IL-10-producing regulatory B cells restrain the T follicular helper cell response in primary Sjögren’s syndrome

Abstract

Increased numbers of T follicular helper (Tfh) cells have been implicated in the development of autoimmune diseases including primary Sjögren’s syndrome (pSS), but how the Tfh cell response is regulated during autoimmune pathogenesis remains largely unclear. Here, we first found negative correlations between IL-10⁺ regulatory B (Breg) cell numbers and Tfh cell responses and disease activity in patients with pSS and mice with experimental Sjögren’s syndrome (ESS). Moreover, we detected high expression of IL-10 receptor on Tfh cells and their precursors in both humans and mice. In culture, IL-10 suppressed human and murine Tfh cell differentiation by promoting STAT5 phosphorylation. By using an adoptive transfer approach and two-photon live imaging, we found significantly increased numbers of Tfh cells with enhanced T cell homing into B cell follicles in the draining cervical lymph nodes of RAG-2−/− mice transferred with IL-10-deficient B cells during ESS development compared with those of RAG-2−/− mice transferred with wild-type B cells. In ESS mice, CD19⁺CD1d^hiCD5⁺ Breg cells with decreased IL-10 production exhibited severely impaired suppressive effects on T cell proliferation. Consistently, CD19⁺CD24⁺CD38^hi Breg cells from pSS patients showed significantly reduced IL-10 production with defective inhibitory function in the suppression of autologous Tfh cell expansion. Furthermore, the adoptive transfer of IL-10-producing Breg cells markedly suppressed the Tfh cell response and ameliorated ESS progression in ESS mice. Together, these findings demonstrate a critical role for IL-10-producing Breg cells in restraining the effector Tfh cell response during pSS development.

Introduction

Primary Sjögren’s syndrome (pSS) is a progressive autoimmune disease characterized by exocrinopathy involving the lacrimal and salivary glands (SG), leading to the clinical manifestations of severe dry-eye and dry-mouth symptoms.¹ In the development of pSS, T cell-mediated tissue pathology has long been recognized as a result of the predominantly CD4⁺ T cell infiltrate in salivary lymphoepithelial lesions.^2,3 Moreover, a highly activated CD4⁺ T cell signature has been reported in the peripheral blood of patients with pSS,⁴ and among these cells, significantly increased numbers of Th17 cells and T follicular helper (Tfh) cells have been identified.^5–7 Using an established mouse model of experimental SS (ESS) that highly recapitulates key features of human pSS, we previously identified a pathogenic function of Th17 cells during ESS development.^8,9 Recently, a pivotal role for Tfh cells was demonstrated, as revealed by the findings that mice specifically deficient in Tfh cells exhibit markedly attenuated autoantibody production and disease pathology upon ESS induction.¹⁰ Tfh cells have been shown to be closely involved in the development of autoimmune diseases,^11,12 including systemic lupus erythematosus and new-onset rheumatoid arthritis, by migrating into follicles to provide robust B cell help and promote pathogenic autoantibody production.^13,14 Although increasing evidence indicates a role for B cell-derived inducible costimulatory ligand in promoting Tfh cell differentiation,^15,16 whether B cells can also regulate Tfh cell tolerance in pSS remains largely elusive.

Interleukin 10 (IL-10) has been identified as a potent regulator of T cells.^17,18 Early studies have suggested a critical role for IL-10-producing regulatory B (Breg) cells in maintaining immune tolerance in both inflammatory reactions and autoimmune pathogenesis.^18,19 In particular, human circulating IL-10⁺ B cells, which are enriched in the CD19⁺CD24⁺CD38^hi Breg cell population, perform a regulatory function by limiting Th1 and Th17 cell differentiation in culture.²⁰ Interestingly, impaired IL-10 production in B cells has been reported in the peripheral blood of patients with SLE or RA.^21–23 Notably, several studies, including our recent findings, have highlighted the therapeutic potential of adoptively transferred IL-10-producing Breg cells with a CD1d^hiCD5⁺ phenotype in the suppression of autoimmune inflammation and disease pathology in animal models of RA and SLE, mainly via the inhibition of Th1 and Th17 cell-mediated responses in an IL-10-dependent manner.^24,25 Although IL-10 receptor (IL-10R) signaling has been reported to skew the differentiation of tonsillar Tfh cells into T follicular regulatory (Tfr) cells in healthy children,²⁶ lymphocytic choriomeningitis virus-infected mice with an IL-10R deficiency have lower frequencies of virus-specific Tfh cells than infected wild-type mice.²⁷ Thus, it remains to be investigated whether and how IL-10⁺ Breg cells are involved in regulating the Tfh cell response in the pathogenesis of pSS.

Here, we found significantly reduced IL-10⁺ B cell numbers in both pSS patients and ESS mice, which negatively correlated with enhanced Tfh cell responses and increased disease activity. In culture, human and murine IL-10-producing Breg cells potently suppressed Tfh cell differentiation via STAT5 phosphorylation. In ESS mice, IL-10 deficiency resulted in markedly increased follicular homing by CD4⁺ T cells and Tfh cell expansion in the draining cervical lymph nodes (CLN) during disease progression. Furthermore, the adoptive transfer of in vitro-expanded Breg cells effectively suppressed the Tfh cell response and ameliorated disease pathology in ESS mice in an IL-10-dependent manner. Together, our findings have demonstrated a critical role for IL-10-producing Breg cells in restraining the Tfh cell response in the pathogenesis of pSS, which may facilitate the development of new therapeutic strategies for treating patients with pSS.

