INTRODUCTION

Although the intracellular signaling pathways of insulin receptor and insulin-like growth factor-1 (IGF-1)¹ receptor have been well investigated, those of other members of insulin receptor family including leukocyte tyrosine kinase (LTK) are poorly understood. LTK is a receptor tyrosine kinase, which belongs to the insulin receptor family and is mainly expressed in pre-B cell, brain, placenta, and several hematopoietic cell lines (1, 2, 3, 4). Recently ALK tyrosine kinase, which has 64% homology with LTK in the amino acid sequence, was cloned by the positional cloning strategy from the breakpoint of the t(2;5)(p23;q35) chromosomal translocation observed in anaplastic large cell lymphomas (5). This gene rearrangement results in generation of fusion protein composed of nucleophosmin and the catalytic domain of ALK. LTK and ALK are structurally closely related to each other and may form a new subfamily in the insulin receptor superfamily.

We previously analyzed the function of LTK by using chimeric receptors composed of the extracellular domain of epidermal growth factor receptor (EGFR) and the transmembrane and the cytoplasmic domains of LTK, and showed that 293 cells which stably express the chimeric receptor display an enhanced growth rate in response to EGF. EGF induces the autophosphorylation of the chimeric receptor and its association with Shc, Grb2, and Sos, but the interaction of the Grb2-Sos complex with LTK is mediated by Shc, which binds to tyrosine 862 at the carboxyl-terminal domain of LTK (6). In this study, we showed that LTK also utilizes IRS-1 as a substrate. While the major binding site for Shc is tyrosine 862 at the carboxyl-terminal domain, tyrosine 485 at the juxtamembrane domain of LTK is responsible for IRS-1 phosphorylation. Both of these tyrosines are located in the NPXY motifs that are recently identified as the consensus sequence of binding sites for the PTB domains of Shc and IRS-1 (7, 8, 9, 10).

Many growth factors such as insulin (11), IGF-1 (12, 13), interleukin 4 (IL4) (14), IL9 (15), IL13 (16), and growth hormone (17, 18) have been shown to utilize insulin receptor substrate-1 (IRS-1) as a substrate. Although IRS-1-deficient mice showed mild growth retardation and glucose intolerance (19, 20), recently identified insulin receptor substrate 2 (IRS-2) is suggested to compensate the function of IRS-1 (21). Thus, it is revealed to be difficult to analyze the biological function of IRS-1 by using IRS-1-disrupted mice. Recent studies have shown that the phosphotyrosine-binding (PTB) domains of IRS-1 and Shc bind to the NPXY motif located at the juxtamembrane domain of insulin receptor  subunit (7). The mutation at tyrosine 960 in the NPXY motif abolishes phosphorylation of both IRS-1 and Shc (22, 23). In the insulin receptor signaling system, therefore, it is difficult to analyze the IRS-1-specific signals. In contrast, the binding sites of LTK for Shc and IRS-1 are separated; therefore, we have found that the LTK signaling system provides a useful tool to investigate the function of IRS-1 and Shc distinctively.

By comparing the growth of cells that express mutant chimeric receptors containing a mutation at each NPXY site, we showed that both NPXY sequences, which are binding sites for Shc and IRS-1, contribute equally to activation of the Ras pathway and generation of mitogenic signals, while only tyrosine 485, which is critical for IRS-1 phosphorylation, generates cell survival signal. In this respect, at least in LTK signaling, IRS-1 is suggested to play a critical role in preventing cells from apoptotic death. Our results provide a useful insight in understanding the distinctive roles of Shc and IRS-1 in the signal transduction system of the insulin receptor family.

MATERIALS AND METHODS

Anti-LTK monoclonal antibody 1D3-1, which recognizes the carboxyl-terminal domain of LTK, was produced as described previously (4). The monoclonal antibody directed against the extracellular domain of EGFR (Ab-1) and anti-Grb2/Ash monoclonal antibody (3F2) were purchased from Oncogene Science Inc. and MBL Inc., respectively. The rabbit anti-Shc antibody and the anti-phosphotyrosine monoclonal antibody 4G10 were purchased from Upstate Biotechnology Inc. Anti-IRS-1 antibody B51 was provided by T. Kadowaki and J36 was a gift from M. Nishiyama. Anti-Syp antibody was a gift from T. Pawson.

