Cholera toxin enhances interleukin‐17A production in both CD4⁺ and CD8⁺ cells via a cAMP/protein kinase A‐mediated interleukin‐17A promoter activation

Summary

Cholera toxin (CT) is a bacterial component that increases intracellular cAMP levels in host cells and suppresses T‐cell activation. Recently, CT was reported to induce T helper type 17‐skewing dendritic cells and activate interleukin‐17A (IL‐17A) production in CD4⁺ T cells through a cAMP‐dependent pathway. However, the underlying mechanism by which cAMP regulates IL‐17A production in T cells is not completely defined. In this study, we took advantage of a small molecule protein kinase A (PKA) inhibitor (H89) and different cAMP analogues: a PKA‐specific activator (N6‐benzoyl‐adenosine‐cAMP), an exchange protein activated by cAMP‐specific activator (Rp‐8‐chlorophenylthio‐2′‐O‐methyl cAMP), and a PKA inhibitor (Rp‐8‐bromo‐cAMP), to elucidate the signalling cascade of cAMP in IL‐17A regulation in T cells. We found that CT induced IL‐17A production and IL‐17A promoter activity in activated CD4⁺ T cells through a cAMP/PKA pathway. Moreover, this regulation was via cAMP‐response element binding protein (CREB) ‐mediated transcriptional activation by using the transfection of an IL‐17A promoter–luciferase reporter construct and CREB small interfering RNA in Jurkat cells. Also, we showed that CREB bound to the CRE motif located at −183 of the IL‐17A promoter in vitro. Most interestingly, not only in CD4⁺ T cells, CT also enhanced cAMP/PKA‐dependent IL‐17A production and CREB phosphorylation in CD8⁺ T cells. In conclusion, our data suggest that CT induces an IL‐17A‐dominated immune microenvironment through the cAMP/PKA/CREB signalling pathway. Our study also highlights the potentials of CT and cAMP in modulating T helper type 17 responses in vivo.

Abbreviations

BNZ‐cAMP

N6‐benzoyl‐adenosine‐cAMP

cAMP

cyclic AMP

CRE

cAMP response element

CREB

cAMP response element‐binding protein

CT

cholera toxin

EMSA

electrophoretic mobility shift assay

EPAC

exchange protein activated by cAMP

IL‐1β

interleukin‐1β

O‐Me‐cAMP

Rp‐8‐chlorophenylthio‐2′‐O‐methyl cAMP

PARP

poly (ADP‐ribose) polymerase

PKA

protein kinase A

Rp‐Br‐cAMP

Rp‐8‐Bromo‐cAMP

siRNA

small interfering RNA

Th17

T helper type 17

Treg

regulatory T

Introduction

During infection, bacterial pathogens manipulate critical intracellular pathways in host cells to benefit their own growth and dissemination, such as metabolism, cell cycle, cytoskeletal organization and host immune defence.1 Bacterial toxin is one of the bacterial components that pathogens use to hijack host systems.2, 3 Bacterial toxins may modulate the immune response through suppression of immune cell activation, enhancement of antigen uptake or release of cytokines.2 As bacterial toxins can easily cross the cell membrane to induce their immunomodulatory effects, they have been widely studied in immune therapy and vaccine development.4, 5

Cholera toxin (CT) from Vibrio cholerae is one of the best‐studied bacterial toxins. CT is an ADP‐ribosylating enterotoxin, consisting of catalytic A and B subunits, where the CT‐B subunit can bind to the GM1 ganglioside receptor in all nucleated cells. The released CT‐A subunit interacts with the host's Gα subunit and results in the stimulation of host adenylate cyclase activity, which elevates the intracellular cAMP levels in various cells.6 The activity of the CT‐A subunit is also associated with its immunomodulatory function. CT has been reported to induce interleukin‐1β (IL‐1β), IL‐6 and IL‐10 but inhibit IL‐12 and tumour necrosis factor‐α in lipopolysaccharide‐stimulated bone‐marrow‐derived dendritic cells.7 CT also directly modulates cytokine production from CD4⁺ T cells, which selectively activates IL‐17A production but suppresses IL‐2, IL‐10, interferon‐γ and tumour necrosis factor‐α. It is worth noting that this IL‐17A induction is also mediated by elevated intracellular cAMP levels.8

