Abstract
Pulmonary fibrosis is characterised by fibroblast accumulation and alveolar epithelium denudation. Increased apoptosis of alveolar epithelial cells and decreased apoptosis of fibroblasts may play an important role in the pathogenesis of disease. Inflammatory cells can modulate apoptosis of other cell types, both by removal of apoptotic debris and by cytokine production, thus preserving a pro-fibrotic environment. In the present review, some of the mechanisms by which apoptosis may contribute to the pathogenesis of idiopathic pulmonary fibrosis are described.
Idiopathic pulmonary fibrosis (IPF) is a chronic diffuse lung disease characterised by progressive deterioration in lung function ultimately leading to death. The histological pattern of IPF is usual interstitial pneumonia (UIP), described as the patchy presence of denuded alveolar epithelium, fibroblastic foci and distortion of lung architecture leading to honeycombing with minimal inflammation. It has been proposed that epithelium–fibroblast interactions may lead to alveolar cell loss and the initiation of the fibrotic process 1.
Apoptosis, or programmed cell death, is an important physiological process for the development and the maintenance of tissue homeostasis, ensuring a balance between cellular proliferation and turnover in nearly all tissue types.
Apoptosis may participate in the development of lung disease via three different mechanisms: 1) increased apoptosis of epithelial cells leading to inefficient re-epithelialisation 2; 2) resistance to apoptosis of fibroblasts leading to increased fibrosis 2; and 3) ineffective removal of apoptotic cells (efferocytosis) by granulocytes sustaining a persistent inflammatory state 3.
Although significant progress has been made with regard to the understanding of the mechanisms involved in the development of pulmonary fibrosis, the pathogenesis of the disease is not yet clear. An important part of the present knowledge has arisen from animal models, such as the bleomycin-induced pulmonary fibrosis model. However, this model has certain limitations, specifically the differences in chronicity and pathogenesis between this model and IPF. In the murine bleomycin model, acute alveolitis develops with significant inflammation followed by fibrosis in a short period of time, unlike the indolent fibrosis and minimal inflammation seen in human IPF 4.
The aim of the present review is to characterise the importance of apoptosis as a potential pathogenic mechanism in the development of pulmonary fibrosis and its relationship with other pathogenic processes.
MECHANISMS OF APOPTOSIS
Apoptotic cells undergo various morphological changes, including cell shrinkage, membrane blebbing, cleavage of chromosomal DNA and the release of the membrane-bound apoptotic bodies. Apoptosis mechanisms involve: 1) the initiation phase, during which apoptotic stimuli lead to caspase activation; and 2) the execution phase, during which caspases induce cell death.
The caspase cascade can be activated by different pathways (fig. 1⇓) 5.
Different pathways leading to caspase cascade activation and apoptosis. FasL: Fas ligand; TNF: tumour necrosis factor; ROS: reactive oxygen species; TNFR: TNF receptor; FADD: Fas activated death domain; TRADD: TNF-α activated death domain; APAF: apoptotic protease activating factor; IAP: inhibitor of apoptosis.
The extrinsic or death receptor pathway
The extrinsic or death receptor pathway involves the activation of death receptors present in the cell membrane, such as Fas and tumour necrosis factor (TNF) receptor 1. Connection of the death ligand to its death receptor leads to activation of an adaptor protein called activated death domain and the subsequent activation of procaspase-8 or -10. Activated caspases induce apoptosis. The extrinsic pathway can be inhibited by at least three mechanisms: 1) FLIP (Fas-activated death domain (FADD)-like interleukin-1 converting enzyme (FLICE)-like inhibitor of apoptosis protein), which binds to procaspase-8 without activating it; 2) the binding of decoy receptors that antagonise membrane death receptors for death ligand binding; and 3) by the heat shock proteins 6.
The intrinsic pathway
The intrinsic pathway is due to an increase of mitochondrial permeability and is activated by cellular “stresses”. Various stress factors lead to reduced expression of anti-apoptotic mitochondrial proteins (e.g. B-cell lymphoma (Bcl)-2 and Bcl-x) and to increased expression of pro-apoptotic mitochondrial proteins (e.g. Bak, Bax, Bim) 7. Other pro-apoptotic factors such as Smac (second mitochondria-derived activator of caspase)/DIABLO (direct inhibitor of apoptosis protein-binding protein with low pI), apoptosis-inducing factor and endonuclease G are also released from the mitochondria 8.
