Cot kinase induces cyclooxygenase-2 expression in T cells through activation of the nuclear factor of activated T cells.

Cyclooxygenase-2 (COX-2) is induced in human T lymphocytes upon T cell receptor triggering. Here we report that Cot kinase, a mitogen-activated protein kinase kinase kinase involved in T cell activation, up-regulates COX-2 gene expression in Jurkat T cells. Induction of COX-2 promoter activity by Cot kinase occurred mainly through activation of the nuclear factor of activated T cells (NFAT). Mutation of the distal (-105/-97) and proximal (-76/-61) NFAT response elements in the COX-2 promoter abolished the activation induced by Cot kinase. Even more, coexpression of a dominant negative version of NFAT inhibited Cot kinase-mediated COX-2 promoter activation, whereas cotransfection of a constitutively active version of the calcium-dependent phosphatase calcineurin synergizes with Cot kinase in the up-regulation of COX-2 promoter-driven transcription. Strikingly, Cot kinase increased transactivation mediated by a GAL4-NFAT fusion protein containing the N-terminal transactivation domain of NFATp. In contrast to phorbol ester plus calcium ionophore A23187, Cot kinase increases both COX-2 promoter activity and NFAT-mediated transactivation in a cyclosporin A-independent manner. These data indicate that Cot kinase up-regulates COX-2 promoter-driven transcription through the NFAT response elements, being the Cot kinase-induced NFAT-dependent transactivation presumably implicated in this up-regulation.

In unstimulated T cells NFAT transcription factor is heavily phosphorylated and localized in the cytosol. Upon T cell activation, it becomes dephosphorylated by calcium/calcineurin phosphatase and translocates into the nucleus (for review, see Ref. 35). Stimulation of T cells by phorbol esters and calcium ionophore A23187 regulates not only the nuclear localization of NFAT, but also its transactivation activity (35,36).
In the present study, we have analyzed the regulation of COX-2 by Cot kinase in human T cells, showing that Cot induces transcription of COX-2 in these cells. Regulation of COX-2 expression by Cot kinase occurs mainly through activation of NFAT. Both distal and proximal NFAT response elements in the COX-2 promoter are essential for Cot kinasemediated induction. We also provide evidence that neither NF-B nor c-Jun is involved in this activation. Finally, we have determined that Cot kinase increases transactivation mediated by the N-terminal domain of NFAT. Involvement of Cot kinaseinduced NFAT transactivation in the COX-2 promoter activation is discussed further.
The plasmid pCMV-p65 contains the cDNA of the human p65 NF-B protein in the pCDNA3 expression vector and was a gift from Dr. Alcamí. The B luciferase reporter contains three copies of the ␤ consensus sequence of the immunoglobulins chain (37). The fulllength human NFATc (p1SH107c, NFATwt) and the dominant negative NFATc (p1SH102C⌬418, dnNFAT) expression plasmids were generously provided by Dr. Crabtree (38). The ⌬CAM-AI plasmid encoding a deletion mutant of a calcineurin catalytic subunit has been described previously (39). The pCMV-TAM67 plasmid encoding a dominant negative mutant variant of c-Jun and the plasmid pRSV c-Jun encoding the wild type c-Jun (c-Junwt) were a gift from Balduino Burgering. The GAL4-hNFAT1 (1-415) contains the first 415 amino acids of the human NFAT1/p fused to the DNA binding domain of the yeast GAL4 transcription factor (amino acids 1-147). This plasmid was constructed by subcloning a PCR fragment of the hNFAT1 (40) into the pABGALlinker plasmid (41) digested with XhoI. The 5Ј-primer used was GGctc-gagATGAACGCCCCCGAGCGGCAGC; the 3Ј-primer, CCCgtcgacT-TACTGCACCTCGATCCGCAGCTCG, contains a SalI site (lowercase letters). The sequence was confirmed by automatic sequencing. The GAL4-DBD is the parental vector pABGAL-linker plasmid. The GAL4luciferase reporter plasmid includes five copies of GAL4 DNA binding sites fused to the luciferase gene (42).
