The BP patients fulfilled both inclusion criteria: (i) clinical blistering or erosions on the skin and (ii) circulating autoantibodies against COL17 as detected by BP180-NC16A enzyme-linked immunosorbent assay (ELISA)/CLEIA (MBL, Nagoya, Japan). BP-IgG was purified from plasma obtained by apheresis in a severe BP patient. Total IgG was purified using a protein G affinity column according to the manufacturer’s instructions (GE Healthcare, Amersham, UK). In accordance with the Hokkaido University Hospital bylaws and standard operating procedures approved by the Hokkaido University Hospital Review Board, we obtained patient consent for experimental procedures to be performed at Hokkaido University Hospital. A full review and approval by an ethics committee were not required, according to local guidelines. The studies were conducted in accordance with the Helsinki guidelines.

Anti-COL17 NC16A IgG was purified from total IgG using a protein G affinity column by means of the HiTrap HNS-activated HP column according to the manufacturer’s instructions (GE Healthcare). Briefly, GST fusion COL17 NC16A was produced as previously described (26). The recombinant protein was coupled with the HiTrap NHS-activated HP column. The titer of anti-COL17-specific IgG was measured by indirect immunofluorescent staining. Indirect immunofluorescent staining using anti-COL17-specific IgG (concentration 0.1 mg/ml) demonstrated titers greater than 1:32,000.

Two different IVIgs (Nihon Pharmaceutical Co., Ltd., Tokyo, Japan, and Baxter International Inc., Deerfield, IL, USA; diluted with PBS) were purchased and used for this study. The concentrations of each agent for the treatment are given in Section “Results” and the figure legends.

To deplete anti-idiotypic antibodies against anti-COL17 NC16A IgG, anti-COL17 NC16A IgG was coupled to a HiTrap NHS-activated HP column according to the manufacturer’s instructions (GE Healthcare). IVIg was passed through the column to remove anti-idiotypic antibodies, and then flow-through fractions (the “IVIg-depleted” sample) and the elution fraction (the “idiotype” sample) were used for the depletion assay. To evaluate the depletion efficacy, 96-well microtiter plates (Maxisorp; Nunc, Roskilde, Denmark) were coated with purified anti-COL17 NC16A IgG (500 ng/well), normal human IgG and PBS. Nonspecific binding was reduced by blocking with protein-free blocking buffer (Thermo Fisher Scientific, Rockford, IL, USA) at room temperature (RT) for 1 h. Plates were subsequently incubated with biotin-conjugated IVIg (1 mg/ml). IVIg was conjugated with biotin using a biotin labeling kit according to the manufacturer’s instructions (Dojindo, Kumamoto, Japan). Finally, plates were incubated with HRP-conjugated streptavidin (Thermo Fisher Scientific) for IgG subclasses at RT for 0.5 h.

DJM-1 cells isolated from human skin squamous cell carcinoma (27) were cultured in DMEM. To investigate the depletion of COL17, cells were cultured to approximately 40% confluence (24). DJM-1 cells were pretreated with agents for 1 h, and then BP-IgG (concentration 1 mg/ml) was added to the culture media for 4-h incubation. In some experiments, DJM-1 cells were treated with BP-IgG for 4 h followed by IVIg for 1 h.

For Western blot analysis of whole-cell lysates, cells were lysed in RIPA buffer (Thermo Fisher Scientific) containing a protease inhibitor cocktail (Sigma Aldrich), and the lysates were centrifuged. Each fraction was subjected to SDS–PAGE in 6% polyacrylamide gel. The gels were transferred onto nitrocellulose membranes. Blotting was performed using rabbit anti-COL17 (1:2,000 dilution) (26), rabbit anti-β-tubulin (Abcam, Tokyo, Japan, 1:20,000 dilution), and anti-integrin α6 (Santa Cruz, Dallas, TX, USA, 1:500 dilution) as the primary antibodies, followed by incubation with HRP-conjugated goat anti-rabbit or anti-mouse IgG (Life Technologies, 1:5,000 dilution). Signals were visualized with Clarity Western ECL Substrate (Bio-Rad Laboratories, Hercules, CA, USA).

BP180-NC16A ELISA was performed according to the manufacturer’s instructions with minor modifications (MBL). Briefly, the ELISA plate was incubated with 100 µl of IVIg (5 mg/ml), bovine serum albumin (BSA, 5 mg/ml), or PBS. Subsequently, the plate was incubated with 1 mg/ml BP-IgG with/without PBS washing. The ELISA index value was calculated according to the manufacturer’s instructions. ELISA using sera from BP patients and healthy volunteers was performed without washing the plate after IVIg incubation.

BP-IgG stimulation leads to reductions in COL17 amount and in cell adhesion to the culture plate (24). Cells were placed on a vortex mixer for 20 min after BP-IgG stimulation with/without IVIg pretreatment. The adhesion of DJM-1 cells to the bottom of the culture plate was assayed by determining the number of adherent cells after vibration. After PBS washing, cells that remained on the bottom of the culture plate were treated with 0.25% trypsin for 5 min at 37°C. The released cells were counted using a blood cell counter under a microscope.

Ex vivo autoantibody-induced, neutrophil-dependent dermal–epidermal separation was performed as described (28, 29). Briefly, 5-µm cryosections from normal human skin were incubated with BP-IgG (2 mg/ml) in the presence or absence of 5 mg/ml IVIg at 37°C for 1 h. After washing with PBS, the slides were covered with a second slide that was taped at each end to form a chamber. Subsequently, 10⁷ cells/ml of freshly isolated normal human leukocyte suspension was injected into the chamber and incubated at 37°C for 5 h. Sections were subsequently stained with H&E. We tested the experiments using two different blood donors.

