Bruton’s Tyrosine Kinase-mediated Interleukin-2 Gene Activation in Mast Cells

Cross-linking of the high affinity IgE receptor (FcεRI) on mast cells induces secretion of cytokines, including interleukin (IL)-2, through transcriptional activation of cytokine genes. Previously, defects in the gene coding for Bruton’s tyrosine kinase (Btk) were shown to result in defective cytokine production in mast cells, and thereby mice carrying btk mutations exhibited diminished anaphylactic reactions in response to IgE and antigen. In this study, we provide evidence that the transcription factors involved in the IL-2 gene expression in T cells are also required for maximal activation of the IL-2 gene in FcεRI-stimulated mast cells. Among them, AP-1 (Jun/Fos) and NF-AT were identified as candidate transcription factors that are regulated by Btk. Consistent with our previous data indicating that Btk regulates stress-activated protein kinases, c-Jun N-terminal kinase (JNK), c-Jun and other JNK-regulatable transcription factors are activated by FcεRI cross-linking in a Btk-dependent manner. Further, FcεRI-induced IL-2 gene activation is dependent on c-Jun and a component, SEK1, of its upstream activation pathway. Collectively, these data demonstrate that Btk regulates the transcription of the IL-2 gene through the JNK-regulatable transcription factors in FcεRI-stimulated mast cells.


Cross-linking of the high affinity IgE receptor (Fc⑀RI) on mast cells induces secretion of cytokines, including interleukin (IL)-2, through transcriptional activation of cytokine genes. Previously, defects in the gene coding for Bruton's tyrosine kinase (Btk) were shown to result in defective cytokine production in mast cells, and thereby mice carrying btk mutations exhibited diminished anaphylactic reactions in response to IgE and antigen. In this study, we provide evidence that the transcription factors involved in the IL-2 gene expression in T cells are also required for maximal activation of the IL-2 gene in Fc⑀RI-stimulated mast cells. Among them, AP-1 (Jun/Fos) and NF-AT were identified as candidate transcription factors that are regulated by Btk. Consistent with our previous data indicating that Btk regulates stress-activated protein kinases, c-Jun N-terminal kinase (JNK), c-Jun and other JNK-regulatable transcription factors are activated by Fc⑀RI cross-linking in a
Btk-dependent manner. Further, Fc⑀RI-induced IL-2 gene activation is dependent on c-Jun and a component, SEK1, of its upstream activation pathway. Collectively, these data demonstrate that Btk regulates the transcription of the IL-2 gene through the JNK-regulatable transcription factors in Fc⑀RI-stimulated mast cells.
Bruton's tyrosine kinase (Btk) plays an indispensable function in B cell development, as exemplified in human (X-linked agammaglobulinemia) and mouse (xid) immunodeficiencies (11)(12)(13)(14). Btk is also implicated in signal transduction for several immune cell and cytokine receptors (15,16). In contrast with Lyn and Syk, Btk and Emt (also known as Itk and Tsk) do not physically associate with Fc⑀RI. However, these Tec family protein-tyrosine kinases are activated by Fc⑀RI cross-linking (17,18). Recently, Btk was shown to play an important role in the Fc⑀RI signaling system. 2 Thus, xid (with a substitution of Arg-28 in Btk with Cys) and btk null mice exhibited diminished anaphylactic reactions. Correspondingly, Fc⑀RI-induced cytokine production was severely impaired in mast cells derived from these mutant mice. Defects in the Fc⑀RI-induced production of interleukin (IL)-2 and tumor necrosis factor ␣ (TNF-␣) in the mutant cells were found in the transcription of these cytokine genes. 2 Fc⑀RI cross-linking induces activation of three major mitogen-activated protein (MAP) kinases, i.e. ERK1/2, JNK1/2, and p38 (20 -25). Activities of MAP kinases are regulated through a unique set of protein kinases in a cascade, e.g. Raf-1 3 MEK1/2 3 ERK1/2 and MEKK1 3 SEK1 (also known as MKK4, MEK4, or JNKK) 3 JNK1/2 (26 -28). Phosphorylation targets of MAP kinases include transcription factors. Typically, JNK phosphorylates c-Jun, a component of the AP-1 transcription factor, at the critical serine residues (Ser-63 and Ser-73) in the activation domain of c-Jun (29,30), causing increased c-Jun transcriptional activity (31)(32)(33). More recently, similar phosphorylation and transcriptional activation by JNK were shown for other transcription factors such as ATF2 (34,35), Elk-1 (36), and Sap-1a (37). xid and btk null mutations affect JNK activities severely and, to a lesser extent, p38 in Fc⑀RI-stimulated mast cells, while ERK activities are not significantly changed (25).
