Epithelia are formed by a continuous layer of interconnected cells encapsulating organs and lining cavities. Epithelial cells are anchored to the basal lamina or basement membrane, a thin layer of extracellular matrix that provides structural support and signalling cues and sits on top of the underlying stromal tissue, which provides nutritional support and contains nerve terminals and immune cells that exchange signals with the epithelial sheet, capable of actively orchestrating and maintaining adaptive responses in health and disease (Lambrecht and Hammad, 2012).

For decades, researchers relied on microscopy-based morphological criteria to define different epithelial cell types that, combined with tissue architecture, determine the balance between different epithelial functions: protective, absorptive, and/or secretory. As an example, airway ciliated cells were first described in 1837, followed in 1852 by description of cells lacking cilia, loaded with granules, with a narrow stem connected to the basement membrane by a circular structure (goblet cells) and two cell types lacking access to the airway lumen: spherical cells, adjacent to the basement membrane (basal cells) and two layers of elongated cells (intermediate cells) (Figure 1). Remarkably, these early studies already were able to appreciate cell type similitudes between different tissues and proposed basal cells were precursors of the other airway epithelial cell types (for comprehensive historical perspective of airway cell type discoveries, see (Widdicombe, 2019).

Later, development of molecular markers and transgenics offered functional criteria to further define these cell types, genealogies, and functions, and how all these depend on tissue architecture. The gut is an example with clear spatial segregation of division, differentiation, tissue-specific functions, and death. Intestinal stem cells residing at the bottom of the crypts give rise to transient-amplifying cells (Duckworth, 2021). These are pluripotent cells that sequentially differentiate into absorptive (enterocytes) and secretory lineages. The latter gives rise to different cell subtypes achieving pleiotropic functions: i) antimicrobial peptide-producing paneth cells, not present in the colon; ii) mucus secreting goblet cells; iii) enteroendocrine cells releasing hormones, and chemosensory tuft cells (Fre et al., 2005). Epithelial cell differentiation is linked to migration upwards from the crypt to the villus or surface epithelium; except for paneth cells in the small intestine, which remain at the crypt bottom in close connection with stem cells (Garabedian et al., 1997). Cell migration and compartmentalization of crypts and villus is regulated by the activation and/or gradient between different pathways (Wnt, EGF, Notch or BMP), in most cases due to the contribution of pericryptal cells and the sub-epithelial microenvironment (Reynolds et al., 2014; Chen et al., 2019). Finally, differentiated cells at the villus tip will be extruded into the lumen where they finally die, to allow the renewal of the epithelial layer or epithelial turnover (Watson et al., 2009) (Figure 2).

In recent years, single cell and spatial transcriptomics have redefined and expanded cell types in virtually all tissues analysed, highlighting commonalities and tissue-specific features, echoing Waymouth Reid’s conclusion that “it is extremely probable that several varieties of such [secreting] structures exist” and greatly contributing to the description of cellular complexity of the intestinal and respiratory epithelia. Single-cell RNA sequencing (sc-RNA-Seq) has confirmed the suggested variability in terms of cell composition and heterogeneity between organs and regions, e.g., small intestine vs. colon, crypt vs. villi, different airway regions, like trachea, airways or alveoli (McKinley et al., 2017; Beumer et al., 2018; Montoro et al., 2018; Moor et al., 2018; Plasschaert et al., 2018; Burclaff et al., 2022).

Additionally, sc-RNA-Seq has identified previously uncharacterised rare types of tissue-specific cells (e.g., lung ionocytes (Montoro et al., 2018; Plasschaert et al., 2018) and ones shared by different epithelia (e.g., tuft cells in airways, gastrointestinal tract, and other tissues (Elmentaite et al., 2021). Importantly, these techniques have demonstrated that the transcriptional profile between different cell subtypes, and thus our distinction between secretory and absorptive (gut) or ciliated types (airways), is not as clear as previously thought. Examples of this are colonic deep secretory cells contributing to the stem cell niche but with classical markers of differentiated goblet cells (Parikh et al., 2019) or mucous-ciliated and suprabasal cells in the airways (Elmentaite et al., 2021). Moreover, these techniques enable tracking of cells transitioning between states (trajectories), identifying new regulatory roles for Sox4, Foxm1, Mxd3, Batf2 in enterocytes (Haber et al., 2017) or Foxi1 in airway ionocytes (Montoro et al., 2018); as well as segregated populations within a given trajectory, such as tuft-2 cells displaying immunological functions (Haber et al., 2017). By enabling comparison between airway states (development, homeostasis, disease) these techniques have shed light on disease mechanisms like a general upregulation of secretory gene expression in all asthmatic airway epithelial types, in addition to a novel intermediate mucous-ciliated cell state expressing markers of both classic cell types that, with goblet cell hyperplasia, contributes to mucous hyperplasia in asthma. In the gut, the same approach has also identified defective mucus maturation in goblet cell as a potential driver of IBD and colorectal cancer, in addition to a new pH-sensing absorptive cell type, pericryptal stromal signalling, lymphocyte imbalance, and platelet aggregation as key contributors to barrier dysfunction in IBD (Regev et al., 2017; Beumer et al., 2018; Kinchen et al., 2018; Huang et al., 2019a; Parikh et al., 2019; Vieira Braga et al., 2019; Deprez et al., 2020; Jackson et al., 2020; Travaglini et al., 2020; Beumer and Clevers, 2021; Elmentaite et al., 2021; Haniffa et al., 2021; Tang et al., 2022).

All this demonstrates how recent advances in genomics, cell lineage tracing, and sc-RNA-Seq have revealed not only the need to redefine the meaning of cell identity, but also have uncovered new cell types involved in epithelial homeostasis and disease.