Statistical calculations were performed using SigmaPlot (Version 12.0, Systat Software, Chicago, IL, USA). To compare the parameters in the COL17-depletion assay and COL17-ELISA, one-way ANOVA test was used. A comparison of the COL17-ELISA using patient sera with/without IVIg was performed using the Wilcoxon signed-rank test. A p-value of <0.05 was considered statistically significant. The graphs present the median ± SD.

When DJM-1 cells are treated with BP-IgG, the amount of COL17 is decreased as determined by Western blotting (COL17-depletion assay) (24). In this study, we examined the effects of IVIg by means of a COL17-depletion assay. For a COL17-depletion assay, 40% confluent cells were incubated with BP-IgG (concentration: 1 mg/ml) for 4 h. The amount of COL17 relative to β-tubulin was determined by Western blotting. DJM-1 cells were treated with BP-IgG before (the pretreatment sample) or after (the posttreatment sample) the addition of serially diluted IVIg. BP-IgG without IVIg induced an approximately 50–60% reduction of COL17 in DJM-1 cells. By contrast, BP-IgG pretreated with 5 mg/ml IVIg restored the amount of COL17 by 70–90% (p < 0.05, Figure 1A). IVIg posttreatment after 4 h of BP-IgG stimulation did not influence COL17 depletion (Figure 1B). Given that IVIg has direct effects on keratinocytes, such as on caspase expression (30), we incubated BP-IgG with IVIg in advance and then added them to the culture medium together (simultaneous sample). The COL17 depletion was restored similar to that in Figure 1A (Figure 1C). The amount of integrin α6, which is a transmembrane protein that is a hemidesmosomal component, was unchanged (Figure 1D). The same concentration (5 mg/ml) of BSA and normal human IgG did not influence the amount of COL17 under any conditions (Figure 1D).

The highest concentration of IVIg (5 mg/ml) was designed based on the estimated serum concentration with standard daily IVIg treatment (0.4 g/kg/day) (3).

Furthermore, cell adhesion was found to be reduced by approximately 60% under BP-IgG stimulation (Figure 2). Pretreatment with IVIg restored cell adhesion. Cells remaining in the culture plate were significantly increased with 5 mg/ml IVIg pretreatment (p < 0.05, Figure 2). Pretreatment with 5 mg/ml IVIg restored adhesion by 90% compared with normal IgG stimulation.

We subsequently investigated the mechanism of action of IVIg in this COL17-depletion assay. The titer of BP-IgG used for the COL17-depletion assay was measured with IVIg (5 mg/ml), BSA (5 mg/ml) or PBS by ELISA. ELISA plates were preincubated with either IVIg or BSA, and BP-IgG was subsequently added. To rule out the possibility of IVIg masking the antigen, the ELISA plate was washed with PBS after IVIg pretreatment, and then BP-IgG was added. Interestingly, 5 mg/ml IVIg (without washing) significantly inhibited the binding of BP-IgG to the ELISA plate (p < 0.05, Figure 3A). IVIg pretreatment (without washing) reduced the average titer by 30% compared with PBS pretreatment (Figure 3A, black bar). This inhibition was not observed using IVIg (with washing) or the same concentration of BSA. These results suggest that IVIg impairs BP-IgG’s ability to bind to COL17, in which IVIg may function as an anti-idiotypic antibody. To confirm our results, samples from 10 BP patients were evaluated in the same manner. The serum titers for all 10 patients were found to be reduced by approximately 40% in the presence of IVIg (without washing) (p < 0.05, Figure 3B).

Next, to confirm that anti-idiotypic antibodies were actually present in the IVIg, we performed experiments to identify them directly. Anti-idiotypic antibodies in the IVIg were depleted by anti-COL17 NC16A IgG (IVIg, IVIg depleted, and idiotype fractions, Figure 4A). IVIg contained approximately 0.02% of anti-idiotypic antibodies against COL17 IgG (Figure 4B). The depletion efficiency was more than 90% (Figure 4C). The titers of BP-IgG were measured in the presence of IVIg (IVIg 5 mg/ml and idiotype 2.6 µg/ml) and normal IgG (5 mg/ml). The titers were found to be decreased in the setting of BP-IgG with IVIg and idiotype, but not BP-IgG with IVIg depleted (Figure 4D). Next, a COL17-depletion assay was performed using IVIg (5 mg/ml) and anti-idiotypic antibodies (2.6 µg/ml). IVIg depleted did not show blocking effects on COL17-depletion (Figure 4E). Furthermore, adhesion was also tested using IVIg and anti-idiotypic antibodies. IVIg depleted did not restore the adhesive strength compared with IVIg and idiotype (Figure 4F).

According to our results, the therapeutic effects of IVIg potentially involve anti-idiotypic antibodies. We next examined the blocking effects of IVIg using an ex vivo assay, demonstrating the dermal–epidermal separation induced by BP-IgG in the presence of leukocytes (28, 29). In case of BP-IgG with PBS, dermal–epidermal separation was observed 5 h after incubation with normal human leukocytes (Figure 5, upper left). Next, we added IVIg, BSA and normal human IgG and incubated them with BP-IgG on the cryosections for 1 h, followed by incubation with normal human leukocytes for 5 h. IVIg depleted (5 mg/ml, upper right) and normal human IgG (5 mg/ml, lower middle) did not block the dermal–epidermal separations. By contrast, IVIg (5 mg/ml, upper middle) and the idiotype (2.6 µg/ml, lower left) protected the dermal–epidermal separation.

There is great concern about differences arising from differences in IVIg lots or in companies. Therefore, we tested an IVIg from the different company. An IVIg product from this different company was found to prevent the COL17 depletion of keratinocytes induced by BP-IgG (Figure 6A). Furthermore, we found the same blocking effects of IVIg using an ex vivo assay (Figure 6B).