Key role of transcription factors network in proliferative vitreoretinal diseases development Authors: Clément Duveau, Yosra Raiss El Harrak, Azine Datlibagi, Jason Perret, François Willermain, Christine Delporte, Elie Motulsky Published: 2026 ABSTRACT AbstractProliferative vitreoretinal diseases (PVDs) encompass severe ocular disorders such as proliferative vitreoretinopathy (PVR), proliferative diabetic retinopathy (PDR), and epiretinal membranes (ERM), characterized by the formation of fibrovascular membranes that often lead to retinal detachment and vision loss. A central mechanism driving these conditions is the epithelial-to-mesenchymal transition (EMT) of retinal pigment epithelial (RPE) cells, orchestrated by a network of transcription factors (TFs). Among these, zinc finger E-box binding homeobox 1 (ZEB1) emerges as a pivotal regulator by repressing epithelial markers like E-cadherin and inducing mesenchymal markers such as N-cadherin and vimentin, thereby promoting cell migration and fibrotic membrane formation. nuclear factor of activated T cells 5 (NFAT5) contributes to this process by mediating osmotic stress responses and upregulating inflammatory cytokines, which further act upon EMT and fibrosis. activator protein 1 (AP-1) and hypoxia inducible factor 1 subunit alpha (HIF-1α) participate in driving inflammation, extracellular matrix (ECM) remodeling, and angiogenesis. While HIF-1α triggers vascular endothelial growth factor (VEGF) expression under hypoxic conditions, AP-1 modulates matrix metalloproteinases (MMPs) essential for ECM degradation and remodeling. Additional TFs, including Kruppel-like factor 4 (KLF4) and microphthalmia-associated transcription factor (MITF), are vital in maintaining RPE cell identity. Their downregulation under pathological conditions disrupts epithelial integrity and predisposes cells to undergo EMT. Moreover, β-catenin, through its role in the wingless-related integration site (Wnt) signaling pathway, reinforces EMT and ECM remodeling, further enhancing fibrotic progression. Adipocyte enhancer-binding protein 1 (AEBP1) and ZFP36 ring finger protein like 1 (ZFP36L1) also regulate inflammatory responses and ECM dynamics, providing novel post-transcriptional targets for therapeutic intervention. Overall, the synergistic interactions among these TFs create complex feedback loops that amplify pathological changes in PVDs. Targeting these molecular pathways offers promising avenues for developing multi-targeted therapies aimed at saving vision-threatening disease while reducing invasive surgical interventions.Keywords: Review, Proliferative vitreoretinal diseases, Retina, Transcription factors, Molecular biology Introduction Proliferative vitreoretinal diseases (PVDs), including proliferative vitreoretinopathy (PVR), proliferative diabetic retinopathy (PDR), and epiretinal membranes (ERM), represent a group of severe ocular disorders characterized by the formation of fibrovascular or avascular membranes on or beneath the retinal surface. These conditions frequently lead to complications such as retinal detachment, fibrosis, and significant vision loss. PVR is a particularly challenging pathology that develops in response to retinal detachment, trauma or surgery, involving the formation of contractile membranes that distort the retinal architecture. Surgical interventions for PVR often face high failure rates due to the aggressive fibrotic responses that ensue [1]. In PDR, chronic hyperglycemia associated with diabetes leads to vascular damage, ischemia and the release of angiogenic factors such as VEGF which promote fibrovascular membrane formation. The angiogenesis-driven nature of PDR may ultimately lead to exudation, neovascularization, retinal hemorrhages and ultimately to tractional retinal detachment being responsible of severe visual loss [2]. Although ERMs are less aggressive than PVR and PDR, they nonetheless contribute to visual distortion and retinal wrinkling, often following retinal injury or surgical intervention [3]. Another related condition, neovascular age-related macular degeneration (nAMD), shares key pathogenic pathways with PVDs, particularly the development of fibrovascular membranes driven by chronic inflammation and oxidative stress [4]. A unifying feature of these diseases is the epithelial-mesenchymal transition (EMT) of retinal pigment epithelial (RPE) cells. During EMT, RPE cells lose their apical-basal polarity, and are characterized by a downregulation of epithelial markers such as E-cadherin and an upregulation of mesenchymal markers such as N-cadherin and vimentin, thereby acquiring motility and invasive abilities. This process is orchestrated by some TFs such as ZEB1, HIF-1α, NFAT5, AP-1, β-catenin, among others. These TFs regulate critical cellular behaviors, including proliferation, migration, extracellular matrix (ECM) remodeling and inflammatory responses, making them central players in the pathogenesis of PVDs [5–8]. The limitations of current treatment which include the invasive nature of surgical membrane peeling and its associated risks, have prompted researchers to explore alternative therapeutic strategies targeting the molecular drivers of PVDs. Understanding the roles and interactions of the TFs involved in PVDs could pave the way for innovative treatments that disrupt pathological processes at their source, thereby reducing the need for surgical intervention and improving patient outcomes. In this review, we focus on nine transcription factors (ZEB1, NFAT5, AP-1, HIF-1α, KLF4, MITF, β-catenin, AEBP1, and ZFP36L1) based on their recurrent involvement in key pathological processes underlying PVDs, namely EMT, inflammation, fibrosis, ECM remodeling, and angiogenesis. These TFs were selected after a comprehensive analysis of current literature identifying their roles in regulating RPE cell phenotype, and/or their interactions with signaling pathways dysregulated in PVDs, such as transforming growth factor beta (TGF-β), Wnt/β-catenin, and nuclear factor-kappa B (NF-κB). The selection was based on a PubMed literature search using combinations of keywords including “transcription factor”, “RPE”, “epithelial-mesenchymal transition”, “retinal fibrosis” and specific disease terms such as “PVR”, “PDR”, and “epiretinal membrane”. Transcription factors were retained if they were repeatedly identified in original studies or reviews as modulators of key PVD-related processes, mechanistically linked to RPE cell transformation, and supported by experimental or transcriptomic evidence. Importantly, these TFs represent a diverse set of molecular regulators with some acting as drivers of EMT and fibrosis (e.g., ZEB1, AEBP1, β-catenin), serving as as stress-responsive mediators of inflammation and hypoxia (e.g., HIF-1α, NFAT5, AP-1), or preserving epithelial identity but downregulated during disease progression (e.g., MITF, KLF4). In addition, ZFP36L1 was included as a post-transcriptional regulator that modulates TF activity and cytokine mRNA stability, adding another regulatory layer relevant to the chronic inflammatory microenvironment of PVDs. Unlike previous reviews that generally address a single TFs in a particular pathway, our work provides an integrative and network-level perspective of TFs in PVDs. We specially highlight TFs interactions, feedback loops and upstreams regulators that converge to drive EMT and fibrosis in RPE. In particular, we emphasize the role of novel or underexplored regulators (such as NFAT5 and ZFP36L1) in retinal inflammation and post-transcriptional control. By synthesizing TF-centered mechanisms across molecular, cellular, and clinical layers, this paper aims to provide a comprehensive overview of TFs involved in the pathogenesis of PVDs and their signaling pathways mechanisms. These insights could potentially lead to the development of therapies that save vision without the need of surgery. Major transcription factors involved in PVDs ZEB1: a master regulator of EMT and fibrosis in PVDs Zinc finger E-box-binding homeobox 1 (ZEB1) is a critical TF in the pathogenesis of PVDs. As a central regulator of EMT, ZEB1 drives key processes such as cellular dedifferentiation, migration, invasion and ECM remodeling, making it a cornerstone in the development of fibrotic and fibrovascular complications associated with PVDs [9]. ZEB1 has been particularly implicated in PVDs, where its