Granzyme B pathway
Cytotoxic lymphocytes, after encountering infected or malignancy-transformed host cells, secrete perforin, a protein capable of creating pores in the cell membrane of the infected cell, through which granzyme B, a serine protease, invades the host cell and induces apoptosis 9.
CELLULAR APOPTOTIC PROFILE IN PULMONARY FIBROSIS
Alveolar epithelial cells
Alveolar epithelium contains a continuous layer of two cell types: the flattened type-I pneumocytes, which cover 95% of the alveolar surface, and the type-II pneumocytes. Type-II pneumocytes can proliferate and generate both type-II and type-I cells. Moreover, they can secrete a number of molecules, such as surfactant-associated products, cytokines and growth factors, enzymes and matrix proteins 1.
Alveolar epithelial cell (AEC) apoptosis, with subsequent fibrosis, is documented in the bleomycin-induced fibrosis model of IPF 10. Additionally, the activation of the Fas–Fas ligand (FasL) pathway also promotes pulmonary fibrosis 11. Furthermore, AEC apoptosis is induced by transferring myofibroblasts from fibrotic lungs that overexpress FasL molecules into healthy murine lungs; likewise, FasL-deficient myofibroblasts do not express this cytotoxic potential 12. Blockade of either the Fas–FasL pathway or caspase inactivation by caspase inhibitors has been shown to prevent AEC apoptosis and pulmonary fibrosis 13, 14.
In lung biopsies of patients with IPF, AECs exhibit positive signals for apoptosis 15, such as p53 and p21, especially in areas adjacent to fibroblasts 16. p53 (a regulator of the cell cycle and trigger of apoptosis in response to DNA damage) is over expressed in AECs of patients with IPF 15, perhaps due to decreased degradation 17. p21 gene (an inducer of G1 arrest and DNA repair) is shown to protect AECs from apoptosis in the bleomycin-induced murine lung model of fibrosis 18. Apoptotic hyperplastic epithelial cells are present in patients with IPF, and the expression of p53, p21, Bax and caspase-3 appears to be upregulated, while Bcl-2 is downregulated (table 1⇓) 19.
Factors associated with increased alveolar epithelial cell apoptosis in idiopathic pulmonary fibrosis
Fibroblasts
Fibroblasts synthesise and deposit collagen and other extracellular matrix components within the fibrotic lesions of UIP 1. During normal wound healing, fibroblast number is reduced by apoptosis 20.
IPF fibroblasts have been reported to be resistant to Fas-mediated apoptosis 21. Resistance is attributed to the increased expression of X-linked inhibitor of apoptosis and FLIP proteins 22, as well as to decreased levels of surface Fas and increased levels of soluble Fas 23. The expression of discoid-domain receptor I, a tyrosine kinase whose ligand is collagen, is increased in fibroblasts of mice with bleomycin-induced fibrosis and it promotes resistance to apoptosis 24.
Transforming growth factor (TGF)-β1, the main pro-fibrotic cytokine, independently activates the two important anti-apoptotic pathways in lung fibroblasts: the focal adhesion kinase (FAK) pathway by activation of Smad3 and the phosphatidylo-inositol 3 kinase (PI3K)/Akt pathway via activation of p38 mitogen-activated protein kinase 25.
Thy-1, a cell surface glucoprotein, has been associated with increased apoptosis of lung fibroblasts. It has been shown that the myofibroblasts of fibroblastic foci are Thy-1 negative and that Thy-1 negativity offers apoptosis resistance (table 2⇓) 26.
Factors associated with decreased fibroblast apoptosis in idiopathic pulmonary fibrosis
Macrophages
Macrophage activation has been implicated in the pathogenesis of fibrotic lung diseases, such as the Hermansky–Pudlak syndrome 27. Bleomycin has been shown to induce alveolar macrophage apoptosis in experimental models 28, 29, while intratracheal administration of apoptotic macrophages to the lungs of rats causes increased macrophage infiltration and apoptosis, and increased collagen deposition 30, 31. Increased macrophage expression of Bcl-x and Bax proteins, as well as caspases-1 and -3 in bleomycin-induced fibrosis 14, 32, and Bcl-2 and Fas expression in alveolar macrophages of patients with IPF, has been reported 33, 34.