Materials and methods

Mice

For C57BL/6 mice, female wild-type (WT), IL-10−/−, recombination activating gene 2 (RAG-2)−/− (CD45.2) and B6.SJL-Ptprca Pepcb/BoyJ (CD45.1) mice between 6 and 8 weeks of age were purchased from The Jackson Laboratory (Bar Harbor, ME, USA). Nonobese diabetic (NOD) mice were also purchased from The Jackson Laboratory, and all NOD mice used between 3 and 8 weeks of age had not yet developed diabetes, as determined by measuring blood glucose levels. The RAG-2−/− mice were maintained under specific pathogen-free conditions. All animal experiments were performed following an institutionally approved protocol in accordance with the guidelines of the Committee on the Use of Live Animals in Teaching and Research of the University of Hong Kong (Reference No. 4800-18, 4712-18 and 3618-15).

ESS induction in mice

The experimental SS (ESS) model was induced in 8-week-old female mice as we previously described.^8,28 In brief, proteins extracted from the bilateral salivary glands (SG) of normal mice were emulsified in an equal volume of Freund’s complete adjuvant (Sigma-Aldrich) at a concentration of 2 mg/mL. For disease induction, each mouse received subcutaneous injections on the back with 0.2 mL of the emulsion on day 0. On day 14, a booster injection was carried out with a dose of 1 mg/mL SG proteins emulsified in Freund’s incomplete adjuvant (Sigma-Aldrich).

Histological assessment and evaluation of disease stages during ESS development

SG tissue removed from immunized mice was frozen in OCT compound (Sakura), and sections were cut at 5 μm thickness. For histopathological scoring, whole tissue sections stained with hematoxylin and eosin (H&E) were assessed by NanoZoomer (Hamamatsu) with the infiltrated area measured in accordance with the EULAR consensus guidelines.²⁹ After SG protein immunization, mice in the early disease stage were identified by reduced saliva secretion and negative histology for glandular infiltration (HS = 0) but serological positivity for anti-SSA autoantibody production; mice in the acute ESS stage exhibited mild infiltration in the submandibular gland (0 < HS ≤ 2), while the development of multiple foci in the exocrine glands was categorized as chronic stage (HS = 3).

Mouse cell culture

For murine CD4⁺ T and CD19⁺ B cell isolation, total or naive CD4⁺ T cells were purified by CD4 MicroBeads (Miltenyi Biotec) and the naive CD62L⁺ CD4⁺ T Cell Isolation Kit, respectively, while B cells were purified by B220 Microbeads (Miltenyi Biotec). Enriched cells were resuspended in culture medium at a density of 5 × 10⁶/mL and subjected to purification by sorting using a BD Influx™ cell sorter (BD Biosciences). Purified murine naive CD4⁺ T cells were seeded in a culture plate precoated with anti-CD3 (clone 17A2, BioLegend) and anti-CD28 (clone E18, BioLegend) antibodies (1 μg/mL). Cells were cultured in RPMI 1640 medium containing l-glutamine and NaHCO₃ (Sigma-Aldrich) supplemented with 100 U/mg/mL penicillin/streptomycin and 10% FCS (Thermo Fisher Scientific). For Tfh polarization, recombinant murine IL-6 (25 ng/mL) and IL-21 (10 ng/mL, R&D Systems) and 5 μg/mL anti-IFN-γ (clone AN-18), anti-IL-4 (clone 11B11) and anti-TGF-β (clone TW7-20B9) neutralizing antibodies (BioLegend) were added to a culture with a cell density of 2 × 10⁶/mL. In some experiments, recombinant mouse IL-10 (240 ng/mL), IL-10 in combination with a STAT5 inhibitor (iSTAT5, N′-((4-Oxo-4H-chromen-3-yl) methylene) nicotinohydrazide, 10 μm) or an anti-IL-10Rβ antibody (10 μg/mL, clone MM0360-8R24) was added into cultures of naive CD4⁺ T cells (2 × 10⁶) for Tfh differentiation. Furthermore, naive CD4⁺ T cells (2 × 10⁶) were cultured alone or cocultured with sorted CD19⁺CD1d^hiCD5⁺ Breg cells (0.5 × 10⁶) from WT or IL-10−/− mice for 72 h before flow cytometric analysis. The suppression rate was calculated by the following formula: % Suppression = (T cell culture alone − T cell coculture)/T cell culture alone × 100%.

