. T-test, Mann–Whitney test or non-parametric Kruskal-Wallis test (KW test) was used and p-values below 0.05 were considered as statistically significant.
Discussion
Our study demonstrates that the effects of both pirfenidone and nintedanib can be evaluated on cultured cells derived from control or IPF lung. These drugs affected not only the proliferation rate of the cells but they also inhibited myofibroblastic ultrastructural features, affected contraction of three-dimensional collagen gels and the invasive capabilities of the cells. Originally myofibroblasts were discovered during EM investigations of a healing wound [
21]. Although α-SMA is the most common marker for myofibroblasts, it is not specific since other types of cells, such as smooth muscle cells, are known to express α-SMA. So far, there is no specific marker available for the myofibroblast, and thus an EM assessment is still needed for the ultimate identification of this cell population. A typical ultrastructural feature for the myofibroblast is a fibronexus (FNX) which is composed of intracellular α-SMA and the associated extracellular fibronectin [
22]. As far as we are aware, this is the first study in which myofibroblasts cultured from IPF patients have been examined by TEM and IEM focusing on FNX in an experimental induction model which has included exposure to anti-fibrosis drugs.
Many of previous studies on pirfenidone or nintedanib have been conducted using the bleomycin-induced fibrosis model in experimental animals, like mice and hamsters. These have revealed that pirfenidone decreased the hydroxyproline level as well as reduced the extent of fibrosis and the numbers of myofibroblasts in lung [
23,
24]. Nakayama and co-workers used commercial human fibroblasts and noted that pirfenidone diminished the expression of heat shock protein 47 (HSP47) and collagen I after TGFβ treatment [
25]. Conte and co-workers reported that primary fibroblasts collected from the human lung responded to pirfenidone in vitro [
26]. Pirfenidone has been found to reduce fibroblast proliferation, TGFβ induced α-SMA and procollagen-I mRNA and protein levels, and it also inhibited the expression of factors in the TGFβ pathway [
26,
27]. Similarly, it has been shown that pirfenidone inhibited collagen gel contraction and TGFβ1 induced α-SMA production of keloid derived fibroblasts [
28]. The results of all of the above studies are consistent with and supported by the present data showing that pirfenidone decreased contraction capabilities, invasion and proliferation of the cells as well as the amount of α-SMA assessed by Western analysis in cell lines composed of fibroblasts and myofibroblasts. Furthermore, our findings were confirmed by an examination of individual myofibroblasts by IEM and TEM, in which the typical ultrastructural features of myofibroblasts were diminished after exposure to either nintedanib or pirfenidone.
In the present study, both nintedanib and pirfenidone inhibited TGFβ1 induced myofibroblast transformation. Previously it has been shown that nintedanib can induce simultaneous inhibition of several targets, i.e. platelet derived growth factor (PDGF), vascular endothelial growth factor (VEGF) and fibroblast growth factor (FGF). Two previous studies examining nintedanib have used the bleomycin-induced fibrosis model of rats and fibroblastic cell lines cultured from the patients with IPF, sarcoidosis and normal lung [
29] and bleomycin- and silica-induced fibrosis models of mice and normal human lung fibroblasts [
30]. Chaudhary and co-authors administered BIBF1000, a compound that resembles nintedanib, and demonstrated that the drug decreased the levels of fibrosis, TGFβ1, procollagen type I, fibronectin and connective tissue growth factor (CTGF) in lung tissue. In addition, they investigated the effect of BIBF1000 on primary lung fibroblasts revealing that the therapy decreased amount of α-SMA when this was estimated by Western analysis which they interpreted as a reduction in the numbers of fibroblasts differentiating into myofibroblasts [
29]. The study of Wollin and others showed that nintedanib inhibited the PDGF-induced phosphorylations of PDGF-receptors α and β as well as the proliferation of human lung fibroblasts. The α-SMA gene expression was decreased in TGFβ induced fibroblasts. In lung tissue of mice, nintedanib reduced total lung collagen levels, as well as the extent of inflammation and fibrosis and prevented granuloma formation [
30]. A recent study suggested that nintedanib can activate autophagic pathways in fibroblasts [
31]. The results of those studies are in agreement with ours which showed that nintedanib inhibited TGFβ induced myofibroblast transformation, contraction and invasion of cells and that both pirfenidone and nintedanib could reduce the ultrastructural features of individual myofibroblasts by TEM and IEM.
