The IL-7R plays an essential role in γδ T cell development by inducing V-J recombination of the TCRγ locus through STAT5. Although tyrosine residues in the intracellular domain of the mouse IL-7R α-chain (IL-7Rα) have been implicated in STAT5 activation, it is still unknown whether they are essential for γδ T cell development. In this study, we showed that those IL-7Rα tyrosine residues are not essential for γδ T cell development, because phenylalanine replacement of four intracellular tyrosine residues (IL-7R-FFFF) partially rescued γδ T cell development of IL-7Rα−/− progenitors. To examine signaling pathways activated by IL-7R-FFFF, we introduced a chimeric receptor consisting of the human IL-4R α-chain and mouse IL-7R-FFFF (4R/7R-FFFF) into an IL-7-dependent pre-B cell line and found that 4R/7R-FFFF induced TCRγ germline transcription and STAT5 activation. Treatment of cells with MEK1/2 inhibitors significantly decreased levels of TCRγ germline transcription and STAT5 tyrosine phosphorylation mediated by 4R/7R-FFFF, suggesting that MEK1/2 plays an alternative role in STAT5 activation by IL-7R. MEK1/2 associated with STAT5 and induced STAT5 tyrosine phosphorylation and DNA binding activity. Furthermore, MEK1 directly phosphorylated a STAT5 tyrosine residue in vitro. Finally, active MEK1 partially rescued TCRγ germline transcription by IL-7R in a pre-T cell line. These results demonstrate that MEK1/2 induces TCRγ germline transcription by phosphorylating STAT5 through IL-7R-FFFF and suggest a potential role for MAPK in IL-7R tyrosine-independent activation of STAT5.
Interleukin-7 plays important roles in the growth and differentiation of lymphocytes. IL-7R consists of two polypeptides, a unique IL-7R α-chain (IL-7Rα)3 and a common γ-chain. IL-7−/− and IL-7Rα−/− mice exhibit severely impaired expansion of early lymphocytes (1, 2). Notably, while IL-7Rα−/− mice have fewer B cells and αβ T cells in the periphery, they totally lack γδ T cells (3, 4). IL-7 binding to IL-7R triggers phosphorylation and activation of receptor-associated JAK1 and JAK3 tyrosine kinases (5) and, following their activation, JAK kinases phosphorylate IL-7Rα Tyr449. STAT and PI3K are then recruited to the phosphorylated Tyr449 residue and subsequently activated by the JAK kinases (6, 7). In addition, IL-7R activates the src family protein tyrosine kinases MAPK and Pyk2 (8, 9, 10, 11, 12). Among STAT proteins, IL-7R mainly activates STAT5 and, to a lesser extent, STAT1 and STAT3 (13). IL-7R transmits at least two types of signal in T and B cells. One is for survival and proliferation. The IL-7R supports cell survival by Bcl-2 induction (14) and cell proliferation by PI3K activation (15, 16). The other is for V(D)J recombination. The IL-7R controls the chromatin accessibility of the IgH and TCRγ loci (17, 18, 19, 20, 21).
STAT5 plays essential roles in signal transduction of many cytokines and growth factors, including IL-7 (22). The mouse genome exhibits two highly conserved STAT5 proteins, STAT5a and STAT5b. Relevant to lymphocyte development, STAT5 deficiency results in a severe combined immunodeficiency phenotype similar to that seen in mice lacking IL-7Rα, JAK3, or the common γ-chain (23). Notably, STAT5a/b−/− as well as IL-7Rα−/− mice completely lack TCRγ rearrangement and γδ T cells. Following IL-7 stimulation, activated STAT5 interacts with consensus motifs in Jγ germline promoters (19). Moreover, constitutively active (CA) STAT5 induces germline transcription and V-J recombination at the TCRγ locus and rescues γδ T cell development of IL-7Rα−/− progenitors. These results demonstrate that STAT5 controls chromatin accessibility of the TCRγ locus.
There are four tyrosine residues in the IL-7Rα intracellular domain, three of which are conserved between human and mouse. Among them, Tyr449 plays a crucial role in proliferation and survival of lymphocyte precursors by triggering STAT5 and PI3K activation (6, 7). However, there are conflicting reports as to whether the Tyr449 is required for TCRγ recombination and γδ T cell development (14, 24). Recently, it was reported that IL-7Rα-Phe449 knock-in mice exhibit reduced but detectable levels of γδ T cells (25). Therefore, it is still enigmatic whether STAT5 is activated by an Tyr449-independent signal.
