Despite accumulating evidence indicating that neurotransmitters released by the sympathetic nervous system can modulate the activity of innate immune cells, we still know very little about how norepinephrine impacts signaling pathways in dendritic cells (DC) and the consequence of that in DC-driven T cell differentiation. In this article, we demonstrate that β2-adrenergic receptor (β2AR) activation in LPS-stimulated DC does not impair their ability to promote T cell proliferation; however, it diminishes IL-12p70 secretion, leading to a shift in the IL-12p70/IL-23 ratio. Although β2AR stimulation in DC induces protein kinase A–dependent cAMP-responsive element–binding protein phosphorylation, the effect of changing the profile of cytokines produced upon LPS challenge occurs in a protein kinase A–independent manner and, rather, is associated with inhibition of the NF-κB and AP-1 signaling pathways. Moreover, as a consequence of the inverted IL-12p70/IL-23 ratio following β2AR stimulation, LPS-stimulated DC promoted the generation of CD4+ T cells that, upon TCR engagement, produced lower amounts of IFN-γ and higher levels of IL-17. These findings provide new insights into molecular and cellular mechanisms by which β2AR stimulation in murine DC can influence the generation of adaptive immune responses and may explain some aspects of how sympathetic nervous system activity can modulate immune function.
It is well established that a cross-talk exists between the nervous and immune systems, with important implications for homeostasis maintenance (1, 2). All immune organs receive innervation from tyrosine hydroxylase-positive fibers, suggesting that the release of neurotransmitters, such as norepinephrine (NE), by the sympathetic nervous system (SNS) could impact the function of immune cells (3, 4). Indeed, an increase in NE release in the spleen was described following the activation of Ag-specific Th cells and B cells, indicating that SNS activation occurs during an ongoing adaptive immune response and, thus, may influence its course (5). Particularly in the spleen and lymph nodes, the sympathetic innervation is restricted to the T cell zones where dendritic cells (DC) are also present (3, 4). It was reported that immune cells, including T lymphocytes and DC, express α- and β-adrenergic receptors (6–8), which allows them to respond to NE. Overall, accumulated evidence suggests that NE release due to SNS activation could influence some immune processes, such as Ag presentation and T cell activation, proliferation, and differentiation.
DC are potent APC and play an important role in linking innate and adaptive immune responses. The production of a distinct set of cytokines by DC upon different types of stimulation is a central feature in determining the generation of diverse effector adaptive immune responses. Usually, DC stimulated with LPS produce higher levels of IL-12p70 than IL-23, which is associated with Th1 differentiation. In contrast, higher levels of IL-23 than IL-12p70 are produced by zymosan-activated DC, which can instead facilitate the orchestration of Th17-mediated immune responses (9, 10).
With regard to DC, the effects of adrenergic receptor stimulation are being revealed. It is known that stimulation of different adrenergic receptors expressed by DC modifies their migration capabilities (11), cytokine secretion (12–14), Ag uptake (15), and ability to perform cross-presentation (14). Specifically with regard to cytokine secretion, it was shown that stimulation of β2-adrenergic receptor (β2AR) modulates cytokine production by activated DC, primarily by inhibiting some proinflammatory cytokines, such as TNF-α, IL-12, and IL-6, and by increasing IL-10 and IL-33 release by these cells (13, 14, 16). Although previous studies showed that β2AR agonists impair the ability of LPS-activated DC to promote Th1 differentiation, favoring instead the generation of Th2 cells (17), Th17 cells, and induced Foxp3+ T cells (18, 19), our understanding of how β2AR stimulation can impact DC-driven naive CD4+ T cell differentiation remains limited. Moreover, we know very little about how β2AR stimulation alters signaling pathways in DC to modulate LPS-induced cytokine secretion.
