Chemokine-mediated angiogenesis: an essential link in the evolution of airway fibrosis
http://www.100md.com
《临床调查学报》
1Department of Medicine, Pulmonary Sciences and Critical Care Medicine and
2Denver Health Medical Center, University of Colorado Health Sciences Center, Denver, Colorado, USA.
Abstract
Angiogenesis may be an important factor in the development of fibrotic lung disease. Prior studies have strongly suggested a role for angiogenic vascular remodeling in pulmonary fibrosis, and emerging evidence indicates that new vessel formation is critical in airway fibrosis. Bronchiolitis obliterans syndrome is a fibrotic occlusion of distal airways that is largely responsible for the morbidity and mortality of patients after lung transplantation. In this issue, Belperio et al. demonstrate a role for CXC chemokine receptor 2 in the regulation of angiogenesis-mediated airway fibroproliferation. By integrating an understanding of neovascularization into the study of events that occur between inflammation and fibrosis, it becomes increasingly possible to rationally design therapies that can halt conditions of maladaptive fibrosis.
See the related article beginning on page 1150
Neovascularization is an important component of fibrotic responses (1). In this issue of the JCI, Belperio and colleagues extend this relationship to the development of chronic lung transplant rejection (2). Using bronchoalveolar lavage fluid from patients with pending or established bronchiolitis obliterans syndrome (BOS) and tracheal allograft tissue from a mouse model of obliterative airway disease, the authors make a convincing case for the central role of CXC chemokine receptor 2 (CXCR2) regulation of angiogenesis-mediated airway fibroproliferation.
Airway inflammation and fibrosis in the evolution of BOS
Chronic allograft rejection is the chief factor limiting long-term survival following lung transplantation. BOS is the pathological correlate of chronic rejection and primarily affects the respiratory and terminal bronchioles, which culminates in a fibrotic occlusion of the distal airways (3). The cumulative incidence of BOS at 5 years after lung transplant is between 50% and 80%, and 5-year survival after BOS onset is only 30–50%. The International Society of Heart and Lung Transplantation Registry has noted that the development of BOS within the first year after transplantation is the single most important factor influencing 5-year mortality among patients undergoing lung transplantation (3). As a fibrotic disease, BOS is poorly responsive to standard immunosuppression employed by transplant physicians. Similarly, pulmonary fibrosis, which affects the lung interstitium rather than the conducting airways, responds poorly to immunotherapy and has long been associated with pathologic angiogenesis (4).
It is a generally recognized phenomenon that inflammation is an initiating event that precedes the progression to fibrosis in several lung diseases, including BOS and idiopathic pulmonary fibrosis. While fibrosis may be a frequent sequel of an acute or subacute inflammatory event, it is also clear that inflammation does not always result in fibrosis. The long-term effect of interstitial or airway fibrosis is irreversible lung architectural remodeling. Key questions regarding the mechanisms of airway remodeling are: (a) What are the specific inflammatory initiators and (b) What is the sequence of events that culminates in fibroproliferation In lung transplantation, the answer to the first question most certainly involves the response to alloantigen triggering of innate and adaptive immune responses. The answer to the second question is probably less well understood but is perhaps of greater importance in the development of therapies that reach beyond immunosuppression. Lung transplant clinicians well appreciate that acute rejection treated early may respond excellently to immunosuppressive therapies but that late intervention is rarely successful. Unfortunately, it is not always possible to intervene early, and occasionally even apparently early intervention with high-dose steroids or T cell–depleting strategies cannot halt the decline in lung function once fibroproliferation is initiated.
The potential role of CXC chemokines in angiofibroproliferative BOS
The study by Belperio and colleagues (2) firmly establishes that neovascularization is an important contributor to the process of fibroproliferation in airway fibrosis. The investigators present a cohesive and clearly argued interpretation of experimental data from human BOS patients and a well-characterized murine model of tracheal transplant rejection. Their findings make a convincing case for the central role of CXCR2-dependent Glu-Leu-Arg–positive (ELR+) chemokine regulation of angiogenesis-mediated BOS fibroproliferation. The study extends their previously published observations based on the BOS model, which implicated monocyte chemoattractant protein-1/CXCR2–mediated mononuclear phagocytic infiltration in the progression of BOS (5). CXC chemokines, 1 of the 4 subfamilies of multifunctional growth- and immune-mediating molecules, regulate acute inflammatory responses and vascular remodeling. CXC chemokines contain an ELR+ motif immediately before the first cysteine residue at the NH2 terminus. This group includes CXC ligand 1 (CXCL1), -2, -3, -5, -6, and -8. Angiogenesis (both physiological and pathological) is potently stimulated by engagement of these ligands with G protein–coupled CXCR2 on endothelial cells. By contrast, IFN-regulated ELR– chemokines such as CXCL4, -9, -10, and -11 function as angio-static regulators through CXCR3 and direct interaction with angiogenic factors such as VEGF (6).
