The cGMP/Protein Kinase G Pathway Contributes to Dihydropyridine-sensitive Calcium Response and Cytokine Production in TH2 Lymphocytes*

Th2 lymphocytes differ from other CD4+ T lymphocytes not only by their effector tasks but also by their T cell receptor (TCR)-dependent signaling pathways. We previously showed that dihydropyridine receptors (DHPR) involved in TCR-induced calcium inflow were selectively expressed in Th2 cells. In this report, we studied whether cGMP-dependent protein kinase G (PKG) activation was implicated in the regulation of DHPR-dependent calcium response and cytokine production in Th2 lymphocytes. The contribution of cGMP in Th2 signaling was supported by the following results: 1) TCR activation elicited cGMP production, which triggered calcium increase responsible for nuclear factor of activated T cell translocation and Il4 gene expression; 2) guanylate cyclase activation by nitric oxide donors increased intracellular cGMP concentration and induced calcium inflow and IL-4 production; 3) reciprocally, guanylate cyclase inhibition reduced calcium response and Th2 cytokine production associated with TCR activation. In addition, DHPR blockade abolished cGMP-induced [Ca2+]i increase, indicating that TCR-induced DHP-sensitive calcium inflow is dependent on cGMP in Th2 cells. Th2 lymphocytes from PKG1-deficient mice displayed impaired calcium signaling and IL-4 production, as did wild-type Th2 cells treated with PKG inhibitors. Altogether, our data indicate that, in Th2 cells, cGMP is produced upon TCR engagement and activates PKG, which controls DHP-sensitive calcium inflow and Th2 cytokine production.

TCR engagement results in phosphorylation events and scaffolding of various adapters and enzymes. This TCR signaling complex, called signalosome, integrates signaling pathways into transcriptional events (12). One of these pathways is orchestrated by the second messenger calcium. Activated phospholipase C ␥1 cleaves phosphatidylinositol (4, 5)-biphosphate into inositol (1,4,5)-triphosphate (IP3) and diacylglycerol (DAG). DAG leads to protein kinase C and Ras-dependent mitogen-activated protein kinase activation (13), whereas IP3 mobilizes intracellular calcium stores. Calcium mobilization is followed by a calcium inflow through calcium release-activated calcium (CRAC) channels (reviewed in Ref. 14) and subsequent activation of the nuclear factor of activated T cells (NFAT). Once dephosphorylated by calcineurin, a calcium-activated phosphatase, NFAT translocates to the nucleus and binds together with other transcription factors to target genes, including cytokine genes (discussed in Ref. 12).
In this study, we wondered whether cGMP was involved in the control of calcium response and cytokine production in Th2 lymphocytes. Our data show that TCR-dependent cGMP production results in PKG activation, which promotes DHP-sensitive calcium inflow and IL-4 synthesis.
To analyze cytokine production, T cells were distributed (2.5 ϫ 10 4 cells/well) into 96-well flat-bottom plates containing either antigenpresenting cells (2.5 ϫ 10 4 cells/well) ϩ OVA peptide in the absence of exogenous cytokines and antibodies or plate-bound anti-TCR mAb (1 g/ml) (28). Cytokine release was quantified by ELISA in supernatants collected 24 h later. cGMP intracellular concentration was determined by using ELISA according to the manufacturer's instructions (cGMP Enzymeimmunoassay Biotrak System; Amersham Biosciences).
Analysis of Intracellular Calcium Concentration-T cells were preincubated for 12 h with or without inhibitors, washed, loaded with 5 M Fura-2 AM (28,29), and stimulated with either soluble anti-TCR mAb (1 g/ml) ϩ anti-CD28 mAb (1 g/ml) or 8-Br-cGMP (200 M) or 8-Br-cAMP (200 M) or SNP (150 M) or spermine NONOate (300 M). [Ca 2ϩ ] i increase was recorded at the single cell level by microspectrofluorometry as previously described (29). Images were recorded every 4 s for at least 400 s. Ionomycin was used as a positive control for each sample. Calcium concentration was calculated according to Ref. 30.
