{"gene":"CLDN5","run_date":"2026-06-09T22:57:18","timeline":{"discoveries":[{"year":1999,"finding":"CLDN5/TMVCF localizes exclusively to tight junction strands in endothelial cells of blood vessels (not epithelial cells), and when cDNA is introduced into mouse L fibroblasts, it reconstitutes extracellular face-associated TJ strands, establishing CLDN5 as an endothelial cell-specific structural component of tight junction strands.","method":"Immunofluorescence microscopy, immunoreplica electron microscopy, cDNA transfection into L fibroblasts (reconstitution assay)","journal":"The Journal of cell biology","confidence":"High","confidence_rationale":"Tier 1 / Strong — direct reconstitution in heterologous cells combined with immunoelectron microscopy localization; foundational study replicated broadly across subsequent literature","pmids":["10508865"],"is_preprint":false},{"year":2012,"finding":"The ETS transcription factor ERG directly regulates CLDN5 gene expression in endothelial cells; ERG knockdown reduces CLDN5 expression, increases endothelial permeability, and promotes stress fiber and gap formation, placing ERG upstream of CLDN5 in the transcriptional control of barrier function.","method":"ERG siRNA knockdown in endothelial cells, permeability assays, stress fiber/gap quantification, transcriptional reporter assays","journal":"The Journal of biological chemistry","confidence":"Medium","confidence_rationale":"Tier 2 / Moderate — genetic loss-of-function (knockdown) with defined cellular permeability phenotype and transcriptional downstream target identification, single lab","pmids":["22235125"],"is_preprint":false},{"year":2011,"finding":"TNF-α reduces CLDN5 promoter activity and mRNA expression in brain endothelial cells via NF-κB signaling; overexpression of the NF-κB subunit p65 (RelA) alone is sufficient to repress the Cldn5 promoter, and a conserved promoter region is required for this downregulation.","method":"Promoter-reporter assays, p65 overexpression, TNF-α treatment of primary brain endothelial cells, qPCR, western blot","journal":"Cytokine","confidence":"Medium","confidence_rationale":"Tier 2 / Moderate — promoter activity assays and gain-of-function p65 overexpression with two orthogonal readouts (promoter activity and mRNA), single lab","pmids":["22138107"],"is_preprint":false},{"year":2014,"finding":"IL-1β–induced CLDN5 transcriptional repression in brain microvascular endothelial cells requires non-muscle MLCK (nmMlck): nmMlck mediates nuclear translocation of β-catenin and FoxO1, which then repress Cldn5 transcription, and primary BMVECs from nmMlck-null mice are protected from this repression.","method":"Primary BMVEC isolation from nmMlck-null mice (genetic loss-of-function), IL-1β treatment, nuclear translocation assays for β-catenin and FoxO1, Cldn5 mRNA/protein quantification, barrier permeability assays","journal":"Journal of cell science","confidence":"High","confidence_rationale":"Tier 2 / Strong — genetic knockout cells with specific phenotypic rescue, multiple orthogonal methods (nuclear fractionation, transcription, permeability), single lab but rigorous design","pmids":["24522189"],"is_preprint":false},{"year":2020,"finding":"Under hypoxia, CAV1 (caveolin-1) mediates redistribution of membranous CLDN5 into the cytosol in brain microvascular endothelial cells; autophagy then degrades CAV1 and cytosolic aggregated CLDN5. Blocking autophagy genetically or pharmacologically aggravates cytosolic CLDN5 accumulation and worsens blood-brain barrier breakdown.","method":"In vivo zebrafish BBB models, in vitro BMEC culture, CAV1 genetic manipulation, autophagy blockade (chemical and genetic), STED super-resolution microscopy, TEER measurements","journal":"Autophagy","confidence":"High","confidence_rationale":"Tier 2 / Strong — multiple orthogonal methods (in vivo zebrafish, in vitro cells, genetic and chemical autophagy blockade, super-resolution imaging, TEER), two model systems","pmids":["33280500"],"is_preprint":false},{"year":2022,"finding":"A de novo missense mutation G60R in the first extracellular loop of CLDN5 converts its function from a purely barrier-forming tight junction protein to an anion-selective channel: stably transfected cell lines with G60R CLDN5 still form tight junctions but display attenuated small-molecule barrier with higher Cl− and lower Na+ permeability, representing the first gain-of-function mutation in the claudin gene family.","method":"Stable transfection of wild-type and G60R CLDN5 cell lines, ion permeability electrophysiology, protein structural modeling, sequence alignment, barrier function assays","journal":"Brain : a journal of neurology","confidence":"High","confidence_rationale":"Tier 1 / Moderate — reconstitution of mutant protein in stably transfected cells with direct electrophysiological measurement of ion permeability and barrier function; single lab but multiple orthogonal methods","pmids":["35714222"],"is_preprint":false},{"year":2022,"finding":"In podocytes, CLDN5 deletion reduces ZO1 expression and induces nuclear translocation of ZONAB, which transcriptionally downregulates WIF1 (WNT inhibitory factor-1), leading to activation of the WNT signaling pathway; podocyte-derived WIF1 also acts in paracrine on tubular epithelial cells to suppress fibrosis.","method":"Podocyte-specific Cldn5 knockout mice, diabetic nephropathy and ureteral obstruction mouse models, ZONAB nuclear localization assays, WIF1 transcriptional analysis, systemic WIF1 delivery rescue experiments","journal":"Nature communications","confidence":"High","confidence_rationale":"Tier 2 / Strong — conditional knockout mouse model, epistasis via WIF1 rescue, multiple in vivo models, paracrine function established by co-culture and in vivo rescue","pmids":["35332151"],"is_preprint":false},{"year":2023,"finding":"Blue light exposure causes rapid ADAM17-mediated degradation of endothelial CLDN5, disrupting tight junctions and the inner blood-retinal barrier. Under basal conditions ADAM17 is sequestered by the inhibitory G protein GNAZ; blue light releases ADAM17 from GNAZ. GNAZ knockdown phenocopies blue light effects (ADAM17 hyperactivation, CLDN5 downregulation, barrier permeability). Pharmacological or genetic inhibition of ADAM17 prevents CLDN5 degradation.","method":"In vitro endothelial cell blue-light exposure, GNAZ knockdown, ADAM17 pharmacological and genetic inhibition, western blot for CLDN5 protein levels, in vivo mouse iBRB leakage, electroretinography","journal":"Fluids and barriers of the CNS","confidence":"High","confidence_rationale":"Tier 2 / Strong — genetic knockdown, pharmacological inhibition, and in vivo model with multiple functional readouts (barrier integrity, electroretinogram); two model systems","pmids":["37095509"],"is_preprint":false},{"year":2024,"finding":"DLL4-NOTCH signaling regulates endothelial CLDN5 expression and blood-brain barrier integrity through the NOTCH-NICD-RBPJ-CLDN5 pathway; DLL4 insufficiency (Dll4+/LacZ mice) leads to downregulation of CLDN5, persistent abnormalities in brain microvasculature, and increased vascular permeability both in vivo and in vitro.","method":"Dll4+/LacZ and Dll4+/+ mouse models, human brain microvascular endothelial cell in vitro studies, in vivo vascular permeability assays, NOTCH pathway component analysis","journal":"The Journal of physiology","confidence":"Medium","confidence_rationale":"Tier 2 / Moderate — genetic heterozygous mouse model plus in vitro corroboration, pathway placed via NOTCH-NICD-RBPJ intermediates, single lab","pmids":["38632887"],"is_preprint":false},{"year":2020,"finding":"Serotonin/5-HT1A signaling enhances endothelial CLDN5 expression in brain microvascular endothelial cells; the 5-HT1A receptor is expressed in BMVECs and mural cells, and PKA is aberrantly activated in both cell