Regulatory B cells suppress imiquimod-induced, psoriasis-like skin inflammation

Interleukin-10-producing CD1dhiCD5+ regulatory B cells play regulatory roles in mouse imiquimod-induced psoriasislike skin inflammation, where adoptively transferred B10 cells ameliorate this condition.

Abstract

Psoriasis is an inflammatory cutaneous disorder characterized by marked epidermal thickening and Th1 and Th17 cell infiltration. At present, the contribution of B cells to the pathogenesis of psoriasis is unclear. In mice, topical application of imiquimod induces inflamed skin lesions and serves as an experimental animal model for human psoriasis. In this study, we showed that imiquimod-induced skin inflammation was more severe in CD19^−/− than WT mice. These inflammatory responses were negatively regulated by a unique IL-10-producing CD1d^hiCD5⁺ regulatory B cell subset (B10 cells) that was absent in CD19^−/− mice and represented only 1–2% of splenic B220⁺ cells in WT mice. Splenic B10 cells entered the circulation and migrated to draining LNs during imiquimod-induced skin inflammation, thereby suppressing IFN-γ and IL-17 production. Furthermore, adoptive transfer of these B10 cells from WT mice reduced inflammation in CD19^−/− mice. The present findings provide direct evidence that B10 cells regulate imiquimod-induced skin inflammation and offer insights into regulatory B cell-based therapies for the treatment of psoriasis.

Introduction

Psoriasis is a cutaneous disorder characterized by widespread erythematous plaques with adherent scales that affect ∼2% of the general population [1]. Although its pathogenesis remains unclear, recent studies have revealed that Th17 and Th1 cells play critical roles in disease development. Histologically, psoriasis is characterized by marked thickening of the epidermis with an inflammatory infiltrate predominantly composed of Th17 and Th1 cells [2]. These infiltrated cells stimulate keratinocytes to produce cytokines, which further amplify the inflammatory response [3].

The application of imiquimod occasionally leads to the development of psoriasis in humans [4, 5]. Imiquimod is a potent agonist for TLR7 and -8 and has been used therapeutically in the treatment of actinic keratosis, genital warts, and a variety of other skin disorders [6]. In mice, the daily topical application of imiquimod for 6 days previously induced the appearance of inflamed skin lesions resembling human psoriasis [7]. This experimental method has therefore served as a convenient animal model of the disease.

B cells not only play a central role in humoral immune responses but also regulate CD4⁺ T cell responses to foreign and self-antigens [8, 9], function as APCs [10], produce cytokines [11], provide costimulatory signals [12], and promote naïve CD4⁺ T cell differentiation into Th1 or Th2 subsets [11]. B cells and specific B cell subsets can negatively regulate immune responses in mice, demonstrating the existence of regulatory B cells [13]. A specific subset of regulatory B cells with a CD1d^hiCD5⁺ phenotype has been identified recently in mice and shown to regulate contact hypersensitivity responses, an EAE model of multiple sclerosis and a DSS-induced colitis model of ulcerative colitis, in an IL-10-dependent manner [14–16]. As multiple regulatory B cell subsets are likely to exist, this subset was named B10 cells to distinguish it from other possible regulatory B cell subsets. B10 cells appear to produce only IL-10 and are responsible for most B cell IL-10 production [17]. B10 cell regulatory functions are antigen-restricted in vivo, and the adoptive transfer of antigen-primed B10 cells reduces inflammation during contact hypersensitivity responses and ameliorates the severity of EAE and DSS-induced colitis [14–16].

Treatment with human rIL-10 reduces the severity of psoriasis in humans [18, 19]. Furthermore, B cell depletion in humans, using the chimeric human anti-CD20 mAb rituximab, resulted in the development of psoriasis [20, 21], suggesting that B cells and IL-10 play important inhibitory roles in this development. The phenotypically unique regulatory B10 cell subset exists within the spleen of naïve WT mice at 1–2% of the total B cell count, whereas CD19^−/− mice have few, if any, B10 cells [14]. Therefore, in the present study, we examined the importance of regulatory B10 cells in an imiquimod-induced psoriasis model in CD19^−/− and WT mice.

MATERIALS AND METHODS

Animals

WT C57BL/6 and IL-10^−/− (B6.129P2-Il10^tm1Cgn/J) mice were purchased from The Jackson Laboratory (Bar Harbor, ME, USA). CD19^−/− (C57BL/6×129) mice were generated as described [22] and backcrossed seven to 12 generations onto the C57BL/6 background before use in this study. All mice were bred in a specific pathogen-free barrier facility and used at 8–12 weeks of age. All studies were approved by the Committee on Animal Experimentation of The University of Tokyo (Japan).

Induction and evaluation of imiquimod-induced skin inflammation

Imiquimod-induced skin inflammation was induced as described previously [7]. In brief, mice received a daily topical dose of 62.5 mg imiquimod cream (5%; Mochida Pharmaceuticals, Tokyo, Japan) on a shaved back for 6 consecutive days. Control mice were treated similarly with a control vehicle cream.

