Claudins in Gastrointestinal Disorders

Claudins in Gastrointestinal Disorders

Samhar Samer Alouch¹, Amjad Ajam¹, Abdul Razzak Chalab Cham¹, Mohamed Naim Khalil¹, Ghazi Alabdul Razak¹, Khaled S M Elshaer¹, Abdulrazzaq Qattea¹, Abdulrahman Comert¹, Cham Jazieh¹, Abdulaziz Dahhan¹, Isra Mansoor Khan¹, Mohamad Yman Barghout¹, Tarek Ziad Arabi¹, Dana A. Almazroua¹, Atif Ahmed², Amani Alkofide³, Abderrahman Ouban¹, Rateb Abbarah¹

1. College of Medicine, Alfaisal University, Riyadh, Saudi Arabia
2. Seattle Children's Hospitals, University of Washington, Seattle, WA, USA
3. Department of Pediatric Hematology Oncology, King Faisal Specialist Hospital and Research Center, Riyadh, Saudi Arabia

*Correspondence to: Abderrahman Ouban, College of Medicine, Alfaisal University, Riyadh, Saudi Arabia.

Copyright.

© 2026 Abderrahman Ouban, This is an open access article distributed under the Creative Commons Attribution   License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Received: 04 March 2026

Published: 10 March 2026

DOI: https://doi.org/10.5281/zenodo.18932938

Abstract

The gastrointestinal (GI) tract relies on a dynamic epithelial barrier, with claudins serving as pivotal determinants of paracellular permeability and cellular signaling. Dysregulation of claudins has been increasingly implicated in the pathogenesis of diverse GI disorders, including malignancies, inflammatory diseases, infections, and barrier dysfunction syndromes. This narrative review synthesizes current evidence on claudin biology in health and disease, highlighting how isoform-specific expression and localization shape barrier integrity and influence signaling pathways such as Wnt/β-catenin, Src kinase, and ERK. In cancer, aberrant claudin expression exhibits dual roles: tumor-suppressive when maintaining adhesion, yet oncogenic when mislocalized or overexpressed, driving epithelial-to-mesenchymal transition, stemness, and therapeutic resistance. Inflammatory and immune-mediated conditions such as inflammatory bowel disease, celiac disease, and eosinophilic esophagitis demonstrate characteristic shifts toward increased pore-forming claudins (e.g., claudin-2) and decreased sealing claudins (e.g., claudin-7, -8), thereby exacerbating barrier leakiness and immune activation. Infectious and metabolic disorders further exploit claudins, as exemplified by hepatitis C virus utilizing claudin-1 as a coreceptor. Beyond pathogenesis, claudins hold promise as diagnostic biomarkers and therapeutic targets: monoclonal antibodies against claudin-1, claudin-binding toxins, and anti-claudin-18.2 therapies are under investigation, with potential applications in gastroesophageal, gastric, pancreatic, and hepatic cancers. Emerging imaging tools leveraging claudin affinity further underscore their translational potential. By integrating findings from human studies, animal models, and mechanistic experiments, this review delineates the multifaceted roles of claudins in GI disorders and evaluates opportunities and challenges in harnessing these proteins for clinical intervention.

Keywords: claudins, gastrointestinal, cancer, inflammatory bowel disease, celiac disease

Introduction

The gastrointestinal (GI) tract is lined by a continuous epithelial barrier that separates the host’s internal milieu from the external environment. At the core of this barrier are tight junctions (TJs) – multi-protein complexes that seal the paracellular space at the apical end of epithelial cells (1). Claudins are a family of tetraspan transmembrane proteins and key structural components of TJs (2). In mammals, at least 27 claudin proteins (claudin-1 through claudin-27) have been identified (2), each with distinct tissue distributions and functions. By forming homotypic and heterotypic strands between adjacent cells, claudins establish charge- and size-selective paracellular channels that control the permeability of ions and solutes across epithelia (3). Beyond their barrier role, claudins are increasingly recognized as dynamic regulators of cellular signaling, polarity, proliferation, and differentiation (1). Aberrant expression or localization of claudins can thus have profound effects on tissue homeostasis, potentially triggering pathological pathways such as inflammation and neoplasia (4).

Over the past two decades, dysregulation of claudins has been implicated in a wide spectrum of GI disorders, ranging from malignancies (esophageal, gastric, colorectal, hepatic, pancreatic cancers) to chronic inflammatory conditions (inflammatory bowel disease, celiac disease, eosinophilic esophagitis), infections and metabolic diseases (e.g. hepatitis C virus infection leading to liver fibrosis), and other barrier-related disorders (such as gastroesophageal reflux disease). Changes in claudin expression are often accompanied by disruption of the tight junction barrier, altered epithelial permeability, and downstream activation of immune responses (5). In parallel, claudins can act as signaling platforms; for example, certain claudins modulate Wnt/β-catenin, Src kinase, or other pathways that influence epithelial-to-mesenchymal transition (EMT), stem cell behavior, or apoptosis resistance (6, 7). These properties place claudins at the intersection of key processes: barrier integrity, immune surveillance, and cell fate determination that collectively drive disease pathogenesis in the GI tract.

