IL-10 from marginal zone precursor B cells controls the differentiation of Th17, Tfh and Tfr cells in transplantation tolerance

Abstract

B cells are known to control CD4 T cell differentiation in secondary lymphoid tissues. We hypothesized that IL-10 expression by marginal zone precursor (MZP) regulatory B cells controls the differentiation and positioning of effector and regulatory T cells during tolerization. Costimulatory blockade with donor-specific transfusion (DST) and anti-CD40L mAb in C57BL/6 mice induced tolerance to allogeneic cardiac allograft. B cell depletion or IL-10 deficiency in B cells prevented tolerance, resulting in decreased follicular regulatory CD4⁺ T cells (Tfr) and increased IL-21 expression by T follicular helper (Tfh) cells in the B cell and T cell zones. IL-21 acted with IL-6 to induce CCR6⁺ Th17 that caused rejection. Deficiency or blockade of IL-6, IL-21, IL-21R, or CCR6 prevented B cell depletion-induced acute cellular rejection; while agonistic mCCL20-Ig induced rejection. Adoptive transfer of IL-10^+/+ MZP in tolerogen treated CD19-Cre^+/−::IL-10^fl/fl mice rescued the localization of Tfh and Tfr cells in the B cell follicle and prevented allograft rejection. MZP B cell IL-10 is necessary for tolerance and controls the differentiation and position of Th17, Tfh and Tfr cells in secondary lymphoid tissues. This has implications for understanding tolerance induction and how B cell depletion may prevent tolerance.

Introduction

B cells are well characterized for their important functions in transplantation such as production of alloantibody, presentation of alloantigen to T cells, secretion of pro- and anti-inflammatory cytokines, and lymphoid organogenesis. B cells produce TNFα, GM-CSF, IFNγ, IL-6, IL-12 and lymphotoxin alpha (LT-α), and these secreted cytokines potentiate the effector function of B cells and help T cell responses [1]. B cell mediated chronic allograft rejection can occur in the complete absence of alloantibody; and B cells affect secondary lymphoid organ structure and T cell cytokine production, leading to increased T cell infiltration of the graft [2]. Together these studies suggest that B cells perform many non-humoral functions with pro-inflammatory characteristics.

Subsets of B cells also act as regulatory cells (Breg) and influence autoimmunity and transplantation tolerance [3,4]. B cells have been shown to suppress the immune response by mediating CD95 induced deletion of alloantigen specific CD8 T cells [5], inducing anergy [6], inducing regulatory CD4 T cells, inhibiting self-reactive CD4 T cells in a CD40-CD40L dependent manner [7], or regulating dendritic cells (DCs) [8]. Several distinct types of regulatory B cells have been identified, including CD1d^hiCD5⁺ B10 B cells [9], CD19⁺CD21^hiCD23^hiCD24^hi transitional T2 B cells [10], TIM-1⁺ B cells [11], a subset of CD138⁺IL-35 producing B cells [12], CD19⁺CD23⁺sIgM^hisIgD^hiCD21/CD35^hi MZP B cells[3], and human CD19⁺CD38^hiCD24^hi B cells [13]. These Bregs are known to suppress the differentiation of Th1 and Th17 cells [14], induce differentiation of CD4⁺Foxp3⁺ T cells [14,15], and inhibit the antigen presenting cell (APC) function of DCs [16]. Breg suppress the immune response by secretion of IL-10 and/or TGF-β [3,14,17], IL-35 [12], and granzymes [18]. CD39, CD73 and GITR have also been shown to play important roles in the regulatory function of Breg [19,20]. These studies demonstrate that B cells perform many non-humoral functions with anti-inflammatory characteristics.

Co-stimulatory blockade with CD40L–CD40 has been used in the induction of transplantation tolerance in mice and immunosuppression in non-human primates [21,22]. CD40 is expressed on B cells, macrophages, DCs, epithelial cells, hematopoietic progenitors and activated T cells; whereas activated T cells, activated B cells and activated platelets express CD40L [23]. Under inflammatory conditions, peripheral blood monocytes, mononuclear phagocytes, smooth muscle cells and vascular endothelial cells also express CD40L [23]. Blockade of CD40L with anti-CD40L mAb or genetic deficiency of CD40 or CD40L inhibits the effector immune response and prevents allograft rejection. There are several mechanisms proposed to explain CD40-CD40L blockade induced transplantation tolerance, including effects on B cell, NK cell, and CD4 and CD8 T cell immunity [3,24–26]. However, the contribution of B cells in regulating effector and regulatory CD4 T cells has not been thoroughly investigated.

Using heterotopic allogeneic cardiac transplantation model, we showed that B cell depletion prevented the co-stimulatory blockade induced tolerance [3]. Furthermore, we showed that costimulatory blockade increased IL-10 in the MZP B cells but not in other B cell subsets including the known CD5⁺CD1d^hi regulatory B cells (Breg) [3]. Detailed analysis showed that co-stimulatory blockade increased IL-21R expression in MZP B cells, and IL-21R⁺ MZP B cells produced significantly increased IL-10. Depletion of B cells or B cell specific IL-10 deficiency inhibited the generation of tolerance, leading to acute cellular rejection of allograft [3]. Adoptive transfer of IL-10 sufficient MZP B cells into the B cell specific IL-10^−/− mice (B-IL-10^−/−) rescued costimulatory blockade induced tolerance. Since MZP Breg and IL-10 are essential for tolerance, we hypothesized that Tfh and Tfr would be most proximately influenced by changes in B cells.