Imaging of NFAT Translocation-NFAT nuclear translocation was evaluated by confocal microscopy as previously described (31,32). Three-round primed Th2 cells were cultured on plate-bound anti-TCR (5 g/ml) ϩ anti-CD28 (3 g/ml) mAbs in the presence or in the absence of PKG inhibitors (KT5823 or Rp-8-pCPTcGMPS). Alternatively, cells were stimulated with either spermine NONOate (1 mM) or 8-BrcGMP (200 M) for 4 h. Cells were then collected and plated onto Lab-Tek chambers (Calbiochem). They were fixed with 4% paraformaldehyde for 10 min and permeabilized with 0.2% Triton X-100 in phosphate-buffered saline for 10 min. Staining was performed with mouse anti-NFATc1 mAb (BD Biosciences) for 45 min. After extensive washing, NFATc1 was revealed using Alexa-488-labeled goat anti-mouse antibody (Molecular Probes, Eugene, OR). Cells were additionally treated with RNaseA (Sigma) for 10 min, and nuclei were stained with propidium iodide. Samples were mounted and examined with a Carl Zeiss LSM 510 confocal microscope using a ϫ63 Plan-Apochromat objective (1.4 oil). We evaluated two parameters: the nuclear versus cytoplasmic NFAT ratio, analyzed by the ImageJ analysis software (NIH), and the percentage of cells that display NFAT nuclear translocation (defined as the frequency of cells in which nuclear/cytoplasmic NFAT ratio was superior to mean ϩ 3 S.D. of the ratio in unstimulated cells).
Flow Cytometric Analysis of IL-4 and IL-5 Intracellular Cytokine Synthesis-Th2 cells were stimulated on plate-bound anti-TCR mAb in the absence or in the presence of either nicardipine or LY-83,583 (10 M). After 12 h of culture, cells were collected, resuspended at 10 6 /ml, and stimulated with phorbol 12-myristate 13-acetate (50 ng/ml, Sigma) ϩ ionomycin (0.5 g/ml, Sigma) for 4 h. Two hours before cell harvest, 10 g/ml of brefeldin A (Sigma) was added. Cells were then fixed with 4% paraformaldehyde (Fluka Chemie, Buchs, Switzerland). Intracytoplasmic staining was performed as previously described (33). Briefly, after washing and 10 min of incubation in saponin medium alone, cells were incubated with phycoerythrin-conjugated rat anti-mouse IL-4 (11B11) or anti-mouse IL-5 (TRFK5) antibodies (BD Biosciences) for 30 min at 4°C. Cells were washed first in saponin buffer and then with phosphate-buffered saline to allow membrane closure. Data were collected on 20,000 CD4 ϩ cells on an XL Coulter cytometer (Coultronics, Margency, France) and analyzed using the CellQuest software (BD Biosciences).
Statistical Analysis-Results are expressed as the mean Ϯ S.D. Overall differences between variables were initially shown by using a Kruskal-Wallis test and subsequently confirmed by the Mann-Whitney U test.

RESULTS
cGMP Induces a Calcium Response, NFAT Translocation to the Nucleus, and IL4 Gene Expression in Th2 Lymphocytes-In B lymphocyte, cGMP was shown to be responsible for DHPR activation and subsequent calcium influx upon B cell receptor stimulation (21). Because we have recently shown that Th2 cells expressed DHPR (18), we investigated the role of cGMP in TCR-induced DHP-sensitive calcium response in these cells. DO11.10 cells were differentiated along the Th2 pathway by three weekly stimulations with antigen-presenting cells loaded with the OVA 323-339 peptide in the presence of IL-4 ϩ antiinterferon ␥ antibody (18). TCR stimulation for 2 (not shown) or 5 min (Fig. 1A) triggered a 2-to 4-fold increase in intracellular cGMP concentration, depending on the experiments. As assessed by microspectrofluorometry on Fura-2AM-loaded Th2 lymphocytes, cell-permeant 8-Br-cGMP promoted a sustained increase in [Ca 2ϩ ] i (Fig. 1B). Abolition of the cGMP-induced [Ca 2ϩ ] i increase in the absence of extracellular calcium (not shown) or in the presence of the DHPR antagonist (nicardipine) pointed out that cGMP promoted a DHP-sensitive calcium inflow (Fig. 1B).
