Research ArticleIMMUNE REGULATION

Regulatory T cells induce activation rather than suppression of human basophils

See allHide authors and affiliations

Science Immunology  25 May 2018:
Vol. 3, Issue 23, eaan0829
DOI: 10.1126/sciimmunol.aan0829

Unexpected basophil activation

Basophils are granulocytes that exist at a relatively rare frequency in the blood but are critical mediators of allergic and inflammatory responses. Regulatory T cells (Tregs) are known to suppress the functions of different immune cells, and Sharma et al. now examine how Tregs regulate basophil functions. Unexpectedly, they observed that resting human basophils are activated and not suppressed in the presence of Tregs. These activated basophils express activation markers (CD69, CD203c, and CD13) and release IL-4, IL-8, and IL-13. Treg-induced activation of basophils involves IL-3 and STAT5 but was not contact-dependent. These results now describe an activating effect mediated by Tregs and provide insight into how basophils are regulated.

Abstract

Basophils are a rare granulocyte population that has been associated with allergic and inflammatory responses. It is essential to understand the regulatory mechanisms by which basophils are kept in check, considering the impact of dysregulated basophil function on immune responses under different pathological conditions. Among immunoregulatory cells, CD4+CD25+FoxP3+ regulatory T cells (Tregs) are the key players that maintain immune tolerance. The mechanisms by which Tregs regulate and suppress diverse immune cell subsets have been studied extensively, but the impact of Tregs on basophil functions is not well understood. We report that human basophils are refractory to Treg-mediated suppression and found that Tregs stimulate resting basophils to induce the expression of activation markers including CD69, CD203c, and CD13 and the release of basophil cytokines including IL-13, IL-8, and IL-4. Mechanistically, Tregs could induce human basophil activation via IL-3 and STAT5 activation, whereas cellular contact was dispensable. Inhibition of either IL-3–IL-3 receptor interactions or STAT5 phosphorylation abrogated Treg-mediated activation of basophils. These results provide evidence of direct positive effects that human Tregs have on basophil activation and reveal a previously unrecognized feature of this cell subset well known for immunosuppressive functions.

INTRODUCTION

Basophils are rare granulocytes (about 0.5% of total leukocytes) that are important for the protection against helminth parasites. In addition, basophils mediate T helper 2 (TH2) responses, support B cell differentiation, and enhance humoral responses (1, 2). Activated basophils produce several immunoregulatory cytokines such as interleukin-4 (IL-4), IL-13, IL-6, TSLP (thymic stromal lymphopoietin), and BAFF (B cell–activating factor) (1, 2). Through these mediators, basophils regulate T and B cell responses and establish a critical link between innate and adaptive immunity (3, 4).

Although rare in number, basophils are implicated in various pathological conditions, in part, due to their ability to undergo rapid activation in response to a wide range of stimuli and release of diverse immune mediators. Basophils are well known for their pathogenic role in allergic conditions through secretion of mediators of hypersensitivity reactions including histamine and leukotriene (1, 2, 5). Many inflammatory skin diseases are characterized by basophil infiltration and activation (6). Basophils are also reported to play a role in the pathogenesis of autoimmune and other inflammatory diseases (711). Therefore, considering the impact of dysregulated functions of basophils on the immune response in various diseases, it is essential to understand the regulatory mechanisms by which basophils are kept in check.

In this regard, regulatory T cells (Tregs) have been widely studied for their role in immune tolerance and in the maintenance of immune homeostasis (1214). Tregs modulate autoimmune and inflammatory responses by exerting direct suppressive effects on various immune cells including dendritic cells (DCs), T cells, macrophages, monocytes, B cells, natural killer cells, and mast cells (1519). These functions are mediated by inhibitory surface molecules [cytotoxic T lymphocyte antigen-4 (CTLA-4) and lymphocyte-activation gene-3 (LAG-3)], inhibitory cytokines [transforming growth factor–β (TGF-β), IL-10, and IL-35], and cytotoxic molecules (granzyme and perforin). Although regulatory mechanisms by which Tregs suppress different immune cell subsets have been extensively studied, the impact of Tregs on basophil functions is not well understood. In view of emerging reports on the role of basophils in various pathological conditions, we investigated the interaction of Tregs with human basophils and its effects on basophil functions.

Here, we report that in contrast to other immune cells, human basophils are refractory to Treg-mediated suppression. Tregs did not inhibit anti–immunoglobulin E (IgE)–mediated activation of basophils. We found that Tregs induce activation of resting basophils characterized by significantly enhanced expression of activation markers and secretion of several basophil cytokines such as IL-13, IL-8, and IL-4. Mechanistically, these stimulatory functions of Tregs were mediated via IL-3 released from the activated Tregs that triggered the signal transducer and activator of transcription 5 (STAT5) pathway in basophils, whereas cell-to-cell contact was dispensable. Inhibition of either IL-3–IL-3 receptor (CD123) interaction or STAT5 phosphorylation abrogated Treg-induced activation of basophils. These results demonstrate the direct positive effects of human Tregs on basophil activation and uncover previously unrecognized features of these immunosuppressor cells.