To examine tyrosine-independent signaling of the IL-7Rα, we dissected signal pathways responsible for STAT5 activation and TCRγ germline transcription in cells expressing mutant IL-7Rα with four tyrosine residues in the intracellular domain replaced with four phenylalanines (FFFF) (IL-7R-FFFF). We found that IL-7Rα tyrosine residues are not essential for γδ T cell development and that MEK1/2 plays an alternative role in STAT5 activation by IL-7R. Furthermore, MEK1/2 directly associated with STAT5 and phosphorylated a critical tyrosine residue. These observations indicate an IL-7R tyrosine-independent STAT5 activation pathway by MAPK in IL-7R signaling.
Materials and Methods
IL-7Rα−/− mice (4) were bred and maintained under specific pathogen-free conditions in our animal facility. Embryonic day 0 was defined by the vaginal plug. Experiments were performed according to guidelines for animal use and experimentation by Kyoto University (Kyoto, Japan).
An IL-7-dependent mouse pre-B cell line, preBR1 (26) (a gift from Dr. A. Kudo at Tokyo Institute of Technology, Tokyo, Japan) and its transfectants were cultured in RPMI 1640 medium containing 50 μM 2-ME, 1× nonessential amino acids, 10% FBS, and 10% conditioned medium of J558-mouse IL-7 cells as an IL-7 source (27). A mouse pre-T cell line, Scid.adh-TAC:CD3ε (28) (a gift from Dr. D. L. Wiest at Fox Chase Cancer Center, Philadelphia, PA), was cultured in RPMI 1640 medium containing 50 μM 2-ME, 1× nonessential amino acids, 1 mM sodium pyruvate, 2 mM glutamine, and 10% FBS. The retroviral packaging cell line Plat-E (29) (a gift from Dr. T. Kitamura at University of Tokyo, Tokyo, Japan) and HEK293T cells were maintained in DMEM containing 10% FBS.
Abs and flow cytometry
PE-conjugated mAbs specific to TCRγδ (clone GL3) and CD4 (clone GK1.5), together with allophycocyanin-conjugated mAbs specific to TCRβ (clone H57-597) and CD8α (clone 53.6.7) were purchased from eBioscience. The phosphorylated STAT5 mAb was from Upstate Biotechnology. PE-conjugated anti-human IL-4R α-chain (IL-4Rα) mAb and antisera against STAT5 were from R&D Systems. Antisera against MEK1/2 and phosphorylated MEK1/2 (Ser 217/221) were from Cell Signaling Technology. Antisera against GST and MEK1 were from GE Healthcare Bio-Science and Santa Cruz Biotechnology, respectively. The hemagglutinin (HA) (clone 3F10) mAb was from Roche. Erythrocyte-reactive mAb (TER119) (30) was provided by Dr. T. Kina of Kyoto University.
Flow cytometric analysis was done as described (19). Viable cells were analyzed by a FACSCalibur device with CellQuest software version 3.1 (BD Biosciences). Debris and dead cells were excluded from the analysis by forward and side scatter and propidium iodide gating.
Plasmid constructs and retrovirus production
cDNAs encoding IL-7R-FFFF, CA-MEK2 (31), and STAT5 mutants (STAT5a-Y694F, STAT5a-S725A, STAT5a-S779A, STAT5a-S725/779A and STAT5b-S730A) were generated using a QuikChange site-directed mutagenesis kit (Stratagene).
Wild-type (WT) mouse STAT5a and CA-STAT5a (32) (gifts from Dr. T. Kitamura at University of Tokyo), HA-tagged MEK1 and HA-tagged CA-MEK1 (31) (gifts from Dr. N. G. Ahn at University of Colorado, Boulder, CO), CA-MEK2, CA-MEK5 (33), and CA-MEK6 (gifts from Dr. E. Nishida at Kyoto University), and CA-MEK7 (34) (a gift from Dr. M. Kracht at Hannover Medical School, Hannover, Germany) were subcloned into the pCAGGS expression vector (35). To prepare GST-STAT5 fusion proteins, STAT5a cDNA encoding amino acid residues 494–793 was subcloned by PCR into pGEX-4T-3 vector (GE Healthcare Bio-Science).
cDNAs encoding full-length IL-7Rα WT (IL-7R-WT) and truncated IL-7Rα (IL-7R-311 and IL-7R-328) were generated and subcloned by PCR into the pMX-IG retrovirus vector (36). For chimeric human IL-4Rα and mouse IL-7Rα (4R/7R) receptors, we replaced the region encoding the extracellular domain of mouse IL-7Rα cDNA with that of human IL-4Rα. cDNAs encoding 4R/7R and STAT5 mutants were subcloned into the pMXs-IP retrovirus vector (36). cDNAs encoding HA-tagged CA-MEK1 and a dominant negative (DN) MEK1 (37) (gifts from Dr. N. G. Ahn at University of Colorado) were subcloned into the pMXs-IB vector (36). High-titer retrovirus was obtained with the Plat-E packaging cell line as described (29).