induction promotes RPE migration and fibrotic membrane formation, as demonstrated in both human epiretinal membranes and murine model of PVR [9–11]. Furthermore, ZEB1 expression has been shown to be upregulated and likely contributing to RPE dedifferentiation in an in vitro model of PVD based on ARPE19 cells (spontaneously arising retinal pigment epithelia cell line derived from the normal eyes of a 19-year-old male) knockdown for apoptosis-stimulating proteins of p53 (ASPP2) [12]. Role in EMT ZEB1 promotes EMT by binding to E-box elements in the promoters of target genes, directly silencing epithelial markers (such as E-cadherin) and enhancing mesenchymal markers (like N-cadherin and vimentin) gene expression. The E-cadherin-to-N-cadherin switch is a hallmark of EMT, conferring motility and invasiveness to RPE cells. Indeed, in PVDs, the loss of E-cadherin disrupts cell-cell adhesion, facilitating RPE cell detachment and migration, which are critical steps in the formation of fibrotic membranes, particularly in PVR [10]. Crosstalk with other TFs In addition to repressing epithelial markers, ZEB1 downregulates the expression of TFs critical for maintaining RPE identity, such as microphthalmia-associated transcription factor (MITF) and Orthodenticle Homeobox 2 (OTX2) [11]. Both MITF and OTX2 are essential for preserving the epithelial phenotype of RPE cells and suppressing mesenchymal characteristics. The loss of these TFs expression in response to ZEB1 activity leads to a shift toward a proliferative, migratory phenotype. ZEB1 also inhibits the expression of p21, a cyclin-dependent kinase inhibitor, thereby promoting RPE cell proliferation, a key feature in the early stages of PVR [11]. Regulation by upstream signals ZEB1 expression is regulated by several upstream signaling pathways that are frequently dysregulated in PVDs. TGF-β signaling is one of the primary inducers of ZEB1 expression, linking inflammation and fibrosis. TGF-β, a well-known driver of fibrosis, induces ZEB1 expression, which in turn represses anti-fibrotic signals and promotes ECM production [10]. Furthermore, hypoxia and oxidative stress, common features in PVDs such as PDR, stabilize HIF-1α, which induces ZEB1 expression. ZEB1 has been reported to intersect with β-catenin signaling, thereby amplifying EMT and promoting fibrotic remodeling [13]. This interplay positions ZEB1 as a critical integrator of hypoxic and inflammatory signals in the retinal microenvironment, creating a feedback loop that sustains EMT. Inflammatory cytokines such as interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α), which are elevated in PVR and PDR, also enhance ZEB1 expression through NF-κB and Smad-dependent pathways [14]. NFAT5: a regulator of osmotic and inflammatory stress in PVDs Nuclear Factor of Activated T-Cells 5 (NFAT5), also known as tonicity-responsive enhancer binding protein (TonEBP), is a TF playing a critical role in cellular responses to osmotic stress and inflammation [15]. The role of NFAT5 has been predominantly studied in the context of hyperosmotic stress in kidney cells. However, recent research highlights its significant involvement in the osmoadaptative response of RPE to hyperosmotic stress in diabetic retinal models [16, 17], thereby suggesting its implication in PVDs, such as PDR. NFAT5 may contribute to the development of fibrotic and fibrovascular membranes through its regulation of EMT, inflammatory pathways and ECM remodeling [18]. Role in hyperosmotic stress NFAT5 is uniquely adapted to respond to osmotic stress, a condition characterized by imbalances in intracellular and extracellular fluid tonicity. Hyperosmotic stress is a hallmark of diabetic retinopathy and PDR, where high glucose levels disrupt the retinal microenvironment [19]. NFAT5 is activated under these conditions through the Mitogen-Activated Protein Kinase (MAPK) and Extracellular Signal-Regulated Kinase 1/2 (ERK1/2) signaling pathways, leading to the transcription of osmoprotective genes that help cells adapt to osmotic stress [16, 18]. However, in the pathological context of PVDs, this NFAT5 activation is maladaptive and may enhance cellular proliferation and migration, therefore promoting the formation of fibrotic membranes. Regulation of EMT and fibrosis NFAT5 activation leads to the upregulation of mesenchymal markers such as vimentin and fibronectin, while repressing the epithelial marker E-cadherin, thereby promoting a migratory and invasive phenotype [20]. In RPE cells, this EMT-driven phenotype could contribute to the detachment, migration, and invasion of RPE cells, culminating in the formation of contractile membranes that distort the retina, as seen in PVR [21]. Under pathological conditions, NFAT5 interacts with other EMT-regulating pathways, such as the TGF-β signaling cascade, to enhance fibrotic responses. For instance, NFAT5 induces the expression of connective tissue growth factor (CTGF), a downstream effector of TGF-β, which promotes collagen deposition and ECM stiffening [17]. These changes could be critical in the development of fibrotic membranes in PVR. Additionally, NFAT5 influences the hypoxic responses of RPE cells by upregulating HIF-1α expression and promoting VEGF expression [18, 22]. This interaction could create a feedback loop where hypoxia and osmotic stress amplify EMT and fibrosis. Inflammatory amplification In addition to its role in EMT, NFAT5 acts as a mediator of inflammation. Under hyperosmotic conditions, NFAT5 upregulates pro-inflammatory cytokines, including TNF-α and interleukin-1 beta (IL-1β) in corneal cells [23]. These cytokines contribute to chronic inflammation in the retinal microenvironment, which drives angiogenesis, ECM remodeling, and RPE cell proliferation [17]. Furthermore, the combined activation of NFAT5 and HIF-1α described earlier enhances the production of ECM components and pro-fibrotic cytokines [18], and may exacerbate tissue remodeling in the RPE. Moreover, NFAT5 also interacts with other inflammatory TFs, such as AP-1 and NF-κB, amplifying inflammatory signaling cascades [24]. These interactions could be crucial in the pathogenesis of PVDs. In addition, chronic inflammation sustained by NFAT5 activity creates a pathological environment conducive to fibrovascular membrane formation in PDR [25]. The elevated cytokine levels recruit immune cells and promote vascular leakage, further aggravating retinal ischemia and hypoxia [17]. This perpetuates the activation of NFAT5, HIF-1α, and AP-1, creating a self-reinforcing loop that accelerates disease progression. AP-1: a central regulator of inflammation, ECM remodeling and EMT in PVDs Activator protein-1 (AP-1) is a TF composed of proteins from the Jun, Fos and ATF families that regulates cellular responses to stress, inflammation and tissue remodeling [26]. AP-1 plays a pivotal role in driving EMT, ECM remodeling and inflammatory responses [27]. By integrating multiple stress signals, AP-1 orchestrates pathological processes that contribute to the progression of these vision-threatening diseases. The functional role of AP-1 in retinal fibrosis has been confirmed in vivo, where antisense oligonucleotides targeting c-fos mRNA significantly reduced RPE proliferation and the progression of PVR in rabbit [28]. Role in inflammation AP-1 is a key mediator of inflammation, activated in response to oxidative stress, cytokines and growth factors. Upon activation, AP-1 upregulates the transcription of pro-inflammatory cytokines which are elevated in PVR and PDR [29]. These cytokines amplify retinal inflammation, recruit immune cells, and create a pathological microenvironment that exacerbates disease progression. Furthermore, reactive oxygen species (ROS) generated under diabetic or hypoxic conditions stimulate upstream kinases, such as c-Jun N-terminal kinase (JNK) and p38 MAPK, leading to the phosphorylation and activation of AP-1 [30]. Once activated, AP-1 drives the expression of genes involved in the inflammatory response, perpetuating a cycle of oxidative damage and inflammation that could contribute to retinal ischemia and fibrosis. Regulation of ECM remodeling AP-1 also plays a central role in ECM remodeling, a key feature of PDR, by regulating the expression of matrix metalloproteinases (MMPs), particularly MMP-9 [29, 31], facilitating the degradation and remodeling of the ECM. This process is essential for the