If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
To whom correspondence should be addressed: Div. of Basic Sciences, Fred Hutchinson Cancer Research Center, 1100 Fairview Ave. N., B2-152, P.O. Box 19024, Seattle, WA 98109-1024. Tel.: 206-667-4436; Fax: 206-667-6522;
* This work was supported by Public Health Service Grants CA40987 and CA20551 (to L. R. R.) from the National Cancer Institute.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Fms is a tyrosine kinase-containing receptor for macrophage colony-stimulating factor (M-CSF) that regulates survival, growth, and differentiation of cells along the monocyte/macrophage lineage. M-CSF stimulation of murine myeloid FDC-P1 cells expressing Fms resulted in the tyrosine phosphorylation of a number of signal transduction proteins, including an unidentified 100-kDa protein. This 100-kDa protein associated with the tyrosine phosphatase SHP-2 but not with the related phosphatase SHP-1. The kinetics of tyrosine phosphorylation of p100 and SHP-2 suggest that p100 may be a direct substrate of SHP-2. p100 bound directly to the SH2 domains of both SHP-2 and the p85 subunit of phosphatidylinositol 3′-kinase. The 100-kDa protein did not appear to bind directly to Fms, Ship, Cbl, Shc, or Grb2, although all of these proteins were coimmunoprecipitated with p85 after M-CSF stimulation. Association of p100 with SHP-2 and p85 did not require the major autophosphorylation sites on Fms nor binding of p85 to Fms. A tyrosine phosphorylated protein of 100 kDa also coprecipitated with SHP-2 from several other myeloid cell lines after M-CSF stimulation but was not seen in immunoprecipitates from Rat2 fibroblasts expressing Fms. Stimulation of FDC-P1 cells with additional cytokines also resulted in coprecipitation of a 100-kDa protein with SHP-2. p100 may therefore be a common component of the signaling pathways of cytokine receptors in myeloid cells.
Recent studies on the protein tyrosine phosphatases SHP-1 and SHP-2 have established the importance of these phosphatases to intracellular signaling in hematopoietic cells. These two cytoplasmic tyrosine phosphatases are very similar in structure. Each has two tandem phosphotyrosine-binding SH2
). For instance, SHP-1 binds to the EPO receptor after EPO stimulation and dephosphorylates the receptor-associated tyrosine kinase JAK2. Cells expressing mutant EPO receptors that cannot bind SHP-1 are hypersensitive to EPO and display prolonged autophosphorylation of JAK2 (
The mechanism by which SHP-2 positively regulates signaling is not fully understood. The SH2 domains of SHP-2 may allow it to act in part as an adapter linking two or more signaling proteins. For instance, after binding to the activated platelet-derived growth factor receptor or EPO-R, SHP-2 is itself tyrosine phosphorylated and binds to the SH2 domains of the adapter protein Grb2. Grb2 is constitutively bound to the Ras guanine nucleotide-releasing factor Sos, and by linking Grb2 and the receptor, SHP-2 effectively recruits Sos to the membrane where it can activate Ras (
). However, the phosphatase activity of SHP-2 is also a necessary component of these pathways, because overexpression of a catalytically inactive SHP-2 has been shown to block mitogen-activated protein kinase activation in response to insulin (
) are important for protein interactions. The SH2 domains of Grb2 have been shown to bind to tyrosine 697. Grb2 binding to Fms probably results in translocation of Sos to the membrane and subsequent Ras activation (
), thereby activating PI 3′-kinase-mediated signaling events. Tyrosine 807 within the kinase catalytic domain of Fms does not bind any known signaling proteins, but is important for c-myc induction in NIH 3T3 cells (
In this report, we have investigated the roles of SHP-1 and SHP-2 in Fms signaling in FDC-P1 cells. We find that SHP-2 binds to a novel 100-kDa tyrosine phosphorylated protein after stimulation by M-CSF and several other cytokines. This 100-kDa protein associates directly with the SH2 domains of both SHP-2 and the 85-kDa subunit of PI 3′-kinase and is associated with Fms, Ship, Cbl, Shc, and Grb2 in p85 immune complexes.