types in schizophrenic prefrontal cortex, coinciding with site-selective CLDN5 breakdown.","method":"Immunohistochemistry on post-mortem human brain tissue, two-dimensional co-culture of BMVECs and pericytes, pharmacological 5-HT1A receptor stimulation, CLDN5 expression quantification","journal":"International journal of molecular sciences","confidence":"Medium","confidence_rationale":"Tier 3 / Moderate — pharmacological modulation in co-culture with human tissue immunohistochemistry confirmation; single lab, mechanistic depth limited by abstract","pmids":["33383868"],"is_preprint":false},{"year":2017,"finding":"The human CLDN5 gene contains a high-frequency SNP (rs885985) creating two ORFs (303 or 218 amino acid isoforms), but only the 218 aa form is detected by immunoblot in human lung tissue. When forcibly expressed in transfected cells, the long 303 aa form is retained in intracellular compartments and does not localize to the plasma membrane, in contrast to the 218 aa form which localizes to intercellular junctions.","method":"Genotyping of human lung tissue, immunoblot, transfection of long- and short-form CLDN5 constructs in cells with subcellular localization imaging","journal":"Annals of the New York Academy of Sciences","confidence":"Medium","confidence_rationale":"Tier 2 / Moderate — direct protein expression experiment in transfected cells plus human tissue immunoblot; single lab, two orthogonal approaches","pmids":["28445614"],"is_preprint":false},{"year":2025,"finding":"CLDN5 in adipocytes (non-tight junction context) affects subcellular localization of Y-box protein 3 (YBX3); CLDN5 deficiency causes YBX3 mislocalization, which reduces IL10 expression (YBX3 directly binds the IL10 promoter and 3'-UTR); secreted IL10 then acts in paracrine via IL10R on neighboring thermogenic adipocytes to regulate thermogenesis and energy expenditure.","method":"Adipocyte-specific Cldn5 knockout mice, YBX3 localization assays, IL10 promoter binding assays, IL10R paracrine signaling experiments, metabolic phenotyping (thermogenesis, energy expenditure, glucose tolerance)","journal":"Nature communications","confidence":"High","confidence_rationale":"Tier 2 / Strong — conditional knockout mouse, promoter binding assay, paracrine signaling rescue, and metabolic phenotyping providing multiple orthogonal lines of evidence in a single rigorous study","pmids":["40610440"],"is_preprint":false},{"year":2026,"finding":"CLDN5 in podocytes forms a stable complex with β1-integrin via its intracellular loop and C-terminal domains; CLDN5 prevents HUWE1-mediated ubiquitination at lysine K774 of β1-integrin, thereby protecting β1-integrin from proteasomal degradation and ensuring its proper membrane localization. CLDN5 deletion impairs podocyte adhesion, spreading, and mechanical stress resistance in vitro, and worsens renal injury in vivo.","method":"Super-resolution imaging (colocalization at podocyte-GBM interface), Co-IP (stable complex), domain-mapping mutagenesis (intracellular loop and C-terminal domains), ubiquitination assays (HUWE1, K774), proteasome inhibition, Cldn5-KO mice with hypertensive and adriamycin injury models","journal":"The Journal of biological chemistry","confidence":"High","confidence_rationale":"Tier 1 / Strong — Co-IP with domain mutagenesis, ubiquitination site identification, proteasomal degradation rescue, super-resolution localization, and in vivo KO model; multiple orthogonal methods in single rigorous study","pmids":["41539562"],"is_preprint":false},{"year":2025,"finding":"Molecular dynamics free energy calculations show that multi-pore Claudin-5 TJ strand models create electrostatic barriers to ion permeation higher than in single-pore architectures, and only the multi-Pore I structural model recapitulates the anion-selective permeability phenotype of the G60R variant, providing structural validation of the paracellular pore mechanism.","method":"Multi-microsecond all-atom molecular dynamics simulations, free energy calculations for water and ion permeation across multi-protomer (16-subunit) claudin-5 TJ models","journal":"bioRxiv","confidence":"Low","confidence_rationale":"Tier 4 / Weak — computational/simulation only, no experimental validation reported; preprint","pmids":[],"is_preprint":true}],"current_model":"CLDN5 is an endothelial cell-enriched four-transmembrane tight junction protein that forms paracellular barrier strands (reconstituted in fibroblasts) and is transcriptionally regulated by NF-κB/p65, ERG, and NOTCH-RBPJ pathways; at the BBB its membrane levels are controlled by CAV1-mediated redistribution and autophagy-mediated degradation, and its tight-junction barrier function can be converted to anion-selective channel activity by the gain-of-function G60R mutation in its first extracellular loop; beyond endothelial junctions, CLDN5 in podocytes stabilizes β1-integrin by blocking HUWE1-mediated ubiquitination and activates a ZO1-ZONAB-WIF1 axis to suppress WNT signaling, while in adipocytes it controls thermogenesis by regulating YBX3 localization and paracrine IL10 signaling."},"narrative":{"mechanistic_narrative":"CLDN5 is an endothelial cell-enriched, four-transmembrane claudin that serves as a structural component of tight junction strands and establishes paracellular barriers, most prominently the blood-brain and inner blood-retinal barriers [PMID:10508865]. Its barrier-forming activity was defined by reconstitution of extracellular face-associated tight junction strands when CLDN5 cDNA is introduced into fibroblasts lacking endogenous junctions [PMID:10508865]. CLDN5 abundance at endothelial junctions is set by convergent transcriptional inputs — it is activated by the ETS factor ERG [PMID:22235125] and by DLL4-NOTCH signaling through an NICD-RBPJ axis [PMID:38632887], and repressed by inflammatory NF-κB/p65 signaling downstream of TNF-α [PMID:22138107] and by an IL-1β–nmMLCK–β-catenin/FoxO1 pathway [PMID:24522189]. Post-transcriptionally, junctional CLDN5 is removed by CAV1-mediated redistribution into the cytosol followed by autophagic degradation under hypoxia [PMID:33280500] and by ADAM17-mediated proteolysis that is normally restrained by the inhibitory G protein GNAZ [PMID:37095509]. A de novo G60R mutation in the first extracellular loop converts CLDN5 from a pure barrier protein into an anion-selective paracellular channel, the first gain-of-function mutation described in the claudin family [PMID:35714222]. Beyond endothelial junctions, CLDN5 has distinct non-junctional roles: in podocytes it stabilizes β1-integrin by blocking HUWE1-mediated ubiquitination at K774 [PMID:41539562] and governs a ZO1-ZONAB-WIF1 axis that restrains WNT signaling [PMID:35332151], and in adipocytes it controls YBX3 localization and IL10-dependent paracrine thermogenic signaling [PMID:40610440].","teleology":[{"year":1999,"claim":"Established CLDN5 as an endothelial-specific structural building block of tight junction strands rather than a passive marker, by showing it is sufficient to nucleate junction strands in cells that have none.","evidence":"Immunofluorescence and immunoreplica EM localization plus cDNA reconstitution in L fibroblasts","pmids":["10508865"],"confidence":"High","gaps":["Did not define the heterotypic claudin partners or ZO adaptors required in vivo","No mechanism for how strand architecture sets ion or size selectivity"]},{"year":2011,"claim":"Identified inflammatory transcriptional repression as a route to barrier loss, showing NF-κB/p65 is sufficient to silence the Cldn5 promoter downstream of TNF-α.","evidence":"Promoter-reporter assays, p65 overexpression, TNF-α treatment of brain endothelial cells with qPCR/western readouts","pmids":["22138107"],"confidence":"Medium","gaps":["Exact p65 binding element within the conserved promoter region not mapped","Single lab, did