IL-4 signaling through the IL-4Rα chain regulates the development and proliferation of the Th2 lineage of effector CD4+ T cells. Analyses of the IL-4R in factor-dependent cell lines led to the development of two apparently conflicting models of the primary structural determinants of IL-4R-mediated proliferative signaling. In one model, proliferation was dependent on the first conserved tyrosine in the cytoplasmic tail (Y1), while in the second, proliferation was independent of cytoplasmic tyrosines. We found that in activated primary T cells, mutation of only the Y1 residue resulted in a modest decrease in IL-4-induced S phase entry, a further decrease in cell-cycle completion, and a complete failure of IL-4 to induce p70S6 kinase phosphorylation. Consistent with a role for the PI3K/mammalian target of rapamycin pathway in mediating cytokine acceleration of G2/M transit, pretreatment of activated T cells with rapamycin resulted in only a modest decrease in IL-4-induced S phase entry, but a total block of cell-cycle completion. Strikingly, IL-4Rα chains that lacked all cytoplasmic tyrosines were competent to signal for STAT5 phosphorylation, mediated efficient S phase entry, and promoted cell-cycle progression. The ability of tyrosine-deficient IL-4Rs to mediate proliferative signaling and STAT phosphorylation was absolutely dependent on the presence of an intact ID-1 region. These findings show that IL-4Rα lacking cytoplasmic tyrosine residues is competent to induce ID-1-dependent proliferation, and indicate that IL-4 can promote G2/M progression via activation of the mammalian target of rapamycin pathway initiated at the Y1 residue.
A normal immune response is dependent on both the development of specific Th subsets and their clonal expansion. The proliferation of lymphocytes is a highly coordinated process in which the successful integration of multiple signaling pathways is critical. Activation through the Ag receptor allows for resting T cells to exit the G0 phase of the cell cycle. Entry into the S phase of the cell cycle is dependent on further cytokine stimulation, in the absence of which cells fail to enter S phase and undergo apoptosis (1, 2, 3, 4). In activated T cells, signaling through cytokine receptors promotes S phase entry by inducing the up-regulation of proteins that promote cell-cycle progression and the down-regulation of cell-cycle inhibitors (5, 6).
IL-4 is a pleiotropic cytokine whose actions promote the growth and survival of B cells, mast cells, and activated T cells (7). To exert these effects, IL- 4 signals through heterodimeric receptor complexes containing the IL-4R α-chain (IL-4Rα) and an accessory chain, either γ common or IL-13Rα (7, 8, 9). Binding of IL-4 to its receptor results in the activation of resident Jak and the subsequent phosphorylation of conserved tyrosine residues on the IL-4Rα chain. These phosphoryltyrosines act as scaffolding sites for molecules that contain Src homology 2 (SH2)5 or protein tyrosine-binding domains (7). IL-4 stimulation leads to the recruitment and phosphorylation of insulin receptor substrate-2 (IRS-2), STAT6, and, in activated T cells, STAT5 (7, 10, 11).
Mutagenic analyses of the IL-4Rα in cell lines led to the development of two apparently conflicting models of how the IL-4R induces proliferation. In one model, proliferation requires a conserved region of the IL-4R that shares homology with the insulin receptor and the insulin-like growth factor-1 receptor. This region, designated the I4R (insulin and IL-4R) motif, contains a highly conserved tyrosine at position 497 (Y1) (numbering given for the mature human IL-4Rα (huIL-4Rα) chain, in which number 1 is the N terminus). Factor-dependent cell lines engineered to express IL-4Rα chains that were truncated or substituted so as to be deficient for theY1 residue were impaired in their ability to support IL-4-mediated proliferation (12, 13). The defect in proliferative capacity mediated by Y1-deficient receptors was associated with the loss of IL-4-induced IRS-1/2 phosphorylation and an impairment in the induction of pathways associated with phospho-IRS-1/2, including PI3K (13). Activated PI3K induces pathways that contribute to proliferation, including Akt and the mammalian target of rapamycin (mTOR) (14).