Reverse Transcription-PCR Assay-The human leukemia T cell line Jurkat was maintained in RPMI 1640 medium supplemented with 10% fetal calf serum (Life Technologies, Inc.), 50 g/ml gentamycin, and 2 mM L-glutamine (complete medium). Jurkat cells were electroporated with 10 g/ml pEF-BOS or pEF-BOS trunc-Cot as described by Ballester et al. (14). In these conditions the electroporation efficiency, as tested by green fluorescence protein expression, was about 30 -40%. 30 min after electroporation, transfected cells were stimulated overnight with 1 M calcium ionophore and 15 ng/ml TPA, and total RNA was extracted using the TRIzol protocol (Life Technologies, Inc.). 1 g of total RNA from control or stimulated cells was used to perform the reverse transcription reaction. The specific primers for either human COX-2 or glyceraldehyde-3-phosphate dehydrogenase for PCR amplification has been described previously (20). The PCR was amplified 20 -35 denaturation cycles at 94°C for 1 min, annealing at 60°C for 1 min, and extension at 72°C for 1 min. Amplified cDNA fragments were separated by agarose gel electrophoresis, and bands were visualized by ethidium bromide staining.
Promoter-driven Transcription and Transactivation Assays-Jurkat cells were cotransfected unless otherwise indicated, with 0.75 g/ml of different COX-2 promoter-luciferase reporter constructs together with 0.35 g/ml pEF-BOS, pEF-BOS Cot, pEF-BOS trunc-Cot, or pEF-BOS inac-Cot constructs utilizing Lipofectin reagent (Life Technologies, Inc.), according to the manufacturer's instructions. All transfections were normalized to the same amount of DNA. 20 h after transfection, cells were stimulated with 10 g/ml soluble ␣-CD3, isolated as described by Ballester et al. (13); 1 g/ml soluble ␣-CD28 (9.3 antibody, generously donated by Dr. C. June); 1 M calcium ionophore A23187 (Sigma), in the presence or absence of phorbol esters (50 ng/ml PDBu (Sigma) or 15 ng/ml TPA (Sigma). Cyclosporin A (CsA, 100 ng/ml, Sandoz) was added to the cells 1 h before stimulation. Cells were collected by centrifugation, and luciferase activity was determined by the luciferase assay kit (Promega), according to the manufacturer's instructions.
For transactivation assays, Jurkat cells were cotransfected with 0.3 g/ml GAL4-DBD or GAL4-hNFAT1 (1-415) expression vector and 0.5 g/ml GAL4 luciferase reporter plasmid together with 0.35 g/ml of the different pEF-BOS constructs. Transfected cells were stimulated or not as indicated above, and luciferase activity was measured after 48 h. Protein measurements were performed with the D c protein assay (Bio-Rad).

FIG. 1. Cot kinase activity induces COX-2 mRNA levels in Jurkat cells.
Total RNA from Jurkat cells electroporated with 10 g/ml pEF-BOS trunc-Cot or 10 g/ml pEF-BOS stimulated or not with 1 M calcium ionophore (Ion) and 15 ng/ml TPA, was used for reverse transcription-PCR analysis to measure COX-2 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA levels. Amplified cDNA fragments were separated by agarose gel electrophoresis, and bands were visualized by ethidium bromide staining. Data shown correspond to a number of cycles in which the amount of amplified product is proportional to the abundance of starting material. The figure shows one of the three experiments performed.

Cot Kinase Up-regulates COX-2 Gene Expression and COX-2
Promoter Activity-We have reported previously that T cell activation induced by phorbol esters plus calcium ionophore, or ␣-CD3 plus ␣-CD28, increases COX-2 mRNA levels in primary resting human T lymphocytes as well as in Jurkat T cells (20,21). In agreement with these data, COX-2 mRNA was increased in Jurkat cells transfected with the empty pEF-BOS vector upon treatment with calcium ionophore plus TPA (Fig.  1). Interestingly, cotransfection of Jurkat cells with an active form of Cot kinase (pEF-BOS trunc-Cot) increased the levels of COX-2 transcript. Thus, overexpression of truncated Cot kinase was sufficient to detect increased levels of COX-2 transcripts, although maximal induction was observed in cells overexpressing Cot kinase upon calcium ionophore plus TPA stimulation (Fig. 1).