not test in vivo barrier consequences"]},{"year":2012,"claim":"Placed the ETS factor ERG upstream as a positive transcriptional driver of CLDN5 required for endothelial barrier maintenance.","evidence":"ERG siRNA knockdown with permeability assays, stress-fiber/gap quantification, and reporter assays","pmids":["22235125"],"confidence":"Medium","gaps":["Direct ERG occupancy of the CLDN5 promoter not resolved beyond reporter assays","Relationship to NF-κB repression not integrated"]},{"year":2014,"claim":"Resolved the signaling chain linking IL-1β to CLDN5 repression, identifying nmMLCK as the obligatory transducer driving nuclear β-catenin/FoxO1 to silence Cldn5.","evidence":"Primary BMVECs from nmMlck-null mice, IL-1β treatment, nuclear translocation assays, transcription and permeability readouts","pmids":["24522189"],"confidence":"High","gaps":["Whether β-catenin/FoxO1 bind the Cldn5 promoter directly not shown","How a cytoskeletal kinase controls transcription factor nuclear entry not mechanistically detailed"]},{"year":2020,"claim":"Defined a post-transcriptional turnover route for junctional CLDN5, showing CAV1 redistributes membranous CLDN5 to the cytosol and autophagy clears the aggregates, with autophagy blockade worsening barrier breakdown.","evidence":"Zebrafish BBB and in vitro BMEC models, CAV1 manipulation, genetic/chemical autophagy blockade, STED imaging, TEER","pmids":["33280500"],"confidence":"High","gaps":["Trigger coupling hypoxia to CAV1 engagement of CLDN5 not defined","Whether degraded CLDN5 is recycled or replaced not addressed"]},{"year":2020,"claim":"Linked GPCR signaling to CLDN5 levels, showing serotonin/5-HT1A signaling enhances endothelial CLDN5 and implicating aberrant PKA activation in disease-associated barrier breakdown.","evidence":"Post-mortem human brain immunohistochemistry, BMVEC-pericyte co-culture, pharmacological 5-HT1A stimulation","pmids":["33383868"],"confidence":"Medium","gaps":["Causal chain from 5-HT1A/PKA to CLDN5 transcription or stability not dissected","Correlative human tissue data, not interventional"]},{"year":2022,"claim":"Demonstrated CLDN5 can be a conduction element, not just a seal, by showing the G60R extracellular-loop mutation creates an anion-selective paracellular channel — the first claudin gain-of-function.","evidence":"Stable WT and G60R CLDN5 cell lines, ion permeability electrophysiology, structural modeling, barrier assays","pmids":["35714222"],"confidence":"High","gaps":["Atomic structure of the conducting pore not solved","In vivo neurological consequences of altered ion flux not established"]},{"year":2022,"claim":"Uncovered a non-junctional signaling role in podocytes, where CLDN5 sustains ZO1 and restrains ZONAB-driven WIF1 repression to keep WNT signaling off, with paracrine WIF1 limiting tubular fibrosis.","evidence":"Podocyte-specific Cldn5 knockout mice, diabetic nephropathy and obstruction models, ZONAB localization, WIF1 rescue","pmids":["35332151"],"confidence":"High","gaps":["How CLDN5 controls ZO1 levels and ZONAB sequestration mechanistically not defined","Whether the same axis operates in endothelial cells not tested"]},{"year":2023,"claim":"Established a light-triggered proteolytic mechanism of barrier disruption: blue light releases ADAM17 from its inhibitory partner GNAZ, leading to ADAM17-mediated CLDN5 degradation and inner blood-retinal barrier loss.","evidence":"Endothelial blue-light exposure, GNAZ knockdown, ADAM17 genetic/pharmacological inhibition, in vivo iBRB leakage, electroretinography","pmids":["37095509"],"confidence":"High","gaps":["Whether ADAM17 cleaves CLDN5 directly or acts via an intermediate not resolved","Photoreceptive sensor upstream of GNAZ release not identified"]},{"year":2024,"claim":"Added DLL4-NOTCH-RBPJ as a developmental/maintenance transcriptional input to CLDN5 and BBB integrity, shown by CLDN5 downregulation and microvascular abnormalities in Dll4 haploinsufficient mice.","evidence":"Dll4+/LacZ mice plus human BMEC studies, vascular permeability assays, NOTCH pathway analysis","pmids":["38632887"],"confidence":"Medium","gaps":["Direct RBPJ occupancy at the CLDN5 locus not demonstrated","Integration with ERG and NF-κB inputs not addressed"]},{"year":2025,"claim":"Extended CLDN5 function to metabolic control, showing adipocyte CLDN5 governs YBX3 localization and IL10 transcription to drive paracrine thermogenic signaling.","evidence":"Adipocyte-specific Cldn5 knockout mice, YBX3 localization, IL10 promoter/3'-UTR binding, IL10R paracrine experiments, metabolic phenotyping","pmids":["40610440"],"confidence":"High","gaps":["How a junctional protein controls cytoplasmic-nuclear YBX3 partitioning not mechanistically defined","Whether CLDN5 binds YBX3 directly not shown"]},{"year":2026,"claim":"Defined a direct protein-stabilization role, showing podocyte CLDN5 binds β1-integrin via its intracellular loop and C-terminus and blocks HUWE1-mediated ubiquitination at K774 to prevent proteasomal degradation.","evidence":"Super-resolution colocalization, Co-IP, domain-mapping mutagenesis, ubiquitination and proteasome-inhibition assays, Cldn5-KO injury mouse models","pmids":["41539562"],"confidence":"High","gaps":["Whether CLDN5 competes with HUWE1 for the same site or sterically shields K774 not resolved","Generalizability of integrin protection beyond podocytes not tested"]},{"year":null,"claim":"How a single tight-junction protein toggles between paracellular barrier/channel function and its diverse non-junctional roles (integrin stabilization, WNT restraint, metabolic signaling) through distinct binding partners and domains remains unresolved.","evidence":"","pmids":[],"confidence":"Low","gaps":["No atomic structure of CLDN5 strands or the G60R pore confirmed experimentally","Determinants directing CLDN5 to junctional versus non-junctional functions unknown","Whether transcriptional, trafficking, and proteolytic regulators act combinatorially in the same cell not integrated"]}],"mechanism_profile":{"molecular_activity":[{"term_id":"GO:0005198","term_label":"structural molecule activity","supporting_discovery_ids":[0]},{"term_id":"GO:0005215","term_label":"transporter activity","supporting_discovery_ids":[5]},{"term_id":"GO:0098631","term_label":"cell adhesion mediator activity","supporting_discovery_ids":[12]}],"localization":[{"term_id":"GO:0005886","term_label":"plasma membrane","supporting_discovery_ids":[0,10]},{"term_id":"GO:0005829","term_label":"cytosol","supporting_discovery_ids":[4]}],"pathway":[{"term_id":"R-HSA-1500931","term_label":"Cell-Cell communication","supporting_discovery_ids":[0]},{"term_id":"R-HSA-162582","term_label":"Signal Transduction","supporting_discovery_ids":[6,8]},{"term_id":"R-HSA-168256","term_label":"Immune System","supporting_discovery_ids":[2,3]}],"complexes":["tight junction strand"],"partners":["ITGB1","HUWE1","ZO1","CAV1","ADAM17","GNAZ","YBX3"],"other_free_text":[]}},"prefetch_data":{"uniprot":{"accession":"O00501","full_name":"Claudin-5","aliases":["Transmembrane protein deleted in VCFS","TMDVCF"],"length_aa":218,"mass_kda":23.1,"function":"Plays a major role in tight junction-specific obliteration of the intercellular space","subcellular_location":"Cell junction, tight junction; Cell membrane","url":"https://www.uniprot.org/uniprotkb/O00501/entry"},"depmap":{"release":"DepMap","has_data":true,"is_common_essential":false,"resolved_as":"","url":"https://depmap.org/portal/gene/CLDN5","classification":"Not Classified","n_dependent_lines":1,"n_total_lines":1208,"dependency_fraction":0.0008278145695364238},"opencell":{"profiled":false,"resolved_as":"","ensg_id":"","cell_line_id":"","localizations":[],"interactors":[],"url":"https://opencell.sf.czbiohub.org/search/CLDN5","total_profiled":1310},"omim":[{"mim_id":"621447","title":"ZINC FINGER PROTEIN 