This narrative review provides a comprehensive overview of the role of claudins in GI physiology and disease. We begin by summarizing claudin structure, isoform diversity, and normal functions in the gut, highlighting how their junctional and non-junctional roles are regulated. We then explore, by thematic categories, how claudin dysregulation contributes to GI disease pathogenesis, from cancers to inflammatory and infectious conditions, focusing on mechanisms like tight junction disruption, EMT promotion, stemness, and immune modulation. Relevant evidence from human clinical studies, animal models, and in vitro experiments is integrated to illustrate these concepts. Finally, we discuss the clinical implications of claudins: their potential as diagnostic/prognostic biomarkers and as therapeutic targets, including emerging strategies such as monoclonal antibodies against claudins, claudin-binding toxins, and imaging agents. Through this thematic approach, we aim to illuminate the multifaceted contributions of claudins to GI disease and the opportunities and challenges in targeting these proteins for clinical benefit.

Physiologic Function of Claudins

At tight junctions, claudins are the primary determinants of barrier properties. Different segments of the GI tract express specific combinations of claudins to balance permeability and sealing. For example, claudin-2 and claudin-15 are known pore-forming claudins that create cation-selective channels, contributing to the “leaky” epithelia of the proximal small intestine (8, 9). Claudin-2 is predominantly localized at TJ regions of intestinal epithelia (especially colon surface cells) and increases paracellular conductance to Na+ and water (5, 10). Claudin-15 similarly permits paracellular Na+flux; mice lacking claudin-15 develop Na+-malabsorption and osmotic imbalances leading to gut dilation (11-13). On the other hand, claudins-1, -3, -4, -5, and -7 are classical “sealing” claudins – they have neutral ECL charges and form tight, ion-indiscriminate seals (3). Claudins-3 and -4, for instance, are abundantly expressed on gastric and colonic epithelial TJs and help establish a high-resistance barrier that limits passive ion leak (14, 15). Claudin-8 is another sealing claudin, enriched at colonic epithelium TJs where it prevents back-leak of luminal sodium in the distal colon (16, 17). Dynamic regulation of claudin phosphorylation and localization allows epithelia to adjust permeability in response to physiological stimuli; e.g. activation of myosin light chain kinase (MLCK) by inflammatory cytokines can trigger endocytosis of claudin-3 and -4, acutely weakening the tight junction barrier (18, 19).

Notably, claudins are not confined to tight junctions: they can also localize along lateral membranes or even be released as extracellular vesicles, mediating non-junctional functions. For example, claudin-7 is found not only at TJs but also on basolateral membranes of colonocytes (20). There, it forms complexes with adhesion molecules like integrin α2 and EpCAM, impacting cell-matrix interactions and signaling cascades (21). Loss of claudin-7 disrupts these complexes and has been shown to activate matrix metalloproteinases (MMPs) and NF-κB signaling, leading to inflammation and compromised epithelial integrity (21, 22). Paradoxically, claudin-7 can also cooperate with EpCAM in cancer stem cells to form a co-transcriptional regulator (the “EpIC” complex) that enhances β-catenin signaling, illustrating a context-dependent signaling role for basolateral claudin-7 in promoting tumorigenesis (23). Claudin-1 offers another example of non-junctional activity: besides sealing TJs in the colon, claudin-1 is often mislocalized to the cytoplasm or nucleus in colon cancer cells, where it has been linked to Wnt/β-catenin pathway activation and increased EMT and invasive behavior (4). Indeed, claudin-1 overexpression in experimental colitis models led to heightened inflammation and impaired mucosal healing via upregulation of ERK and Notch signaling (24). Thus, claudins can influence cellular behavior both through their biophysical barrier function and by serving as platforms or modulators of signaling pathways.

Claudin Isoforms in the GI Tract

Over two dozen claudin genes are expressed in mammals, but their distribution in the GI tract varies widely (3). Claudin-1 is expressed broadly along the GI epithelium, from esophagus to colon, contributing to baseline barrier function. Claudin-3 and claudin-4 are also ubiquitous in GI epithelia and are typically co-expressed; they are highly abundant in the gastric mucosa and colon and moderately in the esophagus and small intestine (1). Claudin-5 is best known in endothelial barriers (e.g. blood–brain barrier), but in the gut it is present in the epithelial tight junctions of the colon as well  [49][50]. Claudin-2 is relatively sparse in the healthy colon (mainly in superficial epithelial cells) but is constitutively expressed in the small intestine (intestinal crypts) where it facilitates paracellular nutrient and water transport[24][26]. Claudin-7 and claudin-8 are enriched in the colon – claudin-7 localizes to both TJ and lateral surfaces of colonic cells[33], whereas claudin-8 is found in colonic and distal small intestinal TJs, helping tighten the distal gut barrier[31]. Claudin-12 is an enigmatic family member that lacks a PDZ-binding motif and has a more cell-type specific pattern: it is detected in the intestine (especially ileum) and has been implicated in vitamin D–dependent calcium absorption across enterocytes[51]. Claudin-15 is largely restricted to the small intestine, where it forms Na^+ channels important for paracellular nutrient uptake; claudin-15 knockout mice develop a “megintestine” phenotype with dilated small bowel due to malabsorption[26][27]. In the stomach, a unique claudin is predominant: claudin-18. The CLDN18 gene encodes two tissue-specific splice isoforms – claudin-18.1 in lung and claudin-18.2 in gastric epithelium[52]. Gastric claudin-18.2 is essential for establishing the stomach’s acid-resistant barrier, as it limits proton permeability across gastric mucosa[52][53]. Claudin-18.2 is normally absent from the intestine, so its expression in other locations (e.g. esophagus, pancreas) is often indicative of disease or neoplastic transformation, as discussed later[54][55].

