Disruption of desmosome assembly by monovalent human pemphigus vulgaris monoclonal antibodies

Abstract

The intercellular interactions of the desmosomal cadherins, desmoglein and desmocollin, are required for epidermal cell adhesion. Pemphigus vulgaris (PV) is a potentially fatal autoimmune blistering disease characterized by autoantibodies against desmoglein (Dsg) 3. During calcium-induced desmosome assembly, treatment of primary human keratinocytes with pathogenic monovalent anti-Dsg3 mAbs produced from a PV patient causes a decrease of Dsg3 and desmoplakin but not desmocollin (Dsc) 3 in the Triton-insoluble fraction of cell lysates within 2 hours. Immunofluorescence and antibody ELISA studies suggest that pathogenic mAbs cause internalization of cell-surface Dsg3 but not Dsc3 through early endosomes. Electron microscopy demonstrated a lack of well-formed desmosomes in keratinocytes treated with pathogenic compared to nonpathogenic mAbs. In contrast, pathogenic mAbs caused late depletion of Dsg3 from preformed desmosomes at 24 hours, with effects on multiple desmosomal proteins including Dsc3 and plakoglobin. Together, these studies indicate that pathogenic PV mAbs specifically cause internalization of newly synthesized Dsg3 during desmosome assembly, correlating with their pathogenic activity. Monovalent human PV anti-Dsg mAbs reproduce the effects of polyclonal PV IgG on Dsg3 and will facilitate future studies to further dissect the cellular mechanisms for the loss of cell adhesion in pemphigus.

Figure 1. Localization of desmosomal cadherins during calcium-induced desmosome assembly in primary human epidermal keratinocytes

PHEK seeded on glass coverslips in basal DK-SFM media (0.07 mM calcium) were shifted to DK-SFM media containing 1.2 mM calcium for the indicated amount of time. Cells were processed for single or double-label immunofluorescence using primary antibodies specific for Dsg3 (H-145), desmoplakin, or Dsc3 (U114). Cell nuclei were stained with DAPI. Scale bar, 10 microns.

Figure 2. Pathogenic PV mAb causes loss of Dsg3 and desmoplakin but not Dsc3 from the Triton X-100 insoluble fraction of keratinocyte lysates during calcium-induced desmosome assembly

PHEK were treated with a pathogenic PV mAb (P1), nonpathogenic PV mAb (NP1), negative control mAb (Neg), or PBS for the indicated amount of time during calcium-induced desmosome assembly. Equal amounts of protein from the Triton X-100 soluble and insoluble fractions were analyzed by SDS-PAGE, followed by immunoblotting for (A) Dsg3 (5G11) and plakoglobin, or (B) Dsc3 (U114) and desmoplakin, with keratin and beta-tubulin as loading controls.

Figure 2. Pathogenic PV mAb causes loss of Dsg3 and desmoplakin but not Dsc3 from the Triton X-100 insoluble fraction of keratinocyte lysates during calcium-induced desmosome assembly

PHEK were treated with a pathogenic PV mAb (P1), nonpathogenic PV mAb (NP1), negative control mAb (Neg), or PBS for the indicated amount of time during calcium-induced desmosome assembly. Equal amounts of protein from the Triton X-100 soluble and insoluble fractions were analyzed by SDS-PAGE, followed by immunoblotting for (A) Dsg3 (5G11) and plakoglobin, or (B) Dsc3 (U114) and desmoplakin, with keratin and beta-tubulin as loading controls.

Figure 3. Pathogenic PV mAb causes delayed depletion of both Dsg3 and Dsc3 from keratinocytes demonstrating preformed desmosomes

PHEK were first incubated with 0.4 mM calcium media overnight to stimulate desmosome assembly, followed by treatment with a pathogenic (P1) or nonpathogenic (NP1) PV mAb for the indicated amount of time. Equal amounts of protein from the Triton X-100 insoluble fractions were analyzed by SDS-PAGE, followed by immunoblotting for Dsg3 (5G11), Dsc3 (U114), or plakoglobin, with keratin as a loading control.

Figure 4. Pathogenic PV mAb causes loss of cell surface Dsg3 and desmoplakin but not Dsc3 during desmosome assembly

PHEK were treated with pathogenic (P1) or nonpathogenic (NP1) PV mAb during calcium-induced desmosome assembly. At the specified time points, cells were processed for single or double-label immunofluorescence using primary antibodies against (A) Dsg3 (5G11), desmoplakin, or (B) both Dsg3 (5G11) and Dsc3 (H-50) after 4 hours of mAb treatment. Scale bar, 10 microns.