787; ZNF787","url":"https://www.omim.org/entry/621447"},{"mim_id":"609131","title":"CLAUDIN 7; CLDN7","url":"https://www.omim.org/entry/609131"},{"mim_id":"603718","title":"CLAUDIN 1; CLDN1","url":"https://www.omim.org/entry/603718"},{"mim_id":"602910","title":"CLAUDIN 3; CLDN3","url":"https://www.omim.org/entry/602910"},{"mim_id":"602909","title":"CLAUDIN 4; CLDN4","url":"https://www.omim.org/entry/602909"}],"hpa":{"profiled":true,"resolved_as":"","reliability":"","locations":[],"tissue_specificity":"Tissue enhanced","tissue_distribution":"Detected in many","driving_tissues":[{"tissue":"adipose tissue","ntpm":424.4},{"tissue":"breast","ntpm":412.2},{"tissue":"choroid plexus","ntpm":406.9},{"tissue":"lung","ntpm":384.3}],"url":"https://www.proteinatlas.org/search/CLDN5"},"hgnc":{"alias_symbol":["CPETRL1","BEC1"],"prev_symbol":["AWAL","TMVCF"]},"alphafold":{"accession":"O00501","domains":[{"cath_id":"1.20.140.150","chopping":"2-28_75-154_162-191","consensus_level":"high","plddt":90.7592,"start":2,"end":191}],"viewer_url":"https://alphafold.ebi.ac.uk/entry/O00501","model_url":"https://alphafold.ebi.ac.uk/files/AF-O00501-F1-model_v6.cif","pae_url":"https://alphafold.ebi.ac.uk/files/AF-O00501-F1-predicted_aligned_error_v6.png","plddt_mean":83.75},"mouse_models":{"mgi_url":"https://www.informatics.jax.org/marker/summary?nomen=CLDN5","jax_strain_url":"https://www.jax.org/strain/search?query=CLDN5"},"sequence":{"accession":"O00501","fasta_url":"https://rest.uniprot.org/uniprotkb/O00501.fasta","uniprot_url":"https://www.uniprot.org/uniprotkb/O00501/entry","alphafold_viewer_url":"https://alphafold.ebi.ac.uk/entry/O00501"}},"corpus_meta":[{"pmid":"10508865","id":"PMC_10508865","title":"Endothelial claudin: claudin-5/TMVCF constitutes tight junction strands in endothelial cells.","date":"1999","source":"The Journal of cell biology","url":"https://pubmed.ncbi.nlm.nih.gov/10508865","citation_count":709,"is_preprint":false},{"pmid":"33280500","id":"PMC_33280500","title":"Autophagy alleviates hypoxia-induced blood-brain barrier injury via regulation of CLDN5 (claudin 5).","date":"2020","source":"Autophagy","url":"https://pubmed.ncbi.nlm.nih.gov/33280500","citation_count":152,"is_preprint":false},{"pmid":"22138107","id":"PMC_22138107","title":"TNF-alpha induced NFκB signaling and p65 (RelA) overexpression repress Cldn5 promoter in mouse brain endothelial cells.","date":"2011","source":"Cytokine","url":"https://pubmed.ncbi.nlm.nih.gov/22138107","citation_count":107,"is_preprint":false},{"pmid":"36978081","id":"PMC_36978081","title":"The CLDN5 gene at the blood-brain barrier in health and disease.","date":"2023","source":"Fluids and barriers of the CNS","url":"https://pubmed.ncbi.nlm.nih.gov/36978081","citation_count":85,"is_preprint":false},{"pmid":"22235125","id":"PMC_22235125","title":"ETS-related gene (ERG) controls endothelial cell permeability via transcriptional regulation of the claudin 5 (CLDN5) gene.","date":"2012","source":"The Journal of biological chemistry","url":"https://pubmed.ncbi.nlm.nih.gov/22235125","citation_count":77,"is_preprint":false},{"pmid":"24522189","id":"PMC_24522189","title":"Non-muscle Mlck is required for β-catenin- and FoxO1-dependent downregulation of Cldn5 in IL-1β-mediated barrier dysfunction in brain endothelial cells.","date":"2014","source":"Journal of cell science","url":"https://pubmed.ncbi.nlm.nih.gov/24522189","citation_count":69,"is_preprint":false},{"pmid":"35318077","id":"PMC_35318077","title":"Angelica sinensis polysaccharide improves rheumatoid arthritis by modifying the expression of intestinal Cldn5, Slit3 and Rgs18 through gut microbiota.","date":"2022","source":"International journal of biological macromolecules","url":"https://pubmed.ncbi.nlm.nih.gov/35318077","citation_count":60,"is_preprint":false},{"pmid":"33838873","id":"PMC_33838873","title":"Brain DNA Methylation Patterns in CLDN5 Associated With Cognitive Decline.","date":"2021","source":"Biological psychiatry","url":"https://pubmed.ncbi.nlm.nih.gov/33838873","citation_count":51,"is_preprint":false},{"pmid":"35332151","id":"PMC_35332151","title":"Loss of CLDN5 in podocytes deregulates WIF1 to activate WNT signaling and contributes to kidney disease.","date":"2022","source":"Nature communications","url":"https://pubmed.ncbi.nlm.nih.gov/35332151","citation_count":39,"is_preprint":false},{"pmid":"26060683","id":"PMC_26060683","title":"Polymorphism of the CLDN5 gene and Schizophrenia in an Iranian Population.","date":"2014","source":"Iranian journal of public health","url":"https://pubmed.ncbi.nlm.nih.gov/26060683","citation_count":28,"is_preprint":false},{"pmid":"36472886","id":"PMC_36472886","title":"Inflammation and Blood-Brain Barrier in Depression: Interaction of CLDN5 and IL6 Gene Variants in Stress-Induced Depression.","date":"2023","source":"The international journal of neuropsychopharmacology","url":"https://pubmed.ncbi.nlm.nih.gov/36472886","citation_count":28,"is_preprint":false},{"pmid":"35714222","id":"PMC_35714222","title":"Recurrent de novo mutations in CLDN5 induce an anion-selective blood-brain barrier and alternating hemiplegia.","date":"2022","source":"Brain : a journal of neurology","url":"https://pubmed.ncbi.nlm.nih.gov/35714222","citation_count":23,"is_preprint":false},{"pmid":"34156149","id":"PMC_34156149","title":"Super-resolved local recruitment of CLDN5 to filtration slits implicates a direct relationship with podocyte foot process effacement.","date":"2021","source":"Journal of cellular and molecular medicine","url":"https://pubmed.ncbi.nlm.nih.gov/34156149","citation_count":22,"is_preprint":false},{"pmid":"29328410","id":"PMC_29328410","title":"CLDN5 affects lncRNAs acting as ceRNA dynamics contributing to regulating blood‑brain barrier permeability in tumor brain metastasis.","date":"2018","source":"Oncology reports","url":"https://pubmed.ncbi.nlm.nih.gov/29328410","citation_count":19,"is_preprint":false},{"pmid":"36477332","id":"PMC_36477332","title":"Variants in CLDN5 cause a syndrome characterized by seizures, microcephaly and brain calcifications.","date":"2023","source":"Brain : a journal of neurology","url":"https://pubmed.ncbi.nlm.nih.gov/36477332","citation_count":18,"is_preprint":false},{"pmid":"28490769","id":"PMC_28490769","title":"Increased cerebral expressions of MMPs, CLDN5, OCLN, ZO1 and AQPs are associated with brain edema following fatal heat stroke.","date":"2017","source":"Scientific reports","url":"https://pubmed.ncbi.nlm.nih.gov/28490769","citation_count":16,"is_preprint":false},{"pmid":"16181776","id":"PMC_16181776","title":"A study of the combined effect of the CLDN5 locus and the genes for the phospholipid metabolism pathway in schizophrenia.","date":"2005","source":"Prostaglandins, leukotrienes, and essential fatty acids","url":"https://pubmed.ncbi.nlm.nih.gov/16181776","citation_count":15,"is_preprint":false},{"pmid":"37095509","id":"PMC_37095509","title":"Blue light exposure collapses the inner blood-retinal barrier by accelerating endothelial CLDN5 degradation through the disturbance of GNAZ and the activation of ADAM17.","date":"2023","source":"Fluids and barriers of the CNS","url":"https://pubmed.ncbi.nlm.nih.gov/37095509","citation_count":15,"is_preprint":false},{"pmid":"38228255","id":"PMC_38228255","title":"CLDN5: From structure and regulation to roles in tumors and other diseases beyond CNS disorders.","date":"2024","source":"Pharmacological research","url":"https://pubmed.ncbi.nlm.nih.gov/38228255","citation_count":14,"is_preprint":false},{"pmid":"39674501","id":"PMC_39674501","title":"CD34+CLDN5+ tumor associated senescent endothelial cells through IGF2-IGF2R signaling increased cholangiocellular phenotype in hepatocellular carcinoma.","date":"2024","source":"Journal of advanced research","url":"https://pubmed.ncbi.nlm.nih.gov/39674501","citation_count":14,"is_preprint":false},{"pmid":"32170581","id":"PMC_32170581","title":"Tight