ELISA

For autoantibody detection, peptides with sequences of the second extracellular loop of the murine M3 muscarinic receptor (M3R^2nd, QYFVGKRTVPPGECFIQFLSEP) and SSA/Ro60^274-290 (QEMPLTALLRNLGKMT) peptides were synthesized (SBS Genetec Co., Ltd). Serum levels of anti-M3R^2nd and anti-SSA autoantibodies were determined by ELISA as previously described.²⁸

Immunofluorescence microscopy

For immunofluorescence microscopy, frozen sections of salivary or labial gland tissue were stained with monoclonal antibodies against mouse IgG, CD4, GL-7, CD19, B220 and CXCR5 (BioLegend), while nuclei were counterstained with Hoechst 33258 (Calbiochem). A rat IgG antibody was used for control staining. The tissue sections were also incubated with a rabbit anti-mouse AQP-5 antibody (Millipore) followed by a goat anti-rabbit IgG-FITC (Millipore) antibody. To determine the distribution of transferred IL-10⁺ B cells in vivo, B cells were stimulated in culture and captured by a cytokine secretion assay (Miltenyi Biotec), followed by SNARF-1 (Invitrogen) labeling before intravenous injection. Stained tissue sections were analyzed by the Carl Zeiss LSM 780. An AxioVision digital imaging system was used to optimize the signal-to-noise ratio.

To determine the CD4⁺ T cell homing to B cell follicles, the homing coefficient was calculated as previously described.³⁰ Briefly, T cells were quantified in the T cell zone and T-B borders, and the homing coefficient equals the ratio of the T cell densities of these two regions according to the following equation:

$$\mathrm{Homing}\mathrm{coefficient}=\frac{\mathrm{T}\mathrm{cell}\mathrm{density}\mathrm{in}\mathrm{the}\mathrm{follicles}}{\mathrm{T}\mathrm{cell}\mathrm{density}\mathrm{in}\mathrm{the}\mathrm{T\; -\; B}\mathrm{borders}}$$

Two-photon live imaging of T cell motility in vivo

The imaging system was composed of a Coherent Chameleon Vision II laser and an Olympus FVMPE-RS motorized upright microscope equipped with a water immersion lens (×25, NA 1.05, Olympus). The motorized stage on which live mice were imaged was enclosed in a customized chamber that was heated to 37 °C at equilibrium. To visualize CD4^± T cells and B cells at T-B cell borders in situ, CD45.1^± T cells (10⁷ per Boy/J mouse) were labeled with carboxyl fluorescein succinimidyl ester (CFSE, Sigma-Aldrich, green), and CD45.2^± B cells (10⁷ per C57BL/6 mouse) were labeled with CMTMR (Invitrogen, red) before being adoptively transferred into C57BL/6 RAG-2−/− mice. The experimental configurations for imaging CFSE and CMTMR were set at 780–920 nm. To capture T cell migration at T-B borders in three dimensions, z-stack imaging (30 μm depth with 1 μm per step) was conducted at a time resolution of 27 s per frame. Each imaging sequence was 35 min in duration. After acquisition, four-dimensional data sets were analyzed using Imaris software (Bitplane, version 9.2) with cell-tracking modules. Cell tracks that lasted for <2 min were excluded from the analysis. The time-lapse image sequences and videos, which were played back at 20 frames per second, were generated and exported from Imaris.

Human samples

A total of 42 pSS patients (all females, mean age: 48 ± 13 years) were recruited from Shenzhen People’s Hospital, Shenzhen, China, and enrolled in this study (Supplementary Table 1). A total of 24 healthy donors (3 males and 21 females, mean age: 40 ± 8 years) were studied in parallel as controls. pSS diagnosis was based on the 2002 American-European Consensus Group classification criteria.³¹ Subjects overlapping with any other autoimmune diseases or not fulfilling the diagnostic criteria were excluded. Informed consent was obtained from all patients. Ethics approval was obtained from the Medical Ethical Committee of Shenzhen People’s Hospital, Shenzhen, China.

Human cell culture

Sorted human naive CD45RO⁻CD4⁺ T cells were seeded in a culture plate precoated with anti-CD3 (clone OKT3) and anti-CD28 (clone CD28.2) antibodies (1 μg/mL). For Tfh polarization, recombinant human IL-6 (25 ng/mL) and IL-21 (10 ng/mL, R&D Systems) and 5 μg/mL anti-IFN-γ (clone Β27), anti-IL-4 (clone MP4-25D2) and anti-TGF-β (clone TW7-28G11) neutralizing antibodies (BioLegend) were added to a culture with a cell density of 2 × 10⁶/mL. In some experiments, recombinant human IL-10 (250 ng/mL, clone JES3-9D7), an anti-IL-10 neutralizing antibody (5 μg/mL) or IL-10 combined with iSTAT5 (10 μm) were added into cultures of naive CD4⁺ T cells (2 × 10⁶) under Tfh differentiation conditions. In cocultures, naive CD4⁺ T cells (10⁵) were cultured with sorted CD19⁺CD24⁺CD38^hi human Breg cells (0.25 × 10⁵) from healthy donors or pSS patients for 72 h before flow cytometric analysis.

Flow cytometric analysis

Surface markers were identified with anti-mouse and anti-human monoclonal antibodies (Supplementary Table 2). Before performing intracellular staining for IL-10 using the BD Fixation/Permeabilization Solution Kit, collected cells were stimulated with phorbol myristate acetate (PMA, 50 ng/mL; Sigma-Aldrich), ionomycin (500 ng/mL; Sigma-Aldrich) and monensin (BioLegend) for 4 h in culture. Intranuclear staining was performed with anti-Bcl-6 and anti-Blimp-1 antibodies using a FoxP3-staining buffer set according to the recommendations of the manufacturer (eBioscience). Stained cells were analyzed with a FACS Fortessa flow cytometer (BD Biosciences), and acquired data were analyzed with FlowJo software (TreeStar, Ashland, OR).

