Pemphigus vulgaris is an autoimmune blistering disease caused by IgG targeting desmoglein 3 (Dsg3), an adhesion molecule of keratinocytes. Anti-Dsg3 IgG production is prevented in healthy individuals, but it is unclear how Dsg3-specific B cells are regulated. To clarify the immunological condition regulating Dsg3-specific B cells, a pathogenic anti-Dsg3 Ig (AK23) knock-in mouse was generated. AK23 knock-in B cells developed normally without undergoing deletion or acquiring an anergic phenotype in vivo. The knock-in B cells showed Ca2+ influx upon IgM cross-linking and differentiated into AK23-IgG+ B cells after LPS and IL-4 stimulation in vitro that induced a pemphigus phenotype after adoptive transfer into Rag2−/− mice. However, the knock-in mouse itself produced AK23-IgM but little IgG without blisters in vivo. Dsg3 immunization and skin inflammation caused AK23-IgG production and a pemphigus phenotype in vivo. Furthermore, Fcgr2b deficiency or haploinsufficiency spontaneously induced AK23-IgG production and a pemphigus phenotype with poor survival rates in AK23 knock-in mice. To assess Fcgr2b involvement in Ig class-switch efficiency, postswitch transcripts of B cells were quantified and significantly higher in Fcgr2b−/− and Fcgr2b+/− mice than wild-type mice in a gene dose-dependent manner. Finally, RNA sequencing revealed reduced expression of FCGR2B and FcγRIIB-related genes in patient B cells. These results indicated that Dsg3-specific B cells do not spontaneously perform pathogenic class switching in vivo, and pemphigus phenotype induction was prevented under normal conditions. Attenuated FcγRIIB signaling is also one of the drivers for pathogenic class switching and is consistent with immunological features identified from clinical samples. This study unveiled a characteristic immune state silencing autoreactive B cells in mice.

Tissue injury caused by autoreactive B lymphocytes is prevented by immunoregulatory functions of the immune system. Clonal deletion (1), anergy (2), and receptor editing (3, 4) are widely accepted mechanisms to regulate autoreactive B cells. Because such mechanisms are achieved during key processes that are necessary for B cells to develop and function properly, harmful activities of autoreactive B cells can be efficiently constrained. In anergy, for example, activation of autoreactive B cells after Ag recognition by BCR is inhibited (2, 5, 6), whereas Ig autoreactivity can be reversed via Ab gene rearrangement, also referred to as “receptor editing” (3, 4, 7). The most basic function of B cells is efficient Ab production, and its essential steps include somatic hypermutation, class switching of Ig, differentiation into memory B cells and Ab-producing plasma cells, and so forth (810). The immune processes to regulate autoreactive B cells involved in these steps must be efficient to prevent autoimmunity. Indeed, the limited accumulation of IgG+ anti-DNA plasma cells was reported as a crucial process to prevent autoantibody-mediated glomerulonephritis (11). Nevertheless, regulatory actions to some of the steps necessary for Ab production have not been elucidated.

Pemphigus vulgaris (PV) is a life-threatening autoimmune blistering disease induced by IgG against desmoglein 3 (Dsg3), a cadherin-type cell adhesion molecule expressed in the stratified squamous epithelium (12, 13). Because IgG interferes with cell–cell adhesion, keratinocytes become dissociated, resulting in blister formation in the skin and oral mucosa. Anti-Dsg3 IgG is detected in patients with PV but not in healthy individuals (14); however, the mechanisms underlying autoantibody production are not fully understood. A PV mouse model was previously established by immunizing Dsg3−/− mice with recombinant Dsg3 (rDsg3); because Dsg3−/− mice lack tolerance against Dsg3, immunization with rDsg3 results in potent Dsg3-specific immune reactions (15). Upon adoptive transfer of lymphocytes from immunized Dsg3−/− mice to Rag2−/− mice, Dsg3-specific lymphocytes expand and produce anti-Dsg3 IgG, resulting in a PV phenotype. A series of anti-Dsg3 Ab clones were isolated from this model for further characterization and epitope mapping (16).

AK23 is a mouse IgG1 Ab clone that recognizes Dsg3 and causes blisters in vivo (16). To investigate how Dsg3-specific B cells are regulated in vivo, AK23 IgM transgenic (AK23IgMtg) mice were generated (17). These mice have AK23 H and L chains in B cells, which produce IgM with AK23 Ag specificity. Although AK23 IgM-expressing B cells and circulating AK23 IgM were detected in transgenic mice (17), no profound IgM deposition on the palate was observed, and the mice did not develop a PV phenotype. Electron microscopy revealed AK23 IgM deposition primarily at the periphery of desmosomes, whereas AK23 IgG accumulated at the center of desmosomes (17). These results implied that AK23 IgM was nonpathogenic in vivo because it could not access Dsg3 at the desmosome core, likely due to the larger size of IgM compared with IgG. Because IgM transgenic mice are incapable of Ig class switching, BCR knock-in mice that allow Ig class switching are required to accurately investigate how pathogenic autoreactive B cells are regulated.

In this study, we generated a Dsg3-specific AK23 BCR knock-in mouse to explore the immunological state that regulates pathogenic autoreactive B cells in a PV mouse model. The data of this study proposed a characteristic immune state that prevents Ig class switching in autoreactive B cells. Pathogenic class switching is accelerated by Dsg3 immunization and skin inflammation. FcγRIIB is also crucially associated with this state. This was also supported by the finding from clinical samples of downregulation of FcγRIIB signaling in B cells of patients with pemphigus.

AK23IgMtg mice were generated by crossing AK23 IgM H chain transgenic mice with AK23 L chain transgenic mice, both of which have been described previously (17) and were on a C57BL/6 genetic background. C57BL/6-Rag-2−/− mice (stock no. RAGN12) were purchased from Taconic (Germantown, NY). C57BL/6-Fcgr2b−/− mice were purchased from Oriental Bio Service (Kyoto, Japan). C57BL/6 wild-type (WT) mice were purchased from Sankyo Labo Service Corporation (Tokyo, Japan). B6.SJL (CD45.1) WT mice (stock no. 002014) were obtained from The Jackson Laboratory. All mice were maintained under specific pathogen-free conditions at our animal facility. All animal experiments were approved by the animal care and use committee of Keio University and RIKEN and were performed in accordance with institutional guidelines.

The targeting vector used to create AK23 IgH knock-in mice was constructed by modifying the vector used to generate anti–myelin oligodendrocyte glycoprotein (anti-MOG) IgH knock-in mice (18); the vector was kindly provided by Antonio Iglesias (F. Hoffmann-La Roche Ltd, Basel, Switzerland). We replaced the anti-MOG VDJ region with the AK23 VDJ sequence derived from the vector used to generate AK23IgMtg mice (17). The vector contained a neomycin resistance gene cassette and homologous sequences derived from a cosmid vector (Fig. 1A). The targeting vector was linearized and electroporated into v6.5 mouse hybrid embryonic stem (ES) cells (C57BL/6 × 129/Sv), and 240 neomycin-resistant clones were selected. Five out of 240 clones were determined as transgene positive by PCR screening. Southern blot analysis was performed using Sac1-digested ES cell DNA and hybridization to a 357-bp digoxigenin-labeled external probe (Fig. 1A and 1B). One correctly targeted clone was obtained, and the insertion was confirmed by sequencing. The positive ES clone was aggregated with eight-cell BDF2 embryos and implanted into pseudopregnant hosts. The resulting germline-transmitting chimeras were obtained and crossed with C57BL/6 mice to produce AK23 H chain knock-in mice (AK23Hki). After backcrossing AK23Hki mice to C57BL/6 >10 times, they were crossed with AK23 L chain transgenic mice (17) for generating AK23 Ig knock-in mice (AK23HkiLtg) on a C57BL/6 genetic background.

FIGURE 1.

Dsg3-specific BCR knock-in mice do not develop PV phenotype in vivo. (A) Scheme of the AK23 H chain knock-in vector. The probe used for Southern blot analysis in (B) is indicated with a thick black line. (B) Southern blot analysis of knocked-in ES cells. The WT and knocked-in alleles were detected as 4-kb and 2-kb fragments, respectively. Arrow indicates the fragment of knocked-in allele. (C) Dsg3 specificities of splenocytes from AK23HkiLtg (n = 2), AK23Hki (n = 2), and WT (n = 2) mice by flow cytometry. (D) Splenocytes from AK23Hki (n = 2) versus WT mice (n = 2) and from AK23HkiLtg (n = 3) versus WT (n = 3) mice, respectively, determined by flow cytometry. (E) Serum anti-Dsg3 IgM and IgG1 levels in WT (n = 6), AK23Hki (n = 7), and AK23HkiLtg (n = 6) mice were determined by ELISA. Serum from AK23IgMtg mice (n = 3) and PV model mice, which was established by adoptive transfer of lymphocytes from Dsg3−/− mice immunized with rDsg3 to Rag2−/− mice (n = 3), was used as a positive control for anti-Dsg3 IgM and IgG1 quantification, respectively. (F) Skin phenotype, histopathology, and immunofluorescence in AK23HkiLtg and WT mice. Dashed lines indicate basement membrane zones. Scale bar, 20 μm. Positive controls (PCs) for IgM and IgG1 Ab detection are the spleen of AK23HkiLtg mice and the palate of PV model mice, respectively. Data are shown as the mean ± SEM. *p < 0.05, **p < 0.01; Student t test.

FIGURE 1.