8-Br-cAMP, another cyclic nucleotide known to promote Th2 cytokine production, had no effect on Th2 cell calcium response (Fig. 1B).
We then wondered whether the 8-BrcGMP-induced [Ca 2ϩ ] i increase was sufficient to induce NFAT translocation to the nucleus. Th2 cells incubated for 4 h in the presence of 8-Br-cGMP exhibited a significant NFAT translocation as shown by confocal microscopy (Fig. 1, C and D). Depending on the experiments, NFAT was found to be translocated to the nucleus in 2-11% of resting Th2 lymphocytes, 30 -50% of cGMPtreated cells, and 41-75% of anti-CD3 ϩ anti-CD28-stimulated lymphocytes (not shown). Because Il4 gene expression is regulated by the calcium-dependent calcineurin pathway (9), we assessed whether cGMP triggered Il4 gene expression. 8-Br-cGMP induced a strong IL-4 mRNA increase. The inhibitory effect of cyclosporin A testified to the involvement of calcium-calcineurin pathway (Fig. 1E). Contrasting with TCR-stimulated Th2 cells, IL-4 was not detected at the protein level in supernatants collected 24 h after cGMP stimulation (not shown). The lower calcium response (Fig. 1, C and D) and Il4 gene expression (Fig.  1E) induced by cGMP could account for this difference.
An Inhibitor of Guanylate Cyclases, LY-83,583 Strongly Reduces Calcium Response and Type-2 Cytokine Production upon TCR Stimulation-[Ca 2ϩ ] i increase induced by stimulation of Th2 lymphocytes with a combination of both anti-TCR and anti-CD28 mAbs was abolished by LY-83,583 ( Fig. 2A). The calcium response was also inhibited either in the absence of extracellular calcium or in the presence of a DHPR antag-   MAY 5, 2006 • VOLUME 281 • NUMBER 18 onist (nicardipine) (Fig. 2A). Data were confirmed with another DHPR antagonist, R(ϩ) BayK 8644 (not shown). Treatment of TCR-activated Th2 cells with LY-83,583 as well as nicardipine resulted in dose-dependent inhibition of IL-4 (Fig. 2B) and IL-5 (Fig. 2C) release. Intracellular IL-4 and IL-5 staining revealed that LY-83,583 and nicardipine dramatically reduced the number of IL-4-and IL-5-producing cells (Fig. 2D).

NO Donors Generate cGMP, Increase [Ca 2ϩ ] i , Induce NFAT Nuclear Translocation and IL-4 Production-
The role of guanylate cyclases was further examined by using NO donors that activate soluble guanylate cyclases (34). SNP induced a 2-fold increase in intracellular cGMP concentration, in the same range as TCR stimulation did (Fig. 3A). NONOate, a more potent NO donor than SNP, induced a 10-fold increase in intracellular cGMP concentration when compared with unstimulated cells (Fig. 3A). Zaprinast, an inhibitor of phosphodiesterase 5 (known to specifically promote cGMP degradation), did not enhance cGMP production, indicating that cGMP degradation did not minimize cGMP production (Fig. 3A). Both NO donors, SNP and NONOate, triggered a sustained [Ca 2ϩ ] i increase (Fig. 3B). However, the calcium response induced by NONOate was more abrupt and higher than the one induced by SNP. The guanylate cyclase inhibitor LY-83,583 delayed and reduced SNP-induced calcium response (Fig. 3B). The SNP-induced [Ca 2ϩ ] i increase was inhibited either in the absence of extracellular cal-cium or in the presence of DHPR antagonist nicardipine (Fig. 3B). SNP induced Il4 gene expression, which was abolished in the presence of cyclosporin A (Fig. 3C). We failed to detect any SNP-induced IL-4 protein secretion (not shown). According to its high potency to enhance cGMP production, NONOate induced NFAT nuclear translocation (Fig. 3D) in 60 -70% of cells (not shown) and detectable IL-4 production (Fig. 3E). By contrast, the dose-dependent increase in IL-4 production triggered by NONOate was not observed in Th2 cells from PKG1 Ϫ/Ϫ mice, suggesting a prominent role of PKG in mediating NONOate effect (Fig. 3E). Altogether, these data showed that guanylate cyclase activation resulted in cGMP production and a subsequent DHPsensitive extracellular calcium inflow, responsible for IL-4 expression.