Statistical analysis

Statistical comparisons were performed as indicated in figure legends using two-sided tests. The quantification of the fluorescence intensity obtained by an AxioVision digital imaging system was analyzed using ImageJ (https://imagej.nih.gov/ij) and normalized to the density of IgG isotype control staining. GraphPad Prism 6 was used for the overall statistical analysis in this study. The results are expressed as the mean ± SD. Pearson’s correlation coefficient was used to analyze the correlation between two parameters. Multiple comparisons were evaluated by one-way ANOVA, and single comparisons were evaluated by an unpaired t-test. P < 0.05 was considered to be statistically significant.

Results

Decreased IL-10⁺ B cell numbers negatively correlate with enhanced Tfh cell numbers and increased disease activity in pSS patients and ESS mice

To assess IL-10⁺ B cell and Tfh cell responses during pSS development, we first examined circulating IL-10⁺ B cells in the peripheral blood mononuclear cells (PBMCs) of pSS patients by flow cytometry (Fig. 1 and Supplementary Fig. S1a).³¹ Consistent with previous findings,²⁰ IL-10⁺ B cells were found to be highly enriched in the CD19⁺CD24⁺CD38^hi Breg cell subset (Fig. 1a, Supplementary Fig. S1b). However, we detected substantially reduced IL-10 production in Breg cells from pSS patients compared with those from healthy donors (HD, Fig. 1b). Notably, both the frequencies and absolute numbers of IL-10⁺ B cells were significantly lower in the pSS patients with higher disease activity than in those with lower disease activity, indicating a negative correlation exists between IL-10⁺ B cell counts and the EULAR-SS Disease Activity Index (ESSDAI, Fig. 1c and Supplementary Fig. S1c). Moreover, we observed a strong but negative correlation between circulating Tfh cells and IL-10⁺ B cells in the pSS patients with active disease (Fig. 1d and Supplementary Fig. S1d). In ESS mice, CD1d^hiCD5⁺ Breg cells, the major B cell subset for IL-10 production, were found to produce significantly less IL-10 than CD1d^hiCD5⁺ Breg cells in naive control mice (Fig. 1e–f). Consistent with the findings observed in pSS patients, IL-10⁺ B cell numbers showed a marked reduction, while Tfh cell numbers steadily increased in the CLN during disease progression (Fig. 1g and Supplementary Fig. S1e), indicating that IL-10-producing Breg cells may be associated with a dysregulated Tfh cell response during pSS development.

Fig. 1.

Decreased IL-10⁺ B cell numbers negatively correlate with increased Tfh cell numbers and disease activity in both human pSS and murine ESS. a, b Representative flow cytometric profiles of IL-10 expression in human CD24⁺CD38⁻, CD24^dimCD38^dim, CD24⁺CD38^hi and CD24⁻CD38⁺ B cell subsets from healthy donors (HD) and pSS patients after a 4 h culture with PMA, ionomycin and monensin (PIM, n = 8 per group). c Correlation between the absolute IL-10⁺ B cell numbers in the peripheral blood of pSS patients and the EULAR-SS Disease Activity Index (ESSDAI) analyzed by Pearson’s correlation coefficient (n = 30). d Correlation between circulating T follicular helper (cTfh) cells and IL-10⁺ B cells in the PBMCs of patients with pSS analyzed by Pearson’s correlation coefficient (n = 32). e, f Representative flow cytometric profiles of IL-10 expression in CD1d^hiCD5⁻, CD1d^loCD5⁺, CD1d^hiCD5⁺ and CD1d^loCD5⁺ B cell subsets from naive control and ESS mice after a 4 h of culture with PIM (n = 5). g Kinetic changes in IL-10⁺ B cell and Tfh cell numbers in the draining cervical lymph nodes (CLN) of ESS mice at disease onset and during the acute and chronic stages (n = 7). Data were derived from at least three independent experiments. At least human PBMC 5 × 10⁵ events were acquired for flow cytometric analyses. Data are presented as the mean ± SD; *P < 0.01

IL-10 inhibits Tfh cell differentiation by promoting STAT5 phosphorylation

Previous studies have shown that IL-10 receptor expression is detected in both human and murine naive CD4⁺ T cells and that expression increases upon TCR engagement.³² Here, we found high levels of IL-10R expression in CCR7^loPD-1^hiCXCR5⁺ Tfh precursor cells and Tfh cells from both humans and mice (Supplementary Fig. S2a).¹³ In culture, recombinant IL-10 elicited a potent suppressive effect on human Tfh cell differentiation (Fig. 2a, Supplementary Fig. S2b). Similarly, IL-10 also inhibited murine Tfh cell differentiation and IL-21 production, an effect significantly abolished by blocking IL-10Rβ (Fig. 2b, c; Supplementary Fig. S2c–d). While examining IL-10R-mediated signal transduction, we found rapidly induced STAT5 phosphorylation at Y694 in naive CD4⁺ T cells upon IL-10 stimulation (Fig. 2d), while treatment with a pSTAT5 inhibitor largely abrogated the IL-10-mediated inhibitory effects on Tfh differentiation (Fig. 2e). Moreover, the administration of a pSTAT5 inhibitor upon ESS induction also resulted in accelerated disease development with augmented Tfh cell responses (data not shown). We further assessed the transcript expression of phenotypic markers and transcription factors of Tfh differentiation in the presence or absence of iSTAT5 upon IL-10 stimulation.³³ Interestingly, the IL-10/pSTAT5 axis was found to significantly inhibit the expression of Ascl2 (Fig. 2f), a transcription factor that controls the expression of the chemokine receptor CXCR5 in CD4⁺ T cells,³⁴ indicating that IL-10 may also modulate the follicular homing of CD4⁺ T cells upon activation. To determine the role of IL-10 in regulating the Tfh cell response during ESS development, naive CD62L⁺CD44⁻CD4⁺ T cells from Boy/J (CD45.1) mice were purified and adoptively transferred into wild-type (WT) or IL-10−/− C57BL/6 (CD45.2) recipient mice (Fig. 2g). Upon ESS induction, both enhanced follicular homing and profound expansion of CD45.1⁺ Tfh cells were observed in the CLN of IL-10−/− ESS recipient mice compared with those in WT recipients (Fig. 2h, i). Thus, these results demonstrate that IL-10 regulates the Tfh cell response in vivo.