Cell-cell junctions weave single epithelial cells into a functioning and dynamic monolayer that acts as a polarized barrier while selectively allowing transepithelial movement of water, ions, and macromolecules. Physiological transepithelial transport is classified as transcellular (mediated by transporters in apical and basolateral membranes) and paracellular transport (mainly determined by tight junctions). In the later, pore and leak pathways act in an interdependent manner (Weber et al., 2010). All these aspects have been nicely reviewed recently (Horowitz et al., 2023). Conversely, in damaged epithelia, transport becomes unrestricted and unselective, even allowing passage of bacteria from the lumen to the underlying tissue.

According to their location, composition, and function, epithelial intercellular junctions are classified as tight junctions (TJs), adherens junctions (AJs) or desmosomes, but all have common features like transmembrane components that physically link neighbour cells, in complex with cytoplasmic scaffolding and adaptor proteins linking the junctions to the cytoskeleton, which confers them mechanosensitivity (Garcia et al., 2018; Beutel et al., 2019; Pannekoek et al., 2019; Angulo-Urarte et al., 2020; Haas et al., 2022). TJs are formed by homotypic claudin and occludin contacts at the apex of lateral membranes between contacting cells. TJs form a regulable belt around a cell, separating the apical and basolateral membrane domains, while also sealing the paracellular pathway to control water and solute diffusion. The cytoplasmic side of TJs binds to adaptor proteins (ZO-1, -2, -3, cingulin) that interact with microtubules and the cytoskeleton. AJs are formed by the calcium-dependent extracellular trans binding of cadherins and force-dependent cytoplasmic binding to actin and microtubules via catenins. AJs are essential for cell-cell adhesion and epithelial mechanical responses, detailed later. Desmosomes are strong intercellular junctions based on cadherins desmoglein and desmocolin, bound to intermediate filaments via catenins plakoglobin and plakophilin. Moreover, junctions act also as signalling hubs, in close interconnection with Rho GTPases (Citi et al., 2014). Small GTPases are frequently found inactive, bound to GDP. After GDP-GTP replacement by Guanine Exchange Factors (GEFs), GTPases are recruited and interact with effector proteins, regulating essential cell functions controlling cell-cell adhesion and barrier function like mechanotransduction, vesicle trafficking, or junctional component dynamics (Braga, 2018). In summary, junctional integrity is essential for epithelial function, and its disruption is a key aspect of diseases like asthma and IBD.

In fact, mechanotransduction between epithelial cells determines tissue homeostasis at different levels. External forces (breathing, circulation flow, peristaltic movements), GTPases, and cytoskeletal contractility control long-term biological outcomes at cell (identity, proliferation, migration, extrusion) and tissue levels (folding, compartmentalization) in development, differentiation, homeostasis, and repair at the cell (identity, proliferation, migration and extrusion) and tissue levels (folding, compartmentalization) (Mahoney et al., 2014; Zhao et al., 2014; Goodwin and Nelson, 2021; Alvarez and Smutny, 2022; Perez-Gonzalez et al., 2022; He et al., 2023). Architecture of the gut epithelium represents a good example in this context; thus, myosin contractility initiates crypt invagination, the GTPase Rac1 controls crypt-villus compartmentalisation, and mechanical tension drives homeostatic intestinal cell migration from crypts to villus (Sumigray et al., 2018; Yui et al., 2018; Krndija et al., 2019; Perez-Gonzalez et al., 2021; Yang et al., 2021; Perez-Gonzalez et al., 2022).

As mentioned, mechanical forces also regulate cell identities, frequently via the transcriptional regulator YAP and its interplay with other signalling pathways, with remarkable tissue-specific features. In the gut, stiffening decreases stemness and promotes YAP-dependent gut stem cell differentiation into goblet cells (He et al., 2023); whereas in the lung, YAP is essential to maintain tissue organization and prevent stem cell loss and excessive goblet cell differentiation and mucin hypersecretion during homeostasis (Mahoney et al., 2014; Zhao et al., 2014; Hicks-Berthet et al., 2021). These differences could be partly explained by YAP being essential in all regenerative scenarios and lung homeostasis, but not in gut homeostasis (Camargo et al., 2007; Barry et al., 2013; Zhao et al., 2014; Yui et al., 2018; Hicks-Berthet et al., 2021).

Mechanical forces also regulate cell numbers in shorter time scales. Cell stretching signals through E-cadherin and Piezo1 to increase nuclear levels of YAP and β-catenin and CDK1 activity, driving cell cycle re-entry (Streichan et al., 2014; Benham-Pyle et al., 2015; Gudipaty et al., 2017; Uroz et al., 2018). Conversely, crowding or compression arrests cell cycle and restores homeostatic cell numbers via cell extrusion, an evolutionarily conserved mechanism where a supracellular actomyosin cable formed around the unwanted cell ratchets in and down, resulting in seamless cell eviction without compromising barrier function (Rosenblatt et al., 2001; Eisenhoffer et al., 2012; McClatchey and Yap, 2012; Puliafito et al., 2012). In that sense, extrusion also works as an innate defence mechanism against external aggression, with healthy cells collectively squeezing cells infected by bacteria or viruses, thus limiting pathogen spreading in the monolayer and ensuring epithelial barrier function (Bastounis et al., 2021; Hippee et al., 2021; Lin et al., 2021; Moshiri et al., 2023).

Both the gut and the respiratory epithelium luminal surface are protected by mucus, a selective barrier to particles and molecules that is built around a family of polymeric glycoproteins called mucins. Mucus that coats the epithelium is a complex hydrogel biopolymer barrier, present not only in the airways and the gastrointestinal tract, but also in the reproductive tract and eyes (Lieleg and Ribbeck, 2011). During homeostasis, the protective mucus layer is produced by the goblet cells that are equipped with specific biological machinery for the secretion of mucins. Notably, some respiratory diseases are characterised by changes in goblet cells function (like asthma or COPD, see Table 1) and we will discuss later the importance of mucus production dysregulation in the pathology of asthma and IBD.

