2. DNA Methylation in Fibrosis
Extensive alterations in DNA methylation profiles are known to be involved in the pathogenesis of pulmonary fibrosis. DNA methylation microarray demonstrated a higher DNA methyltransferase expression in lung tissue samples of IPF patients [
81]. IPF fibroblasts exhibited significant heterogeneity in global DNA methylation patterns compared to non-fibrotic control cells [
82]. Additionally, a study identified extensive alterations in gene expression-associated DNA methylation that were involved in fibroproliferation [
83]. Recently, a genome-wide DNA methylation study found that CpGs methylation was responsible for cell adhesion, molecule binding, chemical homeostasis, surfactant homeostasis, and receptor binding categories in the pathogenesis of IPF [
84].
Global methylation patterns in IPF are very different from that of the control samples and significantly overlap with methylation changes observed in lung adenocarcinoma samples [
85]. DNA methylation biomarkers also overlap between IPF, other ILD, cancer, and COPD; hence, it is unlikely that molecular discrimination between these diseases can be achieved using a single marker [
86]. Interestingly, low-methylation lung squamous cell carcinoma (SCC) significantly correlated with IPF to show a bad outcome [
87]. Cigarette smoking-associated aberrant methylation contributed to pulmonary fibrosis [
88].
Many cell types are well known to be involved in the process of pulmonary fibrosis. The essential cells driving the development and progression of pulmonary fibrosis are epithelial, fibroblasts and myofibroblasts, and alveolar macrophages. Epigenetic regulation has a variety of ways. This discussion focuses on DNA methylation regulation in fibroblasts, epithelial, and macrophages.
2.1. DNA Methylation and EMT
EMT is a cellular and molecular process in which epithelial cells lose their epithelial identity (apical-basal polarity and adhesion), which is characterized by down-regulated epithelial markers, including E-cadherin, occludin, and claudin-1. In contrast, fibroblast-specific genes, such as α-SMA, N-cadherin, fibroblast-specific protein 1 (FSP-1), and type I collagen, are up-regulated [
10,
89]. A hallmark of EMT is the repression of E-cadherin: a transmembrane glycoprotein encoded by the epithelial marker gene CDH1. DNA methylation might be involved in regulating EMT. During EMT induction, EMT-transcriptional factors (EMT-TFs), including ZEB1, SNAIL, and TWIST, interact with DNMTs and undergo the hypermethylation of CpG islands in the CDH1 promoter [
90,
91]. The DNA hypermethylation of the CDH1 through DNMTs was correlated with tumor progression [
90,
92]. In addition, the knockdown of DNMTs was equated with the transfection of siDNMT1, DNMT3a, or DNMT3b, which reversed the TGF-β1-induced suppression of E-cadherin expression and the induction of α-SMA, vimentin, and fibronectin expression [
93].
In normal lung epithelial wound healing, this process ends with the apoptosis of myofibroblasts and inflammation reduction. Yet, aberrant responses to tissue injury may turn into lung fibrosis. Bidirectional EMT cross-talk assists the pro-fibrogenic positive feedback loop, while epithelial cells become “vulnerable and sensitive to apoptosis” and myofibroblasts become “apoptosis-resistant and immortal”, resulting in fibrosis progression instead of wound resolution [
94]. DNA methylation mediated the down-regulation of genes involved in apoptosis. The hypermethylation of the Caspase 8 promoter by DNMT1 and DNMT3b was associated with apoptosis resistance in cancer cells [
95,
96]. Cisneros et al. showed the silencing expression of pro-apoptotic p14ARF in the fibroblast of the patient with IPF due to hypermethylation [
97]. Along with these changes, DNA methylation was linked with the dysregulation of apoptosis and EMT, leading to the development of lung fibrosis processes (
Figure 2).
Figure 2. (1) Induction of EMT. Down-regulation of CDH1 (anti-fibrotic factor) is mediated by EMT-TF, which recruits DNMTs to the CDH1 promoter performing focal hypermethylation of the CpG islands in the CDH1 promoter, leading to decreased gene expression of CDH1. (2) Reduction in apoptotic activity. Hypermethylation mediated by DNMTs of the pro-apoptotic factors, including Caspase 8 and p14 ARF, could underlie resistance to fibroblast apoptosis.
2.2. DNA Methylation and Myofibroblast Differentiation
Under normal physiology, in response to epithelial injury, the cell releases various pro-inflammation and fibrotic cytokines that are responsible for local inflammation and the activation of fibroblasts. Fibroblasts are mesenchymal cells that are recruited to and accumulate at the injured site and undergo a change in phenotype to highly proliferative and contractile myofibroblasts.