The elucidation of VEGF-independent, chemokine-mediated angiogenesis is pivotal to our evolving understanding of the cascade of remodeling events that results in inexorable progression of fibroproliferative BOS. The present study also demonstrates a temporal separation of signaling responses to CXCR2 activation. CXCR2 engagement in early murine tracheal allograft BOS results in neutrophil trafficking, but at 3 weeks results in neutrophil-independent angiofibroproliferation (2). Moreover, treatment of CXCR2–/– mice with threshold doses of cyclosporin, insufficient to prevent BOS in wild-type mice but sufficient to inhibit early monocyte infiltration, resulted in a dramatic reduction in allograft fibro-obliteration, which points to a clinically relevant strategy for further investigation.
How does pathological angiogenesis fit into what is already known or generally posited for lung fibrosis Briefly, there is evidence to suggest that the epithelium can be an early target as well as a source of alloantigen (7). Following exposure to a new antigen or toxin, there is an influx of neutrophils and mononuclear cells (Figure 1A). This period of cellular infiltration is associated with an increased production of chemokines and cytokines. In this cascade, multiple changes occur in the epithelia, including apoptosis, hyperplasia, dedifferentiation, and metaplasia (8-12). Epithelial changes may stimulate underlying myofibroblasts in a paracrine fashion, with resultant collagen matrix deposition (13-15). In the midst of these changes, neovascularization is observed (Figure 1B). Perhaps as a consequence of increased vascularity and chemoattractant gradients, migratory cell populations (e.g., fibrocytes; ref. 16) traffic to the site of inflammation, possibly as part of a tissue repair response. As this process progresses, the cellular infiltrates diminish (8, 9), and airway remodeling proceeds in a unidirectional fashion, with extensive subepithelial, luminal, and/or interstitial fibrosis. Our group has recently established using an orthotopic (in situ) mouse tracheal transplant model that it is possible to reverse alloimmune injury that remains unmitigated for 7 days, but after 10 days of unprotected immune airway injury, fibroproliferative diseases cannot be ameliorated (8). Thus, it appears that a temporal sequence of events leading from inflammation to fibrosis may be closer to being fully elucidated.
In addition to alloimmune injury, several important potential amplifiers of BOS have been identified. These include chronic gastroesophageal reflux (17), community-acquired viral infection (18), and CMV infection (19), all of which are also associated with chronic graft rejection. Chemokine-mediated inflammation and angiogenesis may be a common pathway for progression to fibroproliferative BOS in response to superimposition of these factors on an immune-injured airway. For example, virulent CMV strains produce a potent ELR+ chemokine analog, UL146 (also known as viral CXCL1). This CXCR2-restricted ligand is capable of inducing neutrophil chemotaxis independent of monocyte activation (20) and could potentially contribute to chemokine-mediated perigraft angiogenesis and fibroproliferation.
Belperio and colleagues’ important observations (2) hold great promise for translational application in the therapy of BOS and other fibrotic lung diseases that involve pathological angiofibroproliferation. Presently, prevention of graft rejection requires intensive, prolonged immunosuppression with corticosteroids, calcineurin antagonists, antimetabolites, and immune modulators (21). This approach, which is focused on the inflammatory component, has clear limitations, with significantly increased risks of opportunistic infection and a limited effect on angiogenesis. An obvious therapeutic strategy would be treatment with a combination of lower doses of traditional immunosuppressants and humanized blocking antibodies against angiogenic chemokines or their receptors, as modeled in the present study (2). If this broad framework holds true for most fibrotic conditions, in the future, it may be possible to more appropriately treat other pulmonary, renal, dermatological, and cardiac diseases in which the inflammation-angiogenesis-fibroproliferative pathways are active.