Several claudins have minimal or no known function in the adult GI tract. Claudin-6 and claudin-9 are expressed primarily in embryonic tissues and certain organs (e.g. skin, testis), with little baseline GI expression – though they can be ectopically expressed in some GI cancers and, notably, serve as alternate HCV receptors in the liver[56][57]. Claudin-11 is specific to myelin sheaths and testicular Sertoli cells, and claudin-14, -16, and -19 are critical in the kidney (especially for renal ion reabsorption in the thick ascending limb) and inner ear, but are not significantly present in gut epithelia[58][59]. Claudin-13 is expressed in rodents but the CLDN13 gene is absent in humans[60]. Meanwhile, claudin-20, -21, -22, and -23 are lesser-studied isoforms with some expression in GI tissues: claudin-23, for instance, is enriched in differentiated colonocytes at the luminal surface where it appears to strengthen the mucosal barrier[61][62]. Indeed, claudin-23 knockout in mice leads to a more permeable colonic epithelium and greater susceptibility to inflammation, indicating a homeostatic barrier role that is disrupted during colitis (when claudin-23 expression is downregulated)[63][64]. The CLDN24, CLDN25, CLDN26, and CLDN27 genes are considered putative claudins – their protein products have been predicted but remain poorly characterized in terms of expression and function[59]. Initial reports suggest claudin-25 (also termed “claudin domain-containing protein 1”) is expressed in humans in some tissues (mutations in CLDN25 were linked to a rare neurologic disorder)[65][59], but its presence in the GI tract is not well established. In summary, a subset of claudins (mostly claudin-1 to -8, -12, -15, -18, -23) plays dominant roles in GI epithelial physiology, whereas others are either extrinsic to the GI system or involved in specialized contexts.

Regulation of Claudins

The expression and junctional localization of claudins are tightly regulated by a variety of mechanisms. Pro-inflammatory cytokines prevalent in GI disease can modulate claudin transcription and trafficking. Tumor necrosis factor-alpha (TNF-α) and interferon-gamma (IFN-γ), for example, synergistically activate MLCK and casein kinase in intestinal epithelial cells, leading to phosphorylation and endocytosis of claudin-4 and other TJ proteins, thereby compromising the barrier[32][29]. Interleukin-13 (a Th2 cytokine overexpressed in allergic disorders like eosinophilic esophagitis) has been shown to downregulate claudin-7 and claudin-1 in the esophageal epithelium, partly via induction of microRNA-155 that targets CLDN7 mRNA[66][67]. IL-9 and IL-23, cytokines elevated in active IBD, drive the loss of claudin-8 in the colon through upregulation of microRNAs (miR-21 and miR-223) that suppress CLDN8 expression[68][69]. Conversely, certain growth factors can increase claudin levels: transforming growth factor-alpha (TGF-α) and epidermal growth factor (EGF) signaling were reported to upregulate claudin-2 in intestinal epithelium via a MEK-ERK-dependent pathway, contributing to the proliferative, less-differentiated state of the crypt epithelium[70][71]. Microbial factors also play a role – the presence of a healthy microbiota and probiotic organisms can enhance barrier claudins. In an elegant murine study, colonization with probiotics in early life was shown to induce claudin-3 expression during gut maturation, an effect dependent on toll-like receptor (TLR) signaling (MyD88 pathway)[28][72]. This finding suggests that microbial-epithelial cross-talk helps fortify TJs (via claudin upregulation) as the gut adapts to the external environment. Diet and luminal metabolites likewise influence claudins: for instance, the short-chain fatty acid butyrate (produced by fiber-fermenting colonic bacteria) has been found to increase claudin-23 expression in intestinal cells through SP1 and AMPK signaling, thereby potentially strengthening the barrier[73]. On the other hand, physical insults like acid exposure in the esophagus can redistribute claudins – in reflux conditions, claudin-4 is often delocalized or reduced in superficial squamous cells, weakening the esophageal barrier and rendering it more susceptible to acid injury[74][75]. Overall, claudin biology in the GI tract is governed by a complex network of signals that reflect the physiological state of the tissue. When this regulation is perturbed – whether by genetic, environmental, or inflammatory stimuli – the resulting claudin abnormalities can set the stage for disease, as discussed in the following sections.

Claudins in Gastrointestinal Cancers

Dysregulated claudin expression is a hallmark of many GI malignancies, often emerging early in preneoplastic lesions and evolving through tumor progression (Figure 1). Claudins can act as double-edged swords in cancer: some function like tumor suppressors by maintaining adhesion and polarity (their loss facilitates invasion), whereas others may actively promote tumorigenic signaling when overexpressed. The net effect is context-dependent, varying by claudin subtype, tissue of origin, and stage of disease[76][77]. Here we review claudin alterations in major GI cancers (esophageal, gastric, colorectal, hepatic, and pancreatic), emphasizing how these changes affect tumor biology such as barrier integrity, EMT, stemness, and therapy resistance.

