The Anti-apoptotic Effect of Notch-1 Requires p56lck-dependent, Akt/PKB-mediated Signaling in T Cells*

The Notch family of transmembrane receptors have been implicated in a variety of cellular decisions in different cell types. Here we investigate the mechanism underlying Notch-1-mediated anti-apoptotic function in T cells using model cell lines as the experimental system. Ectopic expression of the intracellular domain of Notch-1/activated Notch (AcN1) increases expression of anti-apoptotic proteins of the inhibitors of apoptosis (IAP) family, the Bcl-2 family, and the FLICE-like inhibitor protein (FLIP) and inhibits death triggered by multiple stimuli that activate intrinsic or extrinsic pathways of apoptosis in human and murine T cell lines. Numb inhibited the AcN1-dependent induction of anti-apoptotic proteins and anti-apoptotic function. Using pharmacological inhibitors and dominant-negative approaches, we describe a functional role for phosphatidylinositol 3-kinase (PI3K)-dependent activation of the serine-threonine kinase Akt/PKB in the regulation of AcN1-mediated anti-apoptotic function and the expression of FLIP and IAP family proteins. Using a cell line deficient for the T cell-specific, Src family protein, the tyrosine kinase p56lck and by reconstitution approaches we demonstrate that p56lck is required for the Notch-1-mediated activation of Akt/PKB function. Furthermore, the Src tyrosine kinase inhibitor, PP2, abrogated ectopically expressed AcN1-mediated anti-apoptotic function and phosphorylation of p56lck. We present evidence that endogenous Notch-1 associates with p56lck and PI3K but that Akt/PKB does not co-immunoprecipitate with the Notch1·p56lck·PI3K complex. Finally, we demonstrate that the Notch1·p56lck·PI3K complex is present in primary T cells that have been activated in vitro and sustained in culture with the cytokine interleukin-2.

The Notch family of transmembrane receptors are important regulators of cell fate determination events in different cell types. Upon activation by ligands usually located on the surface of neighboring cells, Notch undergoes intramembrane proteolysis and release of an intracellular region resulting in signaling (1). The mechanism of Notch-1 signaling is incompletely understood even in well characterized genetic models like Dro-sophila or the worm. There is recent evidence of the interaction of Notch together with T cell antigen receptor in determination of cell fate within the T cell lineage (2,3). Furthermore, Notch-1 has been shown to protect against anoikis (apoptosis induced by matrix withdrawal) or p53-mediated apoptosis in immortalized epithelial cells (4,5), T cell receptor-induced apoptosis in mature cells (6), and dexamethasone-mediated apoptosis in thymocytes (7).
In mammalian cells apoptosis is initiated via extracellular receptor-mediated apoptotic signaling (8,9) or death pathways that converge on the mitochondrion (10). Both pathways culminate in the activation of caspases, a family of proteases that mediate the orderly dismantling of cells. Death receptors of the tumor necrosis factor (TNF) 1 receptor superfamily trigger apoptosis principally via the activation of procaspase 8/10 (8,9), whereas mitochondrial signaling activates caspase-9. Caspaseinduced apoptosis is regulated by endogenous antagonists like FLICE-like inhibitor protein (FLIP) that prevent processing and anti-apoptotic proteins of the inhibitors of apoptosis protein (IAP) family that block function (11). The release of intermediates from the mitochondrial intermembrane space that occurs as an initial step is blocked by anti-apoptotic members of the Bcl-2 family (12).
Mature T cells develop from immature precursors through complex but ordered processes (13). There is considerable evidence that Notch signaling regulates the earliest stages of T cell commitment and promotes differentiation in the T cell lineage (14 -16). During T cell development, negative selection ensures deletion of autoreactive thymocytes and allows maturation of CD4 ϩ or CD8 ϩ T cells (13). Notch involvement in Bcl-2-dependent survival signaling in immature T cells was first reported in this model (7). Thymic maturation into CD4 ϩ or CD8 ϩ subsets of mature T cells is accompanied by the emigration of these cells to peripheral lymphoid organs. Mature T cells remain susceptible to varied apoptotic stimuli, many of which are regulated via pathways distinct from those in thymocytes (17)(18)(19). In this study we ask if Notch-1 regulates apoptosis in model T cell lines and have attempted to identify the molecular mechanism of Notch-1 dependent antiapoptotic function.