In the present study, transplantation tolerance was induced in wild-type C57BL/6 or B cell specific IL-10 deficient mice (B-IL-10^−/− mice) by intravenous injection anti-CD40L mAb and donor-specific transfusion, and transplanted BALB/c cardiac allograft, and investigated the effect of B cell depletion on the differentiation and localization of various subsets of effector and regulatory CD4 T cells in the secondary lymphoid organs. We showed that depletion of B cells resulted in decreased Tfr along with abnormal distribution of IL-21 secreting Tfh. We also showed that depletion of B cells or deficiency of IL-10 in B cells showed altered localization of Tfh and Tfr cells in the secondary lymphoid organs. Reconstitution of wild-type MZP B cells in the B-IL-10^−/− recipients rescued the altered localization of Tfh and Tfr cells and tolerance to allograft. Tfh in turn induced the differentiation of Th17 cells and their CCR6 dependent migration to the allograft, preventing tolerance.

Materials and Methods

Mice

BALB/c (H-2^d), C57BL/6 (H-2^b), and CCR6^−/− (H-2^b) mice 8–10 weeks old were purchased from The Jackson Laboratory. IL-10^fl/fl (H-2^b) mice were from Dr. Christopher Karp (Cincinnati Children’s Hospital Research Foundation, Cincinnati, OH). CD19Cre^+/−:IL-10^fl/fl (B-IL-10^−/−; H-2^b) mice were generated by breeding CD19-Cre^+/+ with IL-10^fl/fl mice [3]. Genotype of B-IL-10^−/− mice were confirmed by PCR analysis using specific-primers and tail DNA. TEa mice (C57BL/6 background; H-2^b) were from Dr. Alexander Rudensky (Memorial Sloan-Kettering Cancer Center, New York, NY). All mice were housed in a specific pathogen-free facility in microisolator cages. All experiments used age- and sex-matched mice in accordance with protocols approved by the Institutional Animal Care and Utilization Committee.

Antibodies and reagents

Anti-mouse CD4 (GK1.5), anti-mouse CD8 (53.67), anti-mouse B220 (RA3–6B2), PE-anti-mouse CD19 (MB19-1), APC anti-mouse B220 (RA3–6B2), Pacific Blue anti-mouse CD45 (30F-11), PE-anti-mouse IgM (II/41), PE anti-mouse CD8 (53–6.7), APC anti-mouse CD4 (GK1.5), FITC anti-mouse CD25 (PC61.5), biotin anti-human/mouse GL-7 (GL7), PE anti-mouse/rat Foxp3 (FKJ-16s) antibodies and isotype control antibodies were purchased from eBioscience (San Diego, CA). FITC anti-mouse CD93 (AA4.1), PE/Cy7 anti-mouse CD21/CD35 (7E9), PE-anti-mouse CD23 (B3-B4), Pacific Blue anti-mouse IgD (11–26c 2A), and biotin anti-mouse IgM (MRM-47) were purchased from Biolegend (San Diego, CA). Rabbit anti-mouse Foxp3 polyclonal antibody, rabbit anti-MIP3a (CCL20), rabbit anti-CXCR5 (EPR8837) and MOMA-1 were purchased from Abcam (San Francisco, CA). FITC anti-mouse CXCR5 antibody was purchased from BD Pharmingen (San Diego, CA). mCCL20-Ig was produced as previously described [27]. Anti-mouse CD20 mAb (clone 5D2, mouse IgG2a) was received from Genentech, Inc. (San Francisco, CA). Cy3-donkey anti-rat, FITC-donkey anti-rat, Cy5-donkey anti-rat, Cy3-goat anti-hamster, Cy5-anti-hamster, FITC-donkey anti-hamster, FITC-goat anti-rabbit, Cy5-goat anti-rabbit antibodies and fluorochrome conjugated streptavidin were purchased from Jackson Immunoresearch Laboratory, Inc. (West Grove, PA). Anti-mouse CD40L (MR-1), anti-mouse IL-6 (MP5–20F3) and control isotype rat IgG1 antibodies were purchased from BioXcell (West Lebanon, NH). Carboxyfluorescein diacetate succinimidyl ester (CFSE) and 4',6-diamidino-2-phenylindole (DAPI) were purchased from Invitrogen (Carlsbad, CA). Neutralizing mouse anti-mouse IL-21 mAb (clone 397.18.2.1) was from Novo Nordisk, Denmark. IL-21R–Ig fusion protein was from Pfizer (San Diego, CA)

Cardiac transplantation and treatment protocols

Transplantation tolerance was induced by administering 1 × 10⁷ BALB/c splenoctyes intravenously (iv) on day −7 prior to transplant and 250 µg anti-CD40L mAb iv on days −7, −4, 0 and +4 to C57BL/6 recipients (10–12 weeks) of sex-matched BALB/c (8–10 weeks) donor vascularized cardiac allografts [3, 25]. B cells were depleted by one iv injection of 100 µg anti-mouse CD20 mAb (5D2) mAb on day+1. Anti-CD20 mAb depleted about ~75% of total CD19⁺ B cells in the spleen and LN and CD19⁺IgM⁺IgD⁺ B cells in the bone marrow with 3 days [3]. Graft function was monitored every other day by abdominal palpation. For immunohistological of allograft and cellular phenotypic analysis of infiltrating cells, mice were euthanized at day 5. Spleen, LN and donor allograft was excised, fixed in OCT medium or processed for single cell suspension.

