Toll-like receptor 9 (TLR9) is highly expressed in B cells, and B cells are important in the pathogenesis of type 1 diabetes (T1D) development. However, the intrinsic effect of TLR9 in B cells on β-cell autoimmunity is not known. To fill this knowledge gap, we generated NOD mice with a B-cell–specific deficiency of TLR9 (TLR9fl/fl/CD19-Cre+ NOD). The B-cell–specific deletion of TLR9 resulted in near-complete protection from T1D development. Diabetes protection was accompanied by an increased proportion of interleukin-10 (IL-10)–producing B cells. We also found that TLR9-deficient B cells were hyporesponsive to both innate and adaptive immune stimuli. This suggested that TLR9 in B cells modulates T1D susceptibility in NOD mice by changing the frequency and function of IL-10–producing B cells. Molecular analysis revealed a network of TLR9 with matrix metalloproteinases, tissue inhibitor of metalloproteinase-1, and CD40, all of which are interconnected with IL-10. Our study has highlighted an important connection of an innate immune molecule in B cells to the immunopathogenesis of T1D. Thus, targeting the TLR9 pathway, specifically in B cells, may provide a novel therapeutic strategy for T1D treatment.
Introduction
Toll-like receptors (TLRs) are innate immune receptors that recognize molecular patterns of microbial pathogens. On binding to the appropriate ligands, TLRs activate signaling pathways that lead to production of proinflammatory cytokines (1,2) and upregulation of costimulatory molecules (3,4). TLR9 has been shown to be expressed on a range of immune cells, including B cells (5,6), dendritic cells (DCs) (7,8), monocytes/macrophages (9,10), T cells (11), epithelial cells (12,13), and endothelial cells (14). TLR9 recognizes guanine-cytosine–rich DNA from pathogens and self-DNA as well as the short single-stranded synthetic DNA 5′-cytosine-phosphate-guanine-3′ (CpG) (15,16). TLR9 stimulation induces downstream signaling, leading to the induction of a range of cytokines and the promotion or suppression of cell survival (17,18). TLR9 plays an important role in systemic autoimmune diseases, such as systemic lupus erythematosus (SLE), in humans and a mouse model of human SLE (19–21). In the absence of TLR9, autoimmune disease was exacerbated in a mouse model of SLE (22,23). Interestingly, a TLR9-deficient NOD mouse model of human type 1 diabetes (T1D) was protected from T1D development (24,25). This protection was partly mediated by impaired interferon-α production by DCs from the TLR9−/− mouse (25) and by enhanced expression and regulatory function of CD73+ T cells as well as by enhanced islet β-cell function (24,26).
Expression of TLRs in B cells provides a cell-intrinsic mechanism for innate signals regulating adaptive immune responses (22,27), and studies have suggested that TLR signaling plays a vital role in B-cell development and activation (6,28,29). Increasing evidence also suggests that B-cell activation by TLR9 ligands is important for optimal antibody responses to microbial antigens and DNAs released from both physiological and pathological dying cells (6,30). To address the intrinsic role of TLR9 in the function of B cells in a spontaneous autoimmune diabetes model system, we generated the NOD mouse model with a B-cell–specific deficiency of TLR9 (TLR9fl/fl/CD19-Cre+ NOD). We hypothesized that TLR9 in B cells modulates T1D development by regulating immune tolerance to islet β-cell autoimmunity. Our results showed that B-cell–specific TLR9 deficiency leads to striking protection from T1D development in NOD mice. Mechanistic studies suggested that TLR9 regulates the interleukin-10 (IL-10)–producing B cells and that TLR9 deficiency in the B cells changes the proportion and function of IL-10–producing B cells that contribute to diabetes protection. The molecular mechanism of TLR9 regulation of IL-10–producing B cells was mediated by downregulation of matrix metalloproteinase (MMP) and upregulation of tissue inhibitor of metalloproteinase (TIMP) gene expression in B cells, leading to protection from diabetes development.
