Emerging role of BMPs/BMPR2 signaling pathway in treatment for pulmonary fibrosis

Abstract

Pulmonary fibrosis is a fatal and chronic lung disease that is characterized by accumulation of thickened scar in the lungs and impairment of gas exchange. The cases with unknown etiology are referred as idiopathic pulmonary fibrosis (IPF). There are currently no effective therapeutics to cure the disease; thus, the investigation of the pathogenesis of IPF is of great importance. Recent studies on bone morphogenic proteins (BMPs) and their receptors have indicated that reduction of BMP signaling in lungs may play a significant role in the development of lung fibrosis. BMPs are members of TGF-β superfamily, and they have been shown to play an anti-fibrotic role in combating TGF-β-mediated pathways. The impact of BMP receptors, in particular BMPR2, on pulmonary fibrosis is growing attraction to researchers. Previous studies on BMPR2 have often focused on pulmonary arterial hypertension (PAH). Given the strong clinical association between PAH and lung fibrosis, understanding BMPs/BMPR2-mediated signaling pathway is important for development of therapeutic strategies to treat IPF. In this review, we comprehensively review recent studies regarding the biological functions of BMPs and their receptors in lungs, especially focusing on their roles in the pathogenesis of pulmonary fibrosis and fibrosis resolution.

1. Introduction

Pulmonary fibrosis (PF) is a chronic, progressive, and irreversible lung disease. Fibrosis in lung tissue is initiated by abnormal lung repair processes followed by recurrent lung inflammatory injuries [1,2]. PF results in thickening and hardening of the lung tissues, interfering with gas exchange. Some cases of PF show no clear origins or causes and are referred as idiopathic pulmonary fibrosis (IPF). PF is characterized by aberrant myofibroblast differentiation and extracellular matrix (ECM) accumulation [3]. The estimated median survival time of IPF is 3–5 years, with a high incidence and mortality in the elderly population. Currently, there are two FDA-approved drugs, pirfenidone and nintedanib, for IPF treatment; however, these two drugs only reduce the progression of PF in patients and can not reverse fibrosis or cure this disease. Lung transplantation is the ultimate therapy for IPF patients [3,4]. Due to the lack of therapeutic agents for IPF, understanding its pathogenesis is extremely important.

Bone morphogenetic proteins (BMPs) belong to the transforming growth factor beta (TGF-β) family [5]. In the BMPs’ canonical pathway, BMPs initiate a signal transduction cascade by binding to cell surface receptors (BMP receptor type I and type II) to induce cellular responses [6–8]. The anti-fibrotic effects of BMPs have been reported in different organs including lungs [9–12]. BMPR2 is a commonly expressed transmembrane serine/threonine receptor of the TGF-β superfamily [13,14]. Activation of BMPR2 by BMPs is thought to play a suppressed role against TGF-β signaling [15].

BMPR2 has been demonstrated to be strongly associated with pulmonary arterial hypertension (PAH), a progressive disease characterized by elevated pressure in the pulmonary artery and right ventricle hypertrophy. Approximately 70 % of inherited PAH patients and 25 % of idiopathic PAH (iPAH) patients have been found to have BMPR2 gene mutations [16,17]. Recent studies indicate that BMPR2 is not only associated with PAH but also possibly involved in the pathogenesis of PF [18]. Therefore, this review will focus on the recent findings on the impact of BMPs and BMPR2 in PF.

2. The pathogenesis of PF

PF is a chronic and progressive fibrotic lung disease that occurs when lung tissues become damaged and scarred [2–4]. Patients of PF generally exhibit shortness of breath, dry cough, and fatigue. Honeycombing changes in lungs can be observed by computed tomography (CT) scans [4,19]. The scarring associated with PF is initiated by repetitive alveolar injuries. Under ordinary circumstances, the lungs undergo a series of self-repair processes after injury. Upon tissue injury, cytokines and chemokines are released to recruit immune cells, such as neutrophils and macrophages, to combat infections and clean injured cells and debris. Immune cells also produce profibrotic factors, including TGF-β1, platelet derived growth factor (PDGF), fibroblast growth factor (FGF), and other factors to activate fibroblast cells and induce their differentiation into myofibroblasts to promote wound healing. Myofibroblasts produce extracellular matrix (ECM) to maintain the integrity of the injured tissue [20,21] (Fig. 1). In successful tissue repair, the ECM is degraded, and the myofibroblasts undergo apoptosis or dedifferentiation to inactive cell types [22–24]. However, abnormal myofibroblast differentiation and ECM accumulation in interstitial lung tissues cause progressive thickened and stiff tissues, ultimately interfering with gas exchange.