Isolation of cells from lymph node (LN), spleen and grafts

Cardiac grafts were perfused with PBS, minced and digested with 1 mg/ml collagenase D (Roche, Indianapolis, IN) and 0.01% DNase I in RPMI medium for 45 minutes at 37°C. Digested suspensions were passed through a 70 µm nylon mesh and centrifuged, cell pellet was suspended in RPMI medium. Single cell suspension were prepared using Lympholyte-M (Cedarlane, Ontario, Canada) density gradient centrifugation as per manufacturers guidelines. Single-cell suspensions were washed with PBS, RBC lysed using ACK lysis buffer (Lonza, Walkerville, MD), and used for the flow cytometry analysis (FACSCantoII or LSR Fortessa, BD Biosciences, Mountainview, CA) or sorting (FACS Aria, BD Biosciences). Spleen, and LN were gently dissociated into single-cell suspensions, and RBC lysed using ACK lysis buffer (Lonza, Walkerville, MD) and used for FACS analysis or FACS sorting.

Cell staining and flow cytometric analysis

Staining of cells was performed with the indicated antibodies with 1µg/10⁶ cells at 4°C for 30 minutes. MZP B cells were purified using a FACS ARIA II sorter (BD Bioscience, Mountainview, CA). Trypan Blue staining of cells showed more than 98% viable cells, and purity was found to be more than 98% for each subset. Data were acquired using the FACSCantoII, LSR Fortessa, or LSRII flow cytometer (BD Biosciences, Mountainview, CA), and analyzed using FlowJo software (Tree Star Inc., Ashland, OR).

Quantitative real-time RT PCR (qRT-PCR)

Cells were lysed in TRIzol reagent (Invitrogen), total RNA was purified, and cDNA was made using oligo d(T)12–14 primer and Omniscript RT kit (Invitrogen) as per the manufacture’s guidelines. mRNA expression was quantified by CFX96 thermal cycler (Bio-Rad, Hercules, CA) or 7900 fast real-time PCR (Applied Biosystem, Carlsbad, CA) using SYBR Green PCR kit (Qiagen, Valencia, CA). PCR consisted of a 15 minutes at 95°C denaturation step, followed by 40 cycles of 15sec at 94°C, 20sec at 56°C, 20sec at 72°C. Relative mRNA expression of specific gene was calculated as: 2^{(Ct of cyclophilin-A - Ct of specific gene)} The primers used for real-time PCR: mouse IL-21, forward 5’-TGAAAGCCTGTGGA AGTGCAAACC-3’ and reverse 5’-AGCAGATTCATCACAGGACACCCA-3’; mouse CCR6, forward 5’-ACCTGTTGAACATGGCCATCACAG-3’ and reverse 5’-AAACCCAAGTGTTGGT GGCATGAG-3’; Tbet, forward 5’-ATCATCACTAAGCAAGGACGGCGA-3’ and reverse 5’-AACATCCTGTAATGGCTTGTGGGC-3’; BCL-6, forward 5’-CCATCACCAACACCAAGGTG CAAT-3’ and reverse 5’- CCATCACCAACACCAAGGTGCAAT-3’. Other primers were published earlier [28].

Histopathological analysis

Cardiac grafts, spleens and LN were harvested, frozen directly in Optimal Cutting Temperature (OCT) compound (Sacura Finetek, Torrance, CA), and stored at −80°C. 8 µm sections were cut with a Leica 1900CM cryomicrotome, fixed with chilled acetone, blocked with 2.5% normal horse serum (Vector Laboratories, Burlingame, CA), stained with the indicated primary antibodies for 30 minutes, stained with conjugated secondary reagents, and mounted with VectaShield mounting solution (Vector Laboratories) with or without DAPI. Images were acquired with a Leica fluorescence microscope (Leica Mikrosysteme, Vertrieb, Germany) and a digital Hamamatsu CCD camera (Hamamatsu Corporation, Bridgewater, NJ). Separate images were collected on the CY3, GFP, and DAPI channels, overlaid, and analyzed with Openlab software (Improvision, Lexington, MA) and Leica MMF software. Quantification of graft infiltrating cells was performed by counting 4–5 fields per tissue section and 3–4 sections per graft.

For hematoxylin and eosin (H&E) staining, PBS buffered formalin fixed tissues were embedded in paraffin, sectioned at 5 µm and H&E staining performed. Parenchymal rejection score of graft was performed as described earlier [29].

Adoptive transfer of CD4⁺ TEa cells

Spleen and LN CD4⁺ T cells from naïve TEa T cell receptor transgenic mice were purified using the mouse CD4⁺ T cell negative selection magnetic-bead based kit (Stem Cell Technologies, Vancouver, BC, Canada) according to the manufacturer’s protocol. Purity of CD4⁺ T cells was typically >85–95%; these cells showed a naïve T cell phenotype (CD62L^hi) with very low expression of CD69, CD44 and CD25 [30]. Cells were stained with CFSE (5 µM), and adoptively transferred (2 × 10⁷ cells/mouse) into recipient C57BL/6 mice +1 day after transplantation. Five days later, CFSE⁺CD4⁺ T cells from spleen and LN were purified by flow cytometric sorting, and mRNA expression was analyzed.

Statistics

Graft survival was graphically expressed using the Kaplan-Meier method and statistical differences were assessed by Log-rank (Mantel-Cox) test using Prism 6 software (GraphPad software, La Jolla, CA). Unpaired, two-tailed Student t test was performed, and p < 0.05 was considered statistically significant.