In the gut, mucus offers moisturising and lubricant properties, protecting the epithelial cells from dehydration and mechanical stress during the passage of luminal content and peristalsis forces (Johansson et al., 2013). It also operates as a surface cleaner, removing debris and bacteria, through binding, collecting, and flushing them away via intestinal flow. The small intestine has a single layer of mucus; while in the stomach and colon, the mucus layer is composed by an inner layer, attached to the epithelium, and an outer layer that interacts with luminal components. The inner layer is impermeable to bacteria and renewed by globlet cells every hour. The outer mucus layer is less dense and is the habitat for commensal bacterial (Hansson, 2019). Notably, in the small intestine, mucus leaves pores that allow the bacteria to penetrate, which is not the case in the large intestine, where the mucus layer is thick and completely avoids the contact with bacteria and the epithelial cells (Paone and Cani, 2020). In the gut, the main mucin is MUC2, which composes the skeleton of the mucus layer. In addition to MUC2, the IgG Fc-binding protein, FCGBP and the intestinal trefoil factor, TFF3 act synergistically to enhance the mucus barrier and exert antibacterial effects, while the metalloenzyme CLCA1 is involved mainly in the stratification and expansion of mucus. Moreover, ZG16, RELMβ, Lypd8, sIgA, and AMP exert bacteriostatic or bactericidal effects under different conditions (Song et al., 2023).

In the airway, mucus is composed of water, different proportions of polymerizing mucin glycoproteins MUC5B and MUC5AC in proximal versus distal regions (Meldrum and Chotirmall, 2021), a range of antimicrobial molecules (defensins, lysozyme, etc.), cellular debris including DNA, and protective factors (trefoil factors) (Thornton et al., 2008). The protective response is driven by microbial sensors in the goblet cells that initiate secretion of mucus, to entrap invading microbes and remove bacteria away through mucociliary clearance (Abdullah et al., 2018). The ciliated cells, which line the surface epithelium of the airways, provide the force necessary for mucociliary clearance by the coordinated beating of their cilia, which confers an escalator motion to bring unwanted material to the mouth to be coughed out. These highly specialized cells are therefore critical to the health and function of the pulmonary system, and often preferential destroyed in favour of mucus producing cells in pathologies like asthma, with mucus hyper-production and -secretion remaining a massive obstacle in asthma treatment.

Many diseases arise from an imbalance between mucus production and elimination. The role of mucus and mucins in diseases of the intestinal and respiratory tracts is excellently reviewed by Hansson and others (Hansson, 2019) and we will describe later the specific importance of mucus regulation in asthma and IBD. We have included in Table 1 other diseases showing specific alteration linked to respiratory or intestinal diseases and the link to different respiratory and digestive diseases. See also (Meldrum and Chotirmall, 2021) for additional information.

It is now clear that the maturation and function of the mucus layer are strongly influenced by the microbiota (Schroeder, 2019). In fact, the consideration of the microbiota as a continuous element of homeostatic regulation of the epithelium has undoubtedly made physicians and researchers to confirm the relationship between the microbes and the epithelial barrier, and to adopt a more holistic view of the disease (Runge and Rosshart, 2021).

On of the main factors that influences the presence of a specific microbiota is the composition of the mucosal layer. Indeed, the mucin glycosylation profile influences the composition of mucus-associated bacteria, selecting specific species (Bergstrom and Xia, 2013). The composition of the mucus not only controls bacteria adhesion, but mucin glycans can also serve as nutrients for specific microorganisms, depending on their glycan-degrading enzyme’s content, highlighting an example of how the host controls the microbiota within the mucus layer (Paone and Cani, 2020). Finally, bacteria can use host glycans to form new polymers used in the creation of their capsule, promoting evasion from the immune system (Martens et al., 2009).

Factors like age, diet, drugs, or disease affect microbiota composition too, even compromising its barrier function. For example, during pulmonary infection, microbial dysbiosis leads to invasion by opportunistic pathogens. These communities disrupt tissue compartments within the airway lumen, including mucus and causing progressive, localized, and chronic infection, particularly in pulmonary diseases (Montassier et al., 2023). Moreover, in asthma, exacerbations are classically induced by infections, like the ones produced by P. aeruginosa, which disrupts pulmonary mucins significantly contributing to disease progression (Meldrum and Chotirmall, 2021). Although the association between IBD and dysbiosis is accepted, whether alterations of the microbiota represent a cause or consequence of the disease is still a matter of discussion (Palm et al., 2014; Forbes et al., 2016; Schaubeck et al., 2016; Becker et al. 2015).

Epithelial cells produce and secrete several molecules that contribute to epithelial integrity and elimination of microorganisms and contaminants, as well as intercellular communication. These molecules, collectively known as “the secretome” support epithelial homeostasis by controlling important cellular processes like proliferation, different mechanisms of cell death, safeguarding of epithelial tight junctions, maintenance of a healthy microbiota, and of course, communication with other cell types, like immune or stromal cells. We will cover some of the key players in the epithelium secretome in the gut and lung and we have summarised their main function in Figure 3.

In the second century, Aretaeus of Cappadocia described asthma as aazein: a short-drawn breath or panting, a death rattle. Aretaeus went on to describe the defining characteristic of all asthmatics, the attack: “…they open the mouth since no house is sufficient for their respiration, they breathily standing, as if desiring to draw in all the air which they possibly can inhale … ”, and if the symptoms abate, he concludes, “the asthmatic escapes death, but in the intervals between severe attacks or even when they are walking on ground level, they bear in mind the symptoms of the disease (Karamanou and Androutsos, 2011).” The haunting trauma of an asthma attack is echoed in Henry Salter’s account, “…not only is asthma not an uncommon disease, but it is one of the direst suffering; the horrors of the asthmatic paroxysm far exceed any acute bodily pain”. The language defining asthma has remained abstract and scantily more informative through the centuries. And it was not till 2017, when the Global Initiative for Asthma (GINA) defined asthma as, “a heterogenous disease usually characterised by chronic inflammation. It is defined by a history of respiratory symptoms such as wheeze, shortness of breath, chest tightness, and cough that vary over time and in intensity, together with variable expiratory airflow limitations.”

