© 2003 American Thoracic Society
When Defenses against Fibroproliferation FailSpotlight on an Axis of ProphylaxisUniversity of Michigan Health System Ann Arbor, Michigan Current concepts regarding the pathogenesis of disorders characterized by fibroproliferation and tissue remodeling emphasize the central importance of dysregulated interactions between epithelial cells and mesenchymal cells. The pulmonary research community has witnessed convergence around this concept by investigators focused on both airway remodeling in asthma (1) and parenchymal fibrosis (2). Much of the attention has been placed on alterations in cellular phenotypes and mediator profiles that promote fibroblast (or smooth muscle cell) proliferation and matrix protein deposition—an example of which is epithelial production of transforming growth factor-ß. Although generation of such "bad" mediators is doubtless integral to fibroproliferative responses, the importance of the failure to appropriately generate "good" mediators that promote repair and prevent fibrosis in the setting of injury should not be underestimated. The article by Marchand-Adam and coworkers (3) in this issue of the Journal (pp. 1156–1161) focuses attention on two such mediators—hepatocyte growth factor and prostaglandin E2—and describes interactions between them that exemplify the exquisite complexity of the lung's homeostatic capability. Although prostaglandin E2 is often regarded as a pro-inflammatory mediator, it actually has broad suppressive effects on inflammatory and immune cells. While it has been reported to have some salutary effects on respiratory epithelium (4), its actions on the epithelial-mesenchymal unit appear to be more prominently focused upon fibroblasts, where it inhibits cell migration (5), proliferation (6), collagen accumulation (7), and differentiation to myofibroblasts (8). Most of these anti-fibrotic actions of prostaglandin E2 depend on its binding to E prostanoid receptors that signal via increased intracellular cyclic AMP. This prostanoid is synthesized via the cyclooxygenase pathway of arachidonic acid metabolism, and is abundantly produced by both fibroblasts and respiratory epithelial cells. A growing body of evidence suggests that the ability to upregulate cyclooxygenase enzyme expression and prostaglandin E2 synthesis applies a brake on the pro-fibrotic responses of fibroblasts to growth and differentiative factors, such as transforming growth factor-ß. The inability of fibroblasts from lungs of patients with fibrotic lung diseases to carry out this normal homeostatic upregulation of prostaglandin E2 generation, demonstrated by several groups (9, 10), is one factor that promotes the altered fibroblast phenotype characteristic of fibroproliferative disorders. Alveolar epithelial cells constitutively elaborate substances that inhibit fibroblast proliferation, and both pharmacologic and genetic studies implicate prostanoids, including prostaglandin E2, as playing an important role in this suppressive capacity (11, 12). It is tempting to speculate that impaired prostaglandin E2 synthetic capacity may help to explain how epithelial cell injury or loss promotes fibrotic responses. Hepatocyte growth factor acts as a "tonic" for epithelial cells throughout the body via its interactions with its cell surface receptor, known as c-met (13). In the lung, it stimulates alveolar epithelial cell migration and proliferation and protects them from apoptosis, and its administration in vivo limits pulmonary fibrosis in response to bleomycin. Mesenchymal cells represent an important source of this mediator. Since apoptotic epithelial cells are found in areas of pulmonary fibrosis, Marchand-Adam and coworkers (3) asked whether the capacity of fibroblasts isolated from fibrotic lungs to generate hepatocyte growth factor is decreased. As compared with cells from non-fibrotic lung, fibroblasts from patients with idiopathic pulmonary fibrosis exhibited a 50% reduction in basal secretion of this epithelial tonic. Because transcription of the hepatocyte growth factor gene is known to be increased by factors that increase intracellular cyclic AMP, such as prostaglandin E2 (14), and because fibrotic fibroblasts under-produce prostaglandin E2 (9, 10), the investigators next explored this link. They confirmed that fibrotic fibroblasts manifest a marked reduction in prostaglandin E2 synthesis. They next established that deficiency of this prostanoid was responsible for the defect in hepatocyte growth factor production by showing that the latter was overcome by the addition of exogenous prostaglandin E2, and also that indomethacin and dexamethasone, both of which inhibit prostaglandin production, reduced hepatocyte growth factor generation in control fibroblasts. These findings identify an important new abnormality contributing to fibrogenesis, and an appealing molecular partner for prostaglandin E2. The prostaglandin E2-hepatocyte growth factor "axis of prophylaxis," as illuminated by Marchand-Adam and coworkers (3) and others, can be envisioned as one central fulcrum, perhaps among many, that dictates whether lung injury culminates in repair or fibrosis. Injury to epithelial cells limits their capacity to elaborate prostaglandin E2. A fibroblast deprived of both its paracrine (epithelial cell-derived) as well as its autocrine (fibroblast-derived) sources of prostaglandin E2 is more easily bullied by growth factors toward a fibroproliferative phenotype, but also exhibits an impaired capacity to generate hepatocyte growth factor. This, in turn, further compromises epithelial cell integrity, and the vicious cycle is complete.
This prostaglandin E2-hepatocyte growth factor axis may also be a target for the actions of other mediators and pharmacologic agents. For example, the authors (3) demonstrate that transforming growth factor-ß reduced hepatocyte growth factor expression in control fibroblasts, providing yet another potential mechanism for the pro-fibrotic effects of this molecule. Interestingly, interferon- The gastric toxicity of conventional nonsteroidal anti-inflammatory drugs involves their inhibition of gastroprotective prostaglandin E2 and hepatocyte growth factor synthesis (16). In parallel fashion, is it possible that the coincidental administration of nonsteroidal drugs (or corticosteroids) during the lung's response to injury (perhaps a viral syndrome?) might provide a milieu conducive to fibroproliferation? It sounds far-fetched. Then again, so did the notion that hormone replacement therapy might promote cardiovascular disease in women (17). FOOTNOTES Conflict of Interest Statement: M.P.-G. has no declared conflict of interest. REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
, which has anti-fibrotic effects, has also been shown to modulate hepatocyte growth factor expression in fibroblasts. By itself, this cytokine exerted no effect; however, it synergistically augmented hepatocyte growth factor production in response to cyclic AMP elevators including prostaglandin E2 (