RESULTS

Human basophils are refractory to Treg-mediated suppression

Because Tregs are classically known for their immunosuppressive functions, we first evaluated the ability of human Tregs to suppress high-affinity receptor for Fc region of IgE (FcεRI)–mediated basophil activation. In our assays, we used activated memory Tregs that have the ability to directly suppress target cells without the need for undergoing differentiation process (20). Stimulation of basophils with anti-IgE antibodies that cross-link FcεRI-bound IgE on the basophils led to a significant increase in the expression of various activation markers, including CD203c, CD13, and CD69, and degranulation-associated markers CD63 and CD107a (Fig. 1, A to H). Tregs did not inhibit anti-IgE–mediated activation of basophils, and the expression of various activation-associated markers on “Treg-educated” basophils was on par with that of anti-IgE–stimulated control basophils and “conventional T cells (Tconv)–educated” basophils (Fig. 1, A to H).

Fig. 1 Human basophils are refractory to Treg-mediated suppression.

Basophils were cultured either alone or with IL-3 or cocultured with anti-CD3/anti-CD28–activated Tregs or Tconv cells in the presence of IL-3 at 1:3 ratio for 24 hours. During the last 1 hour of the culture, basophils were stimulated with anti-IgE antibodies. (A to H) Representative plots (A and B) and mean ± SEM (C to H) of data from six independent experiments using cells from different donors showing the expression [mean fluorescence intensity (MFI) and percent positive cells] of CD203c (C), CD13 (D), CD69 (E and F), CD63 (G), and CD107a (H) on the basophils under various experimental conditions. (I) Amount of histamine released under above experimental conditions (mean ± SEM; n = 6). **P < 0.01, ***P < 0.001, and ****P < 0.0001; ns, not significant by one-way ANOVA with Tukey’s multiple comparison test.

The effect of Tregs on the phenotype of basophils is reproduced in functional assays. We found that Tregs did not inhibit histamine release from anti-IgE–activated basophils (Fig. 1I). These results indicate that human basophils are refractory to Treg-mediated suppression.

Human basophils lack receptors to receive contact-dependent inhibitory signals from Tregs

We examined possible factors contributing to resistance of human basophils to Treg-mediated suppression. Suppression of target immune cells by Tregs is attributed to both contact-dependent and contact-independent mechanisms (15, 2124). CTLA-4 and LAG-3 play a critical role in contact-dependent suppression of target cells by Tregs through interaction with CD80/CD86 and human leukocyte antigen-D related (HLA-DR), respectively (18, 25, 26). However, human basophils from the circulation (Fig. 2A) (2730) and from secondary lymphoid organs such as spleen (Fig. 2B) (31) lacked the expression of HLA-DR and B7 costimulatory molecules under steady state and under stimulation conditions, although lipopolysaccharide-stimulated DCs, used as positive control, were all positive for these molecules (fig. S1). These results imply that the inability of Tregs to suppress activation of human basophils is in part due to the absence of necessary receptors on basophils required to receive inhibitory signals. In addition, no significant differences in the viability of human basophils were observed when they were cocultured with Tregs (Fig. 2, C and D).

Fig. 2 Human basophils are unresponsive to contact-dependent or contact-independent inhibition by Tregs.

(A and B) Expression of HLA-DR and costimulatory molecules CD80 and CD86 (mean ± SEM; n = 4 donors) on resting and IL-3–stimulated basophils from human blood and spleen. (C and D) Percentage of annexin V– and PI-positive basophils that were cultured for 24 hours in medium alone, with IL-3, or with IL-3 plus anti-CD3/anti-CD28–activated Tregs or Tconv. Basophils were also stimulated with anti-IgE during the last 1 hour of the culture, as indicated. Representative plots and mean ± SEM of data from three independent experiments. (E) Expression of TGF-βRII and IL-10Rα (mean ± SEM; n = 6 donors) on human blood basophils. (F) Expression of CD123, CD13, CD69, and CD203c on IL-10Rα+ and IL-10Rα subsets of basophils (mean ± SEM; n = 3 independent experiments using cells from different donors) cultured for 24 hours with IL-3 and various concentrations of IL-10. During the last 1-hour culture, anti-IgE antibodies were added to stimulate basophils. ns, not significant by one-way ANOVA with Tukey’s multiple comparison test (D) or by two-way ANOVA with least significant difference test (F).

Human basophils are unresponsive to contact-independent inhibition by Tregs

Several studies have demonstrated that Tregs mediate immunosuppressive functions by cytokines, in particular, TGF-β and IL-10 (21, 32, 33). The lack of suppressive effects of Tregs on human basophils raised an intriguing question on the expression of receptors for these cytokines. We found that circulating human basophils lack TGF-βRII (Fig. 2E), but IL-10Rα was expressed on a subset of circulating human basophils (40.3 ± 6.6%, n = 6 donors) (Fig. 2E). Similar patterns of expression were observed on human splenic basophils (fig. S2).