We show that ectopic expression of the intracellular domain of Notch-1 (AcN1) confers protection against diverse stimuli that trigger intrinsic and extrinsic pathways of death in model T cell lines. This anti-apoptotic function, we suggest, can result from the AcN1-induced elevated expression of IAP-2, Bcl-x L and FLIP, which represent three major families of anti-apoptotic proteins. The protective effect of AcN1 was attenuated by the Notch antagonist Numb and disruption of Akt/PKB-dependent signaling. Furthermore, we show that loss of function of the T cell-specific, Src family, non-receptor tyrosine kinase p56 lck compromised Akt/PKB phosphorylation and anti-apoptotic function. This allows us to propose a functional hierarchy of interactions between Notch-1, PI3K, and p56 lck that might operate in T cells.

EXPERIMENTAL PROCEDURES
Cells and Reagents-Jurkat, a T lymphoblastoid cell line of human origin, and 2B4, a murine T cell hybridoma, were used in all experiments. The J.CaM1.6 mutant was obtained from ATCC (Manassas, VA). Activated T cells were generated by stimulation in vitro as described before and sustained in interleukin-2 (20). Activated T cells were used on day 5 after stimulation. All reagents were obtained from Calbiochem unless specified otherwise. The mitochondrial dye JC-1 was obtained from Molecular Probes (Eugene, OR). Recombinant TNF, TRAIL, and antibodies to Bid, Bcl-2, Bcl-x L , and FLIP were obtained from R & D Systems (Minneapolis, MN). Antibodies to Akt, BIM, IAP-1, IAP-2, Jagged, p38MAPK, p56 lck , PI3K, and GFP were from Santa Cruz Biotechnology (Santa Cruz, CA), antibodies to phosphorylated p56 lck or Akt were obtained from Cell Signaling Technology (Beverly, MA) and the monoclonal antibody to GFP was from Clontech, BD Biosciences. Antibodies to the various domains of Notch and their sources are as follows: Notch-1 intracellular domain referred to as NICD HB (clone C17.9C6) and to Notch-1 intracellular domain nucleotides 6658 -7131 (clone bTan 20) were obtained from the Developmental Studies Hybridoma Bank (Iowa City, IA); Notch-1 carboxyl terminus referred to as NICD SC (M-20) and to the Notch-1 extracellular domain, residues 20 -150, referred to as Notch-1 EC SC (H-131), were from Santa Cruz Biotechnology and to residues 1299 -1492 in the extracellular domain of Notch-1, referred to as Notch-1 EC UB (clone 8G10), was from Upstate Biotechnology (Lake Placid, NY).
Plasmids and Transfections-The AcN encoding plasmid pcDNA3-ICTX (AcN1; Ref. 21) was a kind gift of J. Aster (Harvard Medical School, Boston). This construct was used to generate stable transfectants. AcN-IRES-GFP was a gift from Annapoorni Rangarajan (Whitehead Institute for Biomedical Research, Cambridge, MA). In addition, we generated the AcN1-GFP construct for which the AcN1 region (spanning amino acids 1759 -2556) was amplified using the following specific primers: FP, 5Ј-GGAATTCATGCGCAAGCGCCGG-3Ј and RP, 5Ј-CGG-GATCCTCACACGTCTGCCTG-3Ј) and cloned into EcoRI and BamHI sites of pEGFPN1. All three constructs, which principally comprise the intracellular domain of Notch-1, yielded the same results in biochemical and functional assays. The dominant-negative mutant of Akt, AH-Akt, was obtained from Julian Downward (Imperial Cancer Research Fund, UK). The Bcl-2 plasmid was a generous gift of Dr. S. Perwez Hussain (NCI, National Institutes of Health). The construct for Numb (22) was obtained from Weimin Zhong (Yale University). The mouse Lck clone NT-18 was obtained from ATCC (Manassas, VA). The open reading frame of Lck was PCR-amplified and cloned into the ScaI and KpnI sites of the pCruz Myc expression vector (Santa Cruz Biotechnology). The sequences of the primers used for the PCR are 5Ј-AAAAGTACTATGG-GCTGTGTCTGC-3Ј (FP) and 5Ј-GGGGTACCTCAAGGCTGGGGCT-G-3Ј (RP). The sequence of all constructs was verified by automated sequencing.
