INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

CREB¹ is a 43-kDa basic leucine zipper (bZip) transcription factor composed of a C-terminal basic DNA-binding domain, an adjacent leucine zipper dimerization domain, and a kinase-inducible transcriptional activation domain. CREB binds to cAMP-responsive element sequence elements (TGANNTCA) both as a homodimer and as a heterodimer in association with other members of the CREB/ATF family, including ATF-1 and the cAMP-responsive element modulator (CREM) (1-6). Phosphorylation of Ser¹³³ within the kinase-inducible transcriptional activation domain of CREB is required to induce the transcriptional activity of the protein. Phosphorylation of Ser¹³³ activates CREB, at least in part, by facilitating its binding to the 256-kDa CREB-binding protein. The CREB·CREB-binding protein complex can, in turn, interact with and activate the basal transcriptional machinery (8, 9). Previous studies have demonstrated that multiple signaling pathways in different cell lineages can mediate the phosphorylation and activation of CREB. These include a protein kinase A (PKA)-dependent pathway that is activated by increased intracellular cAMP (2, 3), a calcium/calmodulin-dependent pathway in which CREB can be phosphorylated by CaM kinases II and IV (10), and a Ras-dependent pathway in which the serine/threonine kinase RSK2 is thought to phosphorylate CREB on Ser¹³³ (11).

Recent studies have demonstrated that transcriptionally active CREB is required for the activation of normal murine T cells following engagement of the T cell antigen receptor (12). Resting T cells contain exclusively unphosphorylated and inactive CREB. TCR cross-linking leads to the rapid and transient phosphorylation of CREB on Ser¹³³. More important, transgenic mice expressing a dominant-negative unphosphorylatable form of CREB display a profound T cell proliferative defect characterized by G₁ cell cycle arrest, markedly decreased IL-2 production, and defective transcriptional induction of multiple Fos and Jun proteins (12). These results were consistent with other reports that demonstrated that T and B cell activation results in CREB phosphorylation and increased CREB DNA-binding activity (13-15), that a CREB-binding site is necessary for the induction of the proliferating cell nuclear antigen gene in response to IL-2 in T cells (16, 17), and that there is a functionally important CREB-binding site in the c-fos promoter (18-20).

Despite the importance of CREB phosphorylation in normal T cell activation, the signaling pathways that regulate CREB phosphorylation (and dephosphorylation) following TCR engagement remained unknown. In the studies described in this report, we have used pharmacological agonists and antagonists, mutant T cell lines, and transient transfection approaches to better define the signaling pathways that regulate CREB phosphorylation in thymocytes and T cells. Our results show that although CREB can be independently phosphorylated by at least three distinct signaling pathways in T cells, TCR cross-linking appears to mediate CREB phosphorylation via a signaling pathway involving the activation of p56^lck, protein kinase C, Ras, Raf-1, MEK, and RSK2.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

Animals-- CD-1 mice were purchased from Charles River Laboratories (Wilmington, MA). Four to eight-week-old animals were used in the studies described in this report. Animals were maintained in the University of Chicago laboratory animal barrier facility in micro-isolator cages. All animal experimentation was carried out according to NIH guidelines and was approved by the animal care committee of the University of Chicago.

Cells-- The JCAM-1 and JCAM-1/p409^lck cell lines were a generous gift from Dr. David Straus (University of Chicago). Jurkat and JCAM-1 T cells were grown in RPMI 1640 medium supplemented with 10% fetal bovine serum (Life Technologies, Inc.), 1 mM glutamate, 100 µg/ml penicillin, and 100 µg/ml streptomycin. JCAM-1/p409^lck cells were grown in the presence of G418 (1 mg/ml) and hygromycin (250 µg/ml). Murine splenic T cells were purified using a commercially available T cell column (R&D Systems, Minneapolis, MN) according to the manufacturer's protocol. The resulting cells were >90% CD3⁺ as assessed by flow cytometry. Single cell suspensions of thymocytes and splenic T cells were cultured in Dulbecco's modified Eagle's medium (Life Technologies, Inc.) containing 10% fetal calf serum, 1 mM glutamate, 100 µg/ml penicillin, 100 µg/ml streptomycin, 0.1 mM nonessential amino acids, and 0.05 mM 2-mercaptoethanol at 37 °C. Splenic T cells and thymocytes (4 × 10⁶ cells/ml) were activated by treatment with plastic-immobilized anti-CD3 mAb 145.2C11 (16 µg/ml), PMA (10 ng/ml), or ionomycin (0.5 µg/ml). Jurkat T cells were activated by treatment with plastic-immobilized OKT3 antibody (10 µg/ml). Experiments using inhibitors involved preincubation with each inhibitor for 30 min at 37 °C prior to activation. The concentrations of inhibitors used were as follows: bisindolylmaleimide I, 12 µM; H-7, 10 µM; H-8, 10 µM; KN-93, 1 µM; W-7, 25 µM, chelerythrine chloride, 4 µM; and PD98059, 50 µM.

