Positive Regulation of Interleukin-4-mediated Proliferation by the SH2-containing Inositol-5′-phosphatase*

The SH2-containing inositol 5′-phosphatase (SHIP) is tyrosine-phosphorylated in response to cytokines such as interleukin (IL)-3, granulocyte-macrophage colony-stimulating factor, and macrophage colony-stimulating factor. SHIP has been shown to modulate negatively these cytokine signalings; however, a potential role in IL-4 signaling remains uncharacterized. It has been recently shown that IL-4 induces tyrosine phosphorylation of SHIP, implicating the phosphatase in IL-4 processes. Tyrosine kinases, Jak1 and Jak3, involved in IL-4 signaling can associate with SHIP, yet only Jak1 can tyrosine-phosphorylate SHIP when co-expressed. In functional studies, cells overexpressing wild type SHIP are found to be hyperproliferative in response to IL-4 in comparison to parental cells. In contrast, cells expressing catalytically inactive form, SHIP(D672A), show reduced proliferation in response to IL-4. These changes in IL-4-induced proliferation correlate with alterations in phosphatidylinositol 3,4,5-triphosphate levels. However, no differential activation of STAT6, Akt, IRS-2, or p70S6k, in response to IL-4, was observed in these cells. These data suggest that the catalytic activity of SHIP acts in a novel manner to influence IL-4 signaling. In addition, these data support recent findings that suggest there are uncharacterized signaling pathways downstream of phosphatidylinositol 3,4,5-triphosphate.

Interleukin (IL) 1 -4 is a critical modulator of cellular re-sponses within the immune system (1). IL-4 signaling is transduced via the IL-4 receptor ␣ chain (IL-4R␣) that heterodimerizes with the common ␥ chain (␥c) upon ligand binding, whereupon the Janus tyrosine kinases (Jaks), constitutively bound to their respective receptor chains, are activated via transphosphorylation (2)(3)(4). Activation of the IL-4R-associated kinases, Jak1 and Jak3, leads to tyrosine phosphorylation of the IL-4R␣ itself. Five conserved tyrosine residues in the cytoplasmic tail of the IL-4R␣ are potential sites of phosphorylation and of subsequent interaction with downstream signaling effectors through SH2-and PTB-containing protein modules. The identification of tyrosine residues critical for activation of signaling pathways and subsequent analysis of molecules that interact with these residues have led to the biochemical characterization of pathways activated by IL-4R engagement. Accumulated evidence supports the idea that, although there appears to be some functional overlap, the IL-4R␣ cytoplasmic region is segregated into distinct functional regions. The region surrounding Tyr-497 is involved primarily in proliferation and protection from apoptosis via the recruitment/activation of the adapter molecules such as insulin receptor substrates 1 and 2 (IRS-1/2), whereas gene activation is regulated via STAT6 recruitment/activation at one of three redundant tyrosine sites at Tyr-575, Tyr-603, or Tyr-631 (5)(6)(7). Although the activation of signaling from the IL-4R has been relatively well studied with the identification of proximal signaling effectors, the intracellular mechanism that limits or terminates IL-4 signaling remains less characterized.
Recently, the polyphosphoinositide phosphatase SHIP has been implicated in the negative modulation of signaling initiated by a number of cytokines and other stimuli (8). SHIP was initially identified and isolated through its recruitment to the immunoreceptor tyrosine-based inhibitory motif of the Fc␥RIIb receptor as well as through its association with Shc in response to a number of stimuli (9 -11). SHIP contains a central inositol 5Ј-phosphatase domain, which acts specifically on both phosphatidylinositol-3,4,5-P 3 and inositol-1,3,4,5-P 4 (12,13). The central phosphatase domain is flanked by a NH 2 -terminal SH2 domain and a COOH-terminal containing two NPXY motifs as well as a number of interspersed proline-rich regions. The NPXY motifs mediate association with proteins such as Shc and Disabled via their respective PTB domains (14,15), whereas the proline-rich sequences may serve as docking sites for SH3-containing proteins such as Grb2 (12,13,16). SHIP is tyrosine-phosphorylated by a broad range of stimuli including insulin, B cell receptor (BCR) stimulation, Fc␥RIIb-BCR cocross-linking, Fc␥RIIb clustering, CDw150, IL-3, IL-2, GM-CSF, G-CSF, and steel factor. SHIP has been recently implicated in IL-4 signaling by the description of IL-4-mediated * This work was supported in part by National Institutes of Health Grants AI33450, CA64628, AI41447, and 2T32DK07328 and a grant from the Asthma and Allergy Foundation of America. 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 1 The abbreviations used are: IL, interleukin; Jak, Janus tyrosine kinase; STAT, signal transducer and activator of transcription; SHIP, SH2-containing inositol 5Ј-phosphatase; IRS, insulin receptor substrate; PIP 3 , phosphatidylinositol 3,4,5-triphosphate; PTB, phosphotyrosine-binding; PI3-K, phosphoinositide 3-kinase; PAGE, polyacrylamide gel electrophoresis; GM-CSF, granulocyte-macrophage colonystimulating factor; WT, wild type; IL-4R, IL-4 receptor; BCR, B cell receptor; GFP, green fluorescent protein; FACS, fluorescence-activated cell sorter; PI(3,4)P 2 , phosphatidylinositol 3,4-bisphosphate; PI(3,4,5)P 3 , phosphatidylinositol 3,4,5-triphosphate; KI, kinase-inactive; PH, pleckstrin homology; PI, phosphatidylinositol. SHIP phosphorylation, although its role remains to be clarified (17).