Today, asthma affects more than 300 million people globally, at a staggering financial cost and a burden to quality of life and remains one of the most common, non-communicable diseases (Papi et al., 2018; Pavord et al., 2018; Porsbjerg et al., 2023). Significant advances have been made in asthma care, as hospital admissions and deaths due to asthma are on the decline since the 1990s. The majority of asthma sufferers present with a type 2 inflammatory response and profile characterised by hyper-production of IL4, IL5, and IL13, increased blood eosinophils and fractional exhaled nitric oxide (FeNO) (Porsbjerg et al., 2023). However, current treatments only manage symptoms and have little-to-no effect on the natural progression of this disease. Even the diagnosis and term itself is umbrella, widely understood by clinicians that asthma could represent manifold pulmonary diseases. The use of inhaled corticosteroids became aggressively prescribed in the late 1980s, which resulted in fewer exacerbations and better control of patient symptoms and mortality. This biased physicians and researchers to approach asthma as a chronic inflammatory disorder, where disease symptoms are to be managed but not cured. This “inflammatory-centric” approach was implemented with wilful disregard that asthma attacks, or airway hyperresponsiveness, the sentinel event of all asthmatics, can occur in individuals without inflammation. Further, the degree of inflammation and the types of inflammation effecting asthmatics (eosinophilic, non-eosinophilic, high-type 2, low-type 2, etc.) is well documented to be highly variable (Hammad and Lambrecht, 2021; Porsbjerg et al., 2023). To this, commissions have been gathered to address the problem of asthma, focusing on the outdated thinking and antiquated research practices governing its treatment and prevention (Papi et al., 2018; Pavord et al., 2018; Porsbjerg et al., 2023). Recently, the Lancet compiled a commission to redefine asthma with the aim, “…to identify entrenched areas of asthma management and treatment in which progress has stalled and to challenge current principles … “We believe that the most important cause of this stagnation is a continued reliance on outdated and unhelpful disease labels, treatment and research frameworks, and monitoring strategies, which have reached the stage of unchallenged veneration and have subsequently stifled new thinking (Pavord et al., 2018).”

It has long been speculated that epithelial loss or damage in asthma studies is due to artefacts from the harvesting and processing protocols while obtaining and analysing tissue samples (e.g., bronchial brushings and biopsies). However, an ever growing number of studies are revealing the loss, damage, and dysregulation of the epithelium in all asthmatics (Payne et al., 2003; Pohunek et al., 2005; van Rijt et al., 2011; Papi et al., 2018; Hammad and Lambrecht, 2021). Loss of the superficial epithelial layer, preferential destruction of ciliated cells, and over expression and activation of EGFR with increases in growth factors, including TGFβ (Hoshino et al., 1998; Shahana et al., 2005; Boxall et al., 2006; Holgate, 2007), are found in the majority of asthma suffers; even occurring in mild, early, and non-fatal asthma. Discussed earlier, damaged epithelium releases a number of soluble mediators promoting remodelling and inflammation (e.g., TSLP, IL25, and IL33), and are not only highly expressed in asthmatic airways, but represent genetic loci identified in a number of genome-wide association (GWA) studies correlating with asthma susceptibility (Cookson, 2004; Allakhverdi et al., 2007; Grotenboer et al., 2013; Moheimani et al., 2016). As an example, Steven Holgate’s group demonstrated that asthmatic children have damaged epithelium with increased expression of EGFR, that was significantly correlated with basement membrane thickness (an important pathological feature of adult asthma), by excessive deposition of collagen III, seen in the absence of eosinophilic inflammation (Fedorov et al., 2005). Moreover, using bronchial biopsies from healthy and asthmatic cohorts, Barbato et al., showed loss of epithelium, increase in angiogenesis, and basement membrane thickening in asthmatic children prior to a mounted inflammatory state (Barbato et al., 2006). In an even earlier study, Marguet and co-workers found increased numbers of epithelial cells in the bronchoalveolar lavage fluid from asthmatic children compared to health controls (Marguet et al., 1999), further suggesting that epithelial loss and damage-not present at birth-is occurring before-or-at disease conception, and likely initiating and sustaining the adaptive response characteristic of most asthmatics.

As reviewed before, the epithelium can act as a barrier through the cooperative action of cell junctions with the cytoskeletal apparatus, essential for barrier function and downstream signalling. Dysregulation of the junctions themselves can orchestrate pro-inflammatory signalling pathways, fuelling an inflammatory cascade and feed-forward mechanisms initiated by the wounded barrier. A consequence of barrier damage, is the release of pro-inflammatory factors (e.g. alarmins), known to elicit a type-2 response resulting in increased IL4 and IL13 in airways that are now appreciated to also perpetuate junction dysfunction by downregulation of claudins, occludin, JAM proteins and ZO-1 (Ahdieh et al., 2001; Ortiz-Zapater et al., 2022a). House dust mite (HDM) extract, one of the major causes of asthma (and asthma exacerbations) in children, contains proteases that are known to cleave junctional proteins including occludin and ZO-1, directly participating in barrier dysfunction. Notably, Tan et al., demonstrated that three chronic HDM experimental asthma mouse models, with distinct inflammatory profiles (eosinophilic, neutrophilic, and mixed granulocytic), all had decreased expression of claudin-5, -8, -18, and -23, ZO-1, and occludin, further suggesting that a dysfunctional epithelium is activating and maintaining inflammatory pathologies rather than inflammation as the initial source of epithelial wounding (Tan et al., 2019). This has been recapitulated in human bronchial epithelial cells in air-liquid interface (ALI) culture systems, and bronchial brushings from asthmatic patients. Downregulation of E-cadherin alone resulted in an EGFR-dependent, type 2-biased inflammatory response, and claudin-18 deficiency was demonstrated to promote barrier dysfunction in asthmatic mice and human epithelial cells (Heijink et al., 2007). An ultra-structural analysis of bronchial biopsies of both allergic and non-allergic asthmatics showed junctions and desmosomes damaged, as well as the destruction of ciliated cells in favour of goblet cell hyperplasia and impaired wound healing, with increased basement membrane thickening (Shahana et al., 2005).