The goal of this study was to investigate the effects of NE and β2AR stimulation on LPS-activated DC function, evaluating their maturation, as well as the profile of cytokines produced upon LPS treatment and its impact on the differentiation of naive CD4+ T cells into IFN-γ– or IL-17–producing T cells. We also sought to determine the signaling pathways underlying β2AR-mediated modulation of cytokine secretion by DC. Our findings provide new insights into the molecular and cellular mechanisms by which β2AR stimulation in DC can influence the generation of adaptive-immune responses.
Materials and Methods
Murine recombinant GM-CSF was purchased from R&D Systems. LPS (ultrapure Escherichia coli O111:B4) and SP600125 (JNK inhibitor) were purchased from InvivoGen. l-NE hydrochloride, fenoterol hydrobromide (β2AR-specific agonist), SKF 86466 hydrochloride (α2aAR-specific antagonist), ICI 118,551 hydrochloride (β2AR-specific antagonist), H-89 dihydrochloride hydrate (protein kinase A [PKA] inhibitor), and 8-CPT-2Me-cAMP (specific exchange protein activated by cAMP [EPAC] activator) were purchased from Sigma-Aldrich.
All experiments were performed using female mice (8–12 wk of age) maintained under specific pathogen–free conditions. C57BL/6 mice were provided by Centro de Desenvolvimento de Modelos Experimentais para Medicina e Biologia, Universidade Federal de São Paulo. IL-10−/− mice were provided by the Department of Immunology, Biomedical Science Institute, University of Sao Paulo. β-arrestin 2−/− mice were purchased from The Jackson Laboratory (Bar Harbor, ME), according to the animal protocol with the full knowledge and permission of the Standing Committee on Animals at Harvard Medical School. This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the Brazilian National Council of Animal Experimentation. The protocol was approved by the Committee on the Ethics of Animal Experimentation at the Federal University of Sao Paulo.
Generation of bone marrow–derived DC
DC were generated from bone marrow of female mice (8–12 wk) using supplemented DMEM and 20 ng/ml GM-CSF (R&D Systems). Cultures were fed on day 4 by the addition of 1 ml supplemented DMEM with 20 ng/ml GM-CSF. On day 7, 500 μl medium was aspirated, and 1 ml fresh supplemented DMEM was added. On day 9, immature DC (iDC) were obtained, and the experiments were performed as described. The purity of CD11c+ cells (≥95%) was confirmed by flow cytometry.
Cytokine production and DC maturation upon LPS stimulation
iDC culture was treated or not with NE (1 μM), fenoterol (1 μM), or SP600125 (10 μM) for 1 h and activated with LPS (5 μg/ml) for 18 h. In some experiments, cells were pretreated with SKF 86466 (1 μM) or ICI 118,551 (1 μM) for 15 min and then stimulated as mentioned above. Supernatants were collected, and IL-12p70, IL-23, and IL-10 production was assessed using an ELISA kit (eBioscience). For DC maturation analysis, cells were collected and preincubated with Fc Block (eBioscience) for 20 min on ice. Cells were then stained with fluorochrome-conjugated Abs against surface markers CD11c, MHC class II, CD40, and CD86 in staining PBS buffer for 30 min on ice and then washed. Cell acquisition was performed on a FACSCanto II instrument using FACSDiva software (BD Biosciences). Data were analyzed using FlowJo software (TreeStar).
Intracellular cAMP measurement
To assess intracellular cAMP levels, 5 × 105 iDC were treated for 15 min with the phosphodiesterase inhibitor IBMX (Sigma-Aldrich), followed by the indicated treatment. Cells were washed, and cAMP was measured using the cAMP-Glo Max Assay Kit (Promega).
