Activation of cGMP-dependent protein kinase Ibeta inhibits interleukin 2 release and proliferation of T cell receptor-stimulated human peripheral T cells.

Several major functions of type I cGMP-dependent protein kinase (cGK I) have been established in smooth muscle cells, platelets, endothelial cells, and cardiac myocytes. Here we demonstrate that cGK Ibeta is endogenously expressed in freshly purified human peripheral blood T lymphocytes and inhibits their proliferation and interleukin 2 release. Incubation of human T cells with the NO donor, sodium nitroprusside, or the membrane-permeant cGMP analogs PET-cGMP and 8-pCPT-cGMP, activated cGK I and produced (i) a distinct pattern of phosphorylation of vasodilator-stimulated phosphoprotein, (ii) stimulation of the mitogen-activated protein kinases ERK1/2 and p38 kinase, and, upon anti-CD3 stimulation, (iii) inhibition of interleukin 2 release and (iv) inhibition of cell proliferation. cGK I was lost during in vitro culturing of primary T cells and was not detectable in transformed T cell lines. The proliferation of these cGK I-deficient cells was not inhibited by even high cGMP concentrations indicating that cGK I, but not cGMP-regulated phosphodiesterases or channels, cAMP-dependent protein kinase, or other potential cGMP mediators, was responsible for inhibition of T cell proliferation. Consistent with this, overexpression of cGK Ibeta, but not an inactive cGK Ibeta mutant, restored cGMP-dependent inhibition of cell proliferation of Jurkat cells. Thus, the NO/cGMP/cGK signaling system is a negative regulator of T cell activation and proliferation and of potential significance for counteracting inflammatory or lymphoproliferative processes.

complex with foreign antigen presented on major histocompatibility complex molecules and (ii) ligation of the CD28 T cell antigen by B7 molecules (CD80 and CD86) on the surface of antigen-presenting cells, resulting in transcription factor activation, interleukin 2 (IL-2) production, and cell proliferation (1)(2)(3)(4). Activation and subsequent proliferation of resting T lymphocytes is an essential process in the T cell-mediated immune response. In vitro, T cells respond to stimulation of the TCR⅐CD3 complex (without costimulation) with IL-2-dependent induction of cell proliferation (2,5). A costimulatory signal from the CD28 pathway synergizes with the signal from the TCR⅐CD3 complex to further increase cell proliferation and IL-2 production, and T cell proliferation becomes largely independent of IL-2 (6). Intracellular signaling pathways, which may mediate T cell activation, include phosphorylation cascades that stimulate MAP kinases including ERK (3,7,8) and p38 kinase (9,10), as well as phosphatidylinositol 3-kinase (11,12) and its target Akt/ protein kinase B (13).
Reports of cyclic nucleotide effects on T cell activation indicate that cAMP is inhibitory, whereas effects of cGMP have not been conclusively elucidated. In fact, lowering of cAMP levels by induction of phosphodiesterase-7 was reported to be required for T cell activation (14). Nitric oxide, which activates soluble guanylate cyclase and increases intracellular cGMP, has inhibitory effects on T cell proliferation (15) and cytokine secretion (16), stimulates c-Jun Nterminal kinase, ERK, and p38 activities in Jurkat T cells (17), and decreases tyrosine phosphorylation of Jak3/STAT5 and proliferation in cultured T cells (18). However, NO can elicit effects via both cGMP-dependent and -independent pathways, and increasing intracellular cGMP may affect not only cGMP-dependent protein kinase but also cGMP-regulated phosphodiesterases or channels and cAMP-dependent protein kinase (19,20).
The present study focuses on the role of cGK in mediating cGMP effects on human peripheral blood T cells. In general, there exist two mammalian isoforms of cGMP-dependent protein kinase, cGK I and cGK II (19 -24), as well as two splice variants of cGK I (cGK I␣ and I␤) (21,22) with differing N-terminal ends. cGK I is expressed in platelets, vascular smooth muscle cells, fibroblasts, certain endothelial cells, lung, cerebellum, and heart, whereas cGK II is expressed in intestinal epithelium and many regions of brain and kidney (19,20,(23)(24)(25)(26). In human T cells, effects of the NO donor sodium nitroprusside (SNP) were compared with those of membrane-permeant, selective activators of cGK such as PET-cGMP and 8-pCPT-cGMP (27). Our results demonstrate that primary human T cells contain endogenous cGK I␤, which mediates inhibitory effects of cGMP on IL-2 production and proliferation. In contrast to the bulk of published studies, only freshly prepared cells were used for examining cGK I effects on T cells because our studies demonstrated that cGMP effects were absent in cGK I-deficient cultured cells; however, cGMP effects could be restored by transfection of these cells with cGK I.
