The role of epithelial-mesenchymal transition in pulmonary fibrosis: lessons from idiopathic pulmonary fibrosis and COVID-19

Abstract

Despite the tremendous advancements in the knowledge of the pathophysiology and clinical aspects of SARS-CoV-2 infection, still many issues remain unanswered, especially in the long-term effects. Mounting evidence suggests that pulmonary fibrosis (PF) is one of the most severe complications associated with COVID-19. Therefore, understanding the molecular mechanisms behind its development is helpful to develop successful therapeutic strategies. Epithelial to mesenchymal transition (EMT) and its cell specific variants endothelial to mesenchymal transition (EndMT) and mesothelial to mesenchymal transition (MMT) are physio-pathologic cellular reprogramming processes induced by several infectious, inflammatory and biomechanical stimuli. Cells undergoing EMT acquire invasive, profibrogenic and proinflammatory activities by secreting several extracellular mediators. Their activity has been implicated in the pathogenesis of PF in a variety of lung disorders, including idiopathic pulmonary fibrosis (IPF) and COVID-19. Aim of this article is to provide an updated survey of the cellular and molecular mechanisms, with emphasis on EMT-related processes, implicated in the genesis of PF in IFP and COVID-19.

Introduction

COVID-19 is an emerging respiratory disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), a new type of beta coronavirus. This novel coronavirus is highly infectious and was able to spread rapidly across the world, leading to an instant pandemic and becoming a major public health issue of the 21st century. According to the estimates of the World Health Organization (WHO), SARS-CoV-2 has resulted in 776 million of infections and more than 7 million of deaths worldwide, and it continues to spread. However, through mass vaccination efforts against SARS-CoV-2, the disease has been brought under control [1,2,3,4]. COVID-19 presents with a range of clinical symptoms, including mild symptoms such as fever, dry cough, sore throat, malaise, vomiting, diarrhea, loss of taste and smell, as well as severe manifestations such as pneumonia, acute respiratory distress syndrome (ARDS), septic shock, multiple organ dysfunction, and even death in some cases of infection. A systemic inflammatory reaction called cytokine storm is believed to be the most important factor responsible for lung and extrapulmonary organ injuries in COVID-19. However, while COVID-19 can affect various organs such as the kidneys, heart, and brain, the lungs are the most affected organ.

Many individuals with COVID-19 experience either no symptoms or mild symptoms, but some patients endure severe COVID-19 and develop conditions such as severe pneumonia, ARDS, and ultimately PF. One of the most severe respiratory manifestations of COVID-19 is a decrease in lung diffusion capacity for gases caused by ARDS and ARDS-induced PF. PF is a group of interstitial lung diseases (ILDs) characterized by scarring and thickening of lung tissue. This condition leads to a decline in lung function and can result in respiratory failure and death due to excessive formation and deposition of extracellular matrix (ECM) in the lung parenchyma. Fibroproliferation and lung remodeling occur as a response to repair of damaged tissue following ARDS and alveolar epithelial cell injury during lung infection with SARS-CoV-2. However, if this process is imbalanced or prolonged, it can result in a fibrotic response [1,2,3, 5,6,7]. Evidence shows that up to 17% of severe or critical COVID-19 patients and more than a third of recovered patients develop PF. Several studies have discussed the mechanisms by which SARS-CoV-2 infection leads to PF [5, 6, 8]. During fibrosis, cellular conversion processes, particularly of epithelial and endothelial cells to a mesenchymal phenotype, can provide a new source of activated fibroblasts that produce ECM components. The epithelial to mesenchymal transition (EMT) process is known to be one of the significant drivers of fibrosis in various pathological conditions, including PF [9, 10]. Different types of PF have been diagnosed. PF is a characteristic aspect of idiopathic pulmonary fibrosis (IPF). IPF is the most common and severe form of these disorders with very limited treatment options [11, 12]. IPF is characterized by the destruction of the normal structure of the lung alveoli and its replacement with progressive and irreversible scarring without a definite etiology, leading to a decrease in lung function and, in about 50% of cases, death from progressive dyspnea within 3–5 years after diagnosis [9, 10]. Aim in this article is to analyze cellular and molecular mechanisms implicated in the genesis of PF in a context of IPF and COVID-19, with emphasis on reprogramming events leading to the acquisition of a mesenchymal-like profibrogenic phenotype.