Lung myofibroblasts are heterogeneous in terms of their origins. The predominant sources of myofibroblasts are resident fibroblasts (lipofibroblasts, matrix fibroblasts, and alveolar niche cells) and pericytes (residing within basal membranes or perivascular linings), along with minor sources such as hematopoietic CXCR4+ fibrocytes, alveolar epithelial cells (AECs), endothelial cells (ECs), and mesenchymal stem cells (MSCs) [
98]. Under the influence of environment-specific factors, cytokines, such as TGF-β, fibroblast-to-myofibroblast transition/transdifferentiation (FMT) changes lung resident fibroblast behavior/phenotypes to another type of normal differentiated cell and the downstream effects at the tissue level [
99]. Myofibroblasts, characterized by the expression of contractile α-SMA, regulate connective tissue remodeling by producing and modifying ECM components, such as fibronectin and collagen [
100]. Furthermore, one of the mechanisms that may account for FMT is DNA methylation.
DNA methylation modulates myofibroblast differentiation (
Figure 3). Several genes and signaling pathways, such as those involving ErbB, focal adhesion, and MAPK, underlie mechanisms for DNA methylation alteration in myofibroblast differentiation and influence the synthesis of ECM [
101]. DNA methylation regulates gene expression to facilitate the formation of fibroblastic foci and lung fibrosis [
47]. DNMT is an essential regulator of α-SMA gene expression during myofibroblast differentiation. The specific effect of DNA methylation in regulating the α-SMA gene is unknown. During the development of cardiac fibrosis, He et al. found that TGF-β could inhibit DNMT1-mediated DNA methylation expression, leading to the overexpression of α-SMA [
102]. TGF-β1 also regulates lung fibroblast differentiation through the hypermethylation of the “fibrosis suppressor” gene, Thymocyte differentiation antigen-1 (
THY-1); therefore, DNMT1-attenuated TGF1-mediated THY-1-inducing α-SMA fiber formation is silenced [
103]. In addition, the expression of α-SMA was significantly increased by adding IL-6 [
104]. A recent study revealed that DNMTl knockdown suppressed the DNA methylation-related myofibroblast phenotype via inhibiting an IL-6/STAT3/NF-κβ positive feedback loop [
105].
Figure 3. DNA methylation regulate myofibroblasts and macrophages. The methylation of promoter DNA can repress gene expression by MBPs, recruiting specific macromolecular complexes that contain histone deacetylases (HDAC), DNMTs, and other transcriptional co-repressors (Co-Rep). Several genes that function as an anti-fibrotic and are associated with myofibroblast differentiation and macrophage M2 polarization have been shown to undergo silencing due to DNA hypermethylation.
MBD protein families have been suggested to play an essential role in myofibroblast differentiation. The hypermethylation of the transcriptional regulator, c8orf4, decreased the capacity of fibrotic lung fibroblasts to up-regulate COX-2 expression and COX-2-derived PGE2 synthesis [
106]. Evidence showed that MeCP2 binds with the I
κB
α promoter and induces the α-SMA promoter [
107,
108]. The inhibition of MeCP2-associated TGF-β1 significantly blocks the expression levels of α-SMA and fibronectin production during myofibroblast differentiation [
109]. In addition, suppressing MeCP2 or MBD2 might reduce pro-fibrotic factors. Wang et al. determined higher MBD2 levels in different types of pulmonary fibrosis patients, including COVID-19, systemic sclerosis-associated interstitial lung disease, and IPF. Therefore, deleting the MBD2 gene attenuated bleomycin-induced lung injury and fibrosis [
110]. MBD2 promoted fibroblast differentiation via repressing the expression of the erythroid differentiation regulator (Erdr) 1 [
111].
Interestingly, other studies showed the anti-fibrotic effects of MBDs protein families. The in vitro study of diffuse cutaneous scleroderma (SSc) showed that the overexpression of MeCP2 via the modulating genes PLAU, NID2, and ADA suppressed fibroblast proliferation, migration, and myofibroblast differentiation [
112]. Perhaps the final effects of DNA methylation on pulmonary fibrosis depends on the hypomethylation of pro-fibrosis genes and hypermethylation of anti-fibrosis genes [
111]. This evidence revealed that the role of the epigenetic gene transcription regulator in myofibroblast differentiation was complex.