Infection of fetal liver cells and hanging drop (HD)-fetal thymic organ culture (FTOC)
Single cell suspensions were prepared from embryonic day 14.5 fetal liver of IL-7Rα−/− mice as described (38). Cells were incubated with TER119 Ab for 30 min at 4°C, and TER119+ cells were then removed by magnetic separation with BioMag goat anti-rat IgG (Qiagen). TER119− cells were infected with retrovirus as described (19). Cells were cultured with 2′-deoxyguanosine-treated, embryonic day 15.5 fetal thymic lobes in HDs for 24 h and then organ cultured as described (38).
Total RNA was isolated with TRIzol reagent (Invitrogen). Random-primed cDNA was quantified in triplicate using an ABI 7500 sequence detector (Applied Biosystems) with TaqMan ribosomal RNA control reagent VIC probe (Applied Biosystems) and SYBR Green Premix Ex Taq (Takara). PCR was conducted for an initial 10 s at 95°C and then 40 cycles at 95°C for 15 s and 60°C for 1 min. cDNA levels were normalized to 18S rRNA. Primer sequences for Jγ4-Cγ4 germline and Vγ2-Jγ1 transcripts were: as follows 5′Cγ4, 5′-GACAAACGCACTGACTCAGACT-3′; 3′Cγ4, 5′-GGATTCCAGAATCTTTTCACCATC-3′; 5′Vγ2, 5′-GTAACCATACACTGGTACCG-3′; and 3′Jγ1, 5′-AGAGGGAATTACTATGAGCT-3′. Primer sequences used for Jγ1-Cγ1 germline transcripts and Vγ3-Jγ1 transcripts were as described (19, 38).
Immunoprecipitation and Western blotting
Cells were lysed in lysis buffer (50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 0.1% SDS, 1% Nonidet P-40, 0.5% sodium deoxycholate, 100 mM NaF, 1 μM Na3VO4, 1 μg/ml aprotinin, 1 μg/ml leupeptin, 10 μg/ml soybean trypsin inhibitor, and 1 μg/ml α1-antitrypsin). Protein concentration was determined using a BCA protein assay kit (Pierce) standardized with BSA. For immunoprecipitation, equal protein amounts were incubated with STAT5- or MEK1-specific Abs at 4°C for 2 h and immunoprecipitated with protein A-Sepharose beads (GE Healthcare Bio-Science). Both whole cell lysates and immunoprecipitates were subjected to electrophoresis on 8% SDS-polyacrylamide gels and transferred to polyvinylidene difluoride membranes (Immobilon-P; Millipore). Chemiluminescence of blots was developed using the ECL Plus detection system (GE Healthcare Bio-Science) and detected by the LAS-3000mini image analyzer (Fujifilm).
In vitro kinase assay
GST-STAT5 fusion protein was expressed in Escherichia coli BL21-CodonPlus (Stratagene) and purified using glutathione-Sepharose (GE Healthcare Bio-Science). Beads were resuspended in kinase buffer (20 mM MOPS (pH 7.2), 25 mM β-glycerol phosphate, 5 mM EGTA, 1 mM Na3VO4, 1 mM DTT, 25 mM MgCl2, and 100 μM ATP), mixed with 0.5 ng of recombinant active MEK1 (Upstate Biotechnology), and incubated at 30°C for 30 min. Samples were immunoblotted with Abs specific to phosphorylated STAT5 or GST.
HEK293T cells were transfected with pCAGGS-STAT5a and pCAGGS-CA-MEK1. Forty-eight hours later, nuclear extracts were prepared and protein levels were quantified using a BCA protein assay kit. To assess STAT5 binding, 10 μg of nuclear extract was incubated with biotin-labeled STAT5 oligonucleotide probe as described (19). Samples were electrophoresed on 4% polyacrylamide gels, DNA-protein complexes were visualized by a chemiluminescent nucleic acid detection module (Pierce) and analyzed by the image analyzer LAS-3000mini.
Treatment of cells with human IL-4 and MEK1/2 inhibitors
Transfectants were washed twice with the cytokine-free medium and cultured in medium containing 1 μg/ml recombinant human IL-4 (hIL-4) (Genzyme). PreBR1 cells were incubated with various concentrations of the MEK1/2 inhibitor U0126, an inactive analog U0124 (both from Calbiochem), or the MEK1/2 inhibitor PD184352 (Toronto Research Chemicals) for 24 h. Scid.adh-TAC:CD3ε cells were treated with U0126 for 3 h. In the HD-FTOC experiment, treatment with U0126 (10 μM) was initiated after FTOC.