migration and invasion of RPE cells that underlies the formation of fibrotic membranes in PVDs. In fibrotic contexts, AP-1 has been reported to upregulate fibronectin and collagen expression in hepatic fibrosis models [32]. Such process may be conserved in RPE-mediated ECM remodeling but will require further experimental confirmation. The dual role of AP-1 in ECM degradation and synthesis highlights its dynamic involvement in both the initial stages of membrane formation and the progression to fibrosis. Role in EMT AP-1 is a critical driver of EMT in RPE cells, a central process to the pathogenesis of PDR [33]. During EMT, AP-1 upregulates mesenchymal markers, such as vimentin and N-cadherin, while suppressing epithelial markers like E-cadherin. This reprogramming enhances the motility and invasiveness of RPE cells, enabling them to migrate and contribute to fibrotic membrane formation [29]. The activation of AP-1 during EMT is mediated by TGF-β and MAPK. TGF-β induces expression of c-Fos and c-Jun subunits in RPE cells, while MAPK pathways, including ERK1/2 and JNK, activate AP-1 through phosphorylation, enhancing their transcriptional activity [34]. The findings corroborate those found in the nucleus pulposus of the intervertebral disk [35]. These interactions place AP-1 at the center of EMT regulation in PVDs. Crosstalk with other TFs AP-1 interacts with several other TFs to amplify its pathological effects. One significant interaction is with β-catenin, a key player in the Wnt signaling pathway. β-catenin enhances AP-1 activity by stabilizing its binding to DNA, leading to the upregulation of genes involved in EMT and angiogenesis in cancer [36, 37]. This synergy between AP-1 and β-catenin may drive RPE cell migration and vascular remodeling, contributing to the formation of fibrovascular membranes in PDR. AP-1 also interacts with NFAT5, a TF activated under osmotic stress, as described above. Together, AP-1 and NFAT5 may regulate the expression of inflammatory and ECM-related genes, creating a feedback loop that sustain chronic inflammation and fibrosis in the retina [24]. Additionally, AP-1 collaborates with HIF-1α under hypoxic conditions to enhance the transcription of VEGF and other angiogenic factors [38]. This process could promote neovascularization in PDR and exacerbate retinal tissue damage. HIF-1α: a master regulator of hypoxia and angiogenesis in PVDs Hypoxia-inducible factor 1 (HIF-1α) is a TF that serves as a key mediator of cellular responses to hypoxia. Its activation under low oxygen conditions is central to the pathogenesis of PVDs, including PDR, PVR and nAMD. Indeed, HIF-1α orchestrates a broad range of processes, including angiogenesis, fibrosis, and inflammation, by regulating the expression of target genes such as VEGF, CTGF and TGF-β [39–42]. In addition, a recent study using a HIF inhibitor (32-134D) showed that pharmacological blockade of HIF-1α reduces retinal vascular leakage, inflammation, and fibrosis in diabetic eye disease mouse model, confirming the pathogenic role of HIF signaling in vivo [43]. Role in hypoxia-induced angiogenesis Under hypoxic conditions, HIF-1α stabilizes and translocates to the nucleus, where it forms a heterodimer with HIF-1β. This complex binds to hypoxia-responsive elements (HREs) in the promoters of target genes, driving the expression of angiogenic factors such as VEGF which in turn, promotes endothelial cell proliferation, migration, and the formation of new blood vessels, leading to pathological retinal neovascularization [44]. In PDR, chronic hyperglycemia and vascular damage create hypoxic microenvironments that activate HIF-1α. The resulting overexpression of VEGF leads to the formation of fragile and leaky blood vessels, contributing to retinal hemorrhages, exudation and tractional retinal detachment [39, 40, 42]. In nAMD, HIF-1α-driven angiogenesis promotes the growth of fibrovascular membranes under the retina, exacerbating retinal damage and vision loss [45]. Contribution to fibrosis In addition to angiogenesis, HIF-1α could play a significant role in fibrosis [46]. Hypoxia-induced activation of HIF-1α upregulates pro-fibrotic factors such as CTGF and TGF-β. These factors stimulate the proliferation of RPE cells and fibroblasts, leading to excessive ECM production and the formation of fibrotic membranes. HIF-1α could also promote the EMT in RPE cells by interacting with EMT regulators like ZEB1 and β-catenin, as shown in some cancers [47, 48]. As explained above, during EMT, RPE cells lose their epithelial characteristics and gain migratory properties, contributing to the formation of contractile membranes that distort the retina in PVR. HIF-1α-induced fibrosis is further amplified by its interactions with inflammatory pathways. For example, HIF-1α upregulates IL-6, which promote chronic inflammation and recruit immune cells to the retinal microenvironment. This inflammatory response exacerbates ECM remodeling and fibrosis, creating a vicious cycle of tissue damage [44]. Regulation of cellular metabolism and survival HIF-1α also regulates cellular metabolism, enabling cells to adapt to hypoxia [49]. By upregulating glycolytic enzymes and downregulating mitochondrial activity, HIF-1α shifts cellular metabolism towards glycolysis, reducing oxygen consumption and preventing oxidative stress-induced cell death. While this adaptation is protective under physiological conditions, it could support in PVDs the survival and proliferation of pathological cell populations, including neovascular endothelial cells and migratory RPE cells [50]. KLF4: a context-dependent modulator in PVDs Kruppel-like factor 4 (KLF4) is a zinc finger TF known for its dual roles in maintaining cellular homeostasis and regulating stress responses [51]. In various context, KLF4 is known to act as a key regulator of epithelial and mesenchymal phenotypes, influencing processes like EMT, fibrosis and inflammation [52] (Fig. 1) but in the context of PVDs, KLF4’s function seems to be complex and context dependent. Involvement of some TFs in the regulation of RPE cell identity AEBP1, adipocyte enhancer-binding protein 1; CRALBP, cellular retinaldehyde-binding protein; CTNNB1, catenin beta 1; ECM, extracellular matrix; EMT, epithelial–mesenchymal transition; HIF-1α, hypoxia-inducible factor 1 alpha; KLF4, Krüppel-like factor 4; MITF, microphthalmia-associated transcription factor; MMP, matrix metalloproteinase; NFAT5, nuclear factor of activated T cells 5; OTX, orthodenticle homeobox; RPE, retinal pigment epithelium; RPE65, retinal pigment epithelium-specific protein 65 kDa; SNAIL, snail family transcriptional repressor 1; SLUG, snail family transcriptional repressor 2; TF, transcription factor; TWIST1, twist family bHLH transcription factor 1; ZEB1, zinc finger E-box-binding homeobox 1; ZFP36L1, zinc finger protein 36 Like 1; ZO-1, zonula occludens-1. Role in maintaining epithelial phenotype Under normal conditions, KLF4 appears to support the epithelial phenotype of RPE cells by promoting epithelial markers such as E-cadherin and suppressing mesenchymal traits. It could help to maintain cellular adhesion, apical-basal polarity and the integrity of epithelial layers, thereby preserving RPE functionality. KLF4 achieves this by directly binding to the promoters of epithelial genes and repressing the expression of TFs involved in EMT, such as Snail family transcriptional repressor 1 (Snail; SNAI1), ZEB1 and Twist family BHLH transcription factor 1 (Twist1) [53]. This protective role of KLF4 is essential for preventing pathological EMT in RPE cells. Loss of KLF4 expression, as observed in PVDs, disrupts epithelial integrity and facilitates the transition to a mesenchymal phenotype, which contributes to cell migration, invasion and fibrosis. The downregulation of KLF4 in RPE cells correlates with the onset of EMT and the formation of fibrotic membranes in PVR [54]. Induction of EMT under pathological stress Despite its role in maintaining epithelial characteristics, KLF4 can paradoxically promote mesenchymal traits under pathological conditions. Indeed, in the presence of TGF-β, KLF4 cooperates with Smad proteins to activate EMT-related genes. This dual role is mediated by post-translational modifications of KLF4 and its interaction with co-regulators that influence its transcriptional activity [55, 56]. KLF4’s involvement in EMT is further modulated by