SARS-CoV-2 plasma membrane receptors and pathogenesis of cytokine storm

Like other coronaviruses, SARS-CoV-2 utilizes the spike (S) protein to bind plasma membrane receptors as a first step of viral entry. Angiotensin-converting enzyme 2 (ACE2) serves as entry receptor for nl63, SARS-CoV, and SARS-CoV-2 within the Coronaviridae family as part of the renin-angiotensin system (RAS). This system regulates several physiological processes, including blood pressure and fluid balance, wound healing, and inflammation. ACE2 is expressed in numerous cell types and tissues, such as the lungs, heart, blood vessels, kidneys, liver, and gastrointestinal tract​​​​​​​. It is primarily present in alveolar cells, bronchial cells, and the vascular endothelium in the lung, which are the most important sites targeted by SARS-CoV-2 [5, 13,14,15]. In addition to ACE2, other possible plasma membrane receptors or cofactors for SARS-CoV-2 cellular entry have been identified. These receptors may be important, particularly in cells where ACE2 expression is reduced or absent; however, limited information has been published about them. Some of these receptors include Cluster of Differentiation (CD)147 (Basigin), CD26 (DPP4), CD209 (CLEC4M), CLEC4G, AXL, KREMEN1, ASGR1 (CLEC4H1), transferrin receptor, Angiotensin II receptor type 2 (AGTR2) and TIM [16,17,18]. In COVID-19 infection, upon binding of the SARS-CoV-2 S protein to the ACE2 receptor on the cell surface, the viral S protein is cleaved by proteolytic enzymes. Host proteases, including Transmembrane Serine Protease 2 (TMPRSS2), Furin, Cathepsin L (CTSL), and the A Disintegrin and Metalloproteinase (ADAM) family sheddases such as ADAM10 and ADAM17, are known to be involved in the viral transmission or infection process. Neuropilin-1 (NRP1) has been shown to facilitate cellular entry of SARS-CoV-2 via endocytosis as a host cell surface mediator by binding to the furin-cleaved S1 segment of the S protein [17, 19]. The cleavage of the S protein following the interaction of SARS-CoV-2 with ACE2 mediates the direct fusion of the viral envelope with the plasma membrane of the target cells, or membrane fusion within endosomes through the host receptor-mediated endocytic route in the absence of proteases in the local environment [20,21,22]. Subsequently, the viral RNA genome is released into the cytoplasm, where it is translated into a larger polyprotein responsible for the production of nonstructural proteins and subgenomic mRNAs for the production of structural proteins. These viral proteins are assembled to form virion particles, which are ultimately released from infected cells through budding at the plasma membrane [23]. Following entry into the host cell, SARS-CoV-2 replication has a cytopathic effect, resulting in the release of damage-associated molecular patterns (DAMPs), that can in turn activate the immune system [24]. For instance, DAMPs, upon interaction with pattern recognition receptors (PRRs) on innate immune cells such as macrophages, monocytes, dendritic cells, neutrophils, and natural killer (NK) cells, can induce immune responses, trigger inflammatory signaling pathways, and promote additional cell injury [25]. Proinflammatory cytokines that have been shown to be secreted following the release of DAMPs from virus-infected cells include interleukin (IL)-1β, IL-2, IL-6, IL-10, interferon-gamma (IFN-γ), tumor necrosis factor alpha (TNF-α), inducible protein 10 (IP-10), granulocyte macrophage colony-stimulating factor (GM-CSF), and monocyte chemoattractant protein-1 (MCP-1). However, in some COVID-19 patients, the secretion of these inflammatory cytokines becomes uncontrolled, a phenomenon commonly referred to as a cytokine storm [2, 13, 26, 27]. It has been demonstrated that the cytokine storm is primarily responsible for acute lung damage, ARDS, and ultimately death. This condition occurs due to fluid buildup in pulmonary alveolar cells, causing significant damage to the endothelial and epithelial surfaces and disrupting the alveolar epithelial-endothelial barrier. The inflammatory and intercellular fluids fill up the alveoli leading to hypoxia and respiratory distress [3, 6, 13, 15]. As mentioned above, ACE2 is part of the RAS. The main role of ACE2 in the RAS is to cleave angiotensin II (Ang II). Biochemical and biomechanical stimuli favor the release of renin, which cleaves angiotensinogen to generate a 10-amino acid peptide called Ang I. Ang I is then converted into an 8-amino acid peptide called Ang II by an enzyme called ACE, located in the endothelium of pulmonary vessels. Ang II causes vasoconstriction, leading to increased blood pressure. Additionally, Ang II may promote inflammation and thrombosis. It has been shown that Ang II can increase inflammation and cause death of alveolar cells, which are crucial for oxygen delivery to the body [14, 28]. Consequently, it can play a significant role in the development of ARDS and ultimately PF. It has demonstrated that ACE2 neutralizes the harmful effects of Ang II by cleaving it into Ang-(1–7). By binding to the MAS receptor (MasR), Ang-(1–7) enhances vasodilation, vascular protection, having anti-fibrotic and anti-inflammatory effects [28]. Therefore, it can be inferred that ACE2 plays an anti-fibrotic role and helps prevent pulmonary epithelial damage and PF. However, the binding of SARS-CoV-2 to ACE2 leads to downregulation of its expression, thereby hindering its homeostatic function of Ang II cleavage, leading to increased inflammation and fibrosis, which can cause acute and chronic lung damage [6, 13,14,15, 29].

COVID-19 and PF

Pathogenesis of PF associated with COVID-19 infection

One of the concerns regarding long-term effects following a COVID-19 infection is the onset of a progressive fibrotic interstitial lung disease, which causes non-reversible lung deterioration [30, 31]. PF is a chronic disease characterized by a disordered wound healing process, damaged alveolar epithelium, the excessive proliferation of fibroblasts/myofibroblasts, and increased deposition of ECM. This is accompanied by a decrease in lung capacity and function [32, 33]. The severity of COVID-19 is associated with an increased risk of PF [34, 35]. The pathogenic mechanisms and anatomical alterations occurring in lung fibrosis in post-COVID-19 share similarities with IPF (Table 1) [32, 36]. Indeed, onset of honeycomb fibrosis, a common feature of IPF, was demonstrated to occur as a sequela of COVID-19 pneumonia [37].

Table 1 Alterations in the expression of fibrosis-related genes in IPF and COVID-19

Key cell types involved in IPF and COVID-19-associated PF

Alveolar epithelial cells (AECs)

The alveolar epithelium of the lung consists of two types of cells: type I alveolar epithelial cells (AECI), which cover more than 95% of the internal surface of the alveolar wall, and type II alveolar epithelial cells (AECII), which are located between AECI and make up the remaining 2–5% [43]. The AECI provides a thin-walled gas exchange surface within the alveolus [44]. AECII are multifunctional cells with secretory and regenerative functions that are important for maintaining lung homeostasis. These functions include biosynthesis and secretion of pulmonary surfactant, as well as regeneration of the alveolar epithelial layer through trans-differentiation into AECI [43, 45].