The destruction of barrier proteins results in the activation of signalling pathways promoting asthmatic inflammation while directly inhibiting barrier function through the decreased expression of junctional proteins providing a viscous feed-forward cycle of wounding, repair, inflammation, and re-wounding. This highlights the need for therapeutics that are targeted to maintain barrier proteins and function in chronic disease, such as asthma and IBD.

There is an established notion that chronic inflammation results in airway hyper-responsiveness, and numerous studies have demonstrated that high doses of oral and inhaled corticosteroids are unable to stop, nor reverse, asthma exacerbations (Childhood Asthma Management Program Research Group et al., 2000; Kips et al., 2000; Guilbert et al., 2006; Porsbjerg et al., 2023). Ultimately, bronchoconstriction is the result of airway remodelling and as we have discussed above, when the epithelium is damaged and junctional proteins disrupted, downstream signalling occurs to respond to assaults; this is true of mechanical forces applied to monolayers. Unique to the lung (and the heart) is that at birth its movements, required for respiration, will not cease until death, causing the lung to be under constant, and constantly changing mechanical forces. Indeed, these forces are required for healthy lung development in utero and after birth, and regulated repair responses (Liu et al., 2016; Li et al., 2018). As earlier discussed, mechanical forces govern epithelial numbers within a monolayer. When crowded regions experience compression, unwanted cells are removed by extrusion to regain homeostatic densities, relieving mechanical stresses (Bagley et al., 2023; Eisenhoffer et al., 2012). Airway epithelium during bronchoconstriction will experience dramatic compressive forces, likely causing excessive cell extrusion, damaging the epithelium, while losing barrier function, and promoting further inflammation (Bagley et al., 2023). Importantly, the mechanically-activated protein YAP1 is well-characterized in airway homeostasis and disease, required for proper airway branching (Lin et al., 2017), maintenance, size regulation, and identity of epithelial cells. Mechanical forces are deeply integrated and unavoidably required for all biological aspects needed for lung development, homeostasis, and pathology, and these mechanically-activated epithelial pathways represent a novel, druggable target in wound repair and disease.

The compressive forces applied to the epithelium during an asthma attack is estimated to be about 30 cm H₂O, at least an order of magnitude greater than the forces felt during normal respiration (Park et al., 2015). Stealing a line from Chris Grainge’s review on airway mechanical compression, “Bronchoconstriction is not only a symptom of asthma but is also a disease modifier” (Veerati et al., 2020). It has now been demonstrated that compressive forces, in vitro and vivo, lead to expression of genes known to elicit pathological responses in lung disease, including early growth response-1 (EGFR-1), platelet-derived growth factor (PDGF), and TGFβ. Stimulation of repair response pathways through EGFR activation and down-stream signalling, leads to the release of growth factors (e.g., TGFβ) and ECM components (collagens) involved in airway remodelling and disease progression (Ressler et al., 2000). Incubating fibroblasts with conditioned medium from compressed airway epithelial cells resulted in increased collagen deposition, all in the absence of an inflammatory component (Tschumperlin et al., 2003). Park et al. nicely demonstrated that repeated compressive forces alone, over a relatively short time period, could elicit mucus production, e.g., Muc5AC, in normal human bronchiole epithelial cells that was dependent upon EGFR and TGFβ2 (Park and Tschumperlin, 2009). This work was confirmed in humans: volunteers underwent methacholine challenges (only three times over 4 days) to induce bronchoconstriction that lead to increases in TGFβ, collagen, and mucus production in airway epithelial cells, also in the absence of an inflammatory response (Grainge et al., 2011). This is important as mucus hyper-production and secretion remains an intractable problem in many pulmonary disorders, including asthma. Indeed, in a study of 93 fatal asthma cases, near all had mucus obstructions in their airways, where half had more than 80% of airways occluded with mucus plugs (Aegerter and Lambrecht, 2023). Now we appreciate that this mucus problem is not simply a result of goblet cell hyperplasia but also, the expression of mucus in bona fide ciliated cells, which is important if we are to develop effective and targeted therapeutics currently missing in today’s clinics. Finally, wounding epithelium itself can induce airway smooth muscle constriction, actively participating in the airway compression-remodelling response. Elegant work by Steven George’s group used ex vivo lung slices from rats and laser ablation to destroy signal airway epithelial cells that resulted in a 70% reduction in airway lumens within seconds of cell wounding, followed by further airway smooth muscle contractions over minutes, again in the absence of an inflammatory response by inflammation (Zhou et al., 2012).