Research Design and Methods
Mice
Mice used in the study were housed in specific pathogen–free conditions with a 12-h dark-light cycle at the Yale University animal facility. TLR9fl/fl C57BL/6 mouse breeders were provided by Mark Shlomchik (University of Pittsburgh) and were backcrossed to our NOD/Caj genetic background for >10 generations. CD19-Cre NOD breeders were purchased from The Jackson Laboratory (Bar Harbor, ME) and bred with our NOD/Caj mice. TLR9fl/fl and CD19-Cre NOD mice were then intercrossed to obtain B-cell–specific TLR9-deficient (TLR9fl/fl/CD19-Cre+) and control (TLR9fl/fl/CD19-Cre−) NOD mice. The total-body TLR9-deficient (TLR9−/−) NOD mice were generated as previously reported (24,26). The use of the animals in this study was approved by the institutional animal care and use committee of Yale University.
Diabetes Incidence Monitoring
Diabetes development in TLR9fl/fl/CD19-Cre+ NOD mice and TLR9fl/fl/CD19-Cre− NOD mice were monitored weekly for diabetes development by testing for glycosuria with glucose strips (Bayer, Whippany, NJ) from 10 to 32 weeks of age. Diabetes was diagnosed by two consecutive positive glycosuria tests 24 h apart and confirmed with a blood glucose reading of 250 mg/dL (13.9 mmol/L) using a FreeStyle glucose meter (Abbott, Chicago, IL).
Monoclonal Antibody Staining, Tetramer Staining, and Flow Cytometry
Mononuclear cells were harvested from spleen, pancreatic lymph nodes (PLNs), mesenteric lymph nodes (MLNs), Peyer patches, and the peritoneal cavity (PC). Cells (2–5 × 106) were first incubated with an Fc blocking antibody (CD16/32, 15 min, 4°C) before staining with antibodies against CD19, B220, CD40, T-cell receptor (TCR)-β, CD5, CD1d, CD21, and CD23 and a viability dye (all from BioLegend) for 30 min at 4°C. For intracellular cytokine staining, the cells (5 × 106) were stimulated with phosphomolybdic acid (50 ng/mL) (Sigma) and ionomycin (500 ng/mL) (Sigma) in the presence of 1 μL/mL BD GolgiPlug for 4 h. Poststimulation, the cells were stained as described above. After washing, fixation (20 min at room temperature), and permeabilization (Intracellular Fixation & Permeabilization Buffer; eBioscience), the cells were then incubated with an Fc blocking antibody (15 min, 4°C) before staining with antibodies against cytokines. Phycoerythrin-conjugated IGRP tetramer (IGRP206–214, VYLKTNVFL; National Institutes of Health Tetramer Core Facility) staining was carried out with pretitrated IGRP tetramer together with antibodies against TCR-Vβ8, CD8α, CD44, CD62L, and isotype control as well as viability dye. All the samples were analyzed with BD LSR II flow cytometry.
Cell Isolation and Cell Depletion
Splenic B cells were isolated using the EasySep Mouse B Cell Isolation Kit (STEMCELL Technologies) according to the manufacturer’s protocol. Splenic CD4+ T cells were isolated by removing CD8+ T cells and antigen presenting cells (APCs). The splenocytes were incubated with hybridoma supernatants containing monoclonal antibodies (mAbs) to CD8 (TB105) and MHC-II (10.2.16) for 30 min at 4°C. After washing with PBS, the cells were further incubated for 45 min on ice with magnetic beads conjugated with goat anti-mouse IgG and IgM (to remove B cells) or goat anti-rat IgG to remove CD8+ and MHC-II+ cells (all the beads were from QIAGEN). CD4+ T cells were then separated using a magnetic plate. The purity was routinely 90–95%, as verified by flow cytometry. To deplete splenic regulatory T cells (Tregs), we removed CD4+CD25+ T cells by EasySep Mouse Regulatory T Cell Isolation Kit (STEMCELL Technologies), and the unbound cells were the Treg-depleted splenocytes that were used for adoptive transfer.
Cell Culture and Transwell Coculture System
Purified B cells (105/well) were stimulated with anti-CD40, with or without anti-IgM or different TLR ligands, for 2 days, and the proliferative response was determined by 3H-thymidine incorporation. In a coculture system, purified B cells, cocultured with purified T subsets from TLR9fl/fl/CD19-Cre− NOD mice, were stimulated with anti-CD3 and anti-CD28. In a transwell coculture system, purified B cells (105/well) from TLR9fl/fl/CD19-Cre+ NOD mice or TLR9fl/fl/CD19-Cre− NOD mice were added to the insert chambers; purified T cells (105/well) from wild-type NOD mice, together with T-cell–depleted APCs (105/well), were plated in the bottom chamber in the presence of anti-CD3 and anti-CD28 (final concentration 1 μg/mL). For the IL-10 blocking transwell culture, anti-IL-10 receptor mAb (1B1.3A) or control rat IgG was added into the culture (insert chamber). All the cultures were incubated for 2 days in triplicate. 3H-thymidine was then added, and cell culture supernatants were collected. Cells were incubated for a further 18 h before harvesting. Proliferation was determined by 3H-thymidine incorporation using a β-counter and presented as counts per minute.