Human embryonic kidney fibroblast 293 cells (Japanese Cancer Research Resources Bank; CRL1573) were grown in Dulbecco's modified Eagle's medium containing 10% fetal calf serum (FCS). Ba/F3 cells were cultured in RPMI 1640 containing 10% FCS and 0.25 ng/ml mouse interleukin 3.

The EGFR-LTK chimeric receptor (EL) cDNA was constructed by ligating the extracellular domain of human EGFR with the transmembrane and the cytoplasmic domains of LTK. Tyrosine-phenylalanine LTK mutants Y485F, Y721F, Y753F, Y779F, Y862F, and Y485F/Y862F were generated as described elsewhere (6). To construct mutant chimeric receptors, the HindIII-EcoRV fragments of the mutant LTK cDNAs which encode the cytoplasmic domain were substituted for that of wild-type EL receptor cDNA.

Transfection into 293 cells was carried out according to the protocol of Chen and Okayama (24). Eighteen hours after transfection, cells were washed once with Dulbecco's modified Eagle's medium and cultured in fresh medium containing 5% FCS for 24 h.

Retrovirus vector was used to transfect cDNAs for chimeric receptors into Ba/F3 cells. Replication-deficient retroviral stocks were prepared by transient hyperexpression in COS7 cells. These constructs were transfected together with -packaging plasmid by the DEAE-dextran method (25). Viral infections were performed by exposing cells to virus stocks with 8 µg of Polybrene/ml at 37 °C for 12 h, and G418-resistant populations were selected in the presence of 1000 µg of G418/ml for at least 3 weeks following 2 days of infection.

Prior to stimulation, cells were starved in RPMI 1640 containing 0.5% FCS for 10 h. Cells were then stimulated with 200 ng/ml EGF for 5 min at 37 °C, washed twice with ice-cold phosphate-buffered saline, and lysed in Triton lysis buffer (0.5% (v/v) Triton X-100, 50 mM Tris-HCl, pH 7.4, 2 mM phenylmethylsulfonyl fluoride, 10 units/ml aprotinin, 1 mM sodium orthovanadate, 1 mM EDTA). Cell lysates were centrifuged and the supernatant was collected. For analysis of total cellular proteins, SDS sample buffer was added directly to lysate and the mixture was denatured for 5 min at 95 °C and subjected to SDS-polyacrylamide gel electrophoresis (SDS-PAGE). Immunoprecipitations were performed at 4 °C for 3 h with specific rabbit or mouse antibodies coupled to the protein A-Sepharose beads. Immunoprecipitates were washed five times in the wash buffer (0.1% (v/v) Triton X-100, 50 mM Tris-HCl pH 7.4), resuspended in SDS sample buffer, and denatured for 5 min at 95 °C prior to loading on the gel. Proteins separated on SDS-PAGE were transferred electrophoretically to a polyvinylidene difluoride membrane (Immobilon, Millipore). The filters were preincubated for 1 h with 1% bovine serum albumin in Tris-buffered saline-Triton X-100 (TBST) buffer (10 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.05% Tween 20), incubated for 2 h at room temperature with the specific antibody, washed three times with TBST buffer, and incubated for 1 h with the alkaline-phosphatase-conjugated goat anti-mouse or goat anti-rabbit antibody. The color reaction was performed with the ProtoBlot system (Promega). For in vitro MAP kinase assay, cell lysates were immunoprecipitated with C92, anti-MAP kinase antibody and the immunoprecipitates were subjected to the in vitro kinase reaction using myelin basic protein (MBP, Sigma) as a substrate, as described previously (26).

To examine a time-dependent cell growth rate, 2 × 10⁵ cells/ml were cultured in RPMI 1640 containing 10% FCS with or without EGF (20 ng/ml) and were counted every 2 days.

For the [³H]thymidine incorporation assay, cells grown in medium containing serum and mouse interleukin 3 (mIL3) were starved in medium containing 0.5% of FCS without mIL3 for 2 h. Then 20 ng/ml EGF were added to medium and incubated for 18 h. Then cells were labeled with 1 mCi of [³H]thymidine for 3 h after growth factor stimulation. The amount of nucleotide incorporated into DNA was quantitated by scintillation counting. Each experiment was repeated at least three times, and a growth rate was expressed as a ratio over the basal values at no EGF stimulation.