Dsg3-specific BCR knock-in mice do not develop PV phenotype in vivo. (A) Scheme of the AK23 H chain knock-in vector. The probe used for Southern blot analysis in (B) is indicated with a thick black line. (B) Southern blot analysis of knocked-in ES cells. The WT and knocked-in alleles were detected as 4-kb and 2-kb fragments, respectively. Arrow indicates the fragment of knocked-in allele. (C) Dsg3 specificities of splenocytes from AK23HkiLtg (n = 2), AK23Hki (n = 2), and WT (n = 2) mice by flow cytometry. (D) Splenocytes from AK23Hki (n = 2) versus WT mice (n = 2) and from AK23HkiLtg (n = 3) versus WT (n = 3) mice, respectively, determined by flow cytometry. (E) Serum anti-Dsg3 IgM and IgG1 levels in WT (n = 6), AK23Hki (n = 7), and AK23HkiLtg (n = 6) mice were determined by ELISA. Serum from AK23IgMtg mice (n = 3) and PV model mice, which was established by adoptive transfer of lymphocytes from Dsg3−/− mice immunized with rDsg3 to Rag2−/− mice (n = 3), was used as a positive control for anti-Dsg3 IgM and IgG1 quantification, respectively. (F) Skin phenotype, histopathology, and immunofluorescence in AK23HkiLtg and WT mice. Dashed lines indicate basement membrane zones. Scale bar, 20 μm. Positive controls (PCs) for IgM and IgG1 Ab detection are the spleen of AK23HkiLtg mice and the palate of PV model mice, respectively. Data are shown as the mean ± SEM. *p < 0.05, **p < 0.01; Student t test.

Close modal

CD45.2+Rag2−/− mice were crossed with CD45.1+ B6.SJL WT mice several times to generate CD45.1+Rag2−/− mice. CD45.2+ AK23HkiLtg mice, generated with the method described above, were also mated with CD45.2+Rag2−/− mice several times to generate CD45.2+ AK23HkiLtg-Rag2−/− mice. Subsequently, CD45.1+Rag2−/− mice were crossed with CD45.2+ AK23HkiLtg-Rag2−/− mice repeatedly to establish the CD45.1+ AK23HkiLtg-Rag2−/− mouse line.

CD45.2+ WT and CD45.2+Dsg3−/− mice were lethally irradiated. The minimum lethal doses of CD45.2+ WT and CD45.2+Dsg3−/− mice were 9–9.5 Gy and 7.5–8 Gy, respectively. After irradiation, they received bone marrow cells from CD45.1+ AK23HkiLtg-Rag2−/− mice. Congenic markers, CD45.1 and CD45.2, are used to distinguish donor and recipient cells, respectively. The recipients were analyzed at the age of 16–20 wk.

B cells were purified from splenocytes by positive selection using magnetic bead–conjugated anti-mouse B220 mAb (clone RA3-6B2; Miltenyi Biotec, Bergisch Gladbach, Germany). Isolated B cells (5 × 105/ml) were incubated with IL-4 (20 ng/ml) and LPS (10 μg/ml) for 96 h. Then, 8 × 106 B cells were adoptively transferred into Rag2−/− mice via the tail vein. The recipients were analyzed at the age of 16–18 wk.

Palate tissues were embedded in O.C.T. optimal cutting temperature compound (Tissue-Tek; Sakura Finetek, Tokyo, Japan), frozen immediately at −140°C, and cut into 6-μm-thick sections. Sections were incubated with Alexa Fluor 488 anti-mouse IgG1 Ab (Invitrogen, Carlsbad, CA) for 1 h at room temperature, washed in PBS, coverslipped using MOWIOL (Calbiochem, Darmstadt, Germany), and imaged using a fluorescence microscope.

Serum anti-Dsg3 Ig titers were determined by ELISA using plates coated with 3 μg/ml or 5 μg/ml rDsg3 protein as described previously (15). Each serum sample was diluted 1:200, 1:500, and 1:1000. Sera obtained from PV model mice (15) and culture supernatant of AK23 hybridoma cells (16) were used as positive controls for anti-Dsg3 IgG. AK23IgMtg mice (17) were used as positive controls for anti-Dsg3 IgM. The index value for anti-Dsg3 IgG was calculated as previously described (14). Total IgG Ab concentrations were determined using the IgG Mouse ELISA kit (Abcam, Cambridge, UK) according to the manufacturer’s instructions.

Single-cell suspensions were prepared from mouse peripheral blood, spleens, and lymph nodes. Cells were stained with the following Ab conjugates: biotin-conjugated anti-IgMa (DS-1; BD Biosciences, San Jose, CA), anti-IgG1 (B68-2; BD Biosciences), anti–BP-1 (6C3; BioLegend, San Diego, CA), and anti-IgM (RMM-1; BioLegend); FITC-conjugated anti-IgMa (DS-1; BD Biosciences) and anti–E-tag (Bethyl Laboratories, Montgomery, TX); PE-conjugated anti-IgMb (AF-6-78; BD Biosciences), anti-CD23 (B3B4; BD Biosciences), anti-IgM (AF6-78; BD Biosciences), and streptavidin; allophycocyanin-conjugated anti-B220 (RA3-6B2; BD Biosciences), anti-CD24 (M1/69; BioLegend), anti-IgD (11-26c.2a; BioLegend), and streptavidin; allophycocyanin–cyanine 7 (Cy7)-conjugated anti-CD21 (7E9; BioLegend); PE/Cy7-conjugated anti-IgM (RMM-1; BioLegend); Pacific Blue–conjugated anti-B220 (RA3-6B2; BioLegend), and streptavidin.

To determine Dsg3 binding to B cells, we purified B cells (Fig. 1C) by positive selection using magnetic beads (MACS; Miltenyi Biotec) conjugated with an anti-mouse B220 mAb (clone RA3-6B2; Miltenyi Biotec) according to the manufacturer’s instructions. Subsequently, cells were incubated with 10 μg of rDsg3-E-His in 100 μl of RPMI on ice for 1 h and stained with anti–E-tag-FITC (Bethyl Laboratories). After excluding dead cells by staining with 7-aminoactinomycin D, cells were analyzed with a FACSCanto II flow cytometer (BD Biosciences), and the resulting data were analyzed using FlowJo software (BD Biosciences, Ashland, OR).

Splenocytes were stained with anti–CD45.1-PE/Cy7 and anti–B220-allophycocyanin (BioLegend) at 4°C for 30 min in PBS. Then, cells were incubated in HBSS containing 10 µM Fluo-4 (Molecular Probes, Eugene, OR) at room temperature for 30 min. After stimulation with 20 µg/ml polyclonal anti-IgM F(ab′)2 fragment, AffiniPure F(ab′)2 Fragment Goat Anti-Mouse IgM, μ Chain Specific (115-006-020; Jackson ImmunoResearch, West Grove, PA), calcium mobilization was measured using the FACSCanto II system (BD Biosciences).

A recombinant baculoprotein of mouse Dsg3 that included the extracellular domain of mouse Dsg3, an E-tag, and an His-tag was used for immunization as described previously (15). Mice were primed by s.c. injection of purified rDsg3 (5 μg) in CFA. Subsequently, mice were boosted twice with rDsg3 (5 μg) in IFA, followed by two i.p. injections of rDsg3 (5 μg) without adjuvant, once per week.

A total of 80–100 μl of 0.5% 2,4-dinitro-1-fluorobenzene (DNFB; Nacalai Tesque, Kyoto, Japan) in acetone/olive oil (3:1) was applied to the shaved abdomens or backs of mice weekly up to four to seven times, depending on the experiment.

Vaccinia virus strain VV-WR was a kind gift from Masayuki Saijo (National Institute of Infectious Diseases, Department of Virology I, Tokyo, Japan). Mice were infected with VV-WR by skin scarification. Briefly, under anesthesia, abdomens or backs of mice were shaved and scratched with a 27-gauge needle. VV-WR (1 × 108 PFU/ml, 150–200 μl/mouse) was applied to the scarred skin areas every 2 or 3 wk.

The study protocols were reviewed and approved by the institutional review boards of the Keio University School of Medicine (approval no. 20120180) and RIKEN [approval no. H27-5(7)], and were conducted following the principles established by the Declaration of Helsinki. Written informed consent was obtained from all patients. B lymphocytes in peripheral blood from 8 patients with pemphigus (six patients with PV and two patients with pemphigus foliaceus) and nine healthy donors were identified as CD3CD19+ lymphocytes and sorted with a FACSAria III flow cytometer (BD Biosciences). Total RNA was extracted from the sorted lymphocytes using TRIzol reagent (Invitrogen). The detailed clinical information of the patients is shown in Supplemental Table I. RNA-sequencing libraries were prepared using the NEBNext Ultra RNA Library Prep Kit for Illumina (New England Biolabs, Ipswich, MA) according to the manufacturer’s instructions and sequenced using a HiSeq 2500 system (Illumina) on a 50-base single-end read mode. The sequence reads were mapped to the University of California, Santa Cruz Genome Browser hg37 reference genome using TopHat2 version 2.0.8 and botwie2 version 2.1.0 with default parameters, and gene annotation was provided by National Center for Biotechnology Information RefSeq.

The transcript abundances were estimated using Cufflinks (version 2.1.1). The negative binomial model-based method edgeR (3.10.0) was used for differential expression analysis from raw count data. Normalized trimmed means of M-values were visualized in a heatmap.

Affected biological pathways and functions (p < 0.05) in pemphigus peripheral B cells were identified through analyses by Canonical Pathway and Diseases and Biological Functions provided in Ingenuity Pathway Analysis software (Qiagen, Germantown, MD) (19).

Splenic IgM+ cells were isolated by positive selection using magnetic beads (MACS; Miltenyi Biotec). Total RNA of the IgM+ cell was extracted using the RNeasy Mini Kit (Qiagen), and cDNA synthesis was performed using TaqMan Reverse Transcription Reagent (Applied Biosystems, Foster City, CA). Reactions were prepared using the Universal SYBR Select Master Mix, and quantitative PCR was performed using a StepOnePlus Real-Time PCR System (Applied Biosystems) and the following primers: HPRT forward, 5′-AGCCTAAGATGAGCGCAAGT-3′; HPRT reverse, 5′-TTACTAGGCAGATGGCCACA-3′; postswitch transcript (PST) γ1 forward, 5′-ACCTGGGAATGTATGGTTGTGGCTT-3′; PST γ1 reverse, 5′-ATGGAGTTAGTTTGGGCAGCA-3′.

All data are shown as mean ± SEM. Data were analyzed by Student t test unless specified otherwise. Significance was set at p < 0.05.