Transcriptional regulation of IL-2 and TNF-␣ genes has been extensively characterized in T cells. IL-2 promoter activity depends on the 300-base pair region upstream of the transcription initiation site (38 -40). Several cis-acting elements that are binding sites for the nuclear factor of activated T cells (NF-AT), nuclear factor-B (NF-B), CD28RC, AP-1, and Oct have been identified (41,42). Cooperation of these transcription factors is required for maximal activation of the IL-2 promoter. NF-AT, a family of related proteins, exists in the cytoplasm before activation and translocates to the nucleus in response to an increase in intracellular concentrations of Ca 2ϩ (43). This translocation depends on dephosphorylation of NF-AT by a Ca 2ϩ -and calmodulin-dependent phosphatase, calcineurin, which is a major target of the immunosuppressive drugs cyclosporin A and FK506. Translocated NF-AT forms a complex with a ubiquitous nuclear factor AP-1 to exert its transcriptional action (reviewed in Ref. 44). NF-B is composed of Rel family members of p50 and p65 (reviewed in Ref. 45). p50/p65 heterodimers seem to stimulate the transcription of the IL-2 gene, while p50/p50 homodimers inhibit IL-2 gene transcription (46). The IL-2 promoter region contains two conventional and two noncanonical AP-1 binding sites. The latter AP-1 sites exist in a close apposition to the distal NF-AT and proximal Oct sites. AP-1 sites are bound by Jun/Fos or Jun/Jun dimers. The binding activity for the proximal Oct (Octp) site is constitutive, although the Octp site-dependent transcription requires T cell stimulation (40). Similar to the IL-2 gene, transcriptional activation of the TNF-␣ gene induced by T cell receptor stimulation involves NF-ATp and ATF-2/Jun heterodimer proteins bound to the 3 site and a cyclic AMP response element, respectively, in the 200-base pair TNF-␣ promoter (47)(48)(49).
In this study, we determined which cis-regulatory elements in the IL-2 gene promoter are involved in Fc⑀RI-induced activation of the IL-2 gene. Among the transcription factors involved in activation of the IL-2 gene, c-Jun and other JNKactivable transcription factors are shown to be regulated by Btk. Since Btk regulates JNK activity, we investigated the possible involvement of the MEKK1 3 SEK1 3 JNK pathway in the IL-2 gene activation in mast cells.

EXPERIMENTAL PROCEDURES
Reagents-B6/129 F2 mice were purchased from Jackson Laboratory. btk null mice (50) originally provided by Drs. Wasif N. Khan and Frederick W. Alt were bred in an animal facility in La Jolla Institute for Allergy and Immunology. Recombinant rat stem cell factor was kindly donated by Amgen. Anti-Btk antibody was described previously (51). Antibodies against MEK-4 (SEK1) and MEKK1 were from Santa Cruz Biotechnology, Inc. Anti-phospho-c-Jun(Ser-63) and anti-c-Jun antibodies were purchased from New England Biolabs, and anti-EE antibody was a kind gift of Dr. Arpad Molnar.
Cells-Bone marrow cells from femurs of B6/129 F2 (wild-type (WT)) and btk null mice with the same genetic background were cultured in RPMI 1640 medium supplemented with 10% fetal calf serum, 50 M 2-mercaptoethanol, 2 mM glutamine, and IL-3 (culture supernatants of mouse IL-3 gene transfected cells). For transient transfection experiments, mast cells (BMMC for bone marrow-derived mast cells) after 4 weeks of culture were used. At this point, more than 95% of the live cells are immature mast cells. The lack of Btk expression in btk null mice was routinely monitored by immunoblotting of spleen cell lysates with anti-Btk.