Transfections-- Exponentially growing Jurkat T cells (10⁷) were transiently transfected using a commercially available eletroporator (Bio-Rad; 250 V, 950 microfarads) with 25 µg of an expression vector encoding Gal4-CREB either alone or together with eukaryotic expression vectors encoding wild-type RSK2 (pMT2-HARSK2), a catalytically inactive mutant RSK2 (pMT2-HARSK2(KR100)) (a kind gift from Dr. M. E. Greenberg, Harvard Medical School), constitutively active p21^v-Ha-ras, a dominant-negative N17Ras, or a dominant-negative Raf-1 (21) containing an ATP-binding site mutation in the catalytic domain. Following electroporation, the cells were cultured for 48 h at 37 °C (1 × 10⁶ cells/ml), divided into aliquots, and then activated as described above. Activated cell extracts were used for Western blot analyses as described below.

Antibodies and Pharmacological Reagents-- Anti-murine CD3 (145.2C11) and anti-human CD3 (OKT3) antibodies were obtained from Pharmingen (San Diego, CA). PMA, ionomycin, forskolin, H-7 (1-(5-isoquinolinylsulfonyl)-2-methylpiperazine dihydrochloride), H-8 (N-(2-[methylamino]ethyl)-5-isoquiolinesulfonamide dihydrochloride), bisindolylmaleimide I, chelerythrine chloride, KN-93 (2-[N-(2-hydroxyethyl)-N-(4-methoxybenzenesulfonyl)]amino-N-(4-chlorocinnamyl)-N-methylbenzylamine), and W-7 (N-6-aminohexyl-5-chloro-1-naphthalenesulfonamide HCl) were purchased from Calbiochem.

Western Blot Analysis-- Protein samples corresponding to 2 × 10⁶ T cells were lysed in 50 µl of sample buffer (25 mM Tris-HCl (pH 6.7), 2% SDS, 10% glycerol, and 0.008% bromphenol blue). Proteins were resolved by SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes as described previously (21, 22). Phosphorylated CREB was detected using two different anti-phosphopeptide antibodies that have each been shown previously to react specifically and exclusively with the Ser¹³³-phosphorylated form of CREB: (i) a rabbit polyclonal IgG antibody, -pCREB (1:3000 dilution; a generous gift from Dr. M. E. Greenberg) (23), and (ii) a commercially available rabbit polyclonal IgG prepared against a Ser¹³³ phospho-CREB peptide (amino acids 123-136, KRREILSRRP(pS)YRK; 1:3000 dilution; Upstate Biotechnology, Inc.). Immunoreactivity was detected with horseradish peroxidase-conjugated goat anti-rabbit Ig (1:2500 dilution; Life Technologies, Inc.) using an enhanced chemiluminescence system (Kirkegaard & Perry Laboratories, Inc., Gaithersburg, MD). A rabbit polyclonal antibody that recognizes both phosphorylated and unphosphorylated forms of CREB (-CREB; 1:3000 dilution; a generous gift from Dr. M. E. Greenberg) was used to detect the total levels of CREB protein using identical Western blotting conditions.

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

Activation-induced Ser¹³³ Phosphorylation of CREB in Normal T Cells-- To assess changes in CREB phosphorylation in response to different activation signals in T cells, we performed Western blot analyses using two different rabbit polyclonal antibodies that have been shown previously to specifically and exclusively recognize the Ser¹³³-phosphorylated form of CREB (-pCREB) (23). In an initial series of experiments, purified murine splenic T cells were stimulated for 1-120 min with immobilized anti-CD3 mAb (-CD3), and the resulting cell lysates were subjected to Western blot analysis with the -pCREB antibody. T cell activation following TCR cross-linking by treatment with -CD3 induced the rapid but transient phosphorylation of CREB Ser¹³³ (Fig. 1A). Phosphorylation was observed as early as 1-2 min after TCR cross-linking, peaked within 5 min, and declined steadily over the next 2 h. Of note, total levels of CREB protein as assessed by Western blotting with an -CREB antibody were unchanged during -CD3-mediated T cell activation (Fig. 1A).