Moreover, mice deficient in macrophage colony-stimulating factor (M-CSF), a factor associated with the survival of macrophages, develop less pulmonary fibrosis and have a decreased number of macrophages in their lungs after bleomycin installation. M-CSF levels are significantly higher in patients with IPF compared with normal subjects 35. Macrophages can also inhibit myofibroblast apoptosis through the release of insulin-like growth factor-1 by macrophages (table 3⇓) 36.
Putative mechanisms of macrophage participation in apoptosis in idiopathic pulmonary fibrosis
Neutrophils
Although impaired neutrophil apoptosis has been associated with a number of respiratory diseases 37, its role in the pathogenesis of IPF is not clearly understood. The presence of increased numbers of neutrophils in the lungs of patients with IPF is associated with dismal prognosis 38. The anti-apoptotic protein Bcl-2 is significantly upregulated in the neutrophils of patients with IPF 34.
Corticosteroids, although minimally active as a therapeutic option in IPF, have been shown to significantly increase the rate of neutrophil apoptosis, while simultaneously reducing the rate of alveolitis and subsequent fibrosis in rats after instillation of bleomycin 39. However, others have reported that neutrophilia is not associated with poorer clinical course and that bleomycin-induced fibrosis is enhanced in the absence of neutrophils 40, 41.
The uncertainty of whether the neutrophilic accumulation in the end stages of disease is a significant pathogenic mechanism or an epiphenomenon of advanced fibrotic disease highlights current uncertainties about the pathogenesis of IPF. Currently, inflammation is not thought to play a major pathogenic role in IPF 1, 2. In advanced disease, an influx of neutrophils might result from either sub-clinical infection in destroyed lung or the fibrogenetic secretion of chemotactic factors. However, it is also possible that neutrophils may play a primary pathogenic role in acute exacerbations of IPF 42, which may account for a neutrophil influx in some cases.
INTERRELATIONSHIPS BETWEEN APOPTOSIS AND OTHER PULMONARY FIBROSIS MECHANISMS
Pathogenesis of pulmonary fibrosis remains a complicated issue. Numerous pathogenic mechanisms are implicated and interrelate, resulting in significant difficulty in discriminating which are the primary and secondary events in the development of the disease. In a recent review, it was proposed that injury due to various agents (chronic viral infections, cigarette smoke, gastro-oesophageal reflux disease and exposure to environmental pollutants) activates multiple pathways (oxidant–antioxidant, coagulation, inflammation and T-helper cell (Th) type 1/Th2 cytokines) and causes an imbalance between pro-fibrotic (TGF-β, connective tissue growth factor and thrombin) and anti-fibrotic molecules (interferon-γ and prostaglandin E2). This imbalance induces a change in cellular function, expressed as altered apoptotic cell behaviour and proliferation, epithelial-mesenchymal transition and increased extracellular matrix production, ultimately leading to fibrosis 42. In the following section, the interrelationships of apoptosis with other fibrotic mechanisms will be reviewed.
Vasoconstrictors
It has been shown in vitro and in animal models that vasoconstrictors participate in the development of fibrosis. Two of the most extensively studied vasoconstrictors are endothelin-1 and angiotensin-II. Although usually associated with resistance to apoptosis 43, endothelin-1 promotes epithelial-mesenchymal transition (the acquisition of mesenchymal phenotypic characteristics by epithelial cells) in IPF 44, while angiotensin peptides actively participate in apoptosis.
Angiotensin peptides are locally produced both by apoptotic AECs and by adjacent myofibroblasts 45. AEC apoptosis is induced via Fas activation by angiotensin-II, which is produced by human lung fibroblasts 46, 47, and it can be attenuated by the administration of the angiotensin-converting enzyme inhibitor captopril 48, by the selective angiotensin type 1 receptor antagonist losartan 49 or by antisense oligonucleotides against angiotensinogen mRNA 50.
Oxidative stress
Oxidative stress has been thought to significantly contribute to epithelial cell damage in IPF 51. Although the source of reactive oxygen species (ROS) in the lung is considered to be inflammatory cells, myofibroblasts can also produce hydrogen peroxide, which can, in turn, induce epithelial cell death 52. Bleomycin induces AEC apoptosis and fibrosis by an increase of ROS 53, and free radical scavengers decrease the rate of apoptosis and pulmonary fibrosis 54. Bleomycin-induced apoptosis has been associated with p53 translocation from the cytosol to the nucleus, caused by ROS in macrophages 55.