Jurkat-E6 cells were obtained from American Tissue Culture Collection and A3.01 human T lymphoma cells from the Center for Disease Control (Atlanta, GA), fetal calf serum and cell culture media were from Life Technologies, Inc., nitrocellulose membranes were from Schleicher & Schuell, and polyvinylene difluoride membranes from Millipore (Eschborn, Germany). Standard chemicals were obtained from Sigma.
Isolation of Human Peripheral Blood Mononuclear Cells, Purified T Lymphocytes, and Platelets-PBMC were isolated from freshly drawn heparinized human blood using Ficoll gradient centrifugation (31) and then resuspended in RPMI 1640 medium. T cells were purified from PBMC using anti-CD3-conjugated Dynabeads, or CD4 ϩ and CD8 ϩ T cell subsets were obtained by incubating PBMC with anti-CD4-or anti-CD8-conjugated Dynabeads at 4°C for 1 h and then following the supplier's protocol. Affinity purification of T cells was performed using either fluorochrome-conjugated antibodies against T cell antigens together with fluorescence-activated cell sorting (see below) or using T cell immunoaffinity columns and a protocol provided by Cedar Lane (Hornby, Canada) for negative-selection of T cells (removal of B cells and monocytes). Purified human platelets were prepared as described elsewhere (32).
For experiments, PBMC or purified T cells were incubated at 37°C in RPMI 1640 medium containing 5% fetal calf serum, 100 units/ml penicillin, 100 g/ml streptomycin, or in the case of long term culturing (less than 2 weeks), in the same buffer except with 10% fetal calf serum plus 100 units/ml IL-2, and an initial addition of 100 ng/ml anti-CD3 antibody.
Fluorescence-activated Cell Sorting of Human T Lymphocytes-PBMC were resuspended in phosphate-buffered saline containing 0.1% bovine serum albumin and then incubated with fluorochrome-conjugated anti-CD3 and anti-CD41a antibodies at 4°C for 1 h. After washing to remove excess antibodies, cells were sorted to obtain pure T cells devoid of platelets (CD3 ϩ /CD41a Ϫ ) separate from a T cell fraction with adherent platelets (CD3 ϩ /CD41a ϩ ). T cell purity was determined by distinguishing T cells from other cells, using fluorochrome-conjugated antibodies directed against cell-specific antigens such as CD2, CD4, CD8 (T cells), CD19 (B cells), CD14 (monocytes), and CD41a (platelets), Briefly, affinity-purified cells were incubated with a combination of up to three differently labeled antibodies at 4°C for 1 h and then washed with phosphate-buffered saline containing 0.1% bovine serum albumin, and analyzed using a FACScan cell sorter (Becton-Dickenson, Hamburg, Germany) together with LysisII and Cellquest software. Purified T cells contained at least 95% T lymphocytes, less than 2% monocytes or B cells, and either no detectable platelets (FACS-sorted cells) or less than 5% platelets (immunoaffinity column-purified cells). For FACS calibration purposes, fluorescein isothiocyanate-and phycoerythrinlabeled isotype-matched control antibodies were used.

Analysis of cGK Activation by Detection of VASP Phosphorylation in Human PBMC, Purified T Cells, and Platelets Either in Lysates or
Intact Cells-In vitro cGK I activity was determined by analyzing VASP phosphorylation in cell lysates of affinity-purified T lymphocytes as well as platelets, following a previously published protocol (33). Briefly, purified T cells were lysed in 10 mM Hepes, pH 7.4, 5 mM MgCl 2 , 0.2 mM EDTA, and a protease inhibitor mix (Complete TM , Roche Molecular Biochemicals), and then lysates were passed through 25 gauge needles to shear chromosomal DNA. In vitro kinase reactions with cell lysates were carried out in buffer (20 mM Tris-Cl, pH 7.4, 10 mM MgCl 2 , 5 mM ␤-mercaptoethanol, 0.1% bovine serum albumin, and 50 M ATP), in the absence or presence of 0.5 M PET-cGMP and 0.1 g of purified recombinant dephospho-VASP, at 30°C for 30 min. Subsequently, samples were processed for SDS-PAGE and Western blot analysis of VASP phosphorylation as described below.