ELISA
Different isotypes of serum Igs were assessed using direct ELISA (Southern Biotech), and the concentrations were determined by standard curve for each isotype (Southern Biotech). IL-10 content of the cell culture supernatants and in serum was determined using the ELISA MAX Standard Set Mouse IL-10 kit (BioLegend) on the basis of the manufacturer’s protocol.
Adoptive Transfer
Purified splenic T cells (5 × 106) from diabetic NOD mice, together with sorted splenic B cells (5 × 106) from 6–7-week-old TLR9fl/fl/CD19-Cre+ or TLR9fl/fl/CD19-Cre− NOD mice, were injected (i.v.) into 5–6-week-old female Rag−/− NOD mice. Another set of adoptive transfer experiments was carried out by transferring Treg-depleted splenocytes (10 × 106) from 6–7-week-old TLR9fl/fl/CD19-Cre+ or TLR9fl/fl/CD19-Cre− NOD mice into 5–6-week-old female Rag−/− NOD mice. All the recipients were monitored weekly for glycosuria, and the experiments were terminated 12–14 weeks after the cell transfer unless the mice developed diabetes, which was confirmed by blood glucose >250 mg/dL (13.9 mmol/L).
Microarray Hybridization and Data Analysis
Total RNA was extracted from sorted splenic B cells from TLR9fl/fl/CD19-Cre+ or TLR9fl/fl/CD19-Cre− NOD mice (7–8-week-old females) with RNeasy Mini Kit (QIAGEN). Complementary RNA synthesis and whole-genome Illumina microarray chip analysis were carried out at the Yale Center for Genomic Analysis. Gene expression data were normalized to control probes and performed using Partek Genomics Suite 6.6, and any outliers were removed during quality control.
Quantitative Real-Time PCR
Splenic B cells from TLR9fl/fl/CD19-Cre+ or TLR9fl/fl/CD19-Cre− NOD mice (7–8 weeks old) were purified using EasySep Mouse B Cell Isolation Kit, and total cellular RNA was isolated with RNeasy Mini Kit (QIAGEN). cDNA was synthesized using SuperScript III First-Strand Synthesis Kit with random hexamers (Invitrogen). Quantitative real-time PCR (qPCR) was performed with a Bio-Rad iQ5 qPCR detection system. The relative mRNA levels of IL-10, CD40, TIMP-1, MMP19, and MMP9 were determined using the 2−ΔΔCt method by normalization with the housekeeping gene GAPDH.
Statistical Analysis
Statistical analysis was performed using GraphPad Prism 7 software. Diabetes incidence was compared using log-rank test. In vitro assays were analyzed with Student t test or ANOVA, and P < 0.05 was considered to be statistically significant.
Data and Resource Availability
The data sets generated and/or analyzed during the current study are available from the corresponding author upon request. The rodent models, antibodies, compounds, or other noncommercial reagents generated and used during the current study are available from the corresponding author upon request.
Results
B-Cell–Specific TLR9 Deficiency Protects Against T1D Development
We have previously reported that total-body TLR9-deficient NOD mice are protected from T1D development (24,26). TLR9 is highly expressed in B cells and has been shown to play an important role in systemic autoimmune disorders, such as SLE (31,32). To determine the intrinsic role of TLR9 in B cells in immune pathogenesis of organ-specific autoimmunity, such as T1D, we first investigated spontaneous diabetes development in TLR9fl/fl/CD19-Cre− and TLR9fl/fl/CD19-Cre+ NOD mice. Interestingly, B-cell–specific deletion of TLR9 led to striking protection from T1D development (Fig. 1A) and insulitis (Supplementary Fig. 1).
