Background
GDF-15 is a protein belonging to the TGF-beta family, which includes several proteins involved in tissue homeostasis, differentiation, remodeling and repair [
1]. As a pleiotropic cytokine it is involved in the stress response program of different cell types after cellular injury. Under normal conditions, GDF-15 is only weakly expressed in most tissues [
2]. However GDF-15 is strongly upregulated in disease states such as acute injury, tissue hypoxia, inflammation and oxidative stress [
3‐
6].
In the cardiovascular system, GDF-15 is expressed in cardiomyocytes and other cell types including macrophages, endothelial cells, vascular smooth muscle cells, and adipocytes [
1,
7,
8]. In endothelial cells (ECs) it has been shown that GDF-15 inhibits proliferation, migration and invasion in vitro and in vivo [
9‐
11]. A recent study demonstrated that the inhibitory effect of GDF-15 on EC proliferation was only present at higher concentrations (50 ng/ml), whereas at ten times lower concentrations (5 ng/ml), GDF-15 caused endothelial cell proliferation and was proangiogenic [
12]. At present little is known about the expression of GDF-15 in the lung. In situ hybridization studies in rats have revealed expression of GDF-15 in bronchial epithelial cells [
1]. GDF-15 is potently induced in animal models of lung injury. Bleomycin administration in adult mice and prolonged hyperoxic exposure in neonate mice resulted in GDF-15 induction [
5].
Pulmonary arterial hypertension (PAH) is a life-threatening disease characterized by a marked and sustained elevation of pulmonary artery pressure that results in right ventricular (RV) failure and death [
13]. Histologically, remodeling of pulmonary arteries show various degrees of medial hypertrophy and endothelial cell growth, which ultimately lead to the obliteration of precapillary arteries [
14,
15]. The mechanisms resulting in pulmonary vascular remodeling are complex and incompletely understood. Several members of the TGF-β superfamily have been implicated in this process [
16] while the role of GDF-15 in the pathophysiology of PAH is not clear. In a recent study we demonstrated elevated serum levels of GDF-15 in patients with idiopathic pulmonary arterial hypertension (IPAH) [
17]. Furthermore, it has been shown that GDF-15 serum levels are increased in scleroderma patients with pulmonary hypertension and GDF-15 protein was predominantly located in monocytes infiltrating the lung tissue [
18].
In the present study we investigated the expression of GDF-15 in human normal lungs and in lung tissue from patients with PAH. In addition, we conducted in vitro-studies to elucidate the possible role of GDF-15 in the pulmonary vasculature.
Discussion
In the present study we demonstrated that GDF-15 is expressed in human lung tissue, arising predominantly in macrophages and pulmonary endothelial cells. Compared to normal lung, GDF-15 appears upregulated in lung tissue of patients with PAH, especially in areas of active vascular remodeling, i.e. plexiform lesions. Since GDF-15 protein influences proliferation and apoptosis of pulmonary endothelial cells, it might play a role in the evolution and homeostasis of plexiform lesions in PAH patients.
GDF-15 is a stress-responsive cytokine that is upregulated under pathologic conditions involving various stimuli such as tissue hypoxia, inflammation, or enhanced oxidative stress [
3‐
6]. Under physiologic conditions GDF-15 is only weakly expressed in most tissues and organs [
34]. It is therefore unsurprising that we only detected a weak immunostaining signal for GDF-15 in human normal lung tissue with almost no expression in the airways like bronchial and alveolar epithelial cells. As demonstrated in previous studies [
18], GDF-15 was strongly expressed in alveolar macrophages which might indicate a role of this protein in innate immunity [
2]. Interestingly, our immunostaining experiments clearly demonstrated strong expression of GDF-15 in the vascular compartment of PAH patients, particularly in the intima of pulmonary arteries. GDF-15 staining was observed in pulmonary vessels of all sizes, beginning from the microvasculature up to large pulmonary vessels. The endothelial expression pattern was observed in normal lung as well as in lungs from PAH patients, suggesting a physiological role for GDF-15 in pulmonary endothelial cells. To date little is known about the functional role of GDF-15 in endothelial cells. A previous study demonstrated inhibitory effects of GDF-15 on proliferation, migration and invasion of endothelial cells
in vitro as well as anti-angiogenic effects
in vivo using a matrigel-plug-assay [
11]. In contrast to these findings, a recently published paper demonstrated both angiogenic and anti-angiogenic properties of GDF-15 [
12], which were concentration-dependent. GDF-15 elicited pro-angiogenic effects at low concentrations, whereas paradoxical effects were observed at higher concentrations (100 ng/ml). In accordance with this finding we too were able to demonstrate concentration-dependent pro- as well as anti-angiogenic effects of recombinant GDF-15 protein on pulmonary endothelial cells
in vitro. That different concentrations of a cytokine could result in different cellular responses is well-known for members of the TGF-β-family. For instance, TGF-β1 exerts bi-functional effects on endothelial cells, regarding activation, proliferation and migration. At low concentrations TGF-β1 has a stimulating effect, whereas higher concentrations inhibit these processes [
35]. It is challenging to speculate the active amount of GDF-15 in the pulmonary vasculature. However, addi-tional autocrine and paracrine pathways may determine the local concentration of GDF-in the vascular compartment. Furthermore, a variety of activating or disabling regulators may interfere with the intra- and extracellular storage as well as the stability of GDF-15 in lung compartments.