In the second model, IL-4Rα chains from which the receptor tyrosines are deleted can induce proliferation. In these studies, IL-4-induced proliferation was dependent on a conserved acidic and serine-rich region of the IL-4R termed the ID-1 region (15). The ID-1 region is similar to an acidic region found in IL-2Rβ, which contributes to Src kinase association with the receptor (16). However, little is known about the mechanisms by which the ID-1 region couples the IL-4R to proliferative signaling. Previous analyses of the mechanisms of IL-4R-induced proliferation were largely worked out in long-term cell lines (7). In CD4+ T cells, however, IL-4-induced mitogenesis is dependent on concomitant TCR signaling. TCR activation induces pathways that result in functional alterations of IL-4R signaling. In developing and committed Th2 cells, for example, IL-4 stimulation induces STAT5 signaling (10, 11). Our laboratory has provided evidence that TCR signaling leads to new programming of the IL-4R complex (17). This reprogramming permits tyrosine-deficient receptors to induce STAT6 signaling and Th2 development, events that were dependent on an intact ID-1 region (17). Together, these findings suggest that the structural elements of the IL-4Rα that are necessary for proliferative signaling in activated T cells may differ from what has been found in model cell lines.
To investigate the determinants of IL-4Rα primary structure that are essential for IL-4-induced cell-cycle progression in activated lymphocytes, we reconstituted CD4+ T cells from IL-4R-deficient mice with a panel of mutant receptors that allowed for the simultaneous analysis of both models of IL-4R-mediated proliferation. We determined that tyrosine-deficient receptors on activated T cells were fully competent to induce STAT5 phosphorylation, S phase entry, and cell-cycle progression in response to IL-4 signaling. With a tyrosine-deficient receptor, these responses were dependent on the presence of an intact ID-1 region. Surprisingly, cells expressing full-length receptors in which only the Y1 residue was mutated exhibited a modest impairment in IL-4-induced S phase entry and a further block to cell-cycle completion. Furthermore, a specific inhibitor of mTOR attenuated S phase entry to only a modest degree under these conditions, but fully blocked cell-cycle completion, providing evidence of a role for PI3K/mTOR signaling in promoting transit through the G2/M phases. These data suggest that signaling through the Y1 residue, most likely through mTOR, counteracts negative signaling input from other regions of the IL-4R. In the absence of this negative regulation, a tyrosine-deficient receptor containing an intact ID-1 region is highly competent to initiate IL-4-induced mitogenesis.
Materials and Methods
IL-4Rα-deficient mice on a BALB/c background (gift from N. Noben-Trauth, George Washington University, Washington, DC) were bred with DO11.10 TCR transgenic BALB/c mice. Resultant IL-4Rα−/− and DO11.10 transgenic mice were genotyped by PCR (9). STAT6-deficient mice on a BALB/c background were obtained from The Jackson Laboratory and bred with DO11.10 TCR transgenic BALB/c mice. Resultant mice were genotyped by PCR. Mice were maintained in microisolator cages under specific pathogen-free conditions and used with institutional approval in accordance with applicable regulations.
Cell culture and reagents
Purified mouse rIL-4 as well as purified Abs against IFN-γ and IL-12 were obtained from BD Pharmingen. huIL-2 was a gift from Biologic Response Modifiers Program. Abs against Thy-1.1 were obtained from BD Pharmingen. Alexa647-conjugated anti-BrdU and streptavidin were obtained from Molecular Probes. Abs against phospho-p70S6 kinase, phospho-Akt, and phospho-STAT5 (p-STAT5) were obtained from Cell Signaling Technology. Splenocytes were cultured in complete IMDM supplemented with 10% FBS, 2 mM l-glutamine, 1× penicillin/streptomycin, and 50 μM 2-ME (referred to hereafter as IMDM/10F). Phoenix cells were cultured in complete DMEM supplemented as for IMDM/10F and with 0.1× MEM amino acids. Inhibitors of ERK activation, U0126 and PD98059, were obtained and used, as described previously (17), at concentrations of 10 and 50 μM, respectively.
Retroviral transduction and Th2 polarization
Single cell suspensions were prepared from spleen and lymph nodes, as described (18). Retrovirus-containing supernatants were collected 48 h after transfection of the ΦNX ecotropic packaging cell line with retrovector plasmids, and centrifuged (1 h, 10,000 × g) with peptide-activated splenocytes from IL-4R-deficient mice. The generation of MSCV-IRES-GFP (MiG) retrovectors encoding wild-type and mutated murine IL-4Rα (mIL-4Rα) cDNAs has been previously described (17). IL-4Rα cDNAs were recloned into the MSCV-IRES-Thy-1.1 (MiT) retrovector (19). Transduced cells were cultured in the presence of purified rhuIL-2 (10 ng/ml), anti-IFN-γ (1 μg/ml), anti-IL-12 (1 μg/ml), and mIL-4 (5 ng/ml) or anti-IL-4 (1 μg/ml).