View larger version (30K): [in this window] [in a new window]   Fig. 1.   Activation-induced phosphorylation of CREB on Ser133 in normal T cells. Purified splenic T cells were activated by treatment with immobilized -CD3 mAb (A) or PMA, ionomycin, or PMA + ionomycin (B) for the times shown (in minutes). Protein samples corresponding to 2 × 106 cells were subjected to immunoblot analysis using two different polyclonal rabbit IgG antibodies specific for the Ser133-phosphorylated form of CREB (-pCREB) or a polyclonal rabbit Ig that recognizes both phosphorylated and unphosphorylated forms of CREB (-CREB). In B, the -CREB samples contain protein from the (PMA + ionomycin)-treated T cells. Uns, unstimulated.

Effects of Protein Kinase Inhibitors on -CD3-induced CREB Phosphorylation in T Cells-- The experiments described above suggested that multiple pathways could result in CREB phosphorylation in T cells. To determine which of these pathways is required for the TCR-mediated phosphorylation of CREB on Ser¹³³, we assessed the effects of specific protein kinase inhibitors on CREB phosphorylation following TCR cross-linking with an -CD3 mAb. Bisindolylmaleimide I has been shown to specifically inhibit PKC activity in intact cells (32, 33). As shown in Fig. 2A, bisindolylmaleimide I profoundly inhibited -CD3-mediated phosphorylation of CREB. The effect of bisindolylmaleimide I appeared to be specific for PKC because, in control experiments, the same dose of bisindolylmaleimide I inhibited PKC-induced CREB phosphorylation induced by treatment with PMA, but had no effect on PKA-mediated CREB phosphorylation induced by treatment with forskolin (Fig. 2A). Further evidence for the critical role of PKC in TCR-mediated CREB phosphorylation came from experiments in which two additional PKC inhibitors, H-7 and chelerythrine chloride, were also shown to inhibit -CD3-induced phosphorylation of CREB on Ser¹³³ (Fig. 2B). These results were also in accord with a recent study that demonstrated that depletion of PKC activity by prolonged treatment of T cells with PMA abrogated CREB Ser¹³³ phosphorylation in response to TCR cross-linking (7).

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Effects of protein kinase inhibitors on -CD3-induced phosphorylation of CREB Ser133 in splenic T cells. Purified splenic T cells were preincubated with medium alone (-CD3, PMA, forskolin, or ionomycin) or the indicated protein kinase inhibitors for 30 min. The cells were then activated by treatment with immobilized -CD3 mAb, PMA, or forskolin (A); immobilized -CD3 mAb or PMA (B); immobilized -CD3 mAb or forskolin (C); and immobilized -CD3 or ionomycin (D). Ser133 phospho-CREB (all blots in A, C, and D and -pCREB in B) and total CREB (-CREB in B) were detected by Western blot analysis as described in the legend to Fig. 1. Uns, unstimulated.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Effects of protein kinase inhibitors on -CD3-induced phosphorylation of CREB in thymocytes. Freshly isolated murine thymocytes were preincubated with medium alone (-CD3, forskolin, or ionomycin) or the indicated protein kinase inhibitors at 37 °C for 30 min. The cells were then activated by treatment with immobilized -CD3 mAb or forskolin (A), immobilized -CD3 mAb (B), or ionomycin (C). Ser133 phospho-CREB was detected by Western blot analysis as described in the legend to Fig. 1. In all experiments, total levels of CREB protein in each sample were shown to be equivalent by Western blot analysis with an -CREB antibody (data not shown). Uns, unstimulated.

TCR-induced Phosphorylation of CREB Requires the Protein-tyrosine Kinase p56^lck-- Although studies of normal peripheral T cells and thymocytes provide the most accurate assessment of the role of specific signaling pathways in T cell activation, these cells are difficult to transfect and to genetically manipulate. Thus, it would be useful to identify an immortalized T cell line in which CREB phosphorylation could be induced by TCR cross-linking. Recent studies have demonstrated that stimulation of Jurkat T cells with the -CD3 mAb OKT3 results in the phosphorylation of CREB specifically and exclusively on Ser¹³³ (7). We have confirmed these findings and have shown that treatment of Jurkat cells with -CD3 mAb, PMA, or ionomycin induces CREB phosphorylation on Ser¹³³ (Fig. 4, A and B). Thus, wild-type and mutant Jurkat cells represent a useful model system for studying CREB phosphorylation in a cultured cell line.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Requirement for p56lck in TCR-induced phosphorylation of CREB in Jurkat T cells. Jurkat T cells (A-C), JCAM-1 p56lck-deficient T cells (B and C), or JCAM-1/p409lck cells stably transfected with a p56lck expression vector (B) were activated by treatment with PMA, ionomycin, or PMA + ionomycin (A and C) or immobilized OKT3 (B) for the times shown (in minutes). Ser133 phospho-CREB (-pCREB) and total CREB protein (-CREB) were detected by Western blot analysis as described in the legend to Fig. 1. Uns, unstimulated.