Quantitative real-time PCR analysis
Total RNA of sorted CD11c+ iDC and CD11c+ LPS-stimulated DC (LPS-DC), treated or not with fenoterol at different time points, was obtained using the TRIzol (Invitrogen) method, following the manufacturer’s instructions. cDNA was synthesized from 2 μg RNA with a SuperScript II Reverse Transcriptase kit (Invitrogen). Gene expression was evaluated by quantitative real-time PCR using the SYBR Green-based system of detection (Applied Biosystems) and the GAPDH gene as the reference for normalizations. Relative expression was calculated by the ∆∆Ct method to compare expression levels of the same transcript in different samples or by the ∆Ct method to compare expression levels of several transcripts in the same sample. The following primers were used: Gapdh: forward, 5′-AAATGGTGAAGGTCGGTGTG-3′ and reverse, 5′-TGAGGGGTCGTTGATGG-3′; Adra1a: forward, 5′-TGGCTGCCATTCTTCCTCGTGA-3′ and reverse, 5′-TTCTTGAACTCCTGGCTGGAGC-3′; Adra1a: forward, 5′-TGGCTGCCATTCTTCCTCGTGA-3′ and reverse, 5′-TTCTTGAACTCCTGGCTGGAGC-3′; Adra1b: forward, 5′-TCTTCATCGCTCTCCCACTT-3′ and reverse, 5′-AGGCAGCTGTTGAAGTAGCC-3′; Adra2a: forward, 5′-GTGACACTGACGGCTGGTTTG-3′ and reverse, 5′-GTGACACTGACGGCTGGTTTG-3′; Adrb1: forward, 5′-GCTGATCTGGTCATGGGATT-3′ and reverse, 5′-AAGTCCAGAGCTCGCAGAAG-3′; Adrb2: forward, 5′-GGGAACGACAGCGACTTCTT-3′ and reverse, 5′-GCCAGGACGATAACCGACAT-3′; Il12b: forward, 5′-TTGAACTGGCGTTGGAAGCACG-3′ and reverse, 5′-CCACCTGTGAGTTCTTCAAAGGC-3′; Il12a: forward, 5′-ACGAGAGTTGCCTGGCTACTAG-3′ and reverse, 5′-CCTCATAGATGCTACCAAGGCAC-3′; and Il23a: forward, 5′-AATAATGTGCCCCGTATCCAG-3′ and reverse, 5′-GCTCCCCTTTGAAGATGTCAG-3′.
Western blotting analysis
After stimulation, cells were washed twice with cold PBS, and the total protein and cytoplasmic/nuclear protein extracts were obtained using RIPA buffer containing phosphatase and protease inhibitors (Pierce-Thermo Scientific) and the Cell Fractionation Kit (Cell Signaling Technology), respectively. Protein quantification was performed using BCA Protein Assay (Pierce-Thermo Scientific). From the total protein extracts, the total and IκBα, cAMP-responsive element–binding protein (CREB), and c-Jun phosphorylation were determined using anti–phospho-IκBα (Ser32) Ab, anti–phospho-CREB (Ser133) Ab, and anti–phospho-c-Jun (Ser63) Ab, respectively (all from Cell Signaling Technology). NF-κBp65 translocation was evaluated in fractionated protein extracts using anti–NF-κBp65 Ab (Cell Signaling Technology). Anti-GAPDH (Cell Signaling Technology) and anti-H3 (Millipore) were used to assay endogenous control to total/cytoplasmic protein and nuclear protein extracts, respectively. Chemoluminescence was detected by photoilluminator Alliance 4.7 and luminosity by UVIband software (both from UVItec Cambridge). The quantification of band intensities was performed using ImageJ.
In vitro naive CD4+ T cell proliferation and differentiation
iDC were treated with fenoterol (1 μM) for 1 h, followed by LPS (5 μg/ml) stimulation for 6 h. Cells were washed with PBS to remove the reagents and cocultured with sorted naive T cells (CD4+CD62L+) stained with Cell Proliferation Dye eFluor 670 (eBioscience) plus soluble anti-CD3 Ab at 1 μg/ml (eBioscience) at a DC/T cell ratio of 1:4. After 3 d, cells were recovered and stained with anti-CD4 for proliferation analysis using FACS. The proliferation analysis was performed using the proliferation platform built into FlowJo software. Division index is defined as the average number of cell divisions that a cell in the original population has undergone. This is an average even for cells that never divided (i.e., includes the undivided peak). Supernatants were collected for IL-17A and IFN-γ production measurement using an ELISA kit (eBioscience).