Three days after antigenic activation, lymphocytes (average ∼95% CD4+ T lymphocytes) were washed and plated in serum-free IMDM for 8 h. To analyze S phase entry, lymphocytes transduced with MiG retrovectors were washed again and then plated in IMDM-10 or IMDM-10 containing either IL-4 or IL-2 for 15 h. BrdU was added, and samples were cultured for an additional 6 h before being collected and fixed with 4% paraformaldehyde. In this assay system, cytokine-driven BrdU incorporation is first observed at 15 h after stimulation. Fixed cells were washed with staining buffer (1% FBS/PBS) and frozen in 10% DMSO/90% FBS for at least 1 h. Cells were thawed, washed once with staining buffer, and fixed again with 4% paraformaldehyde before being permeabilized in 0.5% saponin/1% FBS/PBS. Permeabilized cells were treated with DNase I before staining with fluorochrome-conjugated anti-BrdU. To analyze cell-cycle completion, lymphocytes transduced with MiT retrovectors were stained with CFSE after the 8-h serum starvation. Briefly, cells were stained with 5 μM CFSE (Molecular Probes) for 5 min at 37°C. The reaction was quenched with an equal volume of FBS, and the samples were washed three times with IMDM-10 before being plated in medium or medium containing IL-4 or IL-2. Samples were stained with Abs against Thy-1.1 and analyzed via flow cytometry 48 h after cytokine stimulation. In studies using inhibitor treatment, samples were pretreated for 15 min with LY294002 (50 μM) or rapamycin (100 nM) (both from Calbiochem), stimulated with cytokine, and assayed for BrdU incorporation or cell-cycle completion.
Cells were rested overnight in serum-free medium and then stimulated for 15 min with the indicated cytokines. Whole cell extracts were prepared using radioimmunoprecipitation assay buffer (1× PBS/1% Nonidet P-40/0.5% sodium deoxycholate/0.1% SDS) supplemented with 1 mM Na3VO4, 0.5 mM NaF, microcystin (Calbiochem), and a 1/100 dilution of a protease inhibitor mixture (Sigma-Aldrich). Samples were resolved by SDS-PAGE and transferred onto polyvinylidene difluoride membranes. Western blotting with anti-phospho-p70S6 kinase was performed according to the manufacturer’s instructions (Cell Signaling Technology). Equivalent protein loading was determined by reprobing stripped membranes with Abs that recognized total p70S6 kinase (Santa Cruz Biotechnology).
Intracellular staining for phosphoproteins
Cells were rested overnight in serum-free medium, stimulated for 15 min with IL-4, and then fixed by the addition of prewarmed paraformaldehyde to a final concentration of 0.5%, followed by permeabilization in 95% ice-cold methanol for 30 min. Cells were stained with Abs against p-AKT or p-STAT5 (Cell Signaling Technology) for 30 min at 20–22°C, followed by staining with PE-conjugated anti-rabbit IgG. Cells were analyzed by flow cytometry immediately after staining.
IL-4R tyrosines are dispensable for IL-4-mediated S phase entry
Cytokines such as IL-4 promote the expansion of T cells that have been activated through their TCR in part by inducing transit through the G1/S phase checkpoint. Studies performed in long-term cell lines have led to the development of two models of IL-4R-mediated proliferation. In one, this process requires the Y1 residue, whereas in the alternative model proliferation is independent of cytoplasmic tyrosines, but dependent on the ID-1 region (7, 15). To investigate the primary structural determinants of IL-4R-induced mitogenesis in activated T cells, we developed a flow cytometric assay to measure the percentage of cells that entered S phase directly after cytokine treatment. In this assay, cells are assayed for cytokine-mediated proliferation 3 days after antigenic stimulation, eliminating the need for comitogenic stimulation and thereby focusing only on signals that emanate from cytokine receptors (Fig. 1 A).