To score the severity of inflammation of the back skin, an objective scoring system was developed based on the clinical Psoriasis Area and Severity Index, except that for the mouse model, the affected skin area was not taken into account in the overall score, as described previously [7]. Erythema, scaling, and thickening were scored independently on a scale from 0 to 4, with 0 representing none; 1, slight; 2, moderate; 3, marked; and 4, very marked. The cumulative score (erythema plus scaling plus thickening) served as a measure of the severity of inflammation (scale 0–12). In some experiments, mice were treated with the IL-10R (1B1.3a; BioLegend, San Diego, CA, USA) or control mAb (250 μg) on Days 0 and 3 after the induction of imiquimod-induced skin inflammation.

Histological analysis

The mice were sacrificed 6 days after the induction of imiquimod-induced skin inflammation. Skin samples were removed, and segments were fixed in 10% buffered formalin. After paraffin embedding, 5 μm-thick sections were cut and stained with H&E. For immunohistochemistry, paraffin-embedded tissues were cut into 6-μm sections, deparaffinized in xylene, and dehydrated in PBS. Deparaffinized sections were treated with endogenous peroxidase-blocking reagent (Dako Cytomation A/S, Copenhagen, Denmark) and proteinase K (Dako Cytomation A/S) for 6 min at room temperature. Sections were then incubated with rat mAb specific for mouse CD4 (#2H9; Relia Tech, Wolfenbutel, Germany), CD8 (D-9; Santa Cruz Biotechnology, Santa Cruz, CA, USA), and B220 (RA3-6B2; BD PharMingen, San Diego, CA, USA). Rat IgG (Southern Biotechnology Associates, Birmingham, AL, USA) was used as a control for nonspecific staining. Sections were then incubated sequentially (20 min at 37°C) with a biotinylated rabbit anti-rat IgG secondary antibody and then a HRP-conjugated avidin-biotin complex (Vectastain ABC kit; Vector Laboratories, Burlingame, CA, USA). Sections were developed with 3,3′-diaminobenzidine tetrahydrochloride and hydrogen peroxide and counterstained with methyl green. Stained cells were counted in 10 random grids under high-magnification (×400) power fields of a light microscope. Each section was examined independently by two investigators in a blinded manner.

Cell isolation and B cell purification

Single-cell suspensions of spleen and draining LNs (pooled bilateral axial and inguinal LNs) were generated by gentle dissection. PBMCs were isolated from heparinized blood after centrifugation over a discontinuous Lymphoprep (Axis-Shield PoC As, Oslo, Norway) gradient. B220 mAb-coated microbeads (Miltenyi Biotec, Auburn, CA, USA) were used to purify B cells by positive selection, according to the manufacturer's instructions. When necessary, the cells were enriched a second time using a fresh MACS column to obtain >95% B220⁺ cell purity.

RNA isolation and real-time RT-PCR

Total RNA was isolated from skin specimens, spleen, and draining LN suspensions with RNeasy spin columns (Qiagen, Crawley, UK). Total RNA from each sample was reverse-transcribed into cDNA. Expression of IL-17A and IFN-γ was analyzed using a real-time PCR quantification method, according to the manufacturer's instructions (Applied Biosystems, Foster City, CA, USA).

Antibodies and immunofluorescence analysis

Anti-mouse mAb, with specificities against B220 (RA3-6B2), CD19 (1D3), CD5 (53-7.3), CD1d (1B1), CD4 (H129.19), and CD8 (53-6.7), were obtained from BD PharMingen. Intracellular staining used mAb reactive with FoxP3 (FJK-16s; eBioscience, San Diego, CA, USA) and the Cytofix/Cytoperm kit (BD PharMingen). Single-cell leukocyte suspensions were stained on ice using predetermined optimal concentrations of each antibody for 20–60 min and fixed as described [23]. Cells with the light-scatter properties of lymphocytes were analyzed by immunofluorescence staining and a FACSVerse flow cytometer (Becton Dickinson, San Jose, CA, USA). Background staining was determined using unreactive isotype-matched control mAb (Caltag Laboratories, San Francisco, CA, USA) with gates positioned to exclude ≥98% of unreactive cells.

Flow cytometry analysis of intracellular IL-10 synthesis

Intracellular IL-10 analysis by flow cytometry was as described [14]. Briefly, isolated leukocytes or purified cells were resuspended (2×10⁶ cells/ml) in complete medium (RPMI 1640 containing 10% FCS, 200 μg/ml penicillin, 200 U/ml streptomycin, 4 mM L-glutamine, and 5×10⁻⁵ M 2-ME; all from Invitrogen, Carlsbad, CA, USA) in the presence of LPS (10 μμg/ml; Sigma, St. Louis, MO, USA), PMA (50 ng/ml; Sigma), ionomycin (500 ng/ml; Sigma), and monensin (2 μM; eBioscience) in 24-well, flat-bottom plates for 5 h at 37°C. Anti-mouse IL-10 mAb (JES5-16E3) was obtained from eBioscience. Cells were fixed and permeabilized using the Cytofix/Cytoperm kit. Permeabilized cells were stained with IL-10 mAb.