Results

B cell depletion alters Th17 induction

We previously showed that depletion of B cells in tolerogen (DST + anti-CD40L) treated mice led to reduced migration of Foxp3⁺ Treg cells to the allograft and acute cellular rejection [3]. In that study, rejecting allografts had increased parenchymal rejection scores and perivascular infiltration of leukocytes. Such acute perivascular cellular infiltration in the rejecting cardiac allograft has been associated with Th17 mediated rejection [31]. Th17 cells are known to express the inflammatory cytokines IL-17 and IL-21, and use the chemokine receptor CCR6 to migrate into the inflamed tissues. To elucidate the mechanisms of how B cell or B cell specific IL-10 depletion affected the development and differentiation of CD4⁺ T cells, total CD4⁺ T cells from allograft, LN and spleen were purified from tolerogen treated and/or B cell depleted rejecting mice after five days of transplantation, and expression of several key effector molecules was analyzed. B cell depletion increased the frequency of IL-21 producing CD4 T cells in LN, spleen and allograft (Figure 1A). B cell depletion increased expression of IL-17 and IL-21 mRNA in both spleen and allograft infiltrating CD4⁺ T cells, and significantly increased CCR6 mRNA in the allograft (Figure 1B). These data suggested that B cell depletion enhanced IL-21 producing CD4⁺ T cells and thereby induced the development of Th17 cells.

Figure 1. B cell depletion alters Th17 differentiation.

(A) C57BL/6 recipient mice given tolerogen (DST + anti-CD40L mAb), BALB/c cardiac allografts on d0, and anti-mCD20 mAb or control IgG on d+1. Expression of intracellular IL-21 in CD4 T cells analyzed in LN, spleen and graft and d5. Allograft infiltrating cells were pooled from 3–4 mice/group. Data representative of one of two independent experiments. (B) CD4⁺CD45⁺ cells from recipient spleen and donor hearts purified by flow cytometry sorting, and expression of mRNA analyzed by qRT-PCR on d5. Data are cumulative of three independent experiments. In each experiments cells pooled from 3–4 mice. Mean and s.e.m shown.

B cell depletion induces differentiation of alloantigen specific Th17

To specifically show that B cell depletion enhanced the differentiation of Th17 cells, alloantigen-specific, CFSE labeled, naïve TEa T cell receptor transgenic CD4⁺ T cells were adoptively transferred into tolerogen treated C57BL/6 mice, with or without B cell depletion. TEa T cells are specific for donor I-E^d presented by recipient I-A^b [32]. After 5 days, CFSE⁺CD4⁺ T cells from spleen and LN were purified; and Th1, Th2, Th17 and Treg specific mRNA expression was analyzed. The results showed that B cell depletion led to increased IL-17, IL-21 and CCR6 expression, and no change in Foxp3, IL-4 or IL-10 expression observed in these alloantigen-specific CD4 T cells in spleen and LN (Figures 2A and 2B, respectively). It has been shown that Th17 cytokines can inhibit the differentiation of Th1 in vivo and in vitro [33]. Our results also showed a strong Th17 response and reduced levels of Th1 specific transcription factor T-bet in the B cell depleted mice (Figures 2A and 2B). These results strongly supported the notion that depletion of B cells induced the differentiation of inflammatory Th17 cells in spleen and LN.

Figure 2. B cell depletion induces alloantigen specific Th17 cells. (A and B).

C57BL/6 recipients given tolerogen, anti-mCD20 mAb or control IgG, and CFSE-labeled TEa CD4⁺ T cells adoptively transferred on d+1. After 5 days, CD4⁺CFSE⁺ T cells from (A) spleen and (B) LN purified using flow cytometry, and mRNA expression analyzed by qRT-PCR. Purified cells pooled for RNA isolation from 3–4 mice/group. Data are cumulative of three independent experiments. Mean and s.e.m shown.

B cell depletion alters Tfh position and CCR6, CCR7 and IL-21 expression

CD4⁺CD44⁺CXCR5⁺ Tfh secrete large amounts of IL-21 in the germinal center [34], which helps in the development and differentiation of B cells [35]. IL-21 also helps in the differentiation of Th17 cells outside the follicle [36]. Since the results above showed Th17 induction due to B cell depletion, we investigated if B cell depletion altered Tfh frequency and distribution and IL-21 expression. Depletion of B cells in tolerogen treated mice reduced the percentage of Tfh in the spleen (Figure 3A), commensurate with decreased Bcl-6 mRNA expression (Figure 3B). It is important to note that tolerogen treatment alone did not change the frequency of Tfh cells in spleen (Figure 3C), showing that B cell depletion was required to perturb both the frequency and location of Tfh cells in spleen.

Figure 3. B cell depletion alters the Tfh cell location and expression of IL-21, CCR6, and CCR7.

(A) C57BL/6 mice given tolerogen and anti-CD20 mAb, and after 5 days percentages of CD4⁺CD44⁺CXCR5⁺ Tfh cells in spleen analyzed by gating on CD4⁺ cells (left). Mean percentage of CD4⁺CD44⁺CXCR5⁺ cells plotted (right). Mean and s.e.m shown. 5–6 mice/group. (B) C57BL/6 recipients given tolerogen plus anti-mCD20 mAb or control IgG, and CFSE-labeled TEa CD4⁺ T cells adoptively transferred on d+1. After 5 days, CD4⁺CFSE⁺ T cells from LN and spleen purified and BCL6 mRNA expression analyzed by qRT-PCR. Purified cells pooled for RNA isolation from 3–4 mice/group. (C) C57BL/6 mice treated as indicated. After 5 days, percentages of CD44⁺CXCR5⁺ Tfh cells in spleen analyzed by gating on CD4⁺ cells. 4 mice/group. (D) C57BL/6 mice given tolerogen and anti-CD20 mAb. After 5 days, CD4⁺CD44⁺CXCR5⁺ cells purified from spleen, and IL-21, CCR6 and CCR7 mRNA expression analyzed by qRT-PCR. Purified cells pooled for RNA isolation from 3–4 mice/group. Error bars are standard deviation. (E) C57BL/6 mice given tolerogen, anti-CD20 mAb, and purified CFSE labeled CD4⁺CD44⁺CXCR5⁺ Tfh cells adoptively transferred. After 5 days, spleens harvested, frozen tissue sections stained for CD4 and B220, and location of CFSE⁺ Tfh cells in spleen analyzed (top). Magnification 400x. Quantitative analysis of CFSE⁺ Tfh cells in T cell and B cell zones (bottom). 3 mice/group, 3–5 sections/spleen, 4–5 follicles/section. (F) C57BL/6 mice given tolerogen and anti-CD20 mAb. After 5 days, spleens harvested, 8µM frozen tissue sections stained for CD4, CXCR5, B220 and Foxp3, and location of CD4⁺CXCR5⁺Foxp3⁺ Tfr cells in the B cell follicle analyzed. Tfr cells quantitated and plotted. 3 mice/group, 3–5 sections/spleen, 3–4 follicles/section.