Footnotes
Nonstandard abbreviations used: BOS, bronchiolitis obliterans syndrome; CXCL, CXC chemokine ligand; CXCR2, CXC chemokine receptor 2.
Conflict of interest: The authors have declared that no conflict of interest exists.
References
Kalluri, R., and Sukhatme, V.P. 2000. . Fibrosis and angiogenesis. Curr. Opin. Nephrol. Hypertens. 9::413-418.
Belperio, J. et al. 2005. . Role of CXCR2/CXCR2 ligands in vascular remodeling during bronchiolitis obliterans syndrome. J. Clin. Invest. 115::1150-1162. doi:10.1172/JCI200524233.
Estenne, M., and Hertz, M.I. 2002. . Bronchiolitis obliterans after human lung transplantation. Am. J. Respir. Crit. Care Med. 166::440-444.
Turner-Warwick, M. 1963. . Precapillary systemic-pulmonary anastomoses. Thorax. 18::225-237.
Belperio, J.A. et al. 2001. . Critical role for the chemokine MCP-1/CCR2 in the pathogenesis of bronchiolitis obliterans syndrome. J. Clin. Invest. 108::547-556. doi:10.1172/JCI200112214.
Belperio, J.A. et al. 2000. . CXC chemokines in angiogenesis. J. Leukoc. Biol. 68::1-8.
Fernandez, F.G. et al. 2004. . Airway epithelium is the primary target of allograft rejection in murine obliterative airway disease. Am. J. Transplant. 4::319-325.
Murakawa, T. et al. 2005. . Simultaneous LFA-1 and CD40 ligand antagonism prevents airway remodeling in orthotopic airway transplantation: implications for the role of respiratory epithelium as a modulator of fibrosis. J. Immunol. 174::3869-3879.
Minamoto, K., and Pinsky, D.J. 2002. . Recipient iNOS but not eNOS deficiency reduces luminal narrowing in tracheal allografts. J. Exp. Med. 196::1321-1333.
Neuringer, I.P. et al. 2002. . Epithelial kinetics in mouse heterotopic tracheal allografts. Am. J. Transplant. 2::410-419.
Reader, J.R. et al. 2003. . Pathogenesis of mucous cell metaplasia in a murine asthma model. Am. J. Pathol. 162::2069-2078.
Kasper, M., and Haroske, G. 1996. . Alterations in the alveolar epithelium after injury leading to pulmonary fibrosis. Histol. Histopathol. 11::463-483.
Howat, W.J., Holgate, S.T., and Lackie, P.M. 2002. . TGF-beta isoform release and activation during in vitro bronchial epithelial wound repair. Am. J. Physiol. Lung Cell Mol. Physiol. 282::L115-L123.
Morishima, Y. et al. 2001. . Triggering the induction of myofibroblast and fibrogenesis by airway epithelial shedding. Am. J. Respir. Cell Mol. Biol. 24::1-11.
Zhang, S., Smartt, H., Holgate, S.T., and Roche, W.R. 1999. . Growth factors secreted by bronchial epithelial cells control myofibroblast proliferation: an in vitro co-culture model of airway remodeling in asthma. Lab. Invest. 79::395-405.
Phillips, R.J. et al. 2004. . Circulating fibrocytes traffic to the lungs in response to CXCL12 and mediate fibrosis. J. Clin. Invest. 114::438-446. doi:10.1172/JCI200420997.
Palmer, S.M. et al. 2000. . Gastroesophageal reflux as a reversible cause of allograft dysfunction after lung transplantation. Chest. 118::1214-1217.
Chakinala, M.M., and Walter, M.J. 2004. . Community acquired respiratory viral infections after lung transplantation: clinical features and long-term consequences. Semin. Thorac. Cardiovasc. Surg. 16::342-349.
Zamora, M.R. 2004. . Cytomegalovirus and lung transplantation. Am. J. Transplant. 4::1219-1226.
Penfold, M.E. et al. 1999. . Cytomegalovirus encodes a potent alpha chemokine. Proc. Natl. Acad. Sci. U. S. A. 96::9839-9844.
Knoop, C., Haverich, A., and Fischer, S. 2004. . Immunosuppressive therapy after human lung transplantation. Eur. Respir. J. 23::159-171.
