Attenuation of Bleomycin-induced Pulmonary Fibrosis by Follistatin

Pulmonary fibrosis is a progressive disorder, characterized by the loss of alveolar structure through the apoptosis of epithelial and endothelial cells, proliferation of fibroblasts, and excessive deposition of extracellular matrix (1). This disease is associated with poor outcome, and mean survival ranges from 3.2 to 5 years after diagnosis (2). Pulmonary fibrosis may result from a wide variety of causes, but the mechanism of this pathologic process is not completely understood. It is thought that various factors induce acute injury in alveolar walls, and that continuing and recurring injuries over many years contribute to the temporal and spatial heterogeneity of tissue fibrosis.

Transforming growth factor (TGF)-ß is a potent profibrogenic cytokine in various tissues, including the liver, kidney, and pancreas (3, 4). For pulmonary fibrosis, TGF-ß is thought to be the most important cytokine that stimulates the production of extracellular matrix, fibroblast proliferation, and induction of myofibroblast differentiation (5-12). In an animal model, blocking the action of TGF-ß attenuates fibrosis (9-12).

Activins, members of the TGF-ß superfamily, are pluripotent cytokines exerting a wide range of bioactivities in various cell types. Through these biological effects, activin is involved in tissue repair and fibrotic processes. In the lung, activin is thought to play a significant role, not only in developmental processes (13), but also in pulmonary fibrosis. In samples obtained from patients with pulmonary fibrosis, activin was detected in macrophages, smooth muscle cells, and hyperplastic epithelial cells (14). In bleomycin (BLM)-induced lung injury of the animal model, alveolar macrophages infiltrated into fibrotic area expressed immunoreactive activin (15). Fibroblasts are thought to be one of the target cells of activin. They proliferate and differentiate into myofibroblasts in response to activin (16). Based on these lines of evidence, activin is thought to play a significant role in pathophysiology of pulmonary fibrosis. However, this notion has not been proven experimentally.

The actions of activins are modified by various factors. The most important factor that modulates the action of activin is an activin-binding protein, follistatin. Follistatin binds to activins with high affinity and blocks their actions (17, 18). Follistatin is expressed on the surface of the target cells of activin by binding to the extracellular matrix (19). Activin trapped by follistatin is internalized by endocytosis and subsequently degraded by proteolysis (20). The activin-follistatin system is a complex regulatory system that controls diverse cellular functions.

In the present study, we examined the involvement of the activin-follistatin system in BLM-induced lung injury and fibrosis. We report here that activin A was upregulated at an early phase of the lung injury. Follistatin treatment was effective in reducing infiltration of inflammatory cells at an early phase and resulted in reduction of fibrotic changes at a late phase.

METHODS

Animal Treatment

Animal studies were conducted according to the guidelines of the Committee of Experimental Animal Research of Gunma University. Wistar rats were instilled with BLM intratracheally (see online supplement). Control animals received saline.

On Days 3, 7, 14, 21, and 28, rats were killed and right lobes were removed to be frozen for RNA analysis. Left lobes were removed for histologic analysis. Control animals were killed on Day 3.

For analysis of the efficacy of follistatin, recombinant human (rh)-follistatin was administered intraperitoneally. Protocol for follislatin administration is presented in the online supplement.

RNA Extraction and mRNA Analysis

Total RNA was extracted using TRIzol (Invitrogen, Carlsbad, CA). First stranded cDNA was made from 2

Immunohistochemical Analysis

Immunohistochemistry was performed as described (30). Details of all antibody sources used in this study are provided in the online supplement.

Bronchoalveolar Lavage

For analysis of the effect of follistatin treatment to lung injury, bronchoalveolar lavage (BAL) was performed. On Day 7, BAL was performed three times by inserting 8-ml aliquots of saline into the lung and retrieving the fluid. Total and differential cell counts were performed. Protein contents and lactate dehydrogenase levels in BAL fluid (BALF) were measured. Interleukin (IL)-1ß and monocyte chemotactic protein-1/CC chemokine ligand 2 (MCP-1/CCL2) levels were measured using ELISA (R&D Systems, Minneapolis, MN, and Biosource, Camarillo, CA).

Hydroxyproline Assay

To assess the effect of follistatin treatment on pulmonary fibrosis, Day 28-frozen lung samples were used. Hydroxyproline assay was performed as described previously (21). The values were expressed as hydroxyproline content per protein content of the samples.