Analysis of Th cell differentiation

Isolated leukocytes were activated by plate-bound anti-mouse CD3 (2 μg/ml; 145-2C11; BD PharMingen) and soluble anti-mouse CD28 mAb (1 μg/ml; 37.51; BD PharMingen) and cultured in the presence of murine IL-6 (50 ng/ml; R&D Systems, Minneapolis, MN, USA) and human TGF-β1 (5 ng/ml; R&D Systems) with murine IL-23 (10 ng/ml; R&D Systems) for Th17 cell and IL-12 (1 ng/ml; R&D Systems) for Th1 cell differentiation [24]. After 2 days in culture, cells were restimulated with PMA (50 ng/ml), ionomycin (500 ng/ml), and monensin (2 μM) for 4 h. Cells were fixed and permeabilized using the Cytofix/Cytoperm kit. Detection of IL-17- and IFN-γ-producing cells was determined by intracellular cytokine staining with anti-mouse IL-17A (eBio17B7; eBioscience) or anti-IFN-γ mAb (XMG1.2; eBioscience).

Cell sorting and adoptive transfers

Splenic CD1d^hiCD5⁺ B cells were selected using a FACSAria flow cytometer (Becton Dickinson) with purities of 85–95%. After isolation, 2 × 10⁶ CD1d^hiCD5⁺ or 2 × 10⁶ non-CD1d^hiCD5⁺ B cells were transferred by i.v. injection into CD19^−/− mice, 2 days before or 2 days after the induction of imiquimod-induced skin inflammation.

Statistical analysis

All data are expressed as mean ± sem values. The Mann-Whitney U-test was used to determine the level of significance of differences in sample means, and the Bonferroni test was used for multiple comparisons.

RESULTS

Increased severity of imiquimod-induced skin inflammation in CD19^−/− mice

To assess whether CD19 expression played a role in the pathogenesis of imiquimod-induced skin inflammation, we treated CD19^−/− and WT mice with imiquimod cream for 6 days and quantitatively evaluated the severity of skin inflammation by scoring erythema, scaling, and thickening. A typical example on Day 6 is shown in Fig. 1A. Skin inflammation was first observed in imiquimod-treated WT mice and imiquimod-treated CD19^−/− mice on Day 2 (Fig. 1B). Imquimod-treated CD19^−/− mice had significantly more severe erythema and scaling than imiquimod-treated WT mice from Day 5 to 6, whereas thickening was significantly more severe in imiquimod-treated CD19^−/− mice than imiquimod-treated WT mice from Day 3 to 6. Total scores were significantly higher in imiquimod-treated CD19^−/− mice from Day 2 to 6, showing that they are more susceptible to imiquimod-induced skin inflammation.

Figure 1. Increased severity of imiquimod-induced skin inflammation in CD19^−/− mice.

The shaved back skin of WT and CD19^−/− mice was treated daily with imiquimod cream or control cream. (A) Representative phenotypical presentation of mouse back skin after 6 days of imiquimod treatment. (B) The severity of imiquimod-induced skin inflammation. Erythema, scaling, and thickness of back skin were scored daily from 0 to 4. The total score was calculated by summing individual scores for erythema, scales, and thickness in each mouse. Values represent means (±sem) from greater than/equal to five mice/group. Significant differences between sample means are indicated: *P < 0.05; **P < 0.01.

To further evaluate disease severity, the degree of skin inflammation was also assessed histopathologically. Following the 6-day period of imiquimod treatment, skin samples were harvested for histopathologic evaluation. Imiquimod treatment induced hyperkeratosis, parakeratosis, acanthosis, spongiosis, and elongation of the rete ridges, which are typical histopathological findings of human psoriasis (Fig. 2A). Although these findings were seen in both groups, they were more severe in CD19^−/− mice. Imiquimod treatment significantly increased CD4⁺ and CD8⁺ T cell numbers in both groups (Fig. 2B), and the numbers of these cells were significantly lower in WT mice treated with imiquimod than in CD19^−/− mice treated with imiquimod (P<0.05). There were no significant differences in the number of B cells between the two groups. These results show that skin inflammation was more severe in CD19^−/− than WT mice.

Figure 2. CD19^−/− enhanced the severity of imiquimod-induced skin inflammation.

Skin sections were harvested from WT and CD19^−/− mice after 6 days of imiquimod treatment. (A) Imiquimod treatment induced hyperkeratosis (white asterisks), parakeratosis (black arrow), acanthosis (black asterisks), spongiosis (green arrows), and elongation of the rete ridges (white arrows). Representative skin sections from WT and CD19^−/− mice stained with H&E. Original magnifications, ×200. (B) The numbers of CD4⁺ T cells, CD8⁺ T cells, and B220⁺ B cells/field of view (×400) were counted. Values represent means (±sem) from greater than/equal to four mice/group. Significant differences between sample means are indicated: *P < 0.05; **P < 0.01**.