In both mouse and human systems, IL‐17A/T helper type 17 (Th17) cells have been implicated in the pathogenesis of certain autoimmune diseases such as multiple sclerosis and rheumatoid arthritis.9 However, in mucosal defence, IL‐17A plays a crucial role in the induction of various antimicrobial peptides and cytokines (human β‐defensin, S100A7/8, and CCL20),10, 11 and also in the recruitment of neutrophils to inflamed tissue.12, 13 Therefore, how to fine‐tune Th17 cell effector function has become critical. To date, the mechanism by which some cytokines regulate IL‐17A expression has been well studied, such as Janus kinase/signal transducer and activator of transcription signalling induced by IL‐2, IL‐6, IL‐21 and IL‐2314, 15, 16 and mammalian target of rapamycin signalling pathways induced by IL‐1β.17, 18 We and other laboratories have reported that cAMP is involved in IL‐17A induction by Th17 cells under either prostaglandin E₂ 19, 20 or CT treatment.8 However, the downstream signals of cAMP in the IL‐17A regulation is unclear.

In this study, we used a small molecule protein kinase A (PKA) inhibitor (H89) and different cAMP analogues, including a PKA‐specific activator [N6‐benzoyl‐adenosine‐cAMP (BNZ‐cAMP)], an exchange protein activated by cAMP (EPAC) ‐specific activator [Rp‐8‐chlorophenylthio‐2′‐O‐methyl cAMP (Rp‐Br‐cAMP)], and a PKA inhibitor (Rp‐8‐bromo‐cAMP), to elucidate the mechanism by which cAMP mediates IL‐17A production in T cells. We found that IL‐17A induction by CT is PKA dependent. Furthermore, we used an IL‐17A promoter‐reporter system and Jurkat cell line to demonstrate that the PKA/cAMP‐response element binding protein (CREB) pathway, at least in part, involves CT‐induced IL‐17A promoter activation. We also found that this cAMP/PKA‐mediated pathway was responsible for CT‐induced IL‐17A production not only in CD4⁺ but also in CD8⁺ T cells. Our findings highlight that this cAMP/PKA/CREB pathway may be a therapeutic target in IL‐17‐related diseases.

Materials and methods

Cell line and reagents

Jurkat cells are an immortalized line of T lymphocyte cells (a kind gift from Dr Fu‐Tong Liu; University of California, Davis, CA). Cholera toxin was purchased from List Biological Laboratories (Campbell, CA). PKA‐specific cAMP (BNZ‐cAMP) and PKA antagonist (Rp‐Br‐cAMP) were from Calbiochem (San Diego, CA); EPAC‐specific cAMP (Rp‐8‐chlorophenylthio‐2′‐O‐methyl cAMP (O‐Me‐cAMP) was from Enzo Life Sciences (Farmingdale, NY); H89 was from Sigma (St Louis, MO). Anti‐CREB, anti‐pCREB, and anti‐poly(ADP ribose) polymerase (PARP) antibodies were purchased from Cell Signaling (Danvers, MA) and anti‐RORγt antibodies were from Santa Cruz Biotechnology (Dallas, TX).

Purification and culture of total CD4⁺ T cells and Jurkat cells

The use of human cells in the study was periodically reviewed and approved by the University Human Subject Research Review Committee. Peripheral blood mononuclear cells were purchased from Allcells, LLC (Alameda, CA) or isolated from the leukopak filters from the Delta Blood Bank in Stockton, CA. Total CD4⁺ T cells were positively selected from peripheral blood mononuclear cells using magnetic CD4 microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer's protocol. The purity of CD4⁺ T cells was determined to be 98–99% by FACScan. CD4⁺ T cells were cultured in serum‐free X‐VIVO15 media along with beads coated with anti‐CD3 and anti‐CD28 antibodies (CD3/CD28 bead; 10 cells per bead; DynaBeads Human T‐Activator; Invitrogen, Carlsbad, CA). Jurkat cells were cultured in complete RPMI‐1640 medium supplemented with 10% fetal bovine serum, 1% sodium pyruvate and 1% penicillin/streptomycin.

Examination of IL‐17A production by ELISA

Interleukin‐17A was measured in cell culture supernatants by ELISA (human IL‐17 DuoSet, R&D Systems, Minneapolis, MN) according to the manufacturer's protocol.

DNA plasmid constructs and small‐interfering RNA

The construction of the IL‐17A promoter (229 bp) and promoter with two cAMP‐response element (CRE) mutations in the pGL‐3 basic vector was described previously.8 Constructs with single mutations of the CRE motifs, CRE1 (−183~−178) or CRE2 (−111~−104), were generated using the QuickChange II site‐directed mutagenesis kit (Agilent Technologies, Santa Clara, CA). The constructs were termed IL17p/WT (wild‐type), IL17p/CREmt1, IL17p/CREmt2 and IL17p/CREmt1/2. Primers used to create the mutations were described previously.8 CREB1‐specific small interfering RNA (siRNA) non‐targeted oligomer controls were purchased from Life Technologies (Carlsbad, CA).