DNA Constructs-WT and K430R btk cDNAs in expression vectors, pMX-puro (a retroviral vector) and pME18S, have been described (25,52). An IL-2 gene promoter construct (IL-2Luc, encompassing nucleotides Ϫ321 to ϩ46) fused with the firefly luciferase gene and some mutated versions of this construct, i.e. mut NF-ATdLuc, mut NF-BLuc, mut AP-1pLuc, and mut OctpLuc, were described previously (53). Several more mutants of the IL-2 promoter were generated by two-step polymerase chain reactions in this study (Fig. 1A) and cloned into the KpnI/HindIII site of a luciferase reporter vector, pUC00Luc (54). Mutations were confirmed by nucleotide sequencing. Luciferase reporter constructs in the pGL2-promoter vector (Promega), containing multiple copies of the NF-ATd/AP-1 (Ϫ290 to Ϫ261, seven times), NF-B (Ϫ211 to Ϫ192, eight times), or AP-1/Octp (Ϫ97 to Ϫ64, seven times) sites, were described (54). pG5E1bLuc (55), a reporter plasmid containing five copies of GAL4 binding sites fused with the firefly luciferase gene, and GAL4 DNA binding domain expression vectors (56)  Stable Transfectants-Mast cells derived from btk null mice after 2 weeks of culture in IL-3 were incubated in the presence of rat stem cell factor as well as IL-3 for another 2 weeks to accelerate cell proliferation before retroviral infection. pMX-puro vectors containing no cDNAs, WT btk cDNA, or K430R btk cDNA were transduced into BOSC-23 packaging cells (64) with LipofectAMINE (Life Technologies, Inc.) according to the manufacturer's instructions. Culture supernatants were recovered 48 h after transfection. Virus titers (1-5 ϫ 10 5 colony-forming units/ml) in the supernatants were measured by infecting NIH/3T3 cells and selecting in 1 g/ml puromycin. Mast cells cultured in the presence of stem cell factor were infected with the retroviruses in the presence of 10 g/ml polybrene. Forty-eight h after infection, selection in puromycin (1.5 g/ml) was started to obtain sufficient numbers of drug-resistant cells. Drug was omitted from the culture 2 days before transient transfection. Successful transfection by WT btk cDNA was routinely confirmed by restored secretion of IL-2 and TNF-␣ from immunologically stimulated transfectants. 2 Transient Transfection and Luciferase Assays-BMMC or stably transfected mast cells (1-1.5 ϫ 10 7 cells) prepared as described above were transfected with 5-10-g reporter plasmids together with or without the indicated combinations of expression plasmids by electroporation at 400 V and 950 microfarads using a Bio-Rad Gene Pulser II system. Twenty-four h after transfection, cells were started to be sensitized overnight with anti-dinitrophenyl monoclonal IgE antibody. Forty-one h after transfection, cells were left unstimulated or stimulated with antigen, 30 ng/ml dinitrophenyl conjugates of human serum albumin for 7 h before cell harvest. Cells were lysed in 0.2% Triton X-100 in 100 mM potassium phosphate (pH 7.8)/1 mM dithiothreitol. Luminescence of cleared lysates after the addition of ATP and luciferin solutions was measured using a Monolight 2010 luminometer (Analytical Luminescence Laboratory).
Immunoblotting-Cells were lysed in 1% Nonidet P-40-containing lysis buffer (65), and cleared lysates were analyzed by SDS-polyacrylamide gel electrophoresis followed by electroblotting. Blots were blocked and incubated consecutively with primary antibody, horseradish peroxidase-conjugated secondary antibody, and enhanced chemiluminescence reagents (Amersham Pharmacia Biotech or NEN Life Science Products). Fluorescent bands were visualized by exposure to Reflection TM autoradiography films (NEN Life Science Products).