Imiquimod treatment reduces the number of splenic B cells

To determine whether imiquimod treatment altered the populations of T cells and B cells, the numbers of CD4⁺, CD8⁺, and B220⁺ cells in the spleen and draining LNs were assessed on Day 6 by flow cytometry. The numbers of CD4⁺ and CD8⁺ T cells in the spleen did not change during imiquimod-induced skin inflammation in WT or CD19^−/− mice (Fig. 3A). Although imiquimod treatment did not affect the numbers of CD4⁺ and CD8⁺ T cells in the draining LNs in WT mice, these cells were increased significantly in the draining LNs of imiquimod-treated CD19^−/− mice compared with control-treated CD19^−/− mice (Fig. 3B). WT mice treated with imiquimod had significantly reduced numbers of B cells in the spleen relative to control-treated WT mice (P<0.05). B cell numbers in the draining LNs were increased in imiquimod-treated WT mice compared with control-treated WT mice, although the differences were not statistically significant. By contrast, imiquimod treatment did not affect B cell numbers in the spleen or draining LNs of CD19^−/− mice. Thus, imiquimod treatment diminished the numbers of B cells in the spleen of WT mice.

Figure 3. The effect of CD19^−/− on the numbers of CD4⁺ T cells, CD8⁺ T cells, and B220⁺ cells in the spleen and draining LNs during imiquimod-induced skin inflammation.

The numbers of CD4⁺ T cells, CD8⁺ T cells, and B220⁺ cells in the spleen (A) and draining LNs (B) of naïve (−) or imiquimod-treated (+) mice. (C) The numbers of CD4⁺FoxP3⁺ T cells in the spleen and draining LNs of naïve or imiquimod-treated mice. Values represent means (±sem) from greater than/equal to four mice/group. Significant differences between sample means are indicated: *P < 0.05; **P < 0.01.

The effects of CD19^−/− on the numbers of CD4⁺FoxP3⁺ Tregs in the spleen and draining LNs were also assessed after 6 days of imiquimod treatment. Treg numbers in the spleen and draining LNs were increased significantly during imiquimod-induced skin inflammation in both groups (Fig. 3C). Furthermore, imiquimod-treated CD19^−/− mice had significantly more Tregs in the spleen and draining LNs than imiquimod-treated WT mice (P<0.05). Thus, imiquimod treatment significantly increased the numbers of Tregs in the spleen and draining LNs, which did not correlate with the observed, increased disease severity in CD19^−/− mice.

Splenic B10 cell numbers decrease during imiquimod-induced skin inflammation

Although cytoplasmic IL-10 production was not detected in resting B cells from WT mice, spleen B cells, competent to express cytoplasmic IL-10 following stimulation with LPS, PMA, ionomycin, and monensin, were predominantly found within the CD1d^hiCD5⁺ B cell subset in WT mice (Fig. 4A), as described previously [14, 17]. By contrast, IL-10-producing B cells were less common within the non-CD1d^hiCD5⁺ B cell subset. Moreover, IL-10-producing B cells are predominantly CD21^int/hi, CD23^lo, and CD24^hi [14, 17]. Therefore, spleen IL-10-producing B cells are phenotypically distinct from CD21^intCD23⁺ follicular, CD1d^hiCD21⁺CD23⁻ marginal zone, CD1d⁺CD21⁺CD23⁺ T2-marginal zone precursor, and CD5⁺ B-1a B cell subsets, although they share some overlapping phenotypic makers. After stimulation for 5 h with LPS, PMA, and ionomycin, the proportions and absolute numbers of spleen IL-10-producing B cells were 2.7- and 5.5-fold higher, respectively, in control-treated WT than in control-treated CD19^−/− mice (P<0.01; Fig. 4B). Furthermore, the proportions and absolute numbers of spleen CD1d^hiCD5⁺ B cells were 4.6- and 8.8-fold higher, respectively, in control-treated WT mice than in control-treated CD19^−/− mice (P<0.01; Fig. 4C). Thus, B10 cell numbers were inversely paralleled with the severity of imiquimod-induced skin inflammation in WT and CD19^−/− mice.

Figure 4. IL-10-producing splenic B10 cells disappeared during imiquimod-induced skin inflammation.