Complex cellular interactions between epithelial cells, fibroblasts, immune, and endothelial cells drive the dynamic process of fibrogenesis and scar formation in IPF [45]. Although the pathogenesis of IPF has not been fully identified [46], evidence suggests that AECs, also known as pneumocytes, play a central role in the pathogenesis of IPF [47], and epithelial cell dysfunction is considered the main component in the pathophysiology of IPF. Injured AECs can establish crosstalk between epithelial cells and fibroblasts by secreting several profibrotic cytokines, leading to the activation of myofibroblasts and deposition of ECM [48]. It has been shown that a continuous damage leads to the inability of AECII to repair the damaged alveolar epithelium [45]. Dysfunction and death of AECII, which cause dysregulated repair of the alveolar surface, pathogenic activation of fibroblasts, and myofibroblast accumulation, are recognized as initiating events in the development and progression of IPF [45, 46].

IPF and post-COVID-19 pulmonary fibrosis (PCPF) share numerous similarities. These include the presence of shared mediators that contribute to fibrosis in both conditions, genetic overlaps, similar cytopathic changes, gene expression and cytokine secretion patterns, common clinical and radiological manifestations, as well as the effectiveness of shared treatment approaches for both types of fibrosis [49]. Several cell types, such as lung epithelial cells, immune cells, and lung fibroblasts, are involved in the PF induced by SARS-CoV-2. ACE2 is predominantly expressed in AECs. Indeed, AECII are highly vulnerable to SARS-CoV-2 infection. This is a first step of a process leading to the activation of the innate/adaptive immune system and resulting in severe lung damage [6, 32, 36, 50]. Hence, the development of PF can be regarded as a potential outcome of SARS-CoV-2 infection (Fig. 1) [51].

Fig. 1

An overview of the mechanism of pulmonary fibrosis in patients with SARS-CoV-2 infection. SARS-CoV-2 virus entry into the lung alveoli is followed by the binding of specific surface receptors such as ACE2, leading to the infection of lung resident cells—including AECII and lung macrophages. In response to SARS-CoV-2 infection, these cells release a variety of inflammatory cytokines and chemokines, which in turn attract additional immune cells, leading to a further expansion of the inflammatory response. This process also facilitates the activation of resident resting fibroblasts, the recruitment of circulating fibroblasts precursors, and ultimately results in the deposition of ECM proteins, such as collagen and fibronectin. Furthermore, endothelial dysfunction induced by COVID-19 contributes to inflammation and increases vascular permeability. AECI and AECII; type I and type II alveolar epithelial cells, ACE2; Angiotensin-converting enzyme 2, ECM; extracellular matrix

Fibroblasts

Fibroblasts are the most abundant cell type of the connective tissues and are key players in lung fibrosis [52]. During fibrogenesis, fibroblast are activated and responsible for the production and deposition of ECM proteins in scar tissue [53]. Histopathologically, the key morphological marker of scar lesions in IPF is the presence of distinct and complex structures called fibroblast foci, predominantly in peripheral and basilar areas. These areas serve as the main accumulation sites of ECM-producing fibroblasts/myofibroblasts and collagen synthesis [54,55,56]. It has been demonstrated that myofibroblasts in fibroblast foci can stem from various cellular sources, including lung resident fibroblasts, bone marrow-derived fibroblasts, lung resident mesenchymal cells, bone marrow-derived CD34 + progenitor cells (fibrocytes), and microvascular pericytes. Additionally, AECs, endothelial cells and mesothelial cells (MCs) undergoing specific dedifferentiation processes i.e. EMT, Endothelial-Mesenchymal Transition (EndMT) and mesothelial to mesenchymal transition (MMT), respectively, can be regarded as potential sources of myofibroblasts in IPF [10, 54, 57].

The inflammatory-driven mesenchyme/fibroblast response results in dysregulated lung repair in COVID-19 and decreases alveolar repair. The main pathological feature in pneumonia of COVID-19 patients is an increase in lung fibroblasts, particularly myofibroblasts [36]. The hyperproliferation of myofibroblasts has been observed in patients experiencing severe forms of COVID-19 [58, 59]. The Wnt/β-catenin pathway, through increased production of IL-1β, plays a crucial role in the activation of fibroblasts during COVID-19-induced PF.

COVID-19 infection stimulates several key signaling pathways, including Notch, Nuclear factor kappa B (NF-κB), phosphatidylinositol-3-kinase (PI3K)/Akt, and hypoxia inducible factor (HIF), leading to increased fibroblast activation and differentiation into myofibroblasts. This process also results in increased ECM deposition and inflammatory processes, ultimately contributing to the development of PF [60]. It has demonstrated that COVID-19 and IPF exhibit similar gene expression patterns, including fibrosis- and inflammation-associated genes [61].

Endothelial cells

Endothelial cells, which line the inner surface of the vascular lumen, are an important component of the protective barrier of the vessel wall [62]. Endothelial cells make up 30% of the lung cell population [63]. Maintaining endothelial integrity is crucial for preserving pulmonary homeostasis [64]. It has been demonstrated that endothelial cells may participate in PF through functions such as nitric oxide (NO) production and/or conversion to myofibroblasts [65]. Following lung injury, overexpression of Sterol regulatory element-binding protein 2 (SREBP2) in endothelial cell plays a main role in the induction of mesenchymal and fibrosis-related genes such as transforming growth factor-beta (TGF-β), Wnt, Snail family transcriptional repressor 1 (Snai1), α-smooth muscle actin (α-SMA), vimentin, and N-cadherin [63]. Endothelial cell dysfunction play a causal role in the progression of PF [63, 66]. Vascular alterations, such as microvascular leakage and increased endothelial permeability, have been observed both in IPF and in ARDS caused by COVID-19 [6]. SARS-CoV-2 may impact on endothelial cells through two mechanisms: direct virus-induced endothelial cell infection or indirect injury orchestrated by circulating inflammatory molecules induced during the immune responses [67, 68]. The transcription factor Kruppel-like factor 2 (KLF2) plays a crucial role in maintaining vascular homeostasis by regulating endothelial cell function. It has been shown that patients with COVID-19 exhibit endothelial dysfunction, which is associated with a downregulation of KLF2 expression [69]. Endothelial dysfunction is linked to increased secretion of inflammatory cytokines, alterations in pulmonary blood pressure and endothelial junctions, and the development of PF [70,71,72].