Esophagus (Esophageal Cancer and Barrett’s Dysplasia)

The esophagus provides a clear example of claudin changes during the metaplasia–dysplasia–carcinoma sequence. In the normal esophageal squamous epithelium, claudin expression is relatively low, except for barrier-formers like claudin-1 at the basal layer[78]. In Barrett’s esophagus (a metaplastic condition in which the normal squamous lining is replaced by intestinal-type columnar epithelium), claudin levels shift to a profile resembling intestinal/gastric mucosa. Several studies have reported significant upregulation of claudin-3 and claudin-4 in Barrett’s esophagus, compared to normal squamous mucosa[47][48]. Strong claudin-4 immunoreactivity is seen in most Barrett’s biopsies, including both non-dysplastic and dysplastic segments[79][80]. Claudin-3 and -4 remain elevated during progression to esophageal adenocarcinoma (EAC); in fact, EAC often shows equal or higher claudin-3/4 expression than the precursor Barrett’s lesion[47][79]. This persistent overexpression of claudin-3 and -4 may confer a selective advantage to the metaplastic and neoplastic cells – possibly by tightening the epithelial sheet to resist acid bile injury or by engaging signaling pathways that aid survival. Notably, claudin-4 is a receptor for CPE toxin, but in Barrett’s context its role might relate to regulating paracellular permeability to H^+; claudin-4 overexpression could help these columnar cells better withstand acidic reflux. In contrast, claudin-2 is normally absent in squamous esophagus but becomes induced in Barrett’s dysplasia/EAC. Abu-Farsakh et al. found claudin-2 to be highly expressed in EAC and its high-grade dysplasia precursors, more so than in Barrett’s intestinal metaplasia[81]. Thus, claudin-2 appears upregulated relatively late (during dysplasia and carcinoma) in the esophageal adenoma-carcinoma sequence[81]. Since claudin-2 forms cation-leaky channels, its ectopic expression might paradoxically reduce barrier function in tumors; however, in cancer cells claudin-2 can also activate pro-tumorigenic signals (e.g. EGFR/ERK) that enhance proliferation[82][71]. Indeed, claudin-2 is reported to promote EAC cell growth in vitro by activating EGFR, analogous to its role in colon cancer (discussed below).

Claudin-7 shows a unique pattern in esophageal neoplasia. In Barrett’s metaplasia and low-grade dysplasia, claudin-7 expression is uniformly present and often strong[83][84]. This suggests claudin-7 may be induced as part of the intestinal differentiation program in Barrett’s epithelium. However, as dysplasia progresses to intramucosal carcinoma and invasive EAC, claudin-7 levels tend to drop: only ~50% of EAC cases retain strong claudin-7 immunostaining[84]. The remainder show loss or no change relative to normal[84]. In practical terms, claudin-7 could be a marker distinguishing early Barrett’s lesions (where it’s high) from invasive EAC (where it’s often reduced to baseline)[84]. Functionally, loss of claudin-7 in advancing EAC might contribute to EMT and invasion, since claudin-7 normally helps stabilize cell-cell and cell-matrix contacts. Esophageal squamous cell carcinoma (ESCC), which arises from the squamous mucosa (often linked to smoking or alcohol exposure), exhibits its own claudin aberrations. One study noted claudin-1 is significantly increased in ESCC compared to adjacent normal epithelium[78]. Claudin-1 overexpression in ESCC correlates with more invasive behavior; mechanistically, it may facilitate tumor cell migration by modulating MMPs and interacting with components of the motility machinery[85]. Taken together, esophageal cancers exemplify how claudin dysregulation accompanies neoplastic transformation: a general trend is that tightening claudins (3,4,7) are gained in metaplasia, potentially as an adaptive response to chronic injury (acid reflux), but as carcinoma develops, some claudins (7, and possibly 4 in squamous cases) are lost to enable invasion, while pore-forming claudins (2) are aberrantly upregulated, aiding proliferative signaling and perhaps creating a more permeable barrier that fosters interaction with stromal factors.

Beyond their roles in pathogenesis, claudins in Barrett’s and EAC have clinical significance. Claudin-18 (particularly the -18.2 isoform, normally stomach-specific) has been detected in a subset of Barrett’s and EAC tissues[86][87]. In one study, CLDN18 was the most highly expressed TJ protein in Barrett’s metaplasia, and experimentally introducing claudin-18 into esophageal cell lines increased their resistance to acid and lowered paracellular permeability to H^+[87]. This indicates claudin-18 might protect Barrett’s epithelium against acid injury – but its presence in EAC could provide a therapeutic opportunity, as discussed later (anti-claudin-18.2 therapies). In ESCC, high claudin-1 expression has been associated with poor differentiation and metastasis[85], suggesting it could serve as a prognostic marker or a target for therapy (e.g., blocking claudin-1 to inhibit invasion). Overall, the esophagus showcases the dynamic changes of claudins during carcinogenesis, reflecting shifts in barrier requirements and signaling states from normal epithelium to metaplasia to cancer.