oxidative stress, during which, its expression is altered, leading to a shift in its regulatory activity. This change could enable RPE cells to lose epithelial characteristics, gain migratory and invasive properties, and contribute to the formation of contractile membranes in PVR [57, 58]. In addition, recent in vivo evidence further supports the functional relevance of KLF4 in retinal fibrosis. Specifically, a preclinical study showed that overexpression of KLF4 in RPE cells, as part of a defined cocktail of reprogramming factors, reversed EMT-associated transcriptional changes, reduced fibrosis, and improved retinal morphology and function in aged and stressed mouse models. These results indicate that KLF4 regulates EMT and fibrotic pathways in the RPE in vivo [59]. Role in fibrosis and ECM remodeling KLF4 seems to play a significant role in ECM remodeling in PVDs. It influences the expression of ECM components, such as collagen, fibronectin and MMPs, which degrade ECM. In fibrotic conditions, KLF4 interacts with TGF-β signaling to upregulate pro-fibrotic genes and increase ECM deposition. This interaction creates a feedback loop that amplifies fibrotic responses and promotes the stiffening of the ECM, that could contribute to retinal detachment in PVR [57]. Additionally, KLF4 may modulate cytoskeletal dynamics of RPE, which are crucial for cell migration and contraction, as shown in other cell types such as vascular smooth muscle cells [60]. Regulation by upstream signals KLF4 expression is regulated by several upstream signaling pathways that are dysregulated in PVDs. TGF-β is a primary regulator, inducing KLF4 under fibrotic conditions and altering its activity to promote EMT and fibrosis [57]. Oxidative stress and inflammatory cytokines, such as IL-6 and TNF-α, also influence KLF4 expression, shifting its function toward pro-fibrotic and pro-inflammatory roles [55]. Additionally, Wnt/β-catenin signaling modulates KLF4 activity. β-catenin interacts with KLF4 to regulate genes involved in EMT and ECM remodeling, further reinforcing the mesenchymal phenotype. These interactions highlight the dynamic regulation of KLF4 and its integration into broader signaling networks that appear to drive PVD progression [61]. MITF: a critical regulator of RPE identity and EMT in PVDs Microphthalmia-associated transcription factor (MITF) is a pivotal regulator of RPE cell identity, differentiation and homeostasis. Its role in maintaining the epithelial phenotype of RPE cells makes it a key player in the pathogenesis of PVDs (Fig. 1). Under pathological conditions, MITF is downregulated, contributing to EMT, fibrosis, and the formation of fibrotic membranes characteristic of these diseases [62]. Furthermore, recent data have shown that MITF counteracts TGF-β-induced EMT in RPE cells. Indeed, deletion of death associated protein like 1 (DAPL1), an upstream regulator of MITF, promotes EMT both in vitro and in vivo in mice by downregulating MITF expression. These findings support the protective role of MITF against retinal fibrotic remodeling [63]. Role in maintaining epithelial phenotype MITF is essential for the maintenance of epithelial characteristics in RPE cells. It regulates the expression of epithelial markers such as E-cadherin and tight junction proteins, which are critical for cell-cell adhesion and the structural integrity of the RPE monolayer [64]. Additionally, MITF supports the expression of genes involved in melanogenesis, oxidative stress resistance and cellular differentiation, preserving the specialized functions of RPE cells [65]. Under normal conditions, MITF prevents the mesenchymal reprogramming of RPE cells by suppressing the expression of EMT-related TFs such as ZEB1, Snail and Twist1. This regulatory role ensures that RPE cells maintain their epithelial phenotype and resist pathological transitions that contribute to disease progression [64]. Downregulation in PVDs In PVDs, MITF expression is significantly downregulated, which destabilizes the epithelial phenotype of RPE cells. This reduction in MITF mRNA levels is triggered by pathological signals such as TGF-β, oxidative stress and hypoxia. TGF-β suppresses MITF expression through Smad-dependent pathways, promoting the loss of epithelial markers and the acquisition of mesenchymal traits [66]. Similarly, hypoxia-induced activation of HIF-1α downregulates MITF expression, further driving EMT and fibrosis [67]. The loss of MITF expression in PVDs facilitates the detachment, migration and invasion of RPE cells. These cells contribute to the formation of contractile membranes that distort the retinal architecture and exacerbate vision loss, particularly in PVR. Furthermore, the downregulation of MITF expression enhances cellular proliferation, which is a key feature in the early stages of membrane formation [66]. Regulation of EMT and fibrosis MITF serves as a critical barrier to EMT and fibrosis. Its downregulation in PVDs removes this barrier, allowing TFs like ZEB1 and Snail to drive mesenchymal reprogramming. During EMT, RPE cells lose epithelial characteristics such as apical-basal polarity and cell-cell adhesion while gaining motility and invasive capabilities [9]. These changes enable RPE cells to migrate and play a part in the fibrotic membrane formation. Additionally, MITF may indirectly suppress ECM-related gene expression such as collagen and fibronectin by inhibiting EMT progression [63]. Its loss under pathological conditions could lead to excessive ECM production, contributing to membrane stiffening and fibrosis. The downregulation of MITF also enhances the activity of MMPs, facilitating ECM degradation and remodeling [68]. These processes may be central to the progression of PVDs. β-Catenin: a central player in EMT and fibrosis in PVDs β-catenin (CTNNB1) is a multifunctional protein involved in cell adhesion and transcriptional regulation, primarily through its role in the canonical Wnt/β-catenin signaling pathway. In PVDs, β-catenin seems to contribute significantly to pathological processes, including EMT, fibrosis, and angiogenesis. Dysregulated canonical β-catenin signaling alters cellular behavior and could enable RPE cells to migrate and form fibrotic membranes that distort the retina [69]. In vivo studies confirmed that aberrant activation of β-catenin promotes EMT and retinal fibrosis in PVR. In particular, overexpression of methyltransferase 3, N6-adenosine-methyltransferase complex catalytic cubunit (METTL3) in a rabbit model of PVR reduced β-catenin levels and attenuated fibrotic membrane formation, highlighting the central role of this pathway in disease progression [70]. Role in cell adhesion and EMT In its structural role, β-catenin is a key component of adherens junctions, linking cadherins to the actin cytoskeleton and maintaining epithelial integrity. In healthy RPE cells, β-catenin is primarily localized at cell-cell junctions, stabilizing the epithelial phenotype. However, under pathological conditions, such as oxidative stress or inflammation, β-catenin is redistributed from the membrane to the cytoplasm and nucleus, triggering its transcriptional activity via the canonical Wnt pathway [71]. Nuclear accumulation of β-catenin in RPE cells and epiretinal membranes has recently been shown in patient samples from PVDs, further supporting its role in EMT and fibrotic remodeling [72]. During EMT, β-catenin plays a pivotal role by activating mesenchymal gene expression while suppressing epithelial markers such as E-cadherin [73]. This translocation is a hallmark of EMT, enabling RPE cells to lose their epithelial characteristics, gain migratory and invasive properties, and contribute to fibrotic membrane formation. β-catenin’s transcriptional activity, mediated through its interaction with T-cell factor/lymphoid enhancer factor (TCF/LEF), is a typical feature of the canonical Wnt signaling cascade, promoting the expression of genes such as vimentin [74]. Role in ECM remodeling and fibrosis β-catenin is also a major regulator of ECM remodeling. Through its interaction with TCF/LEF, it promotes the expression of MMPs and fibronectin genes involved in fibrosis and ECM remodeling [74]. The activation of canonical β-catenin signaling enhances the expression of MMPs, such as MMP-2 and MMP-9, which degrade ECM components and facilitate tissue remodeling. Simultaneously, β-catenin may also upregulate