Selman, M., and Pardo, A. 2002. . Idiopathic pulmonary fibrosis: an epithelial/fibroblastic cross-talk disorder. Respir. Res. [serial online]. 3::3-http://respiratory-research.com/content/3/1/3.(Ivor S. Douglas, and Mark)
2Denver Health Medical Center, University of Colorado Health Sciences Center, Denver, Colorado, USA.
Abstract
Angiogenesis may be an important factor in the development of fibrotic lung disease. Prior studies have strongly suggested a role for angiogenic vascular remodeling in pulmonary fibrosis, and emerging evidence indicates that new vessel formation is critical in airway fibrosis. Bronchiolitis obliterans syndrome is a fibrotic occlusion of distal airways that is largely responsible for the morbidity and mortality of patients after lung transplantation. In this issue, Belperio et al. demonstrate a role for CXC chemokine receptor 2 in the regulation of angiogenesis-mediated airway fibroproliferation. By integrating an understanding of neovascularization into the study of events that occur between inflammation and fibrosis, it becomes increasingly possible to rationally design therapies that can halt conditions of maladaptive fibrosis.
See the related article beginning on page 1150
Neovascularization is an important component of fibrotic responses (1). In this issue of the JCI, Belperio and colleagues extend this relationship to the development of chronic lung transplant rejection (2). Using bronchoalveolar lavage fluid from patients with pending or established bronchiolitis obliterans syndrome (BOS) and tracheal allograft tissue from a mouse model of obliterative airway disease, the authors make a convincing case for the central role of CXC chemokine receptor 2 (CXCR2) regulation of angiogenesis-mediated airway fibroproliferation.
Airway inflammation and fibrosis in the evolution of BOS
Chronic allograft rejection is the chief factor limiting long-term survival following lung transplantation. BOS is the pathological correlate of chronic rejection and primarily affects the respiratory and terminal bronchioles, which culminates in a fibrotic occlusion of the distal airways (3). The cumulative incidence of BOS at 5 years after lung transplant is between 50% and 80%, and 5-year survival after BOS onset is only 30–50%. The International Society of Heart and Lung Transplantation Registry has noted that the development of BOS within the first year after transplantation is the single most important factor influencing 5-year mortality among patients undergoing lung transplantation (3). As a fibrotic disease, BOS is poorly responsive to standard immunosuppression employed by transplant physicians. Similarly, pulmonary fibrosis, which affects the lung interstitium rather than the conducting airways, responds poorly to immunotherapy and has long been associated with pathologic angiogenesis (4).
It is a generally recognized phenomenon that inflammation is an initiating event that precedes the progression to fibrosis in several lung diseases, including BOS and idiopathic pulmonary fibrosis. While fibrosis may be a frequent sequel of an acute or subacute inflammatory event, it is also clear that inflammation does not always result in fibrosis. The long-term effect of interstitial or airway fibrosis is irreversible lung architectural remodeling. Key questions regarding the mechanisms of airway remodeling are: (a) What are the specific inflammatory initiators and (b) What is the sequence of events that culminates in fibroproliferation In lung transplantation, the answer to the first question most certainly involves the response to alloantigen triggering of innate and adaptive immune responses. The answer to the second question is probably less well understood but is perhaps of greater importance in the development of therapies that reach beyond immunosuppression. Lung transplant clinicians well appreciate that acute rejection treated early may respond excellently to immunosuppressive therapies but that late intervention is rarely successful. Unfortunately, it is not always possible to intervene early, and occasionally even apparently early intervention with high-dose steroids or T cell–depleting strategies cannot halt the decline in lung function once fibroproliferation is initiated.
The potential role of CXC chemokines in angiofibroproliferative BOS
The study by Belperio and colleagues (2) firmly establishes that neovascularization is an important contributor to the process of fibroproliferation in airway fibrosis. The investigators present a cohesive and clearly argued interpretation of experimental data from human BOS patients and a well-characterized murine model of tracheal transplant rejection. Their findings make a convincing case for the central role of CXCR2-dependent Glu-Leu-Arg–positive (ELR+) chemokine regulation of angiogenesis-mediated BOS fibroproliferation. The study extends their previously published observations based on the BOS model, which implicated monocyte chemoattractant protein-1/CXCR2–mediated mononuclear phagocytic infiltration in the progression of BOS (5). CXC chemokines, 1 of the 4 subfamilies of multifunctional growth- and immune-mediating molecules, regulate acute inflammatory responses and vascular remodeling. CXC chemokines contain an ELR+ motif immediately before the first cysteine residue at the NH2 terminus. This group includes CXC ligand 1 (CXCL1), -2, -3, -5, -6, and -8. Angiogenesis (both physiological and pathological) is potently stimulated by engagement of these ligands with G protein–coupled CXCR2 on endothelial cells. By contrast, IFN-regulated ELR– chemokines such as CXCL4, -9, -10, and -11 function as angio-static regulators through CXCR3 and direct interaction with angiogenic factors such as VEGF (6).