To look for a potential role for tyrosine phosphatases SHP-1 and SHP-2 in Fms signaling in myeloid progenitor cells such as FDC-P1, we initially tried to coimmunoprecipitate these phosphatases with antibody to Fms. Using immunoblots with antibodies specific for SHP-1 or SHP-2, neither phosphatase was found to be associated with Fms after M-CSF stimulation (data not shown). Antibodies to SHP-1 or SHP-2 also failed to coimmunoprecipitate Fms. These results are consistent with those of Yi and Ihle (
), who reported that SHP-1 does not associate directly with Fms. Although Fms was not present in immunoprecipitates of these phosphatases, a tyrosine phosphorylated protein of about 100 kDa was coimmunoprecipitated with SHP-2 from FDFms cells after M-CSF stimulation (Fig. 1). The 100-kDa protein was by far the most prominent phosphoprotein associated with SHP-2. In contrast, no tyrosine phosphorylated proteins were detected in SHP-1 immunoprecipitates either before or after M-CSF stimulation.
Time course experiments were performed to look at the kinetics of p100 phosphorylation and association with SHP-2. FDFms cells were treated with M-CSF for 15 s to 4.5 min, and cell lysates were incubated with an antibody that recognized both SHP-1 and SHP-2 tyrosine phosphatases. A phosphotyrosine blot of the time course experiment showed that p100 was phosphorylated within 15 s and was maximally phosphorylated after 1–1.5 min of M-CSF treatment at 37 °C (Fig.2). After about 1.5 min of M-CSF stimulation, p100 appeared to be rapidly dephosphorylated or at least dissociated from SHP-2 (Fig. 2A). A phosphotyrosine blot of whole cell lysates showed that p100 was among the first proteins tyrosine phosphorylated after M-CSF stimulation and again decreased in intensity after about 1.5 min (Fig. 2B). This suggests that p100 was dephosphorylated at this time and not merely dissociated from SHP-2. Interestingly, SHP-2 itself was not tyrosine phosphorylated immediately upon M-CSF stimulation, but was tyrosine phosphorylated after about 1.5 min. Thus, tyrosine phosphorylation of SHP-2 was not required for its association with p100. Tyrosine phosphorylation of SHP-2 just preceded dephosphorylation of p100. Because tyrosine phosphorylation of SHP-2 has been shown to activate its phosphatase domain (
), this suggests that SHP-2 may directly dephosphorylate p100. In contrast to SHP-2, SHP-1 was constitutively tyrosine phosphorylated in these cells (Figs. 1 and 2A), and M-CSF treatment did not increase the level of phosphotyrosine on SHP-1 (Fig. 2).
p100 always appeared as a broad band on SDS-PAGE gels, suggesting that it might be a glycoprotein. However, p100 failed to bind to lectin columns (data not shown), so the broad appearance of the p100 band on SDS-PAGE gels is likely caused by some other modification. The p100 protein also gradually increased in apparent molecular mass with time after M-CSF stimulation. After 15 s of M-CSF stimulation at 37 °C, the protein had an apparent molecular mass of about 90 kDa. After 10 min, the apparent molecular mass had increased to about 100 kDa (Fig. 2A). This increase in molecular mass was not likely due to an increase in tyrosine phosphorylation, because the level of phosphotyrosine seemed to be decreasing at the same time that the molecular mass of p100 was increasing (Fig. 2B). The increase in mass could therefore be due to serine or threonine phosphorylation, ubiquitination, or another modification.
To determine whether p100 was associated with other proteins known to be involved in Fms signaling, antibodies to Fms, Ship, Cbl, p85, SHP-2, Shc, or Grb2 were used to immunoprecipitate these proteins from M-CSF-stimulated or unstimulated FDFms cells, and proteins associated via phosphotyrosine interactions were eluted with phenyl phosphate. As shown in Fig. 3A, a 100-kDa tyrosine phosphorylated protein was coprecipitated with both SHP-2 and with p85. p85 was also associated with tyrosine phosphorylated Fms, Ship, Cbl, and Shc, as well as unphosphorylated Grb2 (Fig. 3A and data not shown); however, antibodies directed against these proteins did not coprecipitate significant amounts of p100. No additional coprecipitating tyrosine phosphorylated proteins were seen when the immune complexes were eluted with SDS sample buffer (data not shown). This suggested that p100 might bind directly to both SHP-2 and p85 but did not associate directly with the other proteins tested.
Tyrosine phosphorylation of p85 was not detectable on this blot (Fig.3A). This was expected, because the level of phosphotyrosine on p85 is usually very low after M-CSF stimulation of FDC-P1 cells.