AP-1 and NF-AT Elements Are Crucial for IL-2 Gene Transcription in Fc⑀RI-stimulated Mast Cells-
We and others recently showed that rodent mast cells, including BMMC, secrete IL-2 in addition to numerous other cytokines upon Fc⑀RI crosslinking (66,67). Transcriptional activation of the IL-2 gene turned out to be a major mechanism for this phenomenon. 2 In the murine IL-2 promoter region, several cis-regulatory elements have been identified (41,42). Therefore, it is interesting to determine which element is important for IL-2 gene transcription in mast cells. To this end, expression of IL-2 promoter-luciferase constructs with mutations at each cis-element was measured in transiently transfected BMMC derived from B6/129 F2 or CBA/J mice (Fig. 1). Mutations in the distal NF-AT (NF-ATd, positions Ϫ285 to Ϫ281), AP-1 next to the NF-ATd site (NF-ATd/AP-1, Ϫ277 to Ϫ274), distal Oct (Octd, sites drastically reduced Fc⑀RI-induced transcriptional activation of the IL-2 promoter, while the effect of mutations at the distal AP-1 site (positions Ϫ185 to Ϫ183) was relatively small ( 1 ⁄5 (B6/129 F2) or 2 ⁄3 (CBA/J) of the activity of the WT promoter). These results indicate that most of the known cis-regulatory elements are required for maximal activation of the IL-2 promoter in mast cells as well as in T cells. The most severe impairments in Fc⑀RI-mediated IL-2 promoter activation were observed with the mutant promoters containing mutated NF-ATd, NF-ATd/AP-1, NF-ATp, or AP-1/Octp sites in BMMC from both mouse strains (Figs. 1, B and C), suggesting that these AP-1 and NF-AT elements are crucial for Fc⑀RI-mediated IL-2 gene transcription in mast cells. These data are consistent with the accumulated observations on T cells that NF-AT family and AP-1 proteins play essential roles in the transcription of the IL-2 gene (44).
Btk Regulates IL-2 Gene Transcription through AP-1 and/or NF-AT Sites-We recently showed that Fc⑀RI-induced transcriptional activation of the IL-2 as well as TNF-␣ genes is severely compromised in xid and btk null mast cells. 2 To further confirm this observation, we measured transcriptional activities of the IL-2 promoter in btk null BMMC that were stably transfected with the empty vector, WT btk, or kinasedead (K430R) btk cDNAs. Comparable expression of WT and K430R Btk in the transfectants was confirmed by immunoblotting ( Fig. 2A). Upon Fc⑀RI cross-linking, the transcriptional activity was strongly induced in WT btk transfected cells compared with those in vector or kinase-dead btk transfected cells ( Fig. 2A). This result demonstrates that Btk regulates expression of the IL-2 gene and that the intact enzymatic activity of Btk is required for this activity. This Btk function could be ascribed to its effects on mast cell differentiation. However, this possibility is unlikely for two reasons. First, btk null BMMC is phenotypically indistinguishable from WT BMMC. 2 Second, transient expression of WT Btk, but not vector or K430R Btk, in btk null BMMC restored the Fc⑀RI-induced transcriptional activation of the IL-2 gene (Fig. 2B). In these experiments, cells were harvested for luciferase assays only 48 h after transfection. This short period of time would not be long enough for cells to differentiate. These data together with the results shown in Fig. 1 raise the possibility that Btk regulates IL-2 gene transcription by regulating the transcription factors that target the critical AP-1 and/or NF-AT sites. Therefore, we examined the transcriptional activity of individual cis-elements of the IL-2 promoter in btk null BMMC. For these transient transfection experiments, we used the basic enhancerless SV40 promoter-luciferase fusion vectors containing multiple copies of either the NF-ATd/AP-1 or AP-1/Octp sequences. Robust transcriptional activation was seen of NF-ATd/AP-1-Luc and AP-1/Octp-Luc constructs before Fc⑀RI stimulation of WT btk-transfected cells, and Fc⑀RI cross-linking further enhanced their transcriptional activities (Fig. 2C). In contrast, a similar luciferase construct containing eight copies of the NF-B site exhibited a high basal transcriptional activity, but no further enhancement by Fc⑀RI cross-linking was observed. Transcriptional activities of these reporter constructs were much smaller in vector or kinase-dead btk transfected cells. These data suggest that Btk-regulated IL-2 gene transcription is mediated by AP-1-and/or NF-AT-dependent mechanisms.