Pleural mesothelial cells (PMCs)

MCs, which cover the surfaces of the pleural, pericardial, and peritoneal cavities, have an epithelial-like morphology and functions, although they originate from the embryonic mesoderm layer [73, 74]. The pleural mesothelium is composed of a monolayer of MCs situated on a basement membrane. These cells serve diverse functions, including facilitating frictionless movements between adjacent parietal and visceral surfaces through the creation of a lubricating and non-adherent surface [75, 76]. Additionally, they play roles in transporting fluids and suspended particles within serous cavities, presenting antigens, modulating immune responses, and promoting leukocyte migration. Furthermore, these cells exhibit the ability to synthesize pro-inflammatory cytokines, growth factors, and ECM proteins, thereby contributing to the reparative processes of serosal tissue [74, 77, 78].

In addition to AECs and endothelial cells, emerging evidence suggests that PMCs, through their profibrotic properties, may also contribute to the pathogenesis of IPF [79, 80]. After injury to the lung parenchyma and under the influence of certain stimuli, PMCs can migrate to the lung parenchyma, contribute to fibrotic activities, and thus play a role in the pathogenesis of PF [79, 80]. The presence of Wilms’ tumor 1 (Wt1)-expressing PMCs and calretinin-positive PMCs in the lung parenchyma of patients with IPF has been observed, and their numbers have shown a correlation with the severity of IPF [73, 81, 82]. It has been demonstrated that calpain, a Ca²⁺-dependent cysteine protease, mediates the invasion of PMCs into lung parenchyma, and that calpain inhibitors can reduce pulmonary fibrosis limiting PMC invasion. Both bleomycin and TGF-β1 increased calpain activity in PMCs, leading to increased focal adhesion (FA) turnover, collagen-I synthesis, and cell proliferation [83]. Invading PMCs contribute to the promotion of subpleural pulmonary fibrosis by undergoing MMT and interacting with resident lung fibroblasts [84]. Indeed, conditioned-PMC medium induced the differentiation of lung fibroblasts into myofibroblasts, and conditioned-fibroblast medium promoted collagen-I synthesis and MMT in PMCs. The activation of the TGF-β1/Smad2/3, Wnt/β-catenin, and CD147 (extracellular matrix metalloproteinase inducer) signaling pathways were involved in these interactions between PMCs and lung fibroblasts. Moreover, it has been demonstrated that in a model of subpleural pulmonary fibrosis, bleomycin increases in the permeability of PMC membrane through the activation of VEGF/Src signaling. This is followed by the exacerbation of inflammation in the subpleural region and pleural barrier damage, which ultimately contributes to pulmonary fibrosis [85].

Following SARS-CoV-2 infection, PMCs have been shown to potentially can promote fibrotic reactions and regulate the immune response. SARS-CoV-2 can induce structural changes in the pleura by disrupting the mesothelial monolayer. In severe COVID-19, WT1/cytokeratin-positive cells, i.e. bona fide MCs where demonstrated to infiltrate the sub-mesothelial stroma. An in vitro study showed that MeT5A cells, a cell line of human PMCs, express SARS-CoV-2-specific receptors and co-receptors, including ACE2, TMPRSS2, ADAM17, and NRP1, and after infection produce viral particles. In addition, a wide range of interferons, inflammatory cytokines, and metalloproteases are produced following infection of MeT5A cells with SARS-CoV-2 [76].

Macrophages

Due to their ability to produce a wide array of cytokines and chemokines, macrophages are considered as key driver in the pathogenesis of fibrotic diseases [86]. There are two types of macrophages in the lungs: alveolar macrophages (AMs) including tissue-resident AMs and monocyte-derived AMs, and interstitial macrophages (IMs) [87]. AMs are the first line of defense against harmful infectious agents initiating inflammatory reactions and recruiting neutrophils [88]. Macrophages undergo M1 (classical) or M2 (alternative) activation depending on which stimuli they are exposed. During the early inflammatory stages, M1 macrophages secrete NO and release proinflammatory cytokines and chemokines. M1 macrophages under the control of interferon regulatory factor 5 (IRF-5), with the expression of high levels of inducible nitric oxide synthase (iNOS) and proinflammatory cytokines, generate a persistent and sustained inflammatory response acting a trigger in the initial stages of IPF [49]. If the inflammatory response is unresolved, macrophages acquire an M2 polarization, characterized by phagocytic activity and the expression of scavenging molecules, mannose, and galactose receptors. In this context, M2 macrophages enter the fibrotic areas, where they release TGF-β1 and Platelet-derived growth factor (PDGF), which subsequently trigger the activation of fibroblasts and the proliferation of myofibroblasts [49].

Role of cytokines in IPF and COVID19-associated PF

Infections of AMs by SARS-CoV-2 might induce a cytokine storm, which could result in the dysfunction of lung tissue [89, 90]. The cytokine storm consists in an increased production of pro-inflammatory cytokines such as IL-6, IL-1, IL-17, TNF-α, and TGF-β, as well as chemokines like CXCL1, CXCL2, and CXCL8 [2, 91]. The cytokine storm may induce a general immune system dysfunction with multiple organ damage [50].