Stomach (Gastric Cancer and Precursors)

The gastric epithelium normally expresses a distinct claudin profile, dominated by claudin-18.2 (a lineage marker of gastric mucosa) and supplemented by claudin-4, -5, and -7 in the glandular tight junctions[88][49]. Preneoplastic changes in the stomach, such as chronic atrophic gastritis and intestinal metaplasia (often caused by Helicobacter pylori infection), are accompanied by significant claudin alterations. Intestinal metaplasia (IM) of the stomach entails the replacement of gastric glands with intestine-like epithelium; fittingly, studies report increased expression of claudin-1, -3, -4, -5, and -7 in IM, reflecting a shift toward an intestinal TJ phenotype[88][49]. In parallel, claudin-18 is markedly downregulated in areas of IM[88][49] – since claudin-18 is a gastric lineage protein, its loss is expected when gastric cells transdifferentiate into intestinal type. These changes have functional implications: loss of claudin-18 may weaken the gastric barrier, potentially allowing more exposure of underlying mucosa to irritants or carcinogens, whereas upregulation of claudin-3/4/5/7 might initially compensate by sealing the altered epithelium. Progressing along the neoplastic cascade, dysplasia and early gastric carcinoma show further claudin dysregulation. In gastric dysplasia (especially arising on a background of IM), claudin-3 and -5 remain elevated, and claudin-4 and -7 become even more strongly upregulated (often showing diffuse intense staining in high-grade dysplasia)[49][89]. Intriguingly, claudin-18 expression appears to rebound in gastric dysplasia/carcinoma – in fact, claudin-18 (particularly the 18.2 isoform) can be highly expressed in certain gastric cancers, notably the diffuse-type gastric adenocarcinoma[89]. One study noted that claudin-18.2 was especially elevated in diffuse gastric carcinoma lesions, whereas it was lower in IM and only modest in intestinal-type cancers[89]. This seemingly paradoxical re-expression of claudin-18 in diffuse cancers correlates with recent molecular insights: a significant subset of diffuse gastric cancers harbor CLDN18–ARHGAP fusions or mutations that dysregulate claudin-18[90]. In a mouse model, claudin-18.2 knockout led to chronic gastritis and spontaneous gastric tumors accompanied by activation of oncogenic pathways like YAP/Hippo and CD44[53][90]. Thus, claudin-18 may act as a tumor suppressor in the stomach (its loss can promote tumorigenesis through unleashing proliferation signals), yet paradoxically, the residual cancer cells in diffuse gastric carcinoma often overexpress claudin-18.2 on their surface – making it a convenient therapeutic target, as they still “wear” this gastric antigen[54][55].

Besides claudin-18, claudin-4 is frequently aberrant in gastric cancer. Some gastric tumors, particularly of the intestinal type, show increased claudin-4 expression which has been associated with more aggressive features[91][92]. For example, overexpression of claudin-4 in gastric cancer cell lines can enhance invasion and metastatic behavior via activating β-catenin and EMT programs[91]. On the other hand, other studies have found loss of claudin-4 in advanced gastric cancers correlating with poor differentiation and worse prognosis[91][92]. These conflicting reports suggest claudin-4’s role may depend on tumor context (molecular subtype, etc.). It is notable that claudin-4’s binding partner, claudin-3, is generally maintained or increased in gastric neoplasia, possibly to uphold some barrier function in hyperpermeable tumor tissue. Claudin-1 is another inducible claudin in gastric intestinal metaplasia/dysplasia – its expression is low in normal stomach, but rises in IM and remains elevated in many gastric cancers[88][49]. Given claudin-1’s known links to EMT, its upregulation may facilitate the progression of gastric lesions to invasive cancer. Consistently, claudin-1 upregulation (alongside claudin-3 and -5) has been observed at the IM → dysplasia transition[49]. Meanwhile, claudin-5, an endothelial-type claudin, is abnormally present in IM and dysplasia[88][49]; in gastric adenocarcinoma, claudin-5 levels vary, but one study showed it tends to decrease again in invasive carcinoma despite the earlier rise[88]. If claudin-5 is lost in cancer, it might contribute to increased metastatic potential by allowing tumor cells to traverse endothelial barriers more easily (claudin-5 loss in blood vessels is known to increase permeability[93][94], and tumor-derived signals can downregulate claudin-5 to promote angiogenic leakage and intravasation).

Clinically, claudin expression patterns are being explored as biomarkers in gastric cancer. Perhaps the most significant is claudin-18.2, which is present in up to ~30–40% of gastroesophageal adenocarcinomas. Claudin-18.2 has emerged as a target for antibody therapy (as discussed in Clinical Implications), and its detection by immunohistochemistry is used to screen patients for eligibility in trials[95]. The fact that claudin-18.2 is normally confined to gastric mucosa means that its expression in other contexts (like metastatic disease or other organs) strongly suggests a gastric origin; this can help in diagnostic dilemmas (e.g., distinguishing a gastric primary tumor vs. metastasis). Furthermore, the presence of claudin-18.2 in diffuse gastric cancer cells, despite their scattered, discohesive nature, raises interesting questions about its function: it might support small cell clusters or survival in ascitic environments (since diffuse cells lack E-cadherin, claudin-18 could be one of the remaining adhesion molecules). In summary, gastric cancers undergo a swing in claudin composition: gastric-to-intestinal transdifferentiation (IM) leads to loss of native claudin-18 and gain of intestinal claudins, but as malignancy ensues, some gastric cancers re-express claudin-18.2 and alter claudin-1/3/4/5/7 levels in ways that likely influence their invasiveness and interactions with the tumor microenvironment.