The elucidation of VEGF-independent, chemokine-mediated angiogenesis is pivotal to our evolving understanding of the cascade of remodeling events that results in inexorable progression of fibroproliferative BOS. The present study also demonstrates a temporal separation of signaling responses to CXCR2 activation. CXCR2 engagement in early murine tracheal allograft BOS results in neutrophil trafficking, but at 3 weeks results in neutrophil-independent angiofibroproliferation (2). Moreover, treatment of CXCR2–/– mice with threshold doses of cyclosporin, insufficient to prevent BOS in wild-type mice but sufficient to inhibit early monocyte infiltration, resulted in a dramatic reduction in allograft fibro-obliteration, which points to a clinically relevant strategy for further investigation.
How does pathological angiogenesis fit into what is already known or generally posited for lung fibrosis Briefly, there is evidence to suggest that the epithelium can be an early target as well as a source of alloantigen (7). Following exposure to a new antigen or toxin, there is an influx of neutrophils and mononuclear cells (Figure 1A). This period of cellular infiltration is associated with an increased production of chemokines and cytokines. In this cascade, multiple changes occur in the epithelia, including apoptosis, hyperplasia, dedifferentiation, and metaplasia (8-12). Epithelial changes may stimulate underlying myofibroblasts in a paracrine fashion, with resultant collagen matrix deposition (13-15). In the midst of these changes, neovascularization is observed (Figure 1B). Perhaps as a consequence of increased vascularity and chemoattractant gradients, migratory cell populations (e.g., fibrocytes; ref. 16) traffic to the site of inflammation, possibly as part of a tissue repair response. As this process progresses, the cellular infiltrates diminish (8, 9), and airway remodeling proceeds in a unidirectional fashion, with extensive subepithelial, luminal, and/or interstitial fibrosis. Our group has recently established using an orthotopic (in situ) mouse tracheal transplant model that it is possible to reverse alloimmune injury that remains unmitigated for 7 days, but after 10 days of unprotected immune airway injury, fibroproliferative diseases cannot be ameliorated (8). Thus, it appears that a temporal sequence of events leading from inflammation to fibrosis may be closer to being fully elucidated.
In addition to alloimmune injury, several important potential amplifiers of BOS have been identified. These include chronic gastroesophageal reflux (17), community-acquired viral infection (18), and CMV infection (19), all of which are also associated with chronic graft rejection. Chemokine-mediated inflammation and angiogenesis may be a common pathway for progression to fibroproliferative BOS in response to superimposition of these factors on an immune-injured airway. For example, virulent CMV strains produce a potent ELR+ chemokine analog, UL146 (also known as viral CXCL1). This CXCR2-restricted ligand is capable of inducing neutrophil chemotaxis independent of monocyte activation (20) and could potentially contribute to chemokine-mediated perigraft angiogenesis and fibroproliferation.
Belperio and colleagues’ important observations (2) hold great promise for translational application in the therapy of BOS and other fibrotic lung diseases that involve pathological angiofibroproliferation. Presently, prevention of graft rejection requires intensive, prolonged immunosuppression with corticosteroids, calcineurin antagonists, antimetabolites, and immune modulators (21). This approach, which is focused on the inflammatory component, has clear limitations, with significantly increased risks of opportunistic infection and a limited effect on angiogenesis. An obvious therapeutic strategy would be treatment with a combination of lower doses of traditional immunosuppressants and humanized blocking antibodies against angiogenic chemokines or their receptors, as modeled in the present study (2). If this broad framework holds true for most fibrotic conditions, in the future, it may be possible to more appropriately treat other pulmonary, renal, dermatological, and cardiac diseases in which the inflammation-angiogenesis-fibroproliferative pathways are active.
Footnotes
Nonstandard abbreviations used: BOS, bronchiolitis obliterans syndrome; CXCL, CXC chemokine ligand; CXCR2, CXC chemokine receptor 2.