pro-fibrotic factors, including collagen and fibronectin, contributing to ECM deposition and stiffening [75]. This dual role in ECM degradation and deposition makes β-catenin a critical driver of fibrosis. In PVR, its activity (primarily mediated through canonical Wnt/β-catenin signaling) could enable RPE cells to migrate into the vitreous cavity and secrete ECM components, forming contractile membranes that exert traction on the retina. Similarly, in PDR, β-catenin-mediated ECM remodeling supports the formation of fibrovascular membranes, exacerbating retinal detachment and vision loss [76]. Role in angiogenesis In addition to its role in EMT and fibrosis, β-catenin contributes to angiogenesis. Through the canonical Wnt/β-catenin signaling pathway, β-catenin activates genes involved in endothelial cell proliferation and migration, including VEGF. Elevated VEGF levels drive neovascularization, leading to the formation of fragile and leaky blood vessels in the retina [77]. β-catenin also regulates the expression of angiopoietins and other vascular remodeling factors, further supporting pathological neovascularization. These processes are amplified under hypoxic conditions, where it interacts with HIF-1α to enhance VEGF expression and favor angiogenesis [78]. Regulation by upstream signals The activity of β-catenin is tightly regulated by upstream signaling pathways that are frequently dysregulated in PVDs. Canonical Wnt signaling stabilizes β-catenin by preventing its phosphorylation and degradation by the ubiquitin-proteasome system. This stabilization allows β-catenin to accumulate in the cytoplasm and translocate to the nucleus, where it drives the expression of target genes involved in EMT, fibrosis and angiogenesis. Similarly, oxidative stress stabilizes β-catenin, promoting its nuclear translocation and transcriptional activity [71]. Other pathological signals, such as TGF-β and oxidative stress, further enhance β-catenin activity. TGF-β cooperates with β-catenin (mainly through convergence with the canonical Wnt pathway) to upregulate mesenchymal markers and ECM components, creating a feedback loop that amplifies EMT and fibrosis.While non-canonical Wnt pathways (such as Wnt/Planar cell polarity and Wnt/Ca2+) may influence cytoskeletal reorganization and inflammation, their involvement in PVDs are less defined and do not typically engage β-catenin. In contrast, the canonical Wnt/β-catenin signaling pathway remains the main driver of fibrotic and angiogenic responses in RPE cells [71]. AEBP1: a regulator of fibrosis, inflammation, and ECM remodeling in PVDs Adipocyte enhancer-binding protein 1 (AEBP1) is a TF and ECM-associated protein involved in regulating fibrosis, inflammation and ECM remodeling. In PVDs, AEBP1 plays a key role in promoting pathological processes that lead to retinal damage, fibrotic membrane formation and vision loss [79]. A single-cell transcriptomic analysis of human fibrovascular membranes from patients with PDR revealed AEBP1 as one of the top upregulated genes in myofibroblast-like cells. The study identified AEBP1 as a key modulator of fibrogenesis, strengthening its implication in retinal fibrosis in vivo in the human context [79]. In addition, dysregulated AEBP1 expression contributes to ECM deposition, chronic inflammation and the EMT of RPE cells [79, 80]. Role in ECM remodeling and fibrosis AEBP1 is a significant regulator of ECM remodeling, promoting both ECM synthesis and degradation. It directly upregulates in cancer prostatic cells collagen I, collagen III and fibronectin [81], which are the main components of fibrotic membranes in PVR. Furthermore, TGF-β induces AEBP1 expression, which in turn amplifies TGF-β signaling by upregulating downstream effectors such as CTGF. This positive feedback loop subserves ECM deposition and stiffening that could contribute to fibrotic membrane formation and tractional retinal detachment in PVDs [79]. Role in inflammation In addition to its role in ECM remodeling, AEBP1 is a critical regulator of inflammation in PVDs. As shown in other cell types [80, 82], it may enhance the expression of pro-inflammatory cytokines, including IL-6, TNF-α and IL-1β. These cytokines recruit immune cells to the retinal microenvironment, exacerbating chronic inflammation and promoting tissue damage. AEBP1 also interacts with the nuclear factor-kappa B (NF-κB) pathway, a master regulator of inflammation. It activates NF-κB, which in turn upregulates inflammatory mediators [82] and therefore may perpetuates inflammation in the retina. This inflammatory environment plays a part the breakdown of the blood-retinal barrier, vascular leakage and the formation of fibrovascular membranes in PDR. Regulation of EMT AEBP1 seems to play a central role in driving EMT in RPE cells. Following TGF-β signaling, AEBP1 upregulates mesenchymal markers such as N-cadherin and vimentin while suppressing epithelial markers like E-cadherin. This EMT transition enhances RPE cell motility and invasiveness, allowing these cells to migrate and to form a fibrotic membrane [79]. Additionally, AEBP1 modulates cytoskeletal dynamics, which are critical for cellular migration in stromal ovarian cancer [83]. By regulating actin remodeling and focal adhesion complexes, AEBP1 may enable RPE cells to exert traction forces on the ECM that leads to retinal detachment in advanced PVR cases. ZFP36L1: a post-transcriptional regulator in PVDs Zinc finger protein 36, C3H type-like 1 (ZFP36L1), also known as BRF1 (Butyrate Response Factor 1), is a member of the tristetraprolin (TTP) family of RNA-binding proteins. ZFP36L1 plays a crucial role in regulating mRNA stability, particularly for mRNAs containing AU-rich elements (AREs) in their 3’ untranslated regions (UTRs) [84]. In the context of PVDs, ZFP36L1 could contribute to the modulation of inflammation, fibrosis and angiogenesis by targeting the mRNAs of pro-inflammatory cytokines, growth factors and other key regulatory proteins (Fig. 1). Role in mRNA stability and inflammation ZFP36L1 binds to AREs within the 3’ UTRs of target mRNAs, promoting their deadenylation and subsequent degradation. Among its targets are mRNAs encoding pro-inflammatory cytokines such as TNF-α, IL-6 and VEGF [84]. By destabilizing these mRNAs, ZFP36L1 acts as a negative regulator of inflammation and angiogenesis (Fig. 2). Role of ZFP36L1 in mRNA stability and inflammation ARE 3′ UTR mRNA, AU-rich element in the 3′ untranslated region of messenger ribonucleic acid; CTGF, connective tissue growth factor; HIF-1α, hypoxia-inducible factor 1 alpha; IL-1β, interleukin-1 beta; IL-6, interleukin-6; KLF4, KrüppellLike factor 4; MK2, MAPK-activated protein kinase 2; p38 MAPK, p38 mitogen-activated protein kinase; TGFβ, transforming growth factor beta; TNFα, tumor necrosis factor alpha; VEGF, vascular endothelial growth factor; ZEB1, zinc Finger E-box-binding homeobox 1; ZEB2, zinc finger E-box-binding homeobox 2; ZFP36L1, zinc finger protein 36 like 1. In PVDs, chronic inflammation is a significant contributor to disease progression. Elevated levels of pro-inflammatory cytokines in the vitreous humor and retina lead to the activation of RPE cells, gliosis and neovascularization [85]. Downregulation or functional impairment of ZFP36L1 can result in increased stability and accumulation of these cytokine mRNAs, exacerbating inflammation and promoting pathological changes [84]. Regulation of EMT ZFP36L1 influences EMT and fibrosis by targeting mRNAs encoding factors involved in these processes. For instance, ZFP36L1 can bind to and destabilize mRNAs of TGF-β and CTGF, both key mediators in fibrosis and ECM remodeling [86]. By reducing the levels of these pro-fibrotic factors, ZFP36L1 could help to inhibit EMT in RPE cells and curb the subsequent formation of fibrotic membranes characteristic of PVR. Interaction with signaling pathways ZFP36L1’s activity is modulated by various signaling pathways relevant to PVDs. Phosphorylation of ZFP36L1 by kinases such as MAPK-activated protein kinase 2 (MK2) can inhibit its mRNA-destabilizing function [87]. Inflammatory stimuli like TNF-α and IL-1β activate the p38 MAPK pathway, leading to phosphorylation and inactivation of ZFP36L1, thus prolonging the half-life of pro-inflammatory mRNAs [88]. Additionally, hypoxia has been reported to upregulate ZFP36L1 expression in human renal carcinoma cells [89], while to inhibit