DSG3

DSG3 encodes desmoglein-3, a desmosomal cadherin essential for Ca²⁺-dependent cell–cell adhesion in stratified mucosal epithelia, where it anchors desmosomes to the cytoskeleton and transduces signals that regulate adhesion strength, cell proliferation, and migration. The extracellular domain mediates homophilic adhesion and is the target of pathogenic pemphigus vulgaris autoantibodies (PMID:8040292), while the intracytoplasmic cadherin-like subdomain directly binds plakoglobin to link desmosomes to intermediate filaments (PMID:7738346); an extradesmosomal pool associates with E-cadherin, β-catenin, and Src to regulate desmosome maturation and downstream Src/EGFR/ERK, p38 MAPK, and Akt/GSK3β/β-catenin signaling (PMID:26115704, PMID:31178865, PMID:37143675). Homozygous loss-of-function deletion of DSG3 causes acantholytic blistering of oral and laryngeal mucosa, confirming its non-redundant requirement for mucosal epithelial integrity in humans (PMID:37850634).

1. 1994 High

   Identifying the molecular target of pemphigus vulgaris pathogenesis: the properly folded Dsg3 extracellular domain was shown to carry all conformational epitopes required for pathogenic autoantibody binding and blister induction, establishing Dsg3 as the critical disease autoantigen.

   Evidence Recombinant Dsg3 ectodomain (baculovirus/COS7) immunoadsorbed pathogenic IgG and prevented blistering in neonatal mouse passive-transfer model

   PMID:8040292

   Open questions at the time

   • Epitope fine-mapping not resolved
   • Whether Dsg3 antibodies alone are sufficient for mucosal PV or require anti-Dsg1 synergy not addressed
2. 1995 High

   Defining how Dsg3 connects to the desmosomal plaque: the C-terminal 87 amino acids of the intracytoplasmic cadherin-like subdomain directly bind plakoglobin, while the desmoglein-specific repeat domains are dispensable, establishing the minimal plakoglobin-binding interface.

   Evidence Chimeric E-cadherin/Dsg3 constructs tested by co-IP in HaCaT cells and cell-free in vitro transcription/translation binding

   PMID:7738346

   Open questions at the time

   • Structural basis of the Dsg3–plakoglobin interface not resolved
   • Whether other desmosomal plaque proteins bind the same subdomain not tested
3. 1996 High

   Establishing that ER transit is required for Dsg3 conformational maturation: proper folding of Dsg3 requires signal-peptide-directed ER translocation, whereas prosequence cleavage and glycosylation are dispensable, clarifying quality-control steps for functional Dsg3 biosynthesis.

   Evidence Signal-peptide-deleted and protease-site-mutant baculovirus constructs assayed for immunoadsorption of PV sera

   PMID:8823357

   Open questions at the time

   • ER chaperones involved not identified
   • Whether ER maturation defects contribute to disease in vivo not explored
4. 1999 High

   Revealing the mechanism of autoantibody-induced adhesion loss: PV-IgG rapidly depletes Dsg3 from the detergent-soluble membrane pool, generating Dsg3-depleted desmosomes that initially retain other desmosomal components, with subsequent loss of the cytoskeletal Dsg3 pool, establishing a two-step depletion model.

   Evidence Sequential detergent fractionation and immunofluorescence in DJM-1 squamous carcinoma cells treated with PV-IgG

   PMID:9886266

   Open questions at the time

   • Endocytic route of internalized Dsg3 not characterized
   • Signaling events driving depletion not defined
5. 2007 High

   Demonstrating that polyclonal antibody diversity drives PV pathogenesis: individual monoclonal anti-Dsg3 antibodies each deplete Dsg3 to characteristic limits and act synergistically in combination, explaining why polyclonal PV-IgG is more pathogenic than individual specificities.

   Evidence Four monoclonal antibodies tested singly and in combination on DJM-1 cells and normal keratinocytes with dispase adhesion assay and fractionation

   PMID:17428808

   Open questions at the time

   • Which epitope combinations are most pathogenic in vivo not determined
   • Role of antibody valence/isotype not dissected
6. 2010 High

   Ordering the cellular events of Dsg3 loss into three phases: rapid internalization of non-junctional Dsg3, retrograde transport from cell contacts into cytoplasmic vesicles, and final depletion of the cytoskeletal pool; exogenous Dsg3 expression rescues adhesion by restoring biosynthetic supply.

   Evidence Live and fixed imaging of primary keratinocytes with PV-IgG, biochemical fractionation, and Dsg3 overexpression rescue

   PMID:21160493

   Open questions at the time

   • Molecular machinery mediating retrograde transport not identified
   • Whether endocytic route involves clathrin or caveolae not resolved
7. 2013 High

   Uncovering a signaling function for Dsg3 beyond adhesion: Dsg3 sequesters plakoglobin at the membrane, preventing its nuclear translocation and TCF/LEF-dependent transcription of c-Myc, cyclin D1, and MMP-7, thereby linking Dsg3 levels to cell cycle control and invasive behavior in cancer.