TLR9 Intrinsically Regulates the Frequency and Function of B-Cell Subsets
To investigate the effect of TLR9 deletion on B cells in a steady state, we evaluated the phenotype and function of the B cells from 6–8-week-old female TLR9fl/fl/CD19-Cre+ and TLR9fl/fl/CD19-Cre− NOD mice ex vivo (Supplementary Fig. 2, gating strategy). There was a significant increase in CD1d+CD5+ B cells in the PLNs and MLNs from TLR9fl/fl/CD19-Cre+ NOD mice compared with TLR9fl/fl/CD19-Cre− control mice (Fig. 1B and C). CD1d+CD5+ B cells are often also IL-10–producing B-regulatory cells (B10-regs) (33,34). Next, we assessed the B10-regs and found that they were increased in all the lymphoid tissues examined from TLR9fl/fl/CD19-Cre+ NOD mice compared with TLR9fl/fl/CD19-Cre− controls (Fig. 1B and D). However, a statistically significant increase in the B10-regs was found only in the PC of TLR9fl/fl/CD19-Cre+ NOD mice (Fig. 1D). Interestingly, more IL-10–producing B cells from TLR9fl/fl/CD19-Cre+ NOD mice expressed CD40 compared with TLR9fl/fl/CD19-Cre− mice, specifically in the PC (Fig. 1E). The absolute cell numbers of different B-cell subsets were also changed accordingly (Supplementary Fig. 3A–E). We also evaluated the B10-regs in 30–32-week-old nondiabetic female mice and found a similar phenotype (Fig. 1F–H).
Phenotypic analysis of young mice also revealed a significant increase in marginal zone (MZ) B cells (CD21highCD23low) in PLNs of TLR9fl/fl/CD19-Cre+ NOD mice compared with the controls (Supplementary Fig. 4A), whereas transitional B cells (CD21highCD23high) in PLNs of TLR9fl/fl/CD19-Cre+ NOD mice were significantly reduced (Supplementary Fig. 4B). Akin to the phenotypic changes in TLR9fl/fl/CD19-Cre+ NOD mice at a young age, the altered MZ and transitional B cells in PLNs were also seen in the older nondiabetic female mice at 30–32 weeks (Supplementary Fig. 4C and D). Interestingly, unlike the young mice, MZ B cells (CD21highCD23low) from the spleen and MLNs of the older mice were decreased in TLR9fl/fl/CD19-Cre+ NOD mice compared with the control TLR9fl/fl/CD19-Cre− NOD mice (Supplementary Fig. 4C), whereas transitional B cells (CD21highCD23high) from the PC of old mice were increased in TLR9fl/fl/CD19-Cre+ NOD mice compared with controls (Supplementary Fig. 4D). It is noteworthy that the changes of B cells in PLNs remained stable regardless of the age of the mice. In addition, in relation to IgM and IgD B-cell subsets, we showed reduced IgM+IgD− B cells and increased IgM+IgD+ B cells in the spleen (Supplementary Fig. 5A and B). B-cell–specific TLR9 deficiency did not affect class switching overall; however, TLR9fl/fl/CD19-Cre+ NOD mice showed a significant reduction of IgG, which was attributed to the reduction of IgG1 (Supplementary Fig. 6). Phenotypic analysis of splenic plasmacytes revealed that there was no difference in Blimp1+ plasmacytes between TLR9fl/fl/CD19-Cre+ NOD mice and control mice (Supplementary Fig. 7A and B), but Blimp1− plasmacytes were significantly reduced in TLR9fl/fl/CD19-Cre+ NOD mice (Supplementary Fig. 7C). Taken together, our data suggest that B-cell–specific deletion of TLR9 significantly changes the proportion of B-cell subsets in the various lymphoid organs, which may lead to an immune tolerant state and contribute to diabetes protection.