Fig. 2.

IL-10 inhibits human and murine Tfh cell differentiation. a Purified human naive CD45RO⁻CD4⁺ T cells were cultured for Tfh polarization in the absence or presence of IL-10 and further analyzed by flow cytometry (n = 3). b Murine naive CD62L⁺CD44⁻CD4⁺ T cells were purified and cultured for Tfh polarization in the presence of IL-10 and further analyzed by flow cytometry. c Murine naive CD4⁺ T cells were cultured for Tfh polarization with medium, IL-10 or IL-10 plus an anti-IL-10 receptor beta chain (α−IL10Rβ) blocking antibody, and IL-21-producing Tfh cells were analyzed by flow cytometry (n = 3). d The expression of STAT5 phosphorylated at Y694 (pSTAT5) in murine naive CD4⁺ T cells was determined by flow cytometry after IL-10 stimulation. e, f Murine naive CD4⁺ T cells were cultured for Tfh polarization with IL-10 alone or in combination with a STAT5 inhibitor (iSTAT5) and analyzed by flow cytometry (72 h culture) or real-time PCR (24 h culture). g–i Naive CD4⁺ T cells from Boy/J (CD45.1) mice were transferred into WT or IL-10−/− C57BL/6 (CD45.2) mice, which were immunized for ESS induction. The follicular homing of CD45.1⁺CD4⁺ T cells in the CLN was determined (h), and CD45.1⁺ Tfh cells were analyzed (i) (n = 5). Data were derived from at least three independent experiments. Data are presented as the mean ± SD; n.s. not significant; *P < 0.05; and **P < 0.01

IL-10-producing B cells restrain the Tfh cell response during ESS development

To determine the role of IL-10⁺ B cells, the major cellular source of IL-10 in mice (Supplementary Fig. S3a–b), in regulating the Tfh cell response during ESS development, we transferred CD45.2⁺ B cells from either WT or IL-10−/− C57BL/6 mice with congenic CD45.1⁺CD4⁺ T cells into immunodeficient RAG-2−/− mice and induced ESS (Fig. 3a), in which T cells and myeloid cells are capable of producing IL-10. Using two-photon live imaging, we first observed significantly accelerated migration of CD4⁺ T cells towards B cell follicles, as measured by the parameters of centroid velocity and follicular displacement,³⁰ in the CLN of the RAG-2−/− ESS mice transferred with IL-10−/− B cells (Fig. 3b–e and Supplementary Videos 1 and 2). When compared with the control mice transferred with WT B cells, the RAG-2−/− mice transferred with IL-10−/− B cells exhibited significantly higher frequencies and numbers of Tfh cells (Fig. 3f), and these cells accumulated in the GC area during ESS development (Fig. 3g). Subsequently, these mice also exhibited markedly enhanced GC B cell and plasma cell generation (Fig. 3h, i).

Fig. 3.

IL-10-producing B cells restrain the Tfh cell response during ESS development. a RAG-2−/− mice were cotransferred with naive CD45.1⁺CD4⁺ T cells from Boy/J mice and CD45.2⁺CD19⁺ B cells from WT or IL-10−/− C57BL/6 mice, followed by immunization for ESS induction. b–e After transfer, labeled CD45.1⁺CD4⁺ T cells (green) and CD45.2⁺ WT or IL-10−/− B cells (red) in the CLN of RAG-2−/− mice were visualized at 48 h post immunization by two-photon live imaging (b), and the velocity of the T cells was measured (c) (499 vs 531 tracks of T cells over 35 min). x–y–z displacement (μm) plots of individual cell traces with starting positions realigned at the same origin are shown (d), and mean squared displacement (MSD) over 30 min was calculated (e) (26 tracks per group). f CD45.1⁺ Tfh cells in immunized RAG-2−/− mice were analyzed (n = 5). g The distribution and number of CD4⁺ T cells (green) in the associated GC area (GL-7, red) of immunized RAG-2-/- mice were assessed (symbols represent individual follicles, n = 3). h, i Plasma cells and GC B cells in immunized RAG-2−/− mice were analyzed by flow cytometry (n = 5). j The saliva secretion of RAG-2−/− mice was measured at 7 days post immunization (n = 5). k CD19⁺CD1d^hiCD5⁺ Breg cells from naive WT or IL-10−/− mice were expanded in vitro, sorted, and cocultured with naive CD4⁺ T cells under Tfh-polarizing conditions, and differentiated Tfh cells were analyzed. l naive CD4⁺ T cells were labeled with CFSE and cocultured with Breg cells from control or ESS mice for 3 days; proliferating CFSE^lo T cells were assessed by flow cytometry. Data were derived from at least three independent experiments. Data are presented as the mean ± SD; *P < 0.05; **P < 0.01; ***P < 0.001; and ****P < 0.0001