   Evidence RNAi knockdown of DSG3 in head-and-neck cancer cells with co-IP, TCF/LEF reporter, cell cycle analysis, and in vivo xenograft

   PMID:23737966

   Open questions at the time

   • Whether this signaling axis operates in normal keratinocytes not tested
   • Direct versus indirect control of plakoglobin nuclear entry not resolved
8. 2015 High

   Identifying an extradesmosomal Dsg3 signaling complex: Dsg3 forms a complex with E-cadherin, β-catenin, and Src outside desmosomes; Src phosphorylates Dsg3 and E-cadherin and is required for Dsg3 recruitment to the cytoskeletal pool and for desmosome hyper-adhesion, establishing Src as a key regulatory kinase of Dsg3 function.

   Evidence Reciprocal co-IP, Src inhibition, E-cadherin silencing, detergent fractionation, and tyrosine phosphorylation assays in primary keratinocytes

   PMID:26115704

   Open questions at the time

   • Direct Src phosphorylation sites on Dsg3 not mapped
   • Structural basis of the E-cadherin/Dsg3 interaction not determined
9. 2019 High

   Establishing Dsg3 isoform specificity in PV signaling: CRISPR knockout of Dsg3, but not Dsg2, protects keratinocytes from PV-IgG-induced loss of adhesion and prevents downstream Src-dependent EGFR and ERK activation, positioning Dsg3 as the obligate initiating sensor of pathogenic antibody signaling.

   Evidence CRISPR/Cas9 Dsg3 and Dsg2 KO in HaCaT cells with dispase assay, Ca²⁺ influx, and phospho-Western blots

   PMID:31178865

   Open questions at the time

   • Whether compensatory upregulation of other desmogleins occurs in KO cells not fully addressed
   • Mechanism by which Dsg3 activates Src upon antibody binding not resolved
10. 2021 Medium

    Connecting mechanical forces to Dsg3-mediated signaling: external mechanical stress counteracts anti-Dsg3 antibody-induced monolayer fragmentation by activating RhoA, strengthening cortical actin, and suppressing p38 MAPK phosphorylation, revealing mechano-signaling crosstalk downstream of Dsg3.

    Evidence Mechanical stress applied to keratinocyte monolayers with fragmentation assay, p38 MAPK and RhoA activity measurements

    PMID:33724731

    Open questions at the time

    • Mechanosensor coupling Dsg3 to RhoA not identified
    • Genetic validation of p38 MAPK requirement not performed
11. 2023 High

    Resolving epitope-specific signaling hierarchies: EC1-targeting antibody AK23 activates p38MAPK, Akt, and Src and depletes Dsg3, whereas EC5-targeting 2G4 activates p38MAPK and Akt but not Src and does not deplete Dsg3; pharmacological epistasis places p38MAPK upstream of Src and Akt, establishing that antibody epitope position dictates distinct intracellular signaling cascades.

    Evidence Multiple monoclonal antibodies with dispase assay, STED microscopy, GTPase GLISA, pharmacological inhibition of p38 and Src

    PMID:37143675

    Open questions at the time

    • How epitope position transmits distinct conformational signals across the membrane not determined
    • Whether these hierarchies apply to polyclonal patient sera in vivo not tested
12. 2023 High

    Demonstrating two mechanistically distinct Dsg3 adhesion pools: a cell-surface pool whose binding depends on actin filaments and a cell-contact pool whose adhesion is actin-independent but regulated by PKCα-controlled intermediate filament anchorage, resolving how different cytoskeletal systems differentially support Dsg3 function.

    Evidence Hybrid STED/single-molecule force spectroscopy AFM on living keratinocytes with Latrunculin B and PMA treatments

    PMID:36602635

    Open questions at the time

    • Adaptor proteins linking each pool to its respective cytoskeleton not identified
    • Whether PV-IgG preferentially targets one pool not tested
13. 2023 Medium

    Establishing in vivo genetic requirement: homozygous loss-of-function deletion of DSG3 causes acantholytic blistering of oral and laryngeal mucosa (ABOLM) in humans, proving Dsg3 is non-redundantly required for mucosal epithelial integrity.

    Evidence Chromosomal microarray in dizygotic twins with ABOLM phenotype; autoimmune mechanism excluded

    PMID:37850634

    Open questions at the time

    • No functional rescue or cell biology performed
    • Genotype-phenotype correlation across different DSG3 mutations not established
14. 2025 Medium

    Extending Dsg3 signaling to Wnt/β-catenin in cancer: STAT3 transcriptionally activates DSG3, which then promotes AKT phosphorylation, inhibits GSK3β, and drives β-catenin nuclear translocation to upregulate SOX2 and MMP7, enhancing stemness and metastasis in bladder cancer.

    Evidence DSG3 knockdown in bladder cancer cells with Western blot pathway analysis, in vivo xenograft, and lung metastasis model

    PMID:40605005

    Open questions at the time

    • Independent replication in other cancer types lacking
    • Direct mechanism by which Dsg3 activates AKT not defined
    • Whether this pathway operates in normal urothelium not addressed

• Key unresolved questions include: the structural basis of how antibody binding to distinct Dsg3 extracellular domains transmits conformational changes across the membrane to activate specific intracellular kinase cascades; the identity of the endocytic machinery and trafficking adaptors mediating Dsg3 internalization and retrograde transport; and whether the extradesmosomal Dsg3/E-cadherin/Src complex is a regulated signaling platform or a transient biosynthetic intermediate.
• No structural model of full-length Dsg3 exists
• Endocytic route (clathrin vs. caveolae) unresolved
• Stoichiometry and regulation of the extradesmosomal complex unknown