To date, the mechanism of SHIP action has been most extensively investigated in the context of BCR-Fc␥RIIb signaling. Upon cross-linking of Fc␥RIIb to the BCR, Ca 2ϩ mobilization is inhibited via SHIP recruitment to the immunoreceptor tyrosine-based inhibitory motif of Fc␥RIIb (18). Studies suggest that by altering the relative composition of phosphoinositide pools, SHIP can differentially regulate membrane targeting of proteins. Indeed, an emerging picture from multiple systems suggests protein modules such as PH domains and FYVE domains bind phosphorylated inositol groups with varying specificity and thereby regulate protein localization and function (19,20). In the case of BCR-Fc␥RIIb signaling, levels of PIP 3 are the critical regulators of Ca 2ϩ signaling at least in part by their ability to support the activation of the PH domain containing Tec family kinases which activate phospholipase C␥2 and subsequent inositol-1,4,5-P 3 production (21,22). SHIP inhibits Ca 2ϩ mobilization by reducing the recruitment and thereby activation of the Tec kinases to the plasma membrane via shifting the PI pool away from PIP 3 . In a similar fashion, studies suggest that SHIP also negatively regulates activation of the PH-containing serine/threonine kinase, Akt, in response to Fc␥RIIb-BCR cross-linking and IL-3 stimulation in myeloid cells through the presumed negative modulation of membrane targeting (23,24). It is likely that along with reduction in PIP 3 mediated by SHIP, the concomitant increase in PI(3,4)P 2 may in turn increase the membrane targeting of other yet identified proteins that preferentially bind PI(3,4)P 2 . Thus SHIP may lead to the shuttling of various proteins in and out of membrane compartments based at least partially on their lipid binding specificity.
All reports to date have suggested that SHIP performs a negative regulatory role in cytokine signaling. Elucidation of the role of SHIP in vivo has been aided by the analysis of SHIP homozygous null mice (25). Hematopoietic stem cells from SHIP(Ϫ/Ϫ) mice displayed increased colony formation in response to IL-3, GM-CSF, G-CSF, and steel factor, suggesting that SHIP acts as a negative regulator of either proliferation and/or survival in response to these particular cytokines. Interestingly, SHIP(Ϫ/Ϫ) mice display decreased survival secondary to lung infiltration from cells of the myeloid and granulocytic compartment correlating with increased colony formation observed in response to this subset of cytokines. Conversely, analysis of other cellular compartments suggests that SHIP may not solely act as a negative regulator but potentially may serve other functions. In support of this notion, the null mice exhibit decreased bone marrow cellularity and a substantial reduction in pre-B cells with a concomitant reduction in peripheral B cells (25,26). This observation suggests a potential positive role for SHIP. In addition, SHIP recruitment attenuates Fc␥RIIb-induced B cell apoptosis via an unclear mechanism, independent of its inhibitory effect on Ca 2ϩ mobilization (27). This latter result suggests SHIP may differentially effect signals emanating from the same receptor, acting as inhibitor on the one hand or as a positive regulator on the other, depending on the signaling pathways examined.