Previous studies have suggested that engagement of M3 muscarinic receptor (M3R), within which the second extracellular loop (M3R^2nd) serves as a functional domain,³⁵ upon cholinergic activation promotes saliva secretion by inducing the apical polarization of aquaporin 5 (AQP-5) in salivary acinar cells.³⁶ Due to the structural conservation of the extracellular loops in the transmembrane domains of M3R between humans and mice,³⁷ anti-M3R autoantibodies from pSS patients have been shown to induce salivary hypofunction in mice.^38,39 Here, we detected significantly higher titers of serum anti-M3R^2nd IgG in IL-10−/− B cell-transferred RAG-2−/− mice during ESS development, accompanied by strong IgG deposition and diminished apical AQP-5 distribution in the acini of the SG (Supplementary Fig. S3c–d); these mice also exhibited accelerated salivary hypofunction (Fig. 3j).

To assess the functional interaction between IL-10⁺ B cells and Tfh cells, we first examined the distribution of IL-10⁺ B cells in the lymphoid organs. Interestingly, the transferred IL-10⁺ B cells were mainly located at the T-B cell borders in both the draining CLN and spleen of ESS mice (Supplementary Fig. S3e), which is where Tfh cells home to provide B cell help.³⁰ In addition, both human and murine IL-10⁺ B cells expressed high levels of CXCR5 (Supplementary Fig. S3f), suggesting the potential for interactions between IL-10⁺ B cells and Tfh cells or Tfh precursors in situ. Moreover, when purified Breg cells were cocultured with naive CD4⁺ T cells under Tfh-polarizing conditions, the Breg cells from WT mice exerted potent suppressive effects on Tfh differentiation in an IL-10-dependent manner (Fig. 3k, Supplementary Fig. S3g). However, the Breg cells with reduced IL-10 production from ESS mice exhibited diminished inhibitory effects on CD4⁺ T cell expansion (Fig. 3l), which may account for the dysregulated Tfh cell response during ESS progression.

To investigate whether the same mechanism also operates in humans, we first confirmed CD19⁺ B cells as the major cellular source of IL-10 secretion in humans by flow cytometric analysis (Fig. 4a). Furthermore, sorted human Breg cells from HD were cocultured with autologous naive CD4⁺ T cells under Tfh-polarizing conditions (Supplementary Fig. S4a). Consistently, human Breg cells exerted IL-10-mediated suppressive effects on Tfh cell differentiation, an effect abolished by pSTAT5 inhibition (Fig. 4b). However, Breg cells from pSS patients, which have an impaired IL-10-producing capacity, showed significantly reduced suppressive effects on Tfh differentiation (Fig. 4c) and exhibited a strong but negative correlation with disease activity (Fig. 4d, Supplementary Fig. S4b). Together, these results demonstrate a critical role for IL-10-producing Breg cells in restraining Tfh cell responses during pSS progression in both humans and mice.

Fig. 4.

Impaired Breg cell suppressive function contributes to the enhanced Tfh cell response in pSS patients. a Representative flow cytometric profiles of CD4⁺ and CD19⁺ populations among the IL-10-producing lymphocytes in the PBMC populations from HD (n = 8) are shown. b Human CD19⁺CD24⁺CD38^hi Breg cells among the PBMCs of HD were cocultured with autologous naive CD4⁺ T cells under Tfh-polarizing conditions. An anti-human IL-10 neutralizing antibody or iSTAT5 were used to block signal transduction. Tfh cell differentiation was analyzed (n = 7). c, d A representative cytometric profile shows the generation of Tfh cells in cocultures containing sorted Breg cells from HD or pSS patients, and the suppression rates of pSS-derived Breg cells were calculated and correlated with the ESSDAI (n = 28). Data were derived from at least three independent experiments. Data are presented as the mean ± SD; *P < 0.05; and **P < 0.01

IL-10-producing Breg cells exert therapeutic effects on ESS development

Since B cells exhibited reduced IL-10 production and impaired function in restraining Tfh cells during pSS development in both humans and mice, we sought to determine whether the transfer of IL-10-producing Breg cells could attenuate ESS development in mice. After expansion in culture,⁴⁰ Breg cells from naive WT or IL-10−/− mice were adoptively transferred in a single dose of 5 × 10⁵ cells into ESS mice with disease at the early stage. ESS mice receiving treatments with the PBS vehicle or Breg cells from IL-10−/− mice served as controls and exhibited a significantly reduced saliva flow rate, but ESS mice treated with the IL-10-producing Breg cells from WT mice showed no exacerbation in salivary dysfunction (Fig. 5a). Moreover, the ameliorative effect on salivary hypofunction was abolished in ESS mice treated with IL-10R blockade (Supplementary Fig. S5a). Severe lymphocytic infiltration around the perivascular and periductal areas and tissue destruction were observed in diseased ESS mice, whereas no or only mild lymphocyte aggregation was detected in the SG of the ESS mice treated with the WT Breg cells (Fig. 5b–d), and in these mice, the numbers of infiltrating Tfh cells in the SG were also remarkably reduced (Fig. 5e).