(A) CD1d^hiCD5⁺ B cells are the predominant IL-10-producing B cell subset. Splenocytes from WT mice were cultured with LPS, PMA, ionomycin, and monensin for 5 h before permeabilization and staining using B220, CD1d, CD5, and IL-10 mAb. IL-10 production by B220⁺ B cells within the CD1d^hiCD5⁺ and non-CD1d^hiCD5⁺ subpopulations is shown with the proportions of IL-10⁺ cells within the indicated gates. (B) Splenic IL-10-producing B cell proportions and absolute numbers during imiquimod-induced skin inflammation in WT and CD19^−/− mice. Splenocytes were isolated from naïve or imiquimod-treated mice, and B220⁺ splenocytes were purified. Purified B cells were incubated in the presence of LPS, PMA, ionomycin, and monensin for 5 h. B cells were stained with B220 mAb. After permeabilization, the cells were stained with IL-10 mAb. Representative results demonstrate the proportion of IL-10-producing cells of the total B220⁺ B cells within the indicated gates. Bar graphs indicate mean (±sem) percentages and numbers of B cells producing IL-10. (C) CD1d^hiCD5⁺ B cell proportions and absolute numbers during imiquimod-induced skin inflammation in WT and CD19^−/− mice. Splenocytes were isolated from naïve or imiquimod-treated mice and analyzed by flow cytometry for CD1d, CD5, and B220 immunofluorescence. Representative results demonstrate the proportion of CD1d^hiCD5⁺ B cells of the total B220⁺ B cells within the indicated gates. Bar graphs indicate mean (±sem) percentages and numbers of CD1d^hiCD5⁺ B cells. (B and C) Significant differences between sample means are indicated: **P < 0.01.

B10 cells and the spleen CD1d^hiCD5⁺ B cell subpopulation were previously shown to increase significantly during EAE and DSS-induced colitis in mice [16, 25]. To determine whether B10 cell numbers changed during imiquimod-induced skin inflammation in the present study, they were quantified after 6 days of imiquimod treatment. Remarkably, spleen IL-10-producing B cell proportions and numbers were 63% and 86% lower, respectively, in imiquimod-treated WT mice than in control-treated WT mice (Fig. 4B; P<0.01 for both). Furthermore, CD1d^hiCD5⁺ B cell proportions and numbers were 78% and 91% lower, respectively, in imiquimod-treated WT mice than in control-treated WT mice (Fig. 4C; P<0.01 for both). Decreased B cell IL-10 production paralleled CD1d^hiCD5⁺ B cell proportions. Imiquimod treatment did not affect the proportions or numbers of spleen IL-10-producing B cells and CD1d^hiCD5⁺ B cells in CD19^−/− mice. Thus, there was a decrease in splenic B10 cell numbers following imiquimod treatment in WT mice.

Increased IL-10-producing B cells within the draining LNs and blood during imiquimod-induced skin inflammation

Although naïve B cells from peripheral LNs and blood produce little IL-10, the proportions of IL-10-producing B cells are increased during the course of contact hypersensitivity responses, EAE and DSS-induced colitis [14, 16, 25]. Therefore, we assessed whether splenic IL-10-producing B cells appeared in the draining LNs and blood during imiquimod-induced skin inflammation. Control-treated WT B cells from the draining LNs and blood exhibited little, if any, IL-10 production (Fig. 5). However, the proportions and numbers of IL-10-producing B cells within the draining LNs were 3.4- and 3.2-fold higher, respectively, in imiquimod-treated WT mice than in control-treated WT mice (Fig. 5A; P<0.01 for both). Furthermore, circulating, IL-10-producing B cell proportions were 11.9-fold higher in imiquimod-treated WT mice than in control-treated WT mice (Fig. 5B; P<0.01). IL-10-producing B cells were not detected in the inflamed skin by double-immunostaining for B220 and IL-10 (data not shown). Imiquimod treatment did not affect the proportions or numbers of IL-10-producing B cells from draining LNs and blood in CD19^−/− mice. Thus, IL-10-producing B cells were increased in the draining LNs and blood during imiquimod-induced skin inflammation.

Figure 5. Changes in B10 cell frequencies and numbers within draining LNs and blood during imiquimod-induced skin inflammation.

Mononuclear cells were isolated from draining LNs (A) or blood (B) in naïve or imiquimod-treated mice. Representative results demonstrate the frequencies of IL-10-producing cells after LPS, PMA, ionomycin, and monensin for 5 h within the indicated gates among total B220⁺ B cells. Bar graphs indicate mean (±sem) percentages and numbers of B cells that produced IL-10. Significant differences between sample means are indicated: **P < 0.01.

We next examined CD1d and CD5 expression in IL-10-producing B cells from draining LNs and blood in WT mice during imiquimod-induced skin inflammation. CD1d and CD5 were expressed at higher levels in IL-10⁺ than IL-10⁻ B cells (Fig. 6). Thus, IL-10-producing B cells in the draining LNs and blood have the phenotype of regulatory B10 cells.

Figure 6. Phenotypes of IL-10-producing B cells in the draining LNs and blood during imiquimod-induced skin inflammation.