Activation of Ras, Raf-1, and MEK Is Required for CREB Phosphorylation in Response to TCR Engagement-- Ligand binding to many growth factor receptors results in the activation of non-receptor tyrosine kinases, leading to stimulation of a Ras-dependent kinase cascade that includes sequential phosphorylation and activation of Raf, MEK (mitogen-activated protein kinase/extracellular signal-regulated protein kinase kinase), mitogen-activated protein kinase (MAPK), and ribosomal protein S6 kinase (pp90^rsk or RSK) (38, 39). Activated MAPKs as well as members of the pp90^rsk family are known to translocate to the nucleus and to phosphorylate several transcription factors (39, 40). In PC12 cells, a Ras-dependent Ser/Thr protein kinase has been shown to phosphorylate CREB in response to nerve growth factor stimulation (11). Subsequent analyses showed this CREB kinase to be identical to the Ser/Thr protein kinase RSK2 (41). The p21^ras pathway is known to be rapidly activated in response to TCR engagement by both PKC-dependent and protein-tyrosine kinase-dependent pathways (42, 43), thus raising the possibility that Ras and RSK2 might be important for CREB phosphorylation in response to TCR engagement. To more directly determine the role of p21^ras in TCR-induced CREB phosphorylation, we co-transfected Jurkat T cells with an expression vector encoding a recombinant form of CREB (Gal4-CREB) and expression vectors encoding either a constitutively active Ras (CA-Ras) or a dominant-negative mutant Ras protein, N17Ras (DN-Ras). The difference in the size of Gal4-CREB and endogenous CREB made it possible to distinguish the two molecules by SDS-polyacrylamide gel electrophoresis. Co-transfection with CA-Ras and Gal4-CREB resulted in CREB phosphorylation, even in the absence of TCR engagement (Fig. 5A). More important, expression of DN-Ras markedly inhibited -CD3-induced phosphorylation of Gal4-CREB (Fig. 5A). Moreover, -CD3 (but not CA-Ras)-induced CREB phosphorylation was inhibited by the PKC inhibitor bisindolylmaleimide I, suggesting that Ras lies downstream of PKC in the CREB activation pathway in T cells (Fig. 5A). These results were not due simply to differences in transfection efficiencies or to levels of expression of Gal4-CREB in the different transfected cell cultures because identical results were observed in at least three independent transfection experiments.

View larger version (26K): [in this window] [in a new window]   Fig. 5.   Roles of Ras and Raf-1 in the -CD3-mediated phosphorylation of Gal4-CREB Ser133. A, Jurkat T cells were transiently transfected with expression vectors encoding Gal4-CREB, DN-Ras, or CA-Ras as indicated. Phosphorylation of Gal4-CREB Ser133 in the presence or absence of bisindolylmaleimide I (Bis) was detected by Western blot analysis as described in the legend to Fig. 1. B, Jurkat T cells were transiently transfected with expression vectors encoding Gal4-CREB and a dominant-negative form of Raf-1 (DN-raf) as indicated. Following transfection, the cultures were divided into two aliquots. One aliquot was activated with immobilized -CD3 mAb, whereas the other was treated with forskolin. Phosphorylation of Gal4-CREB Ser133 (pGal4-CREB) was detected by Western blot analysis as described in the legend to Fig. 1. Uns, unstimulated.

View larger version (23K): [in this window] [in a new window]   Fig. 6.   MEK inhibitor PD98059 inhibits -CD3-mediated phosphorylation of CREB Ser133. Purified splenic T cells were preincubated with medium alone (unstimulated (Uns)) or with the MEK inhibitor PD98059 for 30 min and then activated with immobilized -CD3 mAb or forskolin. Ser133 phospo-CREB (-pCREB) and total CREB protein (-CREB) were detected by Western blot analysis as described in the legend to Fig. 1.