Fig. 5.

Adoptive transfer of IL-10-producing Breg cells ameliorates ESS development in mice. a ESS mice with established disease were treated with a single dose of PBS vehicle or Breg cells from WT or IL-10−/− mice (5 × 10⁵ cells per mouse). The salivary flow rates of naive controls and treated ESS mice were measured (n = 9 per group). b H&E staining for the histological examination of submandibular and lacrimal tissue destruction in ESS mice treated with PBS or Breg cells is shown. c, d Lymphocytic foci in the submandibular glands were assessed by measuring the infiltrated area and histological scores (n = 9). e CXCR5⁺CD4⁺ T cells within the lymphocytic infiltrates in the submandibular glands were detected by immunofluorescence staining. Data were derived from three independent experiments. Data are presented as the mean ± SD; **P < 0.01; ***P < 0.001; and ****P < 0.0001

We next evaluated the Tfh cell responses in ESS mice given various treatments. Although extensive T cell homing to the B cell follicles was observed during disease progression, ESS mice treated with the IL-10-producing Breg cells exhibited highly restricted CD4⁺ T cell homing patterns (Fig. 6a). Further phenotypic analysis revealed that a profound reduction in the Tfh cell numbers in the CLN of the treated ESS mice occurred in an IL-10-dependent manner (Fig. 6b and Supplementary Fig. S5b–c). Moreover, treatment with the IL-10-producing Breg cells also resulted in a substantial reduction in GC B cell numbers in ESS mice (Fig. 6c). Similar findings were also observed in NOD mice with SS-like symptoms (Supplementary Fig. S5d–h). In addition, the IL-10-producing Breg cells suppressed Th1 and Th17 cells in the CLN of ESS mice, while CXCR5^-FoxP3^± Treg and CXCR5^±FoxP3^± Tfr cells were found to be comparable between mice with or without Breg cell treatment (Supplementary Fig. S6a–b). To examine the frequency of autoreactive B cells from recipient mice, we purified follicular B cells from recipient Boy/J ESS mice treated with congenic WT or IL-10−/− CD45.2⁺ Breg cells (Supplementary Fig. S6c) and performed an ELISpot assay. Quantitatively, the frequencies of anti-M3R^2nd and anti-SSA IgG-secreting B cells were significantly lower in the ESS mice treated with WT Breg cells (Fig. 6d and Supplementary Fig. S6d–e). Notably, reduced IgG deposition in the salivary acini accompanied by increased apical localization of AQP-5 in the glandular epithelial cells upon M3R activation was observed in the treated ESS mice with reduced frequencies of anti-M3R^2nd IgG-secreting B cells (Fig. 6e, f). Thus, these data show that the transfer of IL-10-producing Breg cells can effectively attenuate salivary dysfunction and suppress disease progression in ESS mice.

Fig. 6.

Adoptive transfer of IL-10-producing Breg cells suppresses the Tfh cell response and autoantibody production in ESS mice. a The homing coefficient of CD4⁺ T cells migrating to the associated B cell follicle in treated ESS mice was calculated. Each symbol represents one follicle and its associated T-B cell border (30 follicles from 4 mice per group). b, c Representative flow cytometric profiles of Tfh cells and GC B cells in the CLN of naive control and ESS mice treated with PBS or Breg cells (n = 8 per group) are shown. d Autoreactive B cells against the mM3R^2nd epitope in the CLN of naive control and ESS mice were detected by an ELISpot assay (n = 5). e, f Representative pictures show the distribution of AQP-5 (green) at the apical membrane (arrowheads) of salivary epithelial cells from ESS mice treated with PBS or Breg cells, and the mean fluorescence intensity of salivary-deposited IgG (red) per acinus was quantified and normalized to the mean fluorescence intensity of isotype IgG staining. Symbols represent individual acini (308–455 acini from 6 mice per group); scale bar, 20 μm. Data were derived from five experiments and are shown as the mean ± SD; **P < 0.01; ***P < 0.001; and ****P < 0.0001

Discussion

Recent studies have highlighted a pivotal function of Tfh cells in promoting and maintaining GC reactions in autoimmune pathogenesis, but the regulatory mechanism controlling Tfh cell responses remains less clear. In this study, we first revealed that decreased numbers of IL-10-producing B cells negatively correlated with an enhanced Tfh cell response and increased disease activity in patients with pSS and mice with ESS. In culture, IL-10 exhibited potent inhibitory effects on Tfh differentiation by inducing STAT5 phosphorylation. Compared to immunized RAG-2−/− mice transferred with WT B cells, RAG-2−/− mice transferred with IL-10−/− B cells exhibited increased follicular homing of cotransferred CD4⁺ T cells and enhanced Tfh cell expansion in the CLN, leading to accelerated ESS development. Notably, the Breg cells with a reduced IL-10-producing capacity from pSS patients exhibited significantly impaired inhibitory function in the suppression of autologous Tfh cell expansion, suggesting that the IL-10-producing capacity of B cells may also serve as an indicator for the evaluation of disease activity. Together, these results have demonstrated that IL-10-producing Breg cells are critically involved in restraining the Tfh cell response during pSS development.