To analyze if Cox-2 mRNA induction by Cot kinase correlated with an increase in the transcriptional activity mediated by the COX-2 promoter, Jurkat cells were cotransfected with different COX-2 promoter luciferase constructs and expression vectors for wild type or truncated Cot kinase. In agreement with the regulation of COX-2 mRNA levels by Cot kinase activity, expression of truncated Cot kinase strongly increased transcription driven by a construct spanning from Ϫ998 to ϩ 104 base pairs of the COX-2 promoter (P2-1102) ( Fig. 2A). Overexpression of wild type Cot kinase also up-regulated COX-2 promoter activity although to a slightly lower extent than constitutively active truncated Cot kinase ( Fig. 2A). We next decided to investigate whether COX-2 promoter activation after T cell stimulation requires endogenous Cot kinase. For this, Jurkat cells were transfected with the P2-1102 construct together with an inactive form of Cot kinase (pEF-BOS inac- Cot) and stimulated with PDBu plus calcium ionophore, or with ␣-CD3 plus ␣-CD28. Cotransfection of 0.35 g/ml or 0.11 g/ml of the pEF-BOS inac-Cot construct blocked activation of the COX-2 promoter-driven transcription triggered by ␣-CD3 plus ␣-CD28 (Fig. 2B). Treatment with PDBu plus calcium ionophore induced a higher stimulation of the COX-2 promoterdriven transcription, and consequently a higher concentration (0.35 g/ml) of pEF-BOS inac-Cot was required to abolish this induction (Fig. 2B).
In agreement with the up-regulation of COX-2 transcript levels ( Fig. 1), calcium ionophore and PDBu cooperated with truncated Cot kinase in the activation of COX-2 promoter (Fig.  3). Cot kinase also cooperated with ␣-CD3, ␣-CD28, or calcium ionophore but not with PDBu in the induction of COX-2 promoter transcription (data not shown).
Identification of the Cis-acting Elements Involved in COX-2 Promoter Up-regulation by Cot Kinase-It is known that Cot kinase up-regulates gene transcription through activation of NF-B (16 -18). On the other hand, the presence of a functional NF-B site in the COX-2 promoter located at Ϫ223/Ϫ214 has been reported. Therefore, we next decided to analyze the role of this site in the activation of COX-2 promoter by Cot kinase in T cells. For this, we compared the behavior of constructs P2-431 (containing this NF-B site), P2-274 (a deletion lacking this site), and a construct in which the NF-B site was mutated (P2-431 B-mut) in response to Cot kinase overexpression. The results shown in Fig. 4A indicated that these constructs were equally stimulated in response to Cot kinase or PDBu plus calcium ionophore treatment, discarding a role of this NF-B site in the activation of COX-2 promoter by Cot kinase in T cells (Fig. 4A). As a control, Cot kinase effectively induced transcription of luciferase reporter gene p3x-B, containing a trimer of a NF-B response element (not shown). The role of NF-␤ in COX-2 promoter activation was investigated further by cotransfecting p65 NF-B together with P2-431 or P2-274. As a control of NF-B-driven transcription stimulation, p65 NF-B was cotransfected with the reporter p3x-B luciferase. Transfection of p65 did not increase the transcriptional activity of P2-431 COX-2 but increased by about 4.5-fold the driven transcription activity of p3x-B (Fig. 4B). Taken together the above results discard a role of NF-B in the activation of COX-2 promoter by Cot kinase in T cells.