Stable cell lines expressing neo, AcN1 or Bcl-2 were generated by transfection with 2 g of DNA of appropriate plasmids by electroporation. After 48 h, cells were cultured for 10 -14 days in complete medium supplemented with 80 g/ml G418 for selective growth of transfected cells. Selected cells were not subcloned further. A minimum of two separate sets of stable transfectants have been tested in these experiments. Electroporations were performed as described previously (23). Electroporation usually resulted in transfection efficiencies ranging from 20 -25% and were higher in some experiments. Each experiment was performed a minimum of three times unless indicated otherwise.
Assays of Apoptotic Damage-Apoptotic nuclear morphology, phosphatidylserine exposure, cell lysis, and mitochondrial damage were assessed as described previously (23). All flow cytometric analysis was performed on a BD Biosciences FACS® (BD Biosciences). Apoptotic nuclear damage was assessed in GFP-transfected cells as described before (23).
Immunoprecipitations-Immunoprecipitations were performed in transiently transfected populations. 10 7 cell pellets were lysed in 20 mM Tris (pH 7.5) containing 450 mM NaCl, 1 mM EDTA, 10 g/ml aprotinin and leupeptin, 100 g/ml phenylmethylsulfonyl fluoride, and 1% Nonidet P-40 for 15 min at 4°C. Cell debris was cleared by centrifugation at 14,000 rpm for 30 min, and the resulting supernatant was divided into 2 aliquots, one aliquot incubated overnight with Sepharose G beads (Amersham Biosciences, Uppsala, Sweden) preincubated for 1 h with specific antibody and the other aliquot with beads similarly incubated but with an isotype-matched antibody as a control. Lysates were immunoprecipitated overnight at 4°C by gentle rocking, and then beads with bound proteins (immune complex) were separated by centrifugation at 14,000 rpm for 1 min. Immune complexes were washed in lysis buffer, resolved on SDS gels, and the proteins detected by Western blot analysis. In immunoprecipitations using p56 lck , a polyclonal antibody to Akt (raised in rabbit) from the same company was used as negative control. In experiments where the polyclonal antibody (goat) to Notch intracellular domain was used, a polyclonal antibody (raised in goat) recognizing Bim and from the same company was used as the control group. In other experiments, purified rabbit and goat Ig have been used as negative controls with identical results. In experiments where GFP was immunoprecipitated with a monoclonal to GFP, purified monoclonal mouse Ig (Sigma) was used as a control.
Western Blot Analysis of Proteins-Cell lysates of 0.2-0.5 ϫ 10 6 cells were resolved by SDS-PAGE, and protocols recommended by the manufacturers were used for Western blot analysis of BID, Bcl-x L , FLIP, and Notch. Blots were developed by chemiluminescence using Super Signal (Pierce). p38MAPK was used to establish parity of loading in 10 -12% SDS gels, and Akt or spectrin was used as the loading control (LC) for 8 -10% SDS gels. The term LC has been used to indicate these proteins in the various Western blots.

RESULTS
The murine T cell hybridoma 2B4 and the human lymphoblastoid cell line Jurkat have been used in all experiments. All experiments have been performed using both the stable and transient expression of AcN1 in these cell lines. Stable transfectants of Jurkat and 2B4 were derived as described under "Experimental Procedures." For clarity of presentation, the data presented in the figures were derived from either transient or stable transfections in different experiments.
AcN1 Blocks Apoptosis and Induces Expression of Anti-apoptotic Proteins-The 2B4 T cell hybridoma has been used in many studies as a model system to examine the regulation of apoptosis in mature T cells (24,25). Stable transfectants of 2B4 cells constitutively expressing the intracellular domain of Notch-1 (AcN1) (hereafter referred to as 2B4-AcN1S) had significantly increased levels of the anti-apoptotic proteins Bcl-x L , FLIP L , and IAP-2 ( Fig. 1A). Bcl-2 expression is relatively high in 2B4 cells and was not substantially increased in 2B4-AcN1S cells in comparison with cells transfected with the neomycin gene alone. To test whether elevated levels of these proteins correlated with functional outcomes, apoptotic damage was assessed in 2B4-AcN1S cells (closed symbols or filled bars) or 2B4-neoS cells (open symbols or open bars) treated with dexamethasone, etoposide, or cultured on anti-CD3-coated plates to trigger Fas-ligand-Fas signaling ( Fig. 1, B-D). In all cases 2B4-AcN1S-expressing cells were protected from apoptotic damage, albeit to varying extents. Similar results were obtained on transient transfection of AcN1 (data not shown).