PKG, but Not PDE, Inhibitors Suppress Ca 2ϩ Signaling and Th2 Cytokine Production-To assess whether PKG or PDE were important for calcium signaling in differentiated Th2 cells, inhibitors of PKG and PDE were tested on fully differentiated DO11.10 Th2 lymphocytes. cAMP concentration, which favors Th2 cytokine production, is closely regulated by PDE3 and PDE4 (35,36); therefore, we tested increasing concentrations of PDE4 and PDE3 inhibitors. PDE4 (Ro-20-1724, 10 M) and PDE3 (trequinsin, 10 nM) inhibitors, used at the highest non-toxic doses, did not prevent the [Ca 2ϩ ] i increase (Fig. 5A) or IL-4 production induced by TCR stimulation (Fig. 5D). By contrast, TCR-induced calcium response was robustly diminished by PKG inhibitors (Fig. 5B). In agreement with our previous results showing that cAMP did not mediate a calcium influx (Fig. 1B), the neutralization of cAMP-dependent protein kinase activity did not alter TCR-induced calcium increase (Fig.  5B), indicating that cAMP-dependent protein kinase was not involved in the control of calcium response.

DISCUSSION
Our results strongly support a pivotal role for cGMP and PKG in calcium signaling and IL-4 synthesis in Th2 lymphocytes. Contribution of cGMP in Th2 signaling was underlined by the following data: 1) cell-permeant cGMP, or nitric oxide donors known to activate soluble guanylate cyclase, triggered a calcium response and IL-4 expression; 2) addition of guanylate cyclase inhibitor LY-83,583 suppressed both [Ca 2ϩ ] i increase and Th2 cytokine production induced by TCR activation. Moreover, calcium signaling and IL-4 production were defective in Th2 lymphocytes from PKG null mice, supporting a major role for PKG in Th2 cell activation.
We showed that cGMP concentration was increased in stimulated Th2 cells in a similar range as the rise reported in B-cell receptoractivated B lymphocytes (21). Incubation of Th2 lymphocytes with cGMP or SNP induced calcium response and IL-4 mRNA transcripts. This up-regulation of IL-4 expression was due to calcium signaling because it was inhibited by cyclosporin A. However, cGMP or SNP could not induce detectable IL-4 protein. This could be explained by the fact that IL-4 production requires a sustained calcium response in time and amplitude, which may not be induced by cGMP and SNP. Indeed, NFAT nuclear translocation, which reflects the global calcium response, was weaker in cells stimulated by cGMP as compared with TCR-activated Th2 cells. Stimulation of Th2 cells with NONOate, a potent NO donor, induced a stronger calcium response and NFAT translocation than SNP, resulting in detectable IL-4 production. High levels of IL-4 synthesis, as observed in TCRstimulated Th2 cells, involve other calcium-independent pathways that may not be activated by cGMP or SNP. Therefore, the cGMP pathway is likely necessary but not sufficient to induce optimal Th2 cytokine production.