To analyze the role of IL-4Rα signaling motifs in IL-4-induced mitogenesis of activated T cells, we used retroviral infection to reconstitute lymphoid cells from IL-4Rα-deficient mice with cDNAs encoding either wild-type mIL-4R or mutated IL-4Rs (Fig. 1,B). These receptors are expressed on the surface of T cells within 48 h of infection at levels similar to, but not exceeding, the levels expressed on the surface of lymphoid cells from IL-4R competent mice (17, 20). As expected, cells reconstituted with a wild-type receptor were able to induce S phase entry at levels similar to those mediated by IL-4R competent mice with ∼34% of the cells incorporating BrdU in response to IL-4 (Fig. 1, A and C). Cells expressing a full-length receptor in which the Y1 residue was mutated to phenylalanine (Y1F) exhibited a modest decrease in S phase entry, with cells replicating their DNA in response to IL-4 stimulation at ∼60% the efficiency of wild type. Importantly, however, receptors that were completely deficient for cytoplasmic tyrosines (Y012345F and Δ437Y0F) mediated S phase entry with efficiencies approximately that of the wild-type receptor, with 31–34% of the cells incorporating BrdU in response to IL-4 stimulation.
The ability of tyrosine-deficient receptors to mediate S phase entry was dependent on an intact ID-1 as a tyrosine-deficient receptor with a mutated ID-1 region (Δ437(Y0F)ID-1(mut)) was unable to induce S phase entry (Fig. 1,C). As a positive control, transduced cells were also treated with IL-2 and assayed along with IL-4-treated cells for cytokine-induced S phase entry. All transduced samples responded equally well to IL-2 stimulation, indicating that the S phase entry defect observed in samples expressing the Y1F receptor was due solely to defective signaling from the mutated IL-4Rα chain and not a general defect in cells expressing this receptor (Fig. 1 C). Our previous work established that the efficiency of Th2 differentiation was the same for all receptors except the Δ437(Y0F)ID-1(mut) (17). Moreover, a comparable pattern of results was obtained when these assays were performed in the presence of neutralizing anti-IL-4 (Th-null conditions) until the time of the assay (data not shown). These findings indicate that IL-4Rα cytoplasmic tyrosines are dispensable for IL-4-induced S phase entry. For a tyrosine-deficient receptor, only an intact ID-1 region is necessary for mediating G1 to S phase progression. With full-length receptors in which only the Y1 residue is mutated, however, cells were more compromised in their ability to enter S phase after IL-4 stimulation. Cells expressing Y1F receptors induced S phase entry at 60–70% the efficiency of cells expressing a wild-type receptor. These data indicate that the Y1 residue promotes, but is not essential for, G1/S progression when other tyrosines are present in IL-4Rα.
Tyrosine-deficient receptors mediate cell-cycle completion
In this S phase assay system, there is a 15-h lag time before IL-4-induced S phase entry (BrdU incorporation) is observed. This period of time allows for initiation of DNA synthesis, but not cellular division (data not shown). Although cytokines are well established as regulating the G1 to S phase transition, whether cytokine signaling contributes to cell-cycle progression through G2/M phases is not clear. Prior analyses of IL-4R-induced mitogenesis have been performed using assays that, while measuring DNA synthesis, could not discern between defects in S phase entry vs defective cell-cycle completion. It was possible that cells expressing mutant IL-4Rs could enter S phase, but be more compromised in their ability to complete the cell cycle and divide (i.e., transit through the G2/M phases). To further characterize the mitogenic response transduced through the IL-4R, we performed dye partitioning assays to measure the ability of mutated IL-4Rs to promote cell-cycle progression. Cells transduced with the indicated receptors were stained with CFSE 3 days postactivation. Following IL-4 stimulation, cells were cultured for 24–72 h to assay the efficiency of cytokine-induced cell-cycle completion, as evidenced by the execution of a division in response to IL-4 by a significant fraction of cells expressing a wild-type receptor.
Cells from samples expressing a wild-type IL-4R efficiently induced division (∼70% of cells) following IL-4 stimulation (Fig. 2). In contrast, cells expressing IL-4R that lacked cytoplasmic tyrosines completed division only 70–77% as efficiently as cells expressing wild-type receptors, even though these receptors induced S phase entry at similar levels. Of note, IL-4Rs containing only the Y1F mutation were even more compromised in their ability to transit through and complete the cell cycle, manifesting a decrease to less than half the efficiency of wild-type receptors. These data indicate that the Y1 residue is important for mediating the G2/M transition. Strikingly, this defect was most pronounced for receptors in which the tyrosines downstream of the Y1 residue are intact. This defect was limited to signaling through mutated IL-4Rα chains as no defect was observed in positive control samples treated with IL-2 (data not shown).
