DSG3

DSG3 encodes desmoglein 3, a desmosomal cadherin that mediates calcium-dependent keratinocyte cell-cell adhesion and serves as the principal autoantigen in pemphigus vulgaris (PMID:8040292, PMID:9886266). Its intracytoplasmic cadherin-like segment binds plakoglobin directly through its carboxy-terminal 87 residues, independent of other cellular factors (PMID:7738346), and proper folding of its extracellular conformational epitopes requires transit through the endoplasmic reticulum but not glycosylation or proteolytic processing (PMID:8823357). Dsg3 exists in two functionally distinct cytoskeletal pools resolved by single-molecule force spectroscopy: an extradesmosomal cell-surface pool whose adhesion is actin-dependent, and a desmosomal cell-contact pool that is actin-independent and regulated by PKCalpha (PMID:36602635). The extradesmosomal pool assembles into a complex with E-cadherin, beta-catenin, and Src, where Src-dependent tyrosine phosphorylation of Dsg3 and E-cadherin recruits Dsg3 to the cytoskeletal pool and drives desmosome maturation to a calcium-insensitive state (PMID:26115704). Pathogenic PV autoantibodies binding cell-surface Dsg3 trigger sequential desmosome disassembly—rapid internalization of a non-junctional surface pool followed by depletion of detergent-insoluble desmosomal Dsg3 and loss of adhesion—a process reversible by exogenous Dsg3-driven reassembly (PMID:9886266, PMID:21160493); these antibodies act through epitope-specific cascades involving p38MAPK, Src, Akt, EGFR, and Ca2+ influx (PMID:31178865, PMID:37143675). In cancer, DSG3 retains plakoglobin at junctions to sustain TCF/LEF-driven transcription of c-Myc, cyclin D1, and MMP-7 (PMID:23737966), and in bladder cancer acts downstream of STAT3 to activate AKT/GSK3beta/beta-catenin signaling and upregulate SOX2 and MMP7 (PMID:40605005).

1. 1994 High

   Established that the pathogenic PV autoantibody response targets conformational epitopes on the Dsg3 extracellular domain, defining Dsg3 as the relevant antigen and linking antibody binding to blister formation.

   Evidence Baculovirus-expressed PVIg chimera immunoadsorption of patient sera with neonatal mouse passive transfer

   PMID:8040292

   Open questions at the time

   • Did not map the specific epitope residues
   • Did not address downstream signaling triggered by antibody binding
2. 1995 High

   Resolved how Dsg3 couples to the desmosomal plaque by mapping a direct plakoglobin-binding site to the carboxy-terminal 87 residues of the intracytoplasmic cadherin-like segment.

   Evidence Chimeric truncation constructs in HaCaT cells with Co-IP and cell-free in vitro transcription/translation

   PMID:7738346

   Open questions at the time

   • Did not establish the stoichiometry or affinity of the interaction
   • Did not address how this binding regulates desmosome assembly
3. 1996 High

   Defined the biogenesis requirement for Dsg3 antigenicity, showing ER transit but not glycosylation or proteolytic processing produces the conformational epitopes.

   Evidence Signal-peptide and cleavage-site mutagenesis of baculovirus-expressed PVIg with immunoadsorption assay

   PMID:8823357

   Open questions at the time

   • Did not identify the chaperones mediating folding
   • Single expression system
4. 1999 High

   Showed that PV-IgG selectively depletes Dsg3 in two kinetic phases, first from the soluble membrane pool then from desmosomes, while sparing other desmosomal components, distinguishing Dsg3 loss as the primary lesion.

   Evidence Triton fractionation, immunoblotting, and immunofluorescence time-course in DJM-1 cells

   PMID:9886266

   Open questions at the time

   • Did not define the signaling driving internalization
   • Mechanism of depletion (degradation vs internalization) not resolved
5. 2007 High

   Linked antibody pathogenicity to Dsg3-depleting capacity, showing individual monoclonals have characteristic depletion limits that correlate with in vivo pathogenicity and combine cumulatively.

   Evidence Pathogenic/non-pathogenic monoclonal antibody treatment of keratinocytes with fractionation, imaging, and mouse passive transfer

   PMID:17428808

   Open questions at the time

   • Did not map molecular signaling distinguishing pathogenic from non-pathogenic antibodies
6. 2010 High

   Defined the three sequential phases of PV-IgG-induced desmosome disassembly and demonstrated reversibility, establishing internalization and retrograde transport as the mechanistic route to adhesion loss.

   Evidence Live and fixed imaging, fractionation, cell dissociation assay, and exogenous Dsg3 rescue in primary keratinocytes

   PMID:21160493

   Open questions at the time

   • Did not identify the trafficking machinery driving internalization
   • Did not connect phases to specific kinase cascades
7. 2013 High

   Revealed a transcriptional function of DSG3 in cancer by showing it retains plakoglobin at junctions to sustain TCF/LEF activity and proliferation, with knockdown driving plakoglobin nuclear translocation and growth arrest.