The role of SHIP in cytokine signaling has been addressed by utilizing overexpression studies of wild type SHIP and/or by analysis of knockout mice. However, it remains unclear what effect, if any, the catalytic activity of SHIP domain contributes to these observations. For example, the ability of SHIP to interact with Shc has been proposed to allow SHIP to sequester Shc from Grb2 and thereby reduce Ras signaling, allowing SHIP to inhibit signaling independently of its catalytic activity in some systems (28 -30). Therefore, the relative contribution of the catalytic activity of SHIP in relation to its other domains may vary upon the signaling pathway and cell type. In this report, we focus on the potential role of the catalytic activity of SHIP in IL-4 signaling. In order to gain insights into possible mechanisms of SHIP action, we have begun to establish the structural arrangement of SHIP within the IL-4 receptor signaling complex. In doing so, we have implicated Jak1 in SHIP regulation in vivo. Analysis of IL-4-mediated proliferation in cells overexpressing different forms of SHIP demonstrates a novel positive regulatory role for the catalytic activity of SHIP. This regulation correlates with an alteration in PIP 3 levels but surprisingly not with an alteration in Akt, STAT6, IRS-2, or p70 S6k activation. These results point to a positive regulatory role of the catalytic activity of SHIP and identify SHIP as a potentially key modulator of IL-4 responses.
Plasmids-To generate the retrovirus-based expression plasmids for wild type SHIP(WT) and phosphatase-inactive mutant SHIP(D672A) (32), the cDNAs encoding for these SHIP were subcloned into the retroviral vector MSCV-pIRES-GFP (a kind gift of Dr. G. Nolan, Stanford University, CA), which is an internal ribosome entry site bicistronic expression vector allowing concomitant expression of both SHIP and GFP protein. Myc-tagged SHIP constructs were generated by subcloning the coding region of SHIP cDNA into pcDNA3.1 (Invitrogen, Carlsbad, CA). Wild type and kinase inactive mutants of Jak1 and Jak3 were also subcloned into pcDNA3.1.
Transient Transfections, Retroviral Infection, and Establishment of Cell Lines-Transient transfection of 293T cells was performed by the calcium phosphate coprecipitation method. Retroviruses were generated by transient transfection of retroviral DNA into amphotropic packaging cell line Phoenix by use of the calcium phosphate co-precipitation method. Supernatant-containing viruses were collected at 48 h after transfection. Viral stocks were stored at Ϫ80°C until used or were used immediately for infection. After expansion of infected cells, GFP-positive cells were sorted by FACS (FACStar, Becton Dickinson, Mountain View, CA). Protein expression was confirmed by Western blotting after sorting.
Proliferation Assays-32D/IRS-2 cell lines were grown in complete RPMI 1640 medium and 5% WEHI-3-conditioned media 3 days prior to experiment. On the day of the experiment, cells were washed 3 times with complete RPMI 1640 without 5% WEHI-3-conditioned media and seeded at a density of 2.5 ϫ 10 4 cells per 200 l in plate in complete RPMI 1640 with varying concentrations of recombinant murine IL-4 or 5% WEHI-3 conditioned medium in triplicate. One Ci of [ 3 H]thymidine (NEN Life Science Products) was added 40 h after stimulation followed by incubation for 8 h before analysis. [ 3 H]Thymidine incorporation was measured by a scintillation counter.
Apoptosis Analysis-32D/IRS-2 cell lines were grown in RPMI 1640 and 5% WEHI-3-conditioned media for 3 days prior to the experiment. On the day of the experiment, cells were washed 3 times with complete RPMI 1640 without 5% WEHI-3-conditioned media or serum. Cells were cultured at a density of 2.5 ϫ 10 4 cells per 200 l in a 96-well flat-bottom microtiter plate in culture media with varying concentrations of recombinant murine IL-4 (0, 1, 10 ng/ml) in triplicate for 0, 24, and 48 h. For each condition, the number of trypan blue-positive cells per total of 100 were counted as a measure of cell death. The standard deviation of the triplicate per condition was calculated and expressed as the error bar per condition. One of three representative experiments is shown. A second method to analyze cell death/viability was also performed, data not shown. Cells were prepared as in method one and then stimulated with murine IL-4 at 10 ng/ml or 5% WEHI-3-conditioned medium for a total of 24 h. After 24 h of cytokine stimulation, aliquots of cells were harvested and analyzed by Annexin-V-PE and PI staining analyzed by FACStar. The percentage of viable cells was calculated as the number of Annexin-V and PI double negative staining cells divided by the total number of cells analyzed (10,000).