Oxidative stress also affects fibroblast apoptosis: hydrogen peroxide causes apoptosis of fibroblasts stimulated to proliferate or migrate to the wound 56.
Hypoxia
Lack of oxygen significantly disturbs AEC function, triggers apoptosis and promotes inflammation. Hypoxia-inducible factor (HIF)-1 is a major regulator of hypoxic signalling and is overexpressed in patients with IPF.
HIF-1α causes increased epithelial cell apoptosis via Bnip 3L activation, a member of the bcl2 family. Blockade of HIF-1α or Bnip3L (Bcl-2/adenovirus E1B 19-kDa interacting protein 3L) significantly attenuates hypoxia-induced epithelial cell apoptosis 57. Expression of HIF-1α has been shown to be an early event in IPF and to correlate with activation of p53 and increased AEC apoptosis in lungs of patients with IPF and in bleomycin-induced fibrosis. HIF-1α is absent in fibroblastic foci, whereas the anti-apoptotic protein Bcl-2 is increasingly expressed, further supporting the hypothesis of increased AEC apoptosis concomitantly with decreased fibroblast apoptosis 58.
Extracellular matrix
Increased extracellular matrix deposition is a hallmark of IPF and is associated with a dysregulation of matrix metalloproteases (MMPs) and their inhibitors 2.
The integrity of the basement membrane offers survival signals for adherent epithelial cells, while disruption of the adhesion leads to apoptotic cell death. The extracellular matrix may dictate the apoptotic profile of AECs. When cultured on laminin/collagen mixtures, epithelial cells undergo apoptosis, while culture on fibronectin or fibrin causes epithelial-mesenchymal transition 59.
Matrix composition can lead to different apoptotic fibroblast responses 60. Soluble fibronectin peptides can promote fibroblast apoptosis by disruption of adhesion via activation of the integrin-FAK survival pathway 61; however, fibronectin can also reduce fibroblast apoptosis via activation of the PI3K pathway 62.
TGF-β1
Overproduction of TGF-β1, a growth factor essential for wound healing, can result in excessive deposition of scar tissue and fibrosis 63. TGF-β1 directly induces epithelial cell apoptosis via Fas and caspase-3 activation, and by enhancing the FasL−Fas interaction 64, while mice transfection with soluble TGF type-II receptors significantly attenuates the degree of apoptosis and pulmonary fibrosis 65.
Bid, another Bcl-2 family member is required for AEC apoptosis and bleomycin-induced fibrosis in mice after TGF-β1 activation as Bid-deficient mice are protected from developing fibrosis 66. The Bax-mediated, Bid-activated pathway has recently been shown to be involved in the pathogenesis of pulmonary fibrosis. TGF-β1 significantly stimulates Bax and Bid expression and causes the release of MMP-12 and tissue inhibitor of metalloprotease-1 in mice 67.
TGF-β1 causes increased p21 expression in AECs via a TNF-α signalling pathway and absence of p21 expression is associated with increased TGF-β1-induced fibrosis and epithelium apoptosis 68.
TGF-β receptor type-I inhibitors are reported to decrease bleomycin-induced fibrosis and myofibroblast apoptosis 69. As previously mentioned, TGF-β activates the FAK and the PI3K/Akt anti-apoptotic pathways in lung fibroblasts 25.
Inflammation
Inflammation is not considered to be a major pathogenic component in the development of IPF. However, it plays a significant role in acute exacerbations of the disease and it may be necessary for the initiation of the fibrotic process 42.
Inflammation can lead to increased apoptosis. The pro-inflammatory cytokine TNF-α sensitises fibroblasts to Fas-mediated apoptosis and interferon-γ, a Th1-type cytokine, can act synergistically with TNF-α for the enhancement of the apoptotic response 70.
Neutrophils can also induce AEC apoptosis: infiltrating granulocytes in lung biopsies and bronchoalveolar lavage of patients with IPF expressed significantly higher levels of FasL compared with controls, and this was associated with an increased expression of Fas in the adjacent epithelium 33.