Expression of IL-7Rα mutants rescues γδ T cell development of IL-7Rα−/− progenitors
The IL-7Rα intracellular domain exhibits motifs such as the Box 1, an acidic region, and a serine-rich region (8, 39). The mouse IL-7Rα intracellular domain exhibits tyrosine residues at positions 390, 401, 449, and 456. To identify regions important for γδ T cell development, we constructed IL-7Rα cDNAs with deletions of residues 329–459 (IL-7R-328) and 312–459 (IL-7R-311) or point mutations replacing the four tyrosine residues with phenylalanines (IL-7R-FFFF) (Fig. 1,A) and introduced WT and mutant cDNAs into T cell progenitors from IL-7Rα−/− fetal liver using pMX-IG retroviral vectors. To evaluate IL-7Rα expression levels, we checked internal ribosomal entry site-mediated GFP expression by flow cytometry. Of the surviving cells, 77–85% were GFP+ and GFP was expressed at similar levels in WT and mutant cells (Fig. 1,B). After infection, IL-7Rα−/− T progenitors were cultured by HD-FTOC. As previously described (19), IL-7R-WT expression significantly increased the number of thymocytes (Fig. 1,C). By contrast, IL-7R-FFFF or IL-7R-328 expression caused a moderate increase in thymocyte number, but only a few thymocytes were recovered from IL-7R-311-reconstituted lobes. Organ-cultured thymocytes were stained with anti-TCRαβ and anti-TCRγδ or anti-CD4 and anti-CD8 Abs and analyzed by flow cytometry. Although IL-7R-311 failed to support αβ and γδ T cell development, both IL-7R-FFFF and IL-7R-328 induced differentiation of αβ and γδ T cells from IL-7Rα−/− T cell progenitors and, in both cases, the pattern of CD4 and CD8 expression was similar to that seen in adult thymus (Fig. 1 D). However, the decreased numbers of αβ and γδ T cells seen in IL-7R-FFFF- or IL-7R-328-reconstituted lobes indicate that IL-7R-FFFF and IL-7R-328 only partially rescue αβ and γδ T cell development in the absence of the WT protein. These findings suggest that tyrosine residues in the IL-7Rα intracellular domain function in the proliferation and differentiation of αβ and γδ T cells but are not essential for TCRγ rearrangement. They also indicate that a region between aa 311 and 328, which includes first half of the acidic region (8), delivers a critical signal for both αβ and γδ T cell development.
Tyrosine residues in the IL-7Rα intracellular domain are not essential for TCRγ germline transcription and STAT5 phosphorylation
Tyrosine residues have been implicated in STAT5 activation by IL-7R (5). Previously, we showed that IL-7R induces germline transcription and V-J recombination at the TCRγ locus (17, 19, 21). Because Vγ-Jγ recombination frequently takes place not only in γδ T cells but also in αβ T cells, it is difficult to examine germline transcription of the TCRγ locus using T cell lines. Because the TCRγ locus is maintained in a germline configuration in B cells, we first asked whether IL-7 induces Jγ-Cγ germline transcription in an IL-7-dependent pre-B cell line, preBR1. RT-PCR analysis indicated that preBR1 cells expressed high levels of Jγ1-Cγ1 germline transcripts in the presence of IL-7 (data not shown). To define molecular mechanisms underlying germline transcription mediated by IL-7R, we generated a series of 4R/7R chimeric receptors bearing the extracellular domain of human IL-4Rα and the transmembrane and intracellular domains of the mouse IL-7Rα mutants (Fig. 2,A) and introduced them into preBR1 cells by retroviral infection. All transfected preBR1 cells showed similar levels of receptor expression based on flow cytometric analysis using anti-human IL-4Rα Ab (Fig. 2,B). Because human IL-4R responds in a species-specific manner to hIL-4, we stimulated transfectants with hIL-4 and measured Jγ-Cγ germline transcripts by real-time RT-PCR (Fig. 2 C). In the presence of hIL-4, 4R/7R-WT induced high levels of germline transcripts equivalent to endogenous IL-7R stimulated with mouse IL-7 (data not shown). Expression of the mutant forms 4R/7R-FFFF and 4R/7R-328 reduced levels of germline transcripts 60-fold, whereas 4R/7R-311 expression did not activate transcription at all. These data indicate that not only the tyrosine residues but also the region between Lys312 and Leu328 transduces a signal required for TCRγ germline transcription.