We explored whether the IL-10Rα+ subset of human basophils could be modulated by IL-10. Basophils were treated with increasing concentrations of IL-10 (from 100 pg to 10 ng/ml) for 24 hours and then stimulated with anti-IgE during the last hour of culture. Even at the highest concentration (10 ng/ml), IL-10 did not modify the expression of activation markers on IL-10Rα+ or IL-10Rα basophil subsets (Fig. 2F). IL-10 suppressed the expression of various activation markers on DCs, thus confirming the biological activity of this cytokine (fig. S3). Another report also indicated that IL-10 does not affect either IL-3 or IgE-mediated activation of human basophils (34). These results provide a potential explanation for the refractoriness of human basophils to Treg-mediated suppression.

Tregs induce the activation of resting basophils

Because Tregs did not inhibit activation of human basophils, we examined the impact of activated Tregs cocultured with resting basophils in the absence of any exogenous stimulation or cytokines for 24 hours. Tregs induced the activation of human basophils, as indicated by strongly increased expression of the activation markers CD13, CD203c, and CD69 (Fig. 3, A and B). However, expression of CD63 and CD107a, the markers associated with basophil degranulation, remained unaltered (Fig. 3, C and D), indicating that Treg-induced activation of human basophils is not associated with degranulation. We confirmed that enhanced expression of activation markers on basophils upon coculture with Tregs was not due to nonspecific stimulation of basophils by anti-CD3 and anti-CD28 monoclonal antibodies (mAbs) used for the activation of Tregs because the phenotype of basophils was not altered when they were cultured with these mAbs (fig. S4).

Fig. 3 Tregs induce activation of resting human basophils.

Basophils were either cultured for 24 hours alone or cocultured with anti-CD3/anti-CD28–activated Tregs or Tconv. (A and B) Representative plots and mean ± SEM of data from seven independent experiments using cells from different donors showing the expression of CD13, CD203c, and CD69 on the basophils. (C and D) Representative dot plots and mean ± SEM of data from five independent experiments showing the expression of degranulation markers CD63 and CD107a on the basophils. *P < 0.05, **P < 0.01, and ****P < 0.0001; ns, not significant by one-way ANOVA with Tukey’s multiple comparison test.

Treg-induced activation of human basophils was not restricted to phenotypic changes, and functional analysis of basophils also confirmed their activation. Tregs significantly induced the secretion of IL-13 and IL-8 from basophils (Fig. 4, A and B). Although significant, induction of IL-4 was minimal (Fig. 4C). Analysis of cytokines produced by activated Tregs alone at equivalent numbers revealed that they produced minimal amounts of IL-13 [13.1 ± 4.4 pg/ml (mean ± SEM); n = 8], IL-8 (13.5 ± 5.1 pg/ml; n = 8), and IL-4 (7.3 ± 3.0 pg/ml; n = 8 donors). Together, these results demonstrate that Tregs induce the activation of human basophils, as evidenced by phenotypic changes and secretion of cytokines (Figs. 3 and 4).

Fig. 4 Induction of basophil cytokines by activated Tregs.

Basophils were cultured either alone or with anti-CD3/anti-CD28–activated Tregs for 24 hours. (A to C) Supernatants of cocultures were analyzed for the amounts of (pg/ml; mean ± SEM; n = 6 independent experiments using cells from different donors) IL-13 (A), IL-8 (B), and IL-4 (C). *P < 0.05 by two-tailed paired Student’s t test.

Cell adhesion is dispensable for Treg-induced activation of human basophils

We aimed to identify the mechanism(s) by which Tregs induce activation of human basophils. Interaction of LFA-1 (CD11a/CD18) on Tregs with intercellular adhesion molecule-1 (ICAM-1) (CD54) mediates adhesion and aggregation of Tregs on DCs (3537). Previous studies have reported the expression of ICAM-1 on basophil cell lines and circulating human basophils (38, 39). We confirmed the expression of ICAM-1 on resting human basophils, which was enhanced upon IL-3 stimulation (Fig. 5A). ICAM-1 blockade did not abrogate Treg-induced activation of basophils (Fig. 5, B to D), and we confirmed that ICAM-1 blocking antibodies were functional because they inhibited DC-mediated proliferation of CD4+ T cells (fig. S5). IL-3–stimulated human basophils expressed inducible T cell costimulator ligand (ICOSL) (fig. S6A), but ICOSL blockade had no significant effect on Treg-mediated basophil activation (fig. S6B). These results suggest that adhesion of Tregs to basophils is dispensable for Treg-induced activation of basophils.

Fig. 5 LFA-1–ICAM-1 interaction is dispensable for the Treg-induced human basophil activation.

(A) Expression of ICAM-1 (mean ± SEM; n = 5 donors) on the resting and IL-3–stimulated basophils. (B to D) Basophils were either cultured for 24 hours alone or cocultured with anti-CD3/anti-CD28–activated Tregs with or without blocking mAbs to ICAM-1. (B and C) Representative plots indicating the expression of CD13, CD203c, and CD69 on the basophils. (D) Histograms depicting the expression of CD13, CD203c, and CD69 (mean ± SEM; n = 4 independent experiments using cells from different donors) on the basophils. (E) Expression of CD13, CD203c, and CD69 (mean ± SEM; n = 5 independent experiments using cells from different donors) on the basophils cocultured with Tregs either in direct contact or in transwells. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001; ns, not significant by one-way ANOVA with Tukey’s multiple comparison test.