Btk Regulates c-Jun and Other Transcription Factors That Are Activated by JNK-Previous studies demonstrated that
Fc⑀RI cross-linking induces activation of all three major subfamilies of MAP kinases, i.e. ERK1/2, JNK1/2, and p38 (20 -25). Fig. 2. Btk regulates IL-2 gene transcription. A, btk null BMMC stably transfected with vector, WT btk, or K430R btk cDNAs were electroporated with 8 g of IL-2Luc plasmid. Cell stimulation, lysis, and luciferase assays were performed as described in the Fig. 1 legend. -Fold luciferase activities relative to that of unstimulated vector transfectants are shown (top). A representative result is shown out of at least 15 experiments. Expression of Btk protein in btk null BMMC stably transfected with vector, WT btk, or K430R btk cDNAs was detected by immunoblotting. Positions of molecular mass standards and Btk are indicated (bottom). B, btk null BMMC were co-transfected with 8 g of IL-2Luc and 20 g of pME18S, pME-btk, or pME-btk(K430R). Cell stimulation, lysis, and luciferase assays were performed as above. -Fold luciferase activities relative to that of unstimulated vector transfectants are shown. Shown is a representative result out of two experiments. C, btk null BMMC were co-transfected with 8 g of NF-ATd/AP-1-Luc, AP-1/Octp-Luc, or NFB-Luc, together with 20 g of pME18S, pME-btk, or pME-btk(K430R). Luciferase activities were normalized against that of vector-transfected cells without stimulation. Shown is a representative result out of two experiments.
Btk was shown to critically regulate JNK1/2 and, to a lesser extent, p38, while it did not affect ERKs significantly (25). Targets of JNK include the transcription factors, such as c-Jun, Elk-1, and Sap-1a. Therefore, we tested whether activities of these transcription factors are under the control of Btk in mast cells. To this end, btk null-BMMC stably transfected with the empty vector, WT btk, or K430R btk cDNAs were transiently co-transfected with a luciferase reporter plasmid containing five repeats of GAL4 binding sites (pG5E1bLuc) and GAL4 DNA binding domain expression vectors fused with activation domains of these transcription factors. Fc⑀RI-induced luciferase expression by a GAL4-c-Jun construct in WT btk transfectants was at least 2.5-5-fold higher than those in vector or K430R btk transfectants (Fig. 3A). Substitutions of serine residues at positions 63 and 73, the critical phosphorylation sites by JNK (29,30), with alanine in a GAL4-c-Jun construct (GAL4-c-Jun(S63/73A)) abolished the activation of c-Jun, sug-gesting that this c-Jun activation is mediated by phosphorylation at these serine residues by JNK. Therefore, we analyzed the phosphorylation status of c-Jun in WT and btk null BMMC by immunoblotting with antibody specific for the c-Jun peptide with phosphorylation at Ser-63. As shown in Fig. 3B, phosphorylation at Ser-63 was induced in Fc⑀RI-stimulated WT BMMC in a time-dependent manner, whereas the same phosphorylation was drastically reduced in btk null BMMC. The presence of comparable amounts of c-Jun protein in both BMMC was confirmed by immunoblotting with anti-c-Jun antibody. These data demonstrate that Btk regulates the activity of JNK and thereby the phosphorylation and activity of c-Jun.
Similar to GAL4-c-Jun, activities of GAL4-JunB, GAL4-JunD, GAL4-Elk-1, and GAL4-Sap-1a were 2-3-fold higher in Fc⑀RI-stimulated WT btk transfected cells than those in vector or kinase-dead btk transfected cells (Fig. 3, C and D, and data  not shown). Again similar to c-Jun, activities of phosphoryla- tion site mutants of Elk-1 and Sap-1a constructs were suppressed in either cells. These results are consistent with the finding in other studies that these transcription factors are phosphorylated and activated by JNK. Activation of MEF2C, which was recently shown to be a target of p38 (68), in WT btk transfectants was 2.5-fold higher than activation in vector or K430R btk transfectants (data not shown), consistent with our previous finding that Btk also regulates p38, albeit to a lesser extent than JNK (25).
In contrast, Fos activity may not be regulated by Btk, since 1) ERK1/2 activity, a major component of the c-fos activating pathway, is not significantly affected by btk mutations (25) and 2) transcription of a luciferase reporter construct containing a typical AP-1 element derived from the human collagenase gene promoter was not significantly affected by btk mutations (data not shown). Transcriptional activities of c-fos-Luc were slightly enhanced by Fc⑀RI cross-linking in btk null BMMC, but they were not affected by expression of vector, WT Btk, or K430R Btk (data not shown).