In intact cell experiments, PBMC (preincubated at 37°C for 3 h to overcome potential preactivation arising from handling during preparation) or platelets were resuspended in serum-free RPMI medium and stimulated in the absence or presence of agents that elevate cGMP or cAMP levels (see details under "Results" and in the figure legends) and briefly centrifuged to obtain cell pellets which were processed for SDS-PAGE and Western blot analysis of VASP phosphorylation as described below.
Analysis of cGMP-dependent Activation of the MAP Kinases ERK 1/2 and p38, in Human PBMC and Immunoaffinity Column-Purified T Lymphocytes-PBMC or purified T lymphocytes were incubated in serum-free RPMI medium for 3 h at 37°C and then stimulated in the absence or presence of agents which elevate cGMP levels and briefly centrifuged to obtain cell pellets that were processed for SDS-PAGE and Western blot analysis of phosphorylation as described below.
Western Blot Analysis of Protein Kinases (cGK I, ERK1/2, and p38) and VASP Phosphorylation-Intact cells or cell lysates were added to SDS gel loading buffer and heated at 95°C for 10 min, then analyzed by SDS-PAGE (8% gels, for cGK I and VASP) and Western blotting. Western blot nitrocellulose membranes were blocked with 1% hemoglobin in phosphate-buffered saline and divided horizontally to stain the high molecular weight range of proteins with polyclonal anti-cGK I antiserum (diluted 1:3000) and the lower range with monoclonal antiphospho(Ser 239 )-VASP 16C2 antibodies (1 g/ml), followed by either horseradish peroxidase-coupled goat anti-rabbit or goat anti-mouse secondary antibodies (1:5000; Bio-Rad), respectively. Signals were visualized using the ECL system (Amersham Pharmacia Biotech).
For determining ERK and p38 activation, stimulated cells were analyzed by SDS-PAGE (using 10% gels) and blotted to polyvinylene difluoride membranes, which were then blocked with TBST (10 mM Tris-Cl, pH 7.5, 100 mM NaCl, 0.1% Tween 20) containing 3% nonfat dry milk (Bio-Rad). Signals were visualized with polyclonal antibodies that recognize activated dual phosphorylated ERK1/2 or dual phosphorylated p38, followed by horseradish peroxidase-coupled goat anti-rabbit IgG (Bio-Rad). To determine total expression of MAP kinase for verification of equal protein loading, membranes were stripped at 58°C for 1 h in buffer containing 62.5 mM Tris-Cl, pH 6.8, 100 mM ␤-mercapto-ethanol, and 2% SDS and then blocked in 5% nonfat dry milk (Bio-Rad) and incubated with anti-ERK1/2 or anti-p38 phosphorylation-independent antibodies, followed by ECL detection (data not shown).
Determination of T Cell Proliferation-PBMC (2 ϫ 10 5 /well of 24-well plates) were incubated with or without agents that elevate cAMP or cGMP levels, at 37°C for 1 h under serum-free conditions prior to adding anti-CD3 antibody (12F6 or OKT3, diluted 1:2500), alone or together with anti-CD28 antibody (0.1 g/ml), and incubating in 10% serum-containing medium for analysis of proliferation after 5 days. Subsequently, PBMC were counted using a Cell Counter (Coulter, Krefeld, Germany), and the inhibition of anti-CD3 antibody-induced cell proliferation caused by cAMP or cGMP analogs or SNP was determined.
Proliferation of cGK I␤-expressing Jurkat Cells-Jurkat E6.1 cells (5 ϫ 10 5 /ml) were cotransfected with pCMV-cGK I␤ (34) or with pCMV-cGK I␤-K405A mutant (35), and pCMV-EGFP (kindly provided by S. Schneider-Rasp) at a ratio of 3:2; using Fugene-6 reagent (Roche Molecular Biochemicals). After 16 h of transfection, cells were seeded at a density of 200,000 cells/well and subsequently incubated for 72 h in the presence of various concentrations of 8-pCPT-cGMP. Using flow cytometry, the impact of cGK I expression and activation with 8-pCPT-cGMP on Jurkat cell proliferation was determined. In the presence of 8-pCPT-cGMP, the ratio of green fluorescent cells (e.g. containing cGK I and EGFP) to nonfluorescent cells was calculated and divided by the same ratio for cells exposed to medium only (non-8-pCPT-cGMP treated cells). This gave a Jurkat proliferation ratio which was converted to percent.