STAT5 induction by IL-4Rα lacking cytoplasmic tyrosine residues
IL-4Rα chains that were completely deficient in cytoplasmic tyrosines were fully competent to promote efficient S phase entry and only modestly impaired in cell-cycle completion. In contrast, mutation of the ID-1 region abrogated the transduction of all proliferative signaling by a receptor devoid of cytoplasmic tyrosine residues, including a failure to promote S phase entry. Some studies have highlighted a role for STAT6 in IL-4-dependent lymphocyte proliferation, or a requirement for STAT5 in cytokine-mediated mitogenesis of primary T cells (2, 21, 22, 23). Thus, activation of these STAT pathways by IL-4Rs lacking the Y1 residue might be sufficient to induce IL-4-driven proliferation. To investigate the requirement for STAT6 in cytokine-induced S phase progression, we analyzed IL-4-induced incorporation of BrdU into wild-type and STAT6-deficient primary lymphocytes. T cells lacking STAT6 exhibited a substantial G1/S response to IL-4, albeit one in which rates of IL-4-induced S phase entry were decreased to approximately half those of wild-type cells (Fig. 3 A). Thus, although the data are consistent with a role for STAT6 in IL-4-regulated proliferation of primary lymphocytes, they indicate that a failure to activate STAT6 (17) cannot, by itself, account for the failure of the Δ437(Y0F)-ID-1(mut) receptor to trigger any S phase entry.
In activated T cells, IL-4 stimulation induces STAT5 as well as STAT6 activation, but the primary structure requirements for IL-4Rα to induce STAT5 are not clear (10, 11, 24). Peripheral T cells from STAT5a/STAT5b-deficient mice are severely impaired in their proliferative response to IL-2 or IL-4 stimulation (2, 21). STAT5 activation regulates mitogenic signaling by inducing the transcription of many cell-cycle-dependent genes (6). To determine whether tyrosine-deficient receptors could promote the activation of STAT5, we stimulated transduced lymphocytes with IL-4 and assayed for phospho-STAT5 induction. As expected, cells expressing the wild-type IL-4Rα signaled STAT5 phosphorylation after IL-4 stimulation (Fig. 3, B and C). Importantly, cells expressing either the Y1F receptor or the tyrosine-deficient receptors were also able to induce STAT5 phosphorylation, albeit at a somewhat lower biochemical efficiency similar to what we have previously observed for STAT6 induction in activated T cells (17). Furthermore, parallel to what we previously found for STAT6 induction, the ability of tyrosine-deficient receptors to signal STAT5 phosphorylation was dependent on an intact ID-1 region so that no STAT5 induction was observed after stimulation of the Δ437(Y0F)ID-1(mut) receptor. Because the ability of activating signals to program STAT6 induction by IL-4Rα lacking cytoplasmic tyrosines depended on ERK activation, we also investigated whether inhibition of ERK activation would prevent the acquisition of a competence of IL-4 to induce STAT5 (17). When T cells were activated and grown in the presence of an inhibitor of the ERK pathway, their subsequent ability to mobilize STAT5 induction in response to IL-4 was abrogated as compared with vehicle controls (Fig. 3 D). This finding was not due to acute effects of the inhibitor, U0126, because late addition of the compound had little effect on the induction of p-STAT5. Moreover, similar results were obtained with an independent inhibitor of the pathway, PD98059 (data not shown). Collectively, these findings suggest that there is a shared mechanism underlying the functional reprogramming of IL-4Rα complexes in activated lymphocytes, so that IL-4 induction of STAT5 is similar to the STAT6 signaling pathway that is independent of cytoplasmic tyrosines.