The SEK1-JNK-c-Jun Pathway Is Crucial for Fc⑀RI-induced, Btk-dependent IL-2 Gene Transcription in Mast Cells-Upstream of c-Jun, various stimuli lead to JNK activation through a cascade of protein kinases: MEKK1 3 SEK1 3 JNK (26 -28). Therefore, we examined whether c-Jun and this cascade are used to activate the IL-2 gene. First, the effects of a c-Jun dominant negative mutant were tested on IL-2 promoter activity. A c-Jun mutant, Jun in 282 protein, which lacks the DNA binding ability and acts as a dominant negative protein in Bcr-Abl-induced transformation (61), was transiently expressed in WT BMMC. Jun in 282 protein was expressed ϳ2fold more than the endogenous c-Jun protein in the transfected WT BMMC (Fig. 4A, inset). Fc⑀RI-induced transcriptional activation of the IL-2Luc reporter in WT BMMC was suppressed by ϳ60% by Jun in 282 protein compared with the vector transfected cells under our conditions (Fig. 4A). btk null BMMC were also transfected with control or Jun in 282 expression vectors together with btk expression and IL-2Luc plasmids. Jun in 282 protein almost completely abolished the enhancing effect of Btk on Fc⑀RI-induced IL-2 promoter activity (Fig. 4B). These data clearly indicate that c-Jun plays an important role in IL-2 gene transcription in mast cells and that Btk regulates the activity of c-Jun.
SEK1 is a well characterized direct JNK activator, and many stimuli activate JNK via SEK1 (37). Since a kinase-dead mutant, SEK1(K/R), works as a dominant negative mutant, this mutant was transiently expressed together with the IL-2Luc plasmid in WT or btk null BMMC. Expression of glutathione S-transferase-SEK1(K/R) fusion protein was confirmed by immunoblotting (Fig. 4C, inset). Glutathione S-transferase-SEK1(K/R) suppressed IL-2 promoter activity by ϳ60% in Fc⑀RI-stimulated WT BMMC (Fig. 4C). In btk null BMMC, glutathione S-transferase-SEK1(K/R) also suppressed the po-tentiating effect of Btk on IL-2 promoter activity upon Fc⑀RI cross-linking (Fig. 4D). These data collectively demonstrate that SEK1 is also involved in Btk-mediated IL-2 gene transcription in mast cells.
MEKK1 is a representative activator of SEK1, although recent studies identified several other protein kinases as SEK1 activators (69 -76). A constitutively active MEKK1 mutant, ⌬MEKK1, was co-expressed with the IL-2Luc plasmid in btk null BMMC. Compared with the vector control, ⌬MEKK1 induced a very strong activation of IL-2 promoter without Fc⑀RI stimulation, and Fc⑀RI cross-linking further stimulated IL-2 promoter activity (Fig. 4E). Although these data do not prove that MEKK1 is involved in Btk-mediated IL-2 gene activation, they are certainly consistent with the notion that IL-2 gene activation by Fc⑀RI cross-linking involves SEK1 and JNK, downstream of Btk and upstream of c-Jun and other transcription factors.

DISCUSSION
Production of cytokines is regulated at various levels, e.g. mRNA stability, posttranslational modification, and processing of precursor proteins. However, transcriptional activation is a major regulatory mechanism generally adopted by numerous cytokine genes. Similar to T cells, mast cells seem to use this strategy in response to Fc⑀RI cross-linking. As shown in Fig. 1, most of the known cis-elements that are binding sites for transcription factors are required for maximal activation of the IL-2 gene. Since mutations in any one of these elements lead to drastic decreases in IL-2 gene expression, individual transcription factors cooperate for maximal gene activation, a finding also consistent with previous studies on T cells. Among the cis-elements tested, mutations in the two AP-1 (NF-ATd/AP-1 and AP-1/Octp) and two NFAT (NF-ATd and NF-ATp) sites affected the gene expression most severely in mast cells. On the other hand, transcriptional activities of the cis-elements containing the NF-ATd/AP-1 and AP-1/Octp regions were shown to be regulated by Fc⑀RI cross-linking in a Btk-dependent manner. These results suggest that Btk regulates activities of AP-1 and/or NF-AT transcription factors. AP-1 is a complex of Jun and Fos family members. Our present studies have demonstrated that activities of c-Jun and other JNK-activable transcription factors, i.e. JunB, JunD, Elk-1, and Sap-1a, are regulated by Btk. Jun and Fos family members associate in various combinations and