We previously reported that IL-7R signaling induces TCRγ germline transcription by activating STAT5 (19). Thus, we asked whether 4R/7R-FFFF activates STAT5 in preBR1 cells. As shown in Fig. 2,D, hIL-4 induced substantial levels of STAT5 phosphorylation by 4R/7R-WT in preBR1 cells. Weak but reproducible levels of STAT5 phosphorylation were detected in 4R/7R-FFFF and 4R/7R-328 transfectants, whereas 4R/7R-311 failed to induce phosphorylation. These results indicate that tyrosine residues in IL-7Rα are not essential for STAT5 phosphorylation and that the region between Lys312 and Leu328 transmits a weak signal activating STAT5. To examine the kinetics of STAT5 activation by an IL-7R tyrosine-independent signal, we stimulated preBR1 cells expressing 4R/7R-WT or 4R/7R-FFFF with hIL-4 for various time periods. In 4R/7R-WT transfectants, levels of the phosphorylated STAT5 peaked after 30 min and then decreased to barely detectable levels after 60 min (Fig. 2,E). Although 4R/7R-FFFF-expressing cells showed a similar pattern, levels of STAT5 phosphorylation were always lower than those seen in 4R/7R-WT cells, suggesting that the IL-7R-FFFF signal induces a low but detectable level of STAT5 activation. Because STAT5 partially restores αβ and γδ T cell development of IL-7Rα−/− T cell precursors (19), the ability of an IL-7R tyrosine-independent signal to activate STAT5 supports our previous results showing that tyrosine residues in the IL-7Rα intracellular domain are not essential for TCRγ rearrangement (Fig. 1, C and D).
MEK1/2 functions in TCRγ germline transcription mediated by IL-7R-FFFF
Our results suggested that IL-7R-FFFF delivers a signal for γδ T cell development. Because the results with preBR1 cells were consistent with those seen in IL-7Rα−/− progenitors, we used preBR1 cells to dissect 4R/7R-FFFF signaling. It is reported that IL-7R activates the src family tyrosine kinases, p59fyn and p56lck (8, 9, 40) and that p59fyn associates with a region of the intracellular domain of IL-7Rα between Lys312 and Leu328 (8). To evaluate the role of p59fyn and p56lck in TCRγ germline transcription mediated by 4R/7R-FFFF, we introduced DN forms of both proteins into preBR1 cells. These forms did not inhibit transcription mediated by hIL-4, suggesting that p59fyn and p56lck do not function in TCRγ germline transcription by 4R/7R-FFFF (data not shown).
Because IL-7R activates ERK in B and T cells (10, 41, 42), we asked whether the MEK1/2-ERK pathway is involved in TCRγ germline transcription by 4R/7R-FFFF by stimulating preBR1 cells expressing 4R/7R-FFFF with mouse IL-7 or hIL-4 and undertaking Western blot analysis with anti-phosphorylated MEK1/2 Ab. As shown in Fig. 3,A, IL-7 stimulation induced high levels of serine phosphorylation of MEK1/2 through the endogenous IL-7R. Notably, hIL-4 stimulated intermediate levels of MEK1/2 phosphorylation, suggesting that 4R/7R-FFFF activates the MEK1/2-ERK pathway. We then stimulated preBR1 cells expressing 4R/7R-FFFF with hIL-4 in the presence of the MEK1/2 inhibitor U0126 or its inactive analog, U0124, and measured germline transcripts by real-time RT-PCR. U0126 treatment induced a dose-dependent decrease in germline transcripts while U0124 had no effect (Fig. 3,B). Furthermore, treatment with a structurally distinct MEK1/2 inhibitor, PD184352, resulted in a marked reduction in germline transcripts mediated by IL-7R-FFFF (Fig. 3,C). However, neither U0126 nor PD184352 had a significant effect on the viability of preBR1 cells expressing 4R/7R-FFFF (Fig. 3,D, and data not shown). To confirm that MEK1/2 activation is indeed critical to induce TCRγ germline transcription, preBR1 cells expressing 4R/7R-FFFF were transfected with DN-MEK1 and the germline transcription was monitored. Expression of DN-MEK1 suppressed by nearly 20% the levels of germline transcripts mediated by IL-7R-FFFF (Fig. 3 E), suggesting that MEK1/2 functions in TCRγ germline transcription mediated by this form of the receptor.
