From force to fate: Implications of mechanomemory in lung disease

Abstract

Mechanical cues are increasingly recognized as master regulators of cell behavior in development, regeneration and disease. In fibrotic tissues and solid tumors, aberrant extracellular matrix (ECM) stiffening activates mechanotransduction pathways that reprogram gene expression, metabolism and chromatin architecture. This stiffened microenvironment does not merely act as a transient signal but may also create a mechanotransductive imprint that reinforces signaling pathways, establishing a positive feedback loop that modulates cell behavior through mechanical memory. The concept of mechanomemory, defined as the cellular ability to retain and adaptively respond to past mechanical stimuli, has emerged as a critical player in cell biology offering new insights into how cells interpret and perpetuate biomechanical cues. This “memory” of mechanical stress modulates key processes such as cellular proliferation, epithelial-mesenchymal transition and invasion, contributing to disease progression and resistance against conventional therapies. This review explores the potential role of mechanomemory in the progression of lung diseases. We examine how sustained mechanical signals are encoded through molecular pathways, cytoskeletal adaptations, and epigenetic modifications, leading to persistent pathological cell states. We discuss how this imprinted memory may drive key features of lung disease, including the perpetuation of fibrosis and the acquisition of therapeutic resistance. We propose that targeting mechanomemory could open novel pathways for disrupting the vicious cycle of mechanical stress and lung disease while identifying novel areas for future research.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12931-026-03789-x.

Introduction

Virtually every tissue, whether the compliant lung parenchyma, the pulsatile vasculature or a healing skin wound, exposes its resident cells to forces and changes in extracellular matrix (ECM) stiffness that can span several orders of magnitude. Pathological changes in stiffness are now recognized as key drivers of disease progression, where mechanical forces have emerged as crucial regulators of disease progression, from solid tumor growth [1] to fibrotic tissue remodeling [2]. Tissues such as the lung normally exhibit a highly compliant ECM: healthy lung parenchyma has an elastic modulus of 0.5–5 kPa; in contrast, disease states like idiopathic pulmonary fibrosis (IPF) or lung cancer can stiffen lung tissue to 20–30 kPa [3, 4]. In acute lung injury (ALI) and acute respiratory distress syndrome (ARDS), edema, surfactant dysfunction and mechanical ventilation create a highly heterogeneous and dynamically changing mechanical landscape, exposing epithelial and endothelial cells to injurious levels of stretch, shear and surface stress [5].

Since the lung is a multi-scale mechanical organ, the term “mechanical forces” can refer to distinct stimuli operating at different spatial levels. Macroscopic forces, such as the cyclic stretch and shear stress imposed by mechanical ventilation or blood flow are transduced into microscopic cellular deformations. Simultaneously, the passive mechanical properties of the microenvironment (ECM stiffness) dictate how cells resist and respond to these external loads. Although conceptually different, mechanomemory likely integrates inputs from both these dynamic external forces and the static material properties of the tissue through partly distinct persistence mechanisms. For example, tumor cells and stromal cells in a stiff lung tumor niche activate mechanosensors such as integrins, cytoskeletal elements and stretch-activated ion channels to activate downstream pathways like FAK/Src, Rho/ROCK or YAP/TAZ that promote survival, invasion and metastasis [6–8]. Notably, chronic mechanical stress can also modulate therapeutic responses; a rigid matrix and sustained solid stress within tumors can induce signaling adaptations that confer drug resistance [9, 10]. In addition, patients with fibrotic stiff lungs have a higher risk of developing lung cancer [11], underscoring the pathological synergy between a persistently altered mechanical environment and disease evolution. In ALI/ARDS, repetitive overdistension, atelectrauma and cyclic reopening of flooded or collapsed units similarly drive epithelial injury, inflammation and, in some patients, a fibroproliferative response that can leave a chronically stiffened lung [12, 13].

Beyond these immediate mechanoresponsive effects, cells can “remember” past mechanical conditions and maintain altered phenotypes long after the original force is removed [14]. This phenomenon is defined as mechanical memory, or mechanomemory, and it adds a critical temporal dimension to mechanobiology. Seminal studies demonstrate that prolonged exposure to a stiff ECM can imprint lasting changes in gene expression, cytoskeletal organization and chromatin state, biasing cells towards pro-fibrotic or pro-metastatic behavior even after they return to a softer environment [14–16]. In essence, the mechanical history of a cell can be encoded molecularly by self-reinforcing transcriptional loops, sustained cytoskeletal tension, stable ECM remodeling and epigenetic reprogramming, affecting future responses to both mechanical and chemical stimuli. Mechanical memory has rapidly gained recognition as a fundamental aspect of cell biology, with broad implications for diseases such as cancer, fibrosis and tissue injury repair [14, 17–21]. Recognizing mechanomemory broadens the pathological landscape: it may explain why fibrotic lesions or tumor desmoplasia often continue to worsen long after the inciting injury is removed and why standard antifibrotic or chemotherapy regimens that ignore mechanical context yield only partial benefit. Conversely, it reveals new therapeutic entry points: softening an aberrantly stiff ECM, blocking mechanosignaling hubs or erasing epigenetic memory with chromatin-modifying drugs.

Several recent reviews have comprehensively addressed the broader role of mechanical forces in chronic respiratory diseases and the contribution of mechanobiology to the progression and resolution of lung fibrosis [22, 23]. These works provide an important complementary framework to the present review. Rather than revisiting the full spectrum of pulmonary mechanobiology, we focus here on its temporal dimension: how prior mechanical exposures may be encoded, retained and later recalled as mechanical memory.

This review synthesizes current knowledge of mechanomemory, exploring its implications in key lung diseases such as Idiopathic Pulmonary Fibrosis, Acute Lung Injury and lung cancer. We first define mechanomemory and its main molecular layers, then discuss how mechanical history may shape disease-specific cell states, and finally examine emerging therapeutic strategies and future research priorities aimed at disrupting maladaptive mechano-biological feedback.

Methods/literature search strategy

This review was conceived as a narrative, non-systematic review and was developed in accordance with the Scale for the Assessment of Narrative Review Articles (SANRA). A targeted literature search was performed in PubMed/MEDLINE and Web of Science. We used combinations of free-text terms and controlled vocabulary related to mechanical memory and lung disease, including but not limited to “mechanical memory”, “mechanomemory”, “mechanotransduction”, “extracellular matrix stiffness”, “fibrosis”, “idiopathic pulmonary fibrosis”, “acute lung injury”, “acute respiratory distress syndrome”, “ventilator-induced lung injury”, “lung cancer” and “pulmonary metastasis”. Reference lists of key primary papers and reviews were screened to identify additional relevant publications.