Highly Purified Primary Human T Lymphocytes, but Neither Cultured T Cells nor T Cell Lines, Display Endogenous cGK I Expression and
Activity-A rich source of cGK I in human blood is platelets that can easily contaminate crude lymphocyte preparations. Therefore, for examining the expression of cGK in T lymphocytes, it was crucial to use highly purified cells (Fig.  1A). T cells were prepared from freshly isolated PBMC using anti-CD3-conjugated Dynabeads or were isolated from PBMC by labeling with fluorochrome-conjugated anti-CD3 and anti-CD41a (platelet-specific marker) and using FACS to separate pure T cells devoid of platelets (CD3 ϩ /CD41a Ϫ ) from T cells with adherent platelets (CD3 ϩ /CD41a ϩ ). Western blot analysis using polyclonal anti-cGK I antibody (Fig. 1A) detected cGK I in unsorted as well as both of the FACS-sorted T cell fractions, indicating that T cells devoid of platelets (CD41a Ϫ ) indeed contain endogenous cGK I. To estimate the abundance of cGK I expression in T cells, signal intensities from FACS-sorted T cells were compared with those observed in increasing amounts of platelet extracts with a known content of cGK I (38.1 Ϯ 3.9 pmol cGK I/10 9 platelets, equivalent to 3 ng of cGK I protein/ 10 6 cells (32). T cell-specific cGK I expression was estimated at ϳ5 ng/10 6 T cells (Fig. 1A).
Anti-CD8 ϩ -and anti-CD4 ϩ -conjugated Dynabeads used to purify T cell subsets (with less than 5% contaminating platelets; data not shown) demonstrated endogenous cGK I expression in both CD8 ϩ and CD4 ϩ T cells (Fig. 1B). However, transformed T cell lines (Jurkat-E6 or A3.01) contained no detectable cGK I, and that observed in primary T cell cultures was lost after 48 h of culture (Fig. 1, B and C, respectively). None of the T cells investigated (freshly isolated, cultured, or transformed) expressed detectable cGK II by either Western or PCR analysis (data not shown). The cGK I isoform expressed in T cells was identified by reverse transcriptase-PCR amplification of RNA derived from immunoaffinity column-purified T cells using cGK I␣ and cGK I␤ specific primer pairs (see "Ex-perimental Procedures"). As shown in Fig. 1D, only the cGK I␤ splice variant was detected in human T cells.
Evidence that the band identified by Western blot analysis as cGK I in fresh T cells indeed demonstrated that cGK activity was obtained by studying cGK I phosphorylation of the endog-  (32). T cells (unsorted) were purified from PBMC using anti-CD3conjugated Dynabeads. Pure T cells devoid of platelets (CD3 ϩ /CD41 Ϫ ), as well as a T cell fraction with adherent platelets (CD3 ϩ /CD41 ϩ ) were separated using anti-CD3 and anti-CD41a antibodies and FACS sorting. T cells (10 6 cells/lane) and purified platelets were added to SDS gel loading buffer and examined for cGK I expression by Western blot analysis (see "Experimental Procedures"). The data shown are representative of three experiments. B, endogenous expression of cGK I in CD8 ϩ and CD4 ϩ primary T cells but not in transformed T cell lines Jurkat-E6 and A3.01 (10 6 cells/lane). T cell subtypes were purified using anti-CD4 and anti-CD8-conjugated Dynabeads, and separation from monocytes, B cells, and platelets was verified by FACS analysis (not shown). cGK I antibody specificity was demonstrated by preabsorbing (Preabs.) the antibody (Ab) with immobilized recombinant cGK I protein prior to staining of a parallel Western blot (lower panel in B). The data shown are representative of four independent experiments. C, loss of endogenous cGK I expression upon cultivation of purified T cells in vitro. T cells, purified as described for A above, were either freshly prepared (Fresh) or cultured in vitro for 48 h (Cultured) for examination (10 6 cells/lane) of endogenous cGK I expression. Shown is an example of results obtained in three independent experiments. A platelet-derived extract (Std) was used as a molecular weight marker for identifying cGK I in B and C. D, reverse transcriptase-PCR demonstration of cGK I␤ but not cGK I␣ mRNA in immunoaffinity column-purified T cells. Reverse transcribed T cell RNA as well as positive control cDNA standards (cGK I␣ and cGK I␤) were PCR-amplified using cGK I␣ and cGK I␤ specific primer pairs (see "Experimental Procedures"). A control (ctrl) PCR reaction without template gave no product.