Divergent biochemical requirements for the Y1 residue in IL-4Rα
The BrdU incorporation and CFSE dye partitioning results provide evidence that, in an otherwise intact receptor, the Y1 residue contributes to cell-cycle progression. IL-4 binding to the wild-type IL-4R induces the activation of the PI3K pathway (25). PI3K signaling contributes to proliferative signaling in part by inducing the activation of the Akt and mTOR pathways (14). Previous analyses implicated phosphorylation at the Y1 residue as being essential for the IL-4R-mediated induction of these pathways (26, 27). To analyze the IL-4R signaling motifs that contribute to the induction of PI3K signaling pathways in activated lymphocytes, we transduced splenocytes with the indicated receptors and then assayed for IL-4-induced phospho-Akt induction and mTOR-induced phosphorylation of serine 371 of p70S6 kinase under serum-free conditions. In contrast to what we observed for STAT5 induction, only receptors with an intact Y1 reside could signal Akt and p70S6 kinase phosphorylation (Fig. 4). These data indicate that the primary structure requirements for IL-4Rα-mediated STAT5 induction are similar to those previously observed for STAT6, but only the Y1 residue couples the receptor to the Akt and mTOR pathways.
mTOR signaling contributes to G2/M as well as G1/S progression
To further analyze the contributions of IL-4R-induced signaling pathways toward progression through the cell cycle, we used chemical inhibitors of pathways downstream of the PI3K pathway in our BrdU incorporation assays. Cells expressing the panel of IL-4Rs were treated with LY294002, a potent inhibitor of PI3K activation, or rapamycin, a potent and selective inhibitor of mTOR activity, and then assayed for cytokine-induced BrdU incorporation. Cells in which all PI3K activity, including basal activity induced by serum present in the medium, was blocked were severely deficient in their ability to enter S phase following IL-4 stimulation (Fig. 5,A). However, rapamycin-treated cells expressing the wild-type receptor exhibited only a modest decrease in IL-4-induced S phase entry, entering S phase at ∼80% the efficiency of untreated samples (Fig. 5,B). The two tyrosine-deficient receptors were also only modestly hindered in their ability to promote the G1 to S phase transition. Strikingly, the Y1F mutant receptor in the presence of rapamycin exhibited a much more severe block in S phase entry than the Y012345F mutant (Y(0-5)F) in the presence of inhibitor. The efficiency of inhibitor treatment was verified by assaying treated cells for known targets of the signaling pathway, p70S6 kinase (rapamcyin) or Akt (LY294002) phosphorylation (Fig. 5 C, and data not shown). This finding suggests that tyrosines downstream of the Y1 residue contribute toward a negative regulation of IL-4-induced mitogenesis, an inhibitory influence that is lost in the tyrosine-deficient receptors.
Cells expressing IL-4Rα chains that lacked the Y1 residue exhibited defects in their ability to complete the cell cycle following IL-4 stimulation, yet cells stimulated with IL-4 exhibited only a minor impairment in their ability to enter S phase in the presence of rapamycin. These findings suggested that the mTOR pathway may contribute not only to S phase entry, but also to promoting passage through the G2 and M stages of the cell cycle in activated CD4+ T cells. To characterize the role of the mTOR pathway in cell-cycle completion, cells expressing the indicated receptors were stained with CFSE, pretreated with rapamycin, stimulated with IL-4, and analyzed for their ability to divide following a cytokine signal. Treatment of cells with rapamycin resulted in a complete block of cellular division in all samples (Fig. 6). As expected from the BrdU incorporation results, cells treated with LY294002 also did not divide in response to IL-4 (data not shown). Together, these data with activated lymphocytes highlight a novel role for signaling by a rapamycin-sensitive mTOR complex mediating cytokine-induced cell-cycle progression at a stage after checkpoints leading through G1 and into S phase.
By analyzing a panel of IL-4Rα chains mutated in the signaling domains proposed to be essential for proliferation, we have determined which motifs promote specific aspects of proliferative signaling in activated lymphocytes. Strikingly, these analyses reveal that pathways emanating from the Y1 residue can contribute both to S phase entry and cell-cycle completion, defining a previously unrecognized role for IL-4R signaling in G2/M progression. Consistent with this requirement for IL-4 signaling in completion of mitosis, under these conditions assaying a predominantly cytokine-dependent signal, the mTor inhibitor rapamycin abrogated cell division despite allowing most cells to undergo S phase entry. Although the findings confirm that the Y1 residue contributes to IL-4-induced mitogenesis activated by receptors in which downstream tyrosines are intact, two independent IL-4Rα mutants lacking cytoplasmic tyrosines induced S phase entry at levels similar to those mediated by the wild-type receptor. For tyrosine-deficient receptors, IL-4-mediated mitogenesis absolutely depended on the ID-1 region, inasmuch as tyrosine-deficient receptors with mutations in this region were unable to signal for S phase entry or cell-cycle progression. Moreover, all receptors capable of inducing proliferation were able to signal STAT5 activation. The ability of tyrosine-deficient receptors to activate STAT5 phosphorylation was dependent on the ID-1 region and on a functional ERK pathway during and after T cell activation. Each of these features mirrors the pattern we have previously reported for STAT6 (17), thereby suggesting a shared ERK-dependent mechanism for the change in IL-4Rα signaling that follows T cell activation.