Serine phosphorylation of STAT5 is dispensable for the TCRγ germline transcription
The serine residues of human STAT5a, Ser726 (mouse Ser725) and human STAT5b Ser731 (mouse Ser730), are phosphorylated by IL-7 stimulation (43), and human STAT5a Ser780 (mouse Ser779) is phosphorylated by ERK (44). To investigate whether STAT5 serine phosphorylation is required for TCRγ germline transcription by IL-7R-FFFF, we introduced STAT5 serine to alanine mutants (STAT5a-S725A, STAT5a-S779A, STAT5a-S725/779A, and STAT5b-S730A) into 4R/7R-FFFF-expressing preBR1 cells and monitored germline transcripts by real-time PCR. Expression of these STAT5 mutants had no effect on cell viability (data not shown). Furthermore, rather than inhibiting, these mutants enhanced germline transcription by 4R/7R-FFFF (Fig. 3 F). These results suggest that serine phosphorylation of STAT5 is not essential for the TCRγ germline transcription by IL-7R-FFFF.
MEK1/2 induces STAT5 tyrosine phosphorylation
Because IL-7R activates both MAPK and STAT5, we asked whether treatment with the MEK1/2 inhibitor U0126 affects tyrosine phosphorylation of STAT5 by IL-7R. We stimulated preBR1 cells expressing 4R/7R-FFFF with hIL-4 in the presence of U0126 and measured STAT5 tyrosine phosphorylation by Western blot analysis. Compared with untreated controls, hIL-4 stimulation significantly increased levels of STAT5 tyrosine phosphorylation, which was inhibited by U0126 pretreatment (Fig. 3 G).
To determine whether MEK1/2 induces STAT5 tyrosine phosphorylation, we introduced STAT5 and CA-MEK expression vectors into HEK293T cells and analyzed STAT5 tyrosine phosphorylation by Western analysis. Tyrosine phosphorylation was markedly enhanced in HEK293T cells expressing CA-MEK1 and CA-MEK2 (Fig. 4 A). Although U0126 also inhibits MEK5 (33), CA-MEK5 did not induce STAT5 tyrosine phosphorylation.
To determine whether STAT5 interacts with MEK in vivo, we conducted coimmunoprecipitation experiments using HEK293T cells transfected with STAT5 and HA-MEK1 expression vectors. Cell lysates were immunoprecipitated with anti-MEK1 Ab, followed by immunoblotting with anti-STAT5 Ab. STAT5 was coimmunoprecipitated with MEK1 (Fig. 4,B, left column, top). In the absence of HA-MEK1 expression, STAT5 was not immunoprecipitated. Association of STAT5 with HA-MEK1 was further examined in a reverse experiment in which lysates were immunoprecipitated with anti-STAT5 Ab followed by immunoblotting with anti-HA Ab (Fig. 4 B, right column, top). Significant amounts of MEK1 were coimmunoprecipitated with STAT5, suggesting that MEK1 specifically interacts with STAT5 in vivo.
MEK1/2 could indirectly induce STAT5 tyrosine phosphorylation in HEK293T cells by activating endogenous protein tyrosine kinases. To determine whether STAT5 is a direct MEK1/2 substrate, we examined phosphorylation of STAT5 by purified MEK1 in vitro. For this purpose, GST fusions of WT (GST-STAT5a) or mutant (GST-STAT5a-Y694F) STAT5a were incubated with purified active MEK1. As shown in Fig. 4,C, active MEK1 phosphorylated tyrosine residues of GST-STAT5a but not GST-STAT5a-Y694F. These results are consistent with in vivo results (Fig. 4 A) and indicate that MEK1 directly phosphorylates a tyrosine residue of STAT5.
DNA binding activity of STAT5 is induced by active MEK1
The tyrosine phosphorylation at residue 694 of mouse STAT5 regulates its DNA binding activity. To determine whether MEK1 induces DNA binding activity of STAT5, we transfected HEK293T cells with STAT5 and CA-MEK1 expression vectors and assayed nuclear extracts by EMSA using an oligonucleotide probe containing a STAT consensus motif in the mouse Jγ1 germline promoter. DNA binding activity was clearly detected when CA-MEK1 was cotransfected, whereas transfection of STAT5 alone induced no binding activity (Fig. 4 D). Mutation in the STAT motif abrogated this DNA binding activity.
MEK1 induces TCRγ germline transcription in T cells
Next, we asked whether MEK1/2 functions in V-J recombination of the TCRγ locus in T cells. We first determined whether IL-7 activates STAT5 in a pre-T cell line, Scid.adh. As shown in Fig. 5,A, IL-7 stimulation induced STAT5 tyrosine phosphorylation in Scid.adh cells. Next, we analyzed the configuration of the TCRγ locus. Although the γ1, 2, and 3 clusters were deleted, the γ4 cluster was intact in Scid.adh cells (data not shown). To test whether IL-7R induces TCRγ germline transcription in Scid.adh cells, we examined levels of Jγ4-Cγ4 germline transcripts by real-time RT-PCR. Scid.adh cells showed significant levels of germline transcripts when cultured with IL-7 (Fig. 5 B). Active STAT5 alone induced low but detectable levels of the germline transcripts in the absence of IL-7.