Junction-Related CLDN5 and CLDN6 Genes, and Gap Junction-Related GJB6 and GJB7 Genes Are Somatically Mutated in Gastric and Colorectal Cancers.","date":"2020","source":"Pathology oncology research : POR","url":"https://pubmed.ncbi.nlm.nih.gov/32170581","citation_count":14,"is_preprint":false},{"pmid":"39665142","id":"PMC_39665142","title":"Extracellular Vesicles From Preeclampsia Disrupt the Blood-Brain Barrier by Reducing CLDN5.","date":"2024","source":"Arteriosclerosis, thrombosis, and vascular biology","url":"https://pubmed.ncbi.nlm.nih.gov/39665142","citation_count":11,"is_preprint":false},{"pmid":"33383868","id":"PMC_33383868","title":"Serotonin/5-HT1A Signaling in the Neurovascular Unit Regulates Endothelial CLDN5 Expression.","date":"2020","source":"International journal of molecular sciences","url":"https://pubmed.ncbi.nlm.nih.gov/33383868","citation_count":11,"is_preprint":false},{"pmid":"28445614","id":"PMC_28445614","title":"Two common human CLDN5 alleles encode different open reading frames but produce one protein isoform.","date":"2017","source":"Annals of the New York Academy of Sciences","url":"https://pubmed.ncbi.nlm.nih.gov/28445614","citation_count":9,"is_preprint":false},{"pmid":"38632887","id":"PMC_38632887","title":"Delta like 4 regulates cerebrovascular development and endothelial integrity via DLL4-NOTCH-CLDN5 pathway and is vulnerable to neonatal hyperoxia.","date":"2024","source":"The Journal of physiology","url":"https://pubmed.ncbi.nlm.nih.gov/38632887","citation_count":6,"is_preprint":false},{"pmid":"39147772","id":"PMC_39147772","title":"Loss of Cldn5 -and increase in Irf7-in the hippocampus and cerebral cortex of diabetic mice at the early symptomatic stage.","date":"2024","source":"Nutrition & diabetes","url":"https://pubmed.ncbi.nlm.nih.gov/39147772","citation_count":5,"is_preprint":false},{"pmid":"40610440","id":"PMC_40610440","title":"Adipocyte CLDN5 promotes thermogenesis and energy expenditure through regulation of IL10 expression.","date":"2025","source":"Nature communications","url":"https://pubmed.ncbi.nlm.nih.gov/40610440","citation_count":3,"is_preprint":false},{"pmid":"42262642","id":"PMC_42262642","title":"A longitudinal study of CLDN5 DNA methylation and PTSD.","date":"2026","source":"Journal of behavioral medicine","url":"https://pubmed.ncbi.nlm.nih.gov/42262642","citation_count":0,"is_preprint":false},{"pmid":"41774611","id":"PMC_41774611","title":"Investigation of <italic>CLDN5</italic> Gene 3' Untranslated Region rs10314 Polymorphism and Protein Levels in Children and Adolescents with Attention-Deficit Hyperactivity Disorder: A Molecular and Clinical Correlation Study.","date":"2026","source":"Medical principles and practice : international journal of the Kuwait University, Health Science Centre","url":"https://pubmed.ncbi.nlm.nih.gov/41774611","citation_count":0,"is_preprint":false},{"pmid":"41539562","id":"PMC_41539562","title":"CLDN5 as a novel modulator of podocyte adhesion to extracellular matrix via β1-integrin binding.","date":"2026","source":"The Journal of biological chemistry","url":"https://pubmed.ncbi.nlm.nih.gov/41539562","citation_count":0,"is_preprint":false},{"pmid":"41035457","id":"PMC_41035457","title":"High-intensity interval training improves cognitive dysfunction in chronically stressed mice through alleviating homocysteine-induced transcriptional repression of Cldn5.","date":"2025","source":"Neurobiology of stress","url":"https://pubmed.ncbi.nlm.nih.gov/41035457","citation_count":0,"is_preprint":false},{"pmid":null,"id":"bio_10.1101_2025.11.24.25340875","title":"Trauma Exposure, PTSD, and Methylation of the Blood Brain Barrier Claudin-5 Gene","date":"2025-11-25","source":"bioRxiv","url":"https://doi.org/10.1101/2025.11.24.25340875","citation_count":0,"is_preprint":true},{"pmid":null,"id":"bio_10.1101_2025.08.13.670217","title":"Cerebrovascular Claudin-5 Isoform Expression Correlates with Worsened Stroke Outcomes Following Thromboembolic Stroke","date":"2025-08-18","source":"bioRxiv","url":"https://doi.org/10.1101/2025.08.13.670217","citation_count":0,"is_preprint":true},{"pmid":null,"id":"bio_10.1101_2025.10.21.683634","title":"Human cortico-vascular assembloids reveal a CELF2-AHNAK-dependent switch from neuronal to endothelial tropism in glioblastoma cells","date":"2025-10-22","source":"bioRxiv","url":"https://doi.org/10.1101/2025.10.21.683634","citation_count":0,"is_preprint":true},{"pmid":null,"id":"bio_10.1101_2024.12.21.629919","title":"Stress-induced mitochondrial fragmentation in endothelial cells disrupts blood-retinal barrier integrity causing neurodegeneration","date":"2024-12-22","source":"bioRxiv","url":"https://doi.org/10.1101/2024.12.21.629919","citation_count":0,"is_preprint":true},{"pmid":null,"id":"bio_10.1101_2025.03.12.642911","title":"Spatial Transcriptomics of Schizophrenia Insular Cortex Reveals Blood-Brain Barrier Hyperglycolysis and Increased Parenchymal Mitochondrial Respiration","date":"2025-03-13","source":"bioRxiv","url":"https://doi.org/10.1101/2025.03.12.642911","citation_count":0,"is_preprint":true},{"pmid":null,"id":"bio_10.1101_2025.11.17.688818","title":"Vascular Basement Membrane Laminins Modulate Functional Zonation of Cerebral Microvessels","date":"2025-11-17","source":"bioRxiv","url":"https://doi.org/10.1101/2025.11.17.688818","citation_count":0,"is_preprint":true},{"pmid":null,"id":"bio_10.1101_2025.07.10.664153","title":"Generative cerebral vasculature visualization using spatial transcriptomic data","date":"2025-07-15","source":"bioRxiv","url":"https://doi.org/10.1101/2025.07.10.664153","citation_count":0,"is_preprint":true},{"pmid":null,"id":"bio_10.1101_2025.01.29.635460","title":"A multi-pore model of the blood-brain barrier tight junction strands recapitulates the permeability features of wild-type and mutant claudin-5","date":"2025-01-29","source":"bioRxiv","url":"https://doi.org/10.1101/2025.01.29.635460","citation_count":0,"is_preprint":true}],"cost":{"stage1":{"model":"claude-sonnet-4-6","input_tokens":23502,"output_tokens":3996,"usd":0.065223,"stage1_stop_reason":"end_turn"},"stage2":{"model":"claude-opus-4-8","input_tokens":11839,"output_tokens":4313,"usd":0.08351,"stage2_stop_reason":"end_turn"},"total_usd":0.148733,"stage1_batch_id":"msgbatch_01ACAxJQabDsht39JXeWG22d","stage2_batch_id":"msgbatch_01QEU22A9U5sbPTgsMzTmvSf","note":"batch pricing = 50% of standard"},"stage1_raw":"```json\n{\n  \"discoveries\": [\n    {\n      \"year\": 1999,\n      \"finding\": \"CLDN5/TMVCF localizes exclusively to tight junction strands in endothelial cells of blood vessels (not epithelial cells), and when cDNA is introduced into mouse L fibroblasts, it reconstitutes extracellular face-associated TJ strands, establishing CLDN5 as an endothelial cell-specific structural component of tight junction strands.\",\n      \"method\": \"Immunofluorescence microscopy, immunoreplica electron microscopy, cDNA transfection into L fibroblasts (reconstitution assay)\",\n      \"journal\": \"The Journal of cell biology\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 1 / Strong — direct reconstitution in heterologous cells combined with immunoelectron microscopy localization; foundational study replicated broadly across subsequent literature\",\n      \"pmids\": [\"10508865\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2012,\n      \"finding\": \"The ETS transcription factor ERG directly regulates CLDN5 gene expression in endothelial cells; ERG knockdown reduces CLDN5 expression, increases endothelial permeability, and promotes stress fiber and gap formation, placing ERG upstream of CLDN5 in the transcriptional control of barrier function.