In Vivo PIP 3 Assay-32D/IRS-2 cells were IL-3-starved overnight and then washed with phosphate-free minimum essential medium (Life Technologies, Inc.) followed by incubation in the presence of 50 Ci of [ 32 P]orthophosphate (NEN Life Science Products) supplemented with 0.1% bovine serum albumin for 6 h. After stimulation with IL-4, 500 l of ice-cold mixture of 1 N HCl/methanol (1:1) was added and vigorously vortexed. The lipids in organic phase were extracted 3 times with chloroform and washed with ice-cold mixture of 1 N HCl/methanol (1:1) before drying. Dried lipids were resolved by thin layer chromatography in chloroform/acetone/methanol/acetic acid/H 2 O (46:17:15:14:8).
Alteration of PIP 3 Levels in Response to IL-4 by SHIP-To investigate the functional importance of SHIP in IL-4 signaling, we generated stable transformants of 32D/IRS-2 cells expressing wild type SHIP(WT) or phosphatase-inactive SHIP (D672A), which has a point mutation Asp to Ala in the putative phosphatase active site aspartate at position 672. This mutation is reported to render the protein catalytically (32). In order to establish these stable cell lines, retroviral expression plasmids containing GFP with/without SHIP(WT) or SHIP(D672A) ( Fig. 2A) were transfected into amphotropic packaging cell line Phoenix, and supernatants from these cells were used for viral infection of 32D/IRS-2. After the repeated sorting by GFP, the expression of SHIP was examined by FACScan (Fig. 2B) and Western blotting (Fig. 2C). By using this method, we established independent polyclonal cell lines expressing SHIP(WT) or SHIP(D672A). A representative cell line expresses the ectopically expressed proteins ϳ5-fold higher than the endogenous SHIP protein (Fig. 2C).
Previously, it has been shown that PIP 3 level is elevated by PI 3-kinase (PI3-K) in response to IL-4 stimulation (35). SHIP specifically removes the 5Ј-phosphate from PIP 3 , as well as inositol 1,3,4,5-tetraphosphate in vitro (12,13). Furthermore, SHIP(Ϫ/Ϫ) null mice appear to have increased levels of PIP 3 in response to IL-3, suggesting that SHIP in vivo acts downstream of PI3-K to negatively modulate PIP 3 levels (23). We, therefore, sought to determine whether overexpression of either SHIP(WT) or SHIP(D672A) would be capable of altering PIP 3 levels upon IL-4 stimulation in vivo. In vector alone cells, PIP 3 is induced upon IL-4 treatment (Fig. 3). In cells overexpressing SHIP(WT), IL-4-induced levels of PIP 3 are significantly reduced, consistent with the ability of SHIP to dephosphorylate the 5Ј-position of PIP 3 thereby reducing its levels. In contrast, induction of PIP 3 is greater in cells overexpressing SHIP(D672A) when compared with vector alone cells. Furthermore, the PI3-K inhibitor wortmannin blocked the IL-4-induced increase in PIP 3 production in SHIP(D672) cells, suggesting that the catalytic activity of SHIP lies downstream of PI3-K. Taken together, these results suggest an important role for SHIP in regulating PIP 3 levels upon IL-4-mediated PI3-K activation.
SHIP Positively Regulates Cell Proliferation Induced by IL-4 -SHIP has been shown to be a negative regulator of cellular growth induced by several receptor signaling complexes (8). To examine the role of SHIP in regulating IL-4-induced proliferation, cells were starved from IL-3 and then cultured with the various concentrations of IL-4 for 48 h. Proliferation of these cells was then examined by [ 3 H]thymidine incorporation. SHIP(WT) cells exhibited greater proliferation in response to IL-4 stimulation when compared with vector alone cells over a range of concentrations (Fig. 4A). In striking contrast, the proliferative response to IL-4 was suppressed in cells overexpressing SHIP(D672A) when compared with control cells (Fig.  4A). These results indicate that the increased proliferation observed in SHIP(WT) cells is not secondary to nonspecific protein-protein interactions mediated by SHIP protein overexpression but rather requires the catalytic activity of SHIP. Similar effects on cellular proliferation by SHIP in 32D/IRS-2 cells were observed in other sets of independently generated stable cell lines (data not shown). To determine if SHIP also affected the growth of these same cells in response to IL-3, thymidine incorporation was measured in cells grown in 5% WEHI-3-conditioned media. A previous study utilizing the IL-3-dependent cell line DE-AR revealed that overexpression of wild type SHIP had no effect on cell number (33). In agreement with this report, there was no effect on proliferation of 32D/ IRS-2 cells overexpressing SHIP(WT) or SHIP(D672A) at least at the concentration measured (Fig. 4A). These results suggest that the catalytic activity of SHIP functions as a positive regulator of IL-4-mediated proliferation. This effect correlates with a reduced level of PIP 3 (Fig. 3). In addition, the positive role for SHIP catalytic activity with respect to proliferation appears to be specific for IL-4.