ZFP36L1 via the action of miR 1825p in nasopharyngeal carcinoma cells [90]. ZFP36L1 binds to the 3’ UTR of VEGF [89], HIF-1α [91–93], KLF4 [94], ZEB1 [95], ZEB2 [96], TNF-α [84], IL-6 [84], TGF-β [84], CTGF [84] mRNAs, leading to their degradation. This interplay underscores the importance of ZFP36L1 in hypoxia, inflammation and EMT, all participating in PVD pathogenesis. Interactions between TFs and therapeutic implications Understanding the complex interplay between TFs is crucial for unraveling the mechanisms driving PVDs. These interactions, which often involve synergistic or antagonistic relationships, amplify pathological processes such as EMT, inflammation and ECM remodeling. These dynamic networks provide unique opportunities for developing multi-targeted therapeutic strategies (Fig. 3). Here, we explore these interactions in greater depth and highlight their therapeutic implications. It is important to notice that the following therapeutic strategies are primarily based on preclinical studies in cell and animal models, unless otherwise stated. While some of these approaches have shown promise in other fibrotic diseases, their application to PVDs remains largely experimental. TFs as possible therapeutic targets in PVDs. Identifying and understanding the role of TFs that can switch on/off the transcription of genes involved in PVDs, may make them future potential therapeutic targets. DNA, deoxyribonucleic acid; ERM, epiretinal membrane; PDR, proliferative diabetic retinopathy; PVD, proliferative vitreoretinal disease; PVR, proliferative vitreoretinopathy; RRD, rhegmatogenous retinal detachment; TF, transcription factor Synergistic interactions between various TFs in EMT and fibrosis The EMT process, central to PVDs, is driven by the coordinated action of TFs such as ZEB1, β-catenin and HIF-1α (Fig. 4). ZEB1 represses epithelial markers like E-cadherin and upregulates mesenchymal markers such as N-cadherin [10]. This reprogramming is further enhanced by HIF-1α, which not only upregulates ZEB1 but also activates TGF-β signaling, reinforcing fibrotic responses. Under hypoxic conditions, common in PDR, HIF-1α stabilizes and forms a feedback loop with ZEB1 to perpetuate EMT. β-catenin plays a role in this network by activating Wnt signaling, which stabilizes HIF-1α and enhances mesenchymal gene expression [78]. Role of various TFs in EMT process occurring in PVDs. The role and interactions betweens various TFs are represented in this scheme. AEBP1, adipocyte enhancer-binding protein 1; AP-1, activator protein 1; CTNNB1, beta catenin 1; ECM, extracellular matrix; EMT, epithelial–mesenchymal transition; ERK1/2, extracellular signal-regulated kinase 1/2; HIF-1α, hypoxia-inducible factor 1 alpha; IL-1β, interleukin-1 beta; IL-6, interleukin-6; JNK, c-Jun N-terminal kinase; MET, mesenchymal–epithelial transition; MITF, microphthalmia-associated transcription factor; MMP, matrix metalloproteinase; NF-κB, nuclear factor kappa B; NFAT5, nuclear factor of activated T Cells 5; OTX2, orthodenticle homeobox 2; p21, cyclin-dependent kinase inhibitor 1 A; p38 MAPK, p38 mitogen-activated protein kinase; SMAD, SMAD family transcription factors; TGFβ, transforming growth factor Beta; TNFα, tumor necrosis factor alpha; VEGF, vascular endothelial growth factor; Wnt, wingless/integrated signaling pathway; ZEB1, zinc finger E-box-binding homeobox 1; ZFP36L1, zinc finger protein 36 like 1 These interactions between TFs also extend to ECM remodeling. HIF-1α induces the expression of CTGF, a downstream effector that promotes collagen deposition and ECM stiffening [46]. β-catenin amplifies this effect by destabilizing cell-cell adhesion through the downregulation of E-cadherin expression, enabling RPE cells to migrate and invade the surrounding tissue. ZEB1 further supports these processes by repressing MITF expression [97], a TF essential for maintaining epithelial identity in RPE cells (Figs. 1 and 4). Therapeutically, these synergistic interactions can be targeted through dual inhibition strategies, supported by preclinical evidence, including in vitro and animal studies. Blocking HIF-1α and ZEB1 simultaneously may suppress EMT and reduce TGF-β-driven fibrosis. Similarly, Wnt pathway inhibitors that target β-catenin nuclear activity [98] could disrupt its cooperation with HIF-1α, effectively halting both EMT and ECM remodeling. These approaches are supported by preclinical evidence showing the efficacy of such combinations in reducing fibrotic responses in PVD models [99]. Moreover, the activation of ZFP36L1 may also be considered as another mono therapeutic option to promote the degradation of ZEB1 and HIF-1 mRNAs (see above). However, it will be necessary to verify if this therapeutic option does not affect KLF4 mRNA stability and thereby the RPE cells identity (see above and Fig. 3). Upstream stressors driving transcriptional reprogramming in PVDs Beyond the functional roles of individual TF, the pathological activation of these regulators in PVDs must be understood in the context of the upstream stress signals that dominate the retinal microenvironment. Several systemic and local stressors (most notably oxidative stress, hypoxia, hyperglycemia, osmotic imbalance, and chronic inflammation) act as potent initiators of transcriptional reprogramming in RPE cells [100]. Key upstream triggers such as TGF-β1, TNF-α, and cytokines (e.g., IL-1β, IL-6) are well known to drive EMT through activation of canonical transcriptional pathways involving ZEB1, β-catenin, AEBP1, and others [13, 14, 79] (Fig. 4). In parallel, hypoxia stabilizes HIF-1α, enhancing angiogenic responses and metabolic adaptation [40], while oxidative stress and hyperosmolarity have been shown to upregulate AP-1 and NFAT5, further sustaining inflammation and matrix remodeling [15, 16, 26, 29] (Fig. 4). High glucose and advanced glycation end products (AGEs), central in the diabetic context, modulate multiple TFs either directly or through downstream pathways such as p38 MAPK, NF-κB, and Wnt/β-catenin signaling [101, 102]. Importantly, these upstream signals do not act in isolation: they converge and amplify each other, reinforcing pathological feedback loops. For instance, TGF-β-induced oxidative stress can further promote EMT and downregulate epithelial markers such as MITF and KLF4, weakening epithelial identity [55, 62] (Fig. 1). Meanwhile, inflammatory stressors suppress ZFP36L1, a post-transcriptional gatekeeper that normally dampens excessive cytokine production [88] (Fig. 2). Collectively, this network of stress-induced regulation forms the biochemical foundation for transcriptional drift toward EMT, fibrosis, and chronic inflammation in PVDs . Recognizing and integrating these upstream drivers is essential not only to understand the initiation and progression of vitreoretinal fibrosis, but also to identify therapeutic leverage points (such as modulation of TGF-β signaling, oxidative balance, or osmotic stress) that could indirectly influence transcription factor activity and restore epithelial homeostasis. Inflammatory amplification by NFAT5, AP-1, and AEBP1 Inflammation is a key driver of PVD progression, with TFs such as NFAT5, AP-1 and AEBP1 forming a network that sustains chronic inflammatory responses (Fig. 4). NFAT5, activated under osmotic stress, interacts with AP-1 to upregulate pro-inflammatory cytokines [24]. AP-1, a complex of Jun and Fos proteins, regulates the expression of MMPs, which degrade ECM components and facilitate fibrotic tissue remodeling [29, 32]. AEBP1 amplifies these inflammatory signals by interacting with the NF-κB pathway, sustaining cytokine production and promoting ECM deposition, particularly collagen I and fibronectin [81]. The interplay between these TFs creates a pathological loop where inflammation and fibrosis reinforce each other. For example, NFAT5-induced osmotic stress activates AP-1, which, in turn, enhances AEBP1 activity, driving both inflammation and fibrotic membrane formation [23, 82]. This loop may be particularly relevant in PVR and PDR, where chronic inflammation exacerbates retinal damage and promotes pathological angiogenesis. To disrupt this inflammatory network, therapeutic strategies could target upstream regulators such as MAPK to block AP-1 activation or use agents that modulate osmotic stress to reduce NFAT5 activity. AEBP1 can be targeted with TGF-β inhibitors to