Compared to normal lung tissue, increased GDF-15 expression was observed in PAH lungs, with strongest expression being identified in areas of vascular remodeling, especially in the cells forming the plexiform lesions. In comparison, GDF-15 expression was lower in vascular smooth muscle cells, both in normal vessels and in remodeled arterioles with media hypertrophy. No differences in the expression pattern of GDF-15 were seen between lungs of various underlying aetiologies of pulmonary hypertension such as IPAH, and PAH due to Eisenmenger's physiology. A recent study identified expression of GDF-15 protein in pulmonary macrophages of patients with PAH due to scleroderma, but almost no GDF-15 staining in IPAH lungs [
18]. This staining pattern appears to conflict with our results, but may be related to different protocols of tissue preparation and staining. To confirm the expression pattern seen in our immunohistochemical studies we performed laser-assisted microdissection of vascular subcompartments in PAH lungs. We successfully amplified GDF-15 transcripts in plexiform lesions and cells from morphological normal pulmonary arteries of PAH patients. In accordance to the immunohistochemical staining pattern, increased GDF-15 expression was detected in plexiform lesions compared to unremodeled pulmonary arteries. These findings suggest that GDF-15 could be involved in the pathobiology of plexiform lesions as opposed to the muscular compartment. The cellular and cytokine environment of plexiform lesions, which are characterized by disorganized focal proliferation of endothelial channels [
36,
37], is complex and not fully understood. Since a variety of different cytokines and signaling pathways interact with each other, it is difficult to define the precise role of a single cytokine in such a complex milieu. Key players in vascular remodeling of PAH lungs are members of the TGF-β-superfamily, and TGFβ1 has been reported to potentiate intimal hyperplasia in animal models following arterial injury [
38].
Factors triggering expression of GDF-15 in the pulmonary vasculature remain unclear. Since GDF-15 is a stress responsive cytokine speculation remains that inflammation and oxidative stress trigger expression of GDF-15 in plexiform lesions. Indeed, several studies have demonstrated increased oxidative stress and inflammation within plexiform lesions [
39]. Our findings indicate that hypoxia is a potent stimulator of GDF-15 expression in pulmonary endothelial cells. Furthermore shear stress might lead to induction of GDF-15 expression in the pulmonary vasculature. Given that in severe PAH, plexiform lesions tend to form at bifur-cations [
40] where shear stress is likely to be high, we examined whether shear stress affects GDF-15 expression. We were able to demonstrate that shear stress leads to an upregulation of GDF-15 expression in human microvascular endothelial cells. These findings may be significant, regarding the evolution of an apoptosis-resistant endothelial cell phenotype. Previous reports have shown that shear stress has an anti-apoptotic effect on endothelial cells [
41]. Since shear stress is a potent inducer of GDF-15 in endothelial cells it is possible that the anti-apoptotic effect provoked by shear stress is - at least partly - mediated by GDF-15. In our study we were able to demonstrate that GDF-15 caused an induction of Akt phosphorylation and had a prosurvival effect on endothelial cells. This finding is in accordance with documented anti-apoptotic effects of GDF-15 in cardiomyocytes involving the phosphoinositide 3-OH kinase (PI3K) and Akt-dependent signaling pathways [
32]. The net effect of GDF-15 on cell proliferation, apoptosis and pulmonary vascular remodeling is difficult to evaluate, especially as GDF-15 is not the only player among the mediators orchestrating vascular remodeling. Like other members of the TGF-β-family proteins, GDF-15 executes a wide variety of complex and ambiguous functions, depending on cell type, microenvironment and genetic status of the cell.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
NN and DJ planned the concept and study design. HAG coordinated the study and drafted the manuscript. LM and JR carried out the immunohistochemistry and real time PCR. CS and NN performed the cell culture experiments. CB and LM carried out the laser-assisted microdissection experiments. TK performed the GDF-15 Sandwich IRMA. FL and UL made substantial contributions to the analysis and interpretation of the data. TW and MMH participated in the design of the study. MG critically read and corrected the manuscript. All authors read and approved the final manuscript.