Fig. 1. Pathogenesis of pulmonary fibrosis.

Alveolar inflammation and injury induce fibroblast activation and differentiation to myofibroblast and ECM accumulation. Activation of TGF-β pathway and reduced BMP pathway contribute to the pathogenesis of PF.

3. Resolution strategies of PF

Three steps to promote fibrosis resolution in the treatment of PF haven been considered [25]. Briefly, the first step is to determine the initial cause of injury, followed by application of the appropriate treatment to prevent further damage [26,27]. However, the exact causes of lung injury are often not identified in the onset of IPF. The next important step is the removal of the large amount of ECM deposition in the scar area of lungs during the progression of PF [25]. Finally, the most crucial step is the elimination or inactivation of fibrotic myofibroblasts [21,22]. Myofibroblast differentiation is generally considered a critical step in the development of fibrosis [28]. The markers of myofibroblast differentiation are increased α-smooth muscle actin (α-SMA) levels and subsequent ECM protein production, such as collagen and fibronectin (FN). If the transition and proliferation of myofibroblasts are not interrupted in the course of fibrosis resolution, the ECM accumulation will continue to occur and the PF cannot be reversed. Activated myofibroblasts are generally derived from at least three different sources: 1) resident fibroblasts, 2) epithelial and endothelial cells that undergo epithelial-mesenchymal transition (EMT) and endothelial-to-mesenchymal (EndMT) transition, and 3) circulating progenitors [29–32]. The roles of EMT and EndMT in PF are debatable due to lack supporting evidence from clinical histological analysis. BMPs and BMPR2 are involved in fibrosis resolution, and the next section will focus on their impact and possible therapeutic roles in this context.

4. BMPs and BMPR2 canonical pathways

BMPs are the members of TGF-β family [13,14]. There are more than 60 TGF-β superfamily members that have been identified and involved in a wide range of cellular processes [33]. TGF-β superfamily members are classified into TGF-βs, activins, inhibins, BMPs, growth and differentiation factors (GDFs) [34]. In canonical signaling pathways, TGF-β(1/2) or BMP proteins induce phosphorylation of intracellular downstream targets (Smads) to transduce TGF-β/BMP signals. In the TGF-β-mediated pathway, Smad2/3 are phosphorylated by TGF-β receptors (I/II) and combine with Smad4 to regulate expression of target genes (α-SMA and ECM proteins, such as fibronectin and collagens) [35]. BMPs were originally identified as factors that regulate bone formation and development, although it has been reported that not all BMP members demonstrate osteogenic properties [36,37]. Based on their structures and functions, BMPs can be divided into four groups: 1) BMP2/4; 2) BMP5, 6, 7, 8a, and 8b; 3) BMP9/10; 4) BMP12, 13 and 14 [38–40]. BMP1 is a unique protein that does not belong to the TGF-β superfamily [41,42]. The C-terminus of mature BMPs contains an active structural motif consisting of seven conserved cysteine amino acids [43]. BMP proteins are capable of binding to cell surface receptors to induce phosphorylation of the downstream substrates Smad1/5/8. The phosphorylated Smad1/5/8 (P-Smad1/5/8) associate with Smad4, forming a complex that then translocates into the nucleus to modulate downstream anti-fibrotic gene expression, such as ID1 and ID2 [35,44,45]. TGF-β1 exhibits pro-fibrotic roles by inducing myofibroblast differentiation and ECM production through activation of Smad2/3 [35,46]. An increasing number of studies have indicated that BMP proteins likely exhibit anti-fibrotic functions in the pathogenesis of fibrosis in lung and other organs through activation of Smad1/5/8 [9–12,47,48] (Fig. 1). In addition to inducing pro- or anti-fibrotic gene expression through Smad2/3 or Smad1/5/8 by TGF-β or BMPs, the BMP pathway, through separate mechanisms also competes with TGF-β signaling to protect against fibrogenesis [49–51]. Though the Smad1/5/8-mediated canonical pathway has been identified to play a critical role in the anti-fibrotic effect of BMPs, the non-canonical pathway of BMPs have not been well studied.