Chemokine receptors CCR6 and CCR7 are required for the migration into inflamed tissues and localization of T cells in the LN, respectively [37,38]; and CCR7^loPD-1^hi Tfh cells in the circulation indicate active Tfh differentiation in secondary lymphoid organs, and correlate with inflammation and autoimmunity [39]. To investigate the effect of B cell depletion on IL-21 and chemokine receptors expression in Tfh, CD4⁺CD44⁺CXCR5⁺ Tfh cells were purified from tolerogen treated mice, with or without B cell depletion, and further analyzed. B cell depletion enhanced IL-21 and CCR6 and reduced CCR7 expression in Tfh cells (Figure 3D). These results suggest decreased CCR7 and increased CCR6 expression in the Tfh cells may contribute to altered location outside the B cell follicles.

To further investigate changes in Tfh distribution, naïve wild-type, CFSE labeled, CD4⁺CD44⁺CXCR5⁺ Tfh cells were adoptively transferred into mice which had received tolerogen with or without B cell depletion. The results showed that there was decreased migration of Tfh into splenic B cell zones and increased migration into T cell zones in B cell depleted mice in comparison to those treated only with tolerogen (Figure 3E). These data further support the observations that B cell depletion not only altered the distribution and phenotype of the T cells, but also altered chemokine or other directional signals within the LNs to influence T cell differentiation. Follicular regulatory T cells (CD4⁺CXCR5⁺Foxp3⁺; Tfr) are known to suppress the germinal center reaction [40]. Our data showed that depletion of B cells also reduced the number of Tfr located in the B cell follicles (Figure 3F). Together, these results showed that B cell depletion led to several alterations in CD4⁺ T cell differentiation, with increased CCR6⁺ Th17; localization of Tfh cells outside follicles with increased IL-21 and CCR6 and decreased CCR7 expression; and decreased numbers of Tfr cells in the follicles.

B cell IL-10 deficiency alters Th1, Th17, Tfh, and Tfr induction and distribution

IL-10 produced by Tfr cells plays an important role in controlling the inflammatory reaction and maintaining tolerance [40]. In the cardiac transplant model, we previously showed that IL-10 produced by MZP B cells played an important role in tolerance [3]. To understand how IL-10 produced by MZP B cell controls the differentiation and localization of Th1, Th17, Tfh, and Tfr during inflammation and tolerance, we analyzed how B cell IL-10 regulated these T cell subsets in naïve wild type and naïve B-IL-10^−/− mice, as well as in tolerogen treated B-IL-10^−/− mice, with or without adoptive transfer of wild type MZP B cells. The results showed that absence of IL-10 in B cell (naïve B-IL-10^−/− mice) lead to a reduction of Tfh in the follicle compared to IL-10 sufficient mice (naïve wild-type mice) (Figure 4A). Adoptive transfer of wild-type MZP B cells into tolerogen treated B-IL-10^−/−mice rescued the Tfh distribution in the B cell zone so that it was similar to wild-type (Figure 4A). Deficiency of IL-10 in B cells also led to reduced localization of Tfr cells into the B cell follicles (Figure 4B), and adoptive transfer of wild-type MZP B cells into tolerogen treated and transplanted B-IL-10^−/− mice restored the presence of Tfr cells in the B cell follicles (Figure 4B). Furthermore, compared to wild-type naïve mice, deficiency of IL-10 in B cells led to a decreased percentage of CD4⁺GL7⁺Foxp3⁺ Tfr and reduced frequencies of Tfr and Tfh in the follicles (Table 1). In contrast, the frequencies of CD4⁺Bcl-6⁺T-bet⁺ Th1 and RORγτ⁺ CD4⁺ Th17 were increased in the germinal centers (GL7⁺) and non-germinal centers (GL7⁻) of both LN and spleen in the naïve B-IL-10^−/− mice compared to naïve wild-type mice (Table 1), suggesting that B-IL-10^−/− mice had increased inflammatory Th1 and Th17. Together, these data demonstrated that IL-10 producing MZP B cells were required for maintaining the balance of Th1, Th17, Tfh, and Tfr inside and outside the germinal center and transplantation tolerance.

Figure 4. B cell IL-10 deficiency alters Tfh and Tfr location.