IL-10-producing B cells from the draining LNs and blood in imiquimod-treated WT mice expressed CD1d and CD5. Mononuclear cells were isolated from draining LNs (A) or blood (B) in imiquimod-treated WT mice and were cultured with LPS, PMA, ionomycin, and monensin for 5 h before permeabilization and staining with CD1d, CD5, B220, and IL-10 mAb.

B10 cells regulate IFN-γ and IL-17 production during imiquimod-induced skin inflammation

We examined whether the loss of CD19 expression affected cytokine expression during imiquimod-induced skin inflammation by assessing the mRNA expression of several cytokines in WT and CD19^−/− mice. The spleen, draining LNs, and inflamed skin were harvested after 6 days of imiquimod treatment, and the expression of IFN-γ and IL-17A was quantified by real-time RT-PCR. In the spleen, relative mRNA expression of IFN-γ and IL-17A was not altered by imiquimod treatment in WT or CD19^−/− mice (Fig. 7A). However, IFN-γ and IL-17A transcripts of imiquimod-treated mice were increased significantly relative to those of control-treated mice in the draining LNs and inflamed skin (Fig. 7B and C). Furthermore, IFN-γ and IL-17A mRNA expression in imiquimod-treated CD19^−/− mice was enhanced significantly compared with imiquimod-treated WT mice.

Figure 7. Cytokine mRNA expression in splenic lymphocytes (A) draining LN lymphocytes (B), and inflamed skin (C) of WT and CD19^−/− mice.

Skin samples, draining LN, and spleen lymphocytes were harvested from naïve or imiquimod-treated mice. IFN-γ and IL-17A mRNA levels were quantified by real-time PCR analysis and normalized with the internal control GAPDH. Values represent means (±sem) from greater than/equal to four mice/group. Significant differences between sample means are indicated: *P < 0.05; **P < 0.01.

To examine further the involvement of CD19 expression in IFN-γ and IL-17A production, we cultured draining LN cells in the presence of IL-12 for Th1 differentiation and IL-6, TGF-β1, and IL-23 for Th17 differentiation. Imiquimod treatment enhanced the frequencies of IFN-γ-producing CD4⁺ T cells under Th1 cell-inducing conditions and IL-17A-producing CD4⁺ T cells under Th17 cell-inducing conditions in both groups of mice (Fig. 8). The frequencies of IFN-γ-producing CD4⁺ T cells under Th1 cell-inducing conditions and IL-17A-producing CD4⁺ T cells under Th17 cell-inducing conditions were increased significantly in imiquimod-treated CD19^−/− mice compared with imiquimod-treated WT mice. Thus, the loss of CD19 enhanced IFN-γ and IL-17A production in draining LNs and inflamed skin during imiquimod-induced skin inflammation in WT mice.

Figure 8. IFN-γ and IL-17A production by CD4⁺ T cells from the draining LNs during imiquimod-induced skin inflammation.

Numbers indicate percentages of CD4⁺ T cells within the indicated gates. Bar graphs indicate mean (±sem) percentages of CD4⁺ cells that produced IFN-γ or IL-17A. Significant differences between sample means are indicated: *P < 0.05; **P < 0.01.

IL-10 is required to regulate imiquimod-induced skin inflammation

We determined whether the enhanced skin inflammation observed in CD19^−/− mice was dependent on IL-10 by using a function-blocking mAb against the IL-10R. WT mice were treated with an IL-10R-blocking mAb or isotype-matched control mAb on Days 0 and 3. Blocking IL-10R function significantly augmented skin inflammation in imiquimod-treated WT mice compared with control mAb-treated WT mice (Fig. 9). CD19^−/− mice tended to exhibit more severe skin inflammation than WT mice treated with IL-10R-blocking mAb, but the difference was not statistically significant. Thus, the enhanced severity of imiquimod-induced skin inflammation observed in CD19^−/− mice was at least partially IL-10-dependent.

Figure 9. The suppression of imiquimod-induced skin inflammation is IL-10-dependent.

Imiquimod-induced skin inflammation in WT or CD19^−/− mice treated with control or IL-10R-specific mAb on Days 0, 3, and 5. Significant differences between sample means are indicated: *P < 0.05; **P < 0.01.