The role of the distal (dNFAT) and proximal (pNFAT) NFAT sites of the COX-2 promoter in the activation by Cot kinase was investigated by performing cotransfection experiments with P2-274 COX-2 promoter mutated in any of these sites or in both, together with pEF-BOS trunc-Cot. Mutation of either pNFAT or dNFAT sites in the P2-274 COX-2 promoter partially decreased the activation by Cot kinase. Interestingly, mutation of both NFAT sites abolished the induction of P2-274 COX-2 promoter transcription triggered by Cot kinase (Fig. 5A). These data suggest that induction of COX-2 promoter occurs through activation of NFAT transcription factor. Moreover, overexpression of NFATwt synergized with truncated Cot kinase in the induction of COX-2 promoter (around 40-fold). Stimulation by calcium ionophore plus PDBu of these transfected cells increased COX-2 promoter activity further (120-fold) (Fig. 5B). Conversely, overexpression of a dominant negative version of NFAT (dnNFAT) blocked the P2-274 COX-2 promoter-driven transcription induced by truncated Cot kinase as well as that triggered by calcium ionophore plus PDBu stimulation (Fig. 5B).
It has been shown previously that Cot kinase activates AP-1 (10,12,14) so we next decided to study the role of this transcription factor in the up-regulation of the COX-2 promoter by Cot kinase. For this, Jurkat cells were cotransfected with P2-274 and pEF-BOS trunc-Cot in the presence of a dominant negative version of c-Jun, named TAM67. Overexpression of TAM67 did not modify the induction of COX-2 promoter by Cot kinase (Fig. 6). On the contrary, it decreased COX-2 promoterdriven transcription induced by PDBu plus calcium ionophore (Ref. 21 and Fig. 6) as well as the activation triggered by truncated Cot kinase together with PDBu plus calcium ionophore stimulation (Fig. 6). Involvement of the Calcineurin/NFAT Pathway in the Activation of COX-2 Promoter by Cot Kinase-The immunosuppressive drug CsA inhibits calcineurin activity and consequently blocks the nuclear translocation of NFAT (43). Addition of CsA inhibited the stimulation of the COX-2 promoter induced by PDBu plus calcium ionophore as expected. In contrast, the up-regulation of COX-2 promoter induced by Cot kinase was CsA-independent (Fig. 7A). Cotransfection of ⌬CAM-AI (an activated form of calcineurin) increased COX-2 promoter activity about 3-fold (Fig. 7B). Mutation of any of the proximal or distal NFAT sites in P2-274 COX-2 promoter abolished this up-regulation. Strikingly, ⌬CAM-AI synergized with Cot kinase in COX-2 promoter activation (270-fold) (Fig. 7B). Mutation of any of the proximal or distal NFAT sites in P2-274 COX-2 promoter significantly decreased (by 60 -80%) up-regulation of COX-2 transcription triggered by overexpression of ⌬CAM-AI together with truncated Cot kinase. Full abrogation occurs when both sites were mutated.
Cot Kinase Up-regulates the Transactivation Function of NFAT-The above results suggest that Cot kinase was acting on the NFAT pathway in parallel to calcineurin. It has been described that phorbol ester and calcium ionophore stimulation of NFAT does not only involve its nuclear translocation, but also the optimal function of the transactivation domain located at the N terminus domain of NFAT (35,36,44). To study the implication of Cot kinase in the transactivation of NFAT, cotransfection experiments of Cot kinase together with GAL4-NFATp were performed. The GAL4-NFATp fusion protein is constitutively expressed in the nucleus because of the strong nuclear localization signal at the N terminus of GAL4 (45) and contains the N-terminal transactivation domain of NFATp. Overexpression of both truncated or wild type Cot strongly potentiated the function of the NFAT transactivation domain (Fig. 8A). As a control, Cot kinase did not induce activity when cotransfected with GAL4-DBD. Similar to what we observed with the inducibility of the COX-2 promoter, truncated Cot kinase activated more effectively than wild type Cot kinase transactivation mediated by NFAT. Even more, CsA inhibited NFAT transactivation induced by phorbol esters plus calcium ionophore, but not by Cot kinase activity, in the absence of any additional stimulus (Fig. 8B). DISCUSSION COX-2 has been implicated in inflammation processes and is the target of many nonsteroidal anti-inflammatory drugs (19). More recently, COX-2 has been also associated with oncogenic transformation and angiogenesis (46 -57). We have previously described that COX-2 is induced upon antigenic triggering in resting T cells, where it plays a role in controlling the process of T cell activation (20,21). In this report we have found that Cot