To extract the fragmented DNA observed during apoptotic death, 5×10⁶ cells were collected and lysed with 400 µl of lysis buffer (10 mM Tris-HCl, pH 7.5, 10 mM EDTA, 0.2% Triton X-100). The lysate was kept on ice for 10 min and centrifuged at 15,000 rpm for 10 min. The supernatant was subjected to extraction with phenol/chloroform/isoamyl alcohol (25;24;1) followed by ethanol precipitation. The precipitate was dissolved in 20 µl of TE and incubated for 1 h with RNase A (2 µg/ml) at 37 °C. DNA fragments were separated by 2% agarose gel electrophoresis and visualized by ethidium bromide staining.

RESULTS

Recently, it was reported that the PTB domain of IRS-1 binds to the NPXY motif encompassing tyrosine 960 of the insulin receptor (7). Because LTK belongs to the insulin receptor family, we hypothesized that LTK also utilizes IRS-1 as a substrate. To test this hypothesis, we constructed chimeric receptors composed of the extracellular domain of EGF receptor and the transmembrane and cytoplasmic domains of LTK, whose ligand is not identified yet. Since the expression level of IRS-1 in EL3-3, a stable line of 293 cells transfected with the chimeric receptor (6), was low (data not shown), we transiently expressed IRS-1 in EL3-3 cells. Cell lysates in the presence or in the absence of EGF were immunoprecipitated with anti-IRS-1 antibody, followed by immunoblotting with anti-phosphotyrosine antibody. IRS-1 in EL3-3 cells stimulated with EGF was found to be phosphorylated on tyrosine, whereas IRS-1 in mock cells was not (Fig. 1A), indicating that IRS-1 is a substrate of an LTK-mediated signal.

The evidence that the PTB domain of IRS-1 binds to the NPXY motif located at the juxtamembrane domain of insulin receptor led us to hypothesize that the NPXY motif located at the juxtamembrane domain of LTK is responsible for phosphorylation of IRS-1. To examine this, we transiently expressed IRS-1 and LTK in 293 cells. In this system, LTK was autophosphorylated and, at the same time, induced phosphorylation of IRS-1 (Fig. 1B). We then introduced several LTK mutant cDNAs instead of the wild-type LTK cDNA with IRS-1 cDNA. These mutants include receptors containing point mutations at tyrosines 485, 721, 753, 779, and 862 or at both tyrosines 485 and 862 to phenylalanine (Y485F, Y721F, Y753F, Y779F, Y862F, and Y485F/Y862F) or at lysine 544 to methionine (K544M). The K544M LTK mutant is known to be a kinase-inactive form of LTK (27). The expressions of these LTK mutants were confirmed by the immunoblot with anti-LTK monoclonal antibody, 1D3-1. As expected, all these mutants were autophosphorylated on tyrosine residues except for the K544M mutant. In this experiment, Y485F, Y485F/Y862F, and K544M LTK mutants could not phosphorylate IRS-1, although the expression levels of IRS-1 were similar to one another (Fig. 1B). These results indicate that phosphorylation of IRS-1 is dependent on the LTK kinase activity and that tyrosine 485 located at the juxtamembrane domain of LTK is responsible for the phosphorylation of IRS-1.

The binding motifs for the PTB domains of IRS-1 and Shc are aligned in Fig. 1C. Although both of the PTB domains recognize the NPXY sequence, there should be specificity for them. For example, the PTB domain of Shc specifically binds to the NPXY site of TrkA (Tyr-490), while the PTB domain of IRS-1 specifically recognizes the NPXY sequence of the  chain of IL4 receptor (Tyr-497) (7). From our data, we propose that, except for the NPXY sequence, leucine or isoleucine at 8 position seems critical for the specific binding to the PTB domain of IRS-1. Moreover, as Batzer et al. proposed (28), leucine or isoleucine at 5 position is likely to be important for the PTB domain of Shc, but in contrast to this notion, the NPXY motif of insulin receptor, which is the Shc binding site, does not possess leucine or isoleucine at 5 position (Fig. 1C).

To test whether IRS-1 in EL3-3 cells treated with EGF can transmit signals, we examined association between IRS-1 and several signaling molecules that have been shown to associate with IRS-1 in the insulin receptor signaling system. The immunoprecipitates with anti-IRS-1 antibody were subjected to the immunoblotting with appropriate antibodies. In this experiment, Grb2 and Syp were found to associate with IRS-1 in vivo in a ligand-dependent manner (Fig. 2). These data suggest that IRS-1 phosphorylated by LTK utilizes the same signaling molecules as were used in the insulin receptor system.