Some effects of NO result from guanylate cyclase and PKG activation, raising the question of the involvement of NO in Th2 cell activation. NO is produced by many cell types, including dendritic cells, whereas its production in T lymphocytes is unlikely (37). In this report, we showed that cGMP was produced in purified Th2 cells after TCR activation in the absence of antigen-presenting cells, excluding an effect of NO pro-   MAY 5, 2006 • VOLUME 281 • NUMBER 18 duced by antigen-presenting cells. Thus, guanylate cyclase activation may be NO independent in T lymphocytes. As suggested by others (38), protein kinase C might directly activate guanylate cyclase upon TCR stimulation. Indeed, we have previously shown the involvement of protein kinase C in DHP-sensitive calcium inflow and IL-4 production in Th2 lymphocytes (28).

cGMP-dependent Signaling in Th2 Cells
cGMP may activate cGMP-dependent protein kinase and modulate non-selective cation cGMP-gated channels or cyclic nucleotide phosphodiesterases (discussed in Ref. 39). PDE family members encompass 11 isoenzymes with distinct patterns of substrate selectivity (40,41). PDE3 is defined as cGMP inhibited and cAMP selective, PDE4 as cAMP specific, and PDE5 as cGMP specific (reviewed in Ref. 41). These enzymes are co-expressed in T lymphocytes and especially in Th2 lymphocytes (42). Thus, cGMP might have decreased PDE3 activity, resulting in an indirect increase in cAMP concentration that could regulate DHP-sensitive calcium influx through a cAMP-dependent protein kinase-dependent mechanism (discussed in Ref. 43). This hypothesis is unlikely because we showed that neither PDE3 nor PDE4 inhibitors, expected to increase cAMP levels, modified TCR-dependent calcium increase and IL-4 synthesis. These data are in agreement with previous work showing that PDE inhibitors had little effect on IL-4 production by a Th2 clone (42). Furthermore, addition of cAMP, as well as inhibition of cAMP-dependent protein kinase, did not affect calcium response in Th2 lymphocytes.
Contrary to PDE, our data emphasize a major role for PKG in cGMPdependent signaling in Th2 lymphocytes. Indeed, TCR-induced [Ca 2ϩ ] i increase, nuclear NFAT translocation, and IL-4 synthesis were strongly reduced in Th2 cells from PKG1 Ϫ/Ϫ mice. Likewise, PKG inhibitors inhibited both [Ca 2ϩ ] i increase and cytokine production by wild-type Th2 lymphocytes. PKG could also activate a calcium-independent signaling pathway controlling IL-4 production. By analogy to cardiac myocytes (44), it could be speculated that PKG might decrease nuclear exportation of NFAT through c-Jun NH 2 -terminal kinase 2 (JNK2) inhibition, resulting in an enhanced IL-4 transcription in Th2 cells.
Several studies support that DHPR are involved in calcium responses of non-excitable cells, including B lymphocytes (21), NK cells (19), dendritic cells (20), naive T cells (24), Th2 cells (18,28), and epithelial cells (45,46). It has been shown that signaling through the B cell receptor was coupled to a guanylate cyclase/cGMP-dependent pathway that controlled DHPR-dependent calcium entry (21). Here we have shown that in Th2 lymphocytes DHPR antagonization by nicardipine reduced cGMP-mediated calcium response, suggesting a role for DHPR in cGMP signaling. Although the structure of DHPR has not been completely elucidated in lymphocytes, DHPR had been related to the ␣1 calcium channel subunit in B cells (21) as well as in Th2 cells (18) and human T cells (24). In excitable cells, the NO/cGMP/PKG pathway is known to activate or repress DHPR-dependent calcium currents depending upon the cell type (47,48). How this pathway operates in Th2 lymphocytes will require further investigation. Our study, therefore, establishes the following sequence of events that lead to cytokine production in Th2 lymphocytes (Fig. 7). TCR stimulation results in guanylate cyclase activation, responsible for cGMP production. cGMP then activates PKG, which may, directly or not, control DHP-sensitive calcium influx. Intracellular calcium increase induces NFAT translocation to the nucleus and subsequent gene expression. Altogether, our data lend support for a new signaling pathway in Th2 lymphocytes in which cGMP acts as a major second messenger in Th2 calcium response and cytokine production.