In patients with autoimmune diseases, Tfh cells have been shown to promote B-cell hyperreactivity and autoantibody production.¹² Tfr cells, a newly identified subset of Treg cells, have been shown to negatively regulate Tfh cells.⁴¹ In mice deficient for Tfr cells, significant expansion of Tfh cells and elevation of autoantibody levels are detected during ESS development.¹⁰ Here, we showed that IL-10 inhibited the differentiation of both human and murine Tfh cells from naive T cells in culture. Notably, IL-10 deficiency resulted in both significantly increased follicular T cell homing and increased numbers of Tfh cells in the CLN of IL-10−/− mice with ESS, which further supported the notion that IL-10 plays a critical role in restraining the Tfh cell response during the development of pSS. In addition to secreted cytokines, marginal zone B cells were found to potently inhibit splenic but not nodal Tfh cells in a mouse model of atherosclerosis through contact-mediated suppression that occurred in a PD-L1-dependent manner.⁴² Interestingly, the PD-1/PD-L1 axis was also found to negatively regulate splenic Tfr cells in a mouse model of human multiple sclerosis,⁴³ suggesting bidirectional regulation by PD-L1⁺ marginal zone B cells in the spleen.

Previous studies have shown that IL-10-producing Breg cells can maintain Treg cell function and control effector T cell responses, including those involving the Th1/Th17 cell subsets in both humans with RA and RA mouse models.^20,25,44 In the present study, we showed the potent inhibitory function of IL-10-producing Breg cells in the suppression of Tfh cell generation worked by promoting pSTAT5 expression and modulating the transcriptional networks of Tfh cell programming. Since CXCR5 expression in CD4⁺ T cells is essential for cognate GC responses upon immunization,⁴⁵ here we revealed a profound inhibitory effect of the IL-10/pSTAT5 axis on Ascl2, a critical regulator for CXCR5 transcription,³⁴ which results in restricted CD4⁺ T cell motility and CD4⁺ T cell homing into B cell follicles, thereby limiting the cognate T-B cell interactions and autoreactive B cell expansion during ESS development.

Current studies on cytokine therapy for autoimmune diseases have generated promising results. Recently, low-dose IL-2 treatment showed beneficial effects in patients with lupus.⁴⁶ Treatment with an IL-10 fusion protein was shown to be well tolerated and safe in patients with RA.⁴⁷ However, systemic IL-10 administration does not alleviate the disease pathology of Crohn’s disease, possibly owing to insufficient amounts of local IL-10 in the intestinal tissues.⁴⁸ Tolerogenic cell therapies that produce regulatory cytokines have also been reported to have beneficial effects on patients with autoimmune disorders,⁴⁹ in which therapeutic efficacy may be associated with the ability of transferred cells to migrate to the inflammatory milieu. In this study, we demonstrated that the adoptive transfer of a single dose of IL-10-producing Breg cells early in the onset of disease effectively suppressed humoral autoimmunity and ameliorated the disease pathology of ESS. Early studies suggested that Breg cells from SS patients appear to exhibit normal inhibitory effects on TNF-α and IFN-γ production by CD4^± T cells, but the disease activities of enrolled SS patients were not determined.⁵⁰ In this study, although the function of Breg cells from pSS patients with low disease activity was found to be comparable with that of those from healthy donors, we further revealed that the regulatory function of Breg cells from pSS patients with high disease activity (ESSDAI > 5) was significantly impaired, which may account for the dysregulation of the Tfh cell response during pSS progression in patients. Thus, future studies are warranted to determine the therapeutic potential of Breg cell therapy in ESS mice with chronic inflammation. Previous studies, including our own, have shown that treatment with bortezomib, a proteasome inhibitor, has beneficial therapeutic effects on disease outcome, including significantly reduced autoantibody production, in mouse models of autoimmune disease,²⁸ but recent studies have indicated that bortezomib also inhibits the Tfh cell response in Rhesus macaques.⁵¹ Thus, it remains to be further investigated whether proteasome inhibitors can also exert therapeutic effects on autoimmune diseases by suppressing the Tfh cell response. Nevertheless, our current findings suggest that the transfer of functional IL-10-producing Breg cells may serve as a promising strategy for therapeutic intervention in autoimmune diseases.

In summary, our present work identified a critical role for IL-10-producing Breg cells in regulating the Tfh cell response during pSS development, as the transfer of Breg cells could effectively suppress the Tfh cell response and attenuate tissue pathology in ESS mice in an IL-10-dependent manner. Thus, these findings shed new light on the understanding of disease pathogenesis and the development of novel therapeutic strategies for pSS.

Author contributions

L.L. and X.L. designed and conceived the experiments. X.L. performed the experiments. X.W., F.X., K.M., X.W., L.L., D.X., F.W., X.S., Y.Z. and D.L. provided patient samples and analyzed data. All of the authors interpreted the data and discussed the results. L.L. and X.L. prepared the manuscript.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version of this article (10.1038/s41423-019-0227-z) contains supplementary material.