Among cytokine produced, TGF-β plays a crucial role in the pathogenesis of all types of fibrosis. It induces fibroblast activation, differentiation into myofibroblasts, synthesis and secretion of ECM components, including proteoglycans, fibronectin, and glycoproteins by myofibroblasts, and expression of profibrotic genes such as SNAIL1, SLUG, and ZEB1. It also in cooperation with other cytokines inhibits the expression of matrix metalloproteinase-1 (MMP-1) in fibroblasts which degrades the ECM [32, 92,93,94]. Following alveolar damage, TGFβ1, along with other potent cytokines such as PDGF and Fibroblast growth factor (FGF), plays a pivotal role in the pathogenesis of PF by stimulating fibroblast migration. Under the effect of EGF, PDGF, TGFβ1, and IL-1, fibroblasts can undergo differentiation into myofibroblasts. In addition to their role in ECM synthesis similar to fibroblasts, myofibroblasts have the ability to induce inflammatory responses through the secretion of cytokines such as IL-1, IL-6, IL-8, and MCP-1. Furthermore, they also produce repair mediators like TGFβ1 and vascular endothelial growth factor (VEGF) that contribute to the process of tissue repair [32]. TGF-β1 plays a significant role in the development of IPF by activating both classical and non-classical pathways. It interacts with various signaling pathways, including the Wnt/β-catenin pathway, contributes to alteration of redox homeostasis, mitochondrial disorder, EMT, PI3K/Akt pathway, and imbalances in the fibrinolytic system [94]. Various studies have demonstrated that the upregulation of various isoforms of IL-17 cytokines, particularly IL-17 A, plays a significant role in the initiation and progression of IPF. IL-17 A is crucial in several key processes that contribute to fibrosis, including tissue regeneration, inflammation, and the induction of EMT [95]. Considering the crucial role of cytokines in fibrosis pathogenesis, they represent promising therapeutic targets for the treatment of fibrotic conditions [94].

Cytokine storm leads to ARDS and is one of the main causes of death in patients with COVID-19 [2, 91]. COVID-19 infection in epithelial cells of the respiratory tract leads to a cytokine storm characterized by the release of pro-inflammatory cytokines, including IL-9, IL-8, IL-6, IL-7, IL-1β, and TNF-α. Concurrently, the secretion of cytokines such as TNF-α, IL-6, and IL-1β plays a pivotal role in the development of COVID-19-induced PF [58]. On the other hand, specific cytokines and extracellular mediators such as INF-γ, IL-7, MMP-7, and Insulin-like Growth Factor-1 (IGF-1) have been found to possess protective effects, since increased levels of these cytokines are correlated with elevated survival rates and decreased lung damage [96].

Role of oxidative stress in IPF and COVID19-associated PF

Due to the high levels of oxygen in lung tissues, they can be predominantly susceptible to oxidative stress [97]. Exogenous agents like environmental pollution, cigarette smoke, radiation and silica particles, along with endogenous oxidants such as mitochondrial reactive oxygen species (ROS), hydrogen peroxide, superoxide anions, and NO, play a significant role in targeting AECs, pulmonary vascular endothelial cells, and lung macrophages, leading to the generation of ROS and reactive nitrogen species (RNS) [88]. It has been suggested that the oxidant/antioxidant imbalance in the lower respiratory tract and subsequent oxidative stress-induced AECs apoptosis are crucial for the development of IPF [98]. ROS not only can play an important role in the destruction of the alveolar epithelium, but also change the expression of immunoinflammatory mediators and profibrotic genes to cause the differentiation of fibroblasts into myofibroblasts [99]. In addition to lymphocytes, macrophages, and neutrophils in the lung tissues, fibroblasts and myofibroblasts are the most important sources of ROS/ RNS production [100, 101].

ROS-producing nicotinamide adenine dinucleotide phosphate (NADPH) oxidase (NOX) mediates different functions in pathological repair processes in IPF, including myofibroblast differentiation, production and accumulation of ECM, development of fibroblastic foci, and induction of profibrotic gene expression [99,100,101]. Multiple factors could be effective in PF caused by SARS-CoV-2, such as stimulation of immune response, secretion of inflammatory cytokines (IL-6, IL-8, TGF-β, TNF-α), increased neutrophil infiltration, and increased oxidative stress [102, 103]. COVID-19 can be considered a condition involving redox stress, characterized by an imbalance between oxidants and antioxidants [104]. It has been shown that disruption of mitochondrial homeostasis is one of the complications of COVID-19, which underlies oxidative stress, inflammation, and damage to lung tissue [105]. In COVID-19 patients, the underlying cause of oxidative stress and excessive inflammation can be attributed to suppression of nuclear factor erythroid-related factor 2 (Nrf2), a key transcriptional regulator of cellular antioxidant and anti-inflammatory responses [106]. Lung biopsies obtained from individuals affected by COVID-19 have demonstrated a notable suppression of genes associated with the Nrf2 signaling pathway. This highlights a potential link between Nrf2 signaling and the pathogenesis of COVID-19 [107, 108]. In vivo and in vitro studies show that Nrf2 activators with strong anti-fibrotic effects can significantly protect against PF [109]. These activators have also been suggested in the management of the COVID-19 disease [106].