Figure 4. Pathogenic PV mAb causes loss of cell surface Dsg3 and desmoplakin but not Dsc3 during desmosome assembly

PHEK were treated with pathogenic (P1) or nonpathogenic (NP1) PV mAb during calcium-induced desmosome assembly. At the specified time points, cells were processed for single or double-label immunofluorescence using primary antibodies against (A) Dsg3 (5G11), desmoplakin, or (B) both Dsg3 (5G11) and Dsc3 (H-50) after 4 hours of mAb treatment. Scale bar, 10 microns.

Figure 5. Antibody-induced depletion of Dsg3 correlates with PV mAb pathogenicity

PHEK were treated with pathogenic or nonpathogenic PV antibodies for the indicated amount of time during calcium-induced desmosome assembly. (A) Immunoblot analysis of Dsg3 (5G11) in the Triton X-100 soluble and insoluble fractions of keratinocytes treated with different pathogenic (P1 and P2) and nonpathogenic (NP1 and NP2) PV mAbs. (B) Immunofluorescence analysis of cells treated with pathogenic (P1 and P2) and nonpathogenic (NP1 and NP2) PV mAbs. Cells were stained for Dsg3 (5G11, green), and nuclei were stained with DAPI (blue). Scale bar, 10 microns. (C) The relative concentrations of PV mAbs in PHEK culture medium were monitored for 0–48 hrs and measured by Dsg3 ELISA. Data points represent mean values of four independent experiments.

Figure 5. Antibody-induced depletion of Dsg3 correlates with PV mAb pathogenicity

PHEK were treated with pathogenic or nonpathogenic PV antibodies for the indicated amount of time during calcium-induced desmosome assembly. (A) Immunoblot analysis of Dsg3 (5G11) in the Triton X-100 soluble and insoluble fractions of keratinocytes treated with different pathogenic (P1 and P2) and nonpathogenic (NP1 and NP2) PV mAbs. (B) Immunofluorescence analysis of cells treated with pathogenic (P1 and P2) and nonpathogenic (NP1 and NP2) PV mAbs. Cells were stained for Dsg3 (5G11, green), and nuclei were stained with DAPI (blue). Scale bar, 10 microns. (C) The relative concentrations of PV mAbs in PHEK culture medium were monitored for 0–48 hrs and measured by Dsg3 ELISA. Data points represent mean values of four independent experiments.

Figure 5. Antibody-induced depletion of Dsg3 correlates with PV mAb pathogenicity

PHEK were treated with pathogenic or nonpathogenic PV antibodies for the indicated amount of time during calcium-induced desmosome assembly. (A) Immunoblot analysis of Dsg3 (5G11) in the Triton X-100 soluble and insoluble fractions of keratinocytes treated with different pathogenic (P1 and P2) and nonpathogenic (NP1 and NP2) PV mAbs. (B) Immunofluorescence analysis of cells treated with pathogenic (P1 and P2) and nonpathogenic (NP1 and NP2) PV mAbs. Cells were stained for Dsg3 (5G11, green), and nuclei were stained with DAPI (blue). Scale bar, 10 microns. (C) The relative concentrations of PV mAbs in PHEK culture medium were monitored for 0–48 hrs and measured by Dsg3 ELISA. Data points represent mean values of four independent experiments.

Figure 6. Antibody-bound Dsg3 is internalized from the cell surface into early endosomes

PHEK were treated with P1 or NP1 PV mAb during calcium-induced desmosome assembly. After 1 hour, cells were processed for double-label immunofluorescence using Alexa 594-conjugated primary antibodies against the HA tag of the pathogenic PV mAb (shown in red) and Alexa 488 secondary antibodies against primary antibodies binding the early endosomal marker EEA-1 (shown in green). Red arrows, antibody-bound cell surface Dsg3. Yellow arrows, co-localization of internalized P1 mAb with EEA-1. Green arrows, EEA-1 positive/PV mAb negative early endosomes. Scale bar, 10 microns.

Figure 7. Interference with Dsg3 incorporation into the desmosome prevents calcium-induced desmosome assembly

PHEK were treated with a pathogenic (P1) or nonpathogenic (NP1) PV mAb during calcium-induced desmosome assembly. After 1 hour cells were processed for electron microscopy. (A) Long arrows indicate desmosomes displaying typical electron dense plaques and keratin intermediate filament insertion. Scale bar, 2 microns. (B) Higher magnification view of desmosomes in control NP1-treated keratinocytes. Scale bar, 1 micron. (C) One half-desmosome was observed in control NP1-treated keratinocytes. Scale bar, 0.5 microns. (D) In P1-treated cells, multiple points of cell-cell contact without desmosome formation were observed. Scale bar, 2 microns. (E) In rare P1-treated cells, desmosome-like structures were observed lacking well-defined electron dense plaques (short arrows). Scale bar, 2 microns. (F) Higher magnification view of desmosome-like structures in P1-treated cells lacking electron dense plaques. Scale bar, 0.5 microns.