We primarily included peer-reviewed original articles and reviews in English that addressed mechanobiology or mechanical memory in the context of lung disease or general mechanotransduction pathways with clear relevance to pulmonary disease. Case reports, editorials and conference abstracts were not systematically considered unless they contained conceptually important information on mechanical memory.

Mechanical memory

Cells not only respond acutely to the mechanical properties of their environment, but can also retain information from previous mechanical exposures. In this review, we use the term mechanical memory, or mechanomemory, to refer to a persistent cellular state in which prior exposure to a defined mechanical cue modifies subsequent cell behavior after that cue has been removed or changed [15, 24, 25]. Operationally, “brief” priming generally refers to exposures lasting hours to approximately 1–3 days, which often induce reversible responses, whereas “prolonged” priming usually refers to repeated or sustained exposures of several days to weeks, which are more likely to generate durable phenotypic changes [26]. These time windows are model-dependent and should be interpreted in relation to the cell type, matrix dimensionality and mechanical stimulus used.

In lung-relevant systems, physiologic parenchymal stiffness is typically in the 1–2 kPa range, whereas fibrotic lung matrices may reach much higher values, often in the 25–50 kPa range depending on the measurement method [27]. Thus, throughout this review, “soft” and “stiff” are used as operational terms that refer to the mechanical contrast within each experimental system rather than to a universal threshold. Whenever possible, we specify the stiffness range and priming duration used in the original studies.

Because direct evidence for mechanomemory in lung disease is still emerging, Table 1 includes both lung-derived models and non-pulmonary proof-of-principle studies. The latter are included only when they identify mechanisms that are plausibly relevant to pulmonary fibrosis, ventilator-induced injury or lung cancer biology.

Table 1.

Experimental evidence supporting mechanical memory across lung-relevant and proof-of-principle models. The table includes pulmonary cells when available and selected non-pulmonary models that have defined core mechanisms of mechanomemory with potential relevance to lung disease. “Priming” refers to the initial exposure to a defined mechanical cue, most commonly substrate stiffness, stretch, confinement or pressure. Reported persistence varies according to cell lineage, priming duration, matrix dimensionality and the readout used, ranging from transient responses after short exposures to partly irreversible phenotypes after prolonged or repeated mechanical stimulation

Cell type	Experimental finding	Molecular mediators	Persistence
Mesenchymal stem cells (MSCs)	Stiff priming maintains nuclear YAP/TAZ and RUNX2 activity, biasing osteogenesis after 10 days stiff culture	YAP/TAZ, RUNX2	Several days to weeks; partly reversible [26]
	Chronic stiff culture induces fibrogenic traits and upregulation of miR-21	miR-21	≥ 2 weeks [28]
	Stiff exposure alters metabolism, enhancing smooth muscle differentiation	Metabolic–epigenetic crosstalk	Long-term, possibly irreversible [29]
Fibroblasts	Lung fibroblasts stiff-primed (100 kPa) become irreversibly activated into myofibroblasts with α-SMA stress fibers	ACTA2, collagen I, fibronectin	Stable for days–weeks [15, 24]
	Stiff-primed fibroblasts show enlarged nuclei and heterochromatin enrichment	Epigenetic nuclear changes	Persistent [30]
	Stiff priming for 7 days is required to establish memory in cardiac fibroblast	F-actin cytoskeleton	Stable after stimulus removal but pharmacologically reversible [31]
Epithelial cells	Collective migration influenced by prior stiffness: stiff-primed sheets migrate faster on soft matrices	Traction forces, EMT-related cues	Several days [16]
	Single-cell spreading enhanced after stiff priming	Integrin–cytoskeleton signaling	Persistent in 3D matrices [32]
Chondrocytes	Stiff priming induces chromatin condensation that persist after transfer to soft hydrogels	H3K9me3	Up to 10 days, partly reversible with inhibitors [20]
iPSCs	Increased cell area and α-smooth muscle actin after 24 h of stiff priming in iPSC-derived cardiomyocytes	Microtubule-dependent cytoskeletal signaling	Maintained after transfer to softer substrates [31]
Cancer cells	OSCC cells stiff-primed (20 kPa) upregulate nonmuscle myosin-II, acquiring an invasive phenotype that is maintained on softer substrates	Myosin-II, mesenchymal program	Maintained post-transfer [21, 33]
	Breast cancer cells primed in stiff 3D matrix show enhanced secondary outgrowth (bone colonization)	Cytoskeletal tension, RUNX2	Long-term [34]
	Glioblastoma cells exposed to high interstitial pressure gain resistance to mechanical stress	Invasion-associated programs	Facilitates brain infiltration [35]

Molecular mechanisms of mechanomemory

Mechanomemory is unlikely to arise from a single pathway. Rather, current evidence supports a multi-layer process in which mechanical cues are first sensed and translated into signaling responses, then stabilized through structural and nuclear adaptations, and finally consolidated by epigenetic and metabolic programs. In this framework, persistence emerges when these layers become coupled through positive feedback loops (Fig. 1), whereas reversibility depends on how deeply these adaptations become encoded and for how long the cells were mechanically primed. Key pathways and their potential relevance in lung disease are summarized in Table 2.

Fig. 1.

Conceptual framework of mechanomemory in lung disease. Pathological mechanical cues, such as fibrotic or tumour-associated ECM stiffness and injurious stretch during ALI or VILI, are sensed through integrins–FAK/Src signaling, RhoA/ROCK-dependent actomyosin contractility, Piezo/TRP channels and cytoskeletal tension. These early responses may be reinforced by transcriptional feedback loops, persistent cytoskeletal and focal adhesion organization, ECM deposition and crosslinking, and nuclear lamina remodeling. Long-term stabilization can involve altered chromatin accessibility, histone/DNA modifications, mechanosensitive enhancers and miRNA/non-coding RNA regulation. Metabolic reprogramming acts as a cross-cutting reinforcing layer by promoting stiffness-driven glycolysis, lactate accumulation and histone lactylation. These interconnected mechanisms may stabilize mechanically imprinted phenotypes that contribute to fibroblast activation, impaired epithelial repair, invasion/metastasis and therapy resistance

Table 2.