The observed differences in thymidine incorporation data above may have been explained by a differential effect on cell survival among cells overexpressing of SHIP(WT) versus SHIP(D672A). This is unlikely based on two independent observations. First, an analysis of cell death over 0, 24, or 48 h of either 0, 1, or 10 ng/ml of IL-4 stimulation was performed comparing vector alone cells, SHIP(WT)-, and SHIP(D672A)expressing cells. No significant differences in cell viability as measured by trypan blue staining was observed over 48 h among the these SHIP-expressing cells under varying amounts of IL-4 (Fig. 4B). Second, utilizing Annexin-V and PI staining to detect apoptotic cells, at 0 and 24 h of constant IL-4 stimulation at 10 ng/ml there is no difference in the level of apoptosis among the different SHIP-expressing cells (data not shown). Taken together, these results suggest that the observed differences in proliferation between SHIP(WT) and SHIP(D672A) was not the result of differences in the rate of cell death among the these cells in the setting of constant IL-4 stimulation or upon withdrawal of IL-3.
SHIP Does Not Affect Tyrosine Phosphorylation of IRS-2 and STAT6 -In order to gain insight into the mechanism by which SHIP alters IL-4-induced proliferation, the activation of known components of IL-4 signaling were examined. In 32D cells, IRS-1 and IRS-2 in particular have been shown to be critical for IL-4-mediated proliferation (5-7). Upon IL-4 stimulation, IRS-2 binds the IL-4R leading to p85 recruitment to IRS-2. This p85-IRS-2 interaction then leads to PIP 3 induction (36). The potential inositol binding specificities of the PH and/or PTB domains of IRS proteins have not been defined in vivo. However, recent structural evidence suggests that in vitro the PTB domain of IRS-1 does not bind lipids, whereas the PH domain appears to bind preferentially PI(3,4,5)P 3 relative to PI(3,4)P 2 (37). As a result of the observed effect of SHIP catalytic domain on IL-4-mediated proliferation coupled with the current evidence from other studies implicating SHIP in the regulation of membrane targeting of proteins via PH domains, IRS-2 activation in our cell lines was examined. IRS-2 was immunoprecipitated from extracts prepared from cells stimulated with IL-4 for 0, 15, 30, and 90 min and examined for tyrosine phosphorylation.
Overexpression of either SHIP(WT) and SHIP(D672A) did not effect overall IRS-2 tyrosine phosphorylation, suggesting that the catalytic domain of SHIP alone does not regulate IRS-2 phosphorylation (Fig. 5A).
Although the studies above did not demonstrate any alterations in total IRS-2 tyrosine phosphorylation in cells overexpressing SHIP(WT) and SHIP(D672A), this may be a relatively insensitive measure of IRS-2 association with the plasma membrane. We therefore examined IL-4-induced association of IRS-2 with the p85 subunit of PI3-K as measure of membraneactivated p85. IRS-2 immunoprecipitants were examined for the presence of p85. IL-4 induced similar levels of p85 association with IRS-2 in cells overexpressing SHIP(WT), SHIP(D672A), or vector control (Fig. 5). Similarly, immunoprecipitations of p85 demonstrated that IL-4 induced similar levels of IRS-2 association in all three cell types (data not shown). These results suggest that the alteration PIP 3 levels observed among the cells lines were not secondary to a failure to recruit p85. Furthermore, these results demonstrate that although SHIP catalytic activity alone can alter the IL-4 induction of PIP 3 levels, this does not affect IRS-2 phosphorylation or association with the p85 subunit of PI3-K. Along with the inhibition of PIP 3 induction by wortmannin in SHIP(D672A) cells, these results suggest that SHIP acts downstream of PI3-K activation.
Several reports suggest that in addition to IRS-1/2, STAT6 activation also plays an important role in IL-4-mediated proliferation via regulating the expression of the cell cycle inhibitor p27 Kip (38,39). We therefore examined cells expressing different forms of SHIP for the ability of IL-4 to effect the activation of STAT6. Extracts were generated from cells cultured with IL-4 for different times. Activation of STAT6 was examined by immunoprecipitations and Western blotting with anti-phosphotyrosine antibodies. There was no difference in the levels of STAT6 tyrosine phosphorylation induced by IL-4 in cells expressing overexpressing SHIP(WT), SHIP(D672A), or vector control.