Transient transfection of activated CD4⁺ T cells and Jurkat cells

CD4⁺ T cells were activated with plate‐coated anti‐CD3 antibodies and soluble anti‐CD28 antibodies for 16–20 hr. Activated cells were co‐transfected with the indicated plasmids together with the internal control, Renilla luciferase expression vector pRL‐CMV (Promega, Madison, WI), using the Nucleofector kit for stimulated human T cells (Amaxa). Twenty‐four hours after transfection, tranfected cells were treated with CT under CD3/CD28 activation as indicated. The siRNA transfection of Jurkat cells was also performed using the Nucleofector system. Twenty‐four hour after transfection, Jurkat cells were co‐transfected with the indicated plasmids and the internal control, Renilla luciferase expression vector pRL‐TK (Promega), using Lipofectamine 2000 (Life Technologies) according to the manufacturer's protocol. Twenty‐four hours after the transfection, cells were treated with 10 ng/ml CT together with 12‐O‐Tetradecanoylphorbol‐13‐acetate (TPA) and ionomycin (Cell Signaling Technology, Danvers, MA) for 5 hr. Cell extracts were lysed and luciferase activity was quantified in triplicate using a Dual‐Glo Luciferase Assay System (Promega) according to the manufacturer's protocol. The relative IL‐17A promoter activities were displayed as ratios to the pGL‐3 vector control after relative luciferase units were normalized to the internal control, Renilla luciferase activity. The results were averaged from triplicate wells of three separate experiments.

RNA isolation and quantitative RT‐PCR

Total RNA was extracted from cells using TRIzol according to the manufacturer's protocol and cDNA was synthesized from total RNA using Superscript III Reverse Transcriptase (Life Technologies) and Oligo(dT) reverse primer. SYBR Green quantitative real‐time PCR (Roche, Basel, Switzerland) was carried out to quantify the levels of CREB1 expression after normalization with the housekeeping gene, GAPDH. The gene‐specific primer sets were: CREB1, 5′‐CTC CAC AAG TCC AAA CAG TTC A‐3′ and 5′‐CAA TCC TTG GCA CTC CTG GT‐3′; GAPDH, 5′‐TGC ACC ACC AAC TGC TTA GC‐3′ and 5′‐GGC ATG GAC TGT GGT CAT GAG‐3′.

Protein purification and Western blotting

Whole cell lysates were prepared by lysing cells in RIPA buffer with Halt protease inhibitor (Thermo Fisher, Waltham, MA) and phosphatase inhibitor. Nuclear and cytosolic fractions were prepared using an NE‐PER Nuclear and Cytoplasmic Extraction kit (Thermo Scientific). Protein concentrations were measured using the dendritic cell protein assay (BioRad, Hercules, CA). Protein samples were run on 4–12% Bis–Tris SDS–PAGE gels (Life Technologies), followed by a wet transfer to PVDF membranes (USA Scientific, Ocala, FL). The membranes were blocked with 5% non‐fat milk in Tris‐buffered saline + 0·5% Tween‐20 at room temperature for 1 hr and probed with primary antibodies overnight, followed by hybridization with species‐specific horseradish peroxidase‐conjugated secondary antibody. The anti‐PARP antibody was used as a nuclear loading control.

Electrophoretic mobility shift assay and supershift assay

Electrophoretic mobility shift assay (EMSA) was performed with 4 μg of nuclear protein and 3′‐biotinylated IL‐17A CRE1 or CRE2 promoter oligonucleotides using a LightShift Chemiluminescent EMSA kit (Pierce, Rockford, IL). In order to define the transcription factors that bind to CRE probes, anti‐CREB1 antibodies were added to the reaction 1 hr before the addition of the probes. The protein–DNA complexes were run on 6% DNA Retardation gels (Life Technologies). The 3′‐biotin‐labelled oligonucleotides were prepared using the Biotin 3′ End DNA labelling kit (Thermo Scientific). The probe sequences were as follows: CRE1, CTA TGA CCT CAT TGC TAT GAC CTC ATT GCT ATG ACC TCA TTG‐3′ and CRE2, TAA GTG ACC ACA GAA GTA AGT GAC CAC AGA AGT AAG TGA CCA CAG AAG.