\",\n      \"method\": \"ERG siRNA knockdown in endothelial cells, permeability assays, stress fiber/gap quantification, transcriptional reporter assays\",\n      \"journal\": \"The Journal of biological chemistry\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 2 / Moderate — genetic loss-of-function (knockdown) with defined cellular permeability phenotype and transcriptional downstream target identification, single lab\",\n      \"pmids\": [\"22235125\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2011,\n      \"finding\": \"TNF-α reduces CLDN5 promoter activity and mRNA expression in brain endothelial cells via NF-κB signaling; overexpression of the NF-κB subunit p65 (RelA) alone is sufficient to repress the Cldn5 promoter, and a conserved promoter region is required for this downregulation.\",\n      \"method\": \"Promoter-reporter assays, p65 overexpression, TNF-α treatment of primary brain endothelial cells, qPCR, western blot\",\n      \"journal\": \"Cytokine\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 2 / Moderate — promoter activity assays and gain-of-function p65 overexpression with two orthogonal readouts (promoter activity and mRNA), single lab\",\n      \"pmids\": [\"22138107\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2014,\n      \"finding\": \"IL-1β–induced CLDN5 transcriptional repression in brain microvascular endothelial cells requires non-muscle MLCK (nmMlck): nmMlck mediates nuclear translocation of β-catenin and FoxO1, which then repress Cldn5 transcription, and primary BMVECs from nmMlck-null mice are protected from this repression.\",\n      \"method\": \"Primary BMVEC isolation from nmMlck-null mice (genetic loss-of-function), IL-1β treatment, nuclear translocation assays for β-catenin and FoxO1, Cldn5 mRNA/protein quantification, barrier permeability assays\",\n      \"journal\": \"Journal of cell science\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 2 / Strong — genetic knockout cells with specific phenotypic rescue, multiple orthogonal methods (nuclear fractionation, transcription, permeability), single lab but rigorous design\",\n      \"pmids\": [\"24522189\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2020,\n      \"finding\": \"Under hypoxia, CAV1 (caveolin-1) mediates redistribution of membranous CLDN5 into the cytosol in brain microvascular endothelial cells; autophagy then degrades CAV1 and cytosolic aggregated CLDN5. Blocking autophagy genetically or pharmacologically aggravates cytosolic CLDN5 accumulation and worsens blood-brain barrier breakdown.\",\n      \"method\": \"In vivo zebrafish BBB models, in vitro BMEC culture, CAV1 genetic manipulation, autophagy blockade (chemical and genetic), STED super-resolution microscopy, TEER measurements\",\n      \"journal\": \"Autophagy\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 2 / Strong — multiple orthogonal methods (in vivo zebrafish, in vitro cells, genetic and chemical autophagy blockade, super-resolution imaging, TEER), two model systems\",\n      \"pmids\": [\"33280500\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2022,\n      \"finding\": \"A de novo missense mutation G60R in the first extracellular loop of CLDN5 converts its function from a purely barrier-forming tight junction protein to an anion-selective channel: stably transfected cell lines with G60R CLDN5 still form tight junctions but display attenuated small-molecule barrier with higher Cl− and lower Na+ permeability, representing the first gain-of-function mutation in the claudin gene family.\",\n      \"method\": \"Stable transfection of wild-type and G60R CLDN5 cell lines, ion permeability electrophysiology, protein structural modeling, sequence alignment, barrier function assays\",\n      \"journal\": \"Brain : a journal of neurology\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 1 / Moderate — reconstitution of mutant protein in stably transfected cells with direct electrophysiological measurement of ion permeability and barrier function; single lab but multiple orthogonal methods\",\n      \"pmids\": [\"35714222\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2022,\n      \"finding\": \"In podocytes, CLDN5 deletion reduces ZO1 expression and induces nuclear translocation of ZONAB, which transcriptionally downregulates WIF1 (WNT inhibitory factor-1), leading to activation of the WNT signaling pathway; podocyte-derived WIF1 also acts in paracrine on tubular epithelial cells to suppress fibrosis.\",\n      \"method\": \"Podocyte-specific Cldn5 knockout mice, diabetic nephropathy and ureteral obstruction mouse models, ZONAB nuclear localization assays, WIF1 transcriptional analysis, systemic WIF1 delivery rescue experiments\",\n      \"journal\": \"Nature communications\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 2 / Strong — conditional knockout mouse model, epistasis via WIF1 rescue, multiple in vivo models, paracrine function established by co-culture and in vivo rescue\",\n      \"pmids\": [\"35332151\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2023,\n      \"finding\": \"Blue light exposure causes rapid ADAM17-mediated degradation of endothelial CLDN5, disrupting tight junctions and the inner blood-retinal barrier. Under basal conditions ADAM17 is sequestered by the inhibitory G protein GNAZ; blue light releases ADAM17 from GNAZ. GNAZ knockdown phenocopies blue light effects (ADAM17 hyperactivation, CLDN5 downregulation, barrier permeability). Pharmacological or genetic inhibition of ADAM17 prevents CLDN5 degradation.\",\n      \"method\": \"In vitro endothelial cell blue-light exposure, GNAZ knockdown, ADAM17 pharmacological and genetic inhibition, western blot for CLDN5 protein levels, in vivo mouse iBRB leakage, electroretinography\",\n      \"journal\": \"Fluids and barriers of the CNS\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 2 / Strong — genetic knockdown, pharmacological inhibition, and in vivo model with multiple functional readouts (barrier integrity, electroretinogram); two model systems\",\n      \"pmids\": [\"37095509\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2024,\n      \"finding\": \"DLL4-NOTCH signaling regulates endothelial CLDN5 expression and blood-brain barrier integrity through the NOTCH-NICD-RBPJ-CLDN5 pathway; DLL4 insufficiency (Dll4+/LacZ mice) leads to downregulation of CLDN5, persistent abnormalities in brain microvasculature, and increased vascular permeability both in vivo and in vitro.\",\n      \"method\": \"Dll4+/LacZ and Dll4+/+ mouse models, human brain microvascular endothelial cell in vitro studies, in vivo vascular permeability assays, NOTCH pathway component analysis\",\n      \"journal\": \"The Journal of physiology\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 2 / Moderate — genetic heterozygous mouse model plus in vitro corroboration, pathway placed via NOTCH-NICD-RBPJ intermediates, single lab\",\n      \"pmids\": [\"38632887\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2020,\n      \"finding\": \"Serotonin/5-HT1A signaling enhances endothelial CLDN5 expression in brain microvascular endothelial cells; the 5-HT1A receptor is expressed in BMVECs and mural cells, and PKA is aberrantly activated in both cell types in schizophrenic prefrontal cortex, coinciding with site-selective CLDN5 breakdown.\",\n      \"method\": \"Immunohistochemistry on post-mortem human brain tissue, two-dimensional co-culture of BMVECs and pericytes, pharmacological 5-HT1A receptor stimulation, CLDN5 expression quantification\",\n      \"journal\": \"International journal of molecular sciences\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 3 / Moderate — pharmacological modulation in co-culture with human tissue immunohistochemistry confirmation; single lab, mechanistic depth limited by abstract\",\n      \"pmids\": [\"33383868\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2017,\n      \"finding\": \"The human CLDN5 gene contains a high-frequency SNP (rs885985) creating two ORFs (303 or 218 amino acid isoforms), but only the 218 aa form is detected by immunoblot in human lung tissue. When forcibly expressed in transfected cells, the long 303 aa form is retained in intracellular compartments and does not localize to the plasma membrane, in contrast to the 218 aa form which localizes to intercellular junctions.