To investigate a potential role for MEK1 in TCRγ germline transcription, we introduced either 4R/7R-FFFF or CA-MEK1 cDNA into Scid.adh cells using a retroviral vector. As shown in Fig. 5,C, hIL-4 stimulation induced an increase in Jγ4-Cγ4 germline transcripts in Scid.adh cells expressing 4R/7R-FFFF. Interestingly, CA-MEK1 increased TCRγ germline transcripts to levels similar to those seen with 4R/7R-FFFF in the absence of hIL-4. Furthermore, U0126 suppressed TCRγ germline transcription by IL-7R-FFFF dose-dependently in Scid.adh cells (Fig. 5,D). Finally, to confirm regulation of TCRγ germline transcription by MEK1, DN-MEK1 cDNA was introduced into Scid.adh cells expressing 4R/7R-FFFF. As shown in Fig. 5, E and F, DN-MEK1 significantly suppressed levels of Jγ4-Cγ4 germline transcripts induced by 4R/7R-FFFF signaling, although it did not alter cell viability. Taken together, these results suggest that active MEK1 induces TCRγ germline transcription in T cells.
A MEK1/2 inhibitor blocks γδ T cell development by IL-7R-FFFF
Finally, we asked whether MEK1/2 functions in γδ T cell development by IL-7R-FFFF. We introduced IL-7R-FFFF cDNA into IL-7Rα−/− progenitors and cultured them by HD-FTOC in the presence or absence of U0126. As shown in Fig. 6,A, virtually no αβ and γδ T cells developed on days 5 and 7, whereas they were clearly detected on day 13 after FTOC. U0126 treatment significantly inhibited γδ T cell development at day 13. To examine whether U0126 inhibits TCRγ rearrangement even before expression of γδ TCR, total RNA was prepared from organ-cultured thymocytes on days 0, 5, and 7 of FTOC. Real-time RT-PCR revealed that Jγ-Cγ germline and Vγ-Jγ transcripts were detected in organ-cultured thymocytes as early as days 5 and 7 when no distinct population of γδ T cells was detected (Fig. 6 B). In contrast, U0126 treatment significantly suppressed induction of transcripts on days 5 and 7. These results suggest that MEK1/2 functions in Vγ-Jγ recombination and γδ T cell development stimulated by IL-7R-FFFF.
In this study, we first showed that tyrosine residues of IL-7Rα are not essential for STAT5 activation and γδ T cell development. We found that the region between Lys312 and Leu328 of IL-7Rα also delivers a signal activating STAT5 and promoting γδ T cell development. We further demonstrated that MEK1/2 interacts with and activates STAT5 by phosphorylating a critical tyrosine residue. MEK1 induces TCRγ germline transcription in T cells, and MEK1/2 functions in γδ T cell development via IL-7R-FFFF. These results suggest that MEK1/2 delivers a subsidiary signal for γδ T cell development through IL-7R signaling by phosphorylating a critical tyrosine residue of STAT5.
Although the IL-7Rα Tyr449 residue plays a major role in activating STAT5 (7, 16), our results suggest that the tyrosine residues in the intracellular domain of IL-7Rα play important roles in both αβ and γδ T cells but are not essential for STAT5 activation and γδ T cell development. Consistent with our results, it is reported that IL-7Rα transgenes with deletion mutations partially rescue TCRγ rearrangement and γδ T cell development in IL-7Rα−/− mice (24). In contrast, these results are in disagreement with a previous report that Tyr449 is required for γδ T cell development in an in vivo reconstitution system (14). One explanation for this discrepancy may be a difference in experimental systems. Our results are consistent with a recent report that IL-7Rα-Phe449 knock-in mice showed detectable levels of γδ T cells (25).
Several studies suggest that STAT5 can be activated without tyrosine phosphorylation of cytokine receptors (45, 46). In this study, we showed that MEK1/2 activates STAT5 through IL-7R-FFFF. MEK1 and MEK2 are dual-specificity kinases, which phosphorylate ERK on both tyrosine and threonine residues. However, it is unknown whether MEK1/2 phosphorylates tyrosine residues of molecules other than ERK. Our results suggest that MEK1/2 specifically phosphorylates a critical tyrosine residue of STAT5. Consistent with our results, the MEK1 inhibitor PD98059 was shown to block STAT5 activation by the growth hormone receptor (47). Collectively, our results suggest an alternative pathway of tyrosine phosphorylation and activation of STAT5 by MEK1/2. Interestingly, inhibition of MEK1/2 suppresses tyrosine phosphorylation of STAT3 by a G protein-coupled receptor (48). Therefore, this alternative pathway may be conserved among STAT proteins.