Statistics

Data are expressed as mean ± SEM. The P‐value was determined using two‐tailed Student's t‐test and one‐way analysis of variance for multiple groups with Bonferroni multiple comparison test. P‐values < 0·05 were considered significant.

Results

PKA but not EPAC is involved in CT‐induced IL‐17A production in CD4⁺ T cells

cAMP was first identified as a cytosolic secondary messenger of extracellular ligand stimulation.21 In T cells, cAMP transmits signalling from hormone receptors, inflammatory mediators as well as cytokines. cAMP has been thought to inhibit T‐cell activation via its major effector proteins: PKA22 and EPAC.23 To further study the mechanism by which cAMP regulates IL‐17A production, we used various cAMP analogues and antagonists to determine whether the effect of CT on IL‐17A production is PKA dependent. BNZ‐cAMP specifically binds and activates the PKA signalling pathway and mimics the effect of CT to increase IL‐17A expression in activated CD4⁺ T cells (Fig. 1a). This result was similar to previous findings in mouse T cells with prostaglandin E₂.19

Figure 1.

The cholera toxin (CT) ‐induced interleukin‐17A (IL‐17A) production is dependent on protein kinase A (PKA) signalling but not exchange protein activated by cAMP (EPAC). (a) Total CD4⁺ T cells were treated with CT (10 ng/ml) or BNZ‐cAMP (100 μm) for 3 days under CD3/CD28 activation (10 cells per bead). Culture supernatant was analysed for IL‐17A protein production by ELISA (n = 11). All groups were compared with the ‘–’ group (i.e. the group in which cells were not treated by any cAMP agents). Data were expressed as mean ± SEM (**P < 0·01; ***P < 0·001, two‐tailed paired t‐test). (b) CD4⁺ T cells under CD3/CD28 activation were treated with Rp‐Br‐cAMP (100 μm) 1 hr before adding BNZ‐cAMP (100 μm), O‐Me‐cAMP (100 μm) or the combination of BNZ‐cAMP and O‐Me‐cAMP for 3 days. Culture supernatant was analysed for IL‐17A protein production by ELISA (n = 7). Data were expressed as mean ± SEM (**P < 0·01; ***P < 0·001, one‐way analysis of variance with Bonferroni multiple comparison test).

In a study of human dendritic cells, EPAC alone showed no effect on the activation of dendritic cells whereas PKA signalling induced dendritic cell activation and maturation. However, the induction of both EPAC and PKA signalling pathways induced a tolerogenic dendritic cell, different from PKA‐induced dendritic cells.24 Therefore, we used O‐methyl‐cAMP (EPAC‐specific activator) or the combination of O‐methyl‐cAMP and BNZ‐cAMP to study whether EPAC also regulates the IL‐17A expression in CD4⁺ T cells. O‐Me‐cAMP alone did not up‐regulate IL‐17A production and had no synergistic effect with BNZ‐cAMP either (Fig. 1b). We also applied Rp‐Br‐cAMP in our study, a cAMP antagonist, which binds to PKA and blocks PKA catalytic activity. Cells were incubated with Rp‐Br‐cAMP for at least 1 hr before the treatment with BNZ‐cAMP. The induction of IL‐17A by BNZ‐cAMP was blocked by Rp‐Br‐cAMP (Fig. 1b). Again, the combination of O‐Me‐cAMP with BNZ‐cAMP did not reverse the blockade of Rp‐Br‐cAMP on IL‐17A production by BNZ‐cAMP alone (Fig. 1b). Our results confirm that only the cAMP‐PKA pathway is involved in IL‐17A production and EPAC has no effect on cAMP‐induced IL‐17A expression in activated CD4⁺ T cells.

CT‐induced IL‐17A promoter activity is via a PKA‐dependent pathway

In our previous study, we have shown that CT‐induced IL‐17A expression in activated CD4⁺ T cells is transcription‐dependent. Here, we examined whether the CT‐induced IL‐17A promoter activation is also via a PKA‐dependent pathway. A human IL‐17A promoter‐luciferase reporter construct was transfected into activated primary human CD4⁺ T cells to examine the effects of CT and cAMP analogues on CD4⁺ T cells. CT, db‐cAMP and BNZ‐cAMP had similar effects on the induction of IL‐17A promoter activity (Fig. 2a), which suggests that the PKA signalling pathway, at least in part, is involved in the CT‐induced IL‐17A promoter activation.

Figure 2.