Currently, asthma therapy is big business with annual revenues in the billions, which is on the rise, as all these medications can do is manage symptoms of this common disease, not capable of stopping or reversing its progression. This sentiment is not new, and indeed clinical trials targeting the epithelial-derived alarmins, released by wounded epithelial barriers, have shown promising results. The monoclonal antibody inhibiting TSLP, Tezepelumab, has been demonstrated to significantly supress all three type-2 clinical biomarkers for asthmatics: peripheral blood eosinophils and total IgE (Schleich et al., 2024); while the anti-IL25 and IL33 drugs, Brodalumab and Itepekimab, respectively, were less successful (Chan et al., 2022). Promising work by Wawrzyniak and others in primary human cells were able to reconstitute barrier function, damaged by IL4 and IL13 exposure, from asthmatic patients by inhibiting histone deacetylases (upregulated in asthma), resulting in junctional protein synthesis (Wawrzyniak et al., 2017). Lastly, as we have discussed, mucus is a problem in many pulmonary diseases, and asthma is no exception, with little treatment options available. However, using a mouse model of IL13-induced mucus hyperplasia and primary cells from asthmatics, inhibiting the heat shock protein 90 (HSP90), upregulated in asthma with geldanamycin, blocked, and even reverted, mucus hyperproduction and goblet cell hyperplasia (Pezzulo et al., 2019). In fact, there are ongoing clinical trials for HSP90 inhibitors for various disease (Kitson and Moody, 2013).

To expand upon current asthma treatments and experimental approaches, we need pathophysiologically relevant platforms that allow for efficient and effective drug discovery and development. In the 1990s, Martin Sanders began iconoclastic work in the use of precision cut lung slices (PCLSs) to study lung physiology and pathology (Martin et al., 1996) that has snowballed over the last three decades, as more researchers are being introduced to the power of this ex vivo system in basic cell biology and translational studies (Davies et al., 2015; Alsafadi et al., 2017; Huang et al., 2019b; Lam et al., 2023). PCLSs are thin sections of live tissue containing all resident cell-types, while maintaining proper tissue architecture, preserving cell-to-matrix relationships, within complex, interconnected cellular hierarchies, which make up all tissues and organs. Ex vivo lung slices have been successfully used in studies from, mice, rats, pigs, sheep, non-human primates, and humans (Alsafadi et al., 2020). They have been used to study airway and arteriole contraction (Martin et al., 1996), tumour biology within intact tissue (Davies et al., 2015), viral infection (Rosales Gerpe et al., 2018), HDM-induced asthma (Ortiz-Zapater et al., 2022a), and fibrosis (Alsafadi et al., 2017), with seemingly endless potential in novel therapeutic development (Liu et al., 2021; Lam et al., 2023). Importantly, the use of ex vivo tissue slices reduces the ethical burden for in vivo models, because dozens of slices can be obtained from a single lung, decreasing the number of animals needed, and allowing for multiple treatments assessed in a lone animal. And PCLSs are amenable to many live and fixed imaging techniques (including watching an asthma attack in real time), as well as genetic, biochemical, and molecular biology analyses. An important limitation to PCLSs is that viability decreases in culture conditions over time (usually 7–14 days). Therefore, ex vivo modelling of chronic diseases or assessing treatments to reverse established pathologies can be limited, requiring the development of better culturing conditions to overcome this problem. Regardless, the power of PCLSs to bridge disease characterization in animal models, and humans, with translational research and positive clinical outcomes is undeniable.

In order to maintain tissue homeostasis, the intestinal epithelium acts as a physical and immunological barrier separating the lumen, which contains the microbiota, and the host. “Epithelial leakage”, a common feature in IBD, allows for the invasion of luminal components, which can activate immune cells located at the sub-epithelial space contributing to intestinal inflammation (Martini et al., 2017). Abnormalities in epithelial barrier function can be reflected by an increased permeability, which has been observed in small bowel and colon in CD patients (Jenkins et al., 1988), and has been correlated to the degree of inflammation (Jenkins et al., 1987; Sanderson et al., 1987; Suenaert et al., 2002; Turpin et al., 2020). Moreover, it has been shown that increased intestinal permeability in IBD patients in remission can predict the occurrence of relapse or flares (Wyatt et al., 1993; Irvine and Marshall, 2000; Tibble et al., 2000; Vivinus-Nebot et al., 2014). Together, the occurrence of epithelial barrier dysfunction before the outbreak of the inflammatory response supports the hypothesis of epithelial defects as etiological factors in IBD pathogenesis. GWAS studies identified genes linked to altered barrier function to be associated to IBD; including genetic variants of CARD15/NOD2 gene, resulting in severe forms of CD (D'Inca et al., 2006; Buhner et al., 2006). In fact, several genes relevant to epithelial barrier function have been categorized as IBD loci, such as HFN4, CDH1 and LAMB1 in UC (Consortium et al., 2009). In agreement, several animal models demonstrate that epithelia permeability precedes the development of intestinal inflammation, for example, the IL10 KO (Madsen et al., 1999), and the SAMP/YitFc mouse (Olson et al., 2006), as well as the mouse strain deficient for the xenobiotic transporter mdr1a (Resta-Lenert et al., 2005). The etiological role of epithelial leakage in inflammation is further supported by IBD-like phenotypes in patients suffering from monogenic diseases. For instance, the very-early onset IBD called Tufting enteropathy is caused by mutations in EpCAM, leading to cell-cell contact disruption (Sivagnanam et al., 2008). Although these and other data support the epithelial contribution to the onset and progression of IBD, the low IBD-like phenotype penetrance of these monogenic diseases indicates the existence of functional redundancy between different proteins/pathways within the enterocyte, and/or the requirement for non-epithelial factors for the onset of intestinal inflammation. This idea was indeed confirmed by other mouse models targeting TJ proteins, showing that a leaky barrier is not sufficient to trigger intestinal inflammation, such as JAM-a deficient animals (Khounlotham et al., 2012), or transgenics mice with expression of claudin-2 in IECs (Ahmad et al., 2014).