Molecular pathways contributing to mechanical memory: primary roles, mechanisms of persistence and potential implications in lung disease

Pathway/Molecule	Primary role	Role in mechanical memory	Known/potential relevance in lung disease
YAP/TAZ	Nuclear effectors of the Hippo pathway; sense matrix stiffness, cytoskeletal tension	Sustained nuclear localization/activity on stiff substrates [24–26]; drives persistent pro-fibrotic gene expression [30] and drives cancer progression [36–38]	Activated by stretch/stiffness in lung fibroblasts, epithelial and endothelial cells [39–41]: implicated in lung cancer progression [36, 42] and potential implications in VILI, post-ALI fibrosis and ATII proliferation/differentiation
RhoA/ROCK/MKL1	Regulates actin cytoskeleton organization, actomyosin contractility and stress fiber formation	Persistent activation on stiff substrates maintains myofibroblast phenotype [43]; MKL1 links actin dynamics to sustained pro-fibrotic gene expression [44]	Activation/differentiation of myofibroblasts [43]: potential role in sustained VILI responses and fibrosis
RunX1/RunX2	Transcription factor involved in development and disease	Identified as regulator of fibrogenic activation and epigenetic memory [45, 46]/Persistent activation of YAP/TAZ via feedback loop [26]	Potential role in the amplification of fibrogenic responses upon re-injury and cancer progression via YAP/TAZ
Integrins	Transmembrane receptors mediating cell-ECM adhesion and sensing mechanical cues	Mediate the initial sensing of matrix properties that lead to memory imprinting [47, 48]	Response to matrix changes in ALI; involved in mechanotransduction of ventilator-induced forces [49]
miR-21	MicroRNA involved in post-transcriptional gene regulation	Overexpressed in response to mechanical stimuli; sustain long-term fibrotic mechanical memory [28]	Implicated in pulmonary fibrosis [50]; potential role in sustaining pro-fibrotic phenotypes in lung cells after ALI
Histone modifying enzymes	Add/remove epigenetic marks	Mediate chromatin condensation/remodeling that encodes long-term mechanical memory [20, 51], stiffness can alter H3K9me3 patterns [52]	Potential role in epigenetic changes after mechanical stimuli in ALI to induce persistent epigenetic alterations in lung cells
Metabolic Pathways/enzymes	Regulate cellular energy production and biosynthetic processes	Sustained changes in mitochondrial dynamics and ATP production [53]; metabolic byproducts can drive epigenetic memory via feedback loops [54] and lactylation	Mechanical forces in ALI/VILI can alter lung cell metabolism [8]; persistent metabolic shifts could contribute to sustained dysfunction, pro-fibrotic states and cancer progression [55]

From sensing to persistence: transcriptional and signaling feedback

Sustained changes in gene expression are central to mechanomemory. Mechanosensitive transcription factors like YAP/TAZ not only transduce stiffness signals but also induce genes encoding matricellular proteins (e.g. CTGF, CYR61) and cytoskeletal regulators that stiffen the cell and ECM, thereby reinforcing their own activity, creating a self-sustaining circuit [25]. In MSCs, this has been shown to establish feedback loops with factors like RUNX2 to lock in an osteogenic program [26]. Similarly, chronic TGF-β1 signaling, induced by ECM stiffness, can become autonomous through a positive feedback loop where myofibroblasts deposit more ECM, which in turn amplifies TGF-β1 signaling [56]. This feedback is further reinforced by the way TGF-β1 is stored and activated within the ECM. Latent TGF-β complexes are tethered to the matrix through latent TGF-β-binding proteins, creating a reservoir of growth factor that can be activated by proteases or by integrin-mediated mechanical forces [57]. Matrix stiffening lowers the threshold for mechanical TGF-β activation, thereby coupling ECM remodeling, cellular contractility and chronic TGF-β signaling into a self-amplifying loop [58].

Structural encoding of memory: cytoskeleton, adhesions, ECM and nucleus

A second layer of mechanomemory is stored in cellular architecture. Cells primed on stiff substrates retain a robust actin cytoskeleton and enlarged focal adhesions for days, even after being moved to a soft matrix [16]. This sustained cytoskeletal tension maintains active signaling. Concurrently, cells remodel their surroundings, producing a cross-linked, collagen-rich niche that reinforces mechanosignaling [20]. This structural memory extends to the nucleus, where mechanical forces reinforce the nuclear lamina and alter chromatin tethering, stabilizing specific gene expression programs [59]. Thus, structural persistence spans from the cell cortex to the nucleus, linking mechanical history to durable phenotypic changes that are usually taken as a key indicator of mechanical memory [16, 60].

Long-term stabilization: epigenetic and non-coding RNA programs

For memory to be long-lasting, it likely requires epigenetic encoding. Mechanical forces elicit durable changes in chromatin architecture and epigenetic marks. For instance, in MSCs, dynamic stretch can cause global chromatin condensation via the EZH2 enzyme, creating a stiffer nucleus that persists long after the stimulus ceases [59]. Specific histone modifications, such as the redistribution of repressive H3K9me3 marks in chondrocytes [20] or the decrease of H3K27me3 at smooth muscle gene loci in MSCs [60], can lock cells into altered differentiation states. Furthermore, matrix mechanics can modulate global DNA methylation [61] and the expression of non-coding RNAs. Notably, miR-21 levels, elevated by stiff priming, can be maintained for passages, acting as a molecular memory by continuously modulating target genes long after the initial signal has vanished [28, 51]. Recent work has further refined this view by showing that mechanical cues can act through mechanosensitive enhancer landscapes. These “mechanoenhancers” provide a regulatory substrate through which mechanical information may be converted into durable, locus-specific transcriptional programs. Importantly, targeted epigenetic editing of such regions can reprogram the cellular response to the mechanical microenvironment, suggesting that mechanically imprinted enhancer activity may be both a mechanism of persistence and a potential therapeutic entry point [62].

Metabolic reinforcement of mechanomemory

Metabolic reprogramming is best viewed as a reinforcing layer of mechanomemory rather than a purely downstream consequence. Persistent mechanical signals can alter cellular metabolism to reinforce a phenotype. In fibroblast, endothelial cells and cancer cells, stiff environments often drive cells towards a more glycolytic state to support biosynthesis and growth [53], a shift regulated by mechanosensitive transcription factors YAP/TAZ [63]. Nuclear YAP/TAZ can upregulate glycolytic enzymes [54] and sustain nucleotide synthesis [64], creating feed-forward loops that favor cancer progression. This shift resembles a Warburg-like metabolic phenotype, in which cells preferentially increase glycolytic metabolism even when oxygen is available, thereby supporting biosynthesis, growth and stress adaptation. Similar metabolic rewiring can also be driven by mechanosensitive channels such as PIEZO1 and may contribute to therapy resistance [65]. Notably, sustained glycolysis may increase intracellular lactate, which can serve as substrate for histone lactylation, an emerging epigenetic mark induced by lactate [66, 67]. This modification may stabilize mechanically imprinted phenotypes, providing a direct link between metabolic reprogramming and long-term memory.