\",\n      \"method\": \"Genotyping of human lung tissue, immunoblot, transfection of long- and short-form CLDN5 constructs in cells with subcellular localization imaging\",\n      \"journal\": \"Annals of the New York Academy of Sciences\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 2 / Moderate — direct protein expression experiment in transfected cells plus human tissue immunoblot; single lab, two orthogonal approaches\",\n      \"pmids\": [\"28445614\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2025,\n      \"finding\": \"CLDN5 in adipocytes (non-tight junction context) affects subcellular localization of Y-box protein 3 (YBX3); CLDN5 deficiency causes YBX3 mislocalization, which reduces IL10 expression (YBX3 directly binds the IL10 promoter and 3'-UTR); secreted IL10 then acts in paracrine via IL10R on neighboring thermogenic adipocytes to regulate thermogenesis and energy expenditure.\",\n      \"method\": \"Adipocyte-specific Cldn5 knockout mice, YBX3 localization assays, IL10 promoter binding assays, IL10R paracrine signaling experiments, metabolic phenotyping (thermogenesis, energy expenditure, glucose tolerance)\",\n      \"journal\": \"Nature communications\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 2 / Strong — conditional knockout mouse, promoter binding assay, paracrine signaling rescue, and metabolic phenotyping providing multiple orthogonal lines of evidence in a single rigorous study\",\n      \"pmids\": [\"40610440\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2026,\n      \"finding\": \"CLDN5 in podocytes forms a stable complex with β1-integrin via its intracellular loop and C-terminal domains; CLDN5 prevents HUWE1-mediated ubiquitination at lysine K774 of β1-integrin, thereby protecting β1-integrin from proteasomal degradation and ensuring its proper membrane localization. CLDN5 deletion impairs podocyte adhesion, spreading, and mechanical stress resistance in vitro, and worsens renal injury in vivo.\",\n      \"method\": \"Super-resolution imaging (colocalization at podocyte-GBM interface), Co-IP (stable complex), domain-mapping mutagenesis (intracellular loop and C-terminal domains), ubiquitination assays (HUWE1, K774), proteasome inhibition, Cldn5-KO mice with hypertensive and adriamycin injury models\",\n      \"journal\": \"The Journal of biological chemistry\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 1 / Strong — Co-IP with domain mutagenesis, ubiquitination site identification, proteasomal degradation rescue, super-resolution localization, and in vivo KO model; multiple orthogonal methods in single rigorous study\",\n      \"pmids\": [\"41539562\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2025,\n      \"finding\": \"Molecular dynamics free energy calculations show that multi-pore Claudin-5 TJ strand models create electrostatic barriers to ion permeation higher than in single-pore architectures, and only the multi-Pore I structural model recapitulates the anion-selective permeability phenotype of the G60R variant, providing structural validation of the paracellular pore mechanism.\",\n      \"method\": \"Multi-microsecond all-atom molecular dynamics simulations, free energy calculations for water and ion permeation across multi-protomer (16-subunit) claudin-5 TJ models\",\n      \"journal\": \"bioRxiv\",\n      \"confidence\": \"Low\",\n      \"confidence_rationale\": \"Tier 4 / Weak — computational/simulation only, no experimental validation reported; preprint\",\n      \"pmids\": [],\n      \"is_preprint\": true\n    }\n  ],\n  \"current_model\": \"CLDN5 is an endothelial cell-enriched four-transmembrane tight junction protein that forms paracellular barrier strands (reconstituted in fibroblasts) and is transcriptionally regulated by NF-κB/p65, ERG, and NOTCH-RBPJ pathways; at the BBB its membrane levels are controlled by CAV1-mediated redistribution and autophagy-mediated degradation, and its tight-junction barrier function can be converted to anion-selective channel activity by the gain-of-function G60R mutation in its first extracellular loop; beyond endothelial junctions, CLDN5 in podocytes stabilizes β1-integrin by blocking HUWE1-mediated ubiquitination and activates a ZO1-ZONAB-WIF1 axis to suppress WNT signaling, while in adipocytes it controls thermogenesis by regulating YBX3 localization and paracrine IL10 signaling.\"\n}\n```","stage2_raw":"{\n  \"mechanistic_narrative\": \"CLDN5 is an endothelial cell-enriched, four-transmembrane claudin that serves as a structural component of tight junction strands and establishes paracellular barriers, most prominently the blood-brain and inner blood-retinal barriers [#0]. Its barrier-forming activity was defined by reconstitution of extracellular face-associated tight junction strands when CLDN5 cDNA is introduced into fibroblasts lacking endogenous junctions [#0]. CLDN5 abundance at endothelial junctions is set by convergent transcriptional inputs — it is activated by the ETS factor ERG [#1] and by DLL4-NOTCH signaling through an NICD-RBPJ axis [#8], and repressed by inflammatory NF-\\u03baB/p65 signaling downstream of TNF-\\u03b1 [#2] and by an IL-1\\u03b2\\u2013nmMLCK\\u2013\\u03b2-catenin/FoxO1 pathway [#3]. Post-transcriptionally, junctional CLDN5 is removed by CAV1-mediated redistribution into the cytosol followed by autophagic degradation under hypoxia [#4] and by ADAM17-mediated proteolysis that is normally restrained by the inhibitory G protein GNAZ [#7]. A de novo G60R mutation in the first extracellular loop converts CLDN5 from a pure barrier protein into an anion-selective paracellular channel, the first gain-of-function mutation described in the claudin family [#5]. Beyond endothelial junctions, CLDN5 has distinct non-junctional roles: in podocytes it stabilizes \\u03b21-integrin by blocking HUWE1-mediated ubiquitination at K774 [#12] and governs a ZO1-ZONAB-WIF1 axis that restrains WNT signaling [#6], and in adipocytes it controls YBX3 localization and IL10-dependent paracrine thermogenic signaling [#11].\",\n  \"teleology\": [\n    {\n      \"year\": 1999,\n      \"claim\": \"Established CLDN5 as an endothelial-specific structural building block of tight junction strands rather than a passive marker, by showing it is sufficient to nucleate junction strands in cells that have none.\",\n      \"evidence\": \"Immunofluorescence and immunoreplica EM localization plus cDNA reconstitution in L fibroblasts\",\n      \"pmids\": [\"10508865\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\"Did not define the heterotypic claudin partners or ZO adaptors required in vivo\", \"No mechanism for how strand architecture sets ion or size selectivity\"]\n    },\n    {\n      \"year\": 2011,\n      \"claim\": \"Identified inflammatory transcriptional repression as a route to barrier loss, showing NF-\\u03baB/p65 is sufficient to silence the Cldn5 promoter downstream of TNF-\\u03b1.\",\n      \"evidence\": \"Promoter-reporter assays, p65 overexpression, TNF-\\u03b1 treatment of brain endothelial cells with qPCR/western readouts\",\n      \"pmids\": [\"22138107\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\"Exact p65 binding element within the conserved promoter region not mapped\", \"Single lab, did not test in vivo barrier consequences\"]\n    },\n    {\n      \"year\": 2012,\n      \"claim\": \"Placed the ETS factor ERG upstream as a positive transcriptional driver of CLDN5 required for endothelial barrier maintenance.