suppress its pro-fibrotic effects while simultaneously dampening inflammation. Such interventions hold promise for reducing the severity of inflammation and fibrosis in PVDs in preclinical systems. However, their translational potential in human patients remains to be confirmed. Coordination of angiogenesis by HIF-1α, β-catenin, and AP-1 Angiogenesis is a hallmark of PVDs, particularly in PDR and nAMD, where pathological neovascularization contributes to retinal detachment and hemorrhage. This process is orchestrated by a network of TFs, including HIF-1α, β-catenin and AP-1. HIF-1α is the central driver of hypoxia-induced VEGF expression, which promotes the proliferation and migration of endothelial cells, forming new blood vessels. β-catenin enhances this effect by stabilizing HIF-1α and activating the Wnt signaling pathway, which further upregulates VEGF expression [103]. AP-1 plays a part in angiogenesis by responding to oxidative stress and inflammation, directly increasing VEGF transcription and supporting vascular remodeling [104] (Fig. 4). These TFs also interact with the ECM, creating a structural foundation for angiogenesis. HIF-1α induces the production of ECM proteins such as laminin and fibronectin, while β-catenin regulates their remodeling through MMPs. AP-1 further amplifies this process by increasing MMP activity, promoting ECM degradation and facilitating endothelial cell migration. Therapeutic strategies targeting this network include combining anti-VEGF agents with inhibitors of HIF-1α or β-catenin to suppress both angiogenesis and ECM remodeling [105]. Antioxidant therapies that reduce AP-1 activation may complement these approaches by mitigating oxidative stress, a major driver of neovascularization [106]. Preclinical studies have demonstrated the efficacy of these combinations in reducing angiogenesis and fibrosis in PVD models [107, 108]. Although promising, all the strategies outlined above are based on preclinical experimental data. Therefore, future studies are needed to assess their safety and efficacy in clinical trials targeting retinal fibrosis in PVDs. Post-transcriptional modulation by ZFP36L1 ZFP36L1 regulates inflammatory and fibrotic responses at the post-transcriptional level by destabilizing mRNAs encoding cytokines such as IL-6 and TNF-α (Fig. 1). This activity positions ZFP36L1 as a central modulator of the inflammatory environment in PVDs. By degrading pro-inflammatory mRNAs, ZFP36L1 may indirectly suppresses the activation of TFs like NFAT5 and AP-1, reducing the amplification of inflammation and ECM remodeling. Additionally, ZFP36L1 may mitigate pathological remodeling in PVD by promoting the degradation of mRNAs encoding key EMT and hypoxia-inducible factors, such as HIF-1α, ZEB1, and ZEB2. Through this post-transcriptional regulation, ZFP36L1 helps suppress mesenchymal transition, fibrotic membrane formation, and inflammation, highlighting its potential as a therapeutic modulator [92, 96]. Therapeutic activation of ZFP36L1 could provide a unique approach to managing PVDs by simultaneously addressing inflammation, EMT and fibrosis. Combining ZFP36L1 activators with TGF-β inhibitors could enhance their efficacy by targeting both upstream and downstream pathways involved in fibrotic membrane formation. This strategy is supported by studies showing the ability of ZFP36L1 to reduce fibrotic responses in other inflammatory diseases [86], highlighting its potential in PVDs. Feedback loops and crosstalk The interactions among TFs are governed by complex feedback loops and crosstalk that amplify their pathological effects. HIF-1α, for instance, upregulates ZEB1 and TGF-β expression, creating a positive feedback loop that reinforces EMT and fibrosis. Similarly, β-catenin interacts with AP-1 to regulate MMPs and VEGF expression, sustaining ECM remodeling and angiogenesis. NFAT5, through its interaction with AP-1 and AEBP1, perpetuates inflammatory and fibrotic responses, further complicating disease progression. Breaking these feedback loops is a critical therapeutic goal. Combination therapies targeting multiple nodes in these networks are likely to be more effective than single-agent approaches. For instance, dual inhibition of HIF-1α and β-catenin could disrupt the EMT-angiogenesis axis, while targeting AP-1 and NFAT5 could simultaneously reduce inflammation and ECM remodeling. Combined targeting of ZFP36L1 along with AP-1 and/or NFAT5 could also represent an alternative selection of therapeutic targets. Precision medicine approaches that tailor therapies to individual patient’s TF expression profiles may further enhance treatment efficacy. Therapeutic implications of interactions As explained above, the dynamic interplay among TFs in PVDs underscores the need for multi-targeted therapeutic strategies. Dual pathway inhibition, such as combining HIF-1α and VEGF inhibitors or ZEB1 and β-catenin inhibitors, could address the multifactorial nature of these diseases. Targeting upstream regulators, such as MAPK for AP-1 or TGF-β for AEBP1, may provide broader effects by disrupting entire networks rather than individual TFs. Restorative therapies, including the activation of ZFP36L1, offer an additional layer of intervention by restoring homeostasis in inflammatory and fibrotic processes. Finally, integrating antioxidant therapies with TF inhibitors could mitigate the oxidative stress that drives many of these pathological networks, providing synergistic benefits. Nevertheless, it is essential to consider that simultaneous inhibition of multiple pathways may also interfere with physiological retinal functions. In addition, recent findings have challenged the notion that EMT in RPE cells is irreversible, suggesting instead that pharmacologic or epigenetic interventions may allow reprogramming toward an epithelial phenotype. Indeed, in a human stem-cell derived RPE cells, EMT was associated with distinct transcriptomic and epigenomic changes, and was shown to be preventable by nicotinamide, a compound known to support epithelial identity and metabolic integrity. These data suggest that nutritional or metabolic interventions may help maintain epithelial function and curb fibrosis [109]. Furthermore, pharmacologic targeting of myocardin related transcription factor A (MRTF-A), a transcriptional regulator involved in EMT, successfully suppressed mesenchymal transition in cultured RPE cells, reinforcing the idea that EMT is at least partially reversible. Together, these studies open avenues for therapeutic modulation of EMT in early stages of PVDs [110]. As such, all therapeutic approaches must be carefully evaluated to ensure that network disruption does not compromise retinal health or repair capacity. Table 1 summarizes the TFs involved in PVDs, their upstream signals, downstream targets, associated phenotypes, and potential therapeutic interventions based on current literature. Table 1Key transcription factors involved in PVDsUpstream signal / PathwayTFDownstream targets genes/ Molecular effectorsPhenotype / Associated processPotentialtherapeutic interventionReferencesTGF-β, hypoxia, oxidative stress,IL-6,TNF-α,HIF-1 αZEB1E-cadherin,N-cadherin, vimentin, MITF, OTX2, p21,β-cateninEMT, fibrosis,RPE cell migration/invasion, PVR membrane formation,in vitroDual inhibition of ZEB1 + HIF-1α; activation of ZFP36L1[9–14]Hyperosmotic stress,TGF-β, hypoxia, MAPK, ERK1/2, oxidative stressNFAT5E-cadherin, vimentin, fibronectin, CTGF, HIF-1α, VEGF, IL-6, TNF-α, IL-1β, AP-1EMT, fibrosis, inflammation, PVR/PDR membrane formationModulation of osmotic stress; MAPK inhibition; combined with AP-1 inhibition[15–25]Oxidative stress, cytokines, TGF-β,MAPK (JNK, ERK1/2, p38), NFAT5, CTNNB1AP-1IL-6, TNF-α, IL-1β, MMP-2/9, fibronectin, collagen, vimentin,N-cadherin,E-cadherin, VEGF, AEBP1,HIF-1 αEMT, inflammation, angiogenesis, ECM remodeling, fibrosis, PVR/PDR,PVD model, cancer,in vivoMAPK inhibition; AP-1 inhibition; combined with NFAT5 and ZFP36L1 targeting[24, 26–38]Hypoxia,TGF-β, oxidative stress,Wnt/β-catenin, AP-1,NFAT5HIF-1αVEGF,CTGF, TGF-β, IL-6, TNF-α, ZEB1Angiogenesis, EMT, fibrosis, inflammation, PDR/PVR/nAMD membrane formation,PVD model, in vivoHIF-1α inhibition; combined with ZEB1 or β-catenin; anti-VEGF[39–50]TGF-β, oxidative stress, inflammation (IL-6, TNF-α), Wnt/β-catenin, ZFP36L1KLF4E-cadherin, EMT TFs (Snail, ZEB1, Twist1), collagen/fibronectin/MMPsMaintenance