Cholera toxin (CT) induced interleukin‐17A (IL‐17A) promoter activity is through a PKA‐dependent signalling pathway. (a) Primary CD4⁺ T cells were activated with CD3/CD28 antibodies for 16–18 hr and transfected with the indicated plasmids, pGL3 vector (vector) and pGL3‐IL17p/wild‐type (IL17p/WT), and the pRL‐CMV luciferase control plasmid. After 6 hr recovery, cells were treated with CT (10 ng/ml), db‐cAMP (100 μm), or BNZ‐cAMP (100 μm) for an additional 18 hr under CD3/28 activation. Luciferase activity was measured in the cell lysate after re‐stimulation with 12‐O‐Tetradecanoylphorbol‐13‐acetate (TPA) (50 ng/ml) and ionomycin (1 μm) for 4 hr (d, n = 3). (b) Jurkat cells were transfected with the indicated plasmids (vector and IL17p/WT) and the control pRL‐TK plasmid using Lipofectamine 2000. One day after transfection, transfected cells were pre‐treated with H89 and then treated with CT for 4 hr under the stimulation of TPA and ionomycin before lysing the cells for luciferase activity measurement. Luciferase activity is displayed as ratios to the pGL‐3 vector control after normalization to Renilla luciferase activity. Data represent three independent experiments (n = 3). Data were expressed as mean ± SEM (*P < 0·05; **P < 0·01, one‐way analysis of variance with Bonferroni multiple comparison test).

Next, we examined whether this PKA‐dependent IL‐17A promoter activation also occurs in Jurkat cells. As shown in the previous study, CT also enhanced IL‐17A promoter activity in Jurkat cells and the pre‐treatment of a PKA inhibitor H89 significantly inhibited CT‐induced human IL‐17A promoter activity (Fig. 2b). These results confirmed that Jurkat cells and primary human CD4⁺ T cells share the same mechanism to induce the activation of the IL‐17A promoter.

CREB1 is responsible for the CT‐induced IL‐17A promoter activity in CD4⁺ cells

In most conditions, cAMP‐stimulated gene transcription is mediated through the phosphorylation of CREB on Ser133 by PKA.25 We examined whether CT induces CREB phosphorylation in CD4⁺ T cells. We found that the phosphorylation of CREB in the nuclei of activated human primary CD4⁺ T cells was increased (~20 to 50%) after CT treatment (Fig. 3a).

Figure 3.

The phosphorylation of cAMP response element binding protein (CREB) is induced by CT both in activated primary CD4⁺ T cells and Jurkat cells. (a) Primary CD4⁺ T cells were activated by CD3/28 antibodies for 2 hr before cholera toxin (CT) treatment. Extracted nuclear protein was examined by Western blot and the anti‐PARP antibody was used as nuclear loading control. (n = 5). Data were expressed as mean ± SEM (*P < 0·05; **P < 0·01; ***P < 0·001, two‐tailed paired t‐test). (b) Jurkat cells were activated with 12‐O‐Tetradecanoylphorbol‐13‐acetate (TPA) (50 ng/ml) and ionomycin (1 μm) for 1 hr. H89 (3 μm) was added as indicated 1 hr before the treatment of CT. Nuclear extracts from treated cells were subjected to Western blot analysis to detect the phosphorylation of CREB (n = 3). Data were expressed as mean ± SEM (*P < 0·05; **P < 0·01; ***P < 0·001, one‐way analysis of variance with Bonferroni multiple comparison test).

Furthermore, we found that this CT‐mediated pCREB increase (~30 to 50% increase) also occurs in Jurkat cells (Fig. 3b). Our data showed again that Jurkat and primary T cells share a common signalling pathway in the CT‐induced IL‐17A promoter activation. Hence, we can take advantage of this cell line to study the regulation of CT‐mediated IL‐17A promoter activation. First, we found that this CT‐mediated CREB phosphorylation was blocked by H89 pre‐treatment (Fig. 3b). We also examined whether CREB was responsible for CT‐induced IL‐17A promoter activity. To test this, CREB was knocked down using specific siRNA oligonucleotides. Jurkat cells were transfected with a CREB‐specific siRNA and this specific siRNA significantly decreased both mRNA (Fig. 4a) and protein expression of CREB (Fig. 4b) by 50% compared with non‐targeted oligomer transfected controls. Likewise, CT‐induced IL‐17A promoter activity was reduced (Fig. 4c) in CREB‐knockdown cells. Together, these data reveal the crucial role of the cAMP/PKA/CREB cascades in the regulation of CT‐induced IL‐17A promoter activation.