To generate a Dsg3-specific BCR knock-in mouse line, we constructed an IgH knock-in vector wherein the endogenous JH gene cluster was replaced with the VDJ gene segment of the AK23 H chain gene (Fig. 1A). AK23Hki mice were generated by injecting the knock-in vector into v6.5 mouse ES cells (C57BL/6 × 129/Sv) (Fig. 1B); AK23Hki mice were backcrossed to C57BL/6 mice for >10 generations. AK23Hki mice were then crossed with AK23 L chain transgenic mice (17) to generate AK23HkiLtg mice. To confirm Dsg3 reactivity of B cells generated in AK23HkiLtg mice, we performed flow cytometry after incubation of the B cells with rDsg3 protein. As expected, Dsg3-bound B cells were detected as a single population positively shifted in plots (Fig. 1C). When we analyzed whether the targeted allele harbored IgMa or IgMb allotypes originating from 129/Sv or C57BL/6 mice, respectively, we found that B cells isolated from AK23Hki mice and AK23HkiLtg mice expressed the IgMa, but not IgMb, allotype derived from 129/Sv chromosome (Fig. 1D). These findings imply that Dsg3-specific AK23 BCR was successfully expressed in B cells as IgMa due to allelic exclusion in AK23HkiLtg mice.

Next, we assessed tissue deposition of circulating Igs. In contrast to AK23Hki or WT mice, circulating anti-Dsg3 IgM was abundant in AK23HkiLtg mice (Fig. 1E). However, immunofluorescence staining revealed that in vivo IgM deposition on keratinocyte cell surfaces occurred at extremely low levels in the palate of AK23HkiLtg mice (Fig. 1F). Although Dsg3-specific AK23 BCR knock-in B cells maintained the potential for class switching, only marginal levels of circulating anti-Dsg3 IgG1 were detected in AK23HkiLtg mice, which were considerably lower than in PV model mice used hereafter to indicate the previously reported active disease model established by adoptive transfer of lymphocytes from Dsg3−/− mice immunized with rDsg3 to Rag2−/− mice (15) (Fig. 1E). Consistently, in vivo IgG1 deposition in the palate was minimal in AK23HkiLtg mice (Fig. 1F). In line with these results, we did not observe skin erosion, hair loss, acantholysis, or other pemphigus phenotype in AK23HkiLtg mice (Fig. 1F). Furthermore, up to 1 y of age, neither serum anti-Dsg3 IgG1 nor in vivo IgG1 deposition levels spontaneously increased to the levels that induce acantholytic blister, and AK23HkiLtg mice did not develop cutaneous phenotypes such as erosions and hair loss, suggesting that minor skin irritation or other abrasion associated with aging does not cause pemphigus phenotypes in the mice (n = 18; data not shown).

These results imply that despite the Dsg3 reactivity of AK23HkiLtg B cells, differentiation into pathogenic IgG+ B cells was not observed even in the presence of Dsg3, and a pemphigus phenotype was not induced.

Previous studies have shown that partial clonal deletion and anergy occurred even in BCR transgenic mice that produced autoreactive IgM Abs (20). Although most AK23HkiLtg B cells expressed IgM, AK23HkiLtg B cells may undergo clonal deletion or anergy. Thus, we assessed AK23HkiLtg B cell deletion or anergy in the presence and absence of Dsg3. To this end, we transplanted BM cells from AK23HkiLtg-Rag2−/− mice (CD45.1+) into irradiated WT and Dsg3−/− mice (CD45.2+) (Fig. 2A) and assessed AK23HkiLtg B cell development in the presence and absence of Dsg3, respectively. The Hardy fraction, a representative classification, was used for BM analysis (21). We found no significant difference in the ratio of Hardy fractions A to F during AK23HkiLtg B cell development in the two groups (Fig. 2B). We performed similar analyses using splenic B cells, which consist of cells at transitional stages (T1, T2, and T3), follicular B cells, and marginal zone B cells (22). No significant differences in these subpopulations between the two groups were observed (Fig. 2C). These findings imply that AK23HkiLtg B cells are not subjected to Dsg3-dependent clonal deletion.

FIGURE 2.

AK23HkiLtg B cells are not subjected to clonal deletion or anergy. (A) Scheme of bone marrow transfer (BMT) experiments. All experiments were performed with AK23HkiLtg B cells from WT and Dsg3−/− recipients after BMT. (B) AK23HkiLtg B cell development in WT and Dsg3−/− recipients assessed by flow cytometry. Hardy fractions A–C and D–F B cells were identified after gating on CD45.1+B220+CD43high and CD45.1+B220+CD43low cells, respectively. Representative flow cytometry plots and quantitative summaries of the proportions of each developmental stage (n = 3) are shown. (C) Splenic AK23HkiLtg B cell development in WT and Dsg3−/− recipients assessed by flow cytometry. Donor-derived transitional (T1, T2, and T3) and follicular (Fol)/marginal zone (Mz) B cells were identified after gating on CD45.1+Gr-1CD11bCD19+B220highCD93+ and CD45.1+Gr-1CD11bCD19+B220highCD93 cells, respectively. Proportions of these B cells are shown in representative flow cytometry plots and bar graphs (n = 3). (D) Ca2+ influx in splenic AK23HkiLtg B cells from WT and Dsg3−/− recipients after anti-IgM Ab stimulation (arrow) assessed by flow cytometry. Fluorescence intensity (1) and the time (2) between anti-IgM stimulation and fluorescence intensity peak were determined (n = 4). (E) Quantitation of CD80, CD86, surface IgM, and MHC class II expression levels examined by flow cytometry in splenic AK23HkiLtg B cells from WT and Dsg3−/− recipients. WT B cells from WT recipients after BMT from WT mice were used as controls (n = 5). Data are shown as the mean ± SEM. MFI, mean fluorescence intensity.

FIGURE 2.

AK23HkiLtg B cells are not subjected to clonal deletion or anergy. (A) Scheme of bone marrow transfer (BMT) experiments. All experiments were performed with AK23HkiLtg B cells from WT and Dsg3−/− recipients after BMT. (B) AK23HkiLtg B cell development in WT and Dsg3−/− recipients assessed by flow cytometry. Hardy fractions A–C and D–F B cells were identified after gating on CD45.1+B220+CD43high and CD45.1+B220+CD43low cells, respectively. Representative flow cytometry plots and quantitative summaries of the proportions of each developmental stage (n = 3) are shown. (C) Splenic AK23HkiLtg B cell development in WT and Dsg3−/− recipients assessed by flow cytometry. Donor-derived transitional (T1, T2, and T3) and follicular (Fol)/marginal zone (Mz) B cells were identified after gating on CD45.1+Gr-1CD11bCD19+B220highCD93+ and CD45.1+Gr-1CD11bCD19+B220highCD93 cells, respectively. Proportions of these B cells are shown in representative flow cytometry plots and bar graphs (n = 3). (D) Ca2+ influx in splenic AK23HkiLtg B cells from WT and Dsg3−/− recipients after anti-IgM Ab stimulation (arrow) assessed by flow cytometry. Fluorescence intensity (1) and the time (2) between anti-IgM stimulation and fluorescence intensity peak were determined (n = 4). (E) Quantitation of CD80, CD86, surface IgM, and MHC class II expression levels examined by flow cytometry in splenic AK23HkiLtg B cells from WT and Dsg3−/− recipients. WT B cells from WT recipients after BMT from WT mice were used as controls (n = 5). Data are shown as the mean ± SEM. MFI, mean fluorescence intensity.

Close modal

To assess whether AK23HkiLtg B cells are subjected to anergy, we isolated AK23HkiLtg splenic B cells from the BM of chimeric recipients (Fig. 2A) and analyzed their responses to anti-IgM stimulation using Fluo-4, a dye that emits fluorescence upon calcium influx after Ag stimulation. When AK23HkiLtg B cells derived from WT mice were stimulated, calcium influx was observed by flow cytometry (Fig. 2D). To statistically analyze the difference in Ag-initiated B cell reactivity, we considered the difference in fluorescence intensity and the time between stimulation and peak fluorescence intensity; no significant difference was detected (Fig. 2D). In addition, expression levels of cell surface markers such as CD80, CD86, IgM, and MHC class II, which are usually attenuated in the anergic state, were also comparable between these two groups (Fig. 2E).

Next, we analyzed AK23HkiLtg B cells derived from skin-draining lymph nodes (sLNs) in AK23HkiLtg mice to further confirm the phenotypic similarity of AK23HkiLtg B cells between spleen and sLN-derived cells. A substantial population of Dsg3-binding B cells was confirmed in AK23HkiLtg B cells from sLNs by flow cytometry (Supplemental Fig. 1A), and the AK23HkiLtg B cells responded to anti-IgM stimulation with the elevated expression levels of CD86 and MHC class II after the stimulation (Supplemental Fig. 1B). No clear differences in findings associated with deletion and anergy of AK23HkiLtg B cells were observed between spleen and sLNs.

We also investigated phosphorylated components of BCR signaling such as phosphorylated Bruton tyrosine kinase and phosphorylated phospholipase Cγ2 in splenic AK23HkiLtg B cells. After the anti-IgM stimulations, the expression levels of these phosphorylated components in AK23HkiLtg B cells are significantly elevated like they are in WT B cells (Supplemental Fig. 1C), implying no clear defect in the signaling cascade, at least, immediately after BCR.

These results together indicated that AK23HkiLtg B cells were not subjected to clonal deletion or anergy. Instead, AK23HkiLtg B cells remained nonpathogenic in a steady state in secondary lymphoid organs, retaining their potential to respond to BCR stimulation.

To investigate whether AK23HkiLtg B cells produce IgG autoantibodies after class switching, we forced class switching to IgG1 by stimulating B cells with IL-4 and LPS in vitro (Fig. 3A). We found that an AK23HkiLtg B cell subpopulation was converted into IgG1+ B cells, and most of these cells were reactive to rDsg3 protein (Fig. 3B). After adoptive transfer of IgG1-producing AK23HkiLtg B cells into Rag2−/− mice, transplanted mice developed skin erosions, crusted skin lesions, and hair loss in the skin (Fig. 3C). Furthermore, immunofluorescence analysis revealed IgG1 deposition on keratinocyte cell surfaces and acantholytic blister, the two most characteristic features of pemphigus (Fig. 3C). These findings indicate that AK23HkiLtg B cells retained their potential for class switching and that IgG autoantibodies produced by AK23HkiLtg B cells were pathogenic in vivo.

