The endothelial to mesenchymal transition (EndMT) is a newly recognized, fundamental biological process involved in development and tissue regeneration, as well as pathological processes such as the complications of diabetes, fibrosis and pulmonary arterial hypertension. The EndMT process is tightly controlled by diverse signaling networks, similar to the epithelial to mesenchymal transition. Accumulating evidence suggests that microRNAs (miRNAs) are key regulators of this network, with the capacity to target multiple messenger RNAs involved in the EndMT process as well as in the regulation of disease progression. Thus, it is highly important to understand the molecular basis of miRNA control of EndMT. This review highlights the current fund of knowledge regarding the known links between miRNAs and the EndMT process, with a focus on the mechanism that regulates associated signaling pathways and discusses the potential for the EndMT as a therapeutic target to treat many diseases.
Endothelial cells (ECs) line the inner surfaces of the blood vessels and lymphatic vessels in the body. ECs regulate vascular function by sensing and responding to various cues, and play a key role in the maintenance of vascular homeostasis (1). ECs have the capacity to undergo a dynamic cellular phenotypic switching, termed the endothelial to mesenchymal transition (EndMT), in response to local environmental cues throughout the vascular system. Since EndMT was initially described in relation to heart development (2), many studies have demonstrated the importance of the EndMT process during development (3–6). However, an increasing number of studies have demonstrated that EndMT is closely associated with postnatal pathological processes including cancer progression (7), tissue fibrosis (8), pulmonary arterial hypertension (9), neointima formation (10, 11), vascular calcification (12) and atherosclerosis (13), as well as cerebral cavernous malformations (14). The EndMT features are similar to the extensively studied and better understood epithelial to mesenchymal transition (EMT). During EndMT, ECs lose their ability to express endothelial markers, such as vascular endothelial cadherin (VE-cadherin), platelet endothelial cell adhesion molecule (PECAM-1, also known as CD31) and von Willebrand Factor (vWF). Subsequently, ECs lose their endothelial characteristics and display mesenchymal phenotypes characterized by acquisition of a highly invasive and migratory potential and gain of expression of mesenchymal markers such as alpha smooth muscle actin (α-SMA), smooth muscle protein 22 alpha (SM22α), fibronectin, vimentin and fibroblast specific protein-1 (FSP-1) (Fig. 1) (7–15). Although the molecular mechanisms underlying EndMT are complex and still largely unclear, the EndMT has been gradually defined based on studies of EMT processes, which are better understood in terms of molecular and cellular mechanisms (16). The EndMT can be regulated by multiple extracellular cues, microRNAs (miRNAs), transcription factors and various signaling pathways in different tissues and various pathophysiological conditions. Among the many regulators that control the EndMT process, miRNAs are emerging as key regulators of the EndMT program. MiRNAs are a class of endogenous, small non-coding RNAs containing about 22 nucleotides that play an important role in post-transcriptional regulation of gene expression, typically through direct binding to the 3′-untranslated region of messenger RNA (mRNA) (17, 18). A single miRNA has the capacity to target multiple mRNAs; thus, it is not surprising that miRNAs affect the gene regulatory network and are involved in global cellular processes, including development, differentiation, cell death and cell proliferation (17–19). A growing number of studies have revealed that several miRNAs have the capacity to regulate EndMT processes and such regulatory roles of miRNAs in EndMT suggest potential therapeutic targets to prevent and treat many vascular diseases via modulation of miRNA levels. Here, this review highlights the emerging role of miRNAs during the EndMT process and discuss the potential for EndMT as a therapeutic target to treat vascular diseases.
To date, a growing body of evidence shows that EndMT and EMT are primarily controlled by common signaling pathways, including transforming growth factor-beta (TGF-β) signaling, Notch signaling, and proinflammatory signaling cascades (20–22). These signaling pathways can activate or upregulate common transcription factors such as the Twist, Snail, Slug, zinc finger E-box-binding homeobox 1 (ZEB1) and ZEB2 (20–22). These transcription factors upregulate the expression of mesenchymal markers such as N-cadherin, α-SMA, SM22α, calponin, vimentin, fibronectin and FSP-1, although the precise molecular mechanisms are not fully understood. At the cellular level, these transcription factors can initiate transcriptional reprogramming and subsequently, ECs lose their apical-basal polarity and intercellular junctions, becoming mesenchymal-like cells during EndMT (20–23). While transcriptional control of EndMT has previously been studied extensively, an understanding of post-transcriptional control in this context has recently been sought, and investigated in several studies. Emerging studies have shown that miRNAs, key regulators of post-transcriptional regulation, are potent regulators of the EndMT process via targeting of key components associated with EndMT signaling pathways (Fig. 1) (11, 24).