Data were analyzed in Prism software version 6 (GraphPad Software) using the Student t test for comparisons of two groups. One-way ANOVA was used for three or more unpaired groups. The p values <0.05 were considered significant. Data shown are mean ± SEM.
iDC express functional β2AR
We first investigated the adrenergic receptor expression profile in bone marrow–derived DC (BMDC) using real-time PCR. RNA samples were taken from immature or mature (LPS-activated) sorted CD11c+ DC. The relative gene-expression analysis showed that iDC express α- and β-adrenergic receptors, with β2AR being the most highly expressed (Fig. 1A). β2AR is also the most expressed adrenergic receptor in mDC (Supplemental Fig. 1). We also observed that iDC express higher β2AR mRNA and protein levels than mDC (Fig. 1B, 1C). To test whether β2AR expressed by iDC are functional, we measured intracellular cAMP levels upon treatment of these cells with NE or a β2AR agonist. As expected, iDC responded with a marked elevation in intracellular cAMP levels following exposure to 1 μM fenoterol (specific β2AR agonist) or 1 μM NE for 15 min (Fig. 1D). Because NE also activates α2a receptors, which are coupled to Gαi, it is reasonable that we observed a more intense elevation in cAMP levels following treatment with the β2AR agonist compared with that after NE treatment. Importantly, this effect was abrogated when these cells were pretreated with ICI-118.551, a specific β2AR antagonist (Fig. 1D). Canonical β2AR signaling was shown to increase intracellular cAMP levels, which activate cAMP-dependent PKA. PKA activation, in turn, induces CREB phosphorylation and translocation to the cell nucleus, where it competes with NF-κB for its interaction with CREB-binding protein, resulting in inhibition of proinflammatory gene transcription (20, 21). In agreement with our results shown in Fig. 1D, we found that NE and the β2AR agonist induced significant phosphorylation of CREB in iDC, which was dose-dependently inhibited by H89, a specific PKA inhibitor (Fig. 1E). Thus, iDC express functional β2AR, which activate the canonical signaling pathway.
β2AR signaling does not affect DC maturation
Because β2AR are functional on DC, we next investigated whether β2AR agonists could influence DC maturation and their capacity to support T cell proliferation. For this, iDC were treated or not with NE or fenoterol for 1 h and stimulated with LPS for 18 h. DC activation was analyzed by an increase in costimulatory molecules and MHC class II expression using flow cytometry. Although LPS-activated DC showed increased expression of CD40, CD86, and MHC class II, no effect was seen as a result of treatment with NE (Fig. 2A, Supplemental Fig. 2A), fenoterol (Fig. 2B, Supplemental Fig. 2B), or the receptor antagonists. Consistent with this, NE treatment, as well as β2AR signaling, in LPS-stimulated DC (LPS-DC) did not affect their capacity to support naive CD4+ T cell proliferation (Fig. 2C, 2D). Thus, activation of β2AR does not impair DC maturation or its ability to induce naive T cell proliferation.