Colorectal Cancer (CRC)

The normal colonic epithelium is characterized by a tight junction network containing claudin-1, -3, -4, -7, -8, and others, which together maintain a robust barrier against the highly bacterial-rich colonic lumen. In colorectal cancer, this orderly architecture is disrupted. A striking feature of many colonic tumors is the “claudin paradox”: while some claudins (e.g. claudin-1, -2, -3, -4) are overexpressed and linked to oncogenic properties, others (claudin-7, -8, -23) are downregulated, reflecting loss of differentiation and barrier function. Early evidence of claudin involvement in colon tumorigenesis came from colitis-associated cancer models. In patients with long-standing ulcerative colitis, colon biopsies often exhibit increased claudin-1, -2, -3, -4 expression in areas of active inflammation[96]. This upregulation correlates with inflammatory activity and with activation of the Wnt/β-catenin pathway in these cells[96]. Experimentally, chronic colitis models demonstrate that claudin-1 overexpression promotes dysplasia: claudin-1 transgenic mice show exacerbated colonic inflammation, impaired mucosal healing, and a higher incidence of tumors[39][40]. Mechanistically, claudin-1 appears to drive EMT and stemness in colon cells by activating ERK, SRC and Notch signaling, and by coordinating with β-catenin to alter gene expression[11][97]. A study of human CRC found nuclear and cytosolic accumulation of claudin-1 in invasive fronts of tumors, often alongside nuclear β-catenin, suggesting claudin-1 might physically or functionally interact with the Wnt pathway[11][98]. Clinically, high claudin-1 in colorectal tumors has been associated with increased invasion and metastasis[99], yet intriguingly, among patients with resected stage II colon cancer, those whose tumors had lost claudin-1 had significantly worse prognosis (higher recurrence and mortality)[100][101]. This implies that claudin-1’s role is context-specific: while its overexpression can initiate malignant traits, once a cancer is established, a complete loss of claudin-1 may confer additional aggressiveness (possibly by enabling full EMT and detachment of cells).

Claudin-2 is another key player in colon cancer, especially in colitis-associated cancer (CAC). It is uniformly upregulated in IBD tissues and early dysplastic lesions[24][25]. In a mouse model, targeted colonic overexpression of claudin-2 did not increase tumor incidence per se; in fact it made mice more resistant to DSS colitis, by promoting mucosal healing (through enhanced epithelial proliferation and regulatory immune responses)[102][103]. However, in the setting of existing neoplasia, claudin-2 fuels tumor growth: colonic organoid studies showed claudin-2 drives cancer cell proliferation via EGFR/MEK/ERK signaling[82][71]. Moreover, claudin-2 expression is linked to a cancer stem-like phenotype. One group reported that claudin-2 promotes the self-renewal of colon cancer stem cells, supporting tumor initiation and chemoresistance[104]. Thus, claudin-2 may protect colonic epithelium in inflammatory conditions (by keeping the barrier leaky enough to not overburden the epithelium with cytokine stress, and by activating pro-healing pathways), but in cancer, the same molecule can increase the tumor’s robustness and survival.

A consistent finding across many CRC studies is the loss of claudin-7 as tumors progress. Claudin-7 is highly expressed in normal colon (particularly on the lateral membrane and TJs of absorptive enterocytes) and in early adenomas, but is frequently reduced or absent in colorectal carcinomas[105][106]. Hypermethylation of the CLDN7 gene promoter has been documented in colon adenomas and cancers, explaining its transcriptional silencing[107][108]. Loss of claudin-7 is functionally important: claudin-7 knockout mice spontaneously develop colonic inflammation and tumors, due to barrier defects that allow bacterial products to provoke chronic inflammation, and due to destabilization of the claudin-7/claudin-1/integrin complex which impairs epithelial adhesion[109][110]. In cell culture, claudin-7 re-introduction can suppress colon cancer cell motility and EMT by inhibiting SRC and ERK signaling[34][111]. However, akin to its dual nature in esophageal cancer, claudin-7 in colon cancer might have a pro-tumor angle too: as noted, claudin-7 can partner with EpCAM in EpCAM-high cancer stem cells to form a transcriptional co-activator driving EMT and metastasis[112]. It is possible that claudin-7’s location and binding partners dictate the outcome – when at the membrane with integrins, it is tumor-suppressive, but when engaged in a complex with EpCAM and β-catenin in the nucleus of a stem-like cell, it might promote tumor progression. This nuance is an active area of investigation.