Figure 4.

The small interfering RNA (siRNA) knockdown of cAMP response element binding protein (CREB) expression decreases the cholera toxin (CT) ‐induced interleukin‐17A (IL‐17A) promoter activity. Jurkat cells were transfected with siRNA against CREB or control siRNA (Ctrl). Total RNA (24 hr) and cell lysate (48 hr) from transfected cells were examined by using quantitative PCR and Western blot (a, b) The efficiency of siRNA knockdown of CREB was determined by the levels of mRNA and protein. (c) After siRNA transfection, Jurkat cells were then transfected with the IL‐17A promoter–reporter construct and treated with CT as previously described. The luciferase activity is displayed as fold change compared with the corresponding cells without CT treatment. Data represent three independent experiments (n = 3). Data were expressed as mean ± SEM (*P < 0·05; **P < 0·01, two‐tailed unpaired t test).

CREB1 binds to the IL‐17A promoter at CRE −183 but not CRE −111 in vitro

As noted in our previous work, the CRE motifs are the cis‐regulatory elements of CT‐induced IL‐17A promoter activity. The classical CRE sequence is an eight‐base‐pair palindrome (5′‐TGACGTCA‐3′).26, 27 We used TFSEARCH (Threshold score = 85) to find the first CRE motif located at (−183/−178) of the IL‐17A promoter. The other CRE motif (−111/−104) was previously demonstrated to play a role in CREMα‐mediated IL‐17A expression in patients with systemic lupus erythematosus.28 To further confirm which CRE motifs are necessary for the CT‐induced IL‐17A promoter activity, we generated IL‐17A promoter–reporter constructs with individual CRE mutations, termed IL17p/CREmt1 and IL17p/CREmt2. When transfected into Jurkat and primary CD4⁺ T cells, we found that the CRE motif at (−183/−178) but not the one at (−111/−104) was necessary for the CT‐induced IL‐17A promoter activity (Fig. 5a,b). Unexpectedly, we found that the mutation of CRE motif (−111/−104) enhanced the promoter activity and may act as a repressive motif. However, the CT‐induced effect was unaffected by this mutation.

Figure 5.

The direct binding of cAMP response element binding protein (CREB) to the CRE1 motif (–183) of interleukin‐17A (IL‐17A) promoter accounts for cholera toxin (CT) ‐induced IL‐17A production. (a) A schematic representation of the hIL‐17A promoter depicting cAMP‐responsive element motifs used for wild‐type and CRE‐mutated promoter constructs. (b) Primary CD4⁺ T cells and (c) Jurkat cells were transfected with indicated plasmids (IL17p/WT, CRE_mt1, CRE_mt2 and CRE_mt1/2). One day after transfection, cells were treated with CT for 4 hr under the stimulation of 12‐O‐Tetradecanoylphorbol‐13‐acetate (TPA) (50 ng/ml) and ionomycin (1 μm) and the luciferase activity was measured. Luciferase activity is displayed as ratios to the pGL‐3 vector control after normalization to Renilla luciferase activity. Data represent three independent experiments (n = 3). Data were expressed as mean ± SEM (*P < 0·05; **P < 0·01, two‐tailed paired t test). (d, e) Jurkat cells were stimulated with TPA and ionomycin (PI) and treated with or without CT for 2 hr. Nuclear extracts from Jurkat cells with indicated treatments were subjected to EMSA analysis with biotin‐labelled oligonucleotides, CRE1 or CRE2 probes. The band of biotin‐labelled probe and protein complex is indicated by arrows (→). Unlabelled probes were added as competitors and anti‐CREB antibody was added to detect the specific DNA‐binding protein as indicated. Data represent one of three independent experiments with similar results.

To examine if the CRE motifs of the IL‐17A promoter are bound by CREB in response to CT treatment, we generated two biotin‐labelled nucleotide probes – CRE1 (−183) and CRE2 (−111) probes – and incubated them with lysates from Jurkat cells treated or not with CT. As shown in Fig. 5(c,d), CT‐treated lysates increased the band intensity of the shifted band of the CRE1 probes by 47% but not the one of the CRE2 probes. The addition of anti‐CREB antibody decreased the intensity of the original shifted band but did not result in a supershifted band, indicating that this antibody may bind to CREB at its DNA‐binding site and disrupt the protein–DNA interaction. This finding suggests that CREB is the transcription factor that directly binds to the CRE motif at −183 of IL‐17A promoter in response to CT and activates the promoter activity of IL‐17A.