with other transcription factors, including NF-AT (reviewed in Ref. 77). Therefore, one possible explanation for the Btk dependence of transcriptional activities of NF-ATd/AP-1-Luc and AP-1/Octp-Luc in mast cells is that the activity of the transcription factors bound to the AP-1 sites, but not those of the factors bound to the NF-ATd or Octp sites, is dependent on Btk. Although we have not studied NF-AT or Oct as a candidate Btk-regulated transcription factor in further detail, this possibility certainly warrants further investigation. Luciferase activities are shown relative to that in unstimulated vector (pME18S and pSR␣MSVtkneo)-transfected cells. C, WT BMMC were co-transfected with 8 g of IL-2Luc and 40 g of either the empty vector (pEBG) or pEBG-SEK1(K/R). -Fold luciferase activities are shown after normalization against that in unstimulated vector-transfected cells. Expression of endogenous SEK1 and glutathione S-transferase-SEK1 was detected by immunoblotting of total cell lysates with anti-MEK4 (inset). D, btk null BMMC were co-transfected with the indicated combinations of 8 g of IL-2Luc, 20 g of either pME18S or pME-btk, and 40 g of either pEBG or pEBG-SEK1(K/R) plasmids. Luciferase activities are shown relative to that in unstimulated vector (pME18S and pEBG)-transfected cells. E, btk null BMMC were co-transfected with 8 g of IL-2Luc and 40 g of either the empty vector (pCMV5) or ⌬MEKK1 expression vector. Luciferase activities are shown relative to that in unstimulated vector (pCMV5)-transfected cells. ⌬MEKK1 was detected by immunoblotting with anti-EE antibody (inset). A-E, representative results are shown out of two identical experiments.
Btk regulates the enzymatic activities of JNK and p38, but not ERK (25). At least some stimuli use the cascade of protein kinases, MEKK1 3 SEK1 3 JNK1/2 (28). Mast cells seem to use this pathway to transcriptionally activate the IL-2 gene upon Fc⑀RI stimulation. Thus, dominant negative mutants of SEK1 and c-Jun inhibited the Fc⑀RI-induced IL-2 gene expression, while a constitutively active MEKK1 mutant activated the transcription of the IL-2 gene. However, our study does not rule out the possibility that a protein kinase(s) other than MEKK1 might be activated by Fc⑀RI cross-linking to activate SEK1 and JNK. There are several reports on other protein kinases with SEK1-activating ability (69 -76). Furthermore, there are several chromatographically separable activities that activate JNK (78). One such kinase, MKK7, was recently cloned (79,80). This and other JNK kinases might be activated by Fc⑀RI cross-linking and regulated by Btk. These possibilities should be explored in future work.
Btk was shown to be activated through phosphorylation of Tyr-551 by Lyn and other Src family protein-tyrosine kinases (81,82). Activated Btk then autophosphorylates at Tyr-231 in heterologous (83) and mast cells (84) and phosphorylates target proteins such as phospholipase C-␥ (85). On the other hand, the enzymatic activity of Btk is negatively regulated by phosphorylation by protein kinase C (51). However, it is not known how Btk activation leads to JNK activation. One possible scenario is that phospholipase C-␥ activated by Btk results in an increase in [Ca 2ϩ ] i , which then leads to activation of Pyk2, a Ca 2ϩsensitive protein-tyrosine kinase that activates JNK in response to stresses (86). Another possible route to JNK activation is through Ras activation. Indeed, this pathway is shown to be used when pro-B cells were stimulated by IL-3. 3 Alternatively, Btk may phosphoarylate and activate Vav, a tyrosine phosphorylation-dependent guanine nucleotide exchange factor for Rac1 (87), and activated GTP-bound Rac1 leads to the activation of JNK (19). Experiments to test these possibilities are under way.
In conclusion, the current data demonstrate that Btk regulates the transcription of the IL-2 gene via the JNK activation pathway upon Fc⑀RI cross-linking. Since btk mutant mast cells have defective secretion of several other cytokines, i.e. TNF-␣, IL-6, and granulocyte-macrophage colony-stimulating factor, c-Jun and other JNK-regulated transcription factors also seem to activate the transcription of these cytokine genes in a Btkdependent manner. These proinflammatory cytokines are implicated in allergic and other inflammatory diseases.