The signaling pathway downstream of LTK is summarized in Fig. 3. Two independent signaling pathways through LTK activate Ras. Shc binds to tyrosine 862 of LTK and connects LTK and the Grb2-Sos complex in a growth factor-dependent manner (6). Tyrosine 485 of LTK is essential for phosphorylation of IRS-1. IRS-1, then, associates with Syp and Grb2 in a ligand-dependent manner. In this schema, both Shc and IRS-1 can contribute to activation of the Ras pathway, because they associate with Grb2, which is an adaptor molecule linking receptor tyrosine kinases with the Ras signaling pathway. This schema has some similarity to that of insulin receptor in that it utilizes IRS-1 and Shc as major substrates, whereas the difference is that binding sites in LTK for IRS-1 and Shc are separated.

To investigate the roles of tyrosines 485 and 862 in the LTK signaling pathway, we constructed mutant EGFR-LTK chimeric receptors in which each tyrosine or both are substituted for phenylalanine. These mutant receptors were designated as EL-Y485F, EL-Y862F and EL-Y485F/Y862F (Fig. 4A). Stable transfectants of Ba/F3, a mouse interleukin 3 (mIL3)-dependent cell line derived from a mouse pro-B cell that does not express endogenous EGFR, were produced by retrovirus vectors carrying the wild-type and mutant cDNAs (EL, EL-Y485F, EL-Y862F, and EL-Y485F/Y862F). We confirmed the expression of the 140-kDa chimeric receptor by a combination of immunoprecipitation with anti-EGFR antibody and immunoblotting with 1D3-1. The expression levels of these chimeric receptors were approximately similar among a series of the mutants (Fig. 4B). To investigate the ability of EGF to induce autophosphorylation of these receptors in Ba/F3 cells, the stable transfectants were starved for 10 h in medium containing 0.5% FCS, and stimulated with 200 ng/ml EGF for 5 min at 37 °C. Cells were then lysed and immunoprecipitated with anti-EGFR antibody and subjected to the immunoblotting with anti-phosphotyrosine monoclonal antibody 4G10. As shown in Fig. 5A, 140-kDa tyrosine-phosphorylated proteins were detected in cells expressing chimeric molecules when stimulated with EGF, whereas no tyrosine-phosphorylated proteins were detected in mock cells treated with EGF.

To confirm the expressed receptors possess the mutation at the Shc binding site, we examined the ability of these receptors to phosphorylate Shc. As expected from the receptor constructions, EL-Y862F and EL-Y485F/Y862F receptors could not phosphorylate Shc on tyrosine residues (Fig. 5B).

We examined the mitogenic activities of these stable transfectants cultured in the medium containing EGF but not mIL3 using the thymidine incorporation assay in comparison with that of mock cells. The mitogenic activity of EL cells was enhanced in response to EGF. But the incorporation of thymidine into EL-Y485F/Y862F cells were reduced to the level of mock cells. The mitogenic activities of EL-Y485F cells and EL-Y862F cells were approximately the same but significantly lower than that of EL cells (Fig. 6A). Next, we investigated the ability of these mutant receptors to phosphorylate MAP kinase. In this experiment, EGF treatment of EL cells resulted in phosphorylation of MAP kinase on tyrosine residues, whereas MAP kinase was not phosphorylated in mock cells. The degree of tyrosine phosphorylation of MAP kinase in EL-Y485F cells and in EL-Y862F cells were approximately the same but lower than that of EL cells. However, EL-Y485F/Y862F receptor could not phosphorylate MAP kinase at all (Fig. 6B). Furthermore, the kinase activity of MAP kinase in these cell lines was evaluated by an in vitro kinase assay using immunoprecipitates with anti-MAP kinase antibody and MBP as a substrate. We confirmed that the result was consistent with the level of tyrosine phosphorylation of MAP kinase (Fig. 6C). These data indicate that the EGFR-LTK chimeric receptor can activate the Ras pathway and generate mitogenic signals in Ba/F3 cells, whereas the mutant receptor in which two NPXY motifs were disrupted cannot. In summary, these data suggested that, in LTK signaling, IRS-1 and Shc contribute equally to the mitogenic signals and the activation of the Ras pathway.