The level of phosphotyrosine on SHP-2 was highest in the M-CSF-stimulated p85 immunoprecipitates (Fig.3A) and was also visible in M-CSF-stimulated SHP-2 immunoprecipitates eluted with SDS sample buffer (data not shown). The phosphotyrosine blot in Fig. 3A was then reprobed with antibodies to p85 and to SHP-2. Although p85 had not been detected in the phosphotyrosine blot, a significant amount of p85 was present in the SHP-2 immunoprecipitates after M-CSF stimulation, as well as in the p85 immunoprecipitates (Fig. 3B). (It was evident that a fraction of the proteins bound directly to antibodies were also eluted with phenyl phosphate.) Similarly, a significant amount of SHP-2 was pulled down in the p85 immunoprecipitates after M-CSF stimulation (Fig.3C). The SHP-2 present in the p85 immunoprecipitates migrated somewhat more slowly than the SHP-2 that was immunoprecipitated directly. Because tyrosine phosphorylation of the SHP-2 protein was detected in the p85 immunoprecipitates but not in the SHP-2 immunoprecipitates, this suggests that the subset of SHP-2 associated with p85 migrated more slowly because it was tyrosine phosphorylated.
To determine which domains of SHP-2 and p85 were required for association with p100, GST fusion proteins containing different domains of SHP-2 or p85 were prepared. Fusion proteins bound to glutathione-agarose beads were incubated with lysates of M-CSF-stimulated FDFms cells, washed extensively with Nonidet P-40 lysis buffer and eluted with phenyl phosphate. p100 coprecipitated with a GST fusion protein containing both N-terminal and C-terminal SH2 domains of SHP-2 and was efficiently eluted with phenyl phosphate (Fig.4A). However, GST fusion proteins containing only the N-terminal or C-terminal SH2 domain of SHP-2 failed to bring down detectable amounts of p100. This suggests that the association between p100 and SHP-2 may be stabilized by the binding of two phosphotyrosines on p100 to the two SH2 domains of SHP-2. p100 was also coprecipitated with the SH2 domains of p85 and was easily eluted with phenyl phosphate. Either the N-terminal or the C-terminal SH2 domain of p85 was sufficient for association with p100 (Fig. 4B).
The results using GST fusion proteins suggested that tyrosine phosphorylated p100 may bind directly to the SH2 domains of SHP-2 and p85. However, GST fusion proteins might pull down proteins that bind indirectly along with proteins that bind directly. We therefore used far Western blots to demonstrate direct binding of p100 to the SH2 domains of SHP-2 and p85. p85 immunoprecipitates were transferred to nitrocellulose, and the membranes were incubated with GST or GST-SH2 domain fusion proteins. As shown in Fig. 5, both SHP-2 and p85 SH2 domains bound directly to the p100 band that coimmunoprecipitated with p85.
Tyrosine phosphorylated p100 could also be coimmunoprecipitated with SHP-2 from cells expressing Fms mutants lacking four major Fms autophosphorylation sites (Tyr697, Tyr706, Tyr721, or Tyr807). Tyrosine phosphorylation of Fms at these sites was not required for phosphorylation of p100 nor for association of p100 with SHP-2 (Fig. 6A). p100 also associated with p85 in cells expressing Fms Y721F (Fig.6B). This mutant lacks the autophosphorylation site required for association of p85 with Fms (
). Therefore, in cells expressing Fms Y721F, p85 did not coprecipitate Fms, and p85 was not tyrosine phosphorylated. This indicates that tyrosine phosphorylation of p100 and association of p100 with p85 does not require the tyrosine phosphorylation of p85 nor binding of p85 to Fms.
In addition to FDC-P1 cells, several other myeloid cells lines were examined to see whether p100 was a common component of M-CSF signaling. The myeloid progenitor cell lines FDC-P1, 32D, and EML (
) (each expressing an exogenous murine c-fms gene) and BAC1.2F5 (a more differentiated macrophage-like cell line expressing the endogenous c-fms gene) were stimulated with M-CSF and immunoprecipitated with antibody to SHP-2. Coprecipitating proteins were detected by blotting with antibody to phosphotyrosine. In all of the myeloid cell lines tested, a tyrosine phosphorylated p100 protein was coimmunoprecipitated with SHP-2 (Fig. 7). However, no p100 was detected after coprecipitation with SHP-2 in Rat2 fibroblasts expressing murine c-fms. Therefore, p100 appears to be a common component of M-CSF signaling in myeloid cells but not in all cell types.