Other claudins in CRC show varied changes. Claudin-3 and claudin-4 are frequently elevated in primary colorectal adenocarcinomas compared to normal mucosa[113]. They contribute to the well-known “claudin-high” phenotype seen in certain molecular subtypes of CRC (often MSI-low, β-catenin active tumors). High claudin-3/4 might confer some cohesive advantage to tumor cells (holding cells together in glands) and also interact with growth signals; for instance, claudin-3 deficiency in mice led to activation of IL6/Stat3 and Wnt signaling and more rapid tumor formation[114][76], whereas overexpression of claudin-3 in a colon cancer cell line enhanced malignancy via EGFR and PI3K/Akt activation[76][115]. Claudin-5 is not normally prominent in colon epithelial cells, but it can be induced in CRC by factors from tumor-infiltrating lymphocytes (through Notch signaling)[116][117]. There is evidence that claudin-5 upregulation in tumor epithelia might initially serve a homeostatic function to maintain some differentiation; however, if claudin-5 is subsequently downregulated (either in tumor cells or local endothelium via VEGF), it could facilitate metastasis by increasing vascular permeability[118][119]. Claudin-8 is downregulated in IBD and remains low in many CRCs[120][121]. Since claudin-8 normally helps seal the colonic TJ, its loss could increase leakiness and allow growth factors or antigens to permeate the tumor microenvironment. Low claudin-8 has been linked with a more mesenchymal, invasive tumor phenotype (though data in CRC specifically are sparse)[121]. Claudin-12 is reportedly upregulated in sporadic colorectal adenocarcinomas[122], but its role is not understood. Given claudin-12’s involvement in calcium flux, one wonders if altering calcium homeostasis in tumor cells (via claudin-12 upregulation) affects proliferation or differentiation. Finally, claudin-23 is often downregulated in colon cancer, especially those arising from inflammatory backgrounds[123][73]. This is consistent with the observation that CLDN23 is a gene target of gut barrier-promoting pathways and is lost when inflammation triggers DNA methylation or transcriptional repression. Low claudin-23 in colon tumors might contribute to the disorganized junctions observed in poorly differentiated or inflamed cancers, potentially allowing more immune cell infiltration (which can either attack the tumor or, paradoxically, provide growth factors for it).

From a clinical perspective, claudin expression patterns might help in stratifying CRC or guiding therapy. For example, claudin-1 has been explored as a therapeutic target: a recent study demonstrated that an anti-claudin-1 antibody could suppress growth of claudin-1^high colon tumors in mice, especially those with active Wnt signaling[124][100]. Conversely, as mentioned, patients with claudin-1^low tumors have poorer outcomes, so assessing CLDN1 by immunohistochemistry could have prognostic value[101]. Claudin-2 might serve as a marker for colitis-driven neoplasia; it tends to be elevated in early colitis-associated dysplasia[125] and could identify patients at risk. Claudin-18.2, while not normally in colon, can appear in some colon cancers (particularly those with gastric differentiation or pancreatic metastases); its presence would open up the possibility of anti-CLDN18.2 therapy, though primary CRC expression of CLDN18.2 is uncommon. On the diagnostic front, pathologists sometimes use claudin immunostains to determine tumor origin: for instance, claudin-4 is positive in virtually all adenocarcinomas (including CRC) but negative in mesotheliomas, so a metastatic colon cancer in peritoneum can be confirmed with claudin-4 staining[126][127]. In summary, colorectal cancers illustrate both the loss-of-function (barrier weakening) and gain-of-function (oncogenic signaling) aspects of claudin dysregulation, and ongoing research is aiming to translate these insights into biomarkers and targeted treatments.

Liver (Hepatocellular Carcinoma and Cirrhosis)

Hepatocytes in the liver are connected by tight junctions as part of the bile canalicular apparatus, though the composition of liver TJs differs from intestinal epithelia. Key claudins in hepatocyte TJs include claudin-1 and claudin-2, with smaller contributions from claudin-3, -5, and -7 in bile duct epithelium and vasculature. In chronic liver diseases that lead to cirrhosis (fibrosis of the liver), claudin expression is altered both in hepatocytes and cholangiocytes. Immunohistochemical studies on cirrhotic liver tissues have shown that claudin-1 and claudin-7 are upregulated in cirrhotic hepatocytes compared to normal liver[128]. Moreover, if hepatocellular carcinoma (HCC) arises within a cirrhotic liver, those tumor cells exhibit even higher claudin-1/7 levels[128]. This suggests that claudin-1 and -7 may be induced during the regenerative and fibrogenic processes of cirrhosis, possibly as hepatocytes attempt to re-establish cell adhesion in a changing matrix. Claudin-4 is also elevated in cirrhosis, especially in areas of active fibrosis – a study found that claudin-4 (and claudin-7) levels correlated with the grade of fibrosis in chronic hepatitis/cirrhosis, independent of inflammation[129]. One hypothesis is that hepatic stellate cells (the fibrogenic cells of the liver) or inflammatory cytokines (like TGF-β) induce claudin-4/7 in hepatocytes during fibrosis, and these claudins might influence the epithelial-to-mesenchymal transitions that hepatocytes undergo in cirrhosis. Interestingly, claudin-5, which normally is an endothelial TJ protein, can be aberrantly expressed in cholangiocytes in primary biliary cirrhosis, potentially affecting the permeability of biliary epithelium. However, more data exist for claudin-1: it has a recognized role in liver fibrosis beyond the tight junction. A recent study demonstrated that non-junctional claudin-1 in hepatic stellate cells can activate TGF-β and Wnt pathways, contributing to fibrotic progression[130][131]. Blocking claudin-1 with a monoclonal antibody in fibrotic mouse models not only reduced fibrosis but unexpectedly also suppressed HCC development, highlighting claudin-1’s fibrogenic and carcinogenic signaling roles[132][133].