Hypoxia can alter DNA methylation patterns and contribute to the fibrotic process. Global DNA methylation was detected in hypoxic fibroblasts and was associated with increased DNMT1 and DNMT3A expression and decreased THY-1 expression [
113,
114]. Hypoxia altered DNA methylation, which is characterized by the increased expression of TGF-β1 and DNMT1, and significantly decreased RASAL1 [
115]. Chronic hypoxia induces oxidative stress. Recently, the administration of the antioxidant enzyme extracellular superoxide dismutase (EC-SOD) has been shown to alleviate the hypoxia-induced DNA methylation of the Ras association domain family 1 isoform A (RASSF1A) through the Ras/ERK pathway [
116].
2.3. DNA Methylation and Macrophage Polarization
Pulmonary macrophages consist of monocyte-derived alveolar (AMs) and lung tissue-resident alveolar or interstitial macrophages (IMs). Both lung tissue-resident macrophages and monocyte-derived macrophages can be polarized into classically activated macrophages (M1) or alternatively activated macrophages (M2) by their capacity to induce inflammatory or anti-inflammatory immune responses, respectively [
117]. The heterogeneity of macrophages regulates the development of pulmonary fibrosis from the early phases of injury and the fibrotic phase. The M2 macrophage, instead of the M1 phenotype, is involved in the progression of lung fibrosis [
118]. After the resolution of the lung injury, monocyte-derived AM persisted and expressed higher pro-inflammatory and pro-fibrotic functions, and therefore, leading to the depletion of monocyte-derived AM ameliorates lung fibrosis [
119].
DNA methylation alterations in IPF lungs have been well documented, but the role of DNA methylation in macrophages needs to be better defined. Many studies have shown that DNA methylation plays a fundamental role in macrophages, including the regulation of injury-induced inflammation and the control of fibroblast activation in the wound-healing response. Chen et al. examined DNA methylation changes in lung macrophages and found that changes in DNA methylation patterns drove the dysregulation of innate immune cells in cystic fibrosis [
120]. Profiling DNA methylation from healthy subjects and IPF patients showed that aberrant macrophage polarization played a crucial role in developing IPF, but epigenetic alterations were not associated with accelerated aging [
38]. DNMTs are associated with the activation of gene expression during macrophage phenotypic changes. During the development of IPF, DNMTs are associated with inappropriate gene expression activation, resulting in the opposite roles of M1/M2 polarization.
DNMTs can act as pro- and anti-fibrotic by regulating M2 polarization. Yang et al. revealed that DNMT3a and DNMT3b suppress the Proline-Serine-Threonine Phosphatase Interacting Protein 2 (PSTPIP-2) by hypermethylation results, which caused a mixed induction of hepatic macrophage M1 and M2 [
121]. In obesity, DNMT3b regulates M2 macrophage polarization, and the deletion of DNMT3b induces M2 macrophage polarization [
122]. On the contrary, another study provided different results. Qin et al. found that DNMT3b ameliorates the development of bleomycin-induced pulmonary fibrosis by limiting M2 macrophage polarization [
123]. In this case, DNMT3b acts as a negative regulator of M2 macrophage polarization. The role of DNMT3-associated macrophage polarization in organ fibrosis might have different functions. DNMT3 limits M2 macrophage polarization and alleviates the progression of lung fibrosis, yet DNMT3 induces M2 polarization and enhances liver fibrogenesis.
Consistent with the involvement of DNMTs in macrophage phenotype change, MBPs interact with methylated DNA to regulate the expression of multiple genes. However, MeCP2 can accelerate both M1 and M2 polarization. MeCP2 might tend to M1 in the early phase of inflammation. The deletion of MeCP2 down-regulated M1 differentiation and inhibited the expression of iNOS and the secretion of cytokines that are pro-inflammatory in acute lung injury (ALI) [
124]. MeCP2 regulates the gene expression of macrophages; thus, MeCP2 deficiency leads to the dysregulation of macrophage polarization [
125]. MeCP2 accelerates M2 polarization by inhibiting Ship expression and enhancing the phosphatidylinositol 3-kinases/protein kinase B (PI3K/PKB) signaling pathway in bleomycin-induced pulmonary fibrosis [
110]. Furthermore, silencing MeCP2 using siRNA blunted M2 macrophage polarization to elevate IRF4 expression [
75]. Conversely, MeCP2 directly or indirectly promoted the differentiation of M0 macrophages to M1 or M2 macrophages and the polarization of M2 to M1 macrophages [
126]. Therefore, researchers might suggest that MeCP2 stimulates M2 macrophages to promote the progression of lung fibrosis but protect against renal fibrosis via M1 polarization.
Little is known about the impact of TET proteins on macrophage polarization. TET regulates macrophage differentiation and M1-polarization [
127]. Therefore, the inhibition of the TET protein using dimethyloxallyl glycine (DMOG) promotes M2 polarization by up-regulating Arg1, Fizz1, and Ym1 [
128].