Different cytokine and growth factor receptors use many of the same proteins in their signaling pathways. To determine whether p100 might be involved in signaling through other receptors, FDFms cells were stimulated with stem cell factor, IL-3, GM-CSF, or M-CSF and then immunoprecipitated with antibody to SHP-2. A phosphotyrosine blot of the immunoprecipitated proteins showed that a tyrosine phosphorylated protein of 100 kDa was coprecipitated with SHP-2 after stimulation with each of these cytokines (Fig. 8). (Much more tyrosine phosphorylated p100 was detected after stimulation with M-CSF, probably because these cells express much higher levels of Fms than the other receptors.) These results suggest that p100 is involved in the signaling pathways of multiple cytokines.
We have characterized a novel 100-kDa tyrosine phosphorylated protein that bound to the SH2 domains of both SHP-2 and p85 in response to M-CSF and several other cytokines. The 100-kDa protein that bound to SHP-2 is very likely the same protein pulled down with p85, because a GST-SHP-2 fusion protein bound in far Western blots to the p100 protein that coimmunoprecipitated with p85. A tyrosine phosphorylated p100 protein was also coimmunoprecipitated with SHP-2 from a variety of Fms-expressing myeloid cells and in response to stem cell factor, IL-3, and GM-CSF as well as M-CSF. However, because we do not yet have an antibody to p100, we cannot be certain that the p100 protein seen in other myeloid cells and in response to other cytokines is identical to the p100 that bound to SHP-2 and p85 in M-CSF-stimulated FDFms cells. Nevertheless, our results suggest that p100 could be a common component of signaling from multiple receptors in hematopoietic cells. Although a tyrosine phosphorylated p100 was seen in all the myeloid cells examined, the p100 protein was not detected in SHP-2 immunoprecipitates from Rat2 fibroblasts expressing Fms. We have shown previously that the 145-kDa Ship protein is tyrosine phosphorylated in response to M-CSF in myeloid cells but not in fibroblasts (
). Our results with p100 further illustrate that Fms signaling in myeloid cells is at least in part different from that in fibroblasts.
p100 bound directly to the SH2 domains of SHP-2 and p85 but did not bind to immunoprecipitated SHP-1 or to GST fusion proteins containing the SH2 domains of SHP-1 (Figs. 1 and 5). Thus, two very similar tyrosine phosphatases, although simultaneously expressed in the same cells, do not interact with the same signaling protein. Binding to different signaling intermediates may be one mechanism by which SHP-1 and SHP-2 exert different effects on cellular signaling.
The consensus sequences for binding to the SH2 domains of SHP-2 and P85 are significantly different. Binding to the SHP-2 N-terminal SH2 domain is favored by Ala, Iso, Leu, or Val at the +1 and +3 positions relative to the phosphotyrosine (
). It is likely therefore that SHP-2 and p85 bind to different phosphotyrosines on p100. Whereas either the N-terminal or C-terminal SH2 domain of p85 was sufficient to bind p100, both N- and C-terminal SH2 domains of SHP-2 were required for efficient binding to p100. The reason for this difference is unknown. We cannot exclude the possibility that the GST fusion proteins containing single SH2 domains of SHP-2 were improperly folded. However, these fusion proteins were soluble, undegraded, and very similar in sequence to single SHP-2 SH2 domain constructs used by Tauchi et al. to show binding of individual SH2 domains of SHP-2 to the EPO-R, c-Kit, and Grb2 (
). The two SH2 domains of SHP-2 might therefore bind to two different phosphotyrosines on a single p100 protein.
p100 appears to be a novel signaling protein. The 100-kDa protein did not cross-react with antibodies to focal adhesion kinase, rasGAP, the 110-kDa catalytic subunit of PI 3′-kinase, the adapter protein Eps8, the insulin receptor β subunit, Vav, Fps/Fes kinase, or STAT1 (data not shown). Other investigators have seen similar tyrosine phosphorylated proteins after M-CSF stimulation of myeloid cells, but these proteins were not further characterized. Sengupta et al. (
) found that a 99-kDa protein is among the first proteins tyrosine phosphorylated after M-CSF stimulation of BAC1.2F5 cells. Intriguingly, this p99 is more susceptible to dephosphorylation at 37 °C than most other phosphoproteins. This might be particularly true for a phosphoprotein that is bound to a tyrosine phosphatase. Kanagasundaram et al. (
) saw a 95-kDa tyrosine phosphorylated protein that coprecipitated with p85 after M-CSF stimulation of BAC1.2F5 cells. As with the p100 protein described here, this 95-kDa protein exhibits decreasing mobility on SDS-PAGE gels with time after M-CSF treatment. Welham et al. (
) found that after M-CSF stimulation of FDMAC11/4.5 cells, a tyrosine phosphorylated protein of 100 kDa coimmunoprecipitates with p85. These p99, p95, and p100 proteins could well be the same as the p100 protein characterized in this report.