Previous analyses of the role IL-4R signaling domains in proliferation were performed using factor-dependent cell lines. These studies led to the development of two seemingly incompatible paradigms delineating the contributions of IL-4Rα signaling motifs in IL-4-induced mitogenesis. In one model, proliferative signaling was dependent on the Y1 residue, while in the other proliferation proceeded in the absence of receptor tyrosines, but was instead dependent on the ID-1 region (7, 15). Our data led us to propose a model of proliferative signals by the IL-4R in lymphocytes that integrates these two paradigms (Fig. 7). In this model, the Y1F residue initiates signaling cascades whose role is to counteract negative regulatory mechanisms mediated by tyrosines downstream of the Y1 residue. In the absence of these negative regulatory residues, tyrosine-deficient receptors are capable of inducing mitogenesis provided they contain an intact ID-1 region. The ability of tyrosine-deficient receptors to signal proliferation may be linked to the ability of these receptors to induce phospho-STAT5 and STAT6. STAT5 is essential for cytokine-driven proliferation, as illustrated by the severe defect of STAT5a/b-deficient T cells in their IL-2- and IL-4-induced proliferative responses (2, 21). Indeed, one major difference in IL-4 signaling in cell lines vs primary cells is the functional usage of the STAT5 signaling pathway after IL-4 stimulation in activated lymphocytes. Because elimination of the capacity to use STAT6 reduced, but did not eliminate, IL-4-induced S phase entry, we infer that the ability of tyrosine-deficient IL-4Rα coordinately to mobilize p-STAT5 after T cell activation contributes to their ability to signal proliferation in lymphocytes. Although the mechanisms by which TCR signaling reprograms IL-4R function are not fully understood, we have previously published evidence that both ERK and NF-κB signaling contribute to alterations in IL-4R signaling (10, 17). As noted, our evidence suggests that at least the ERK-dependent reprogramming mechanism is shared by the STAT5 and STAT6 pathways. Both STAT5 and STAT6, once activated, would promote cell-cycle progression in part by activating genes that regulate proliferation, including growth factor-independent-1, c-myc, bcl-2, and bcl-x (6, 22, 23).
One basis for the negative regulation of IL-4-induced mitogenesis may be the motif containing the Y5 residue, IXYXXL, which resembles the consensus ITIM (ITIM motif). This sequence can serve as a scaffolding site for tyrosine phosphatases and has been linked to the negative modulation of receptor signaling (28, 29). Phosphopeptides derived from the IL-4Rα ITIM immunoprecipitate the protein phosphatases SH2-containing phosphatase (SHP)-1, SHP-2, and SHIP. SHIP phosphorylation is induced by IL-4 stimulation (28, 30). The ITIM found in the erythropoietin receptor contributes to SHP-1 association with the receptor, an event that dampens proliferative signaling induced by erythropoietin binding (31). Pre-B cells from motheaten mice, which have greatly reduced levels of SHP-1, display enhanced activation of STAT6 compared with wild-type controls in response to IL-4 stimulation (32). The 32D/IRS-2 cells expressing IL-4Rα chains in which the Y5 residue was mutated to F (Y5F) were hyperproliferative when compared with cells expressing the wild-type IL-4R (28). Moreover, phosphopeptides derived from the sequence surrounding the IL-4Rα Y2 residue may be capable of associating with SHP-2, suggesting that this region of the receptor might also contribute to the termination of IL-4R signaling (33). Full-length IL-4Rα chains in which the ITIM is mutated promoted enhanced STAT6, but not IRS-2, phosphorylation (28, 34). Similarly, the duration of IL-4-induced STAT6 phosphorylation was enhanced in cells that lacked functional SHP-1 (34). Therefore, although the activation of STAT5 and STAT6 mediated by tyrosine-deficient receptors is weaker than that induced by wild-type receptors, the loss of the negative regulatory signaling motif may allow for prolonged activation of these transcription factors.