B Cells Without TLR9 Are Hyporesponsive to Both Innate and Adaptive Immune Stimuli
To determine the intrinsic effect of TLR9 on B-cell function, we stimulated splenocytes from TLR9fl/fl/CD19-Cre+ and TLR9fl/fl/CD19-Cre− NOD mice (6–8 weeks) with anti-CD40 (adaptive stimulus) and innate stimuli, including Pam3Csk4, lipopolysaccharide (LPS), and CpG (TLR2, TLR4, and TLR9 agonists, respectively). Splenocytes from TLR9fl/fl/CD19-Cre+ NOD mice showed significantly impaired responses to all the stimuli, regardless of whether the cells were stimulated through adaptive or innate pathways (Fig. 2A–D). To verify that the hyporesponsiveness was due to TLR9 deficiency in B cells, we repeated the experiments using purified splenic B cells from TLR9fl/fl/CD19-Cre+ and TLR9fl/fl/CD19-Cre− NOD mice. We found similarly impaired responses to both innate immune and adaptive stimuli, including costimulation of anti-CD40 and anti-IgM (Fig. 2E–I). To assess whether the hyporesponse was accompanied by immunoregulatory cytokines, we measured IL-10 in the supernatants of the anti-CD40–stimulated cultures. Interestingly, we found a significant increase in IL-10 production in the culture supernatants in which B cells from TLR9fl/fl/CD19-Cre+ were used (Fig. 2J). Moreover, the impaired immune response of the TLR9-deficient B cells was found in the old (30–32 weeks) TLR9fl/fl/CD19-Cre+ NOD mice (Fig. 3A–E). IL-10 in the supernatants of anti-CD40–stimulated cultures also was significantly increased in old (30–32 weeks) TLR9fl/fl/CD19-Cre+ NOD mice (Fig. 3F). Taken together, our data suggest that B-cell–specific deficiency in TLR9 promotes a B-cell hyporesponsive immune tolerant state.
TLR9 Regulates the Fate of IL-10–Producing B Cells and Suppresses T-Cell Responses
To determine whether B-cell–specific deficiency of TLR9 also led to hyporesponsiveness of T cells, we stimulated purified T cells from wild-type NOD mice with anti-CD3 and anti-CD28 mAbs in the presence of B cells from TLR9fl/fl/CD19-Cre+ or TLR9fl/fl/CD19-Cre− NOD mice. Similar to the hyporesponsiveness seen in TLR9-deficient B cells, wild-type NOD splenic T cells also showed impaired response to anti-CD3 and anti-CD28 stimulation when TLR9-deficient B cells were used for the antibody cross-linking (Fig. 4A). Because TLR9-deficient B cells secreted a higher level of IL-10 after stimulation (Fig. 2J), we hypothesized that activated B cells from TLR9fl/fl/CD19-Cre+ NOD mice would have increased regulatory function mediated by IL-10. To test our hypothesis, we performed two sets of experiments using anti-CD40–activated B cells from TLR9fl/fl/CD19-Cre+ and TLR9fl/fl/CD19-Cre− NOD mice, both in coculture and in transwell culture. For the former, we cultured purified B cells with and without anti-CD40, together with purified T cells from wild-type NOD mice, in the presence of anti-CD3 and anti-CD28. Similar to the results shown in Fig. 4A, the T-cell response was reduced in the presence of stimulated B cells from TLR9fl/fl/CD19-Cre+ mice (data not shown). To determine whether IL-10 played a role in the suppressed T-cell response, we measured the secreted IL-10 in the culture supernatants. Similar to the earlier results (Fig. 2J), we detected a higher amount of secreted IL-10 in the cultures where TLR9-deficient B cells were present (Fig. 4B). To determine whether the suppressed T-cell response was indeed due to B cells from TLR9fl/fl/CD19-Cre+ mice rather than cell contact between B cells and T cells, we repeated the experiment using a transwell system (research design and methods). We first purified splenic B cells from TLR9fl/fl/CD19-Cre+ and TLR9fl/fl/CD19-Cre− NOD mice and stimulated with anti-CD40 overnight. After washing, we placed the preactivated B cells in the transwell inserts, with the splenic T cells and B-cell–depleted APCs from wild-type NOD mice seeded in the lower wells, in the presence of anti-CD3 and anti-CD28. In this culture system, there was direct cell-cell contact between T cells and APCs from wild-type NOD mice, while B cells from TLR9fl/fl/CD19-Cre+ and TLR9fl/fl/CD19-Cre− NOD mice were separated from T cells by the transwell insert. We assessed T-cell proliferation by 3H-thymidine incorporation and the content of secreted IL-10 in the supernatant. In line with the coculture results, we found impaired T-cell proliferation when the preactivated B cells in the transwell inserts were from TLR9fl/fl/CD19-Cre+ NOD mice (Fig. 4C). Again, T-cell hyporesponse was accompanied by higher IL-10 secretion by B cells from TLR9fl/fl/CD19-Cre+ NOD mice (Fig. 4D). To further confirm the role of IL-10 in suppressing the T-cell response, we blocked IL-10 signaling using mAb to the IL-10 receptor in the transwell system. As expected, blocking IL-10 signaling significantly enhanced T-cell proliferation, whereas T-cell responses were impaired in the presence of control IgG (Fig. 4E). It is noteworthy that blockade of the IL-10 receptor also enhanced T-cell proliferation when preactivated B cells from TLR9fl/fl/CD19-Cre− NOD mice were present compared with control IgG (Fig. 4E). To determine whether the levels of IL-10 in the systemic circulation were also altered by the absence of B-cell–specific TLR9, we measured IL-10 in the serum samples of TLR9fl/fl/CD19-Cre+ and TLR9fl/fl/CD19-Cre− NOD mice. Interestingly, we found a significant increase in circulating IL-10 in TLR9fl/fl/CD19-Cre+ NOD mice compared with TLR9fl/fl/CD19-Cre− NOD mice, regardless of age (Fig. 4F, 6–8 weeks, and Fig. 4G, 30–32 weeks). Taken together, our results demonstrate that B-cell TLR9 deficiency converted B cells to B10-regs that suppressed T-cell responses.