(A, B) Naïve wild type or naïve B-IL-10^−/− mice, or B-IL-10^−/− recipients given tolerogen and BALB/c allografts. On the day of transplant, IL-10 sufficient MZP B cells adoptively transferred into recipients. Mice euthanized on day 40, and spleens harvested and frozen. (A) Spleen sections (8µM) stained for CD4, B220, CXCR5 and Foxp3. Representative images shown from B-IL-10^−/− + tolerogen + allograft or B-IL-10^−/− + tolerogen + allograft + WT MZP B cell treated mice (upper). Mean numbers of Foxp3⁺CD4⁺ cells in the follicles (CXCR5⁺ zone) of spleen (lower). 3 mice/group, 3–4 follicles/section. Magnification 200X. (B) Spleen sections (8µM) stained for CD4, B220, CXCR5 and Foxp3, and analyzed. Representative images of naïve littermate control or B-IL-10^−/− spleen (upper). Mean CD4⁺CXCR5⁺Foxp3⁻ Tfh cells per follicle (lower). Error bars are standard deviation. 3 mice/group, 3–4 follicles/section. Magnification 200X.

Table 1.

Frequency of Th1, Th2, Th17 and Tfr cells in B-IL-10−/− mice spleen and lymph node.

	Spleen			Lymph node			Spleen			Lymph node
	Germinal Center (gated on GL7+ cells)						Non-Germinal Center (gated on GL7 cells)
	B-IL-10 KO mice	Control mice	p value	B-IL-10 KO mice	Control mice	p value	B-IL-10 KO mice	Control mice	p value	B-IL-10 KO mice	Control mice	p value
CD4+Foxp3+ T cells	10.11 ± 0.5	12.7 ± 0.8	0.013	15.8 ± 1.6	14.8 ± 0.8	0.215	13.3 ± 0.8	14.6 ± 1.3	0.06	14.1 ± 0.2	13.2 ± 0.9	0.201
CD4+Bcl6+Tbet+ (Th1) cells	12 ± 1.4	5.3 ± 0.9	0.004	14.1 ± 0.7	6.1 ± 2.5	0.011	2.4 ± 0.4	1.3 ± 0.1	0.011	3.2 ± 0.6	1.7 ± 1.3	0.174
CD4+IL- 21R+RORyt+ (Th17) cells	0.1 ± 0.03	0.12 ± 0.06	0.353	0.4 ± 0.1	0.07 ± 0.03	0.025	0.004 ± 0.002	0.004 ± 0.001	0.343	0.05 ± 0.02	0.01 ± 0.004	0.019
CD4+GATA3+ (Th2) cells	3.6 ± 0.9	2.0 ±0.4	0.107	7.6 ± 1.1	5.9 ± 1.2	0.167	0.03 ± 0.01	0.04 ± 0.01	0.22	0.8 ± 0.02	0.15 ± 0.1	0.255

Eight weeks old littermate control or B-IL-10−/− mice were sacrificed. Inguinal lymph nodes and spleen cells were stained for the indicated marker and analyzed by flow cytometry. Data shown is the mean percentage of various subsets of CD4+ T cells. Subsets showing significant changes in B-IL-10−/− mice are in bold font. Mean percentage of indicated subsets in different organs and their localization given. (n = 5 mice/group).

Neutralization of IL-6 or IL-21, or deficiency of CCR6, prevents B cell depletion induced allograft rejection

B cell depletion resulted in increased IL-21 expression by Tfh cells and an increased frequency of CCR6⁺ Th17. Since IL-6 and IL-21 are required for the development and differentiation of Th17 cells [41], we determined if IL-6 or IL-21 were mechanistically part of B cell depletion induced allograft rejection. To investigate the role of IL-21 or IL-21R in the co-stimulatory blockade induced tolerance, wild-type C57BL/6 mice were treated with tolerogen and transplanted BALB/c cardiac allograft, and IL-21 was neutralized by intravenous injection of anti-IL-21 blocking mAb, or by blocking IL-21 receptor signaling with soluble IL-21R-Ig. Both approaches prevented B cell depletion induced rejection (Figure 5A). Similarly, neutralization of IL-6 prevented B cell depletion induced allograft rejection (Figure 5B). Together these results demonstrated that blocking cytokines required for Th17 differentiation under B cell depleting conditions prevented graft rejection, supporting the notion that Th17 differentiation after B cell depletion plays a role in rejection. Furthermore, neutralization of IL-6 or IL-21 with mAb also rescued allograft survival in B-IL-10^−/− mice (Figure 5C).

Figure 5. Neutralization of IL-6 or IL-21, or deficiency of CCR6, prevents B cell depletion induced allograft rejection.

(A) C57BL/6 recipients given tolerogen, anti-mCD20 mAb, anti-IL-21 mAb (10 mg/kg i.v. on d −1 and +5), or IL-21R–Ig fusion protein (200 µg/mouse i.v. on d −1, +3 and +5), and allograft survival monitored. Anti-CD20 mAb vs. anti-CD20 mAb + anti-IL-21 mAb (p=0.0017); anti-CD20 mAb vs. anti-CD20 mAb + IL-21R–Ig (p=0.0045); (B) C57BL/6 recipients given tolerogen, and anti-mCD20 mAb with or without anti-IL-6 mAb (1 mg/mouse i.p. on d +1, +3 and +5). Allograft survival monitored. Tolerogen + anti-CD20 mAb vs. tolerogen + anti-CD20 mAb + anti-IL-6 (p=0.0003). (C) B-IL-10^−/− recipients given tolerogen, transplanted with allografts, and given anti-IL-6 mAb (1 mg/mouse i.p. on d +1, +3 and +5) or anti-IL-21 mAb (10 mg/kg i.v. on d −1 and +5). Allograft survival monitored (left) and parenchymal rejection scored (right). B-IL-10^−/− vs. B-IL-10^−/− + anti-IL-21 mAb (p=0.03); B-IL-10^−/− vs. B-IL-10^−/− + anti-IL-6 mAb (p=0.05) (D) C57BL/6 recipients given tolerogen, and anti-mCD20 mAb or control IgG, and allograft survival monitored. Donor grafts and native heart harvested on d5 and expression of CCL20 assayed by immunofluorescence microscopy. Magnification 200X (left). Mean number of CCL20⁺ cells/section, 3 grafts/group, 3–4 sections/graft (right) (E) Wild type C57BL/6 or CCR6^−/− recipients given tolerogen, and anti-mCD20 mAb, and graft survival monitored. WT + anti-CD20 mAb vs. CCR6−/− with or without anti-CD20 mAb (p<0.001). (F) C57BL/6 recipients given tolerogen, anti-mCD20 mAb, and either control IgG or mCCL20-Ig fusion protein (200 µg/mouse i.v., d +1). Graft survival monitored. DST + anti-CD40L mAb vs. DST + anti-CD40L + mCCL20-Ig (p<0.0001); DST + anti-CD40L mAb + anti-CD20 mAb vs. DST + anti-CD40L + anti-CD20 mAb + mCCL20-Ig (p, ns).