FIGURE 3.

Ig class switching in AK23HkiLtg B cells initiates PV phenotype induction. (A) Schematic outline of the experimental procedures. (B) Flow cytometric analysis of splenic B cells from AK23HkiLtg, AK23Hki, and WT mice before (day 0) and after (day 4) in vitro stimulation with IL-4 and LPS to detect Dsg3-specific IgG1+ B cells. (C) Clinical presentation, H&E staining, and immunofluorescence staining of Rag2−/− recipients (n = 2) after adoptive transfer of AK23HkiLtg IgG1+ B cells. PV phenotypes, including skin erosions and hair loss, were recorded. IgG1 deposition on keratinocyte cell surfaces (green) and acantholysis (asterisks) are shown. Scale bar, 20 μm. (D) Scheme of immunization procedure. (E) Serum anti-Dsg3 IgG1 was quantified by ELISA in AK23HkiLtg mice after serial immunization with rDsg3 (n = 5) and PBS (n = 4). Data are pooled from two independent experiments. (F) Skin phenotype of Dsg3-immunized AK23HkiLtg and control mice. IgG1 depositions are shown in green, acantholytic blister is indicated by black arrowheads, and hair loss and skin erosion are indicated by yellow arrowheads. (G) Scheme of DNFB treatment. DNFB was applied weekly to the trunk of AK23HkiLtg, WT, and AK23IgMtg mice. (H) Anti-Dsg3 IgG1 Ab was quantified after DNFB treatment of AK23HkiLtg (n = 8), WT (n = 4), and AK23IgMtg (n = 3) mice by ELISA. Data were pooled from two independent experiments. (I) Skin phenotype of AK23HkiLtg mice after DNFB treatment. Hair loss and erosion are indicated by yellow arrowheads, acantholytic blister is indicated by black arrowheads, and IgG1 depositions are shown in green. Scale bar, 20 μm. *p < 0.05, **p < 0.01. Statistical significance was determined by Mann–Whitney U test (E and H). Data are shown as the mean ± SEM.

FIGURE 3.

Ig class switching in AK23HkiLtg B cells initiates PV phenotype induction. (A) Schematic outline of the experimental procedures. (B) Flow cytometric analysis of splenic B cells from AK23HkiLtg, AK23Hki, and WT mice before (day 0) and after (day 4) in vitro stimulation with IL-4 and LPS to detect Dsg3-specific IgG1+ B cells. (C) Clinical presentation, H&E staining, and immunofluorescence staining of Rag2−/− recipients (n = 2) after adoptive transfer of AK23HkiLtg IgG1+ B cells. PV phenotypes, including skin erosions and hair loss, were recorded. IgG1 deposition on keratinocyte cell surfaces (green) and acantholysis (asterisks) are shown. Scale bar, 20 μm. (D) Scheme of immunization procedure. (E) Serum anti-Dsg3 IgG1 was quantified by ELISA in AK23HkiLtg mice after serial immunization with rDsg3 (n = 5) and PBS (n = 4). Data are pooled from two independent experiments. (F) Skin phenotype of Dsg3-immunized AK23HkiLtg and control mice. IgG1 depositions are shown in green, acantholytic blister is indicated by black arrowheads, and hair loss and skin erosion are indicated by yellow arrowheads. (G) Scheme of DNFB treatment. DNFB was applied weekly to the trunk of AK23HkiLtg, WT, and AK23IgMtg mice. (H) Anti-Dsg3 IgG1 Ab was quantified after DNFB treatment of AK23HkiLtg (n = 8), WT (n = 4), and AK23IgMtg (n = 3) mice by ELISA. Data were pooled from two independent experiments. (I) Skin phenotype of AK23HkiLtg mice after DNFB treatment. Hair loss and erosion are indicated by yellow arrowheads, acantholytic blister is indicated by black arrowheads, and IgG1 depositions are shown in green. Scale bar, 20 μm. *p < 0.05, **p < 0.01. Statistical significance was determined by Mann–Whitney U test (E and H). Data are shown as the mean ± SEM.

Close modal

To assess whether class switching in AK23HkiLtg B cells can be induced in vivo, we repeatedly immunized AK23HkiLtg mice with rDsg3 protein and adjuvant (Fig. 3D). In contrast to mice treated with adjuvant alone, rDsg3-immunized AK23HkiLtg mice started producing anti-Dsg3 IgG1 Abs and developed a PV phenotype at ∼4 wk after treatment (Fig. 3E and 3F). Consistently, acantholytic blister and IgG1 deposition on keratinocyte cell surfaces were histologically detected in rDsg3-immunized AK23HkiLtg mice but not in mice immunized with adjuvant alone (Fig. 3F), indicating that Dsg3 immunization promoted IgM to IgG class switching in Dsg3-specific BCR knock-in B cells in vivo.

Next, to evaluate whether skin inflammation that potentially causes endogenous Dsg3 exposure of the immune system is sufficient to drive class switching in AK23HkiLtg B cells, we employed a hapten-induced dermatitis model where Dsg3-expressing keratinocytes were damaged by forced dermatitis. Repeated application of DNFB on shaved trunk skin induced severe dermatitis and skin erosion (Fig. 3G). In addition, DNFB treatment promoted anti-Dsg3 IgG1 production in AK23HkiLtg mice but not in WT or AK23IgMtg mice (Fig. 3H). In fact, on the one hand, the PV phenotype including erosions and hair loss was observed on the face of AK23HkiLtg mice after DNFB was applied only on trunk skin (Fig. 3I). On the other hand, dermatitis did not develop on the face in WT and AK23IgMtg mice (Fig. 3I and data not shown). In addition, acantholytic blister and IgG1 deposition on the keratinocyte cell surfaces were confirmed only in AK23HkiLtg mice (Fig. 3I).

Viral infections have been implicated in autoimmune disease exacerbation (23). Hence, we assessed the ability of a skin-tropic virus, vaccinia virus, to induce class switching in AK23HkiLtg B cells by damaging Dsg3-bearing keratinocytes. To this end, we inoculated the trunk skin of AK23HkiLtg mice with vaccinia virus (Supplemental Fig. 1D). After repeated inoculation, we observed skin erosion and IgG1 deposition on keratinocyte cell surfaces of AK23HkiLtg mice (Supplemental Fig. 1E, 1F). These results indicate that severe skin damage is one of the exogenous causes of pathogenic class switching in Dsg3-specific B cells and drives autoimmunity in mice. However, our data showed that this harmful process does not occur in AK23HkiLtg B cells under physiological conditions.

Although forced skin inflammation was considered as one of the extrinsic factors to promote class switching in Dsg3-specific B cells in mice, skin damage is not always a causative driver of pemphigus development or exacerbation in the clinical situation (24). To more precisely understand disease pathophysiology toward spontaneous development of the pemphigus phenotype, we shifted to investigation of the possible involvement of an immune-related intrinsic pathway in driving class switching of Dsg3-specific B cells. The FcγRIIB pathway is one of the representatives because altered FcγRIIB function by its polymorphism has been reported to associate with several autoimmune diseases (2529). We investigated pathogenic roles of systemic FcγRIIB signaling in class switching by using the Fcgr2b−/−-AK23HkiLtg mouse on a C57BL/6 genetic background as an initial step.

The body weight of Fcgr2b−/−-AK23HkiLtg mice was significantly lower than that of AK23HkiLtg mice after 6 wk of age (Fig. 4A). Consistent with the weight loss phenotype often observed after the induction of the pemphigus phenotype, anti-Dsg3 IgG Ab levels were also elevated in Fcgr2b−/−-AK23HkiLtg mice after 6 wk of age compared with AK23HkiLtg mice (Fig. 4B). However, we observed no differences in total IgG levels between Fcgr2b−/− and WT mice until 16 wk after birth (Fig. 4C). Importantly, even Fcgr2b+/−-AK23HkiLtg mice, but not AK23HkiLtg mice, developed a PV phenotype including skin erosion and hair loss, as well as acantholytic blister and IgG deposition, including IgG1, IgG2a, and IgG2b, but not IgG3, on keratinocyte cell surfaces (Fig. 4D and Supplemental Fig. 2A–D). In addition, the survival rates of Fcgr2b−/−-AK23HkiLtg and Fcgr2b+/−-AK23HkiLtg mice were significantly lower than those of AK23HkiLtg mice (Fig. 4E).

FIGURE 4.

FcγRIIB that constrains class switching maintains the immunoregulatory state for preventing spontaneous development of pemphigus phenotype in the AK23HkiLtg mouse. (A) Body weight of Fcgr2b−/−-AK23HkiLtg (closed circle; n = 7) and AK23HkiLtg (open circle; n = 6) mice. (B) Anti-Dsg3 IgG Ab levels in Fcgr2b−/−-AK23HkiLtg (closed circle; n = 7) and AK23HkiLtg (open circle; n = 5) mice were determined by ELISA. (C) Nonspecific total IgG Ab concentration (mg/ml) in Fcgr2b−/− (closed circle; n = 6) and WT (open circle; n = 6) mice was determined by ELISA. (D) Skin of Fcgr2b+/−-AK23HkiLtg mice. Hair loss and erosion are indicated by yellow arrowheads, IgG depositions are shown in green (original magnification ×200), and acantholytic blister is indicated by black arrowheads (original magnification ×100). (E) Survival rates of Fcgr2b−/−-AK23HkiLtg (red line; n = 11), Fcgr2b+/−-AK23HkiLtg (black line; n = 10), and AK23HkiLtg (blue line; n = 54) mice were assessed in single-cohort analysis and are shown. (F) Ex vivo PST γ1 expression levels in B cells from Fcgr2b−/− (n = 3), Fcgr2b+/− (n = 3), and WT (n = 3) mice. Statistical significance was determined by two-way ANOVA (A–C), log-rank test (E), or t test (F). Data are shown as the mean ± SEM. *p < 0.05, **p < 0.01.

FIGURE 4.