Cell Culture

MRC5 human fetal lung fibroblasts were purchased from American Type Culture Collection (Manassas, VA) and cultured in Dulbecco's modified Eagle's medium containing 10% fetal calf serum, penicillin, and streptomycin in an atmosphere of 5% CO2 at 37°C. To obtain quiescent cells, cells were cultured in serum-free medium for 24 hours.

Western Blot Analysis

Cells were washed three times with phosphate-buffered saline (PBS). After homogenization, 20 µg of protein from each sample were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to a polyvinylidine difluoride membrane. Western blot was performed as described (31).

Immunocytochemical Analysis

Cells cultured on coverslips were washed with PBS, fixed with 3% paraformaldehyde, permeabilized, and then blocked. Immunocytochemical analysis was performed as described previously (32).

Measurement of Activin A and Collagen

After serum starvation for 24 hours, TGF-ß (Wako, Osaka, Japan) was added to the medium and cells were incubated at 37°C for 2 days. The concentration of activin A and collagen in the the culture medium was then was measured using ELISA.

Statistical Analysis

The differences between averages was compared by t test, in which p < 0.05 was considered significant.

RESULTS

Expression of Activin, Follistatin, and Receptors for Activin after BLM-induced Lung Injury

Intratracheal administration of BLM resulted in acute and severe lung injury. In late-phase, BLM induced marked pulmonary fibrosis. We first examined the involvement of the activin-follistatin system in BLM-induced lung injury and fibrosis. We analyzed the changes in the expression of mRNA for the ß^sub A^ subunit of activin, follistatin, and activin receptor by reverse transcription-PCR. As shown in Figure 1A, ß^sub A^ mRNA was virtually absent in saline-treated control lung. In contrast, mRNA for the ß^sub A^ subunit was upregulated 3 days after BLM instillation and remained elevated until Day 7. On Day 14, the level of ß^sub A^ mRNA was still elevated, but decreased gradually thereafter. Real-time reverse transcription-PCR revealed that ß^sub A^ mRNA was increased significantly in the acute inflammatory and fibro-proliferative phases. Follistatin mRNA was expressed abundantly in normal lung and increased slightly after BLM-induced lung injury and fibrosis (Figures 1A and 1B). Next, localization of activin A, follistatin, and the receptor for activin was studied by immunohistochemical analyses. In control animals, a weak signal of immunoreactive activin A was detected in alveolar macrophages, bronchial epithelial cells, alveolar epithelial cells (Figure 2A), and vascular smooth muscle cells (data not shown). After BLM instillation, numerous inflammatory cells infiltrated the alveolar space and interstitium. High levels of immunoreactive activin A was detected in alveolar macrophages infiltrating these areas (Figure 2B). On Day 28, there were accumulated fibroblasts with weak immunoreactivity of activin A. More intense signals were found in macrophages, which were positive for CD68 (Figure 2C and 2I). Immunoreactivity of activin A in macrophages did not change, but the number of positive cells decreased. In control animals, immunoreactive follistatin was found in macrophages, bronchial epithelial cells, alveolar epithelial cells (Figure 2D), and vascular smooth muscle cells (data not shown). In BLM-induced lung injury and fibrosis, high levels of immunoreactive follistatin were detected in infiltrated macrophages (Figures 2E and 2F) and accumulated fibroblasts (Figure 2F). In control animals, the type IIB activin receptor was detected in alveolar macrophages, bronchial epithelial cells, alveolar epithelial cells (Figure 3A), and vascular smooth muscle cells (data not shown). The intensity of these immunoreactivities did not change after BLM instillation (Figure 3B). In addition, immunoreactivity was found in accumulated fibroblasts on Day 28 (Figure 3C).