We performed Transwell experiments to confirm that cell-to-cell contact is dispensable for stimulating basophil activation by Tregs. Preactivated Tregs were placed in the upper chamber and basophils in the lower chamber of the Transwell system. Tregs in transwells induced significant expression of basophil activation markers (CD13, CD203c, and CD69). The expression levels of activation markers were equivalent to levels induced on basophils cocultured with Tregs without Transwell conditions (Fig. 5E). The leak-proof nature of Transwell was confirmed by the absence of CD4+ T cells in the lower chambers of Transwell (fig. S7).

Soluble mediators released from activated Tregs stimulate basophils

Because cell contact was found not to be required for Treg-induced human basophil activation, we explored whether soluble mediators released by activated Tregs mediate basophil stimulation. To examine this, we cultured resting basophils with cell-free supernatants from either unstimulated or anti-CD3/anti-CD28–stimulated Tregs. Supernatants from unstimulated Tregs did not modify the expression of activation markers on basophils (Fig. 6). However, basophils cultured with the supernatants from stimulated Tregs showed significantly enhanced expression of CD13, CD203c, and CD69 (Fig. 6A) but not degranulation-associated markers CD63 and CD107a (Fig. 6B). The expression pattern of these activation markers was identical to basophils cocultured with activated Tregs (Fig. 3). These results thus confirm a role for soluble mediator(s) released from activated Tregs in stimulating basophil activation.

Fig. 6 Soluble mediators released from activated Tregs stimulate basophils.

Basophils were cultured for 24 hours either alone or with cell-free culture supernatants (SUPs) from unstimulated (U-Tregs) or anti-CD3/anti-CD28–stimulated Tregs (S-Tregs). (A and B) Level of expression of (MFI or percent positive cells) CD13, CD203c, and CD69 (A) and degranulation markers CD63 and CD107a on the basophils (B) (mean ± SEM; n = 3 independent experiments using cells from different donors). *P < 0.05 and **P < 0.01; ns, not significant by one-way ANOVA with Tukey’s multiple comparison test.

Tregs induce human basophil activation by an IL-3–dependent mechanism

Previous reports have shown that IL-3 plays a crucial role not only in the survival (40, 41) but also in the activation of basophils (39, 42, 43). IL-3 directly induces and enhances several activation-linked markers on human basophils, such as CD69, CD203c, and CD13 (39, 44), and stimulates secretion of cytokines including IL-13 (45). The expression of activation markers and secretion of cytokines by human basophils cocultured with Tregs were reminiscent of IL-3 activity on basophils.

Several recent articles identified activated CD4+ T cells as a major source of IL-3 (4648). We found that anti-CD3/anti-CD28–activated Tregs secrete IL-3 in the range of 148 ± 25 pg per million cells (mean ± SEM; n = 11 donors), whereas resting Tregs did not secrete detectable levels of IL-3 (fig. S8). A dose-response study indicated that treatment of human basophils with IL-3 at a concentration equivalent to that secreted by human Tregs could induce activation (fig. S9). These data thus validate that the quantity of IL-3 produced by Tregs is sufficient to induce basophil activation.

To demonstrate unequivocally that Tregs induce human basophil activation by an IL-3–dependent mechanism, we followed two independent approaches: blockade of IL-3 interaction with its receptor (IL-3R) and inhibition of IL-3–mediated signal transduction pathway. First, we blocked IL-3R on basophils before coculture with activated Tregs. Consistent with our hypothesis, the expression of Treg-induced activation markers CD203c, CD13, CD69, and FcεRI on basophils was significantly reduced upon IL-3R blockade (Fig. 7, A and B). In addition, IL-3R blockade also decreased Treg-induced secretion of IL-8 and IL-13 by basophils (Fig. 7C).

Fig. 7 Tregs induce human basophil activation by an IL-3–dependent mechanism.

Basophils were either cultured for 24 hours alone or cocultured with anti-CD3/anti-CD28–stimulated Tregs in the presence of isotype control mAbs or blocking mAbs to IL-3R. (A) Representative plots of CD13, CD203c, CD69, and FcεRI expression on basophils. (B) Histograms (mean ± SEM) depicting the expression of CD13 (n = 8 independent experiments using cells from different donors), CD203c, FcεRI, and CD69 (n = 11 independent experiments) on the basophils. (C) Amounts of secretion of IL-13 and IL-8 (mean ± SEM; n = 5 independent experiments using cells from different donors) by basophils. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 by one-way ANOVA with Tukey’s multiple comparison test.

IL-3 promotes priming of basophils for FcεRI-mediated degranulation and for the release of inflammatory mediators. In response to degranulation signals, Treg-activated basophils showed significantly enhanced expression of degranulation-associated molecules CD107a and CD63 and released histamine at levels similar to that observed with IL-3–FcεRI–activated basophils (fig. S10). These data confirm that Treg-activated basophils are functionally fit and that Tregs license basophils to undergo activation by degranulation signals.