FcγRIIB that constrains class switching maintains the immunoregulatory state for preventing spontaneous development of pemphigus phenotype in the AK23HkiLtg mouse. (A) Body weight of Fcgr2b−/−-AK23HkiLtg (closed circle; n = 7) and AK23HkiLtg (open circle; n = 6) mice. (B) Anti-Dsg3 IgG Ab levels in Fcgr2b−/−-AK23HkiLtg (closed circle; n = 7) and AK23HkiLtg (open circle; n = 5) mice were determined by ELISA. (C) Nonspecific total IgG Ab concentration (mg/ml) in Fcgr2b−/− (closed circle; n = 6) and WT (open circle; n = 6) mice was determined by ELISA. (D) Skin of Fcgr2b+/−-AK23HkiLtg mice. Hair loss and erosion are indicated by yellow arrowheads, IgG depositions are shown in green (original magnification ×200), and acantholytic blister is indicated by black arrowheads (original magnification ×100). (E) Survival rates of Fcgr2b−/−-AK23HkiLtg (red line; n = 11), Fcgr2b+/−-AK23HkiLtg (black line; n = 10), and AK23HkiLtg (blue line; n = 54) mice were assessed in single-cohort analysis and are shown. (F) Ex vivo PST γ1 expression levels in B cells from Fcgr2b−/− (n = 3), Fcgr2b+/− (n = 3), and WT (n = 3) mice. Statistical significance was determined by two-way ANOVA (A–C), log-rank test (E), or t test (F). Data are shown as the mean ± SEM. *p < 0.05, **p < 0.01.

Close modal

To elucidate critical roles of systemic deficiency of FcγRIIB in class switching of B cells, we evaluated the class-switch efficiency of Fcgr2b−/−, Fcgr2b+/−, and WT IgM+ B cells ex vivo. The expression levels of PST γ1, which is a by-product of class-switch recombination to IgG1, were significantly higher in Fcgr2b−/− and Fcgr2b+/− B cells than in WT B cells (Fig. 4F). In addition, PST γ1 levels were significantly higher in Fcgr2b−/− B cells than in Fcgr2b+/− B cells (Fig. 4F). These results imply that FcγRIIB suppresses class-switch recombination in a gene dose-dependent manner, consistent with the differential survival among Fcgr2b+/+-AK23HkiLtg, Fcgr2b+/−-AK23HkiLtg, and Fcgr2b−/−-AK23HkiLtg mice (Fig. 4E). To elucidate whether FcγRIIB deficiency affects Dsg3-specific B cell development, we analyzed B cell development in the BM and spleen of Fcgr2b−/+ AK23HkiLtg mice and AK23HkiLtg mice. Our results showed that there was no difference in B cell development between these mice (Supplemental Fig. 2E, 2F), implying that anti-Dsg3 IgG production promoted in Fcgr2b−/+ AK23HkiLtg mice (Fig. 4D) is not associated with abnormal B cell development. Taken together, FcγRIIB signaling is one of the pivotal intrinsic factors that does not allow development of a pemphigus phenotype, presumably through suppressing pathogenic class switching of Dsg3-specific B cells in AK23HkiLtg mice.

To unveil intrinsic factors associated with B cell pathogenicity from clinical samples in an unbiased manner, we performed genome-wide gene expression analysis of CD19+ B cells isolated from patients with pemphigus (n = 8) and healthy individuals (n = 9). Gene expression analysis identified 84 of 577 immune-related genes as being differentially expressed between two groups (p < 0.05 by edgeR) (Fig. 5A, Supplemental Table I). To narrow the pivotal candidate molecules that are associated with pemphigus pathogenesis, we further analyzed enriched biological functions in patient samples with Ingenuity Pathway Analysis (Fig. 5B). As expected, activation of B lymphocytes was most highly enriched in our dataset, and FCGR2B, SEMA4D, and ZAP70 were included among genes altered in the affected biological function (Fig. 5B). Furthermore, FcγRIIB signaling was also included in the affected biological pathways estimated by pathway analysis based on the identified 84 genes indicating that the expression was altered in patient B cells (Supplemental Fig. 3). FCGR2B and other downstream molecules of FcγRIIB, including DOK1, BTK, and GSK3A, were downregulated in patients with pemphigus compared with healthy individuals (Fig. 5C). Meanwhile, KRAS, the expression of which is suppressed by FcγRIIB signaling (30), was upregulated in patients with pemphigus (Fig. 5C). These results suggested that the attenuated FcγRIIB pathway is one of the immunological features observed in peripheral blood B cells of patients with pemphigus, consistent with mouse data showing protective function of FcγRIIB in pemphigus development.

FIGURE 5.

FcγRIIB pathway is downregulated in peripheral blood B cells of patients with pemphigus. (A) Heatmap of hierarchical clustering of immune-related genes differentially expressed (p < 0.05) in peripheral blood B cells from patients with pemphigus (pemphigus vulgaris, n = 6; pemphigus foliaceus, n = 2) and healthy control individuals (HCs; n = 9). (B) Enriched biological functions in pemphigus peripheral B cells and genes involved in each function are shown. (C) Expression levels (trimmed mean of M [TMM] values) of FcγRIIB and FcγRIIB pathway-related molecules. Data are shown as the mean ± SEM. *p < 0.05, **p < 0.01; Student t test.

FIGURE 5.

FcγRIIB pathway is downregulated in peripheral blood B cells of patients with pemphigus. (A) Heatmap of hierarchical clustering of immune-related genes differentially expressed (p < 0.05) in peripheral blood B cells from patients with pemphigus (pemphigus vulgaris, n = 6; pemphigus foliaceus, n = 2) and healthy control individuals (HCs; n = 9). (B) Enriched biological functions in pemphigus peripheral B cells and genes involved in each function are shown. (C) Expression levels (trimmed mean of M [TMM] values) of FcγRIIB and FcγRIIB pathway-related molecules. Data are shown as the mean ± SEM. *p < 0.05, **p < 0.01; Student t test.

Close modal

Immunological tolerance is crucial for preventing autoimmunity. In fact, interruptions of representative immune tolerance mechanisms, such as deletion, anergy, and receptor editing, lead to autoimmunity (3133). Given that anti-Dsg3 IgM Abs are not pathogenic, their class switching to IgG autoantibodies is essential for pemphigus induction. However, class switching of anti-Dsg3 IgM Abs to IgG autoantibodies is absent under physiological conditions in AK23HkiLtg mice, even though Dsg3-specific knock-in B cells maintain their ability of class switching. Our findings elucidated a characteristic immune state that does not allow Ig class switching even in the presence of corresponding autoantigens under physiological conditions. Pathogenic class switching was forced to occur upon immunization with the rDsg3 protein or upon severe skin inflammation. It was also autonomously induced under systemically reduced FcγRIIB signaling. Our results demonstrate that the silencing regulation that prevents Ig class switching crucially avoids pemphigus phenotype induction and seems to constantly operate under physiological conditions via tolerogenic molecules such as FcγRIIB.

In this study, several factors that drove class switching of Dsg3-specific B cells were experimentally elucidated. In the clinical situation, similar pathogenic conditions, such as contact dermatitis (3438), surgical procedure (3942), trauma (24), and HSV- or varicella zoster virus–induced skin inflammation (4346), have been reported to rarely associate with pemphigus onset. This clinical evidence supports the possibility that this mouse model is able to reflect some aspects of pemphigus pathophysiology, and our data imply that the extrinsic factors such as dermatitis, injury, and virus infection can trigger the pathogenic IgG autoantibody production also in patients.

However, attenuated FcγRIIB signaling is one of the crucial intrinsic factors driving class switching in general and inducing pemphigus phenotype development in AK23HkiLtg mice. In many previous studies, FcγRIIB has been reported as a pivotal molecule to inhibit IgG+ B cell expansion and Ag presentation by raising the threshold for BCR activation and to induce plasma cell apoptosis, among others, mainly related to B cell immune reactions (47, 48), whereas the detailed molecular mechanisms downstream of FcγRIIB that are linked to those immunological phenomena had not been completely elucidated. Because class switching is an immunological event that occurs after BCR activation, it is reasonable to assume that FcγRIIB finely tunes the BCR threshold and modulates Ig class switching. In this study, our data revealed that there was no difference in B cell development between Fcgr2b+/−-AK23HkiLtg mice and AK23HkiLtg mice (Supplemental Fig. 2E, 2F). This result implied that enhanced anti-Dsg3 IgG production observed in Fcgr2b+/−-AK23HkiLtg mice is not associated with abnormal B cell development.

Generally, the pivotal role of FcγRIIB in class switching has not been clearly elucidated. In fact, a previous study demonstrated no influence of FcγRIIB deficiency on the proportion of class-switched IgG+ B cells in anti-DNA BCR knock-in mice (11). In our study, Dsg3-specific IgG+ B cells were undetectable in AK23HkiLtg mice, whereas the anti-Dsg3 IgG titer spontaneously increased in the Fcgr2b−/−-AK23HkiLtg and, more important, Fcgr2b+/−-AK23HkiLtg mice. To investigate the possibility of FcγRIIB involvement in Ig class switching, we analyzed PST instead of looking at switched IgG+ cells by flow cytometry. PST is produced during class-switch recombination and used as a direct indicator of the efficiency of class switching (49, 50). In addition, it is already known that FcγRIIB regulates the proliferation of IgG+ B cells after class switching (48). Flow cytometric analysis only can evaluate the size of the IgG+ B cell bulk population and cannot distinguish the FcγRIIB contribution to class switching from its effect on IgG+ B cell proliferation after class switching. However, assessment of PST in IgM+ B cells that contain B cells under ongoing class switching at least can exclude the influence from enhanced proliferation of IgG+ B cells due to FcγRIIB deficiency. Although the molecular link between FcγRIIB signaling and class-switching inhibition in autoreactive B cells remains to be elucidated, our results with PST demonstrated that FcγRIIB deficiency enhanced Ig class switching even in a nontransgenic B cell population.