SHIP Acts Independently of Akt or p70 S6K Pathway-The serine/threonine kinase Akt is believed to be a major downstream effector of PI3-K activation induced upon mitogen/survival signaling (40,41). Accumulated evidence suggests that Akt is activated by a combination of serine/threonine phosphorylation and targeting to the cellular membrane via an NH 2terminal PH domain. Analogous to SHIP-negative regulation of the PH-containing Tec family of kinases, SHIP negatively regulates Akt activity in some systems (23,24). SHIP deficiency correlates with increased PIP 3 levels and Akt activation, thereby suggesting that PIP 3 leads to preferential membrane targeting and thereby Akt activation, although this has not been formally proven. On the other hand, Akt binds with high affinity to both PIP 3 and PI(3,4)P 2 , which is generated by SHIP. The relative contribution of these lipids to AKT activation is unclear (42-46). As IL-4 previously has been shown to activate Akt in a PI3-K-dependent manner (47-49), Akt activation upon IL-4 stimulation in the cell lines expressing vari-ous forms of SHIP was examined. Unexpectedly, although PIP 3 levels are altered in the cells lines, no differential effect on Akt activation was observed utilizing a phosphospecific antibody to Ser-473 as a surrogate marker of activation (Fig. 6A). In addition, no difference in Akt enzymatic activation in response to IL-4 was observed using in vitro kinase assays (data not shown).
A second PI3-K-dependent pathway is mediated by the serine/threonine kinase p70 S6K , which appears to potentially regulate cell proliferation in parallel but independently of the Akt pathway (50). Since IL-4 is known to activate p70 S6K (51), a possible role for the catalytic activity of SHIP in regulating p70 S6K was explored. Surprisingly, similar to the results with respect to Akt, no differential effect on activation of p70 S6K was observed comparing control, SHIP(WT), and SHIP(D672A) cell lines using a phospho-specific antibody to p70 S6k (Fig. 6B). Taken together, these results suggest that the catalytic activity of SHIP alone is insufficient to regulate both Akt and p70 S6K activation. Furthermore, the effects of the catalytic domain of SHIP on IL-4-mediated proliferation cannot be accounted for by alteration in Akt or p70 s6k activation.
The Role of Jak1 and Jak3 in SHIP Phosphorylation-The kinase(s) responsible for phosphorylation of SHIP in response cytokines have not been identified. Syk and Lck have been implicated in SHIP phosphorylation in BCR-Fc␥RIIb and TCR signaling, respectively (14,52). Two members of the Janus kinase family, Jak1 and Jak3, are activated in response to IL-4 in hematopoietic cells. The induction of SHIP phosphorylation in response to IL-4 suggests that kinases activated by IL-4 may be capable of phosphorylating SHIP. To examine further the mechanisms by which SHIP is phosphorylated and recruited to the IL-4 signaling complex, we examined whether SHIP can be phosphorylated by either of these kinases. 293T cells were transfected with Myc-tagged SHIP together with either wild type Jak1(WT) or Jak3(WT) as well as their catalytic deficient counterparts, Jak1(KI) or Jak3(KI). Cell lysates were immunoprecipitated with anti-Myc antibody, and tyrosine-phosphorylated SHIP was visualized by anti-phosphotyrosine immunoblotting. As shown in Fig. 7A, Jak1(WT), but not Jak3(WT), was capable of mediating SHIP phosphorylation. Moreover, an additional tyrosine-phosphorylated protein was detected upon Jak1/SHIP co-expression. This phosphorylated protein was shown to be Jak1. Interestingly, although Jak1(KI) fails to FIG. 5. Effects of overexpression of SHIP on activation of IL-4R␣-associated proteins. A, cells were treated with IL-4 for indicated times, and cell lysates were prepared. After immunoprecipitation (IP) with anti-IRS-2 antibody, tyrosine phosphorylation of IRS-2 and association of p85 PI 3-kinase subunits were examined by Western blotting (WB) using anti-phosphotyrosine and anti-p85 antibodies, respectively. Equal amounts of IRS-2 in precipitates were examined by Western blotting using anti-IRS-2 antibody. B, after immunoprecipitation with anti-STAT6 antibody, tyrosine phosphorylations of STAT6 were detected by Western blotting with anti-phosphotyrosine antibody. Equal amounts of STAT6 in precipitates were also examined by Western blotting using anti-STAT6 antibody. phosphorylate SHIP, Jak1(KI) can associate with SHIP, suggesting kinase-independent association. Surprisingly, Jak3(WT) and Jak3(KI) also associate with SHIP in a kinase-independent manner, suggesting that although Jak3 may not phosphorylate SHIP, it may serve recruit SHIP recruitment to ythe IL-4R signaling complex.