cGMP-dependent Protein Kinase in T Cells
enous substrate VASP in lysates of T cells (affinity-purified using anti-CD3-conjugated Dynabeads) in response to 0.5 M PET-cGMP, a hydrolysis resistant, cGK-selective cGMP analog (Fig. 2). VASP phosphorylation was determined by Western blot analysis using the anti-phospho(Ser 239 )-VASP antibody 16C2, which detects a 46-kDa VASP phosphorylated on Ser 239 , as well as a 50-kDa VASP phosphorylated on both Ser 239 and Ser 157 (30). The signal for VASP phosphorylation obtained from fresh T cell extracts ( Fig. 2A, fourth lane) was less prominent in comparison with that from an equal amount of platelet extract (Fig. 2C, fourth lane). This appeared to be in part due to lower VASP levels in T cells (determined using a phosphorylationindependent antibody which recognizes total VASP; data not shown), and the signal in fresh T cells could be increased by adding exogenous purified recombinant VASP ( Fig. 2A, compare fourth lane with sixth lane), which is isolated primarily in the de-phospho form ( Fig. 2A, second and fifth lanes). The exogenously added dephospho-VASP ensured saturating amounts of substrate and increased the sensitivity of the in vitro kinase assay in T cells ( Fig. 2A, sixth lane). In contrast to freshly purified T cells, T cells from the same donor cultured for 10 days in vitro no longer demonstrated PET-cGMP-stimulated VASP phosphorylation activity (Fig. 2B).

Distinct Patterns of cGK I Activation in Freshly Isolated Human PBMC versus Platelets, and Loss of cGMP-dependent but Not cAMP-dependent VASP Phosphorylation in Cultured
PBMC-The stimulation of VASP phosphorylation by cGMPelevating agents was also demonstrated in intact PBMC (Western blots; Fig. 3). After 25 min of stimulation, VASP phosphorylation in PBMC in response to 200 M PET-cGMP, a membrane-permeant, hydrolysis-resistant, cGK-selective cGMP analog, and in response to 100 M SNP, a NO donor, was markedly less than VASP phosphorylation in platelets (Fig. 3,  A and B, respectively, showing analysis of equal amounts of protein), as had also been observed in cell lysates above. Examination of a detailed time course of SNP stimulation revealed that peak VASP phosphorylation occurred transiently after 5 min SNP and then rapidly declined (Fig. 3B), unlike the SNP stimulation in platelets that persisted throughout 2 h. YC-1 (50 M), a nitric oxide-independent activator of guanylate cyclase and an inhibitor of several phosphodiesterases, caused strong and long lasting stimulation of VASP phosphorylation in PBMC (Fig. 3). SNP plus YC-1 treatment produced somewhat more VASP phosphorylation than YC-1 alone. cBIMPS, a selective activator of cAK, also strongly stimulated VASP phosphorylation in PBMC and platelets. Whereas the cBIMPS stimulatory effect was preserved in cultured PBMC, none of the effects of cGMP-elevating agents were preserved (Fig. 3A), indicating that their effects were lost along with cGK I and could not be mediated by cAK.  5, A and B) shown to contain cGK I (Figs. 1-3). In cultured PBMC (cultured for 3 days prior to the proliferation assay) that have lost cGK I, the inhibitory effect of cBIMPS was highly preserved; however, no inhibition by 8-pCPT-cGMP was observed, and inhibition by SNP was reduced from 70% in fresh PBMC to 50% (Fig. 5, B and C). Jurkat-E6 cells (Fig.  5D) containing no detectable cGK I (Fig. 1B) also did not respond to 8-pCPT-cGMP, but Jurkat proliferation was inhibited by cBIMPS (60%) and by SNP (40%). Not only the proliferation of T cells stimulated with anti-CD3 antibody alone (Fig. 5B) but also that of T cells costimulated with anti-CD3 plus anti-CD28 (Fig. 5E) was inhibited by 8-pCPT-cGMP (40 and 30% inhibition, respectively).