Effect of rh-Follistatin on Acute Lung Injury

To assess the role of endogenous activin A in lung injury, we administered rh-follistatin (12 µg/kg) or saline to rats with BLM-induced lung injury. BAL was taken from BLM-treated rats on Day 7, when lung injury was in the acute phase. As shown in Figure 4A, large increases in inflammatory cells, including macrophages and neutrophils, were observed in BLM-injected saline-treated rats. The numbers of macrophages and neutrophils were 11.5 ± 2.6 × 10^sup 4^ cells/ml and 0.09 ± 0.02 × 10^sup 4^ cells/ml (mean ± SEM) in vehicle-injected saline-treated rats, respectively, and 124.1 ± 49.8 × 10^sup 4^ cells/ml and 111.4 ± 66.9 × 10^sup 4^ cells/ml in BLM-injected saline-treated rats, respectively. Lung permeability, as assessed by protein content, and cytotoxicity, as assessed by lactate dehydrogenase, were also increased in this group (Figure 4B). In the rh-follistatin-treated group, the number of inflammatory cells was markedly lower than in the saline-treated group (Figure 4A). The effect of follistatin was significant in terms of the number of macrophages (40.9 ± 13.8 × 10^sup 4^ cells/ml) and neutrophils (37.3 ± 11.1 × 10^sup 4^ cells/ml). With regard to biochemical parameters of the BALF, lung permeability in the rh-follistatin-treated group was lower than that in the control group (Figure 4B). The cytotoxic index was also lower, but not significantly so (Figure 4B, p = 0.07). The levels of IL-1bgr; and MCP-1/CCL2 in BALF were measured by ELISA. IL-1ß in BALF was significantly increased after injection of BLM (Figure 4C). IL-1ß was significantly reduced by treatment with rh-follistatin. Likewise, rh-follistatin markedly reduced the level of MCP-1 in BLM-treated rats (Figure 4D). Light microscopic examination revealed a decreased number of infiltrated inflammatory cells and reduced destruction of lung architecture in the rh-follistatin-treated group (Figure 4E).

Effect of rh-Follistatin on Lung Fibrosis

We next examined whether exogenous follistatin affected the lung fibrosis observed in the later phase. BLM-treated rats were killed on Day 28 and histologic changes and hydroxyproline content of the tissue were determined. In control rats, severe fibrosis occurred unanimously on light microscopic examination (Figures 5A-a and 5A-c: BLM [+]/saline^sub 1-13^ group and BLM [+]/saline^sub 8-27^ group, respectively). In contrast, lung fibrosis was markedly improved in rh-follistatin-treated groups over both the acute phase (BLM [+]/Foll^sub 1-13^ group, Figure 5A-b) and the postacute phase (BLM [+]/Foll^sub 8-27^ group, Figure 5A-d). We then measured hydroxyproline content of the tissue to estimate collagen deposit in the lung. Hydroxyproline content was significantly increased in BLM-treated rats. Again, follistatin significantly reduced the hydroxyproline content (Figure 5B). There was no significant difference in the reduction of hydroxyproline content between the acute-phase treatment group and the postacute-phase treatment group.

Effect of TGF-ß on the Expression of Activin A in Cultured Lung Fibroblasts

TGF-ß is thought to be the most important cytokine involved in pulmonary fibrosis, and a variety of effects of this cytokine are crucial in pathogenesis (5-12). To clarify whether the actions of TGF-ß on fibroblasts are mediated at least partly by activin A, we examined the effect of TGF-ß on the expression of activin A in cultured lung fibroblasts. TGF-ß upregulated the expression of ß^sub A^ mRNA in MRC5 lung fibroblasts (Figures 6A and 6B) and also augmented the release of activin A into culture medium (Figure 6C).

Effect of Follistatin on Differentiation and Collagen Production of Lung Fibroblasts Induced by TGF-ß

One of the important actions of TGF-ß in lung fibroblasts is to convert them into activated myofibroblasts. We analyzed the effect of TGF-ß on differentiation of lung fibroblasts and examined whether follistatin was able to block the effect of TGF-ß. We examined the changes in the expression of a-smooth muscle actin, which is a marker of myofibroblasts. As shown in Figures 7A-7C, TGF-ß enhanced the expression of a-smooth muscle actin at both the mRNA and protein levels in lung fibroblasts. Immunocytochemical analysis revealed induction of a-smooth muscle actin-positive filaments in fibroblasts treated with TGF-ß (Figure 7D). Indeed, follistatin inhibited the effect of TGF-ß on the expression of a-smooth muscle actin, presumably via blockade of the activin action (Figures 7A-7D). We also examined the effect of follistatin on TGF-ß-induced collagen production. As shown in Figure 7E, follistatin markedly inhibited collagen production induced by TGF-ß.