Since IRS-1 and Shc use the different binding sites of LTK, we can directly compare the roles of Shc and IRS-1 in the cell growth signals by analyzing the biological feature of EL-Y485F and EL-Y862F cells. We checked the growth and morphology of the stable transfectants in the medium containing EGF but not mIL3. In this experiment, mock cells immediately died within 4 days because Ba/F3 cells cannot survive without mIL3. In contrast, EL cells continuously proliferated in the presence of EGF even if mIL3 was depleted. EL-Y862F cells proliferated slower than EL cells for 96 h and gradually stopped growth. EL-Y485F cells proliferated for 24 h and then rapidly lost their viability and died. EL-Y485F/Y862F cells showed a similar growth tendency to that of EL-Y485F cells (Fig. 7A). When dying, the cell body and nuclei of EL-Y485F and EL-Y485F/Y862F cells were fragmented and they formed apoptotic bodies (Fig. 7B). The low molecular weight DNAs were extracted from these cell lines cultured in the presence of EGF for 3 days and electrophoresed in an agarose gel. Notably, the ladder pattern was observed for DNAs extracted from mock cells, EL-Y485F cells, and EL-Y485F/Y862F cells but not for those of EL cells and EL-Y862F cells (Fig. 7C). Although the MAP kinase activity of EL-Y485F cells were as same as that of EL-Y862F cells and the mitogenic activities of both cell lines were approximately equal, the survival of EL-Y862F cells were significantly prolonged compared with that of EL-Y485F cells in the presence of EGF. These results indicate that tyrosine 485 of LTK is critical for the anti-apoptotic activity of LTK and suggest that this survival signal is transmitted through IRS-1, because tyrosine 485 of LTK is the binding site for IRS-1. Furthermore, since the MAP kinase activities of EL-Y485F and EL-Y862F were comparable but the survival properties of them were completely different from each other, we can conclude that the survival signal pathway of LTK transmitted via tyrosine 485 is distinct from the Ras pathway.

Since tyrosine 485 of LTK is critical for IRS-1 phosphorylation, we hypothesized that the survival signals transmitted via tyrosine 485 of LTK is generated by IRS-1. However, anti-apoptotic effects of IRS-1 have not been reported so far; therefore, we tried to determine which pathway is responsible for transmitting cell survival signals. Except for the Ras pathway, it is reported that IRS-1 transmits signals activating p70^S6 kinase in the insulin receptor signaling (29). Therefore, we supposed that the p70^S6 kinase pathway may be involved in the survival signal of IRS-1. To test this hypothesis, we cultured EL cells and EL-Y862F cells in the medium containing rapamycin, which is an immunosuppressant known to inhibit the p70^S6 kinase activity (30). In this experiment, although rapamycin impaired the growth of these cell lines to some extent (Fig. 8), most cells remained alive and apoptotic bodies were not formed (data not shown). Thus we conclude that the survival signal pathway transmitted through tyrosine 485 of LTK is distinct from the p70^S6 kinase pathway.

DISCUSSION

Growth factor receptors utilize various sets of signaling molecules downstream of them, and those signaling molecules they utilize may determine the specificity of the biological activities of the growth factors. In this study, we showed that LTK utilize two signaling molecules, Shc and IRS-1. Recent studies have shown that both of them possess the PTB domain at the amino-terminal region and the binding sites for these domains are the NPXY motif (7). The LTK receptor has two NPXY motifs, tyrosine 485 at the juxtamembrane domain and tyrosine 862 at the carboxyl-terminal domain. In consistent with these findings, our data indicate that the major binding sites for Shc is tyrosine 862 (6) and tyrosine 485 is critical for IRS-1 phosphorylation. We also checked if IRS-1 directly binds to LTK in vivo, but failed to demonstrate the association between them (data not shown). This is not surprising because the interaction between insulin receptor and IRS-1 is also difficult to demonstrate by coimmunoprecipitation (31, 32), suggesting that the binding of the PTB domain of IRS-1 with the NPXY motif may be weak or transient. The interaction between insulin receptor and IRS-1 was recently shown by the two-hybrid system (33).