of epithelial identity; EMT/fibrosis under pathological stress, ECM remodeling, migration/contraction, in vivo, EMT reversibility/[51–61]TGF-β, hypoxia, oxidative stress,ZEB1,HIF-1αMITFE-cadherin, tight junction proteins, ZEB1, Snail, Twist1, collagen/fibronectin/MMPsMaintenance of RPE identity; downregulation promotes EMT, fibrosis, PVR membrane formation,in vitro,in vivo, EMT reversibility/[9, 62–68]Wnt/β-catenin, TGF-β, oxidative stress, hypoxia, ZEB1β-cateninE-cadherin, vimentin, N-cadherin, MMP-2/9, fibronectin, VEGF, HIF-1 α, AP-1EMT, fibrosis, angiogenesis, ECM remodeling, PVR/PDR,in vivoWnt pathway inhibitors; combined with HIF-1α or AP-1 targeting[69–78]TGF-β,AP-1AEBP1Collagen I/III, fibronectin, MMPs, IL-6, TNF-α, IL-1β, NF-κBEMT, fibrosis, inflammation, ECM remodeling, PVR/PDR membrane formation, cancer,in vivoTGF-β inhibitors; combined with modulation of NFAT5/AP-1[79–83]TNF-α,IL-1β, hypoxia,MK2,miR-1825pZFP36L1mRNAs: TNF-α, IL-6, VEGF, HIF-1α, KLF4, ZEB1/ZEB2, TGF-β, CTGFInhibition of EMT, fibrosis, inflammation, angiogenesis, EMT reversibilityActivation of ZFP36L1; combined with AP-1/NFAT5 targeting; TGF-β inhibitors[84–96] Key transcription factors involved in PVDs TGF-β, hypoxia, oxidative stress, IL-6, TNF-α, HIF-1 α E-cadherin, N-cadherin, vimentin, MITF, OTX2, p21, β-catenin EMT, fibrosis, RPE cell migration/ invasion, PVR membrane formation, in vitro [9– 14] Hyperosmotic stress, TGF-β, hypoxia, MAPK, ERK1/2, oxidative stress Oxidative stress, cytokines, TGF-β, MAPK (JNK, ERK1/2, p38), NFAT5, CTNNB1 IL-6, TNF-α, IL-1β, MMP-2/9, fibronectin, collagen, vimentin, N-cadherin, E-cadherin, VEGF, AEBP1, HIF-1 α EMT, inflammation, angiogenesis, ECM remodeling, fibrosis, PVR/PDR, PVD model, cancer, in vivo Hypoxia, TGF-β, oxidative stress, Wnt/β-catenin, AP-1, NFAT5 VEGF, CTGF, TGF-β, IL-6, TNF-α, ZEB1 Angiogenesis, EMT, fibrosis, inflammation, PDR/PVR/nAMD membrane formation, PVD model, in vivo Maintenance of epithelial identity; EMT/fibrosis under pathological stress, ECM remodeling, migration/ contraction, in vivo, EMT reversibility TGF-β, hypoxia, oxidative stress, ZEB1, HIF-1α Maintenance of RPE identity; downregulation promotes EMT, fibrosis, PVR membrane formation, in vitro, in vivo, EMT reversibility EMT, fibrosis, angiogenesis, ECM remodeling, PVR/PDR, in vivo TGF-β, AP-1 EMT, fibrosis, inflammation, ECM remodeling, PVR/PDR membrane formation, cancer, in vivo TNF-α, IL-1β, hypoxia, MK2, miR-1825p Conclusion TFs play a central role in the pathogenesis of PVDs, driving processes such as EMT, inflammation and angiogenesis. Their complex interplay, involving factors like ZEB1, HIF-1α, NFAT5, AP-1 and β-catenin, underscores the multifactorial nature of PVDs. These interconnected networks amplify pathological changes, making TFs highly attractive therapeutic targets. Current treatments are dominated by invasive surgical approaches, which fail to address the underlying molecular drivers of PVDs. Targeting TFs offers a transformative alternative by disrupting key pathological pathways. and moreover provide prolonged therapeutic intervention that could allow recovery and curb recurrence of PVDs. Strategies such as dual inhibition of HIF-1α and ZEB1 to prevent EMT and fibrosis or targeting β-catenin and AP-1 to mitigate ECM remodeling and inflammation, exemplify the potential for precision-based therapies. Additionally, activation holds promise for reducing inflammatory and fibrotic responses through post-transcriptional regulation. Especially, ZFP36L1 may attenuate the effects of ZEB1 and HIF-1α by reducing their mRNA stability. In addition, strategies aiming at reversing EMT in PVDs could also be considered as potential therapeutic options to restore epithelial characteristics and inhibit fibrotic progression of RPE cells. The future of PVD treatment lies in integrating multi-targeted therapies with patient-specific approaches. By addressing the root molecular mechanisms and leveraging combinations of TF inhibitors, anti-VEGF agents and antioxidants, it may be possible to significantly reduce the reliance on surgery and improve patient outcomes. Continued research into TF interactions will be crucial to realizing these advances and reshaping the management of these vision-threatening conditions. Regarding future perspectives, unresolved questions about the spatial and temporal regulation of TFs involved in PVDs call for more advanced experimental approaches. Ideally, a multi-omics approach combining data from DNA (such as profiling of chromatin accessibility, DNA methylation, histone modification), RNA (such as profiling of single-cell or single nuclei transcriptome, and spatial transcriptome) and proteins (such as profiling single cell proteome) will provide a complete comprehensive view of PVDs for deeper insights into disease mechanisms, major TFs involved, and identification of better drug targets. These analyses could be complemented by multiplex immunofluorescence to validate the expression and localization of TFs and proteins of interest within PVDs samples. In addition, mathematical dynamic modeling of gene regulatory networks, combined with trajectory analysis, will provide further insight into the temporal regulation of molecular mechanisms involved in PVDs. Altogether, these approaches will contribute to a comprehensive and dynamic understanding of transcriptional networks driving retinal fibrosis, inflammation, and angiogenesis occurring in PVDs, paving the way for more precise and personalized therapeutic strategies. Acknowledgements Some figures were partly created using Servier Medical Art (https://smart.servier.com/), licensed under a Creative Commons Attribution 4.0 Unported License (CC BY 4.0). Abbreviations Adipocyte enhancer-binding protein 1 Activator protein 1 Spontaneously arising retinal pigment epithelial cell line derived from the normal eyes of a 19-year-old male Blood-retinal barrier Catenin beta-1 (β-catenin) Connective tissue growth factor 4′,6-diamidino-2-phenylindole Death associated protein like 1 Extracellular matrix Extracellular signal-regulated kinase 1/2 Epithelial-to-mesenchymal transition Epiretinal membrane Hemagglutinin Hypoxia-inducible factor 1-alpha Immunofluorescence Interleukin-1 beta Interleukin-6 Kruppel-like factor 4 Mitogen-activated protein kinase Methyltransferase 3, N6-adenosine-methyltransferase complex catalytic cubunit Microphthalmia-associated transcription factor MAPK-activated protein kinase 2 Matrix metalloproteinase Myocardin related transcription factor A Neovascular age-related macular degeneration Nuclear factor-kappa B Neuroretina cyclin dependent kinase inhibitor 1 A Proliferative diabetic retinopathy Proliferative vitreoretinal disease Proliferative vitreoretinopathy Retinal pigment epithelium Retinal pigment epithelium-specific 65 kDa protein Rhegmatogenous retinal detachment Snail family transcriptional repressor 1 Self-labeling protein tag T-cell factor/lymphoid enhancer factor Transcription factor Transcription factor responsive element Transforming growth factor beta Tunicamycin and Thapsigargin Tumor necrosis factor alpha Twist family BHLH transcription factor 1 Vascular endothelial growth factor Western blot Wingless-related integration site Zinc finger E-box-binding homeobox 1 ZFP36 ring finger protein like 1 Author contributions C.Duveau wrote the manuscript. Y.R.E.H. designed the figures. C.Delporte. E.M., F.W., A.D. and C.Duveau obtained funding. All authors have reviewed and approved the final manuscript. Funding This work was supported by the Université Libre de Bruxelles, the Fonds de la Recherche Scientifique - FNRS (Fund for Scientific Research (Research Credit grant J.0128.23; E.M.); Fund Iris Research from the King Baudouin Foundation (2022-J1820690-226884; F.W.). A.D. benefited from research fellow of the Fonds Erasme pour la recherche médicale, and fundings for Individual Research Projects in Ophthalmology (IRPO) from the Fonds pour la Recherche en Ophtalmologie (FRO) and from Vocatio. C. Duveau. and Y.R.E.H. are Research Fellow of the Fonds de la Recherche Scientifique – FNRS. Data availability Not applicable. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare that they have no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Christine Delporte and Elie Motulsky contributed equally to this work. References Associated Data Data Availability Statement Not applicable.