DISCUSSION

Although there have been various challenges to recover normal architecture and function of the lung in fibrotic tissues, no effective therapy has been found until now. For pulmonary fibrosis, many animal studies have been conducted, with few desirable results being reported. A key pathway that directs injured tissue to fibrosis from normal repair is a pathway involving highly profibrotic cytokine TGF-ß. TGF-ß is thought to be one of the most important stimulators of fibroblasts, leading to pulmonary fibrosis. It promotes differentiation of fibroblasts into myofibroblasts, an active form of fibroblasts, and thereby markedly augments extracellular matrix production. Evidence from both animal and human studies has shown that TGF-ß plays a significant role in fibrogenesis of the lung tissues (7, 22-28). There have been only a few animal studies, the intent of which was to block the action of TGF-ß to treat pulmonary fibrosis (9-12); no such investigations have been conducted on humans.

Activin A has been shown to be involved in several types of tissue injury and fibrosis. In the liver and kidney, activin A plays an inhibitory role in tissue regeneration after an acute injury. Follistatin treatment promotes regeneration by blocking the activin action (29, 30). Other in vitro studies have shown that activin A has profibrotic potency in hepatic and renal fibrosis (31, 32). Indeed, both activin A and TGF-ß stimulate production of the other via an autocrine loop, and both factors contribute to tissue fibrosis (32, 33). In the lung, it has been proposed that activin A plays an important role in pulmonary fibrosis based on results of in vitro studies (14-16). It remains to be shown how the activin-follistatin system is involved in lung fibrosis in vivo. Furthermore, it is not certain whether or not follistatin treatment prevents lung fibrosis. The present results show that activin A was upregulated, during early-phase lung injury, at both the mRNA and protein levels, and then decreased gradually to control levels (Figure 1). Histologically, immunoreactivity of activin A was still observed in macrophages and fibroblasts in the fibrotic lung (Figure 2C). Presumably, follistatin is expressed in a normal condition and strongly inhibits the activin action. After BLM instillation, the effect of activin A may dominate the antagonistic effects of follistatin. Regarding the activin action, however, it remains to be established whether activin A exerts proinflammatory or antiinflammatory effects. Recent studies suggest that activin A plays a novel role in inflammation and repair processes (34, 35). Activin A was strongly induced by proinflammatory cytokines, such as IL-1ß and tumor necrosis factor-a (36), and the release of activin into circulation induced production of the proinflammatory cytokines after lipopolysaccharide treatment (37). These results suggest that activin A plays a proinflammatory role in acute-phase inflammation. On the other hand, activin A was shown to exert an antiinflammatory role through its ability to antagonize the actions of IL-1 and IL-6 (38-40). We speculate that the action of activin in the inflammatory process depends on the type of damaged cells or tissues and the type of injury. Our results clearly demonstrate that, in the lung, rh-follistatin treatment reduced both the number of infiltrating cells in BALF (Figure 4A) and the lung permeability after BLM instillation (Figure 4B). The levels of IL-1ß and MCP-1 in BALF were also reduced (Figures 4C and 4D). IL-1ß is one of the major proinflammatory cytokines in acute lung injury. In addition, it has a potential role in pulmonary fibrosis with subsequent induction of TGF-ß (41). Moreover, reduction of MCP-1/CCL2 inhibited the accumulation of macrophages and other types of cells expressing CCR2 in BALF. These inhibitions resulted in the histologic improvement of damaged lung tissues (Figure 4E). Hence, the proinflammatory effect of activin A is predominant in acute lung injury. In addition, a recent study revealed that TGF-ß, which has been thought to be important in the late phase of tissue repair and pulmonary fibrosis, is active early in acute lung injury (42). In that study, local activation of TGF-ß was found to be critical for the development of pulmonary edema, and blocking its activation was observed to be effective in attenuating alveolar flooding. Because follistatin attenuates the effect of TGF-ß (31-33), we speculate that the efficacy of rh-follistatin results from blocking not only the activin action but also the TGF-ß action during early-phase injury.

In addition to the effects on acute lung injury, rh-follistatin was effective in preventing pulmonary fibrosis. In the late phase of injury, BLM increased the number of fibroblasts and promoted differentiation of fibroblasts into myofibroblasts. Myofibroblasts, an activated form of fibroblasts, produce excessive extracellular matrix, and thereby promote progression of tissue fibrosis. Matsuse and colleagues reported that activin A is a potent profibrotic activator in this process (14-16). In a series of in vitro studies, they showed that activin A stimulates proliferation of human lung fibroblasts (HFL-1), induces differentiation into myofibroblasts (16), and facilitates HFL-1-mediated collagen gel contraction (43). These effects of activin A are abolished by follistatin (16, 43). The present study extended their studies and demonstrates for the first time that rh-follistatin prevents tissue fibrosis in an in vivo animal model (Figures 5A and 5B). Indeed, the effect of rh-follistatin was comparable to the agents that block TGF-ß action (9-12).