There are two types of BMP receptors, type I and type II receptors that form hetero-tetrameric complex on the cell surface. ALK1 (ACVRL1), BMPR1A (ALK3), BMPR1B (ALK6), and ALK2 (ACVR1) are BMP type I receptors. Type II receptors have three subtypes including BMPR2, ACVR2A (ActRIIA), and ACVR2B (ActRIIB) [8,49–51] (Fig. 2). All of subtypes of BMP receptors are known to bind BMP ligands. BMPRs contain a short extracellular domain, a membrane-spanning domain, and an intracellular serine/threonine kinase domain [52]. The structures of ALK3 and ALK6 in the BMPR type I group are highly similar, while ALK2 is distinct from the others [44]. BMPR2 has constitutively protein kinase activity, and it contains a long C-terminal domain positioned next to the kinase domain [50]. It is generally believed that BMPs can bind BMPR1 and BMPR2 complex on the cell surface to initiate intracellular signaling [53]. The affinities of various BMP ligands for different BMP receptors are diverse. BMP2/4 bind to ALK3 and ALK6; the signaling of BMP5, 6, 7 and 8 is through ALK2, 3, and 6; BMP9/10 principally recognizes ALK1 [6,54,55] (Fig. 3). Type II receptors also exhibit various binding affinities to BMPs. For example, BMP2, 4, 9, and 10 bind to BMPR2; BMP6, 9, and 10 bind to ACVR2A; and BMP9 and BMP10 bind to ACVR2B (Fig. 3). BMPR2 can only be recognized by BMPs, while ACVR2A/2B can be activated by other ligands within the TGF-β family, such as activins [55,56]. Some BMPs not only signal through type I and type II BMP receptors, but also can induce downstream signaling through TGF-βRs.

Fig. 2. TGF-β and BMP receptors and their signaling pathways.

BMP type I receptors have 4 subtypes including ALK1 (ACVRL1), ALK2 (ACVR1), ALK3 (BMPR1A) and ALK6 (BMPR1B); type II receptors consist of ACVR2A (ActRIIA), ACVR2B (ActRIIB) and BMPR2B (BMPR2). TGF-β signaling pathway plays a pro-fibrotic role to promote the α-smooth muscle actin (SMA), fibronectin (FN), and collagen expression; while BMP signaling pathway indicates the anti-fibrotic function.

Fig. 3. Binding affinities of various BMPs for BMP receptors.

BMP ligands have different binding affinities for each type I and type II of BMP receptors.

5. Expression of BMPs and BMPR2 in PF

5.1. Altered expression of BMPs in IPF lungs

Single cell RNA-seq data has revealed genetic alternations of BMP ligands and receptors in different lung cell types from both normal and IPF patients (Fig. 4). There is a dramatic increase in gene expression of BMP1 in fibroblasts and myofibroblasts in IPF lungs compared to control subjects, but BMP1 has been shown to not be required for development of lung fibrosis in a murine model of PF [57]. Levels of BMP2, 3, 5, 6 and 7 are significantly altered in alveolar epithelial cells in IPF lungs. The expression of BMP2 and 3 exhibit 2-fold reduction in alveolar type 2 (AT2) cells in IPF patients, whereas gene expression of BMP7 is significantly increased in AT1 cells in IPF patients compared to normal subjects. Furthermore, the expression of BMP4 and BMP5 genes vary in fibroblasts and myofibroblasts in IPF lungs as well. BMP5 is increased 1-fold more in myofibroblasts in IPF lungs. BMP4 gene expression is up-regulated 2-fold in fibroblasts in the IPF lungs (Fig. 4); however, reduction of protein level of BMP4 in IPF lungs has been reported [47]. Guan et al. demonstrated that the protein level of BMP4 was down-regulated in lung tissues and fibroblasts from IPF patients and bleomycin-induced mice lung fibrosis. Thus, this raises a caution to investigators in the field that scRNA-seq data can be used to compare gene levels of BMPs, but the protein levels in IPF lungs should be confirmed by immunoblotting or immunohistology staining. The molecular regulation of BMPs has not been studied. Further investigation may focus on identifying transcriptional or suppression factors for BMPs, and examining DNA methylation of BMPs’ promoter and mRNA stability in the setting of PF.

Fig. 4. Changes of gene expression of BMP ligands and receptors in IPF lungs normalized with control health subjects.

Data is re analyzed from http://ipfcellatlas.com/.