To further authenticate that IL-3 signaling is implicated in the Treg-mediated activation of basophils, we investigated STAT5 pathway (34). Tregs induced phosphorylation of STAT5 in basophils at levels similar to that observed by treatment with IL-3 used at an equivalent concentration produced by stimulated Tregs (Fig. 8, A and B). STAT5 inhibition in basophils led to abrogation of Treg-induced activation markers CD13, CD203c, and CD69 on basophils (Fig. 8C) and the secretion of IL-13 and IL-8 (Fig. 8D). These data together demonstrate that IL-3 plays a critical role in inducing Treg-mediated activation of human basophils.

Fig. 8 Tregs induce activation of basophils by STAT5 pathway.

(A and B) Basophils were either cultured for 6 hours alone or with IL-3 or cocultured with anti-CD3/anti-CD28–stimulated Tregs. (A) Representative plots of phosphorylated STAT5 (pSTAT5) expression in basophils (A) and mean ± SEM of pSTAT5-positive basophils (B) from four independent experiments using cells from different donors. (C and D) Basophils were pretreated with STAT5 inhibitor (STAT5 inh) or DMSO, followed by coculture with Tregs for 24 hours. Expression of CD13, CD203c, and CD69 on the basophils (C) and the amounts of IL-13 and IL-8 in the culture supernatants (D) (mean ± SEM; n = 4 independent experiments). *P < 0.05, **P < 0.01, and ****P < 0.0001; ns, not significant by one-way ANOVA with Tukey’s multiple comparison test.

DISCUSSION

Here, we found that human basophils not only show unique refractoriness toward Treg-mediated–suppressive effects but also undergo activation by Tregs. In mice, basophils and antigen-specific T cells were reported to be engaged in multiple serial interactions of short to moderate duration in the tissues (47). However, mouse and human basophils display differences in the expression of costimulatory and antigen-presenting molecules that play a critical role in the cross-talk between innate and T cells (2730, 4953). Therefore, cognate interaction as in the case of classical antigen-presenting cells (APCs) and Tregs (54) is unlikely in the case of human basophils. Both Tregs and basophils are present in secondary lymphoid organs and in the peripheral tissues (31, 55, 56). In these sites, Tregs constantly receive activation signals from APCs, leading to the secretion of IL-3. Consequently, without involving cognate interaction, IL-3 secreted from activated Tregs could affect basophils both at secondary lymphoid organs and at peripheral tissues.

Basophils play an important role in the protection against helminth parasites, and various lines of evidence provide a pointer that IL-3 is central for these functions by supporting survival of basophils and promoting basophilia (5759). IL-3 synergizes with basophil-derived IL-4 and IL-13 to promote the alternative activation of monocytes (60). Alternative or M2-type monocyte-macrophages are well recognized for their anti-inflammatory functions and for protection against nematode parasite (61). Thus, we propose that in vivo, under physiological conditions, Tregs might contribute to controlling the inflammation and to fighting helminth infection by promoting basophilia, by enhancing the survival and activation of basophils via an IL-3–dependent mechanism, and by inducing IL-4 and IL-13. Although activation of human basophils by Tregs led to induction of these cytokines, the effect on IL-13 secretion was particularly prominent. On the basis of the current evidence, it appears that IL-4 response by human basophils is mostly dependent on IgE/FcεRI cross-linking. However, IL-13 could be induced in basophils by IgE/FcεRI-independent signaling as well (62). In this regard, IL-3 has been reported to provide adequate stimulus to induce IL-13 secretion by human basophils (44, 63). Because Tregs induced the activation of human basophils via an IL-3–dependent mechanism, it provides an explanation for low induction of IL-4 and higher secretion of IL-13 by basophils. Abrogation of IL-13 (and IL-8) production by basophils either upon IL-3R blockade or upon STAT5 inhibition further validates that these cytokines are contributed mainly by basophils.

Our study is limited to polyclonal CD4+CD25+FoxP3+ memory Tregs from the circulation of healthy donors. However, Tregs are highly diverse, and in addition to FoxP3 Tregs (such as Tr1 cells), several distinct subsets of FoxP3+ Tregs have been identified. FoxP3+ human bona fide Tregs in the circulation consist of CD45RAFoxP3hiCD4+ activated and CD45RA+FoxP3loCD4+ resting Tregs (20). Moreover, on the basis of the expression of chemokine receptors CXCR3, CCR4, CCR10, and CCR6, memory Tregs were further classified into various effector TH-like Treg subsets (TH1-Treg, TH2-Treg, TH17-Treg, and TH1/TH17-Treg) that are immunosuppressive and FoxP3+ but secrete effector cytokine(s) of a corresponding TH subset (6468). Therefore, it is likely that IL-3 secretion pattern might vary among these distinct Treg subsets and, as a consequence, their effect on basophil activation. In addition, IL-3 production by Tregs might also be influenced by signals they receive in the microenvironment and, in particular, signals by basophils. Detailed analyses of the influence of basophils on Treg plasticity, phenotype, functions, and cytokine secretion pattern, particularly IL-3, should shed light on these points. Although chronic inflammation promotes pathogenic reprogramming of Tregs into TH cells (69, 70), this might not be the case in our study because basophil-Tregs were cultured only for 24 hours. Another area that requires further exploration is whether IL-3 production and suppressive functions of Tregs are connected or are an independent process. It was suggested that reduced number and/or function of FoxP3+ Tregs in several immunodeficiencies such as immunodysregulation, polyendocrinopathy, enteropathy, X-linked (IPEX), Omenn syndrome, and Wiskott-Aldrich syndrome leads to enhanced TH2 cytokine production and, as a consequence, increased IgE levels (71). Defective functions of Tregs are also reported in diverse autoimmune and inflammatory diseases (72). A comparative analysis of Treg functions and IL-3 production in healthy versus immunodeficient and autoimmune patients and their repercussion on basophil functions needs to be performed.