CD1d^hiCD5⁺ B cells reduce imiquimod-induced skin inflammation

The ability of CD1d^hiCD5⁺ B cells to regulate imiquimod-induced skin inflammation was assessed using adoptive transfer experiments. Splenic CD1d^hiCD5⁺ B cells and non-CD1d^hiCD5⁺ B cells were purified from WT or IL-10^−/− mice. Purified B cells were then transferred into CD19^−/− mice, which were treated with imiquimod, 2 days before or after transfer. Transferring WT CD1d^hiCD5⁺ B cells into CD19^−/− mice, 2 days before induction, significantly reduced the severity of imiquimod-induced skin inflammation (P<0.05 on Days 5–7; Fig. 10). By contrast, the severity of skin inflammation was not reduced significantly in mice that received non-CD1d^hiCD5⁺ B cells from WT mice or CD1d^hiCD5⁺ B cells from IL-10^−/− mice, 2 days before induction. The adoptive transfer of naïve WT CD1d^hiCD5⁺ B cells into CD19^−/− mice, 2 days after induction, did not reduce the severity significantly. Although the transfer of WT CD1d^hiCD5⁺ B cells into CD19^−/− mice, 2 days before induction, did not affect the number of CD4⁺, CD8⁺, and B cells in the draining LNs, it significantly reduced the numbers of CD4⁺ and CD8⁺ T cell infiltration in the inflammed skin (Fig. 11A and B). Moreover, the adoptive transfer of CD1d^hiCD5⁺ B cells from WT into CD19^−/− mice significantly reduced IFN-γ and IL-17A mRNA expression in the draining LNs and inflamed skin (Fig. 11C and D). Thus, the adoptive transfer of splenic B10 cells reduced imiquimod-induced skin inflammation when they were transferred before the induction of this inflammation.

Figure 10. Regulatory CD1d^hiCD5⁺ B10 cells suppress disease symptoms in imiquimod-induced skin inflammation.

CD1d^hiCD5⁺ or non-CD1d^hiCD5⁺ B cells were purified from naïve WT or IL-10^−/− mice by cell sorting. Purified cells were transferred into CD19^−/− mice. Imiquimod was administered to the recipient mice 2 days before or after transfer. Significant differences between PBS-treated CD19^−/− mice versus other groups are indicated: *P < 0.05. Values represent means (±sem) from greater than/equal to three mice/group.

Figure 11. The effect of transferring CD1d^hiCD5⁺ B10 cells into CD19^−/− mice 2 days before induction.

(A) The numbers of CD4⁺ T cells, CD8⁺ T cells, and B220⁺ cells in draining LNs. (B) The numbers of CD4⁺ T cells, CD8⁺ T cells, and B220⁺ B cells/field of view (×400) were counted in the inflamed skin sections. IFN-γ and IL-17A mRNA expression in draining LN lymphocytes (C) and inflamed skin (D). Values represent means (±sem) from greater than/equal to three mice/group. Significant differences between sample means are indicated: *P < 0.05.

DISCUSSION

B cells play positive effector roles and negative regulatory roles during immune responses [26], so their depletion may decrease or increase inflammation, depending on the disease model. It has been demonstrated previously that B cell depletion by anti-CD20 mAb prior to disease initiation is beneficial in mouse models of systemic sclerosis [27], rheumatoid arthritis [28], and type 1 diabetes [9], whereas B cell depletion enhances inflammation in the contact hypersensitivity responses [14]. Furthermore, B cell depletion has two contrasting effects on disease progression in EAE depending on when it occurs using the anti-CD20 mAb [15]. B cell depletion before EAE induction worsens disease symptoms, whereas the adoptive transfer of CD1d^hiCD5⁺ B10 cells, but not other B cells, reduces the severity of EAE. Therefore, increased EAE severity following B cell depletion is likely to result from effective B10 cell subset depletion. By contrast, B cell depletion after the development of EAE symptoms impairs pathogenic T cell expansion and reduces disease severity. Thus, identifying the relative contributions of each B cell subset to disease will be critical for the development of optimal therapeutic strategies. Moreover, these reciprocal regulatory roles of B cells are likely to overlap during the course of disease, with the balance of the opposing influences shaping the normal disease course.

Our experimental findings revealed that the splenic B10 cell subset plays a critical role in the inhibition of imiquimod-induced skin inflammation. Consistently, B cell depletion in humans using rituximab leads to the development of psoriasis in patients with no previous history of psoriasis or psoriatic arthritis [20, 21], suggesting the predominance of regulatory B cells, relative to effector B cells, in psoriasis. However, it has also been reported that rituximab administration in a patient with psoriatic arthritis improved the disease severity [29]. Therefore, it is possible that the balance between opposing positive and negative regulatory B cell functions shapes the course of disease in psoriasis as in mouse EAE. Further studies are needed to determine the contribution of B cells to the pathogenesis of psoriasis.

Most splenic B10 cells disappeared during imiquimod-induced skin inflammation in this study, whereas IL-10-producing B cells with a B10 phenotype were found in the blood and draining LNs. This suggests that splenic B10 cells entered the circulation and migrated to the draining LNs during imiquimod-induced skin inflammation, thereby inhibiting Th17 and Th1 immune responses. However, it is likely that non-CD1d^hiCD5⁺ B10 progenitor cells exist in the draining LNs and blood and are induced to mature during imiquimod-induced skin inflammation. In support of this, IL-10-producing B cells can be generated by stimulating B cells from peripheral LNs and blood with a combination of agonistic anti-CD40 mAb for 48 h plus LPS, PMA, ionomycin, and monensin for the last 5 h of culture [17]. By contrast, splenic B10 cell numbers were previously found to increase during EAE and DSS-induced colitis [16, 25], which may be explained by a more efficient induction of B10 migration following potent TLR stimulation by imiquimod.