AcN1 Blocks Apoptosis in Jurkat Cells-To confirm that the results described above were not a peculiarity of the 2B4 hybridoma, we ectopically expressed AcN1 in Jurkat, a human lymphoblastoid T cell line. This resulted in elevated levels of Bcl-x L and IAP-2 proteins in Jurkat cells stably transfected with AcN1 (J-AcN1S) as compared with Jurkat cells transfected with the control plasmid (J-neoS) ( Fig. 2A). Endogenous FLIP expression is higher in Jurkat than in 2B4 cells; however, the increase in protein expression induced by AcN1 was significant in both cell lines. Bcl-2 on the other hand was not detectable in untransfected Jurkat cells, and its expression could not be induced by AcN1. We established that our assay could detect Bcl-2 protein by detecting the protein in Jurkat cells that stably express bcl-2 ( Fig. 2A, third lane).
The experiments with the 2B4 cell line indicated that AcN1 modulated mitochondrially regulated apoptotic signaling. Since many apoptotic signaling pathways have been characterized in the Jurkat cell line, we used the cell line to test whether AcN1 regulated apoptosis triggered by extrinsic pathways of death as well. J-neoS and J-AcN1S cells were assessed for susceptibility to apoptosis induced by the death receptor ligands TNF or TRAIL. Apoptosis was triggered by treatment with TNF and cycloheximide (the combination referred to as TNF) in all analysis. J-AcN1S cells were protected from TNFinduced apoptotic nuclear damage (Fig. 2B, solid bars), loss of mitochondrial transmembrane potential, which is a measure of mitochondrial integrity (Fig. 2B, dark, hatched bars) or cell lysis (data not shown). AcN1 also protected cells from different concentrations of TRAIL-induced apoptosis at a level equivalent to that achieved by overexpressing Bcl-2 in these cells (Fig.  2C). Furthermore, on transient transfection with AcN1-GFP, Jurkat cells were protected from apoptosis triggered by etoposide ( Fig. 2D) or the PKC inhibitor, staurosporine, at 20 ng/ml staurosporine but not at the higher concentration of staurosporine (Fig. 2E). In Fig. 2, D and E, J-AcN1-GFP-transfected cells are represented by filled symbols and cells transfected with GFP alone by open symbols. Thus, ectopic expression of AcN1 resulted in increased expression of anti-apoptotic proteins and protection from diverse apoptotic stimuli. We then attempted to identify the mechanism by which AcN1 could initiate these events.
AcN1 Activates PI3K-mediated Signaling-Previous experiments have shown that Notch-1 acts through the PI3K pathway to trigger survival in epithelial cells (4). If a similar mechanism operated in T cells, modifications in intermediates of this pathway would, in principle, be detected upon AcN1 ectopic expression. We used pharmacological inhibitors of PI3K and a dominant-negative inhibitor of the serine-threonine kinase Akt (AH-Akt) to test the functional interactions between Notch signaling and the PI3K pathway in T cells. We detected increased phosphorylation Akt-Thr 308 with no change in levels of total Akt in either 2B4 or Jurkat cell lines transiently transfected with AcN1 (Fig. 3A). In functional assays of apoptosis, LY294002 a pharmacological inhibitor of PI3K (26) reversed the anti-apoptotic function of AcN1 in 2B4 cells (Fig. 3B). TNF-induced apoptosis is substantially lower in J-AcN1S cells (Fig. 3C, hatched bars) as compared with TNF-induced apopto- sis in J-neoS cells (Fig. 3C, black bars). Both LY294002 and the structurally distinct inhibitor wortmannin (27) attenuated AcN1 inhibition of TNF-induced apoptosis (Fig. 3C). In the same experiment, UO126, a cell-permeable inhibitor of MEK-1/2, was without effect. Similarly, TRAIL-induced apoptosis was also lower in cells that stably express AcN1S (Fig.  3D, closed circles) and in Bcl-2 expressing cells (Fig. 3D, closed triangles) as compared with cells that were transfected with neomycin alone (Fig. 3D, closed squares). Expectedly, LY294002 reversed the protective effect of AcN1 but not Bcl-2 inhibition of TRAIL-induced death (Fig. 3D, filled symbols). We verified that LY294002 was not toxic to cells (Fig. 3, C and D,  open symbols). In subsequent experiments we used a dominantnegative form of Akt (AH-Akt) to test the outcome of more specific disruption of Akt signaling on AcN1-inhibition of apoptosis. Apoptosis induced by TRAIL is substantially reduced in cells that express AcN1 (Fig. 3E, AcN1) when compared with those expressing neomycin (Fig. 3E, neo). In this assay, the co-expression of AH-Akt attenuated the anti-apoptotic effect of AcN1 (Fig. 3E, AcN1ϩAH-Akt), indicating that disrupting Akt signaling interfered with the anti-apoptotic function of AcN1.