\",\n      \"evidence\": \"ERG siRNA knockdown with permeability assays, stress-fiber/gap quantification, and reporter assays\",\n      \"pmids\": [\"22235125\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\"Direct ERG occupancy of the CLDN5 promoter not resolved beyond reporter assays\", \"Relationship to NF-\\u03baB repression not integrated\"]\n    },\n    {\n      \"year\": 2014,\n      \"claim\": \"Resolved the signaling chain linking IL-1\\u03b2 to CLDN5 repression, identifying nmMLCK as the obligatory transducer driving nuclear \\u03b2-catenin/FoxO1 to silence Cldn5.\",\n      \"evidence\": \"Primary BMVECs from nmMlck-null mice, IL-1\\u03b2 treatment, nuclear translocation assays, transcription and permeability readouts\",\n      \"pmids\": [\"24522189\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\"Whether \\u03b2-catenin/FoxO1 bind the Cldn5 promoter directly not shown\", \"How a cytoskeletal kinase controls transcription factor nuclear entry not mechanistically detailed\"]\n    },\n    {\n      \"year\": 2020,\n      \"claim\": \"Defined a post-transcriptional turnover route for junctional CLDN5, showing CAV1 redistributes membranous CLDN5 to the cytosol and autophagy clears the aggregates, with autophagy blockade worsening barrier breakdown.\",\n      \"evidence\": \"Zebrafish BBB and in vitro BMEC models, CAV1 manipulation, genetic/chemical autophagy blockade, STED imaging, TEER\",\n      \"pmids\": [\"33280500\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\"Trigger coupling hypoxia to CAV1 engagement of CLDN5 not defined\", \"Whether degraded CLDN5 is recycled or replaced not addressed\"]\n    },\n    {\n      \"year\": 2020,\n      \"claim\": \"Linked GPCR signaling to CLDN5 levels, showing serotonin/5-HT1A signaling enhances endothelial CLDN5 and implicating aberrant PKA activation in disease-associated barrier breakdown.\",\n      \"evidence\": \"Post-mortem human brain immunohistochemistry, BMVEC-pericyte co-culture, pharmacological 5-HT1A stimulation\",\n      \"pmids\": [\"33383868\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\"Causal chain from 5-HT1A/PKA to CLDN5 transcription or stability not dissected\", \"Correlative human tissue data, not interventional\"]\n    },\n    {\n      \"year\": 2022,\n      \"claim\": \"Demonstrated CLDN5 can be a conduction element, not just a seal, by showing the G60R extracellular-loop mutation creates an anion-selective paracellular channel — the first claudin gain-of-function.\",\n      \"evidence\": \"Stable WT and G60R CLDN5 cell lines, ion permeability electrophysiology, structural modeling, barrier assays\",\n      \"pmids\": [\"35714222\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\"Atomic structure of the conducting pore not solved\", \"In vivo neurological consequences of altered ion flux not established\"]\n    },\n    {\n      \"year\": 2022,\n      \"claim\": \"Uncovered a non-junctional signaling role in podocytes, where CLDN5 sustains ZO1 and restrains ZONAB-driven WIF1 repression to keep WNT signaling off, with paracrine WIF1 limiting tubular fibrosis.\",\n      \"evidence\": \"Podocyte-specific Cldn5 knockout mice, diabetic nephropathy and obstruction models, ZONAB localization, WIF1 rescue\",\n      \"pmids\": [\"35332151\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\"How CLDN5 controls ZO1 levels and ZONAB sequestration mechanistically not defined\", \"Whether the same axis operates in endothelial cells not tested\"]\n    },\n    {\n      \"year\": 2023,\n      \"claim\": \"Established a light-triggered proteolytic mechanism of barrier disruption: blue light releases ADAM17 from its inhibitory partner GNAZ, leading to ADAM17-mediated CLDN5 degradation and inner blood-retinal barrier loss.\",\n      \"evidence\": \"Endothelial blue-light exposure, GNAZ knockdown, ADAM17 genetic/pharmacological inhibition, in vivo iBRB leakage, electroretinography\",\n      \"pmids\": [\"37095509\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\"Whether ADAM17 cleaves CLDN5 directly or acts via an intermediate not resolved\", \"Photoreceptive sensor upstream of GNAZ release not identified\"]\n    },\n    {\n      \"year\": 2024,\n      \"claim\": \"Added DLL4-NOTCH-RBPJ as a developmental/maintenance transcriptional input to CLDN5 and BBB integrity, shown by CLDN5 downregulation and microvascular abnormalities in Dll4 haploinsufficient mice.\",\n      \"evidence\": \"Dll4+/LacZ mice plus human BMEC studies, vascular permeability assays, NOTCH pathway analysis\",\n      \"pmids\": [\"38632887\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\"Direct RBPJ occupancy at the CLDN5 locus not demonstrated\", \"Integration with ERG and NF-\\u03baB inputs not addressed\"]\n    },\n    {\n      \"year\": 2025,\n      \"claim\": \"Extended CLDN5 function to metabolic control, showing adipocyte CLDN5 governs YBX3 localization and IL10 transcription to drive paracrine thermogenic signaling.\",\n      \"evidence\": \"Adipocyte-specific Cldn5 knockout mice, YBX3 localization, IL10 promoter/3'-UTR binding, IL10R paracrine experiments, metabolic phenotyping\",\n      \"pmids\": [\"40610440\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\"How a junctional protein controls cytoplasmic-nuclear YBX3 partitioning not mechanistically defined\", \"Whether CLDN5 binds YBX3 directly not shown\"]\n    },\n    {\n      \"year\": 2026,\n      \"claim\": \"Defined a direct protein-stabilization role, showing podocyte CLDN5 binds \\u03b21-integrin via its intracellular loop and C-terminus and blocks HUWE1-mediated ubiquitination at K774 to prevent proteasomal degradation.\",\n      \"evidence\": \"Super-resolution colocalization, Co-IP, domain-mapping mutagenesis, ubiquitination and proteasome-inhibition assays, Cldn5-KO injury mouse models\",\n      \"pmids\": [\"41539562\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\"Whether CLDN5 competes with HUWE1 for the same site or sterically shields K774 not resolved\", \"Generalizability of integrin protection beyond podocytes not tested\"]\n    },\n    {\n      \"year\": null,\n      \"claim\": \"How a single tight-junction protein toggles between paracellular barrier/channel function and its diverse non-junctional roles (integrin stabilization, WNT restraint, metabolic signaling) through distinct binding partners and domains remains unresolved.\",\n      \"evidence\": null,\n      \"pmids\": [],\n      \"confidence\": \"Low\",\n      \"gaps\": [\"No atomic structure of CLDN5 strands or the G60R pore confirmed experimentally\", \"Determinants directing CLDN5 to junctional versus non-junctional functions unknown\", \"Whether transcriptional, trafficking, and proteolytic regulators act combinatorially in the same cell not integrated\"]\n    }\n  ],\n  \"mechanism_profile\": {\n    \"molecular_activity\": [\n      {\"term_id\": \"GO:0005198\", \"supporting_discovery_ids\": [0]},\n      {\"term_id\": \"GO:0005215\", \"supporting_discovery_ids\": [5]},\n      {\"term_id\": \"GO:0098631\", \"supporting_discovery_ids\": [12]}\n    ],\n    \"localization\": [\n      {\"term_id\": \"GO:0005886\", \"supporting_discovery_ids\": [0, 10]},\n      {\"term_id\": \"GO:0005829\", \"supporting_discovery_ids\": [4]}\n    ],\n    \"pathway\": [\n      {\"term_id\": \"R-HSA-1500931\", \"supporting_discovery_ids\": [0]},\n      {\"term_id\": \"R-HSA-162582\", \"supporting_discovery_ids\": [6, 8]},\n      {\"term_id\": \"R-HSA-168256\", \"supporting_discovery_ids\": [2, 3]}\n    ],\n    \"complexes\": [\"tight junction strand\"],\n    \"partners\": [\"ITGB1\", \"HUWE1\", \"ZO1\", \"CAV1\", \"ADAM17\", \"GNAZ\", \"YBX3\"],\n    \"other_free_text\": []\n  }\n}","audit_flag":null,"evaluation":{"pairwise":"win","faith_supported":6,"faith_total":6,"faith_pct":100.0}}