Role of RSK2 in TCR-mediated CREB Phosphorylation-- As described above, in PC12 cells, nerve growth factor-mediated activation of Ras leads to the subsequent activation of RSK2, which in turn phosphorylates and activates CREB. To assess the role of RSK2 in the TCR-induced phosphorylation of CREB, we transfected Jurkat T cells with expression vectors encoding wild-type RSK2 or a catalytically inactive mutant of RSK2 (RSK2(KR100)) along with a Gal4-CREB expression vector. TCR-mediated stimulation of Jurkat T cells transfected with Gal4-CREB resulted in low level phosphorylation of Gal4-CREB (Fig. 7A, eighth lane). Phosphorylation of Gal4-CREB was dramatically increased by co-transfection of the wild-type RSK2 expression vector (Fig. 7A, twelfth lane). This increase in Gal4-CREB phosphorylation required a kinase-active form of RSK2 because it was not observed in cells co-transfected with catalytically inactive RSK2(KR100) (Fig. 7A, fourteenth lane). These differences in Gal4-CREB phosphorylation were not due to differences in transfection efficiencies or levels of expression of Gal4-CREB because treatment of an aliquot of each transfected culture with forskolin induced comparable levels of Gal4-CREB phosphorylation, whether or not the cells were co-transfected with the RSK2 expression vector (Fig. 7A, compare ninth, twelfth, and fifteenth lanes). Thus, the effects of RSK2 overexpression on CREB phosphorylation were specific for TCR-mediated activation. In addition, the RSK2-dependent phosphorylation of CREB in response to -CD3 stimulation required activation of PKC because bisindolylmaleimide I, but not H-8, inhibited Gal4-CREB phosphorylation in cells co-transfected with the RSK2 expression vector and stimulated by TCR engagement (Fig. 7B). Taken together, these experiments demonstrated that RSK2 participates in the TCR-mediated phosphorylation of CREB and suggested that RSK2 and PKC may belong to a common signaling pathway. Finally, TCR-mediated Gal4-CREB phosphorylation in Jurkat cells transfected with RSK2 was abrogated by co-transfection with a dominant-negative Ras expression vector (Fig. 7C), confirming the requirement for Ras in RSK2-dependent, -CD3-induced CREB Ser¹³³ phosphorylation.

View larger version (30K): [in this window] [in a new window]   Fig. 7.   RSK2 is required for the -CD3-mediated phosphorylation of Gal4-CREB. Jurkat T cells were transiently transfected with expression vectors encoding Gal4, Gal4-CREB, wild-type RSK2, catalytically inactive RSK2(KR100), or DN-Ras as indicated. Live cells (1 × 106) were activated with immobilized OKT3 (-CD3) or forskolin (100 µM) for 10 min (A). The cells were preincubated with bisindolylmaleimide I (Bis) or H-8 as described under "Experimental Procedures" (B). Phosphorylation of Gal4-CREB Ser133 (pGal4-CREB) was detected by Western blot analysis as described in the legend to Fig. 1. Uns, unstimulated.

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion References

Activation of pre-existing CREB by phosphorylation on Ser¹³³ is one of the early events in the signaling pathways activated by engagement of the T cell antigen receptor. This rapid and transient activation of CREB is critically important for the normal transcriptional induction of specific AP1 family members, the transcriptional activation of the IL-2 gene, and subsequent cell cycle progression and T cell proliferation (12). In the studies described in this report, we have investigated the signaling pathways responsible for CREB phosphorylation in normal peripheral T cells and thymocytes. Our results demonstrate that CREB phosphorylation in these cells can be mediated by at least three distinct pathways: (i) a PKA-dependent pathway that can be simulated by treatment with forskolin and blocked by H-8, (ii) a calcium/calmodulin-dependent pathway that can be simulated by treatment with ionomycin and blocked by W-7, and (iii) a PKC-dependent pathway that can be simulated by treatment with PMA and blocked by chelerythrine chloride, H-7, and bisindolylmaleimide I. Despite the multiplicity of signaling pathways that are capable of mediating CREB phosphorylation in T cells, our experiments are consistent with a model in which CREB phosphorylation and activation are mediated by a single signaling pathway that requires activation of p56^lck, PKC, Ras, Raf-1, MEK, and RSK2.