CT also induces IL‐17A production from CD8⁺ T cells via a PKA‐dependent pathway

Interleukin‐17A is also produced by CD8⁺ T cells29 and innate immune cells, such as γδ T cells,30, 31 NK1·1‐invariant natural killer T cells,32 neutrophils33 and innate lymphoid cells.34, 35, 36 These cells also share some characteristics with Th17 cells. For example, γδ T cells produce IL‐17A in response to IL‐1β and IL‐23.37, 38, 39 CD8⁺ T cells have also been reported to produce IL‐17A in the presence of IL‐1β, IL‐6, IL‐23, and transforming growth factor‐β.40 In order to determine whether CT induces IL‐17A in these other cell types, CD8⁺ T cells were isolated from adult peripheral blood mononuclear cells and stimulated with anti‐CD3 and anti‐CD28 in the presence of CT or other cAMP analogues. Although activated CD8⁺ T cells produced less IL‐17A than CD4⁺ T cells, CT increased IL‐17A production by approximately twofold in CD8⁺ T cells (Fig. 6a). We also examined whether dibutyl‐cAMP (a PKA and EPAC activator), BNZ‐cAMP and O‐Me‐cAMP were also able to induce IL‐17A production from CD8⁺ T cells. Only dibutyl‐cAMP and BNZ‐cAMP had the capacity to induce IL‐17A production (Fig. 6a), confirming the involvement of PKA but not EPAC. In addition, CT also induced more CREB phosphorylation in CD8⁺ T cells (Fig. 6b). These data indicate that CT enhances IL‐17A production in non‐CD4⁺ IL‐17A‐producing cells through the cAMP/PKA/CREB signalling pathway as well.

Figure 6.

Cholera toxin (CT) also induces interleukin‐17A (IL‐17A) production in CD8⁺ T cells via a protein kinase A (PKA) ‐dependent signalling pathway. Total CD8⁺ T cells were stimulated with CT (10 ng/ml), db‐cAMP (100 μm), BNZ‐cAMP (100 μm) and O‐Me‐cAMP (100 μm) for 3 days under CD3/CD28 activation (10 cells per bead). (a) IL‐17A protein expression was measured in cell culture supernatants by ELISA (n = 9). All groups were compared with the ‘–’ group (i.e. the group in which cells were not treated with any cAMP agents). Data were expressed as mean ± SEM (*P < 0·05; **P < 0·01, two‐tailed paired t‐test). (b) Primary CD8⁺ T cells were treated with CT for 1 hr after CD3/CD28 activation. Nuclear extracts from treated cells were subjected to Western blot analysis to detect the phosphorylation of cAMP response element binding protein (CREB). Data shown represents one of three independent experiments with similar results.

Discussion

Cyclic AMP is a universal regulator of normal physiological functions and is also used by many microbial pathogens to paralyse immune responses,41 such as pertussis toxin and CyaA toxin of the respiratory pathogen Bordetella pertussis,42, 43 and cholera toxin of the gastrointestinal pathogen Vibrio cholerae.6 In addition, in vitro HIV infection also dampens the T cells’ proliferative ability through elevated intracellular cAMP levels.44 In contrast to its immune suppression, our group and others found that CT and cAMP also induces Th17 responses by acting on dendritic cells45 or T cells.8 In this study, we demonstrate that CT induces IL‐17A expression by the activation of a cAMP/PKA/CREB cascade, then CREB trans‐activating IL‐17A promoter by binding to the cis‐regulatory element, CRE motif at −183, of the promoter. Furthermore, our study suggest that this PKA‐dependent regulation of IL‐17A expression may be universal as this effect also occurs in other IL‐17A‐producing cells, such as CD8⁺ T cells. This evidence indicates that the net effect of elevated cAMP levels on immune response prefers an IL‐17A‐dominated environment. However, whether this IL‐17A‐dominated response favours host protection or pathogen transmission is unclear and deserves further exploration.

Both PKA and EPAC are major effector molecules of cAMP.46 Binding of cAMP to the regulatory subunit of PKA results in the dissociation of the catalytic subunits, phosphorylating its downstream target proteins and transcription factors, such as CREB family members. Similar to PKA activation, cAMP also binds to EPAC1/2 at the regulatory subunit to release the catalytic subunits Rap1 and Rap2,47 activating downstream signalling such as Ca²⁺ release.48 The cAMP/PKA signalling is generally regarded as an anti‐inflammatory pathway because PKA inhibits TCR signalling,49 and IL‐2 and CD25 expression.50 However, recent findings of cAMP‐induced IL‐17A responses suggest dual roles of cAMP in immune regulation. On the other hand, whether EPAC signalling is involved in T‐cell immunology is unclear. One report showed that EPAC regulates gene expression in Jurkat cells via c‐Jun, although the biological function was unclear.51 However, our results showed that cAMP induced IL‐17A production in T cells through PKA but not EPAC signalling, suggesting that PKA is the major downstream target of cAMP in T cells.