Several miRNAs act to inhibit EndMT by directly targeting transcription factors or inhibiting signaling pathways associated with induction of EndMT. Among the signaling pathways that activate the EndMT process, the TGF-β signaling represents the most well-known inducer of EndMT. Several studies have revealed that TGF-β significantly downregulates the expression of several miRNAs (such as miR-200a, miR-20a, miR-29 and miR-630) leading to activation of the EndMT process (25–28). The miR-200 family is composed of five members; miR-200a, miR-200b, miR-200c, miR-141 and miR-429. Before the study of EndMT, it had been demonstrated that the miR-200 family had inhibitory effects on EMT through targeting of ZEB1 and ZEB2 (29). A recent study investigated the role of miR-200a (which is well known to inhibit the EMT process) as it relates to EndMT (25). It was demonstrated that miR-200a expression was significantly downregulated after treatment with TGF-β1 of human aortic endothelial cells (HAECs), while TGF-β1 treatment upregulated the expression of growth factor receptor-bound 2 (GRB2). Overexpression of miR-200a resulted in significant inhibition of EndMT via downregulation of FSP-1 and α-SMA and upregulation of VE-cadherin and PECAM-1. At the same time, miR-200a targets
Several studies have suggested that the fibroblast growth factor (FGF) signaling modulates TGF-β signaling via regulation of many genes in various cell types (32–34). Fafeur
It has been reported that other miRNAs are also involved in TGF-β-induced EndMT. The MiR-23 negatively regulates TGF-β-induced EndMT in mouse embryonic endothelial cells (MEEC) and identified hyaluronic acid synthase 2 (Has2) as a direct target of miR-23. MiR-23 plays an essential role in cardiac valve formation by regulating Has2 expression (38). Bayoumi
The molecular mechanisms of endothelial dysfunction by high glucose have been elucidated in the pathologic context of various vascular diseases (41). Emerging evidence has demonstrated that high glucose can lead to EndMT via regulation of the expression of miRNAs (such as miR-200b and miR-18a-5p), and contribute to the progression of diabetic complications (42–44). Cao
On the contrary, several miRNAs have the capacity to promote EndMT by directly targeting molecules associated with inhibition of EndMT (45–49). For example, miR-21 expression increased during TGF-β-induced EndMT in HUVECs and inhibition of miR-21 has been shown to partially prevent TGF-β-induced-EndMT. In addition, it was found that miR-21 negatively regulated phosphatase and tensin homolog (PTEN), a well-known target of miR-21, and following activation of the Akt pathway, resulted in promotion of the EndMT.
Many studies have demonstrated that fibroblasts are implicated in a multitude of pathologies, and there is substantial evidence to indicate that they are the central mediator of pathological tissue remodeling (50). Several studies have demonstrated that a large proportion of fibroblasts found in damaged tissues are of endothelial origin via EndMT (7, 51). These results suggest that the EndMT is an attractive prospective target with regard to prevention and treatment of many diseases. Indeed, the EndMT plays an essential role during development (52) and can also contribute to postnatal pathologies associated with many diseases such as fibrosis, neointima formation, diabetic complications, heterotopic ossification, Kawasaki disease and pulmonary arterial hypertension (11, 24, 25, 27, 28, 38, 42–46, 48, 49, 53). Given that EndMT is closely involved in multiple diseases, the blocking of EndMT may represent a useful strategy for implementation in the treatment plans developed to combat human diseases. In addition, the miRNAs play a key role in the maintenance of homeostasis in the entire vasculature as they have the capacity to target multiple protein-encoding genes. Therefore, imbalances in the expression of miRNAs are closely related to the pathogenesis of many diseases via abnormal regulation of their target mRNAs. Thus, a strategy based on restoration of abnormal miRNA expression could have important therapeutic value for the treatment of various diseases. Finally, given the documented close relationship between EndMT and miRNAs as it occurs in various pathologies, modulation of miRNAs involved in EndMT processes could be a new therapeutic strategy for treatment of human diseases (17, 18). The therapeutic potential of modulation of miRNA expression in regulating the EndMT process in several diseases is summarized below.