Integrative perspective

Signaling feedback, structural persistence, epigenetic encoding and metabolic reinforcement should not be interpreted as independent modules. Rather, they likely contribute with different weights depending on cell type, mechanical cue (stiffness versus stretch), dimensionality (2D versus 3D) and priming duration. Most mechanomemory evidence comes from reductionist in vitro systems (often 2D hydrogels with tunable stiffness). While these models are powerful to isolate stiffness-driven signaling, cell-ECM mechanotransduction in 3D can differ substantially due to matrix architecture, confinement, degradability, ligand presentation, local matrix remodeling and the release of bioactive ECM fragments or neo-epitopes that can further influence cell behavior. Therefore, translating stiffness-priming or mechanomemory signatures from 2D to the in vivo lung requires caution and validation in 3D systems (e.g. organoids, precision-cut lung slices, decellularized lung scaffolds) and in vivo models [68].

Emerging experimental tools may help bridge this gap. For example, local photocrosslinking of native tissue matrix now allows investigators to stiffen defined regions within living or ex vivo tissues and examine the downstream cellular consequences in a preserved matrix architecture [69]. Such approaches are particularly relevant for mechanomemory research because they make it possible to test whether focal mechanical “hotspots” are sufficient to initiate persistent cell-state changes, rather than relying exclusively on globally stiff synthetic substrates.

Together, these recent advances indicate that mechanomemory should not be viewed only as prolonged activation of canonical mechanotransduction pathways, but as a multi-layered process involving spatial matrix mechanics, regulatory enhancer landscapes and persistent injury-primed cell states.

Implications of mechanomemory in lung disease

Mechanomemory in IPF: stiffness as cause and consequence

In Idiopathic Pulmonary Fibrosis (IPF), mechanical memory is central to disease pathogenesis. A defining feature of IPF is the excessive and disorganized deposition of the ECM, which is increasingly recognized not just as a consequence of the disease but as an active contributor to its progression [70]. The resulting increase in tissue stiffness directly induces lung fibroblasts activation, perpetuating a vicious cycle of matrix production and further stiffening [41]. This stiff environment can cause fibroblasts to adopt an activated myofibroblast phenotype, characterized by α-smooth muscle actin (α-SMA)-positive stress fibers, heightened contractility and ECM production [27].

Mechanistically, this process involves a complex interplay of signaling pathways. As the matrix stiffens, integrins on cell surfaces transmit higher tension to the cytoskeleton, activating focal adhesions kinases (FAK) and Src. These signals synergize with pro-fibrotic cytokines like TGF-β, forming a feed-forward loop that is a core feature of IPF progression [71]. This loop is further amplified by the activation of mechanosensitive pathways like the YAP/TAZ signaling axis [72]. A sufficiently stiff substrate (Young modulus > 20–25 kPa) can induce myofibroblast differentiation even without exogenous TGF-β, whereas soft substrates (Young modulus < 5 kPa) maintain fibroblasts quiescence [73]. Prolonged exposure to this stiffness leads to long-term mechanical memory in fibroblasts, explaining the sustained cellular activation seen in IPF [18].

Seminal studies support the existence of mechanical memory-like behavior in lung fibroblasts. Marinkovic et al. showed that IPF patient-derived fibroblasts cultured on soft substrates (mimicking a normal lung) had dramatically reduced proliferative and myofibroblastic activity compared to those cultured on stiff matrices [74]. In a complementary study, Novak et al. “switched” pulmonary fibroblasts between rigid 2D culture and softer 3D collagen hydrogels. Cells expanded on stiff surfaces retained aspects of an activated phenotype even after being moved to a softer 3D matrix, although this state was partially reversible. For example, αSMA/Acta2, MMP3, fibronectin and DCN were largely driven by the most recent culture condition, supporting partial reversibility after transfer to a softer 3D matrix. In contrast, collagen 1, collagen 3 and MMP8 showed history-dependent or non-reversible patterns, indicating that functional recovery can coexist with residual transcriptomic imprinting [18].

More recently, in vivo evidence has strengthened the concept that prior fibrotic injury can leave a persistent mesenchymal imprint. RUNX1 has been identified as a mediator of fibrotic activation and epigenetic memory in lung fibroblasts, linking previous injury to enhanced fibroblast responses upon subsequent challenge [46]. This finding provides an important bridge between reductionist in vitro stiffness-priming studies and the possibility that fibrotic lung tissues retain a biologically meaningful memory of prior injury.

This mechanical recall extends beyond fibroblasts. Alveolar epithelial cells, especially in peripheral lung regions exposed to high mechanical strain, are also critically affected. Repetitive stretch injury to these cells is thought to initiate the fibrogenic cascade by inducing epithelial damage, delaying or distorting alveolar repair, and promoting fibroblast-activating signals [75]. On stiff matrices, alveolar type II (AT2) cells show increased susceptibility to injury and aberrant differentiation, promoting a transitional state instead of differentiating into type I cells [76]. In fibrotic mouse models, stiff areas presented senescent AT2 cells with nuclear protein loss [77]. Since past stiffness is known to influence cell migration and transition [16, 32], it is plausible that AT2 cells in a fibrotic environment are primed to perpetuate an aberrant phenotype. Immune and vascular compartments are also likely to participate in these feedback loops. Infiltrating monocytes and macrophages can regulate ECM turnover, secrete profibrotic mediators and amplify fibroblast activation, whereas endothelial dysfunction may contribute to altered barrier function, vascular remodeling and local mechanical heterogeneity [78, 79]. Thus, in IPF, the entire cellular ecosystem including fibroblasts, epithelial cells, endothelial cells and macrophages, responds to and is conditioned by mechanical signals, making tissue stiffness and mechanomemory both a cause and a consequence of pathological cell activation in fibrosis.

Mechanomemory in the pathogenesis and progression of lung cancer

A cardinal feature of many solid tumors, including lung cancer, is a marked increase in the stiffness of the ECM compared to the surrounding healthy tissue. This shift in the mechanical environment is not a passive bystander effect but an active contributor to malignancy [80].

In lung tumors, increased stiffness arises from extensive remodeling of the tumor microenvironment (TME), driven by both malignant cells and stromal cells, particularly cancer-associated fibroblasts (CAFs). Tumor and stromal cells increase deposition of ECM components (including collagen and fibronectin) and promote matrix cross-linking through enzymes such as lysyl oxidases (LOXs), generating a denser and mechanically rigid niche [80]. In turn, this stiffened TME activates mechanotransduction pathways in cancer cells through integrin-mediated signaling, FAK and YAP/TAZ, which promote survival, invasion and aggressive phenotypes [81]. These pathways are central to acute mechanoresponses and are also biologically plausible mediators of mechanomemory, as they can engage transcriptional and cytoskeletal feedback loops capable of sustaining phenotypic changes after prolonged mechanical exposure [19].