CREB is a direct downstream transcriptional factor of PKA. The phosphorylation of CREB in T cells can be activated by PKA, protein kinase Cα/θ (PKCα/θ), Ras, extracellular signal‐regulated kinases1/2 and p80 ribosomal protein S6 Kinase.25, 52, 53, 54 In our data, we also detected CREB phosphorylation in activated Jurkat cells and primary CD4⁺ T cells and this activation may be through PKC activation.54 Recently, Wang et al.55 also showed that this CREB phosphorylation, activated by CD3‐PKC‐θ signalling, also promotes IL‐17A expression by binding to the mouse Il17‐Il17f gene locus. Of interest, our data showed that this alternative activation of CREB by cAMP‐PKA signalling further enhanced IL‐17A promoter activity. These findings suggest that CREB‐mediated IL‐17A expression is evolutionally conserved and multiple pathways can trigger CREB binding to the IL‐17A promoter. However, how these pathways cooperate together in the regulation of CREB function needs to be investigated. Interestingly, CREB also promotes transforming growth factor‐β‐mediated Foxp3 expression and the generation of regulatory T (Treg) cells.56, 57 It has been found that Treg cells have higher levels of intracellular cAMP than conventional T cells and they execute their suppressive function through the transfer of cAMP to target cells.58 Whether higher cAMP is also vital in the maintenance of Treg cell functions and whether cAMP or CREB regulates the balance of Th17 cells and Treg cells via different signalling pathways is not well understood. Our data imply that Th17 and Treg cells not only share the transforming growth factor‐β signalling pathway in common, but also the cAMP signalling pathway in the regulation of a proper immune response.

Based on the result of TFSEARCH and literature evidence, two putative CRE motifs in the IL‐17A promoter were studied for this CREB binding. Our data demonstrated that the CRE motif at −183 of the IL‐17A promoter is responsible for CT‐induced IL‐17A expression and for the CREB binding. Unexpectedly, the CRE mutation at −111 enhanced IL‐17A promoter activity. Although CT‐induced promoter activity was not affected by this mutation, in EMSA, we observed how T‐cell activation induced the shifted band of the CRE2 probe and CT treatment slightly decreased the band intensity. Whether this motif is a negative regulatory element and whether CREB activated by PKA or PKC‐θ may contribute differently in the regulation of IL‐17A expression need to be studied.

We also extended our finding to IL‐17A‐expressing CD8⁺ (Tc17) T cells because of their supportive roles in autoimmune and infectious diseases. Tc17 cells have also been reported to support Th17‐mediated autoimmune encephalomyelitis59 and Tc17 cells are also associated with Th1‐mediated autoimmune diabetes.60 Additionally, Tc17 cells are also important in protection against influenza virus infection.61 In CD4⁺ T‐cell‐depleted mice, Tc17 and other IL‐17A‐secreting cells compensated for the loss of IL‐17A production from CD4⁺ T cells to protect the host against mucosal Candida albicans infection.62 Therefore, our findings provide a potential target that widely mediates IL‐17A expression in human immune cells.

In conclusion, we demonstrated the involvement of cAMP/PKA signalling in CT‐induced IL‐17A production from CD4⁺ T cells; we have also extended these findings to CD8⁺ T cells. Although the promoter‐reporter assays and EMSA analyses showed the direct binding of pCREB to the IL‐17A promoter, additional work is needed to confirm the physiological role of CREB in primary T‐cell systems. Because of the dual roles of cAMP and PKA in immune regulation, the delicate control of cAMP levels is crucial for immune homeostasis. Our findings provide insight into the therapeutic application of cAMP‐elevating drugs to immune‐mediated diseases and the development of Th17‐skewing vaccines.

Disclosures

The authors have no direct financial interest or relationship to the subject matter of this report.

Authorship

H.‐C.T. formulated the hypothesis, designed research, performed all the experiments and data analysis, and wrote this manuscript. S.V. assisted with revising the manuscript. S.L. assisted with Western blotting experiments. R.W. is the principal investigator who is the main contributor to formulating, designing and revising the manuscript.