We also provide evidence that IL-4 signaling promotes transit through the G2/M phases, most likely through the activation of the mTOR pathway. Treatment of the cells with the mTOR inhibitor rapamycin resulted in a modest decrease in S phase entry, but a complete block in cell-cycle progression (Fig. 6). This block in cell-cycle completion is more severe than what was observed for the Y1F receptor, most likely due to rapamycin blocking all mTOR activity, including that which is induced by nutrients and growth factors in the medium. This novel role for cytokine signaling in mediating transition through a G2/M checkpoint would not have been prominent in conventional BrdU/PI cell-cycle analyses or discernible in [3H]thymidine uptake assays in which DNA synthesis is measured 48 h poststimulation. Decreased [3H]thymidine incorporation in such assays can result from a failure to promote S phase entry, an increase in apoptosis, a block at the G2/M checkpoint, or combinations of several of these. Indeed, the block to IL-4-induced proliferation mediated by receptors that lacked the Y1 residue appeared more profound in previous analyses. In those assays, cells would complete multiple divisions before being assayed, so that the decrease in cell-cycle completion we measured with CFSE would be amplified to the larger deficit observed in earlier [3H]thymidine assays. A failure to protect from apoptosis could, in principle, account for the decrease that we observed for cell-cycle progression mediated by the Y1F IL-4Rα mutant. However, propidium iodide assays uncovered no discernible difference among the IL-4Rs with intact ID-1 in their mediation of protection against apoptosis by IL-4 (L. Stephenson and M. Boothby, unpublished observations).
Surprisingly, while our data indicate that the Y1 residue is critical for the induction of the Akt and mTOR pathways following IL-4 stimulation, we demonstrate a less stringent requirement for their activation by IL-4 in cytokine-induced S phase entry of activated T cells. This observation is consistent with a role for PI3K signaling in potentiating, but not initiating, cytokine-induced proliferation mediated by STAT5 signaling in lymphoid cells (35). However, the relative contributions of cytokine-induced PI3K and STAT5 signaling to proliferation of primary cells remain an unsettled issue. Our data support a model in which the PI3K signaling induced by growth factors in serum and after TCR activation is sufficient to potentiate S phase entry induced by IL-4Rα activation of STAT5 and STAT6. For G2/M progression, however, receptors lacking Y1 are less efficient, thereby suggesting a more stringent threshold requirement for PI3K signaling at this stage of the cell cycle and a greater dependence on the ability of a cytokine to enhance PI3K activity. Finally, these studies used antigenic peptide-loaded APCs to initiate T cell activation and cycling, suggesting that the characteristics of TCR signaling and costimulation were physiologically relevant. However, future studies in vivo will elucidate the extent to which these aspects of IL-4Rα signal initiation and cell-cycle regulation influence clonal expansion in immune responses.
We thank J. Price and C. Alford for cytometric assistance; Mary Johns and Haidong Chen for technical assistance; Immunex, P. Marrack, and W. Sha for plasmids and reagents; K. Pennington, M. Sundrud, V. Torres, and D. Unutmaz for helpful discussions; G. Nolan for ΦFNX cells; and N. Noben-Trauth for BALB/c-IL-4Rα−/− mice.
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.
Submitted in partial fulfillment of the requirements of the Ph.D. degree (to L.M.S.) at Vanderbilt University. This work was supported by National Institutes of Health Grant GM42550, the Sandler Family Supporting Foundation through a Senior Investigator Award of the Sandler Program for Asthma Research (to M.B.), Molecular Endocrinology Training Program Training Grant T32 DK07563 (to L.M.S.), and the postdoctoral fellowship program of Korea Science & Engineering Foundation (M01-2004-000-20448-0; to D.-S.P.). Core facilities essential to this work were supported by the Diabetes Research and Training Center (DK20593) and Vanderbilt Ingram Cancer Center (CA68485).
Abbreviations used in this paper: SH2, Src homology 2; huIL, human IL; IRS, insulin receptor substrate; MiG, MSCV-IRES-GFP; mIL, murine IL; MiT, MSCV-IRES-Thy-1.1; mTOR, mammalian target of rapamycin; SHP, SH2-containing phosphatase; p-STAT5, phospho-STAT5.