Direct evidence that mechanical priming drives metastatic behavior through mechanomemory remains stronger in non-pulmonary systems than in lung cancer specifically. Mechanomemory has been proposed as a mechanism by which exposure to mechanical stresses in the primary tumor can prime cancer cells for metastatic success, with these adaptations retained during dissemination and colonization [17]. Supporting proof-of-principle comes from non-pulmonary tumor models (including breast and oral carcinoma), where pre-conditioning on stiff substrates enhances invasive behavior and extravasation capacity [16, 21]. These studies support the concept that mechanical history can bias later metastatic performance, but they should not be interpreted as direct evidence in lung cancer. Under this hypothesis, mechanical interactions within a stiff primary pulmonary TME may select or imprint subclones with enhanced invasive fitness and dissemination potential. Therefore, mechanomemory would not only reflect adaptation to the local lung niche, but could also influence the ability of tumor cells to survive transit and colonize distant organs. A testable prediction is that tumors with stronger signatures of mechanotransduction and matrix remodeling will show greater metastatic propensity or treatment resistance. If validated, this framework would support early interventions targeting tumor-matrix interactions or mechanotransduction pathways at the primary site as a strategy to reduce the emergence and dissemination of mechanically primed metastatic clones.

Mechanomemory in Acute Lung Injury: A new frontier in pathogenesis and therapeutic innovation

Acute Lung Injury (ALI) and its severe form, Acute Respiratory Distress Syndrome (ARDS), are critical illnesses characterized by acute respiratory failure, diffuse alveolar damage and reduced lung compliance, and are associated with substantial mortality and long-term morbidity among survivors [82, 83]. In addition to inflammation and edema, ALI/ARDS creates a highly heterogeneous mechanical environment, in which ventilation, cyclic reopening of unstable units and altered matrix properties expose cells to injurious stretch, shear and interfacial stresses [5, 49, 83, 84].

A substantial body of work supports the role of mechanical stress in lung injury pathogenesis. Pathological cyclic stretch can injure alveolar epithelial cells, alter gene expression and promote inflammatory mediator release [85, 86]. Pulmonary endothelial and stromal cells also respond to injurious mechanical stimuli with inflammatory and pro-fibrotic signaling programs [85–87]. Mechanotransduction pathways including FAK/MAPK/NF-ĸB are activated under these conditions [86], and disruptions of nuclear mechanosensation (including altered nuclear membrane/Lamin-associated responses) modulate the severity of ventilator-associated injury in preclinical models [88]. Consistent with these mechanisms, fibroproliferative repair and matrix remodeling are recognized features of non-resolving ARDS and may contribute to persistent functional impairment [12, 13, 83].

Although direct evidence for long-term mechanomemory in human ARDS is limited, some observations are compatible with a priming/persistence framework. In a preclinical model, prior exposure to mechanical ventilation altered the inflammatory response to a subsequent ventilation insult, suggesting that mechanical history can modify future injury responses [89]. In parallel, work on lung injury-associated mesenchymal epigenetic priming supports the broader concept that prior injury can leave persistent response biases, even if this has not yet been definitively established as ventilation-specific mechanomemory in ARDS [45]. Finally, recent analyses showing that recruitment/derecruitmentt, defined as the cyclic collapse and reopening of unstable or fluid-filled alveolar units during ventilation, can concentrate injurious power in spatially heterogeneous regions, providing a plausible mechanical basis for localized, non-uniform downstream cell states [84].

We hypothesize that, in a subset of patients, repetitive injurious stretch and recruitment/derecruitment may imprint persistent transcriptional, epigenetic and matrix-associated changes in epithelial, endothelial and mesenchymal cells, thereby contributing to prolonged dysfunction after the acute phase. Under this model, mechanically “focused” regions of the lung could become hotspots of persistent maladaptive remodeling. A related testable prediction is that different initial etiologies and ventilation strategies may produce different post-ARDS trajectories by generating distinct spatial stress patterns [49, 90]. If validated, this framework would position mechanomemory as a contributor linking acute mechanical injury to chronic sequelae, and would support studies combining longitudinal sampling with spatial and single-cell profiling to identify persistent mechanically imprinted cell states in ARDS survivors.

Therapeutic opportunities

If mechanical stimuli play a major role in the pathogenesis of respiratory conditions, a therapeutic approach to the underlying mechanisms should be considered, as it opens a novel strategy that could help to improve the outcome of patients. There are several opportunities to exploit mechanodependent pathways in lung diseases (Supplementary Table 1). A conceptual overview of these emerging therapeutic strategies is provided in Fig. 2.

Fig. 2.

Emerging therapeutic opportunities to target lung mechanomemory

Targeting ECM stiffness and remodeling

A primary strategy to counter mechanomemory is softening the abnormally stiff ECM that drives it. This can be approached by inhibiting matrix cross-linking or by actively degrading excess matrix components.

A key target is the lysyl oxidase (LOX) family of enzymes, which catalyze the collagen cross-linking that increases matrix rigidity. While inhibiting LOX has shown promise in preclinical models, clinical translation has been challenging. The monoclonal antibody simtuzumab, which targets LOXL2, failed to improve outcomes for IPF patients in a Phase II trial [91, 92]. Subsequent studies revealed that inhibiting LOXL2 alone could paradoxically enhance fibroblast invasion and myofibroblast differentiation [93], potentially due to compensatory mechanisms from other LOX family members like LOXL1 or LOXL4 [94]. These findings suggest that while normalizing matrix stiffness via LOX inhibition is a compelling approach, it may require a pan-LOX blockade or more context-specific strategies.

Recent work has extended the therapeutic relevance of matrix mechanics from absolute stiffness to stiffness gradients. Durotaxis, defined as directed cell migration along stiffness gradients, was shown to actively contribute to disease progression in vivo in models of lung fibrosis and metastatic cancer. Mechanistically, this process depended on a FAK–paxillin mechanosensory module that links stiffness cues to YAP-dependent transcriptional programs [95]. Importantly, genetic disruption of FAK–paxillin binding using a FAK-FAT^L994E knock-in model, as well as pharmacological inhibition of this interaction with the small molecule JP-153, blocked durotaxis and attenuated disease severity. These findings support anti-durotactic therapy as a mechanistically targeted strategy for lung fibrosis and cancer, complementary to approaches aimed at globally softening the ECM.