To extend the above results, the association of endogenous SHIP with endogenous Jak1 and Jak3 was examined. The potential interaction of SHIP with either Jak1 and/or Jak3 was analyzed in the parental 32D cell line, lacking expression of IRS-1 and IRS-2 (Fig. 7B). Consistent with the above association data in 293T cells, SHIP immunoprecipitates with Jak1 and Jak3 in 32D cells independent of cytokine stimulation, again suggesting that the kinase activity of neither Jak1 nor Jak3 is required for SHIP binding. These data suggest that Jak1 and Jak3 may recruit SHIP to the IL-4 receptor in a cytokine-independent manner. DISCUSSION As with a wide variety of other cytokines, a role for SHIP in IL-4 signal transduction was initially suggested by its tyrosine phosphorylation upon ligand stimulation. Previously, SHIP has exclusively been reported to be a negative regulator of signaling initiated by other cytokines, including IL-3, GM-CSF, CSF, and SCF. Yet it is not clear that SHIP behaves unequivocally as a negative regulatory factor, since evidence exists that SHIP may act as a positive effector as follows: 1) SHIP null mice have a diminished number of B lymphoid cells (25), and 2) SHIP appears to be required for the attenuation of apoptosis upon B cell Fc␥RIIb-BCR cross-linking (27). The mechanisms underlying these two observations remain to be elucidated. The inhibitory nature of SHIP upon signaling cascades activated by cytokines has been founded on studies employing overexpression of SHIP and/or by analysis of homozygous null mice. Although these studies provide important insights into SHIP function, the relative contribution of the domains within SHIP to the observed phenotypes has not been addressed. In particular, the role of the catalytic activity of SHIP in regulation of cytokine remains unknown. Therefore, in this study we have focused on a potential functional contribution of the catalytic activity of SHIP in IL-4-mediated responses.
The data presented within this paper demonstrate that overexpression of wild type SHIP leads to hyperproliferation in response to IL-4, and overexpression of a catalytic inactive mutant form of SHIP (D672A) suppresses IL-4-induced proliferation. As the mutant SHIP(D672A) has complete sequence and domains except the point mutation in catalytic domain, this mutant can suppress the endogenous SHIP protein in a dominant negative manner as described previously (32). Significantly, the catalytic activity of SHIP behaves as a positive regulator of IL-4 proliferation independently of cell death. Taken together, these results are the first example of the catalytic activity of SHIP acting as positive effector of cytokine signaling. The hyperproliferation of 32D cells overexpressing SHIP in response to IL-4 correlates with a down-regulation of PIP 3 levels (Fig. 3). At least at the concentration of IL-3conditioned media tested, there was no alteration in proliferation induced by IL-3. This lack of alteration in IL-3 function does not appear due to a selection bias in these transformants of the IL-3-dependent 32D cell line. In these transformants, IL-3 can induce SHIP phosphorylation and Shc association (Fig. 1, data not shown). Furthermore, in agreement with our results, a previous report has demonstrated that ectopic overexpression of SHIP(WT) in IL-3-dependent cells does not affect cell growth over at least 48 h of culture (33).