As previously discussed, ECs represent a major source of the fibroblasts found in pathological fibrotic tissues via EndMT. In this context, let-7 miRNA plays a critical role in the regulation of EndMT via targeting components of TGF-β signaling in HUVECs and HUAECs. It has also been shown that let-7 miRNA is regulated by FGF signaling (11, 24). It was demonstrated that let-7 miRNA and FGF receptor expression and phosphorylation were suppressed in diabetic condition(s). In addition, it was found that the endogenous antifibrotic peptide N-acetyl-seryl-aspartyl-lysyl-proline (AcSDKP) restored let-7 miRNA levels by means of restoration of FGF receptor expression, as well as phosphorylation to normal levels. Thereby, AcSDKP inhibited EndMT, as evidenced by increased endothelial and reduced mesenchymal marker expression and, in turn, ameliorated diabetic kidney fibrosis. This suggests that AcSDKP could be a potential therapeutic option for diabetic kidney fibrosis as an endogenous antifibrotic molecule via inhibition of EndMT (24). The same group further showed the therapeutic potential of modulation of dysregulated miRNA during EndMT in the pathogenesis of diabetic kidney fibrosis (28). They found that DPP-4 could be induced in the diabetic kidney, and miR-29 expression was inhibited in diabetic mice (which identified DPP-4 as a direct target of miR-29). Thus, they identified the potential of DPP-4 inhibitor, which is generally used to treat diabetes mellitus type 2, for restoration of dysregulated miR-29 levels. As a result, DPP-4 inhibitor linagliptin restored miR-29 levels in diabetic kidney, and also found suppression of DPP-4 activity and protein expression. Thereby, EndMT was inhibited and, in turn, kidney fibrosis was ameliorated in diabetic kidney. These results suggest that linagliptin has potential therapeutic value for the restoration of normal kidney function in diabetic patients with kidney fibrosis via inhibition of EndMT (28).
MiRNAs play a key role in regulating EndMT by targeting multiple components associated with signaling pathways that regulate EndMT. As described above, many studies support that dysregulation of miRNAs in vascular ECs leads to activation of the EndMT process and contributes to the pathogenesis of human diseases. Thus, strategies that restore miRNA expression to physiological levels are attractive therapeutic opportunities for the treatment of human diseases via inhibition of EndMT. Although our current understanding of the molecular mechanisms underlying the miRNA-EndMT axis is advancing, more work is still required to better understand the complex network of miRNAs and their target mRNAs involved in the EndMT process. In addition, understanding of the regulation of the reversible biological process of EndMT is necessary for the prevention and treatment of many diseases. In conclusion, studies of miRNA involved in EndMT will provide new insights into the molecular mechanisms of a broad variety of human pathologies and the identification of potential targets that are able to inhibit EndMT, will provide effective therapeutic drugs for the treatment of human diseases.
This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the minister of Education, Science and Technology (NRF-2016R1A5A1011974 and NRF-2016R1C1B 2006591 to J.K).
The authors have no conflicting interests.
MicroRNAs and target genes regulating the EndMT
|Endothelial cell type||microRNA||Target||Effect on EndMT||Clinical relevance||Reference|
|HUAEC, HUVEC||Let-7||TGFβR1||Inhibit||Neointima formation and fibrosis||11, 24|
|HUVEC||miR-20a||TGFβR1, TGFβR2, SARA||Inhibit||Non determined||26|
|MEEC||miR-23||Has2||Inhibit||Cardiac valve formation||38|
|CEC||miR-532||PRSS23||Inhibit||Acute myocardial infarction||39|
|HRMEC||miR-200b||Smad2, Snail||Inhibit||Diabetic retinopathy||42|
|MS-1||miR-27b||Elk1, Neuropilin 2, Plexin A2, Plexin D1||Promote||Non determined||47|
|LMVEC||miR-130a||BMPR2||Promote||Pulmonary arterial hypertension||48|
|RPMEC||miR-126-5p||Non determined||Promote||Neonatal pulmonary hypertension||49|
HUAEC: human umbilical artery endothelial cell, HUVEC: human umbilical vein endothelial cell, HAEC: human aortic endothelial cell, HD-MVEC: human dermal microvascular endothelial cell, HMVEC: human dermal microvascular endothelial cells, MEEC: mouse embryonic endothelial cell, CEC: cardiac endothelial cell, HRMEC: human retinal microvascular endothelial cell, MHEC: mouse heart endothelial cell, HAVEC: human aortic valvular endothelial cell, MCEC: mouse cardiac endothelial cell, MS-1: mouse pancreatic microvascular endothelial cell, LMVEC: lung microvascular endothelial cell, RPMEC: rat pulmonary microvascular endothelial cell.