An alternative is to degrade or remodel the excessive collagen network directly. The hormone relaxin exhibits potent antifibrotic properties [96], inhibiting new collagen synthesis while upregulating matrix metalloproteinases to accelerate collagen turnover [97]. In rodent models of lung and cardiac fibrosis, relaxin treatment reduced collagen content and improved tissue compliance [98]. However, clinical translation has been hampered by pharmacokinetic issues [99]. A more direct approach involves using enzymes like collagenase, which is already approved for localized fibroproliferative conditions such as Dupuytren’s contracture, a disorder in which collagen-rich fibrotic cords form in the palmar fascia and progressively limit finger extension [100].. Intratracheal delivery of matrix metalloproteinase 9 ameliorated lung injury in neutropenic mice submitted to mechanical ventilation (that show a functional deficiency in this protease) [101]. A targeted or systemic delivery of collagenase could debulk fibrotic scar tissue in organs like the lung, but in practice this is limited by potential off-target damage and tight regulation of proteases in vivo [102].

Experiences from non-pulmonary malignancies where hyaluronan-depleting strategies (PEGPH20) failed to improve overall survival despite stromal modulation [103], suggest that targeting a single matrix component may be insufficient to overcome tumor adaptation or that the stroma may also play a tumor-restraining role [104]. These experiences indicate that matrix-directed therapies in lung fibrosis and lung cancer will probably require careful patient selection and rational combinations.

Inhibiting mechanotransduction pathways

Another major strategy is to target the cellular machinery that “reads” and responds to mechanical cues. By interrupting these mechanotransduction pathways, the goal is to prevent cells from executing pro-fibrotic or pro-survival programs in response to pathological stiffness.

A central player is the Hippo-YAP/TAZ pathway. On stiff substrates, the transcriptional coactivators YAP and TAZ translocate to the nucleus and drive gene programs for cell growth and matrix production. In fibrotic tissues and solid tumors, YAP/TAZ are constitutively active, serving as master regulators of disease progression [72]. Culturing fibroblasts on stiff matrices activates YAP/TAZ, while soft matrices keep them quiescent in the cytoplasm [40]. While direct YAP/TAZ inhibitors for human use remain elusive, the repurposed drug verteporfin, which disrupts the YAP-TEAD interaction, has provided crucial proof-of-concept. Verteporfin has been shown to suppress fibroblasts activation in lung disease models [42, 105], and reduce fibrogenesis in preclinical models of liver [106] and cardiac [107] fibrosis.

Other key mechano-signaling nodes are also being targeted. Focal Adhesion Kinase (FAK), which propagate signals from the matrix, is hyperactivated in cancer and fibrosis, promoting cell migration and survival [108, 109]. FAK inhibitors are in clinical trials for non-small cell lung cancer, but as single agents, they have shown only modest efficacy, suggesting a combination strategy will be necessary [110, 111].

Downstream of FAK is the Rho-ROCK pathway, a key mediator of cellular contractility. In fibrotic diseases, overactive ROCK1/2 contributes to the high contractive force of myofibroblasts and excessive matrix deposition [4, 112]. ROCK inhibitors like fasudil have shown promising antifibrotic effects in preclinical models, reducing fibrosis and myofibroblasts numbers in the lung [113, 114]. These results have spurred the clinical development of next-generation ROCK2-selective inhibitors like zerasudil, which is currently in early-phase studies for IPF.

Approved antifibrotic therapies for IPF, including pirfenidone, nintedanib and the recently approved PDE4B inhibitor nerandomilast, also intersect with pathways relevant to fibroblast activation and matrix remodeling, although they were not originally developed as mechanomemory-targeted therapies [115]. Nintedanib, a multi-tyrosine kinase inhibitor, was recently shown to impair the nuclear translocation of YAP/TAZ [116]. Nerandomilast expands the therapeutic landscape by targeting PDE4B-dependent signaling and reducing FVC decline in patients with fibrotic lung disease. While not designed as “mechanotherapies”, their clinical efficacy demonstrates that breaking the mechano-fibrotic feedback loop via biochemical pathways is a valid therapeutic strategy [117]. Additionally, inhibitors of integrins, such as the oral αvβ6/β1 inhibitor bexotegrast, are showing early promise in IPF by blocking the mechanical activation of latent TGF-β in the fibrotic niche [118].

Reprogramming or erasing mechanical memory

Mechanical memory at the cellular level is thought to be at least partly encoded by epigenetic changes. Sustained forces or a stiff microenvironment can lead to stable alterations in chromatin state, for instance, through modification of histones or DNA methylation, that lock cells into an activated phenotype even after the stimulus is gone [52]. Additionally, the physical properties of the nucleus itself play a significant role: a stiff ECM can cause the nucleus to adapt by increasing Lamin-A expression, which can reinforce certain gene expression patterns [88]. Therefore, a novel class of therapeutic strategies aims not just to block signaling, but to erase or overwrite these memory marks, essentially reprogramming fibrotic cells towards a more normal state.

Histone deacetylase (HDAC) inhibitors work by maintaining a more open chromatin structure, which can reactivate silenced anti-fibrotic genes. Pan-HDAC inhibitors like trichostatin A or vorinostat can block myofibroblasts differentiation [119, 120], and compounds like panobinostat have been shown to reduce collagen accumulation in preclinical models of pulmonary fibrosis [121]. Selective inhibition of specific HDAC subtypes reduced collagen gene expression and fibrotic lesions in mice and suppressed TGF-β induced fibroblasts activation and lung fibrosis, with efficacy comparable to pirfenidone in some readouts [121], suggesting that more targeted approaches are possible.

Bromodomain and extraterminal domain (BET) protein inhibitors represent another epigenetic approach. BET proteins like BRD4 act as “readers” of acetylated histones and co-activate fibrotic gene transcription [122]. Inhibiting these proteins reduces the proliferation and activation of IPF-derived lung fibroblasts and mitigates fibrosis in mouse models [123]. Critically, in a corneal injury model, the BRD4 inhibitor JQ1 was shown to suppress and even reverse the scarring response [124], indicating that the epigenetic memory of a mechanical injury can be pharmacologically erased.

DNA methylation is another layer of epigenetic regulation involved in mechanical memory. Fibroblasts in IPF often display altered DNA methylation patterns, leading to the silencing via hypermethylation of several genes with known and uncharacterized roles in IPF [125]. DNA methyltransferase (DNMT) inhibitors, like 5-azacytidine or 5-aza-2’-deoxycytidine, can demethylate DNA and reactivate silenced genes. In experimental models, 5-azacytidine was found to attenuate lung fibrosis in neonatal rats exposed to hyperoxia via p16 [126] and alleviate the lung inflammation of bleomycin-induced ALI [127].

