Model of mechanism of pemphigus acantholysis. Pathogenic PF IgG may initiate signaling by acting as competitive inhibitors for trans and/or cis-interactions within the EC1/2 region of the desmosome cadherin dsg1 at the desmosome adhesive interface. The signal is likely propagated by modulation of a docking plexus on the cytoplasmic surface of the membrane and may include the cytoplasmic tail of dgs1 as well as associated proteins such as plakoglobin, desmoplakin, and plakophilin. Activation of p38MAPK and HSP25/27 contributes to reorganization of keratin intermediate filaments and the actin cytoskeleton with subsequent loss of keratinocyte cell-cell adhesion and clinical blister formation in the superficial layers of the epidermis. DSC, desmocollin; DSG, desmoglein; DP, desmoplakin; KIF, keratin intermediate filament; PG, plakoglobin; PKP, plakophilin.

Inhibiting p38MAPK prevents clinical blistering in PF passive transfer mice. Neonatal C57BL/6J mice were injected intradermally with either PF IgG (0.1 mg of IgG/g body weight) (A) or PF IgG (0.1 mg of IgG/g body weight) plus SB2023580 (B). After 18 hours, the skin of neonatal mice from the test and control groups was examined clinically. PF IgG-treated mice have a positive Nikolsky’s sign (white arrows), demonstrating loss of epithelial cell-cell adhesion. In contrast, mice treated with the SB203580 and PF IgG have a negative Nikolsky’s sign, indicating that epithelial adhesion remains intact. Inhibiting p38MAPK prevents histological blistering in PF passive transfer mice. Representative skin biopsies of mice treated with control IgG (0.015 mg of IgG/g body weight) (C), PF IgG (0.015 mg of IgG/g body weight) (D), SB203580 and then PF IgG (E), and SB202474 and then PF IgG (F), were fixed in formalin and stained with H&E. Acantholysis leading to blister formation is seen in PF IgG-treated mice, but is blocked in mice treated with the p38MAPK inhibitor SB203580, but not the inactive analog SB202474.

The inhibitors do not prevent systemic absorption of the injected IgG: anti-dsg1 IgG serum levels. Serum samples obtained from control IgG, PF IgG-treated, PF IgG plus SB203580-treated, and PF IgG plus SB202470-treated mice were examined for the presence of anti-dsg1 autoantibodies by using a dsg1 ectodomain-based ELISA. (A) No significant differences in circulating anti-dsg1 IgG were found in mice pretreated with the active p38MAPK inhibitor or the inactive analog when compared to mice treated with PF IgG alone. P values are as follows: i) PF IgG compared to control, P = 0.004; ii) PF IgG + SB203580 compared to PF IgG, P = 0.27; iii) PF IgG + SB202474 compared to PF IgG, P = 0.4. SD is shown by error bars (n = 3 mice per treatment group); P values were calculated by using the Student’s t-test. The inhibitors do not prevent the diffusion of IgG to the epidermis. Perilesional skin biopsies from control IgG (B), PF IgG-treated (C), PF IgG plus SB203580-treated (D), and PF IgG plus SB202470-treated (E) mice were examined for the presence of human anti-dsg1 PF IgG by direct immunofluorescence by using a mouse anti-human Cy-2-conjugated monoclonal antibody. A honeycomb pattern of staining in the epidermis is seen in mice treated with PF IgG, PF IgG plus SB203580, and PF IgG plus SB202474, demonstrating that the inhibitors do notprevent binding of PF autoantibodies to the keratinocyte cell surface.

Increased phosphorylation of p38MAPK in the skin of PF IgG-treated mice. Neonatal C57BL/6 wild-type mice were injected intradermally with control IgG (CON, 0.015 mg of IgG/g body weight), PF IgG (0.015 mg of IgG/g body weight), SB203580 and then PF IgG (PF IgG + SB203580), or the inactive analog SB202474 and then PF IgG (PF IgG + SB202474). Skin biopsies were obtained after 18 hours of treatment and extracted in IEF lysis buffer. A: Samples were equally loaded, separated by 10% SDS-PAGE, transferred to polyvinylidene difluoride, and immunoblotted with antibodies to p38MAPK, phospho-p38MAPK (phospho-p38), or total p38MAPK (p38). Blots were developed by ECL reaction (Amersham Pharmacia). Band intensities of phospho-p38MAPK were normalized to the band intensities of total p38MAPK. Extracts prepared from the blister roof and floor of skin biopsies of PF IgG were similarly analyzed. A: Representative Western blot. B: Signal intensity from the ECL reaction for each band was quantified with a GeneGnome scanner (Syngene Bio Imaging) by using GeneSnap software (n = 3 mice per treatment condition; SD shown by error bars). P values are as follows: i) PF IgG compared to control, P = 0.002; ii) PF IgG + SB203580 compared to PF IgG, P = 0.004; iii) PF IgG + SB202474 compared to PF IgG, P = 0.181; iv) blister roof compared to blister floor, P = 0.03. P values were calculated by using the Student’s t-test. Increased phosphorylation of HSP25 in the skin of PF IgG-treated mice. Increased amounts of the most negatively charged HSP25 isoform (P2) were observed in PF IgG-treated mice and blocked in mice pretreated with SB203580, but not with the inactive analog SB202474. Neonatal C57BL/6 wild-type mice were injected intradermally with control IgG (CON, 0.015 mg of IgG/g body weight), PF IgG (0.015 mg of IgG/g body weight), SB203580 and then PF IgG (PF IgG + SB203580), or SB202474 and then PF IgG (PF IgG + SB202474). Skin biopsies were obtained after 18 hours of treatment and extracted in IEF lysis buffer. C: Skin extracts (30 μg) were prepared and separated in the first dimension by using 7 cm, pH 4 to 7, IPGphor strips (Amersham Pharmacia Biosciences) and in the second dimension by 10% SDS-PAGE, followed by immunoblotting with antibodies to murine HSP25 as described. Blots were developed by ECL reaction (Amersham Pharmacia). C: Representative Western blots of region of the two-dimensional gels showing HSP25 charge isoforms. D: Signal intensity from the ECL reaction for each spot corresponding to the two-dimensional gel HSP25 charge isoforms labeled P0, P1, and P2 were quantified as above with a GeneGnome scanner and GeneSnap software (n = 3 mice per treatment condition, SD shown by error bars). Values were expressed as a percentage of total HSP25 by using the formula Pn/(P0 + P1 + P2), where Pn corresponds to the signal intensity for n = spot 0, 1, or 2 and P0 + P1 + P2 is the summed signal intensity for all three HSP25 isoforms. An increase in the percentage of the most negatively charged HSP25 isoform, P2, is observed in skin extracts from PF IgG versus control-treated mice (P = 0.01). This increase is blocked in mice pretreated with SB203580 (P = 0.001 compared with PF IgG-treated mice), but not the inactive analog SB202474 (P = 0.4 compared with PF IgG-treated mice). P values were calculated by using the Student’s t-test.