kinase is involved in the regulation of COX-2 transcription in T cells. The fact that a kinase-deficient mutant of Cot kinase blocked the activation of COX-2 promoter induced by ␣-CD3 and ␣-CD28, as well as by phorbol esters plus calcium ionophore, indicates that Cot kinase or a Cot-like kinase plays a pivotal role in the up-regulation of COX-2 gene expression upon T cell activation. Cot kinase regulates the activity of several transcription factors induced in T cell activation such as NFAT, AP-1, or NF-kB (10 -18). Previous results indicate that Cot kinase increases interleukin-2 gene expression (14 -16) mainly by up-regulating the transcription driven by NF-B and the composite element NFAT-AP-1 of interleukin-2 promoter (16); Cot kinase also up-regulates the AP-1 response element of the collagenase promoter (12,14,16). It has been shown that COX-2 promoter contains binding sites for these transcription factors, acting as positive regulatory elements of COX-2 transcription in several cell types (22)(23)(24)(25)(26)(27)(28)(29)(30)(31)(32)(33)(34). However, we have shown here that activation of COX-2 promoter in T cells occurs in a NF-kB-independent manner, in agreement with previous results discarding a role of this transcription factor in this cell type. On the other hand, a dominant negative mutant of c-Jun inhibits PDBu plus calcium ionophore activation of COX-2 promoter, pointing to an important role of those signaling pathways leading to AP-1 activation (21). However, this mutant did not affect Cot kinase-mediated induction of Cox-2 promoter, suggesting that Cot could act through a c-Jun-independent pathway to activate this promoter.
Data obtained with the COX-2 promoter deletion constructs or with those with distal and proximal NFAT sites mutated, as well as the overexpression of a dominant negative version of NFAT, indicate that NFAT is required for Cot-mediated or PDBu plus calcium ionophore up-regulation of COX-2 promoter. The striking synergism between wtNFAT or ⌬CAM-AI and Cot kinase and the fact that CsA did not inhibit Cot kinase-mediated Cox-2 promoter induction supports the hypothesis that Cot kinase-induced NFAT-dependent transacti- vation is either downstream of calcineurin, or it represents a parallel pathway (Fig. 9). Here, we show that Cot kinase mediated up-regulation of genes transactivated by NFAT occurs through the increase in the transactivation function of NFAT. The Cot kinase induced-transactivation by NFAT described here can explain why Cot kinase up-regulates COX-2 promoter activity through their NFAT sites in addition to NFAT binding to these sites. Moreover, the up-regulation of transactivation by NFAT and the induction of COX-2 promoter-driven transcription by Cot kinase are both CsA-independent and could also explain the reported CsA-insensitive activation of interleukin-2 and tumor necrosis factor-␣ gene by Cot kinase in T cells (15,14).
NFAT transcription factors are regulated at two different levels, primarily at the level of subcellular localization and secondarily at the level of the intrinsic DNA binding activity. Tsatsanis et al. (15) have reported the regulation of NFAT by Cot kinase at the level of subcellular localization. Thus, they have shown that overexpression of Cot kinase induces the nuclear accumulation of HA-NFATc in the 3T3 fibroblasts. An integration of signal promoted by phorbol esters and calcium ionophore is required to increase the transactivation by NFAT (35,36,44). However, PDBu plus calcium ionophore activation of the NFAT transactivation domain is CsA-sensitive (35,36, and this report) in contrast to what we observed with Cot kinase. In this context, it remains to be established whether the reported accumulation of NFATc by Cot kinase in the nucleus (15) is a consequence of the increase of transactivation function of NFAT triggered by Cot kinase which thereby retained it in the nucleus. Another possibility is that Cot kinase regulates both the NFAT in/out shuttling of the nucleus and transactivation by NFAT by different ways, but with a common finality, to induce promoter transcription through NFAT response elements.
In conclusion, the data shown here indicate that Cot kinase controls COX-2 promoter activity mainly, if not exclusively, through the NFAT response elements. We also provide evidence of a new mechanism of NFAT transcription factor activity up-regulation by Cot kinase which helps to explain the important role of this kinase in the regulation of gene expression after T cell activation. Further analysis remains to be done to define the interactions between Cot kinase and the N-terminal transactivation domain of NFAT.