Although recent studies have revealed that IRS-1 and Shc play essential roles in the signal transduction pathway of insulin receptor, their distinctive roles are not well defined. The major reason for this is that the PTB domains of IRS-1 and Shc recognize the same NPXY motif located at the juxtamembrane domain of insulin receptor  subunit (7) and that the mutation at tyrosine 960 in the NPXY motif disrupts phosphorylation of both IRS-1 and Shc (23). Some reports exert that Shc is the predominant molecule to activate the Ras pathway in insulin receptor signaling (34, 35), but Syp associated with IRS-1 is also reported to up-regulate the Ras activity (36). To analyze the distinct roles of Shc and IRS-1 in insulin receptor families, the signaling pathway of LTK provides a useful model because the binding sites for them are separated. Analyses of EL-Y862F cells and EL-Y485F cells indicate that these receptors transmit approximately the same level of signals concerning DNA synthesis and MAP kinase activation. While EL-Y862F cells could survive for more than a week in the medium containing EGF, even when mIL3 was depleted, EL-Y485F cells rapidly lost their viability and died within 4 days. These differences suggest that while both NPXY motifs can transmit signals inducing MAP kinase activation and DNA synthesis, only tyrosine 485 of LTK transmits cell survival signals. Since we showed in this study that tyrosine 485 of LTK is critical for IRS-1 phosphorylation, it is strongly suggested that IRS-1 plays an essential role in generating survival signals of LTK.

Several signaling pathways are known to exist downstream of IRS-1 via p85 subunit of phosphatidylinositol 3-kinase, Syp, Grb2, and Nck (36, 37, 38). None of these signaling molecules, however, have been shown to be related to death-suppressing phenomena. It is reported that IRS-1 transmits signals activating p70^S6 kinase (29). Because our data suggest that the survival signal pathway of IRS-1 is distinct from the Ras pathway, we supposed that the p70^S6 kinase pathway may be involved in the survival signal of IRS-1, but rapamycin, which is known to inhibit the p70^S6 kinase activity, did not affect the survival of EL cells or EL-Y862F cells. Thus, we can conclude that the survival signal pathway of IRS-1 is distinct from the p70^S6 kinase pathway.

Apoptosis is a form of programmed cell death, characterized by chromatin condensation, cytoplasmic blebbing, and DNA fragmentation. Depletion of survival factors is one of the common cause of apoptosis, which probably plays a critical role in regulating cell numbers or eliminating misplaced cells in animal tissues. Interestingly, insulin and IGF-1 are reported to promote cell survival (39, 40). Moreover, IL4 can suppress apoptosis of B cells induced by hypercross-linking surface immunoglobulin receptors (41). How these factors activate the intracellular survival signals remains to be elucidated. Considering our data, together with the fact that insulin receptor, IGF-1 receptor and IL4 receptor utilize IRS-1 (or IRS-2) as a molecules for signaling, it is possible that these factors activate survival signals through IRS-1 (or IRS-2).

Our other proposal in this study is that not only growth signals but also survival signals are required to induce continuous proliferation of animal cells. The EL-Y485F receptor can activate the Ras pathway and induce mitogenesis by EGF stimulation, but apoptotic death of EL-Y485 cells was induced without the survival signals transmitted through tyrosine 485. Although the cell growth signals have been well investigated, the cell survival signals are poorly understood. Recent studies on cytokine signaling also support the idea that growth signals transmitted through the Ras pathway are not enough to sustain the continuous expansion of hematopoietic cells. For example, EGF cannot induce the continuous growth of Ba/F3 cells that stably express EGFR, whereas it can activate the Ras pathway in these cells (42). Interestingly, studies on interleukin 2 receptor signaling system suggest that cell proliferation and survival signals are mediated by at least three distinct pathways. They are the Ras pathway that leads to c-Fos/c-Jun induction, the c-Myc induction pathway, and the Bcl-2 induction pathway (43). We examined the expression level of Bcl-2 protein in EL cells, EL-Y485F cells, and EL-Y862F cells before and after their culture in EGF for 1 or 2 days, but no change was detected (data not shown).

In this study, we have shown that existence of growth and survival signaling pathways transmitted through two distinct NPXY motifs of LTK. Although both NPXY motifs can transmit growth signals, only tyrosine 485 is critical for anti-apoptotic effects of LTK. Our analyses showed that this survival signal is suggested to generated through IRS-1 and to be distinct from the Ras pathway and the p70^S6 kinase pathway. Taken together, all these results suggest the existence of another unidentified signaling pathway downstream of IRS-1, which is relevant to the anti-apoptotic activity. To determine the mechanism by which IRS-1 generates survival signals, further investigation will be required.