The profibrotic action of TGF-ß is often mediated by some downstream pathways. For example, connective tissue growth factor, is induced by TGF-ß and both synergize to promote fibroblasts proliferation, migration and extracellular matrix deposit (44). The present results clearly demonstrate that activin A acts as one of the downstream factors of TGF-ß activity. Indeed, follistatin inhibited TGF-ß-induced fibroblast differentiation (Figures 7A-7D) and collagen production (Figure 7E). Based on the these results from in vitro experiments, we speculate that follistatin has the potential to inhibit the effects of TGF-ß on lung fibroblasts by blocking the activin activity and contributing to the antifibrotic role in animal models. It is also likely that other effects of TGF-ß in lung fibroblasts (e.g., cell proliferation and migration) are also inhibited by follistatin. Because follistatin inhibits acute lung injury, it is likely that the effect of follistatin on lung fibrosis is due partly to the attenuation of acute inflammation. It is also evident that follistatin directly blocks the activation of fibroblasts and reduces the production of collagen. In addition, in an in vivo model, follistatin reduced collagen deposition in the post-acute-phase treatment group. Therefore, follistatin exerts beneficial actions by acting during acute and late phases.

Follistatin binds to activin with high affinity, which is similar to or exceeds that estimated for activin binding to its receptor (45). For a decade, it has been thought that follistatin binds only to activin. However, a recent study by Glister and colleagues (46) showed that follistatin also binds to other members of the TGF-ß family, such as BMP-4, -6, and -7. However, binding affinities of follistatin were much lower for these TGF-ß family members than that for activin. Glister and colleagues measured the binding affinity of follistatin for BMPs in contrast to that for activin (100%), and showed the relative affinities of follistatin for BMP-4, -6 and -7 were 10, 5, and 1%, respectively. In addition, their study showed that the binding of follistatin and BMP-6 or BMP-7 was not irreversible and had no impact on the downstream effector, Smad1. BMP-4 has been thought to play an important role in branching morphogenesis of the developing lung. A recent study showed that BMP-4 is expressed in adult lung fibroblasts (47). Although the role of BMP-4 in lung injury or fibrosis is not clear, the possibility remains that follistatin blocks the action of BMP-4.

Recently, it was shown that cells from bone marrow or circulation contribute to repair of lung injury or formation of lung fibrosis (48-51). Fibrocytes, a minor component of the circulating pool of leukocytes, which express collagen I and CD45, also expresses chemokine receptors, such as CCR1, CCR2, CCR5, CCR7, and CXCR4 (50, 51). They are attracted by the chemokine signals, traffic to the lung, and then induce fibrosis. Blockade of chemokine signals by genetic depletion of CCR2 or anti-CXCL12 antibody interferes with the recruitment of fibrocytes into fluorescein isothiocyanate- or BLM-injected fibrotic lung, and results in beneficial effects. Because the time course for the expression of activin and the cellular infiltration into BLM-injured lung correlates with the time course for fibrocyte recruitment into the lung in those studies, it is possible that recruitment of the circulating fibrocytes could be a site of action of activin. If so, follistatin may exert its action at least partly by inhibiting the recruitment of fibrocytes through, for example, attenuation of the CCL2 levels in BALF. This issue should be examined in the near future.

In summary, we have demonstrated that the activin-follistatin system is involved in lung injury and fibrosis induced by BLM. Activin A plays a proinflammatory role and a profibrotic role in these processes. rh-Follistatin treatment attenuated the tissue injury and fibrosis by blocking activin activity. Overall, follistatin has potential as a therapeutic tool for pulmonary fibrosis.

Conflict of Interest Statement: None of the authors have a financial relationship with a commercial entity that has an interest in the subject of this manuscript.

Acknowledgment: The authors thank Ms. Mayumi Odagiri for secretarial assistance.

© 2005 American Thoracic Society Provided by ProQuest LLC. All Rights Reserved.