Our results suggest that combination of pirfenidone and nintedanib might provide enhanced efficacy in suppressing the proliferation of fibroblastic cells. However, there are no clinical studies supporting this result. Recent pilot study evaluated safety and pharmacokinetics of combined therapy [
32] and another pilot study evaluated switching the therapy from pirfenidone to nintedanib [
33]. Both of these studied contained only few patients receiving both drugs and further evaluations are required, especially as the adverse effect of these drugs partially overlap.
Clinical trials and practical experience with both agents investigated in the present study have shown that the effect of treatment on each patient is variable i.e. some patients benefit more than others from either pirfenidone or nintedanib [
34]. Currently, there are no biomarkers that can predict which patients will benefit from a particular therapeutic option. Interestingly, we noted in our study that also with these in vitro models, the results were variable when using cells cultured from several different IPF-patients. It is not known whether this in vitro phenomenon reflects the clinical behavior of the disease in certain IPF patients. In order to test this tempting hypothesis it would be necessary to prospectively recruit the IPF-patients prior to initiation of medication with these particular drugs. In vitro experimental exposures should be conducted in parallel with the therapeutic treatment of the patients and then the in vitro results could be compared with the clinical follow-up data. In the future, it may be possible that bronchoalveolar lavage samples could be collected from IPF-patients to determine whether they are likely to benefit from some particular type of therapy.
Pirfenidone and nintedanib reduced proliferation, the amount of α-SMA, collagen gel contraction properties, invasion capabilities and myofibroblastic-like ultrastructural features of cultured stromal cells derived from either healthy lung or from the lungs of patients with IPF. Proliferation was even further reduced if the cells were treated with both rugs simultaneously. Minor differences were observed between the cells obtained from control lung or from IPF as well as in the effects exerted by either pirfenidone or nintedanib. It is possible that some of these differences are caused by variable cell populations as it is known that IPF derived cells have more myofibroblast features than cells derived from normal lung [
11]. Cell culture based in vitro platforms may be useful for screening new IPF drugs in the future. The effects of the therapeutic agents were, however, variable in the samples collected from different individuals, and therefore further studies will be needed to clarify these phenomena and to determine whether the behavior of cells in vitro accurately reflects the clinical effectiveness of a pharmacological treatment.
Acknowledgements
The technical assistance of Biocenter Oulu EM laboratory personnel is gratefully acknowledged. This study was supported by Foundation of the Finnish Anti-Tuberculosis Association, Jalmari and Rauha Ahokas Foundation, Väinö and Laina Kivi Foundation, Swedish-Finnish Cultural Foundation, a state subsidy of the Oulu University Hospital, Swedish Heart-Lung Foundation and Otto A. Malm Foundation.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
STL: conception and design, acquisition, analysis and interpretation of data, drafting the manuscript, final approval of the manuscript. AV: acquisition and analysis of data (electron microscopy and proliferation), final approval of the manuscript. HK: acquisition, analysis and interpretation of data (electron microscopy), revising the manuscript, final approval of the manuscript. EL-B: acquisition, analysis and interpretation of data (patient information), revising the manuscript, final approval of the manuscript. RS: acquisition, analysis and interpretation of data (electron microscopy), revising the manuscript, final approval of the manuscript. SK: acquisition and analysis of data (Western analysis and proliferation assays), revising the manuscript, final approval of the manuscript. UZ: acquisition, analysis and interpretation of data (collagen gel contraction analysis), revising the manuscript, final approval of the manuscript. MCS: conception and design, revising the manuscript, final approval of the manuscript. RK: conception and design, collection of patients, acquisition of data (patient information), analysis and interpretation of data, drafting the manuscript, final approval of the manuscript.