Finally, cell and nucleus changes in response to mechanical stimuli may be manipulated by controlling the amount of cholesterol in the cellular membrane. Treatment with anti-PCSK9 antibodies results in a decrease in the stiffness of the nuclear and cell envelope in lung cancer cells, and ameliorate their increase in invasiveness when exposed to mechanical stress [8].

Disrupting the matrix-cell feedback loop

Because the mechanisms sustaining mechanomemory are interconnected, single-target interventions may be insufficient in established disease. A growing body of work therefore supports the rationale for dual or multi-target approaches aimed at disrupting the matrix-cell feedback loop [72, 128, 129]. For example, the natural product derivative triptolide was found to alleviate bleomycin-induced lung fibrosis in mice through multiple mechanisms simultaneously: it reduced collagen expression, enhanced matrix degradation by modulating MMP levels, suppressed lysyl oxidase and blocked the integrin-FAK-YAP mechanotransduction pathway [130]. This “shotgun” effect foreshadows the potential of rational multi-target therapies.

Looking ahead, pairing an ECM-softening agent with a mechanotransduction inhibitor could be particularly effective. A LOX inhibitor (to reduce collagen cross-linking) could be combined with a YAP or FAK inhibitor (to block the cell’s ability to sense and respond to stiffness) to prevent cells from re-stiffening the matrix. Such combinations are already being explored in cancer, where FAK inhibitors are being added to immunotherapies to break stroma-mediated resistance [131]. Similarly, in cardiac fibrosis, simultaneously inhibiting mechanical stress signaling and a biochemical pathway was more effective than either alone at improving heart function [132].

Combination therapies may also need to be deployed in a sequential or phased manner. A fibrotic organ might first require a “debulking” phase with an ECM-degrading agent like relaxin, followed by a “maintenance” phase with a mechanotransduction blocker like a ROCK inhibitor to prevent recurrence. This approach is analogous to cancer treatments where a tumor is surgically removed and then followed by targeted therapies. While speculative, future IPF trials might test a short-term intervention like high-dose relaxin infusion to rapidly reduce lung stiffness, followed by chronic therapy with a drug like nintedanib, a ROCK inhibitor, or an integrin antagonist to stabilize the softer, less activated state. While this kind of multi-step regimen faces significant regulatory and safety challenges, they may be necessary for established fibrosis. Previous studies and trials suggest that once a fibrotic scar is formed, simply blocking a single pathway is unlikely to dissolve it and an orchestrated reset of the entire microenvironment will likely be required.

A further translational complication is the temporal and spatial heterogeneity of human lung disease. In IPF and other progressive fibrotic lung diseases, regions of advanced, architecturally distorted fibrosis may coexist with areas of early remodeling or relatively preserved parenchyma within the same lung. Therefore, phased therapies aimed at first reducing fibrosis burden and then preventing reactivation of mechanical memory may not act uniformly across the organ. Advanced scarred regions may be less reversible, whereas mechanically active remodeling zones may remain more amenable to matrix-directed, mechanotransduction-targeted or epigenetic interventions. Future trials may therefore need to combine regional imaging, circulating or imaging-based biomarkers and spatial tissue analyses to distinguish irreversible fibrotic regions from mechanically active niches that remain therapeutically modifiable.

Future directions

Fundamental mechanisms of mechanomemory

One of the most profound unanswered questions revolves around the precise mechanisms by which mechanical information is stored and recalled by cells, moving beyond the current understanding of merely irreversible effects [14]. If mechanomemory truly implies the retrieval of stored information, the molecular and cellular pathways facilitating this “memory retrieval” are yet to be definitively identified [133].

Specifically, the field needs to unravel the precise mechanisms linking mechanical signals to both short-term and long-term changes in chromatin structure and gene expression patterns. While it is known that forces transmitted to the nucleus can rapidly alter chromatin conformation and gene expression [134], the detailed cascade of events and the specific molecular players that maintain these changes over extended periods remain elusive. Additionally, identifying the optimal mechanical stimuli or combinations of stimuli (e.g., fluid flow, shear stress, pressure gradients, vibration, spatial constriction, crowding, etc.) that precisely regulate mechanomemory is a significant challenge. While some factors like priming duration and frequency of exposure are known to influence memory intensity and duration [133], a comprehensive understanding of the parameters required for precise control is lacking.

The future of mechanomemory research lies in the synergistic integration of various cutting-edge technologies. Microfluidic platforms are increasingly being adapted to mimic in vivo conditions with precise control over mechanical stimuli [135]. These organ-on-a-chip systems can replicate physiological flow rates, shear stress and stiffness gradients, while allowing for real-time observation of cellular behavior in 3D environments [136]. Furthermore, developing organoids with tunable stiffness and incorporating them into microfabricated components can simulate complex tissue environments with controlled mechanical properties [137]. These models could bridge the gap between traditional 2D cell culture and full in vivo systems, offering a more controllable yet physiologically relevant platform for studying mechanomemory.

The enigma of information storage in mechanomemory highlights the need to move beyond correlation to causality. Future studies should aim to understand the multi-scale integration of mechanical signals, from the initial molecular sensing to persisting modifications. Delineating the precise molecular switches that govern memory formation, maintenance and potential retrieval is still necessary to understand the full therapeutic potential of mechanomemory.

Development of novel approaches to test and treat mechanomemory

Mechanical memory, by definition, is a stable or long-lasting cellular state. Erasing it might not be as simple as a one-time reset. The history of epigenetic cancer therapy is informative: drugs like 5-azacytidine or HDAC inhibitors often require continuous or repeated dosing to maintain a less malignant state [138]. Similarly, a fibrotic organ that has been partially “reprogrammed” might require long-term maintenance therapy to prevent relapse. This raises issues of patient tolerance and cost. In IPF, patients already stay on pirfenidone or nintedanib indefinitely; adding another long-term agent could increase pill burden and risk of adverse effects.