5.2. Altered expression of BMPR2 in IPF lungs

It has been revealed that BMPR2 level is significantly downregulated in fibrotic area in IPF lungs compared to normal subjects [58–60]. This was confirmed in bleomycin-induced experimental lung fibrosis [58,59]. Consistently, the levels of p-Smad1/5/8, the downstream pathway of BMPR2, in IPF lungs were also found to be reduced [59]. BMPR2 protein has also been identified to be decreased in TGF-β1-treated rat lung fibroblasts [59,60]. Moreover, decreased BMPR2 signaling has been observed in endothelial cells (EC) and vascular smooth muscle cells (VSMCs) in fibrotic lungs in a rat model of experimental PF and in IPF patients [58,60]. BMPR2 is well known to be strongly associated with the pathogenesis of PAH. The symptoms and duration of PH and PF are tightly correlated. BMPR2 and p-Smad1/5/8 protein levels have been shown to be significantly reduced in lungs in IPF or IPF+PH patients compared to normal control lung tissues [18,58]. Few studies have been published investigating molecular regulation of BMPR2 expression in lungs. The effect of BMPs on BMPR2 expression has not been reported. Elevated IL-6 and STAT3 may correlate with the reduction of BMPR2 through upregulation of certain microRNAs (miRs) in the setting of lung fibrosis [58,61]. MiR-215–5p was shown to suppress BMPR2 expression in mouse lung epithelial (MLE) cells [62]. The molecular mechanisms of regulation of BMPR2 in PF remains unknown. BMPR2 protein stability has been shown to be regulated by ubiquitination in pulmonary artery endothelial and smooth muscle cells. Smurf1 E3 ligase targets BMPR2 for its ubiquitination and degradation [63,64]. The ubiquitination site, Smurf1 binding site on BMPR2, as well as whether Smurf1 regulation of BMPR2 stability in lung fibroblasts have not been revealed. Identification of Smurf1 binding site and a deubiquitinase for BMPR2 stabilization may be beneficial for development of small molecules to regulate BMPR2 protein stability.

6. The effects of BMPs and BMPRs signaling in PF

6.1. BMP4 & BMP7 protect against PF

Although BMP members have been revealed to play an anti-fibrotic function in several different organs [60,67–70], BMP4 and BMP7 are the most intensively studied in pulmonary fibrosis. Therefore, we will focus on BMP4 and BMP7 in lung fibrosis.

6.1.1. The effects of BMP4 in PF

BMP4 is a key member of the BMP family that plays roles in embryonic development and bone formation [65,66]. A previous study has demonstrated that BMP4 highlights an important factor in cardiovascular diseases [67]. Recently, Peng’s group has discovered that inhalation of BMP4 into lungs can restore normal lung regeneration and improve lung function in a murine lung fibrosis model [68]. The potential therapeutic effect of BMP4 in lung fibrosis was confirmed by another recent study from Guan et al. [47]. The authors showed that overexpression of BMP4 with an AAV system reversed bleomyin-induced murine lung fibrosis and promoted lung fibrosis resolution. Importantly, they showed that BMP4 deficient mice were more susceptible to developing PF in mice. Mechanistically, BMP4 treatment suppresses TGF-β1-induced phosphorylation of Smad2/3, α-SMA expression, and ECM protein production in mouse lung fibroblasts [47]. This data was consistent from a study by Pegorier, T.K. et al. that showed that BMP4 attenuated TGF-β1-mediated MMP13 production in human lung fibroblasts [11]. In addition, gremlin, an antagonist of BMPs including BMP4, has been discovered to be greatly increased in IPF patients, as well as in the context of TGF-β1-treated lung epithelial cells. Gremlin enhanced the fibrotic response and reduced epithelial regeneration through inhibition of BMP4 in the lungs [69–71]. Further, Guan et al. revealed that BMP4 decreased oxidative damage in lung fibroblasts, promoted mitophagy, and restored mitochondrial dynamics in TGF-β1-stimulated lung myofibroblasts [47]. These observations suggest that the anti-fibrotic effects of BMP4 occurs both through attenuation of TGF-β1 pathways and independent modulation of TGF-β1 pathways. However, not all studies support the conclusion that BMP4 antagonizes TGF-β1 pathways. In Pegorier, S. et al.’s study, BMP4 attenuated TGF-β1-induced FN mRNA and protein levels, but had no effect on α-SMA levels in human lung fibroblasts [11], suggesting that BMP4 does not attenuate all the pro-fibrotic effects induced by TGF-β1. Although most studies support that BMP4 exhibits an anti-fibrotic role, an earlier study reported that BMP4 enhanced lung fibroblast differentiation to myofibroblasts [72]. Other studies reported that BMP4 induced EMT in lungs and pancreas, which may contribute to organ fibrosis [73]. The cause of the discrepancy of the effects of BMP4 is unclear. According to the published literatures, it is well accepted that BMP4 exerts antifibrotic effects in the lungs, particularly in lung fibroblasts. Its effect in lung epithelial cells needs more investigation. These findings may guide new research directions for the utilization of BMP4 in IPF therapy.