PF and the EMT process

EMT

EMT is a dynamic biological process in which epithelial cells lose cell-cell adhesion and apical-basal polarity, alter their shape from a cuboidal structure to a spindle-cell shape with dramatic remodeling of the cytoskeleton, and acquire some mesenchymal characteristics, including invasion, migration, and production of ECM [75, 110,111,112,113]. EMT occurs during various physiological and pathological conditions, including normal embryonic development and tissue regeneration, wound healing, organ fibrosis, and tumor progression [114, 115]. EMT has been demonstrated as one of the key cellular mechanisms contributing to fibrogenesis in different organs, including PF [57]. EMT is classified into three functional categories [116, 117]. Type I EMT is occurs during embryogenesis. Type II is related to wound healing, tissue regeneration, and fibrosis, while type III is a mechanism of tumor progression [118, 119]. EMT associated with lung fibrosis is categorized as type II EMT. Several stimuli have been shown to induce EMT in PF, including epithelial injury, cigarette smoking, latent membrane protein 1 (LMP1) of Epstein Barr virus, Epidermal growth factor receptor (EGFR) signaling, and TGF-β. This last cytokine is considered the key inducer of EMT, acting through several intracellular pathways including SMADs, Wnt/β-catenin and Ras/ERK/MAPK, in cross talk with endoplasmic reticulum (ER) stress and unfolded protein response (UPR) [54, 120,121,122]. It has been suggested that EMT is involved in the pathogenesis of IPF. However, studies on its underlying mechanism show different results [9]. In vitro studies have shown the induction of EMT in human AECII in the presence of TGF-β, which can be produced as a result of injury by activated AECs [123]. In biopsies from lung tissues of patients with IPF, the observation of epithelial cells with mesenchymal characteristics confirms the occurrence of EMT [124]. Studies using in vivo models of PF, especially the bleomycin -induced murine model, are conflicting regarding the importance of EMT in the formation of fibroblast populations. Some in vivo studies suggest that AECII undergo EMT, but other lineage-tracing studies report a relatively small contribution of epithelial cells in fibroblast populations through the EMT process [9, 125, 126]. There is further evidence showing that α-SMA, as a key marker for myofibroblasts, does not colocalize with EMT-derived cells, which may indicate that AECs have undergone a partial EMT without transitioning to a myofibroblastic phenotype [125, 127]. It has been shown that AECII having undergone a RAS-induced EMT can express ECM-related genes at low levels, while conditioned media obtained from activated AECII had the ability to activate fibroblasts [128]. Therefore, these findings suggest that AECs do not directly contribute to the generation of myofibroblast populations. Instead, AECs undergoing EMT can create a profibrogenic condition through dysregulated paracrine signaling directed from AECs to underlying fibroblasts. This can ultimately lead to local fibroblast activation and myofibroblast differentiation by secreted factors, a process controlled by continuous ZEB1-mediated expression of tissue plasminogen activator (tPA) in AECII [9, 125, 128, 129]. It is important to note that the murine model of bleomycin-induced PF, does not fully reproduce human IPF due to several limitations including the induction of reversible fibrosis, failure to mimic the histological appearance, and to reproduce the progressive nature of IPF. Therefore, more accurate models are needed to better reproduce the human IPF [9, 10, 130].

EndMT

EndMT is an EMT-like process characterized by the loss of endothelial-specific markers such as cluster of differentiation 31 (CD31) and vascular endothelial cadherin (VE-cadherin), which are junctional endothelial proteins. Cells undergoing EndMT also acquire an elongated phenotype with a front-rear polarity and concurrently gain the expression of specific markers of mesenchymal cells such as α-SMA, N-cadherin, and fibroblast-specific protein 1 (FSP1)/ S100A4 [10, 131, 132]. Similarly to EMT, postnatal EndMT is plays a role in several human diseases, including vascular disorders, cancer, and fibrotic diseases [133]. The occurrence of EndMT has been demonstrated in in vitro and in vivo models of fibrosis, as well as in lung tissue samples from patients with IPF, suggesting a role of this process in the pathogenesis of IPF [132]. For example, it has been reported that the co-localization of CD31 and α-SMA is significantly increased in mouse models and lung tissue of patients with radiation-induced PF compared to control [134]. In the arterial samples from IPF patients, increased expression of mesenchymal markers including N-cadherin, S100A4/FSP1, and vimentin compared to control indicates the dedifferentiation of resident endothelial cells into mesenchymal cells [135]. A lineage-tracing mouse model of bleomycin-induced PF indicated that a significant number of fibroblasts derive from lung capillary endothelial cells [136]. Hypoxia, oxidative stress, inflammation, endothelial cell senescence, injured endothelial cells, and impaired shear stress may have been implicated in the induction of the EndMT process [66, 137]. TGF-β is considered a key regulator of the EndMT process, acting through both canonical and non-canonical intracellular signalling [138, 139].

TGF-β can lead to increased expression of mesenchymal markers through the overexpression of EMT-specific transcription factors, including Snail, Slug, and Twist1 [138, 140]. Several sources for TGF-β have been identified, including circulating inflammatory cells adhering to the endothelium or infiltrating the sub-endothelial space, other inflammatory cells located in the perivascular environment, or latent TGF-β released by mesenchymal cells [141]. Besides TGF-β also proinflammatory cytokines, including IL-1β, and TNF-α, can potently induce EndMT [4].

MMT

It has been demonstrated that inflammatory and infectious stimuli may promote a complex multistep process known as MMT [142]. MMT is a specific type of EMT process in which MCs progressively lose their epithelial-like properties and gain the ability to invade the sub-mesothelial stroma and develop proinflammatory and profibrotic mesenchymal-like features [143].

It is believed that the contribution of PMCs to the pathogenesis of IPF may occur through an MMT process [57, 79]. TGF-β is considered the key player in this process since it can induce the migration of PMCs to the lung parenchyma and their dedifferentiation into myofibroblasts through the Smad2 signaling pathway [79, 144]. The abundance of mesenchymal cells of mesothelial origin has been shown to correlate with the extent of fibrotic changes in a murine model of IPF [81]. Such results indicate PMCs as a source of myofibroblasts and, therefore, their involvement in the pathogenesis of IPF (Fig. 2) [144].