CCR6 plays an important role in the migration of Th17 cells to inflamed tissues, which in turn recruit Th1 and cytotoxic CD8⁺ T cells [38,42]. Immunohistochemical analysis of the allografts showed increased expression of the CCR6 ligand CCL20 in the rejecting grafts (Figure 5D), suggesting a role for CCR6 in recruiting Th17 cells into the rejecting allograft. To investigate the role of CCR6 in B cell depletion-induced acute rejection, CCR6^−/− mice were given tolerogen and transplanted BALB/c allograft. Deficiency of CCR6 in recipient mice prevented B cell depletion induced allograft rejection (Figure 5E). In contrast, administration of mCCL20-Ig fusion protein, an agonist for CCR6 [27], caused acute allograft rejection (Figure 5F). Together, these results demonstrated that neutralization of IL-6 or IL-21, or deficiency of CCR6, prevented B cell depletion induced acute rejection and rescued graft survival in B cell depleted recipients, while stimulation of CCR6 without B cell depletion induced acute rejection.

Discussion

In our prior work we showed that co-stimulation blockade induced a subset of IL-10⁺ MZP B cells in the secondary lymphoid organs [3], and B cells and B cell IL-10 in the MZP subset were required for co-stimulatory blockade induced, alloantigen specific tolerance. Here we extended these findings and showed that B cell function and B cell IL-10 determine the frequency and distribution of Tfr and Tfh during pro-inflammatory and anti-inflammatory responses. Depletion of B cells or B cell IL-10 resulted in decreased frequencies of Tfr and Tfh, a shift of Tfh out of the follicle, and an increase in and altered distribution of IL-21 which in conjunction with IL-6 induced Th17 and allograft rejection.

The costimulatory receptor CD40 is expressed on B cells and regulates several important processes, including isotype switching, memory B cell development and germinal center formation. Interaction of CD40 with its cognate ligand CD40L, expressed on activated T cells, leads to activation of a cascade of the transcription factors Sp1, NFkB and NFAT, which leads to production of IL-10 in B cells [43]. CD40^−/− B cells do not produce IL-10 and lack regulatory function [43]. These studies show that CD40-CD40L signaling helps in inducing IL-10 production in B cells, and IL-10 producing B cells control effector immune responses in the secondary lymphoid tissues and play important roles in the generation of tolerance [3,43].

The detrimental effects of alloantibody in transplantation are well established. Thus, blocking the generation of alloantibody by depletion of B cells has been considered an important strategy. Along with others, we showed that depletion of B cells actually leads to acute allograft rejection [2,3,44]. B cells exert regulatory functions by producing IL-10 [3,45], IL-35, FasL, and TGF-β; and these molecules can induce cell death in effector cells [5,46]. Breg can also prevent the differentiation of effector Th1 and Th17 cells, and promote the differentiation of Tregs [46]. Our present work shows that depletion of B cells increased the differentiation of Th17 in the secondary lymphoid tissues and resulted in migration of effector Th17 and CD8 T cells in a CCR6 dependent manner into the allograft to prevent tolerance [3]. These studies suggest that Breg control alloimmunity by blocking the differentiation as well as migration of effector CD4 T cells from secondary lymphoid tissues to the inflamed allograft.

The germinal center (GC) is a well demarcated transient structure present in the B cell follicles in the spleen and lymph nodes [47]. In the GC, CD4 T cells help B cells for clonal expansion, affinity maturation, isotype switching and differentiation into memory or plasma B cells. Tfh cells in the GC maximize the antigen-specific help to B cells through CD40L. Deficiency or blocking the CD40-CD40L interaction inhibits the GC reaction and helps in the maintenance of tolerance [22,48]. Treatment with anti-CD40 antibody significantly reduces GC frequency and size in rhesus macaques and prevents antibody-mediated allograft rejection [22]. These studies suggest that GCs provides an active cellular and molecular platform that can be modulated for the generation of immunity or alloantigen-specific tolerance.

Stromal fibers or the fibroblastic reticular network controls the migration of T cells in the different cellular compartments of secondary lymphoid organs. These fibers and networks not only help in the migration of T cells, B cells and APCs; but also provide a platform for cognate interactions which play important roles in shaping immunity and tolerance [49]. Reagents that disrupt the stromal fiber network prevent the generation and maintenance of transplantation tolerance [50,51]. Our present results showed that B cell depletion alters the migration and distribution of Tfh and Tfr in specific cellular compartments of the secondary lymphoid organs. These findings suggest that in tolerance induction the location of effector Tfh and regulatory T cells Tfr cells is critical, and regimens that alter the location of these cells may interfere with tolerance.