Multiple subsets of Tregs play an important role in the suppression of allergic responses both during sensitization phase and during effector phase by targeting various cell types (73). However, few reports indicated that the frequency of CD4+CD25+ Tregs is increased in patients with allergic rhinitis, atopic dermatitis, and atopic and bronchial asthma (7476). In addition, the number of Tregs has increased during exacerbation of asthma and allergic rhinitis (75, 76). Although increased Tregs observed in these patients might represent response of the host to curtail pathogenic responses, on the contrary, our current data suggest that enhanced Treg response observed in these patients might even aggravate allergic responses by activating basophils in an IL-3–dependent manner to induce IL-13 and IL-4 that are TH2-associated cytokines and support IgE class switch. The Treg number in pediatric patients with allergic rhinitis and bronchial asthma was reported to correlate positively with total IgE level (75). It is therefore likely that various specialized subsets of regulatory cells participate in the control of different arms of allergic response and that a particular Treg subset alone might not efficiently control allergic responses (7780). For example, in allergic rhinitis patients, deficiency of allergen-specific IL-10–producing Tr1 subset found to be associated with disease severity, although the number and function of CD4+CD25+ Tregs in these patients did not change (81). Therefore, detailed investigation of various Treg subsets is crucial for establishing their correlation with diseases severity in allergic diseases. Considering that CD4+CD25+FoxP3+ Tregs activate basophils and could license them to undergo degranulation, leading to histamine release, we suggest that simultaneous blocking of basophil responses is necessary to attain maximum benefits of Treg-targeted therapeutic approaches for allergic conditions.

MATERIALS AND METHODS

Study design

This research was designed to investigate the regulation of human basophil functions by CD4+CD25+FoxP3+ Tregs and to identify underlying mechanisms. Cells were isolated from the buffy bags of healthy adult blood donors. Because there was no intervention, blinding and randomization were not used. Depending on the number of experimental groups, our experimental design used repeated-measures analysis of variance (ANOVA) or Student’s t test to measure the level of significance. Acquisition of data from multiple independent donors for each of the parameters permitted us to control interindividual variations in the experiments.

Isolation of basophils

Buffy bags of healthy donors were subjected to Ficoll density gradient centrifugation and cellular fractions containing peripheral blood mononuclear cells (PBMCs), and basophils were collected. Basophils were isolated from these cells by using Basophil Isolation kit II (Miltenyi Biotec) and autoMACS. The purity of basophils was 96 ± 0.5%, as analyzed by flow cytometry.

Regulatory T cells

For the isolation of CD4+CD45RO+CD25+ Tregs, untouched CD4+ T cells were first purified (CD4+ T Cell Isolation kit II, Miltenyi Biotec), followed by isolation of untouched memory T cells (CD45RO+), by depletion of naïve T cells using CD45RA microbeads (Miltenyi Biotec). Further, CD25+ Tregs were positively selected from these memory T cells by CD25 microbeads (Miltenyi Biotec). Identity of Tregs was confirmed by phenotype analysis of various markers (fig. S11A) and by functional assays, as detailed below (fig. S11B). CD4+CD45RO+CD25−/lo T cell fraction was used as conventional T cells (Tconv) for the comparison with Tregs (fig. S11). The purity of various T cell populations was 96 ± 3%.

CFSE labeling of CD4+ T cells

CD4+ T cells were washed and resuspended in phosphate-buffered saline (10 × 106 cells/ml). Cells were incubated with carboxyfluorescein succinimidyl ester (CFSE; 5 μM; BD Biosciences) at 37°C in a water bath for 10 min, followed by thorough washing with the medium. Labeled CD4+ T cells were resuspended in X-VIVO medium at a concentration of 106 cells/ml.

Treg suppression assay

For analyzing the suppression of Tconv proliferation by Tregs (fig. S11B), CFSE-labeled Tconv cells were cultured in serum-free X-VIVO 15 medium either alone (0.1 × 106 cells per 200 μl) or with Tregs at 1:3 ratio in the presence of plate-bound anti-CD3 mAb (1 μg/ml) and soluble anti-CD28 mAb (1 μg/ml). After 5 days, proliferation of Tconv cells was analyzed by flow cytometry based on the dilution of CFSE.

Coculture of basophils with Tregs or Tconv

To investigate the effect of Tregs on anti-IgE–mediated activation of human basophils, we cultured basophils (0.1 × 106 cells/well per 200 μl) in 96-well U-bottom plate alone in serum-free X-VIVO 15 medium, with IL-3 (2 ng per 0.1 × 106 cells; ImmunoTools), or with IL-3 plus Tregs or Tconv cells (0.3 × 106 cells/well per 200 μl) at 1:3 ratio for 24 hours. Tregs/Tconv under the coculture conditions were stimulated with plate-bound anti-CD3 mAb (1 μg/ml) and soluble anti-CD28 mAb (1 μg/ml). During the last 1-hour culture, polyclonal, affinity-isolated anti-human IgE (ε-chain–specific; Sigma-Aldrich) antibodies (2 ng per 0.1 × 106 cells) were added to stimulate the basophils.