As mentioned above, increased epithelial permeability is a hallmark of patients suffering from IBD (Teshima et al., 2012). In 2007, Zeissig et al. described upregulation of the pore-formin Claudin-2 and downregulation and/or redistribution of claudin-5, -8 and occludin as the main alterations affecting the apical junctional complex (AJC), and thereby contributing to impaired barrier function in CD (Zeissig et al., 2007). In UC, claudin-2 is also upregulated, while the barrier forming claudin-4 and -7 are downregulated (Oshima et al., 2008), as well as occludin (Heller et al., 2005). In IBD or immune-driven colitis the upregulation of claudin-2 can be attributed, at least partially, to the increased levels of several proinflammatory cyotkines, such as IL13 (Heller et al., 2005). Conversely, Myosin Light Chain Kinase (MLCK) activation causing phosphorylation of MLC and occludin endocytosis contribute to permeability mediated by the leak pathway (Clayburgh et al., 2005; Marchiando et al., 2010; Van Itallie et al., 2010). Previously mentioned, occludin is downregulated in IBD patients (Heller et al., 2005; Kuo et al., 2019), which can be triggered by cytokines such as TNF (Su et al., 2013) or LIGHT (Schwarz et al., 2007). Additionally, the tricellular TJ proteins tricellulin (Krug et al., 2009; Saito et al., 2021) and angulin-1 (Sugawara et al., 2021) also contribute to the leak pathway permeability. Recent studies also pointed to a downregulation of tricellulin expression in UC patients (Krug et al., 2018). Indeed, in vitro studies have shown that the pro-inflammatory milieu in the inflamed gut of IBD patients can lead to alterations on several proteins within the AJC, as upon stimulation with IL1β (Al-Sadi et al., 2008), IL6 (Suzuki et al., 2011), IL4 and IL13 (Ceponis et al., 2000), TNF-α (Ma et al., 2004) or IFN-γ (Madara and Stafford, 1989; Wang et al., 2005).

Regulated cytoskeleton function is crucial for TJ assembly and epithelial barrier function, In fact, transcriptional regulation of ACF7, a cytoskeleton crosslinking protein, is observed in UC patients (Ma et al., 2017). Accordingly, mice with an epithelial-specific knockout of non-muscle MyosinIIA suffer from increased intestinal permeability, low scale mucosal inflammation, and increased susceptibility to experimental colitis (Naydenov et al., 2016). Cell stress can also induce changes in actin dynamics and affect actin-binding proteins, such as Villin-1 and Gelsolin, which in turn control survival of Intestinal Epithelial Cells (IECs) and barrier function (Roy et al., 2018). Recent in vivo studies demonstrated that prenylation of Rac1 and RhoA, tightly associated to the cytoskeleton, significantly contribute to epithelial barrier function in the gut, and this correlated with alterations of its expression and/or subcellular localization in the intestinal epithelium of IBD patients (Lopez-Posadas et al., 2016; Martinez-Sanchez et al., 2022).

Beyond structural defects, IBD is associated with changes in the epithelial secretome. IBD has been associated to defects of goblet cell differentiation, supporting the key role of the mucus in the intestine (Gersemann et al., 2009). Indeed, the composition of gut mucus is altered in IBD, which is depicted by reduced TFF3, expression diminished levels mucin2 and reduced mucus sulfatation (Tytgat et al., 1996). Focusing on alarmins, TSLP and IL33 have been suggested as protective molecules in IBD (Taylor et al., 2009). Although UC patients show reduced expression of TSLP (Tahaghoghi-Hajghorbani et al., 2019); controversial data are available concerning IL33 in CD and UC (Seidelin et al., 2010; Tahaghoghi-Hajghorbani et al., 2019).

The increasing acceptance of the causative role of epithelial-derived mechanisms in IBD pathogenesis is also reflected by the current effort to exploit this in the clinic, both for treatment and diagnosis of chronic inflammatory diseases. Traditional sugar permeability assays (Meddings and Gibbons, 1998; Teshima and Meddings, 2008) are giving way to molecular imaging, such as confocal-laser endomicroscopy (CLE) using a tracer dye to assess intestinal permeability. This technique permits the identification of epithelial gaps (Kiesslich et al., 2007), even correlating to the occurrence of relapses (Kiesslich et al., 2012) and the identification of subclinical lesions in IBD (Lim et al., 2014; Zaidi et al., 2016). Using this CLE, a recent cross-sectional diagnostic study demonstrated the superiority of barrier healing (versus endoscopic/histologic remission) for the prediction of adverse outcomes in CD and UC, validating epithelial leakage as a prognostic marker of the disease (Rath et al., 2023). In order to overcome safety limitations of CLE, a multimodal imaging label-free imaging technique has been used to assess intestinal permeability in UC patients (Quansah et al., 2023). Despite these advanced imaging techniques, there is a clear need for the identification and validation of non-invasive methods for the diagnosis of “leaky gut”. Although several biological markers have been suggested in this context [plasma/serum citrulline, FABP-2, alpha-GST or zonulin; urine claudin-3; or faecal defensins (Bischoff et al., 2014)], none of them has been efficacious in disease prognosis or progression. We believe that the use of CLE (or alternative imaging methods) alone, or in combination with other standard methods (endoscopy/histology), and the identification of biomarkers for impaired intestinal permeability, will allow us to define the functional state of epithelial integrity and contribute to the prediction of IBD flares.

Currently, the clinical management of IBD strives to control symptoms and mucosal healing (Neurath and Travis, 2012). However, the lack of response to therapy and the low safety profile of immunosuppressive drugs implies the need of alternative therapies, and epithelial restoration emerges as a key component to achieve mucosal healing in IBD, with the final objective of achieving sustained clinical remission, reduced rate of surgery and lower incidence of long term complications. Thus, new knowledge of epithelial dysfunction would likely impact IBD clinical management.