cGMP/cGK Inhibition of T Cell Proliferation Partially Results from Reduced IL-2 Expression-IL-2 is an important regulator of T cell proliferation and is released by activated T cells.
Anti-CD3-stimulated IL-2 release was reduced by 8-pCPT-cGMP to 42% of control (from 64 to 37 pg of IL-2/10 7 cells; Fig.  6). To test whether the observed cGMP-dependent inhibition of T cell proliferation was due to reduction of IL-2 levels, recombinant human IL-2 was added to anti-CD3-stimulated PBMC cultures in the presence or absence of 8-pCPT-cGMP. IL-2 (500 units/ml) effectively diminished 8-pCPT-cGMP inhibition of anti-CD3-stimulated proliferation (Fig. 7). Costimulation with anti-CD3 plus anti-CD28 elicits IL-2 independent proliferation, which is also inhibited by 8-pCPT-cGMP (Fig. 5E), but this was, as expected, not reversed by addition of exogenous IL-2 (data not shown).
Reconstitution of cGK Expression in Jurkat-E6 Cells Restores cGK-mediated Inhibition of Cell Proliferation-Jurkat-E6 cells were found to be deficient in cGK I (Fig. 1B), and their proliferation was resistant to inhibition by 8-pCPT-cGMP (Fig. 5D).
To provide clear evidence that cGMP-dependent inhibition of T cell proliferation was indeed mediated by cGK, cytomegalovirus-promoter-driven cDNA constructs encoding either cGK I␤ or an inactive cGK I␤-mutant (cGK I␤ K405A) were transfected into Jurkat cells, and, subsequently, the number of cGK expressing cells were determined in the presence of increasing concentrations of 8-pCPT-cGMP. As observed previously for cGK I-deficient Jurkat cells in Fig. 5D, the proliferation of cGK I␤-K405A-transfected, 8-pCPT-cGMP treated Jurkat cells was not inhibited relative to that of similarly transfected, non-8-pCPT-cGMP-treated cells (Fig. 8). In contrast, the proliferation of cGK I␤ (wild type)-transfected, 8-pCPT-cGMP treated Jurkat cells was inhibited relative to that of the corresponding non-8-pCPT-cGMP-treated cells, indicating that cGMPdependent inhibition of T cell proliferation can indeed be reconstituted by cGK I␤ transfection. DISCUSSION Human T cells devoid of platelets were shown to express endogenous cGK I, which could be stimulated by NO and cGMP analogs to inhibit T cell proliferation. Both CD4 ϩ and CD8 ϩ T cells expressed endogenous cGK I␤; however, neither cell lines such as Jurkat and A3.01 nor cultured T cells retained cGK I. Thus, NO or cGMP effects reported in such cells are unlikely to be mediated by cGK I. In our experiments employing freshly isolated T cells, effects of NO and cGMP analogs were clearly mediated by cGK I and not cAK because these agents were unable to phosphorylate VASP in cultured PBMC that had lost cGK I, but contained cAK that did phosphorylate VASP in response to the cAMP analog cBIMPS. The pattern of VASP phosphorylation by cGK I in response to SNP was particularly different in PBMC compared with platelets. Whereas SNPstimulated VASP phosphorylation persisted for over 2 h in platelets, it lasted only about 5 min in lymphocytes, suggesting the presence of a highly efficient cGMP phosphodiesterase in lymphocytes. Experiments using other types of agonists gave results in agreement with this. PET-cGMP (a hydrolysis-resistant analog), YC-1 (an activator of soluble guanylate cyclase but also an inhibitor of several phosphodiesterases including ones that hydrolyze cGMP (36)), as well as SNP plus YC-1, had much longer lasting effects than SNP alone.