Although neither an increase in the number of B cells nor IL-10-producing B cell infiltration was observed in skin during imiquimod-induced skin inflammation, it is possible that a small number of B10 cells infiltrate the skin and directly suppress the inflammatory response. Moreover, imiquimod treatment reduced not only CD1d^hiCD5⁺ B cells but also non-CD1d^hiCD5⁺ B cells in the spleen of WT mice. Although the combination of agonistic anti-CD40 mAb for 48 h plus LPS, PMA, ionomycin, and monensin for the last 5 h of culture does not lead to IL-10 production from spleen non-CD1d^hiCD5⁺ B cells, it remains possible that imiquimod treatment converts non-CD1d^hiCD5⁺ B cells into IL-10-competent B cells. Imiquimod treatment may also induce B cell apoptosis or plasma cell differentiation, thereby decreasing splenic B cell numbers. It is not yet possible to track B10 cells in vivo, as no specific markers have been identified, so further studies are needed to determine the precise mechanisms by which B10 cells attenuate inflammation in local inflammatory responses.

IL-10 is thought to play a protective role in the development of psoriasis [19]. It is secreted by multiple cell types, including T cells, B cells, monocytes, macrophages, mast cells, and eosinophils, and can suppress Th1 and Th2 polarization and inhibit macrophage antigen presentation and proinflammatory cytokine production [30]. In psoriasis patients, plasma IL-10 levels are decreased significantly compared with healthy individuals and are negatively correlated with the severity of skin lesions [31, 32]. Moreover, IL-10 transcripts are reduced in psoriasis patients relative to healthy individuals [18]. Therefore, the relative IL-10^−/− may be associated with the development of psoriasis. Consistently, the systemic application of human rIL-10 reduces the severity of skin lesions in psoriasis patients [18, 33, 34]. However, a larger, randomized, double-blind, placebo-controlled study demonstrated that systemic human rIL-10 administration resulted in only a temporary improvement [35]. As the serum half-life of rIL-10 is only 2.6 h in humans [36], it is the likely cause of the disappointing results of rIL-10 treatment for psoriasis. The results of this study show that B10 cell transfer ameliorates imiquimod-induced skin inflammation, suggesting that the therapeutic use of regulatory B cells is more promising than rIL-10 application in the treatment of psoriasis.

The current findings demonstrate that B10 cells play a regulatory role in imiquimod-induced skin inflammation and that adoptively transferred B10 cells before induction are sufficient to ameliorate this. By contrast, transferring B10 cells after induction did not reduce the severity significantly. It has been reported that B10 cells predominantly regulate disease initiation, whereas Tregs suppress late-phase disease [14, 15, 25]. As this imiquimod-induced skin inflammation model peaks within 6 days, it may be too late to transfer B10 cells after the induction. Moreover, it is also suggested that antigen-specific B10 cells are required to suppress inflammation [14, 15]. Additional studies will therefore be required to clarify the mechanisms by which B10 cells control imiquimod-induced skin inflammation. Furthermore, CD19^−/− mice had increased Treg numbers significantly in the spleen and draining LNs compared with WT mice, although this did not correlate with the observed, increased disease severity in CD19^−/− mice. It is likely that Tregs increase in response to excessive skin inflammation in CD19^−/− compared with WT mice. However, it is not certain whether Tregs do not play a major role in the inhibition of the imiquimod-induced skin inflammation model, as this model is too acute, or whether the increase is not sufficient to compensate for the loss of B10 cells in CD19^−/− mice. Further studies, assessing the roles of regulatory B cell and Treg studies using a chronic model of psoriasis, are necessary to clarify this.

Currently, psoriasis patients are treated with TNF inhibitors, corticosteroids, and immunosuppressive drugs, but these therapies can cause severe side-effects, such as infection, and may fail in certain cases. Recently, IL-10-producing regulatory B cell subsets have been identified in human blood [37–39], which supports the use of regulatory B cell-based therapies in the treatment of human psoriasis. Further studies are required to investigate the association between regulatory B cell numbers and the severity of psoriasis in humans. It is also necessary to establish efficient methods of expanding regulatory B cells in vitro or in vivo and to examine the effects of regulatory B cell transfer in psoriasis patients. Nonetheless, if the efficacy observed in mice translates to humans, our results may provide new insights and therapeutic approaches for treating psoriasis.

AUTHORSHIP

K.Y., T.F.T., and S. Sato designed the research. K.Y., M.K., and S. Shibata performed the research. K.Y., N.I., T.F.T., and S. Sato contributed new reagents and analytic tools. K.Y., Y.A., Y.T., M.S., and T.K. analyzed the data. K.Y., T.F.T., and S. Sato wrote the paper.

DISCLOSURES

T.F.T. is a paid consultant for MedImmune and Angelica Therapeutics. The authors declare no other financial conflicts of interest.