TLR9 Deficiency in B Cells Confers Protection Against T1D Development
To determine whether TLR9-deficient B cells induce suppression of islet β-cell autoimmunity, we first assessed whether B-cell–specific TLR9 deficiency dampened autoreactive T cells, especially cytotoxic CD8+ T cells, by staining for IGRP-tetramer–specific CD8+ T cells. Our results showed that the absence of B-cell–specific TLR9 did not have a noticeable effect on CD8 T cells, including IGRP-tetramer–specific CD8+ T cells (Supplementary Fig. 8A–C). Next, we assessed whether B-cell–specific TLR9 deficiency affected the proportion and the function of Tregs. Interestingly, we found that the absence of B-cell–specific TLR9 significantly increased the proportion of Tregs (Supplementary Fig. 9). To test whether the increased Tregs contribute to diabetes protection, we adoptively transferred splenoctyes from TLR9fl/fl/CD19-Cre+ and TLR9fl/fl/CD19-Cre− NOD mice to Rag−/− NOD recipients after depleting Tregs. In the absence of Tregs, diabetes development in the Rag−/− NOD recipients that received splenoctyes from TLR9fl/fl/CD19-Cre+ mice was significantly delayed and reduced compared with the Rag−/− NOD recipients that received splenoctyes from TLR9fl/fl/CD19-Cre− mice (Fig. 5A). These results suggest that diabetes suppression in TLR9fl/fl/CD19-Cre+ NOD mice is not due to Tregs, despite their increase. To determine whether TLR9-deficient B cells confer diabetes suppression, we carried out a further set of adoptive transfers of purified splenic T cells from diabetic TLR9fl/fl/CD19-Cre− NOD mice (wild type), together with purified splenic B cells from nondiabetic TLR9fl/fl/CD19-Cre+ or TLR9fl/fl/CD19-Cre− NOD mice, into Rag−/− NOD recipients. Diabetes development was significantly reduced in the Rag−/− NOD recipients that received B cells from TLR9fl/fl/CD19-Cre+ mice compared with the Rag−/− NOD recipients that received B cells from TLR9fl/fl/CD19-Cre− NOD mice (Fig. 5B). To investigate whether IL-10 played a role in diabetes suppression, we isolated splenic B cells from the Rag−/− NOD recipients and assessed IL-10 gene expression, with or without anti-CD40 stimulation, by qPCR. In line with the results shown above, we found significantly higher IL-10 gene expression in the B cells from the Rag−/− NOD recipients that received B cells from TLR9fl/fl/CD19-Cre+ mice compared with the Rag−/− NOD recipients that received B cells from TLR9fl/fl/CD19-Cre− NOD mice (Fig. 5C). This was found both with and without anti-CD40 stimulation, with significantly higher IL-10 gene expression in TLR9-deficient B cells compared with TLR9-sufficient B cells. Our data suggest that B cells from TLR9fl/fl/CD19-Cre+ NOD mice exert a stronger immunosuppressive function mediated by the higher levels of the anti-inflammatory cytokine IL-10.