The mechanism by which SHIP alters IL-4 induced proliferation remains unclear. This alteration failed to correlate with any observed changes in activation of Akt, p70 S6k , IRS-2, and STAT6 activation. The lack of an effect on the PH-containing proteins Akt and p70 S6k is especially surprising given the current evidence that such proteins lie downstream of PI3-K with PIP 3 levels regulating PH domain-mediated membrane targeting. The lack of altered protein activation cannot be accounted for by failure to alter PIP 3 levels in response to IL-4 as shown in Fig. 3, suggesting that varying PIP 3 levels alone is not sufficient to effect activation of at least these proteins in response to IL-4. These results appear to differ from the alteration in Akt activity observed in response to IL-3 in SHIP null mice (23). It is possible that more than one domain of SHIP contributes to inhibition of Akt activation with the catalytic activity only partially contributing to regulation of Akt activation. Similarly, in IL-4 signaling the catalytic activity of SHIP alone may be insufficient to modulate Akt activation, but it is sufficient to mediate positive effects on proliferation. We detected no effect on IL-4-mediated protection from apoptosis in SHIP(WT) versus SHIP(D672A) cells, consistent with PIP 3 Akt activation seen in response to IL-4. Furthermore, it is entirely possible that Akt activation in the case of IL-4 signaling is mediated primarily independently of PI3-K signals, as there FIG. 7. SHIP interacts and is phosphorylated by Jak kinases. A, Jak kinases involvement in SHIP phosphorylation. 293T cells were transfected with expression plasmids as indicated. Cell lysates were prepared at 40 h post-transfection, and Myc-tagged SHIP was immunoprecipitated (IP) with anti-Myc antibody. The immunoprecipitates were subjected to Western blotting (WB) by anti-phosphotyrosine antibody to monitor the phosphorylation of SHIP. Subsequent reprobing with anti-Jak1, Jak3, and Myc were done sequentially. Expression levels of Jak1 and Jak3 were also examined using total lysates prepared above. B, SHIP associates with Jak1 and Jak3. 32D cells were stimulated with IL-4. Jak1 or Jak3 were immunoprecipitated from these cells, and the immunoprecipitates were separated by SDS-PAGE followed by Western blotting using anti-SHIP antibody. The blots were subsequently reprobed with anti-Jak1 and Jak3 antibody. Normal rabbit serum (NRS) was used as control. are other mechanisms, e.g. cAMP, that directly modulate Akt activation (53).
Taken together, these results suggest that the role of SHIP domains involved in the regulation of signaling activated by different cytokines remains unclear. It is reasonable to postulate that different domains of SHIP may act in concert or antagonistically to integrate cytokine responsiveness. Perhaps the catalytic function may have no impact on signaling depending on the cytokine or pathway. Thus, it cannot be excluded that the catalytic activity of SHIP cooperates with other domains to mediate the IL-4-mediated pathways examined. Similarly, the domains involved in the negative regulation of IL-3 signaling cannot be distinguished in the context of the SHIP(Ϫ/Ϫ) mouse. During preparation of this manuscript it has been reported that B cells from SHIP(Ϫ/Ϫ) mice showed enhanced proliferative responsiveness to co-stimulation by IL-4 and CD40 (54). Our results in this report clearly showed that catalytic activity itself is important for the proliferative effect by IL-4 in at least the 32D myeloid cell line. The exact role of SHIP in IL-4 biology remains unclear.
Many cytokines induce SHIP phosphorylation, yet the tyrosine kinases involved have not been identified. In IL-4 signal transduction, Jak1 is necessary to phosphorylate the two previously recognized downstream signaling effectors, Stat6 and IRS-1/2 (55,56). Analysis of Jak-SHIP interactions suggests that Jak1 and Jak3 associates with SHIP independently of cytokine stimulation. Although SHIP can bind to a putative immunoreceptor tyrosine-based inhibitory motif in the IL-4R␣ chain, 2 this tyrosine is not required for the tyrosine phosphorylation of SHIP. It is possible that SHIP can be recruited to the IL-4 receptor through several redundant mechanisms. Interestingly, the site of Jak1-mediated SHIP phosphorylation remains to be determined, but preliminary experiments suggest that two previously identified NPXY motifs are not required for Jak-induced SHIP phosphorylation, suggesting Jak may regulate uncharacterized SHIP interactions. 2 SHIP may affect the recruitment or activation of other PHcontaining proteins in response to IL-4. Candidates for such proteins include the PH-containing RasGAP-interacting proteins FRIP (p56 Dok-2 ) and p62 Dok-1 recently implicated in IL-4 signaling (57). However, as endogenous p56 Dok-2 and p62 Dok-1 were not detectable by either Western blot analysis or by immunoprecipitation in 32D cells, we are analyzing the stable transformant of these proteins by testing the alteration in potential of the Dok activation in this cell system. Although the functions of Dok proteins have not been defined in IL-4 signaling, the finding that p62 Dok-1 is phosphorylated by the Bcr-Abl and v-Abl oncoprotein suggests it may act in cellular proliferation (58 -60). Although inositol phosphate composition of the membrane can alter the recruitment of PH-containing proteins, other structural motifs such as FYVE may also be sensitive to inositol phosphate composition (19,20). Therefore, the mechanism of SHIP regulation of IL-4-induced signals is likely to involve yet to be identified proteins.