A number of SHP-1 or SHP-2-binding proteins have been described that are probably not identical to our p100. Su et al. (
) recently described a 95-kDa protein that is tyrosine phosphorylated after epidermal growth factor receptor stimulation of 293 cells. Unlike our p100 protein, however, this p95 associates with the SHP-1 tyrosine phosphatase. SHPS-1, a recently cloned 115–120 kDa protein, is tyrosine phosphorylated in response to insulin and other mitogens in v-src-transformed rat fibroblasts (
). This is a transmembrane glycoprotein that binds to the SH2 domains of both SHP-1 and SHP-2. SHPS-1 lacks a YXXM consensus sequence, however, and so probably cannot bind to p85. A 115-kDa protein with characteristics similar to SHPS-1 has also been seen in Chinese hamster ovary cells after stimulation with insulin (
). In NIH 3T3 cells expressing the insulin receptor, a tyrosine phosphorylated protein of 120 kDa associates with the SH2 domains of SHP-2 after insulin stimulation. A 115-kDa tyrosine phosphorylated protein also coimmunoprecipitates with SHP-2 from HepG2 and NIH 3T3 cells stimulated with epidermal growth factor (
). In the latter case, however, the 115-kDa protein does not coprecipitate with the SH2 domains of SHP-2. It seems likely that many more potential SHP-2 substrates will be found, and these may include several different families of SHP-2-binding proteins.
The Drosophila homolog of SHP-2, Corkscrew, has been shown to bind a 115-kDa cytosolic protein called DOS (
). Upon activation of the Sevenless, Torso, or DER receptor tyrosine kinases, DOS is tyrosine phosphorylated and binds to the SH2 domains of Corkscrew. Binding of DOS to p85 has not been demonstrated, but a YXXM motif is present in DOS (
We do not know whether our p100 protein is similar to DOS, but the data presented here allow us to make some predictions about its interactions with other signaling proteins in myeloid cells (Fig. 9). Tyrosine phosphorylation of p100 is among the earliest events after M-CSF stimulation of Fms. p100 is probably tyrosine phosphorylated by a kinase other than Fms, because it is phosphorylated after stimulation by several different cytokines and in cells that do not express Fms (Fig. 8 and data not shown). Phosphorylation of p100 allows it to bind to the SH2 domains of SHP-2 and p85. Tyrosine phosphorylation of p100 and binding of p100 to SHP-2 precedes phosphorylation of SHP-2. However, the SHP-2 that is coprecipitated with p85 seems to be more highly tyrosine phosphorylated than the bulk of SHP-2 in the cell. This suggests that binding of the SHP-2·p100 complex to p85 results in tyrosine phosphorylation of SHP-2. Because p85 binds directly to Fms, binding of the SHP-2·p100 complex to p85 could bring SHP-2 into the vicinity of the receptor and perhaps facilitate phosphorylation of SHP-2 by Fms. The p85 immunoprecipitates also include a number of other signaling proteins, including Fms, Ship, Cbl, Shc, Grb2, and Sos (Fig.3 and data not shown). The rapid dephosphorylation of p100 immediately following phosphorylation of SHP-2 suggests that SHP-2 may dephosphorylate p100 directly. The dephosphorylation of p100 occurs relatively rapidly after M-CSF stimulation at a time when the level of phosphotyrosine on some cellular proteins is still increasing (Fig.2B). It is possible that p100, like DOS inDrosophila, is a required positive effector of receptor-mediated signaling. If so, then dephosphorylation of p100 by SHP-2 could be a prerequisite for the formation of subsequent signaling complexes. Additional experiments using antibodies to p100 will be necessary to address these possibilities.
We thank Zhizhuang Zhao for helpful discussions and Ed Giniger, Susan Geier, and David Lucas for critiquing the paper.