Failure to remove excessive numbers of apoptotic cells may induce a persistent inflammatory state. Macrophages are considered to be the professional phagocytes of apoptotic cells, which can engulf and deposit apoptotic remnants 3. Macrophages ingesting apoptotic cells release the anti-inflammatory cytokines interleukin (IL)-10 and TGF-β1, which act locally 71. Furthermore, apoptotic cells can produce IL-10 and TGF-β1, thus enhancing the phagocytic capacity of macrophages 72. Macrophages which phagocytose apoptotic bodies can also release pro-apoptotic factors that induce apoptosis of adjacent cells 73. However, cleavage of apoptotic remnants may also lead to inflammation enhancement 74. Interestingly, dexamethasone has been reported to induce apoptosis of pulmonary inflammatory cells and reduce the extent of bleomycin-induced fibrosis 39.
Pro-coagulant activity
After tissue injury, activation of the coagulation cascade rapidly takes place in order to provide a provisional extracellular matrix for the repair process to occur 1. In pulmonary fibrosis, excess extravascular coagulation has been found, largely due to the production of pro-coagulant molecules, such as tissue factor and plasminogen-activating inhibitors, by AECs 75. Furthermore, anticoagulants can effectively attenuate pulmonary fibrosis in animal models 75 and, in a randomised clinical trial of patients with IPF, addition of anticoagulants to corticosteroids significantly reduced acute exacerbations of IPF and improved survival 76.
Apoptosis of AECs and subsequent basement membrane denudation is a signal for tissue repair and initiation of the coagulation cascade 75. Moreover, it has been shown that pro-coagulants can initiate apoptosis. Activation of protease-activating receptor (PAR)-1, a high-affinity receptor for thrombin, induces alveolar epithelial cell apoptosis in vitro 77. PAR-1 also contributes to fibrogenesis via activation of TGF-β1 78.
AGEING, APOPTOSIS AND PULMONARY FIBROSIS
Pulmonary fibrosis is an age-related process. IPF is a disease affecting older individuals 2. Patients with hereditary forms of pulmonary fibrosis, such as Hermansky–Pudlak syndrome, also develop pulmonary fibrosis as they age 79. Dysregulation of apoptosis pathways is implicated in aging. Senescent cells affect organ repair and structure and promote local inflammation 80.
Activation of the p53 pathway is a cardinal event in the process of cell senescence 80. Activation of p53 has also been shown to occur in the apoptotic process observed in AECs of patients with IPF 15. Increased AEC apoptosis in senescent rats exposed to hyperoxia and oxidative stress has been reported 81. Senescent human fibroblasts are resistant to apoptosis and this resistance is associated with increased expression of the anti-apoptotic factor Bcl-2 and downregulation of caspase-3 82.
Telomerase is a ribonucleoprotein enzyme that maintains telomere length and retains cellular and subsequent organism lifespan. Dysfunctional telomeres trigger cell cycle arrest or apoptosis in mice 83. Mutations in telomerase components, associated with defective telomerase activity and shortened telomeres, have been found in some cases of IPF 84 and telomerase regulation has been implicated in bleomycin-induced fibrosis. After bleomycin administration, AECs show a decreased telomerase activity in vitro, resulting in increased apoptosis, while in vivo telomerase activity increases in order to protect alveolar epithelium from bleomycin-induced apoptosis 85.
However, telomerase activity is required for fibrosis. Lung fibroblasts of mice with bleomycin-induced fibrosis with telomerase deficiency show decreased proliferation and increased apoptotic rates compared with controls, while restoration of telomerase activity results in increased lung fibrosis 86.
CONCLUSION
Accumulating evidence supports the importance of apoptosis as a potential pathogenic mechanism in idiopathic pulmonary fibrosis. Apoptosis can be either beneficial or detrimental to the organism. Different cell types exhibit different apoptotic behaviour. The pro-fibrotic environment induces epithelial cell apoptosis and increased myofibroblast survival. Inflammatory cells can modulate the fibrotic responses of the lung, both by self-apoptosis and removal of apoptotic debris. Targeting specific cell types with anti-apoptotic agents represents a major challenge in therapeutic intervention for debilitating diseases such as idiopathic pulmonary fibrosis. However, apoptosis is one of many mechanisms involved in the development of idiopathic pulmonary fibrosis and the complexity of interactions between these different mechanisms remains far from being resolved.
Statement of interest
None declared.
Acknowledgments
The authors would like to thank V. Fayngersh (Alpert Medical School, Brown University, Providence, RI, USA) for his valuable comments and for linguistic review of the paper.
- Received December 31, 2007.
- Accepted August 6, 2008.
- © ERS Journals Ltd