TLR9 Regulates the Cross Talk Between TIMP-1 and IL-10 in B Cells
To determine the molecular mechanism by which IL-10 is enhanced in B10-regs in the absence of TLR9, we performed microarray analysis using purified splenic B cells, directly ex vivo, from TLR9fl/fl/CD19-Cre+ and TLR9fl/fl/CD19-Cre− NOD mice. Among three downregulated genes, we found that the gene expression of MMP19 was strikingly lower in the splenic cells from TLR9fl/fl/CD19-Cre+ mice (Fig. 5D and Table 1), which was confirmed by qPCR (Fig. 5E). Using Ingenuity Pathway Analysis (IPA) (QIAGEN), we identified that TLR9 and IL-10 interacted with MMP19 and that TIMPs could modulate these interactions (Fig. 5F). In addition to their role in extracellular matrix remodeling, TIMPs also have cytokine functions affecting cell differentiation, apoptosis, and cell growth (35,36). Because TIMPs inhibit the activities of MMPs (37,38), downregulation of MMPs may suggest upregulation of TIMPs. To determine whether TIMP-1 is upregulated in TLR9-deficient B cells, with or without stimulation, we assessed the expression level of TIMP-1 by qPCR. Indeed, TIMP-1 was highly upregulated in the activated B cells from TLR9fl/fl/CD19-Cre+ mice compared with their counterparts from TLR9fl/fl/CD19-Cre− NOD mice (Fig. 5G). The significant upregulation of TIMP-1 was also seen in nonactivated TLR9-deficient B cells, although the scale was lower compared with the activated B cells (Fig. 5G).
Gene symbol . | Reference sequence . | TLR9-sufficient B- cell gene expression . | TLR9-deficient B-cell gene expression . | Fold change . | P value . |
---|---|---|---|---|---|
Gm20257 | NR_045007 | 7.07757 | 5.510315 | 0.33745 | 6.40E-07 |
Mmp19 | NM_001164197 | 8.27076 | 4.308635 | 0.064163 | 1.38E-06 |
Tlr9 | NM_031178 | 9.067313 | 5.851875 | 0.107661 | 1.68E-06 |
Pde3b | NM_011055 | 7.112163 | 8.42422 | 2.482953 | 2.63E-06 |
Snora30 | NR_034045 | 6.635437 | 8.76151 | 4.365277 | 3.91E-06 |
Gene symbol . | Reference sequence . | TLR9-sufficient B- cell gene expression . | TLR9-deficient B-cell gene expression . | Fold change . | P value . |
---|---|---|---|---|---|
Gm20257 | NR_045007 | 7.07757 | 5.510315 | 0.33745 | 6.40E-07 |
Mmp19 | NM_001164197 | 8.27076 | 4.308635 | 0.064163 | 1.38E-06 |
Tlr9 | NM_031178 | 9.067313 | 5.851875 | 0.107661 | 1.68E-06 |
Pde3b | NM_011055 | 7.112163 | 8.42422 | 2.482953 | 2.63E-06 |
Snora30 | NR_034045 | 6.635437 | 8.76151 | 4.365277 | 3.91E-06 |
Data are presented as described in research design and methods.
To further establish the cross talk of TLR9 with TIMP-1/MMP19 and IL-10, we stimulated splenic B cells from wild-type NOD mice with CpG and assessed the gene expression level of TIMP-1, MMP19, and IL-10 by qPCR. Unstimulated B cells were used as controls. In support of the results shown earlier, CpG stimulation resulted in lower gene expression of TIMP-1 and IL-10 (Fig. 6A and B) but higher expression of MMP19 (Fig. 6C) compared with the controls. Interestingly, the expression level of CD40 paralleled IL-10 (Fig. 6D), supporting our earlier results (Fig. 1H). Our results suggest novel cross talk between TLR9 and TIMP-1/MMP19, which regulates IL-10 in B cells.
Discussion
Because TLR9 is pivotal in the development of T1D (26) and is primarily expressed in B cells and plasmacytoid DCs (39,40), we aimed to clarify the role of TLR9 in B cells in development of spontaneous T1D. Our key finding was that B-cell TLR9 plays a critical role in immune tolerance to islet β-cell autoimmunity because TLR9 deficiency in B cells strongly protected NOD mice from T1D development. TLR9 modulated T1D susceptibility through different pathways, which included effects on B-cell subsets, including MZ B-cell, transitional B-cell, and B10-reg subsets. In the absence of TLR9 in B cells, these B-cell subsets, particularly B10-regs, were enhanced both in frequency and in function. At the molecular level, we discovered a novel regulatory network between TLR9 and TIMP-1, MMP19, CD40, and the IL-10 signaling pathway. Our study confirmed an important function of B cells in influencing immune tolerance whereby increased immune tolerance was observed in the absence of B-cell TLR9.