In our experiments, stimulation of cGK I activated ERK and p38 MAP kinases, and in contrast to cAMP (37), cGMP did not inhibit anti-CD3-induced MAP kinase activation (data not shown). NO, cGMP, or cGK I stimulation or inhibition of MAP

cGMP-dependent Protein Kinase in T Cells
kinases has been demonstrated in other cell types including cGK I-transfected baby hamster kidney cells (38), 293T fibroblasts (39), CHO cells (40), Jurkat T cells (17), macrophages (41), mesangial cells (42), and also low passage rat aorta vascular smooth muscle cells (43). Responses to NO, cGMP, or cGK not only varied in different cell types with respect to activation or inhibition of MAP kinases but also proliferation. In many cases, a definitive relationship between cGK and effects on MAP kinases and proliferation is not possible because certain cells may not contain cGK (e.g. Jurkat cells), and this was not always investigated. Also some criteria that have been used to define cGK effects in cells, such as inhibition by agents such as KT5823, may be unreliable because this agent has been shown to actually enhance cGMP effects in certain cell types such as platelets and mesangial cells (44). Our own results in T cells indicate that short term cGK I activation of MAP kinases does not activate T cells, instead cGK I inhibits proliferation of activated T cells. This is in distinct contrast to cAMP-dependent inhibition of T cell proliferation reported to be related to MAP kinase inhibition (37) and thus represents a major difference in cAMP and cGMP effects on T cells.
Selective activation of cGK I by cGMP analogs, as well as by SNP, inhibited proliferation of human T cells, although these were only 40 -70% as effective as the cAMP analog cBIMPS, which caused complete inhibition. In Jurkat T cells and PBMC cultured for 3 days prior to the proliferation assay, i.e. cells that contain no detectable cGK I, the inhibitory effect of 8-pCPT-cGMP was lost, whereas that of cBIMPS was unchanged and that of SNP was only somewhat reduced. These results indicated that cGMP effects are not mediated by cAK and that SNP has additional inhibitory effects on T cell proliferation that are clearly cGK-independent. cGMP inhibition of cell proliferation could be reconstituted in Jurkat cells overexpressing wild type cGK I␤ but not the inactive mutant cGK I␤ϪK405A, further demonstrating that the observed inhibition of proliferation was indeed cGK-dependent and not due to cross-activation of cAK. cGK-mediated inhibition of cellular proliferation has also been described for other cell types such as vascular smooth muscle cells (45,46) and cardiac fibroblasts (47). Furthermore, vectormediated overexpression of eNOS or cGK I has been reported to decrease neointima proliferation in balloon catheter-damaged carotid arteries (48,49).
In our studies, the effect of cGMP analogs on proliferation appeared specific because the effect is absent in cGK I-deficient cells (whereas that of cAMP analogs is intact) but can be reconstituted by transfection of cGK I. This agrees with other publications that also show specific effects of cGMP on cGK. In contrast to the respective normal cells containing endogenous cGK I, platelets of cGK I-deficient mice lose 8-pCPT-cGMP-dependent inhibition of platelet aggregation (50), cGK I-deficient platelets of chronic myelocytic leukemia patients lose 8-pCPT-cGMP-dependent inhibition of intracellular Ca ϩ2 (51), and cGK-deficient endothelial cells lose 8-pCPT-cGMP-dependent inhibition of intracellular Ca ϩ2 and endothelial permeability (52). In contrast, the effect of a cAMP analog (50) or cAMPelevating agent (52) was intact.
Activation of T cells is associated with increased expression and release of IL-2, which stimulates T cell proliferation (4). cGK activation reduced antigen-receptor-stimulated IL-2 release and proliferation, and inhibition of anti-CD3-stimulated cell proliferation could be significantly reduced by addition of exogenous recombinant IL-2. These results indicate that inhibition of IL-2 release represents at least one mechanism by which cGK inhibits T cell proliferation.
Inhibition of T cell activation by the NO/cGMP/cGK signaling pathway may be of potential significance for counteracting inflammatory or lymphoproliferative processes. Studies have shown that NO (53,54), as well as cGMP (18), is involved in inhibition of T cell activation and inflammation. NO has also been reported to inhibit the proliferation of human leukemia cells (55); however, this requires study in greater detail. Because there are also cGMP-independent effects of NO and cGMP can affect other mediators than cGK, the role of cGMP and cGK in inflammation and lymphoproliferative disorders needs to be directly investigated in the future.
In summary, our data show that cGK I␤ is present only in freshly purified T cells, that cGK I-dependent inhibition of T cell proliferation is clearly distinct from that which is cAMP mediated, that cGK I-dependent inhibition of T cell proliferation results in part from inhibition of IL-2 release, and that not all NO inhibitory effects on T cell proliferation are controlled by cGK I. cGK I suppression of T cell proliferation may be an important independent adjunct to the inhibition mediated by cAMP and have implications for diverse aspects of immune function.