Additionally, there are fundamental challenges in translating the concept of memory erasure. Some fibrotic tissue changes may be irreversible if the architecture is destroyed. For example, honeycomb cysts in end-stage IPF won’t regrow normal alveoli just by softening the matrix; so, timing of intervention is crucial. Another factor is disease heterogeneity. Progressive fibrotic lung diseases may arise from different initiating mechanisms, including epithelial injury, immune-mediated inflammation, environmental antigen exposure or autoimmune disease, and the relative contribution of mechanical feedback may vary across these contexts. Identifying those with a strong mechanical component via gene signature could enrich trials for responders. A particularly difficult problem is the lack of biomarkers that specifically report mechanical memory, as opposed to ongoing fibrosis, ECM turnover or fibroblast activation. Circulating ECM fragments and collagen neo-epitopes can provide information about matrix formation and degradation and have been associated with fibrotic progression [139]. In addition, fibroblast activation protein (FAP) and FAP-targeted imaging approaches are emerging as promising tools to detect active fibroblast remodeling in fibrotic lung disease and cancer [140]. However, these markers do not yet establish whether a cell population has retained a mechanically imprinted state after the original mechanical cue has changed or been removed nor whether a therapy has erased cellular memory. On a cellular level, one might examine biopsies for the presence of activated myofibroblasts or nuclear YAP localization as a surrogate of mechanotransduction activity [141, 142]. However, in fibrotic diseases, repeated biopsies are not often feasible, so biomarkers in blood or bronchoalveolar lavages should be prioritized. Still, they may not represent a specific readout of mechanical memory per se. This makes early-phase trials of mechanotransduction-targeted drugs more challenging, since one must rely on general fibrosis markers to confirm target engagement.

Therefore, biomarkers of mechanomemory will likely need to combine readouts of matrix remodeling or fibroblast activation with evidence of persistent mechanotransduction or epigenetic imprinting. Potential candidates include sustained YAP/TAZ nuclear localization, persistent RUNX/AP-1 or TEAD-associated enhancer accessibility, chromatin marks linked to prior mechanical exposure, or transcriptional signatures that remain after mechanical unloading. Such markers should be interpreted as mechanomemory-related only when longitudinal or perturbational data demonstrate persistence beyond ongoing tissue stiffness or active injury.

At present, approaches to normalize ECM stiffness, inhibit mechanotransduction pathways and reprogram the epigenetic memory of mechanical stress have all shown great promise in the lab. Yet, the concept of “erasing” mechanical memory is still in its infancy. The path forward will likely involve carefully designed combination trials, robust biomarkers of mechanics and perhaps, staging of therapy to include an initial matrix-directed intervention followed by sustained signal modulation as summarized in Fig. 3.

Fig. 3.

Conceptual roadmap for future research and clinical translation of mechanical memory. Key priorities include the rational combination of therapies, the development of reliable biomarkers and improved patient stratification by both disease stage and mechanistic origin

Conclusion

Mechanical memory has emerged as a unifying concept linking chronic mechanical stress to long-term cellular changes in diseases ranging from aggressive cancers to fibrosing disorders. Over the past decade, the scientific community has gained a profound appreciation for how the mechanical microenvironment not only triggers immediate biochemical signaling but also imprints stable phenotypic changes in cells. Although this review outlines the principal pathways by which mechanical cues may be memorized at the cellular and epigenetic level, the field still lacks definitive mechanistic studies and longitudinal clinical evidence; rigorous multi-scale studies combining controlled mechanical perturbations with longitudinal single-cell, single-nucleus and spatial profiling will be needed to distinguish transient mechanoresponses from persistent mechanically imprinted cell states. These approaches could identify whether specific epithelial, fibroblast, endothelial or immune cell populations retain transcriptional or chromatin signatures after mechanical unloading or resolution of the initial injury. Nonetheless, encouraging progress has been made in devising strategies to therapeutically target the mechanical axis: approaches to soften or normalize ECM stiffness, inhibit key mechanotransduction pathways (e.g. integrin–FAK–YAP/TAZ signaling), or reprogram the epigenetic marks of mechanical stress have all shown promise in preclinical models. Harnessing such strategies could potentially blunt tumor aggression or halt fibrosis at its roots by resetting aberrant mechanical cues. Yet, translating these innovations into the clinic will require overcoming significant challenges. Robust biomarkers are also needed to measure mechanomemory and track the efficacy of mechanotherapeutic interventions, since assessing mechanical state changes in patients is far less direct than monitoring, for example, a genetic mutation’s activity. Safety and feasibility remain key concerns, especially for agents that globally alter the ECM or epigenome.

The concept of “erasing” or reconditioning a cell’s mechanical memory is still in its infancy, but it represents a bold and necessary frontier for next-generation therapies. By integrating mechanical memory into our understanding of disease pathology, we open the door to novel treatments that could complement traditional biochemical approaches. Preclinical and early-phase translational studies should now test whether combining ECM normalization, mechanotransduction blockade and epigenetic or metabolic reprogramming can more effectively interrupt the self-reinforcing loops that sustain fibrotic or malignant phenotypes. In parallel, improved readouts of mechanical activity and persistent cell-state imprinting will be essential before mechanomemory-targeted strategies can be translated into clinical interventions. Embracing this paradigm shift holds the promise of fundamentally improving outcomes for patients with recalcitrant solid tumors, fibrosis, and other mechanically influenced diseases.

Abbreviations

ALI

Acute lung injury

ARDS

Acute respiratory distress syndrome

AT2

Alveolar type II (cells)

BET

Bromodomain and extraterminal (proteins)

BRD4

Bromodomain-containing protein 4

CAFs

Cancer-associated fibroblasts

DNMT

DNA methyltransferase

ECM

Extracellular matrix

FAK

Focal adhesion kinase

HDAC

Histone deacetylase

IPF

Idiopathic pulmonary fibrosis

LOX

Lysyl oxidase

MAPK

Mitogen-activated protein kinase

MMP

Matrix metalloproteinase

MSCs

Mesenchymal stem cells

NF-κB

Nuclear factor kappa B

PCSK9

Proprotein convertase subtilisin/kexin type 9

PIEZO1

Piezo-type mechanosensitive ion channel component 1

ROCK

Rho-associated coiled-coil containing protein kinase

RUNX2

Runt-related transcription factor 2

TEAD

TEA domain transcription factor

TGF-β1

Transforming growth factor beta 1

TME

Tumor microenvironment

VILI

Ventilator-induced lung injury

YAP/TAZ

Yes-associated protein/transcriptional co-activator with PDZ-binding motif

αvβ6/β1

Integrins alpha-v beta-6/beta-1

Authors’ contributions

Conception and design: M.C. and G.M.A. Literature search and study selection: M.C., N.C.O.Z., and S.M.E.R. Drafting the manuscript: M.C., N.C.O.Z., and S.M.E.R. Critical revision for important intellectual content: All authors.

Funding

Supported by Centro de Investigación Biomédica en Red (CIBER)-Enfermedades Respiratorias (CB17/06/00021), Gobierno del Principado de Asturias (IDE/2024/000797) and Instituto de Salud Carlos III (PI24/01685).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.