Fig. 2

Role of EMT process and its cell specific variants; EndMT and MMT in pulmonary fibrosis. Mesenchymal transition of pleural MCs and endothelial cells of pulmonary capillaries, induced by specific factors (MMT and EndMT processes, respectively), can provide new sources of fibroblasts/myofibroblasts and extracellular matrix production in fibrotic conditions. Moreover, AECII undergo EMT in response to specific inducing factors, mainly through paracrine effects and the release of secretory factors may induce local fibroblasts activation and extracellular matrix deposition. EMT; Epithelial-mesenchymal transition, MMT; Mesothelial-to-mesenchymal transition, EndMT; Endothelial-to-mesenchymal transition, AECI and AECII; type I and type II alveolar epithelial cells, TGF-β; Transforming growth factor-β, TNF-α; Tumor necrosis factor-α, IL-1β; Interleukin-1 beta, SMADs; suppressor of mothers against decapentaplegic, MAPKs; mitogen-activated protein kinases, PI3K/AKT; Phosphatidylinositol 3-kinase /protein kinase B, NF-κB; Nuclear factor kappa B, TAK1; Transforming growth factor-β-activated kinase-1

Impact of dedifferentiation processes of epithelial cells, endothelial cells and PMCs in COVID-19-associated PF

PF is one of the potential consequences of COVID-19 disease [144], and the EMT process has been identified as a crucial driver of the fibrotic condition [145]. It has been demonstrated that ACE2 is abundantly expressed in lung epithelial cells and endothelial cells of blood vessels [145], and these cells may play an important role in fibrosis after COVID-19 disease [146].

Experimental evidence indicates that viral infections, including cytomegalovirus and SARS-CoV-2, are capable of inducing substantial lung injury by harming endothelial and epithelial cells, thus initiating inflammatory and fibrotic responses [147, 148]. Recently, evidence has revealed the role of the EMT process in SARS-CoV-2-induced PF. Transcriptome profiling of single-cell RNA identified early and late subgroups of EndMT and EMT subpopulation and revealed that several genes associated with the mesenchymal state are expressed both in COVID-19 patients and IPF patients, suggesting potential targets that could reverse the myofibroblastic conversion in both diseases [149]. In vitro studies and analysis of samples from COVID-19 patients have demonstrated that SARS-CoV-2 infection can trigger the expression of genes associated with EMT including ZEB-1, Slug, anexelekto (AXL) and vimentin, and decreased expression of genes encoding for epithelial junction proteins [150,151,152,153]. Furthermore, it has been observed that upregulation of ZEB1 and its binding to the ACE2 promoter lead to the inhibition of ACE2 expression [150]. As previously mentioned, inhibition of ACE2 expression is associated with increased fibrosis [14].

On the other hand, it has been demonstrated that only a small subset of AECII could express ACE2, and weak ACE2 expression has been observed in other epithelial cells [154, 155]. In a study performed on lung epithelial cells (BEAS-2B) and lung fibroblasts (MRC-5), it was shown that treatment with SARS-CoV-2 spike proteins led to production of inflammatory cytokines and the induction of EMT through a mechanism linked to the Growth Arrest and DNA-damage-inducible alpha [39].

Pandolfi et al. reported the presence of neutrophil extracellular traps (NETs) in the bronchoalveolar lavage fluid of patients with severe COVID-19 infection, which were associated with EMT induction. Furthermore, this group demonstrated that NETosis, as well as the secretion of inflammatory mediators such as TGF-β, IL-8, and IL-1 by AMs, were required for the complete activation of the EMT expression pattern [151].

An increase in the expression of TGF-β has been reported in patients with COVID-19 [58, 156]. The elevated oxidative stress in epithelial cells leads to an increase in the production and release of TGF-β [157, 158]. Additionally, in response to the engulfment and digestion of apoptotic alveolar cells induced by COVID-19, macrophages infiltrate the lungs and release more TGF-β. By inducing the secretion of IL-4, IL-6, and IL-13, TGF-β stimulates AMs, promoting the development and progression of PF [103, 159]. Following the activation of TGF-β, its downstream signaling factors participate in promoting EMT. For example, the activation of the protein kinase B (AKT) pathway induced by TGF-β leads to an increased expression of β-catenin and collagen type II alpha 1 chain (COL2A1) [58, 160].

One of the main known mechanisms for Nrf2-induced protection against PF is the inhibition of EMT. Nrf2 signaling can affect the development of EMT by targeting specific pathways, such as suppressing the expression high mobility group box 1 (HMGB1), an EMT-promoting transcription factor-like protein. Moreover, Nrf2 signaling can decrease the expression of Snail, and control the abnormal expression of Numb, a phosphotyrosine binding domain (PTB) protein associated with the EMT process [109, 161, 162]. Downregulation of genes related to the Nrf2 pathway has been reported in patients with COVID-19, thus suggesting a functional role of this transcription factor in inhibiting PF caused by COVID-19.

The importance of endothelial cell damage in the initiation of PF after COVID-19 infection has been previously mentioned [146]. The relationship between the significant increase of ACE2-positive endothelial cells and EndMT-related changes in endothelial cells has been reported [146, 163]. Studies suggest the possible involvement of EndMT in COVID-19-induced lung fibrosis and pulmonary artery hypertension [70]. A study utilizing histochemistry and immuno-histochemistry on autopsy samples from lung patients who died from COVID-19 demonstrated that EndMT may play a central role in the development of lung fibrotic lesions in COVID-19 [164].

Although studies have reported the possible involvement of PMCs in SARS-CoV-2 pathology, a direct role of the MMT process in PF caused by COVID-19 has not been demonstrated so far. However, further research is needed to better understand how mesenchymal transition processes from the three mentioned cellular sources contribute to lung fibrosis caused by COVID-19.