Therapeutic strategies based on promoting the integrity of TJs might have a beneficial effect in IBD. For instance, the zonulin inhibitor AT-1001 (lazarotide) impairs TJ disassembly due to cytoskeleton rearrangement and ameliorates experimental colitis in mice (Arrieta et al., 2009; Sturgeon et al., 2017). Inhibition of the pore function from TJ can limit increased paracellular permeability, which can be achieved for example, by inhibiting casein kinase 2 (Raleigh et al., 2011), indeed providing a certain protection against experimental colitis (Raju et al., 2020). On the other hand, the group of JR. Turner has extensively characterized the MLCK-dependent signaling transduction regulating the leak pathway, culminating in the identification of the small molecule divertin (Graham et al., 2019). Divertin blocks the MLCK1 recruitment via the IgG3 domain to the perijunctional actomyosin ring inhibiting occludin endocytosis and promoting barrier function without altering MLCK enzymatic activity (He et al., 2008; Graham et al., 2019). Strikingly, divertin showed a similar therapeutic effect as anti-TNF in immune-mediated mouse experimental intestinal inflammation (Graham et al., 2019).

The accumulated evidence about the causative role of barrier function in IBD implies the need of assessing the impact of current treatments on the intestinal epithelium, and the potential link to success/lack of response in specific patients. One case in this context is the barrier repair observed upon anti-TNF treatment in CD patients (D'Haens et al., 1999; Rutgeerts et al., 2012; Kierkus et al., 2012), which has been mechanistically linked to the inhibition of IEC apoptosis and Notch pathway modulation (Kawamoto et al., 2019). Additionally, mesalamine treatment improved mucosal healing in clinical trials including mild-to-moderate UC patients (Lichtenstein et al., 2011; Bokemeyer et al., 2012; Probert et al., 2014). In this case, epithelial wound healing can also be promoted by increasing epithelial cell migration and proliferation (Baumgart et al., 2005) and impaired cytokine-driven paracellular permeability (Khare et al., 2019). The beneficial effects of corticosteroids was initially thought to be through the regulation of inflammatory factors (Wild et al., 2003). However, it was recently shown that the exposure of intestinal organoids derived from CD patients to prednisolone rescued the modulated expression/distribution of E-cadherin, ILDR-1, Claudin-2, MLCK and phospho-STAT1 upon cytokine treatment (Xu et al., 2021). Elegant work by Zuo et al. described the capacity of tacrolimus to interact with FKBP8, which in turn impairs their interaction with MLCK1 for its recruitment to the acto-myosin ring for the induction of epithelial barrier function (Zuo et al., 2023). Together, this shows the potential contribution of classical immunosuppressive drugs and biologicals to epithelial restoration in the context of IBD.

Many experts have demonstrated that there is a way to confer benefit to the host by administration of probiotics. Probiotics are live organisms that can shape the commensal microbiota and the composition of the mucus. Thus, bacteria such as Bacillus subtilis (Li et al., 2020b; Ahl et al., 2016), or Lactobacillus spp (Bron et al., 2017) or the Lactobacillus reuteri alter mucin production. In fact, Lactobacillus showed a protective effect, increasing the mucus layer thickness (Ahl et al., 2016). There are also many different studies reporting the beneficial effects of the supplementation with A. muciniphila (Wu et al., 2017; van der Lugt et al., 2019). Moreover, bacteria-derived metabolites altering the mucus composition, such as indoleacrylic acid, have shown protective effects in experimental colitis (Wlodarska et al., 2017). Thus, probiotics and their impact on the mucus layer emerge as interesting mechanisms to impact on intestinal epithelial integrity.

Despite attractive strategies, there is still no pharmacological treatment for epithelial restoration in IBD. This is partially because the limitations for primary intestinal epithelial cultures until the development of intestinal organoids, which impeded the segregation of epithelial intrinsic mechanisms. Organoids are multicellular culture systems embedding in an ECM-like matrix mimicking the 3D architecture of the intestinal epithelium, which are valuable surrogates for intestinal tissue. Importantly, the cellular complexity and plasticity of the intestinal epithelium can also be mimicked in intestinal organoids (Basak et al., 2017; Treveil et al., 2020; Martinez-Silgado et al., 2023), and they can be used for genetic manipulation, biobanking (van de Wetering et al., 2015), and translational studies, since they conserve genetic and epigenetics of the original tissue if derived from ASCs (Dotti et al., 2017). The use of organoids has made possible the validation of molecular signatures linked epithelial alterations in disease (Bigorgne et al., 2014), as well as the identification of new targets in epithelial cell biology with a potential direct application in IBD (Bayrer et al., 2018; Glal et al., 2018; Deuring et al., 2019; Li et al., 2020a). Functionally, permeability assays can be applied to intestinal organoid cultures (Bardenbacher et al., 2020; Rallabandi et al., 2020). For example, the restoration of permeability and remission upon low dose naltrexone organoids studies showed the restoration (Lie et al., 2018), or the cytokine-mediated induction of impaired barrier function in human-derived material (Gleeson et al., 2020). Moreover, the use of organoids has made it possible to study epithelial crosstalk with other players within the intestinal tissue, such as the microbiota (Leber et al., 2018; Roodsant et al., 2020) and immune cells. Thus, co-cultures of mononuclear phagocytes and organoids demonstrated that this intercellular communication is involved in epithelial cell differentiation and can be targetable in IBD (Ihara et al., 2018). Moreover, organoid culture has opened the path to the development of stem cell transplantation therapy to treat refractory ulcers in IBD patients (Yui et al., 2012). Technical development in the field has also allowed to overcome inherent limitations: the use of microinjection into the organoid lumen to study host-microbiota interactions (Saxena et al., 2016); or the inverted polarity of apical-out organoids mimicking the open intestinal lumen (Co et al., 2021). Further, the development of gut-on-a-chip models including non-epithelial cells within the gut tissue, such as the enteric nervous system, the endothelium and immune mediators will for sure have an enormous impact on biomedical research (Shin and Kim, 2022). Altogether, organoids are nowadays an indispensable tool for the development of new therapies in IBD in general, as nicely reviewed by Yoo and Donowitz (Yoo and Donowitz, 2019).