We previously demonstrated that STAT5 induces TCRγ recombination and γδ T cell development via IL-7R signaling (19). In the present study, we showed that γδ T cell development is closely correlated with STAT5 activation through the IL-7R. Furthermore, TCRγ recombination and γδ T cell development are blocked in STAT5a/b−/− mice (23). These observations support our model that STAT5 controls accessibility of the TCRγ locus (19, 21). Although IL-7 stimulation also induces activation of STAT1 and STAT3 (13), neither CA-STAT1 nor CA-STAT3 induced TCRγ germline transcription in preBR1 cells (data not shown). In addition, STAT1 and STAT3 do not bind to the Jγ1 promoter (19). Therefore, these results suggest that STAT5 is the only STAT protein inducing TCRγ germline transcription and γδ T cell development in IL-7R signaling.
We demonstrated that the region between Lys312 and Leu328 of IL-7Rα transduces an alternative signal activating STAT5. Although previous studies suggest that this region might recruit p59fyn (8), we found that p59fyn and p56lck were not involved in TCRγ germline transcription via 4R/7R-FFFF (data not shown). Interestingly, the region between Lys312 and Leu328 also contains the acidic region (8). A similar region is important for interaction and activation of JAK1 and JAK3 in the IL-2R β-chain (49). Because JAK3−/− mice are defective in rearrangement of the TCRγ locus (50), the region between Lys312 and Leu328 may be required for recruitment and activation of JAK1/3.
In this study, we demonstrated that MEK1/2 activates STAT5 through IL-7R-FFFF. Although previous studies report that ERK is activated downstream of the IL-7R (10, 41, 42), it is not known how MEK1/2 is activated by the IL-7R. Pyk2 associates with JAK1 and IL-7Rα and is rapidly activated by IL-7 (12). Pyk2 then triggers downstream signals, including the MAPK pathway (51). Therefore, Pyk2 may function in MEK1/2 activation by the IL-7R. In addition, the MEK1/2-ERK pathway is involved in serine phosphorylation of STAT5 (52). Although serine phosphorylation is important for transcriptional activation by STAT1 and STAT3, its effect on STAT5 activity is not yet clear (52, 53, 54). Although STAT5 serine phosphorylation is dispensable for TCRγ germline transcription by IL-7R, we cannot exclude the possibility that MEK1/2 controls TCRγ germline transcription by activating ERK and its downstream factors such as c-Jun, c-Fos, and c-Myc.
Although tyrosine phosphorylation of cytokine receptors is the primary activator of STAT, our study suggests that an alternative activation pathway also exists. STAT5 is a key transcription factor regulating several cytokine-induced biological responses in normal lymphocyte development. Furthermore, dysregulation of the JAK-STAT and MAPK pathways is observed in several hematopoietic malignancies (55, 56). This study should facilitate understanding of the role of an alternative pathway of STAT5 activation by MAPK in normal and malignant hematopoietic development.
We thank Drs. A. Kudo, D. L. Wiest, E. Nishida, T. Kitamura, T. Kina, M. Hattori, T. Matsuda, O. Miura, F. Arai, A. Kitanaka, N. G. Ahn, M. Kracht, and P. Holland for materials, H. Miyachi and S. Kitano for mice, S. Hayashi, M. Kishida, and S. Kamioka for excellent technical assistance, members of the Ikuta laboratory for discussion, and Dr. N. Begum for critically reading the manuscript.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by Grants-In-Aid from the Ministry of Education, Culture, Sports, Science and Technology of Japan and by grants provided by Takeda Science Foundation, Naito Foundation, Sapporo Bioscience Foundation, Astellas Foundation for Research on Metabolic Disorders, and Sumitomo Foundation.
Abbreviations used in this paper: IL-7Rα, IL-7R α-chain; CA, constitutively active; DN, dominant negative; FFFF, four intracellular domain tyrosine residues replaced with four phenylalanine residues; FTOC, fetal thymic organ culture; HA, hemagglutinin; HD, hanging drop; hIL-4, human IL-4; IL-4Rα, IL-4R α-chain; IL-7R-311, IL-7Rα cDNA with aa residues 312–459 deleted; IL-7R-328, IL-7Rα cDNA with aa residues 329–459 deleted; 4R/7R, chimeric human IL-4Rα and mouse IL-7Rα; WT, wild type.