It is intriguing how FcγRIIB, which does not exhibit Ag specificity, can regulate Ag-specific autoimmunity. Thus, we analyzed the effects of FcγRIIB on anti-Dsg3 IgG (antigen-specific) and total IgG (antigen nonspecific) production. In contrast to AK23HkiLtg mice, AK23HkiLtg-Fcgr2b–deficient mice started producing high levels of anti-Dsg3 IgG Abs 6 wk after birth (Fig. 4B). However, there was no difference in the titers of total IgG Abs between Fcgr2b−/− and WT mice until 16 wk after birth (Fig. 4C). The acceleration of class switching with resultant anti-Dsg3 IgG elevation in the early phase in Fcgr2b−/−-AK23HkiLtg mice can be attributed to the autoreactivity of the knocked-in BCR in the mice. Presumably, immunological stresses such as expected autoantigen exposure to the knock-in BCR may have forced class switching, leading to early pathogenic autoantibody production in AK23HkiLtg-Fcgr2b–deficient mice but not in AK23HkiLtg mice. This could be one of the reasons why FcγRIIB deficiency is able to play pathogenic roles in Ag-specific autoimmune diseases, although FcγRIIB itself does not exhibit Ag specificity.

It also should be noted, as shown in this study, that Fcgr2b haploinsufficiency also induced spontaneous development of a pemphigus phenotype. In a previous study, G386C polymorphism in the FCGR2B gene promoter, which reduced susceptibility to autoimmune diseases by upregulating FcγRIIB, was less common in patients with pemphigus, implying the pathogenic involvement of reduced FcγRIIB signaling in pemphigus development (51, 52), consistent with the results of this study. More important, the attenuated FcγRIIB pathway was identified as one of the phenotypic alterations observed in peripheral blood B cells from patients with pemphigus. This finding obtained via unbiased analyses with clinical samples was consistent with the results from the Fcgr2b haploinsufficient pemphigus model, supporting the clinical relevance of this study. Because it was not actually shown whether the effect of Fcgr2b insufficiency on accelerated autoantibody production is B cell intrinsic or not in this system, analysis in which Fcgr2b in B cells is specifically deleted is required to answer the question. However, the results of this study implied that reduction of systemic FcγRIIB signaling could be one of the predisposing factors in pemphigus development.

AK23HkiLtg mice basically do not develop PV phenotypes in steady state up to 1 y of age. However, at a low but certain rate, AK23HkiLtg mice showed body weight loss without skin and mucosal phenotypes and died, showing a relatively low survival rate in 3 mo after birth (Fig. 4E) compared with the expected survival rate of WT mice. We observed that dying AK23HkiLtg mice tended to make high-pitched breathing sounds. It was suspected that these breathing sounds might be associated with the death of AK23HkiLtg mice, but the abnormal breathing sound, at least, is not related to anti-Dsg3 IgG production, and the cause of death remained unclear. In contrast, Fcgr2b−/+- and Fcgr2b−/−-AK23HkiLtg mice died after the manifestation of mucocutaneous phenotypes and basically did not make the abnormal breathing sound, demonstrating that Fcgr2b−/+- and Fcgr2b−/−-AK23HkiLtg mice certainly die of mucocutaneous symptoms in this pemphigus model. It might be interesting in a future study to elucidate why AK23HkiLtg mice make abnormal breathing sounds in some cases and the sounds do not develop in Fcgr2b−/+- and Fcgr2b−/−-AK23HkiLtg mice. Such a study may reveal unknown immunological roles of AK23HkiLtg B cells or AK23 IgG.

It is of interest to elucidate whether several anti-Dsg3 IgG subclasses other than IgG1 play a role in the pathophysiology of the disease models in this study. Because AK23 is originally an anti-Dsg3 IgG1 mAb (16) and IgG1 is the most predominant subclass in PV model mice (53), we mainly investigated IgG1 for elucidating pemphigus phenotypes in this study. In the AK23HkiLtg mice immunized repeatedly with rDsg3 protein to induce a pemphigus phenotype, IgG1 deposition in the palate and a PV phenotype were clearly observed (Fig. 3F), but the titer of circulating anti-Dsg3 IgG1 measured by ELISA seemed low (Fig. 3E). This finding implied that IgG subclasses other than IgG1 may also have a role in PV phenotype development in this model. Indeed, direct immunofluorescence analysis revealed that IgG2a and IgG2b subclasses in addition to IgG1 were deposited on keratinocyte cell surfaces of the Fcgr2b−/+-AK23HkiLtg mouse (Supplemental Fig. 2A–D). These findings suggested that these IgG subclasses may also contribute to the pathophysiology in this disease model. AK23HkiLtg mice might be a useful tool to elucidate which IgG subclasses are likely to be class switched by various in vivo stimuli.

Clonal ignorance is another mechanism for immunological tolerance, preventing autoreactive B cell activation (54, 55). Clonal ignorance occurs regardless of class switching, because insulin-specific B cells were maintained in a state of clonal ignorance but could exist as IgG+ B cells after class switching (56, 57). In rheumatoid factor (RF) BCR knock-in mice (AM14 sd-Tg mice), B cells remained inactive via clonal ignorance; B cells were activated and underwent class switching and secreted large amounts of IgG RF Abs in the MRL/lpr background (58). However, the importance of class switching in disease pathogenesis could not be investigated in this model, because neither RF IgG nor IgM was pathogenic for phenotype induction. In contrast, PV is a unique autoimmune disease in which only IgG but not IgM autoantibodies are pathogenic. Therefore, our model was able to specifically focus on class switching as a tolerance-targeting step, which is completely distinct from previous models.

One limitation of this study could be a high precursor frequency of autoreactive B cells in our BCR knock-in system. Because the entire B cell population in AK23HkiLtg mice express the same anti-Dsg3 BCR, the precursor frequency in the mice is very high. It is reported that high precursor frequency can attenuate immune responses by intraclonal competition (59). Therefore, the experimental condition analyzed in this study may not necessarily reflect the physiological condition of anti-Dsg3 B cells in actual patients with pemphigus. Reducing the precursor frequency by adoptively transferring the AK23HkiLtg B cells into WT mice can provide a more physiological condition reflecting actual immune responses of anti-Dsg3 B cells in patients. This limitation is worth resolving to further confirm the significance of this study.

For development of therapeutic approaches in pemphigus, it is necessary to consider regulatory mechanisms that function in anti-Dsg3 Ab production. One of the representative mechanisms that might contribute to the disease mechanism is a regulatory T (Treg) cell–mediated immune suppression (60). In humoral immunity, follicular regulatory T cells play an important role in immune regulation. A follicular regulatory T cell was discovered as a Treg cell that expresses Bcl6 and CXCR5 in secondary lymphoid tissue and regulates immune reactions in the germinal center, suppressing humoral immune responses against foreign Ags (6163). However, it has not been completely understood how Treg cells regulate humoral autoimmunity (64, 65). In a pemphigus model, it was demonstrated that Treg cells suppressed anti-Dsg3 Ab production and disease development (66). However, it had been difficult to analyze anti-Dsg3 autoimmunity associated with effector and regulatory T cells and B cells in an Ag-specific manner in the previous model. Our new BCR knock-in model enabled us to elucidate at least how Treg cells control the immune reaction of Dsg3-specific B cells at molecular levels. Furthermore, together with a transgenic system of Dsg3-specific T cells (67), it might be possible to understand details of a pemphigus pathomechanism formed by autoreactive T and B cells via Ag-specific investigation. Such fine research in the future would be desirable.

If there is an autoimmune driving force that constantly induces pathogenic class switching for harmful tissue damage and there is an opposite protective force that allows individuals to escape from the harm under physiological conditions, the latter of which might be immunological tolerance. Our results indicated that the characteristic immunoregulatory state does not allow Dsg3-specific B cells to perform pathogenic class switching and prevents pemphigus phenotype induction, and FcγRIIB was found to be one of the contributors to the maintenance in mice. This silencing regulation of autoreactive B cells observed in this study might be considered as a crucial immunoregulatory state to suppress disease progression in pemphigus, a rare disease in which IgM to IgG class switch is indispensable for disease development.

We are grateful to Drs. Antonio Iglesias and Masayuki Saijo for providing the anti-MOG Ig H chain knock-in construct and the vaccinia virus, respectively; Drs. Sayuri Chiba and Aiko Shiohama for their guidance in generating the DNA constructs; Dr. Hidehiro Fukuyama for fruitful scientific advice; Kyoko Hidaka for animal care and genotyping of the mice; Minae Suzuki and Hiroyo Koike for the preparation of cryosections; and Mariko Okajima for laboratory management.

This work was supported by Grants-in-Aid for Scientific Research (B) (11470089, 13557026) from the Japan Society for the Promotion of Science to S.K.; a Grant-in-Aid for Scientific Research (C) (22590437) to H.F.; a Grant-in-Aid for Scientific Research (S) (21229014) to M.A.; Grants-in-Aid for Scientific Research (A) (26253065) to M.A. and (19H01051) to H.T. from the Japan Society for the Promotion of Science; a Grant-in-Aid for Scientific Research on Innovative Areas (18073015) to S.K. from the Ministry of Education, Culture, Sports, Science and Technology, Japan; Health and Labor Sciences Research Grants for Research on Allergic Disease and Immunology (H20-Immunology-General-006) to S.K. from the Ministry of Health, Labour and Welfare of Japan; and Keio University Gijuku Academic Development Funds to H.F.

The online version of this article contains supplemental material.