Conflict of interest: The authors have declared that no conflict of interest exists.
References
Kalluri, R., and Sukhatme, V.P. 2000. . Fibrosis and angiogenesis. Curr. Opin. Nephrol. Hypertens. 9::413-418.
Belperio, J. et al. 2005. . Role of CXCR2/CXCR2 ligands in vascular remodeling during bronchiolitis obliterans syndrome. J. Clin. Invest. 115::1150-1162. doi:10.1172/JCI200524233.
Estenne, M., and Hertz, M.I. 2002. . Bronchiolitis obliterans after human lung transplantation. Am. J. Respir. Crit. Care Med. 166::440-444.
Turner-Warwick, M. 1963. . Precapillary systemic-pulmonary anastomoses. Thorax. 18::225-237.
Belperio, J.A. et al. 2001. . Critical role for the chemokine MCP-1/CCR2 in the pathogenesis of bronchiolitis obliterans syndrome. J. Clin. Invest. 108::547-556. doi:10.1172/JCI200112214.
Belperio, J.A. et al. 2000. . CXC chemokines in angiogenesis. J. Leukoc. Biol. 68::1-8.
Fernandez, F.G. et al. 2004. . Airway epithelium is the primary target of allograft rejection in murine obliterative airway disease. Am. J. Transplant. 4::319-325.
Murakawa, T. et al. 2005. . Simultaneous LFA-1 and CD40 ligand antagonism prevents airway remodeling in orthotopic airway transplantation: implications for the role of respiratory epithelium as a modulator of fibrosis. J. Immunol. 174::3869-3879.
Minamoto, K., and Pinsky, D.J. 2002. . Recipient iNOS but not eNOS deficiency reduces luminal narrowing in tracheal allografts. J. Exp. Med. 196::1321-1333.
Neuringer, I.P. et al. 2002. . Epithelial kinetics in mouse heterotopic tracheal allografts. Am. J. Transplant. 2::410-419.
Reader, J.R. et al. 2003. . Pathogenesis of mucous cell metaplasia in a murine asthma model. Am. J. Pathol. 162::2069-2078.
Kasper, M., and Haroske, G. 1996. . Alterations in the alveolar epithelium after injury leading to pulmonary fibrosis. Histol. Histopathol. 11::463-483.
Howat, W.J., Holgate, S.T., and Lackie, P.M. 2002. . TGF-beta isoform release and activation during in vitro bronchial epithelial wound repair. Am. J. Physiol. Lung Cell Mol. Physiol. 282::L115-L123.
Morishima, Y. et al. 2001. . Triggering the induction of myofibroblast and fibrogenesis by airway epithelial shedding. Am. J. Respir. Cell Mol. Biol. 24::1-11.
Zhang, S., Smartt, H., Holgate, S.T., and Roche, W.R. 1999. . Growth factors secreted by bronchial epithelial cells control myofibroblast proliferation: an in vitro co-culture model of airway remodeling in asthma. Lab. Invest. 79::395-405.
Phillips, R.J. et al. 2004. . Circulating fibrocytes traffic to the lungs in response to CXCL12 and mediate fibrosis. J. Clin. Invest. 114::438-446. doi:10.1172/JCI200420997.
Palmer, S.M. et al. 2000. . Gastroesophageal reflux as a reversible cause of allograft dysfunction after lung transplantation. Chest. 118::1214-1217.
Chakinala, M.M., and Walter, M.J. 2004. . Community acquired respiratory viral infections after lung transplantation: clinical features and long-term consequences. Semin. Thorac. Cardiovasc. Surg. 16::342-349.
Zamora, M.R. 2004. . Cytomegalovirus and lung transplantation. Am. J. Transplant. 4::1219-1226.
Penfold, M.E. et al. 1999. . Cytomegalovirus encodes a potent alpha chemokine. Proc. Natl. Acad. Sci. U. S. A. 96::9839-9844.
Knoop, C., Haverich, A., and Fischer, S. 2004. . Immunosuppressive therapy after human lung transplantation. Eur. Respir. J. 23::159-171.
Selman, M., and Pardo, A. 2002. . Idiopathic pulmonary fibrosis: an epithelial/fibroblastic cross-talk disorder. Respir. Res. [serial online]. 3::3-http://respiratory-research.com/content/3/1/3.(Ivor S. Douglas, and Mark)