Our conclusion that activation of PKC is required for TCR-mediated CREB Ser¹³³ phosphorylation is based on the use of pharmacological inhibitors that could potentially have effects on multiple signal transduction molecules. However, the validity of our conclusions concerning the role of PKC is supported by the following considerations: (i) three different inhibitors of PKC abrogated TCR-mediated CREB phosphorylation, whereas three inhibitors of PKA and calcium/calmodulin kinase pathways had no effect on CREB phosphorylation following TCR cross-linking; (ii) in parallel control experiments, PKC inhibitors had no effect on forskolin- or ionomycin-mediated CREB phosphorylation in the same cells; and (iii) our results in normal T cells and thymocytes are consistent with a previous report that demonstrated that depletion of activated PKC by prolonged treatment of Jurkat T cells with PMA also abrogated TCR-mediated CREB Ser¹³³ phosphorylation (7). Moreover, the previous finding of the importance of PKC in mediating CREB phosphorylation following immunoglobulin cross-linking in B cells suggests that common signaling pathway(s) may regulate CREB activation in response to antigen receptor cross-linking in these two different lymphoid cell lineages (15).

Previous studies have demonstrated a critical role for p21^ras in TCR-mediated induction of the IL-2 gene and T cell proliferation (26-31, 42-44). Our results demonstrate that Ras, Raf-1, and MEK also play critical roles in the TCR-mediated activation of CREB in T cells. Specifically, we have shown that (i) expression of constitutively active Ras resulted in the phosphorylation of CREB even in the absence of TCR signaling; (ii) TCR-induced phosphorylation of CREB was inhibited by expression of dominant-negative mutant Ras or Raf-1 proteins as well as by PD98059, an inhibitor of MEK; and (iii) expression of the dominant-negative Ras protein also abrogated the RSK2-dependent phosphorylation of CREB following TCR engagement. Although our findings do not allow us to definitively conclude whether Ras lies upstream or downstream of RSK2 in the CREB activation pathway, by analogy with the PC12 system (11, 41), we would propose a model in which TCR engagement leads to the activation of Ras, Raf-1, and MEK, which in turn results in RSK2 activation.

At least two signaling pathways have been shown to be capable of activating p21^ras in response to TCR engagement (26-31, 44). The first of these requires protein-tyrosine kinases and can be blocked by inhibitors such as herbimycin. The second pathway appears to be unique to T cells and involves the activation p21^ras by PKC. Our findings that PKC inhibitors block TCR-mediated CREB phosphorylation but fail to block CREB phosphorylation in response to expression of constitutively active p21^ras and that expression of dominant-negative N17Ras also blocks TCR-mediated CREB phosphorylation suggest that PKC-mediated activation of p21^ras is important for TCR-mediated activation of CREB.

When considered in the context of our current understanding of TCR-mediated signaling pathways, our results suggest the following working model for TCR-mediated CREB activation. Cross-linking of TCR leads to the rapid activation of p56^lck, which results in the activation of PKC. Activated PKC in turn leads to the activation of Ras-Raf-1-MEK-MAPK. Activated MAPK then activates RSK2, which phosphorylates CREB on Ser¹³³. It should be emphasized that such a straightforward linear signaling model is likely oversimplified and that several features of the model remain untested or unproved. For example, the pathways by which activated p56^lck leads to PKC activation remain unclear, as do the links between PKC and Ras. Similarly, we have not formally demonstrated the relationship between activated MAPK and RSK2 activation in T cells, nor have we demonstrated that RSK2 directly phosphorylates CREB Ser¹³³ in vivo. Despite these caveats, our results have identified a number of the critical signaling components that regulate CREB activation following stimulation through the T cell antigen receptor, and our working model suggests future experiments designed to more precisely elucidate this important T cell signaling pathway.

The inhibition of T cell activation is important for the treatment of both autoimmune diseases and transplant rejection. Currently available immunosuppressive agents target the calcineurin-dependent activation of NFAT (e.g. cyclosporin A) or the pathway that leads to the activation of NF-B (e.g. glucocorticoids and aspirin). Given the importance of CREB phosphorylation in T cell activation, the CREB activation pathway described herein represents a potentially novel target for the development of immunosuppressive drugs. To obtain specificity, it will likely be important to inhibit distal parts of the pathway that are specific for CREB phosphorylation rather than proximal signaling molecules such as PKC or Ras, which play more generalized roles in cellular homeostasis in many mammalian cell lineages. Nevertheless, our previous finding that expression of a dominant-negative (unphosphorylatable) form of CREB markedly inhibits IL-2 expression and T cell proliferation (12) suggests that specific inhibitors of this pathway may have potent immunosuppressive effects.