IL-21 in the germinal center is required by B cells for their development, proliferation, survival and function. IL-21 signals through the heterodimeric IL-21R plus common cytokine receptor gamma-chain by phosphorylation of transcription factor STAT3. B cell IL-21 induced STAT3 phosphorylation is required for establishment of long-lived antibody response [52]. It has been reported that IL-21 together with CD40 signaling regulates IL-10 production in B cells [53,54]. Apart from B cells, IL-21R signaling in DCs regulates CCR7 expression leading to migration of DCs from inflamed pancreatic tissue to the draining LN to present islet antigens and induce inflammatory CD4⁺ T cells [55]. IL-21 also promotes the differentiation of inflammatory Th17 cells and inhibits Treg suppressive function [56], and plays an important role in systemic lupus erythematosus, multiple sclerosis, rheumatoid arthritis, and type 1 diabetes [56]. Depletion or neutralization of IL-21 regulates systemic lupus encephalitis and rheumatoid arthritis, controls Burkitt’s lymphoma, and reduces graft-versus-host disease by inducing Treg [56]. In non-human primates, blockade of IL-21 and IL-21R prevents acute cardiac rejection by blocking the expansion of antigen-activated T cells [22,57,58]. Our results showed that neutralization of IL-21 or blockade of IL-21R prevented B cell depletion-induced graft rejection, probably by inhibiting the differentiation of Th17 and promoting Treg.

CD4⁺CD44⁺CXCR5⁺ Tfh cells present in the germinal center are the major source of IL-21. Under homeostatic or tolerogenic conditions, Tfh cells are located in the germinal center and provide help to B cells. In the absence of B cells, which are the major consumers of IL-21, our present results showed a change in the distribution of IL-21 producing Tfh cells from the B cell to the T cell zone, which redirected the availability of IL-21 and lead to increased inflammatory Th17 cells. In a non-human primate kidney transplantation model, animals with signs of antibody-mediated rejection have increased IL-21 production in the germinal center of the lymph nodes [22]. Blocking costimulatory signals with CD40 inhibits Tfh cells in the GCs and reduces IL-21 production in the follicles [22]. Our data showed that depletion of B cells or deficiency of IL-10 in B cells led to altered location of Tfh cells outside the follicle and increased expression of IL-17 and IL-21. These findings suggest that altering the distribution and availability of IL-21 in different domains of the spleen and LN by B cell depletion engaged IL-21-related inflammatory responses and induced allograft rejection.

CD4⁺ Tfr cells in the follicles share a similar phenotype with Tfh cells. Tfr cells express transcription repressor Bcl6 and dampen the GC response by controlling Tfh as well as B cells [40,59]. We showed that co-stimulatory blockade increased the IL-10 production in marginal zone precursor B cells, and these IL-10 expressing MZP B cells increased Tfr cells, controlled alloantibody production, and helped in the induction of transplantation tolerance [3]. In costimulatory blockade induced tolerance, dynamic interaction of Tfh, B cells, and Tfr in the secondary lymphoid tissue microenvironment are likely key factors in the generation of tolerance (Figure 6).

Figure 6. Role of IL-10 producing MZP B cells in tolerance and rejection.

Schematic representation of interaction of effector B cells, MZP B cells, Tfh, Teff and Tfr in the secondary lymphoid tissue during co-stimulatory blockade induced tolerance. During tolerance, IL-10 produced by MZP B cells induces the differentiation of Tfr cells which migrate into the T cell zone and control the immune response in the secondary lymphoid tissues. Treg also migrate to the site of inflammation and control the alloimmune response. In the absence of IL-10 producing MZP B cells, Tfh cells down-regulate CXCR5 and up regulate CCR7, leading to migration into the T cell zone. Tfh also produce IL-21 which together with other cytokines in the microenvironment such as IL-6 drives the differentiation of Th17. Th17 cells express CCR6 and migrate to the site of inflammation.

B cell functions are mostly associated with worse transplant outcomes due to acute and chronic antibody-mediated rejection. Randomized controlled trials with anti-CD20 mAb have been performed in renal transplantation [44,60]. In one study, the rituximab regimen did not result in any difference in T cell mediated rejection in the first 6 months; and three-year follow-up showed significantly higher mortality in the rituximab treated patients [60]. In a steroid free maintenance regimen trial, rituximab increased acute cellular rejection of allograft and the trial was halted [44]. Given the importance of non-humoral and regulatory function of B cells, untargeted depletion of B cell may not be efficacious. Our results here suggest mechanisms for how pan-B cell depletion could result in a pro-inflammatory effect with enhanced allograft rejection and interference with tolerance.

Highlights.

1. Depletion of B cells prevents co-stimulatory blockade induced transplantation tolerance.

2. Depletion of B cells or deficiency of IL-10 in B cell induces Th17 cells in spleen and lymph nodes.

3. Deficiency of IL-10 in B cells altered the location of follicular regulatory CD4 T cell (Tfr) and follicular helper CD4 T cell (Tfh) in secondary lymphoid organs.

4. Neutralization of IL-6, IL-21 or deficiency of CCR6 prevents B cell depletion induced cardiac allograft rejection in tolerogen treated mice.

Abbreviations

APC

antigen presenting cells

Breg

regulatory B cell

DC

dendritic cell

DST

donor-specific splenocyte transfusion

H&E

hematoxylin and eosin staining

i.v.

intravenous

LN

lymph node

mAb

monoclonal antibody

MOMA-1

metallophilic macrophages

MST

median survival time

MZ

marginal zone

MZP

marginal zone precursor

PR

parenchymal rejection

Tfh

T follicular helper cell

Tfr

follicular regulatory CD4⁺ T cell