For measuring the viability of cells by annexin V and propidium iodide (PI; Sigma-Aldrich) staining, basophils were cultured for 24 hours in medium alone, with IL-3, or with IL-3 plus anti-CD3/anti-CD28–activated Tregs or Tconv. Basophils were also stimulated with anti-IgE antibodies during the last 1 hour of the culture. As a positive control for annexin V and PI staining, PBMCs were treated with gemcitabine (1 μg per 0.5 × 106 cells/ml; Sigma-Aldrich) for 24 hours or 20% ethanol (200 μl per 0.5 × 106 cells/ml) for 30 min, respectively (fig. S12).

For the analysis of the impact of Tregs on resting basophils, basophils (0.1 × 106 cells/well per 200 μl) were either cultured alone in serum-free X-VIVO 15 medium or cocultured with Tregs (0.3 × 106 cells/well per 200 μl) at 1:3 ratio for 24 hours. Tregs in the coculture were stimulated with plate-bound anti-CD3 mAb and soluble anti-CD28 mAb.

To investigate the implication of either surface molecules or soluble mediators in Treg-induced basophil activation, we cultured basophils (0.1 × 106 cells) alone in serum-free X-VIVO 15 medium, with Tregs (0.3 × 106 cells/well per 200 μl), or with Tregs in the presence of blocking mAbs (10 μg/ml) to ICAM-1, ICOSL, or IL-3 receptor (CD123) or isotype control mAb for 24 hours. Tregs under all coculture conditions were stimulated with anti-CD3 and anti-CD28 mAbs.

To analyze whether soluble mediators released from Tregs induce basophil activation, we cultured Tregs (0.3 × 106 cells) either alone (resting phase) in serum-free X-VIVO 15 medium or stimulated with anti-CD3 and anti-CD28 mAbs for 24 hours. Cell-free culture supernatants were harvested, and basophils (0.1 × 106 cells per well) were cultured for 24 hours either alone in serum-free X-VIVO 15 medium or in the presence of supernatants obtained from unstimulated or stimulated Tregs.

For Transwell experiments, Tregs (0.3 × 106 cells per 100 μl) were preactivated with anti-CD3 and anti-CD28 mAbs for 24 hours. Basophils were placed in the lower chambers of transwells (Transwell→, 0.4 μm; Corning), and preactivated Tregs were added to the upper chambers of Transwell. The cells were cultured for 24 hours.

To analyze STAT5 activation, we cocultured basophils with preactivated Tregs (0.3 × 106 cells per 200 μl) at 1:3 ratio for 6 hours. As a positive control, basophils (0.1 × 106 cells per 200 μl) were treated with IL-3 (100 pg/ml, an equivalent concentration produced by stimulated Tregs) for 6 hours. Cells were stained with anti-STAT5 (pY694) or isotype control mAbs by using Cell Signaling Buffer Set A (Miltenyi Biotec).

To confirm the role of STAT5 in Treg-induced basophil activation, we pretreated basophils (0.1 × 106 cells per 200 μl) with STAT5 inhibitor CAS 285986-31-4 (20 μM; Merck Millipore) or equivalent volume of dimethyl sulfoxide (DMSO) for 2 hours, followed by coculture with Tregs (0.3 × 106 cells) for 24 hours. Tregs in the coculture were stimulated with anti-CD3 and anti-CD28 mAbs.

To determine whether Treg-activated basophils are functionally fit and Tregs license basophils to undergo activation by degranulation signals, we cocultured basophils (0.1 × 106 cells per 200 μl) with Tregs (0.3 × 106 cells per 200 μl) for 24 hours in the presence of anti-CD3 and anti-CD28 mAbs. During the last 1-hour culture, anti-IgE antibodies (2 ng per 0.1 × 106 basophils) were added to stimulate the basophils.

Basophil response was examined by analyzing the expression of activation-associated markers by flow cytometry. Cell-free culture supernatants were collected from basophil-Treg coculture experiments and subjected for the analysis of histamine or cytokines.

Treatment of basophils with IL-10

Basophils (0.1 × 106 cells/well per 200 μl) were cultured in X-VIVO 15 medium alone, with IL-3 (2 ng per 0.1 × 106 cells), or with IL-3 plus increasing concentration of recombinant human IL-10 (rhIL-10; eBioscience) ranging from 100 pg/ml to 10 ng/ml for 24 hours. During the last 1-hour culture, anti-IgE antibodies (2 ng per 0.1 × 106 cells) were added to stimulate basophils, and the expression of activation markers was analyzed by flow cytometry.

