| 1999 |
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. |
Immunofluorescence microscopy, immunoreplica electron microscopy, cDNA transfection into L fibroblasts (reconstitution assay) |
The Journal of cell biology |
High |
10508865
|
| 2012 |
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. |
ERG siRNA knockdown in endothelial cells, permeability assays, stress fiber/gap quantification, transcriptional reporter assays |
The Journal of biological chemistry |
Medium |
22235125
|
| 2011 |
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. |
Promoter-reporter assays, p65 overexpression, TNF-α treatment of primary brain endothelial cells, qPCR, western blot |
Cytokine |
Medium |
22138107
|
| 2014 |
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. |
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 of cell science |
High |
24522189
|
| 2020 |
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. |
In vivo zebrafish BBB models, in vitro BMEC culture, CAV1 genetic manipulation, autophagy blockade (chemical and genetic), STED super-resolution microscopy, TEER measurements |
Autophagy |
High |
33280500
|
| 2022 |
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. |
Stable transfection of wild-type and G60R CLDN5 cell lines, ion permeability electrophysiology, protein structural modeling, sequence alignment, barrier function assays |
Brain : a journal of neurology |
High |
35714222
|
| 2022 |
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. |
Podocyte-specific Cldn5 knockout mice, diabetic nephropathy and ureteral obstruction mouse models, ZONAB nuclear localization assays, WIF1 transcriptional analysis, systemic WIF1 delivery rescue experiments |
Nature communications |
High |
35332151
|
| 2023 |
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. |
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 |
Fluids and barriers of the CNS |
High |
37095509
|
| 2024 |
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. |
Dll4+/LacZ and Dll4+/+ mouse models, human brain microvascular endothelial cell in vitro studies, in vivo vascular permeability assays, NOTCH pathway component analysis |
The Journal of physiology |
Medium |
38632887
|
| 2020 |
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. |
Immunohistochemistry on post-mortem human brain tissue, two-dimensional co-culture of BMVECs and pericytes, pharmacological 5-HT1A receptor stimulation, CLDN5 expression quantification |
International journal of molecular sciences |
Medium |
33383868
|
| 2017 |
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. |
Genotyping of human lung tissue, immunoblot, transfection of long- and short-form CLDN5 constructs in cells with subcellular localization imaging |
Annals of the New York Academy of Sciences |
Medium |
28445614
|
| 2025 |
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. |
Adipocyte-specific Cldn5 knockout mice, YBX3 localization assays, IL10 promoter binding assays, IL10R paracrine signaling experiments, metabolic phenotyping (thermogenesis, energy expenditure, glucose tolerance) |
Nature communications |
High |
40610440
|
| 2026 |
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. |
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 |
The Journal of biological chemistry |
High |
41539562
|
| 2025 |
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. |
Multi-microsecond all-atom molecular dynamics simulations, free energy calculations for water and ion permeation across multi-protomer (16-subunit) claudin-5 TJ models |
bioRxivpreprint |
Low |
|