β2AR signaling in DC modulates LPS-induced expression of IL-12 family cytokine genes and promotes a shift in the IL-12p70/IL-23 ratio
We next investigated the effects of β2AR activation on cytokine production by LPS-DC, focusing on the Th1 and Th17 axis. As expected, LPS stimulation induced higher levels of IL-12p70 than IL-23 in DC (Fig. 3A). NE treatment of LPS-DC downregulated IL-12p70 but did not affect IL-23 production (Fig. 3A), which resulted in a shift in the IL-12p70/IL-23 ratio in favor of IL-23 (Fig. 3B). In addition, IL-12p70 production was restored when LPS-DC were pretreated with a selective β2AR antagonist but not with a selective α2AR antagonist (Fig. 3A, 3B), indicating that the downregulation of IL-12p70 production by NE-treated LPS-DC is mediated by β2AR. This result was further confirmed using fenoterol (β2AR agonist) during DC activation with LPS (Fig. 3C, 3D). It is important to mention that the treatment with the β2AR agonist alone did not induce cytokine production (Supplemental Fig. 3A). In addition, we asked whether β2AR signaling could affect IL-6 production by LPS-DC, another important cytokine in the differentiation of Th17 cells. We found that treatment with the β2AR agonist decreased LPS-induced IL-6 secretion by DC, which was restored when the cells were pretreated with a β2AR antagonist (Supplemental Fig. 3B). Because IL-12p70 and IL-23 are heterodimeric proteins that share the p40 subunit, we analyzed, at different time points, IL-12 family gene expression in LPS-DC treated or not with a β2AR agonist. We found that treatment with the β2AR agonist induced a downregulation of Il12p40 and Il12p35 genes, whereas it upregulated Il23p19 transcripts in LPS-DC (Fig. 3E). It was shown previously that the IL-23p19 promoter region possesses binding sites for CREB and C/AATT enhancer–binding protein β. The latter can be phosphorylated via EPAC. EPAC (via C/AATT enhancer–binding protein β) and phospho-CREB were shown to stimulate p19 transcriptional activity (22). Accordingly, we were able to show that a specific EPAC activator (8-CPT-2Me-cAMP) induced increased IL-23 production in LPS-DC (Supplemental Fig. 4). Thus, phospho-CREB and/or EPAC activation following β2AR signaling could explain the increased LPS-induced p19 transcription in DC. It is important to mention that, although we found a decrease in Il12p40 expression, this subunit is expressed at much higher levels than Il12p35 and Il23p19 (data not shown); therefore, IL-12p40 expression is not a limiting factor for IL-12 or IL-23 production under these experimental conditions. Thus, β2AR activation modulates IL-12 family cytokines at the transcriptional level, leading to a shift in the IL-12p70/IL-23 ratio in LPS-DC.
The downregulation of IL-12p70 mediated by β2AR activation in LPS-DC occurs in a PKA-, EPAC-, and IL-10–independent manner
We further investigated the signaling pathways involved in the β2AR-mediated modulation of IL-12p70 production by LPS-activated DC. Based on our own data (Fig. 4A, 4B) and previous studies (14, 15) showing that NE and β2AR agonists increase IL-10 production by LPS-DC, we examined whether β2AR-mediated downregulation of IL-12p70 production by LPS-DC results from increased IL-10 release. Using iDC generated from IL-10−/− mice, we found that IL-10 deficiency does not affect β2AR-mediated inhibition of LPS-induced IL-12p70 production in DC (Fig. 4C). Next, based on our findings showing that β2AR signaling activates the canonical pathway, resulting in increased intracellular cAMP levels (Fig. 1D) and PKA-dependent CREB phosphorylation (Fig. 1E), we asked whether PKA activation is responsible for the β2AR-mediated IL-12p70 inhibition in LPS-DC. We found that PKA inhibition by H89 treatment did not restore IL-12p70 production in LPS-DC treated with the β2AR agonist (Fig. 4D). Finally, we investigated whether EPAC, which is a PKA-independent cAMP-mediated signaling pathway, is involved in the β2AR-mediated IL-12p70 inhibition in LPS-DC. We found that activation of EPAC-dependent signaling pathways did not mimic the inhibition of IL-12p70 production by LPS-DC that occurs upon β2AR activation (Fig. 4E). Taken together, these results demonstrate that IL-12p70 downregulation induced by β2AR activation in LPS-DC is independent of IL-10, EPAC, and PKA.