EMT inhibition as a rationale of therapy for IPF and COVID-19-induced PF

Evidence shows that the EMT process is a primary driver and pathological mechanism in fibrotic conditions such as IPF and COVID-19-induced PF. Thus, halting or reversing the EMT process could be a valuable clinical strategy [165,166,167]. Given the pathophysiological similarities between IPF and COVID-19-related PF, a similar response to treatments may be hypothesized [49, 51, 168]. The principal aim of EMT-related antifibrotic treatments is to block the factors that trigger this process [165]. Below, we outline some of these strategies explored in studies.

In fibrotic diseases, activation of TGF-β signaling is a well-known inducer of EMT, stimulating both Smad and non-Smad downstream pathways. Consequently, targeting TGF-β signaling may present a promising therapeutic focus for fibrosis [169]. Various strategies have been investigated to block TGF-β signal transduction at different points, including: (1) Inhibition of TGF-β mRNA expression using antisense oligonucleotides; (2) Neutralizing antibodies against TGF-β; (3) Synthetic peptides that occupy the TGF-β receptor binding site or disrupt its interactions; (4) Inhibitors of TGF-β receptor kinase activity; (5) Factors that inhibit the phosphorylation or nuclear translocation of Smad2/3, the key mediators of the TGF-β signaling cascade; and (6) Upregulation of Smad7, a natural inhibitor of TGF-β signaling [57, 165]. However, it is essential to consider that TGF-β is a pleiotropic growth factor with numerous biological functions. Thus, inhibiting TGF-β signaling can impact not only EMT-induced fibrosis but also other TGF-β-mediated activities, potentially resulting in systemic effects [170]. Therefore, the direct targeting of specific biological functions of TGF-β seems to be effective. In this regard, cytokines such as hepatocyte growth factor (HGF), bone morphogenetic proteins (BMP-7, BMP-2), VEGF, and FGF-1 have shown anti-fibrotic effects by inhibiting or reversing TGF-β-induced EMT. Additionally, certain antifibrotic microRNAs, including miR-200c and miR-769-5p, have been found to inhibit the EMT process in endothelium and mesothelium via TGF-β-dependent mechanisms [171,172,173].

Given that several factors—such as cytokines, inflammation, and oxidative stress—can mediate EMT induction by affecting the cellular microenvironment, strategies targeting multiple inducing factors seem more advantageous [165]. Recent studies suggest that therapies utilizing stem cells and their derivatives have potential in restoring the native cellular microenvironment, presenting a promising approach for treating fibrotic diseases. Mesenchymal stem cells (MSCs), for instance, can exert anti-EMT effects in fibrosis through two primary mechanisms: the secretion of HGF and BMP-7 to modulate TGF-β signaling, and the release of immunomodulatory factors that influence microenvironmental inflammation—a key factor in EMT occurrence—by counteracting inflammatory cytokines like TNF-α and IL-1 [57, 165, 174, 175].

Moreover, therapies based on the regulation of epigenetic changes are also being explored in fibrosis treatment, such as using histone deacetylase (HDAC) inhibitors, which can inhibit TGF-β-induced EMT and alleviate fibrosis [173, 176,177,178]. Interestingly, HDAC1-3 specific inhibition favoured increased SARS-CoV-2 infection in MCs which was linked to increased ACE2 expression and dampened IFN response, warning for the use of HDAC inhibitors in SARS-CoV-2 patients [179, 180].

Considering oxidative stress’s role in activating pathological processes related to fibrosis—such as EMT—and the reciprocal regulation of ROS and TGF-β signaling, antioxidant treatments appear promising [100, 181]. Certain natural flavonoids and polyphenols, such as procyanidins and proanthocyanidins, have demonstrated antioxidant and anti-inflammatory properties that can reverse the EMT process [165, 182].

Up to now, because PF is irreversible [183], drug therapy is not expected to provide a complete cure. However, it may help relieve symptoms and slow the decline of lung function [184]. Some effective compounds in the treatment of IPF, along with their potential use in COVID-19-induced PF at different stages, include: EW-7197 (preclinical phase), which blocks the catalytic activity of the TGF-β type I receptor and inhibits TGF-β/Smad2/3 and ROS signaling; GSK3335103 (In vitro and preclinical phase), a new inhibitor of integrin αvβ6 that can attenuate TGF-β signaling; and ZSP1603 (preclinical phase), which blocks PDGF-Rβ and ERK and reduce the expression of TGF-β1 [168]. Pirfenidone and nintedanib are FDA-approved antifibrotic drugs for the treatment of IPF, with favorable results also observed in PF associated with COVID-19 [185]. The inhibitory effect of pirfenidone on TGF-β-induced EMT has been reported in studies [186, 187].

Conclusion

Dedifferentiation processes such as EMT, EndMT and MMT are cellular mechanisms potentially involved in the pathogenesis of several fibrotic diseases. Although the three cell types originating these processes have different functions and cellular specificities, the mechanisms of dedifferentiation are largely overlapping, being based on the local production of proinflammatory /profibrotic extracellular mediators, first of all TGFβ, and the induction of EMT-TFs such as Snail, Slug and ZEB1 acting as master genes of a wide cell reprogramming events eventually leading to a myofibroblastic conversion.

Data from in vitro studies, murine models and clinical samples suggest that these processes play a role in the pathogenesis of PF during both IPF and COVID-19. However, few studies have demonstrated a causal link between EMT/EndMT/MMT and the development of PF. Therefore, more research in this field is required to shed light on the molecular mechanisms underlying and to provide new effective therapeutic strategies to promote EMT reversal and to block the progression of the fibrotic damage.