B10-regs are important in health and disease, both in humans and in mouse models (41–44), and B10 cells are considered to be an indispensable contributor to the maintenance of tolerance and immune homeostasis. Previous studies suggested that CpG, a TLR9 agonist, induces B10-regs (45). Interestingly, we found that the proportions CD1dhighCD5high B cells and CD1dhighCD5high B10-regs in NOD mice are enhanced when B cells are deficient in TLR9. Moreover, the increase of B10-regs found in our study was age independent. B-cell–specific deficiency in TLR9 also altered the proportions of MZ and transitional B cells, namely increased MZ B cells and decreased transitional B cells. While we have confirmed that TLR9 signaling plays an important role in the formation of different effector B-cell subsets (46,47), the role of TLR9 in B cells goes beyond this. Our study demonstrated that B-cell–specific deficiency of TLR9 leads to an immune tolerant state, supported by B-cell hyporesponsiveness to both innate and adaptive stimuli, as well as an impaired ability to promote optimal T-cell responses. Moreover, B cells with B-cell–specific TLR9 deletion were able to suppress both spontaneous and induced T1D development. All these findings were associated with IL-10 production of B cells, regulated by TLR9.
Interestingly, molecular analysis of TLR9-deficient and -sufficient B cells revealed only a limited number of gene expression differences, and the expression of MMP19 was lowest in TLR9-deficient B cells compared with TLR9-sufficient B cells. The MMP gene family includes >20 genes that regulate the breakdown of extracellular matrix in normal physiological and disease processes (48,49). The expression of MMP and its inhibitor TIMP are essential in tissue remodeling and cell signaling (38,50). An earlier study showed that MMP19 is expressed in synovial blood vessels and in circulating T cells of patients with rheumatoid arthritis (51).
Other studies have reported that TIMPs may play a role in the pathological processes in multiple sclerosis in humans (52,53) and experimental autoimmune encephalomyelitis in animal models (54,55). Furthermore, Guedez et al. (35) reported that TIMP-1 regulates IL-10 in B-cell differentiation in lymphoma cell lines. However, little is known about the role of MMPs in the function of TLR9 and B-cell regulation. Using IPA, we identified the network of TLR9-IL-10 with MMP and TIMP-1 in B cells. TLR9-deficient B cells have a marked reduction of MMP19 but highly increased TIMP-1, with or without activation. It is interesting that the expression of CD40 in TLR9-deficient B cells was positively associated with the expression of IL-10 and TIMP-1 expression but negatively associated with the expression of MMP19. Sanchooli et al. (56) reported an elevation of soluble CD154 (CD40 ligand) and TIMP-1 in patients with multiple sclerosis. Our study suggests a novel cross-talk loop of TLR9, MMPs/TIMP-1, IL-10, and CD40. Our results also demonstrate that TLR9 plays a vital role in this network, which leads to immune tolerance to islet β-cell autoimmunity. Thus, TLR9 in B cells acts as a crucial modulator of T1D susceptibility.
B-cell depletion immunotherapy has been used for treating different autoimmune disorders with some success (57,58). However, B-cell depletion also eliminates B10-regs, which are important in immune tolerance to autoimmunity. Our study provides preclinical information for selective molecular therapy by targeting TLR9 in B cells.
This article contains supplementary material online at https://doi.org/10.2337/figshare.13177016.
J.A.P. and J.P. contributed equally to the work.
Article Information
Funding. This work was supported by Juvenile Diabetes Research Foundation International postdoctoral research fellowship PDF-2016-197 (to J.A.P.), National Institute of Diabetes and Digestive and Kidney Diseases grants DK-092882, DK-100500, HD-097808, and P30-DK-945735 (to L.W.), and American Diabetes Association grant 1-14-BS-222 (to L.W.).
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. S.S., J.A.P., and J.P. analyzed the data and conducted experiments. S.S., J.A.P., F.S.W., and L.W. wrote and edited the manuscript. S.S., J.A.P., and L.W. designed the experiments. Y.H., J.H., Y.X., and L.Z. conducted some of the experiments. Y.Z. and H.Z. performed bioinformatics analysis on microarray. L.C. supervised part of the study. L.W. conceived the project. L.W. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.