Inhibition of NF-κB and AP-1 are involved in IL-12p70 downregulation mediated by β2AR activation of LPS-DC
An alternative signaling pathway downstream of β2AR activation, involving NF-κB inhibition through the interaction of β-arrestins 1 and 2 with the IκBα complex in the cell cytoplasm, was described recently (23, 24). This interaction prevents the phosphorylation and degradation of IκBα, resulting in the inhibition of NF-κB (p65) translocation to the cell nucleus (23, 24). To evaluate the influence of β2AR signaling on NF-κB activation in LPS-DC, we performed Western blotting analysis of phospho-IκBα and NF-κB (p65) translocation in iDC and LPS-DC treated or not with fenoterol. As expected, we found that LPS activation induced an increase in IκBα phosphorylation in DC compared with iDC. Treatment with the β2AR agonist before LPS stimulation decreased IκBα phosphorylation in DC (Fig. 5A). Consistent with this, β2AR activation of LPS-DC decreased NF-κB (p65) translocation to the cell nucleus (Fig. 5B). To assess whether β-arrestin 2 was involved in NF-κB–mediated IL-12p70 downregulation in LPS-DC, we analyzed IL-12p70 production in DC from β-arrestin 2−/− mice. As shown in Fig. 5C, fenoterol (at 1 and 0.01 μM) inhibited nearly 50% of the IL-12p70 production by LPS-DC from wild-type (WT) mice. The effect of β2AR signaling on IL-12p70 production by LPS-DC was partially abrogated in β-arrestin 2−/− DC. Collectively, these data indicate that the downregulation of IL-12p70 mediated by β2AR activation of LPS-DC is at least partially dependent on β-arrestin 2, an effect most likely related to the inhibition of NF-κB (p65) activation. Moreover, these data suggest that additional mechanisms are involved in the modulation of IL-12p70 production by β2AR signaling.
In accordance with previous data indicating that AP-1 transcription factor is also important for IL-12 production (25–27), we found that LPS stimulation induces an increase in phospho–c-Jun in DC (Fig. 6A). Indeed, we also found that SP600125, a selective JNK inhibitor, was able to inhibit c-Jun phosphorylation (Fig. 6A) and downregulate IL-12p70 production (Fig. 6B) in LPS-DC. Because we confirmed that AP-1 is important for IL-12p70 production by LPS-DC, we decided to investigate whether β2AR signaling could also impair c-Jun phosphorylation. Fig. 6A shows that, in fact, β2AR activation leads to an inhibition of LPS-induced c-Jun phosphorylation in DC. Moreover, we found that β2AR-mediated blockade of LPS-induced c-Jun phosphorylation in DC is not dependent on PKA because H89 treatment did not restore c-Jun phosphorylation (Fig. 6C). Thus, impairment of AP-1 activation is likely to be an additional pathway through which β2AR signaling could inhibit IL-12p70 production by LPS-DC.
β2AR signaling in LPS-DC impacts CD4+ T cell differentiation, inhibiting IFN-γ and favoring IL-17A production
Finally, we evaluated whether the shift in the IL-12p70/IL-23 ratio in LPS-DC induced by β2AR activation could impact the adaptive-immune response, as measured by in vitro T cell differentiation. CD4+CD62L+ T cells were sorted and cocultured for 3 d with iDC or LPS-DC treated or not with β2AR agonists, plus soluble anti-CD3 Ab. In the coculture of naive T cells with LPS-DC, we detected more IFN-γ, a Th1 cell–derived cytokine, than IL-17A, a cytokine produced by Th17 cells (Fig. 7). In contrast, we observed a decrease in IFN-γ and an increase in IL-17A production when naive T cells were cocultured with LPS-DC that had been treated with a β2AR agonist or NE (Fig. 7). We also measured a cytokine associated with Th2 responses, but we did not detect IL-4 production in any experimental condition (data not shown). Thus, in agreement with its effects on IL-12p70/IL-23 production, β2AR activation of LPS-DC may facilitate Th17 responses in detriment of Th1 differentiation.