Abbreviations used in this article

BM

bone marrow

Cy7

cyanine 7

DNFB

2,4-dinitro-1-fluorobenzene

Dsg3

desmoglein 3

ES

embryonic stem

ki

knock-in

MOG

myelin oligodendrocyte glycoprotein

PST

postswitch transcript

PV

pemphigus vulgaris

rDsg3

recombinant desmoglein 3

RF

rheumatoid factor

sLN

skin-draining lymph node

tg

transgenic

Treg

regulatory T

WT

wild type

1.
Nemazee
D. A.
,
K.
Bürki
.
1989
.
Clonal deletion of B lymphocytes in a transgenic mouse bearing anti-MHC class I antibody genes.
Nature
337
:
562
566
.
2.
Goodnow
C. C. J.
,
J.
Crosbie
,
S.
Adelstein
,
T. B.
Lavoie
,
S. J.
Smith-Gill
,
R. A.
Brink
,
H.
Pritchard-Briscoe
,
J. S.
Wotherspoon
,
R. H.
Loblay
,
K.
Raphael
, et al
1988
.
Altered immunoglobulin expression and functional silencing of self-reactive B lymphocytes in transgenic mice.
Nature
334
:
676
682
.
3.
Gay
D.
,
T.
Saunders
,
S.
Camper
,
M.
Weigert
.
1993
.
Receptor editing: an approach by autoreactive B cells to escape tolerance.
J. Exp. Med.
177
:
999
1008
.
4.
Tiegs
S. L.
,
D. M.
Russell
,
D.
Nemazee
.
1993
.
Receptor editing in self-reactive bone marrow B cells.
J. Exp. Med.
177
:
1009
1020
.
5.
Nossal
G. J.
,
B. L.
Pike
.
1980
.
Clonal anergy: persistence in tolerant mice of antigen-binding B lymphocytes incapable of responding to antigen or mitogen.
Proc. Natl. Acad. Sci. USA
77
:
1602
1606
.
6.
Cooke
M. P.
,
A. W.
Heath
,
K. M.
Shokat
,
Y.
Zeng
,
F. D.
Finkelman
,
P. S.
Linsley
,
M.
Howard
,
C. C.
Goodnow
.
1994
.
Immunoglobulin signal transduction guides the specificity of B cell-T cell interactions and is blocked in tolerant self-reactive B cells.
J. Exp. Med.
179
:
425
438
.
7.
Nemazee
D.
2006
.
Receptor editing in lymphocyte development and central tolerance.
Nat. Rev. Immunol.
6
:
728
740
.
8.
Xu
Z.
,
H.
Zan
,
E. J.
Pone
,
T.
Mai
,
P.
Casali
.
2012
.
Immunoglobulin class-switch DNA recombination: induction, targeting and beyond.
Nat. Rev. Immunol.
12
:
517
531
.
9.
Kurosaki
T.
,
K.
Kometani
,
W.
Ise
.
2015
.
Memory B cells.
Nat. Rev. Immunol.
15
:
149
159
.
10.
Nutt
S. L.
,
P. D.
Hodgkin
,
D. M.
Tarlinton
,
L. M.
Corcoran
.
2015
.
The generation of antibody-secreting plasma cells.
Nat. Rev. Immunol.
15
:
160
171
.
11.
Fukuyama
H.
,
F.
Nimmerjahn
,
J. V.
Ravetch
.
2005
.
The inhibitory Fcγ receptor modulates autoimmunity by limiting the accumulation of immunoglobulin G+ anti-DNA plasma cells.
Nat. Immunol.
6
:
99
106
.
12.
Amagai
M.
,
V.
Klaus-Kovtun
,
J. R.
Stanley
.
1991
.
Autoantibodies against a novel epithelial cadherin in pemphigus vulgaris, a disease of cell adhesion.
Cell
67
:
869
877
.
13.
Stanley
J. R.
,
M.
Amagai
.
2006
.
Pemphigus, bullous impetigo, and the staphylococcal scalded-skin syndrome.
N. Engl. J. Med.
355
:
1800
1810
.
14.
Amagai
M.
,
A.
Komai
,
T.
Hashimoto
,
Y.
Shirakata
,
K.
Hashimoto
,
T.
Yamada
,
Y.
Kitajima
,
K.
Ohya
,
H.
Iwanami
,
T.
Nishikawa
.
1999
.
Usefulness of enzyme-linked immunosorbent assay using recombinant desmogleins 1 and 3 for serodiagnosis of pemphigus.
Br. J. Dermatol.
140
:
351
357
.
15.
Amagai
M.
,
K.
Tsunoda
,
H.
Suzuki
,
K.
Nishifuji
,
S.
Koyasu
,
T.
Nishikawa
.
2000
.
Use of autoantigen-knockout mice in developing an active autoimmune disease model for pemphigus.
J. Clin. Invest.
105
:
625
631
.
16.
Tsunoda
K.
,
T.
Ota
,
M.
Aoki
,
T.
Yamada
,
T.
Nagai
,
T.
Nakagawa
,
S.
Koyasu
,
T.
Nishikawa
,
M.
Amagai
.
2003
.
Induction of pemphigus phenotype by a mouse monoclonal antibody against the amino-terminal adhesive interface of desmoglein 3.
J. Immunol.
170
:
2170
2178
.
17.
Tsunoda
K.
,
T.
Ota
,
M.
Saito
,
T.
Hata
,
A.
Shimizu
,
A.
Ishiko
,
T.
Yamada
,
T.
Nakagawa
,
A. P.
Kowalczyk
,
M.
Amagai
.
2011
.
Pathogenic relevance of IgG and IgM antibodies against desmoglein 3 in blister formation in pemphigus vulgaris.
Am. J. Pathol.
179
:
795
806
.
18.
Litzenburger
T.
,
R.
Fässler
,
J.
Bauer
,
H.
Lassmann
,
C.
Linington
,
H.
Wekerle
,
A.
Iglesias
.
1998
.
B lymphocytes producing demyelinating autoantibodies: development and function in gene-targeted transgenic mice.
J. Exp. Med.
188
:
169
180
.
19.
Krämer
A.
,
J.
Green
,
J.
Pollard
Jr.
,
S.
Tugendreich
.
2014
.
Causal analysis approaches in Ingenuity Pathway Analysis.
Bioinformatics
30
:
523
530
.
20.
Lang
J.
,
D.
Nemazee
.
2000
.
B cell clonal elimination induced by membrane-bound self-antigen may require repeated antigen encounter or cell competition.
Eur. J. Immunol.
30
:
689
696
.
21.
Li
Y. S.
,
K.
Hayakawa
,
R. R.
Hardy
.
1993
.
The regulated expression of B lineage associated genes during B cell differentiation in bone marrow and fetal liver.
J. Exp. Med.
178
:
951
960
.
22.
Allman
D.
,
S.
Pillai
.
2008
.
Peripheral B cell subsets.
Curr. Opin. Immunol.
20
:
149
157
.
23.
Rigante
D.
,
M. B.
Mazzoni
,
S.
Esposito
.
2014
.
The cryptic interplay between systemic lupus erythematosus and infections.
Autoimmun. Rev.
13
:
96
102
.
24.
Daneshpazhooh
M.
,
M.
Fatehnejad
,
Z.
Rahbar
,
K.
Balighi
,
N.
Ghandi
,
M.
Ghiasi
,
R.
Abedini
,
V.
Lajevardi
,
C.
Chams-Davatchi
.
2016
.
Trauma-induced pemphigus: a case series of 36 patients.
J. Dtsch. Dermatol. Ges.
14
:
166
171
.
25.
Kyogoku
C.
,
H. M.
Dijstelbloem
,
N.
Tsuchiya
,
Y.
Hatta
,
H.
Kato
,
A.
Yamaguchi
,
T.
Fukazawa
,
M. D.
Jansen
,
H.
Hashimoto
,
J. G.
van de Winkel
, et al
2002
.
Fcγ receptor gene polymorphisms in Japanese patients with systemic lupus erythematosus: contribution of FCGR2B to genetic susceptibility.
Arthritis Rheum.
46
:
1242
1254
.
26.
Willcocks
L. C.
,
E. J.
Carr
,
H. A.
Niederer
,
T. F.
Rayner
,
T. N.
Williams
,
W.
Yang
,
J. A.
Scott
,
B. C.
Urban
,
N.
Peshu
,
T. J.
Vyse
, et al
2010
.
A defunctioning polymorphism in FCGR2B is associated with protection against malaria but susceptibility to systemic lupus erythematosus.
Proc. Natl. Acad. Sci. USA
107
:
7881
7885
.
27.
Radstake
T. R.
,
B.
Franke
,
M. H.
Wenink
,
K. C.
Nabbe
,
M. J.
Coenen
,
P.
Welsing
,
E.
Bonvini
,
S.
Koenig
,
W. B.
van den Berg
,
P.
Barrera
,
P. L.
van Riel
.
2006
.
The functional variant of the inhibitory Fcγ receptor IIb (CD32B) is associated with the rate of radiologic joint damage and dendritic cell function in rheumatoid arthritis.
Arthritis Rheum.
54
:
3828
3837
.
28.
Zhou
X. J.
,
J. C.
Lv
,
D. F.
Bu
,
L.
Yu
,
Y. R.
Yang
,
J.
Zhao
,
Z.
Cui
,
R.
Yang
,
M. H.
Zhao
,
H.
Zhang
.
2010
.
Copy number variation of FCGR3A rather than FCGR3B and FCGR2B is associated with susceptibility to anti-GBM disease.
Int. Immunol.
22
:
45
51
.
29.
Bruin
M.
,
M.
Bierings
,
C.
Uiterwaal
,
T.
Révész
,
L.
Bode
,
M. E.
Wiesman
,
T.
Kuijpers
,
R.
Tamminga
,
M.
de Haas
.
2004
.
Platelet count, previous infection and FCGR2B genotype predict development of chronic disease in newly diagnosed idiopathic thrombocytopenia in childhood: results of a prospective study.
Br. J. Haematol.
127
:
561
567
.
30.
Tridandapani
S.
,
H.
Phee
,
L.
Shivakumar
,
T. W.
Kelley
,
K. M.
Coggeshall
.
1998
.
Role of SHIP in FcγRIIb-mediated inhibition of Ras activation in B cells.
Mol. Immunol.
35
:
1135
1146
.
31.
Enders
A.
,
P.
Bouillet
,
H.
Puthalakath
,
Y.
Xu
,
D. M.
Tarlinton
,
A.
Strasser
.
2003
.
Loss of the pro-apoptotic BH3-only Bcl-2 family member Bim inhibits BCR stimulation-induced apoptosis and deletion of autoreactive B cells.
J. Exp. Med.