In hepatocellular carcinoma, claudin expression patterns are heterogeneous and can correlate with tumor behavior. Some HCCs (particularly well-differentiated tumors) maintain high levels of claudin-1 at their cell–cell borders, essentially retaining some hepatocyte phenotype. In contrast, more poorly differentiated HCCs often show loss of claudin-1 expression[134]. This loss is associated with vascular invasion and metastasis; essentially, when HCC cells downregulate claudin-1 and other junction proteins, they can more easily dissociate and spread. Clinical studies have found that reduced claudin-1 in HCC tissue is linked to worse overall survival[134]. One analysis reported that HCC patients whose tumors had low claudin-1 had significantly shorter survival, whereas those with high claudin-1 had a more favorable prognosis[134]. This might appear contradictory to the idea of claudin-1 as an oncogene in colon cancer, but it underscores tissue context: in liver, claudin-1 is part of the epithelial architecture keeping cells in a differentiated, less invasive state, so its loss denotes a switch to a malignant phenotype. Conversely, some studies suggest very high claudin-1 in HCC can also be problematic – for example, claudin-1 overexpression was found to enhance invasive activity in an in vitro model by promoting MMP2-mediated matrix degradation[85] (that particular study was in oral SCC cells, but similar principles may apply in HCC). It’s possible that an optimal range of claudin-1 is needed: too much might activate certain pathways, but too little definitely removes restraints on invasion. Other claudins in HCC are less studied; claudin-9 and -6 are expressed in some hepatocytes (and recall, claudin-6/9 are HCV receptors). HCC arising in an HCV-infected liver might therefore have had unique claudin perturbations from the viral entry process. Occludin and claudin-12 have been reported to decrease in HCC, consistent with a general trend of TJ downregulation during hepatocyte transformation.

Another type of liver tumor, cholangiocarcinoma (bile duct cancer), also involves claudin changes. Cholangiocytes normally express claudin-3, -4, -7, among others; cholangiocarcinomas often show strong claudin-4 (similar to many adenocarcinomas) and sometimes claudin-18 if they arise near the hepato-pancreatic region (since pancreatic ducts and bile ducts share some claudin profile). However, detailed data on claudins in cholangiocarcinoma are limited. Clinically, claudin expression might help differentiate HCC from metastatic adenocarcinoma to the liver. For instance, claudin-4 is usually negative in HCC (since hepatocytes don’t express claudin-4), whereas it’s positive in most adenocarcinomas (like colorectal or pancreatic metastases). Pathologists can use a claudin-4 stain to support a diagnosis: a liver tumor negative for claudin-4 and positive for HepPar1 is likely an HCC, while a claudin-4 positive tumor in liver suggests a metastasis[135].

Finally, we must note the interplay with Hepatitis C virus (HCV) – an infectious cause of liver disease that can lead to cirrhosis and HCC. HCV uses claudin-1 (and claudin-6/9) as essential co-receptors to enter hepatocytes[42][43]. This means claudin-1 is not only a structural protein but also part of the viral life cycle in HCV infection. Chronic HCV has been reported to increase claudin-1 expression on hepatocytes, potentially making the cells more permissive to infection spread[57]. Blocking claudin-1 with antibodies can inhibit HCV entry in vitro[136], and such strategies are being considered as antiviral adjuncts. Moreover, HCV core and NS proteins can alter TJ assembly – some studies found HCV infection leads to mislocalization of occludin and claudin-1, which might contribute to the barrier dysfunction seen in HCV-infected livers (this could allow immune cells greater access or foster a pro-fibrotic microenvironment). The HCV-claudin connection has also fueled interest in anti-claudin-1 therapeutics; as mentioned, an antibody developed for its antifibrotic and antitumor effect in liver (by Alentis Therapeutics) was partly motivated by earlier work showing that anti-claudin-1 could block HCV entry[130]. In summary, liver pathologies highlight claudin-1 as a central figure – involved in viral infection, fibrogenesis, and carcinogenesis – with claudin-4, -7 contributing to fibrotic change, and claudin loss being a marker of aggressive HCC.

Pancreas (Pancreatic Cancer and Cystic Neoplasms)

The pancreas, specifically the pancreatic ductal epithelium, shares some claudin features with the biliary tract and intestine. Normal pancreatic ducts express claudin-3, -4, -7, -18 (the latter because pancreatic ducts and acini have some gastric differentiation traits, explaining claudin-18.2 expression). In the context of pancreatic ductal adenocarcinoma (PDAC), claudin dysregulation is pronounced and may contribute to the notorious treatment resistance and invasive nature of this cancer. One of the most consistently overexpressed claudins in PDAC is claudin-4. Nearly all PDAC tumors show strong claudin-4 membrane positivity[137][138] – so much so that claudin-4 has been considered a potential universal marker and therapeutic target for pancreatic cancer. Functional studies indicated that claudin-4 might promote PDAC cell survival and chemo-resistance by activating downstream signals (e.g. via PI3K/Akt). However, claudin-4’s primary impact might be on the barrier between tumor cells: PDAC cells form glandular structures that are often sealed off, creating high intratumoral pressure and poor drug penetration. High claudin-4 could worsen this by tightening cell junctions, contributing to the fibrotic tumor’s ability to exclude chemotherapeutics. Indeed, preclinical work has exploited claudin-4’s presence by using CPE peptide to selectively lyse pancreatic cancer cells or as a carrier to deliver nanoparticles[139][138]. Encouragingly, toxin molecules based on Clostridium perfringens enterotoxin (which binds claudin-4) have shown potent killing of claudin-4^+ PDAC cells in mice[137][44], underscoring how overexpression of a normally innocuous TJ protein can become a cancer vulnerability.

Claudin-18.2 is another notable claudin in pancreatic cancer. While normal pancreas has low claudin-18, it is found in a significant subset of PDACs – particularly those with gastric-type differentiation or certain mole