6.1.2. The effect of BMP7 in PF

BMP7 has been reported to play an anti-fibrotic role in kidney, heart, and liver fibrosis through inhibiting TGF-β1-induced EMT and fibrotic responses in fibroblasts [9,74]. Zeisberg, M. et al. demonstrated that BMP7 promoted mesenchymal to epithelial transition (MET) in rental fibroblasts that facilities renal tissue repair after injury [75]. BMP7 not only induces phosphorylation of Smad1/5, but also suppresses TGF-β1-mediated Smad2/3 phosphorylation [9]. Phosphorylated Smad1/5 competes with phosphorylated Smad2/3 for binding to Smad4 [76]. Pegorier, S. et al. showed that BMP7 attenuated TGF-β1-induced α-SMA expression in human lung fibroblasts [11]. However, Knight’s team discovered that BMP7 could not protect against bleomycin-induced lung fibrosis [77]. Notably, this study did not examine the effect of BMP7 in lung fibroblasts. Additionally, the authors used human recombinant BMP7 in a murine model of PF, it is unclear whether human BMP7 exhibits the same effect as murine BMP7 in the in vivo study. Further, BMP7 has been shown to alleviate silica-induced pulmonary fibrosis in rats. BMP7 is capable of reducing expression of vimentin in the silica-induced lung fibrosis model and inhibiting EMT through suppressing TGF-β/Smad2/3 pathways [78,79]. Although the findings regarding the effects of BMP7 in fibrosis are still debatable, it remains a relatively well-recognized potential therapeutic approach to treat fibrotic diseases [80]. It is expected that further progress will be obtained using BMP7 combined with BMP4.

6.2. The therapeutic potential of BMPR2 in treating PF

Based on the reduction of BMPR2 in fibrotic lungs and anti-fibrotic effects exhibited by BMPs, it has been hypothesized that overexpression of BMPR2 may attenuate TGF-β1-mediated pathways and alleviate the severity of PF. This was examined in lung fibroblasts. Fukihara et al. showed that overexpression of BMPR2 with adenovirus suppressed the TGF-β1-induced p-Smad2/3 pathway. BMPR2 overexpression promoted the effect of BMP7 on inhibition of TGF-β-induced FN lung fibroblasts. Further, overexpression of BMPR2 reduced the phosphorylation of p38 MARK in lung fibroblasts [59]. Therefore, BMPR2 signaling is believed to suppress TGF-β signaling and fibroblast activation. The therapeutic potential of BMPR2 in treating PF needs to be evaluated in more studies using in vivo models of PF. Strategies to increase BMPR2 levels, such as upregulation of BMPR2 expression by AAV or increase in protein stability by decreasing its ubiquitination, may therefore have therapeutic potentials for treating PF.

7. Conclusions and perspectives

BMPs and BMPR2 in PF and PAH have attracted interests from researchers. BMP ligands, particularly BMP4/7, have been found to play roles in alleviation of PF. Previous research works on BMPR2 have been focused on PAH, but recently its function in PF has also been gained attention. Numerous studies have demonstrated that reduced expression of BMPR2 in fibrotic lungs exacerbates PF. The prevailing concept is that BMPR2 plays an anti-fibrotic role. Based on the symptomatic similarities and strong clinical correlation between PAH and PF, and it is believed that upregulation of BMPs and BMPR2 may provide new clinical therapeutic options for treating PF. Inhibitors of BMPs/BMP receptors have been developed and they are useful tools to investigate the molecular mechanisms of BMPs-mediated signaling and biological effects (Table 1) [81–101]; however, small molecules of BMP/BMPR2 activators and stabilizers are not commercially available. This review suggests a new direction for development of drugs to treat PF.

Table 1.

Inhibitors/antagonists of BMPs and BMP receptors.

Data availability

No data was used for the research described in the article.