Isolation of human splenocytes

Spleen pieces were mechanically disaggregated by gentleMACS dissociator (program m_spleen_04, Miltenyi Biotec) and were then filtered through a 70-μm nylon membrane filter (BD Biosciences) to obtain single-cell suspension of splenocytes. Splenocytes were then processed through Ficoll-Paque density gradient centrifugation to remove red blood cells (RBCs). RBC-depleted splenocytes were then washed two times with RPMI 1640. Splenic basophils were distinguished from mast cells (CD117+) and DC subsets (BDCA1-4+), and the expression of HLA-DR, CD80, CD86, IL-10Rα, and TGF-βRII on these CD203c+FcεRI+ splenic basophils was analyzed by flow cytometry.

Culture of DCs

Monocytes from PBMCs were isolated by using CD14 microbeads (Miltenyi Biotec). Monocytes were differentiated to DCs by culturing them with rhIL-4 (500 IU per 106 cells) and rhGM-CSF (recombinant human granulocyte-macrophage colony-stimulating factor; 1000 IU per 106 cells) (both from Miltenyi Biotec) for 6 days. Differentiation of DCs was confirmed by the analysis of surface markers. In addition, DCs were also stimulated with lipopolysaccharide from Escherichia coli O55:B5 (10 ng per 0.5 × 106 cells; Sigma-Aldrich) for 24 hours and analyzed for the expression of maturation-associated markers.

To analyze the effect of IL-10 on DCs, we treated cells with rhIL-10 (10 ng/ml per 0.5 × 106 cells) for 24 hours, followed by analysis of surface markers by flow cytometry. To explore the effect of ICAM-1 blockade toward DC-mediated CD4+ T cell proliferation, we preincubated DCs with ICAM-1 blocking mAb (20 μg/ml) or isotype control mAb for 2 hours, followed by coculture with CFSE-labeled allogeneic CD4+ T cells (0.1 × 106 T cells per well) at 1:20 ratio in a total volume of 200 μl for 5 days.

Measurement of histamine and cytokines

Histamine was measured in the culture supernatants by Histamine EIA kit (Bertin Pharma). IL-3 (Quantikine ELISA kit, R&D Systems), IL-8, IL-13, and IL-4 (ELISA Ready-SET-Go, eBioscience) were analyzed in the culture supernatants by enzyme-linked immunosorbent assay (ELISA).

Statistical analysis

Statistical analyses were performed by GraphPad Prism 6 software. Data were analyzed by one-way ANOVA with Tukey’s multiple comparison tests (α = 0.05) or two-tailed paired Student’s t test, as indicated. Two-way ANOVA with least significant difference was used to analyze Fig. 2F. Data are means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

SUPPLEMENTARY MATERIALS

immunology.sciencemag.org/cgi/content/full/3/23/eaan0829/DC1

Materials and Methods

Fig. S1. Expression of HLA-DR, CD80, and CD86 on human DCs.

Fig. S2. Expression of TGF-βRII and IL-10Rα on human splenic basophils.

Fig. S3. Effect of IL-10 on the phenotype of human DCs.

Fig. S4. Anti-CD3 and anti-CD28 mAbs do not alter the phenotype of basophils.

Fig. S5. Blocking of ICAM-1 on DCs inhibits DC-mediated CD4+ T cell proliferation.

Fig. S6. ICOS-ICOSL interaction is dispensable for the Treg-mediated human basophil activation.

Fig. S7. Representative dot plots depicting the absence of CD4+ T cells in the lower chambers of transwells.

Fig. S8. IL-3 secretion by human Tregs.

Fig. S9. Dose-dependent effects of IL-3 on the phenotype of human basophils.

Fig. S10. Tregs license basophils to undergo activation by degranulation signals.

Fig. S11. The phenotypic and functional features of isolated human Tregs.

Fig. S12. The positive controls for annexin V and PI staining.

Table S1. Raw data from figure graphs (Excel).

REFERENCES AND NOTES

Acknowledgments: We thank N. Rambabu and the staff of Centre d’Histologie, d’Imagerie et de Cytométrie, Centre de Recherche des Cordeliers for the help, A. Fanidi for the statistical advice, and V. Languillat-Fouquet and H. Martelli, Service Chirurgie Pédiatrique, Hôpital Bicêtre, France for providing the spleen sections of patients with spherocytosis. Funding: This work was supported by the European Community’s Seventh Framework Programme (FP7/2007-2013) under grant agreement HEALTH-2010.2.4.5-2 ALLFUN, INSERM, Sorbonne Université, and Université Paris Descartes. M.S., E.S.-V., and A.K. were recipients of fellowship from Indo-French Center for Promotion of Advanced Research. C.G. is a recipient of fellowship from La Fondation pour la Recherche Médicale, France (FDM20150633674). Author contributions: M.S. and J.B. conceptualized the study. M.S., M.D., E.S.-V., and J.B. designed the experiments. M.S., M.D., E.S.-V., C.G., A.K., and M.S.M. performed the experiments. M.S., M.D., E.S.-V., P.B., S.V.K., and J.B. analyzed and interpreted the data. M.S., M.D., E.S.-V., and J.B. performed the statistical analysis. P.B. provided the essential tools. M.S. and J.B. wrote the manuscript, and all authors edited and approved the manuscript. Competing interests: The authors declare that they have no competing interests. Data and materials availability: All data supporting the findings of this study are available within the article.
View Abstract

Navigate This Article