DC are of paramount importance because they drive the generation of adaptive-immune responses through Ag presentation and T cell activation and differentiation. In particular, the set of cytokines produced by stimulated DC encountering pathogenic agents is related to their ability to promote the differentiation of different CD4+ T cell subsets which shall orchestrate the most appropriate effector response to handle each type of infection. In fact, sensing the presence of LPS via TLR4 usually leads DC to produce higher amounts of IL-12p70 than IL-23 (28, 29), driving the generation of Th1-mediated responses that are better able to clear up infections caused by intracellular bacteria. In contrast, upon challenge with zymosan or β-glucans, dectin-1–stimulated DC preferentially secrete IL-23, promoting Th17-mediated responses that are better suited to deal with fungal infections (28, 30). Thus, matching the set of cytokines produced following DC activation with generation of the more appropriate adaptive effector response is critical to efficient control of infections by the host. Uncovering physiological stimuli that may interfere with DC sensing of potential pathogens is instrumental to our understanding of how and why, in some instances, hosts develop adaptive-immune responses that are not adequate to control a particular type of infection. Resting DC are likely to be reached by neurotransmitters and neuropeptides released upon the steady-state tonus of firing sympathetic fibers present in many organs, such as the spleen, lymph nodes, and skin. In stress-related conditions, due to the increased sympathetic tonus, it is possible that resting DC are exposed to these mediators to an even greater degree. In general, we still know very little about how SNS activity can influence the behavior of DC and, in particular, whether neurotransmitters released by the SNS can interfere with the way that DC sense potential pathogens.
In the current study, we demonstrate that, upon β2AR stimulation, instead of producing more IL-12p70 than IL-23, DC display a different response following LPS challenge and secrete about twice as much IL-23 as IL-12p70. This alteration in the IL-12p70/IL-23 ratio impacts DC-driven in vitro T cell differentiation. In fact, we found that, following β2AR stimulation, LPS-activated DC promoted the generation of CD4+ T cells that, upon TCR engagement, produce lower amounts of IFN-γ and higher levels of IL-17.
Consistent with previous findings (12, 14), we observed that the β2AR-mediated shift in the IL-12p70/IL-23 ratio following LPS challenge of DC was due to decreased IL-12p70 production (with no alteration in IL-23 secretion) and was accompanied by increased IL-10 production. Because IL-10, a conceptually anti-inflammatory cytokine, inhibits IL-12 production in LPS-activated immune cells (31, 32), we used BMDC from IL-10−/− mice to investigate whether increased IL-10 production by LPS-DC treated with β2AR agonists was responsible for downregulating IL-12 in these cells. However, we did not find any difference in IL-12 secretion between WT and IL-10−/− LPS-DC that had been treated with β2AR agonists, suggesting an IL-10–independent mechanism.
β2AR belong to the G protein–coupled receptor family and are usually associated with the stimulatory subtype of G protein (Gs) (4, 33). The canonical β2AR signaling pathway activation results in increased adenylyl cyclase activity that leads to augmented intracellular cAMP levels, which, in turn, promote PKA-mediated CREB phosphorylation (4, 33). Indeed, we found that β2AR activation in DC also results in increased cAMP levels and PKA-dependent CREB phosphorylation. However, although H89 treatment (PKA inhibitor) totally prevented CREB phosphorylation, it did not reverse IL-12 downregulation in LPS-DC upon β2AR activation, which suggests a PKA-independent mechanism.
It is known that β2AR activation may also lead to a distinct set of signaling events culminating with the stimulation of MAPK, which is mediated, not by Gs proteins, but by the βγ subunits of pertussis toxin–sensitive inhibitory subtype of G protein (Gi) proteins (34). Recently, β2AR activation was shown to inhibit Ag cross-presentation by DC, and this effect seemed to be mediated through Gi protein (14). Because the switching of the coupling of β2AR from Gs to Gi requires the PKA-dependent phosphorylation of the receptor (14, 34, 35), in our case it is unlikely that β2AR-mediated inhibition of IL-12 secretion by LPS-DC occurs through Gi protein–dependent signaling pathways.