198
:
1119
1126
.
32.
Oliver
P. M.
,
T.
Vass
,
J.
Kappler
,
P.
Marrack
.
2006
.
Loss of the proapoptotic protein, Bim, breaks B cell anergy.
J. Exp. Med.
203
:
731
741
.
33.
Hayashi
K.
,
T.
Nojima
,
R.
Goitsuka
,
D.
Kitamura
.
2004
.
Impaired receptor editing in the primary B cell repertoire of BASH-deficient mice.
J. Immunol.
173
:
5980
5988
.
34.
Brenner
S.
,
R.
Wolf
,
V.
Ruocco
.
1994
.
Contact pemphigus: a subgroup of induced pemphigus.
Int. J. Dermatol.
33
:
843
845
.
35.
Lynfield
Y. L.
,
L. P.
Pertschuk
,
A.
Zimmerman
.
1973
.
Pemphigus erythematosus provoked by allergic contact dermatitis. Occurrence many years after thymoma removal.
Arch. Dermatol.
108
:
690
693
.
36.
Tsankov
N.
,
J.
Kazandjieva
,
M.
Gantcheva
.
1998
.
Contact pemphigus induced by dihydrodiphenyltrichlorethane.
Eur. J. Dermatol.
8
:
442
443
.
37.
Vozza
A.
,
V.
Ruocco
,
S.
Brenner
,
R.
Wolf
.
1996
.
Contact pemphigus.
Int. J. Dermatol.
35
:
199
201
.
38.
Ruocco
V.
,
M.
Pisani
.
1982
.
Induced pemphigus.
Arch. Dermatol. Res.
274
:
123
140
.
39.
Duick
M. G.
,
B.
Zaks
,
R. L.
Moy
,
R. P.
Kaplan
.
2001
.
Mohs micrographic surgery-induced pemphigus.
Dermatol. Surg.
27
:
895
897
.
40.
Mehregan
D. R.
,
R. K.
Roenigk
,
L. E.
Gibson
.
1992
.
Postsurgical pemphigus.
Arch. Dermatol.
128
:
414
415
.
41.
Micali
G.
,
M. R.
Nasca
,
M. L.
Musumeci
,
D.
Innocenzi
.
1998
.
Postsurgical pemphigus.
Dermatology
197
:
192
193
.
42.
Rotunda
A. M.
,
A. R.
Bhupathy
,
R.
Dye
,
T. T.
Soriano
.
2005
.
Pemphigus foliaceus masquerading as postoperative wound infection: report of a case and review of the Koebner and related phenomenon following surgical procedures.
Dermatol. Surg.
31
:
226
231
.
43.
Goon
A. T.
,
Y. K.
Tay
,
S. H.
Tan
.
2001
.
Pemphigus vulgaris following varicella infection.
Clin. Exp. Dermatol.
26
:
661
663
.
44.
Horiguchi
Y.
,
S.
Fujii
,
S.
Takase
,
R.
Mori
.
2005
.
Case of pemphigus vulgaris in oral mucosa showing an extension to the body surface with recurrent varicella.
Int. J. Dermatol.
44
:
951
954
.
45.
Keane
J. T.
,
F. D.
Malkinson
,
J.
Bryant
,
S.
Levin
.
1976
.
Herpesvirus hominis hepatitis and disseminated intravascular coagulation. Occurrence in an adult with pemphigus vulgaris.
Arch. Intern. Med.
136
:
1312
1317
.
46.
Ruocco
E.
,
V.
Ruocco
,
A.
Lo Schiavo
,
G.
Brunetti
,
R.
Wolf
.
2014
.
Viruses and pemphigus: an intriguing never-ending story.
Dermatology
229
:
310
315
.
47.
Smith
K. G.
,
M. R.
Clatworthy
.
2010
.
FcγRIIB in autoimmunity and infection: evolutionary and therapeutic implications. [Published erratum appears in 2010 Nat. Rev. Immunol. 10: 674.]
Nat. Rev. Immunol.
10
:
328
343
.
48.
Nimmerjahn
F.
,
J. V.
Ravetch
.
2008
.
Fcγ receptors as regulators of immune responses.
Nat. Rev. Immunol.
8
:
34
47
.
49.
Lee
C. G.
,
K.
Kinoshita
,
A.
Arudchandran
,
S. M.
Cerritelli
,
R. J.
Crouch
,
T.
Honjo
.
2001
.
Quantitative regulation of class switch recombination by switch region transcription.
J. Exp. Med.
194
:
365
374
.
50.
Tashiro
J.
,
K.
Kinoshita
,
T.
Honjo
.
2001
.
Palindromic but not G-rich sequences are targets of class switch recombination.
Int. Immunol.
13
:
495
505
.
51.
Recke
A.
,
G.
Vidarsson
,
R. J.
Ludwig
,
M.
Freitag
,
S.
Möller
,
R.
Vonthein
,
J.
Schellenberger
,
O.
Haase
,
S.
Görg
,
A.
Nebel
, et al
German AIBD Genetic Study Group
.
2015
.
Allelic and copy-number variations of FcγRs affect granulocyte function and susceptibility for autoimmune blistering diseases.
J. Autoimmun.
61
:
36
44
.
52.
Su
K.
,
J.
Wu
,
J. C.
Edberg
,
X.
Li
,
P.
Ferguson
,
G. S.
Cooper
,
C. D.
Langefeld
,
R. P.
Kimberly
.
2004
.
A promoter haplotype of the immunoreceptor tyrosine-based inhibitory motif-bearing FcγRIIb alters receptor expression and associates with autoimmunity. I. Regulatory FCGR2B polymorphisms and their association with systemic lupus erythematosus.
J. Immunol.
172
:
7186
7191
.
53.
Ohyama
M.
,
M.
Amagai
,
K.
Tsunoda
,
T.
Ota
,
S.
Koyasu
,
J.
Hata
,
A.
Umezawa
,
T.
Nishikawa
.
2002
.
Immunologic and histopathologic characterization of an active disease mouse model for pemphigus vulgaris.
J. Invest. Dermatol.
118
:
199
204
.
54.
Hannum
L. G.
,
D.
Ni
,
A. M.
Haberman
,
M. G.
Weigert
,
M. J.
Shlomchik
.
1996
.
A disease-related rheumatoid factor autoantibody is not tolerized in a normal mouse: implications for the origins of autoantibodies in autoimmune disease.
J. Exp. Med.
184
:
1269
1278
.
55.
Wang
H.
,
M. J.
Shlomchik
.
1999
.
Autoantigen-specific B cell activation in Fas-deficient rheumatoid factor immunoglobulin transgenic mice.
J. Exp. Med.
190
:
639
649
.
56.
Thomas
J. W.
,
C.
Hulbert
.
1996
.
Somatically mutated B cell pool provides precursors for insulin antibodies.
J. Immunol.
157
:
763
771
.
57.
Goodnow
C. C.
1992
.
Transgenic mice and analysis of B-cell tolerance.
Annu. Rev. Immunol.
10
:
489
518
.
58.
Sweet
R. A.
,
S. R.
Christensen
,
M. L.
Harris
,
J.
Shupe
,
J. L.
Sutherland
,
M. J.
Shlomchik
.
2010
.
A new site-directed transgenic rheumatoid factor mouse model demonstrates extrafollicular class switch and plasmablast formation.
Autoimmunity
43
:
607
618
.
59.
Le
T. V.
,
T. H.
Kim
,
D. D.
Chaplin
.
2008
.
Intraclonal competition inhibits the formation of high-affinity antibody-secreting cells.
J. Immunol.
181
:
6027
6037
.
60.
Sakaguchi
S.
,
N.
Mikami
,
J. B.
Wing
,
A.
Tanaka
,
K.
Ichiyama
,
N.
Ohkura
.
2020
.
Regulatory T cells and human disease.
Annu. Rev. Immunol.
38
:
541
566
.
61.
Linterman
M. A.
,
W.
Pierson
,
S. K.
Lee
,
A.
Kallies
,
S.
Kawamoto
,
T. F.
Rayner
,
M.
Srivastava
,
D. P.
Divekar
,
L.
Beaton
,
J. J.
Hogan
, et al
2011
.
Foxp3+ follicular regulatory T cells control the germinal center response.
Nat. Med.
17
:
975
982
.
62.
Chung
Y.
,
S.
Tanaka
,
F.
Chu
,
R. I.
Nurieva
,
G. J.
Martinez
,
S.
Rawal
,
Y. H.
Wang
,
H.
Lim
,
J. M.
Reynolds
,
X. H.
Zhou
, et al
2011
.
Follicular regulatory T cells expressing Foxp3 and Bcl-6 suppress germinal center reactions.
Nat. Med.
17
:
983
988
.
63.
Wollenberg
I.
,
A.
Agua-Doce
,
A.
Hernández
,
C.
Almeida
,
V. G.
Oliveira
,
J.
Faro
,
L.
Graca
.
2011
.
Regulation of the germinal center reaction by Foxp3+ follicular regulatory T cells.
J. Immunol.
187
:
4553
4560
.
64.
Clement
R. L.
,
J.
Daccache
,
M. T.
Mohammed
,
A.
Diallo
,
B. R.
Blazar
,
V. K.
Kuchroo
,
S. B.
Lovitch
,
A. H.
Sharpe
,
P. T.
Sage
.
2019
.
Follicular regulatory T cells control humoral and allergic immunity by restraining early B cell responses.
Nat. Immunol.
20
:
1360
1371
.
65.
Fu
W.
,
X.
Liu
,
X.
Lin
,
H.
Feng
,
L.
Sun
,
S.
Li
,
H.
Chen
,
H.
Tang
,
L.
Lu
,
W.
Jin
,
C.
Dong
.
2018
.
Deficiency in T follicular regulatory cells promotes autoimmunity.
J. Exp. Med.
215
:
815
825
.
66.
Yokoyama
T.
,
S.
Matsuda
,
Y.
Takae
,
N.
Wada
,
T.
Nishikawa
,
M.
Amagai
,
S.
Koyasu
.
2011
.
Antigen-independent development of Foxp3+ regulatory T cells suppressing autoantibody production in experimental pemphigus vulgaris.
Int. Immunol.
23
:
365
373
.
67.
Takahashi
H.
,
M.
Kouno
,
K.
Nagao
,
N.
Wada
,
T.
Hata
,
S.
Nishimoto
,
Y.
Iwakura
,
A.
Yoshimura
,
T.
Yamada
,
M.
Kuwana
, et al
2011
.
Desmoglein 3-specific CD4+ T cells induce pemphigus vulgaris and interface dermatitis in mice.
J. Clin. Invest.
121
:
3677
3688
.

The authors have no financial conflicts of interest.

Supplementary data