   Evidence RNAi in head and neck cancer lines with Co-IP, TCF/LEF reporter, cell-cycle flow cytometry, and xenografts

   PMID:23737966

   Open questions at the time

   • Did not establish whether desmosomal or extradesmosomal Dsg3 mediates this
   • Direct vs indirect effect on plakoglobin localization not dissected
8. 2015 High

   Identified an extradesmosomal Dsg3/E-cadherin/beta-catenin/Src complex in which Src-driven tyrosine phosphorylation controls Dsg3 cytoskeletal recruitment and desmosome maturation to a Ca2+-insensitive state.

   Evidence Reciprocal Co-IP, E-cadherin overexpression/silencing, Src inhibition, phospho-tyrosine blotting, and cohesion assays in keratinocytes

   PMID:26115704

   Open questions at the time

   • Did not establish phosphorylation site stoichiometry
   • Mechanism of Src activation not resolved here
9. 2015 Low

   Proposed a mechanism for Src activation in which Dsg3 displaces inactive Src from the caveolin-1 scaffolding domain, relieving Cav-1-mediated inhibition.

   Evidence Co-IP of Dsg3/Src/Cav-1, overexpression/knockdown, and colocalization analysis (Data in Brief)

   PMID:26858977

   Open questions at the time

   • Single Co-IP and immunofluorescence without functional validation
   • Model proposed but Src kinase activation not directly measured
   • Data article rather than primary research
10. 2019 High

    Used genetic epistasis to confirm Dsg3 specificity and order the antibody-triggered cascade, showing Dsg3 (not Dsg2) is required for adhesion loss and that PV-IgG activates EGFR/ERK via Src while Ca2+ influx is EGFR-independent.

    Evidence CRISPR/Cas9 Dsg3 and Dsg2 knockout HaCaT cells with cohesion assays, phospho-blots, Ca2+ measurements, and inhibitors

    PMID:31178865

    Open questions at the time

    • Did not place all pathway branches in a single hierarchy
    • Relationship between EGFR signaling and Dsg3 depletion not fully resolved
11. 2021 Medium

    Connected mechanotransduction to antibody signaling, showing mechanical stress suppresses anti-Dsg3-induced p38 MAPK activation by enhancing RhoA and cortical actin.

    Evidence Anti-Dsg3 antibody plus mechanical stress in keratinocyte monolayers with p38 phospho-assay, RhoA activity, and actin imaging

    PMID:33724731

    Open questions at the time

    • Single lab
    • Mechanosensor upstream of RhoA not identified
12. 2022 Medium

    Mapped the spatial distribution of Dsg3 across epidermal layers and showed extradesmosomal Dsg3-plakoglobin complexes form desmoplakin-independently in the basal layer, supporting their role in desmosome assembly.

    Evidence STED super-resolution colocalization of Dsg1/Dsg3/desmoplakin/plakoglobin in normal and pemphigus skin

    PMID:35711465

    Open questions at the time

    • Did not functionally test the assembly role
    • Single lab
13. 2023 High

    Demonstrated epitope-specific antibody signaling, showing EC1- versus EC5-targeting monoclonals diverge in Dsg3 depletion and Src activation while converging on p38MAPK-dependent Akt and Src activation.

    Evidence Two pathogenic monoclonals with dispase dissociation, phospho-blots, STED imaging, Ca2+ flux, and Rho/Rac ELISA in keratinocytes

    PMID:37143675

    Open questions at the time

    • Did not extend to additional epitopes
    • Single lab
14. 2023 High

    Distinguished two biophysically distinct Dsg3 adhesion pools, an actin-dependent extradesmosomal pool and an actin-independent, PKCalpha/intermediate-filament-regulated desmosomal pool, on living cells.

    Evidence Hybrid STED/SMFS-AFM single-molecule force spectroscopy with Latrunculin B and PMA perturbation

    PMID:36602635

    Open questions at the time

    • Did not connect the two pools to specific binding partners at the molecular level
    • Single lab
15. 2025 Medium

    Defined a STAT3-DSG3-AKT/GSK3beta/beta-catenin oncogenic axis in bladder cancer driving EMT, stemness, and metastasis via SOX2 and MMP7 upregulation.

    Evidence DSG3 knockdown/overexpression in bladder cancer lines with Co-IP, pathway blots, STAT3 transcription assays, and xenograft/metastasis models

    PMID:40605005

    Open questions at the time

    • No independent replication
    • Single lab
    • Direct vs indirect activation of AKT by DSG3 not resolved
16. 2024 Medium

    Implicated ER stress in antibody-induced adhesion loss, showing PV IgG activates IRE1alpha and PERK and that ER tubules contact internalizing Dsg3 puncta, with ER stress inhibition protecting adhesion.

    Evidence Time-lapse live imaging, ER stress biochemical markers, pharmacological inhibition, and cohesion assay (preprint)

    Open questions at the time

    • Preprint not yet peer-reviewed
    • Mechanism linking ER stress to Dsg3 internalization not resolved
    • Single lab

• How the distinct cytoskeletal pools, kinase cascades, ER-stress responses, and transcriptional functions of Dsg3 are integrated into a single regulatory hierarchy across normal epidermis, pemphigus, and cancer remains unresolved.
• No unified ordering of p38MAPK, Src, EGFR, Akt, Ca2+, and ER stress branches
• Structural basis of Dsg3 trans-adhesion not determined in the corpus
• Whether cancer and pemphigus signaling share common Dsg3-proximal events untested