β-arrestin 1 and 2 are well-known regulators of G protein–coupled receptor signaling, being involved in their internalization and consequent desensitization. More recently, it was also demonstrated that β-arrestin 1 and 2 interact with IκBα (24). This interaction prevents phosphorylation and degradation of IκBα and, thus, inhibits NF-κB translocation to the cell nucleus and attenuates the transcription of NF-κB target genes (23, 24, 36). In addition, β2AR stimulation significantly increases the amount of β-arrestin 2 that is associated with IκBα, which stabilizes IκBα and decreases TNF-α–induced phosphorylation and degradation of endogenous IκBα, with a consequent reduction in the expression of NF-κB target genes, such as IL-6 and IL-8 (23). Interestingly, β2AR-mediated enhancement of the β-arrestin 2–IκBα interaction was not mimicked by forskolin, indicating that this is unlikely to be mediated by the increase in cAMP levels and consequent PKA activation (23).
In fact, we verified that stimulation of β2AR resulted in decreased IκBα phosphorylation following DC activation with LPS. In agreement, we also found that β2AR stimulation inhibited LPS-induced NF-κB p65 translocation to the DC nucleus. We then decided to test whether β-arrestin 2 was involved in the β2AR-mediated downregulation of IL-12 in LPS-activated DC. In experiments using BMDC from β-arrestin 2−/− mice, we found that the absence of β-arrestin 2 partially prevented the β2AR-mediated reduction in IL-12 in LPS-activated DC. These data indicate that, although β2AR stimulation may lead to β-arrestin 2–dependent inhibition of NF-κB activation, with consequent downregulation of IL-12 production by LPS-activated DC, additional mechanisms are likely involved. Also, we cannot ignore that the effect seen on β-arrestin 2−/− DC was only partially due to the fact that β-arrestin 1 can also bind to IκBα and block NF-κB activation (24).
The role of AP-1 in controlling IL-12 production by innate immune cells was described (25, 27). Hence, JNK inhibition was shown to decrease IL-12p70 secretion by LPS-stimulated human monocyte-derived DC (26). In this study, we showed that JNK inhibition blocked LPS-induced c-Jun phosphorylation and impaired IL-12p70 production by DC, corroborating the notion that AP-1 is an important transcription factor promoting IL-12p70 expression after LPS stimulation in DC. Moreover, we also demonstrated that β2AR signaling inhibited AP-1 transcription factor activation, decreasing c-Jun phosphorylation in LPS-DC. Interestingly, β2AR-mediated inhibition of c-Jun phosphorylation was PKA independent, just as was the β2AR-mediated decrease in IL-12 secretion upon LPS challenge. Altogether, these data indicate that β2AR signaling suppresses the NF-κB and JNK signaling pathways, which are important to induce IL-12 production by LPS-DC.
Finally, although it is tempting to speculate about the implications of our findings in terms of the development and progression of autoimmune diseases, we believe that this is not a simple subject because IFN-γ–producing T cells are also involved in many autoimmune disorders. In conclusion, our results unravel that β2AR activation, through NF-κB and AP-1 inhibition, modifies the cytokine profile elicited by LPS stimulation in DC, shifting the IL-12/IL-23 ratio and impairing the development of Th1 cells.
We thank Daniela Teixeira for performing the cell-sorting procedures.
This work was supported by Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) Grant 08/58564-9 and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) Grants 475000/2010-2 and 246252/2012-0 (Ciências sem Fronteriras).
The online version of this article contains supplemental material.
Abbreviations used in this article:
bone marrow–derived DC
cAMP-responsive element–binding protein
exchange protein activated by cAMP
inhibitory subtype of G protein
stimulatory subtype of G protein
protein kinase A
sympathetic nervous system
The authors have no financial conflicts of interest.