Ectopic expression of AH-Akt in 2B4 cells also blocked AcN1mediated induction of IAP-2 and FLIP but did not affect the up-regulation of Bcl-x L (Fig. 3F). J-AcN1S cells treated with 20 M LY294002 for 6 -8 h were assessed for the expression of anti-apoptotic proteins. LY294002 blocked Akt phosphorylation and expression of IAP-2 (Fig. 3G). Again, Bcl-x L expression was not modified by LY294002. Taken together, these experiments using genetic and pharmacological approaches suggest that AcN1 triggered activation of a PI3K signaling pathway that culminated in the expression of the anti-apoptotic proteins FLIP and IAP. However, Bcl-x L expression appeared to be regulated independent of Akt signaling.
Numb Blocks AcN1-induced Signaling in T Cells-We used the Notch antagonist Numb to confirm that the observations in the preceding set of experiments were specifically initiated by Notch signaling. As shown in Fig. 4A, co-transfection of Numb disrupted AcN1 inhibition of dexamethasone or etoposide in the 2B4 cell line. Ectopic expression of Numb alone enhanced Fas-mediated apoptosis (triggered by cross-linking the TcR-CD3 complex) in the 2B4 cell line (Fig. 4B) and TRAIL-induced apoptosis in Jurkat cells (Fig. 4C). Furthermore, Numb blocked AcN1-mediated induction of Bcl-x L , IAP-2, and FLIP in both cell lines (Fig. 4, D and E) and substantially reduced the phosphorylation of Akt (Fig. 4F). Using the GFP-tagged construct of AcN1 (AcN1-GFP) we could determine that AcN1 is principally localized to the nucleus, and this distribution was not disrupted in cells that co-expressed Numb (Fig. 4G).
p56 lck Is a Putative Adapter in the AcN1-mediated Activation of Akt-The experiments thus far suggest that AcN1 signals via the PI3K pathway to activate Akt culminating in cell survival. What is the link between Notch-1 and Akt activation? p56 lck is a member of the Src-like family of non-receptor tyrosine kinases associated with T cell receptor signaling. p56 lck is activated when CD4 or CD8 co-receptors are cross-linked and phosphorylates and activates other intracellular substrates such as Akt via PI3K. Ectopic expression of AcN1 resulted in enhanced phosphorylation of p56 lck in Jurkat and 2B4 cells (Fig. 5A). To explore the functional implications of this phosphorylation, if any, we used the isogenic Jurkat cell line that was defective for p56 lck (J.CaM1.6). In J.CaM1.6 cells AcN1 did not induce phosphorylation of Akt or expression of IAP-2 or FLIP (Fig. 5B). TNF-dependent induction of IAP-2 and increased phosphorylation of Akt indicated that the J.CaM1.6 cell line was not defective in the modulation of these proteins if signaled appropriately (Fig. 5C). The cell line was refractory to AcN1-mediated inhibition of etoposide-induced apoptosis (Fig.  5D). However, reconstitution with p56 lck restored the ability of AcN1 to block etoposide-induced apoptotic nuclear damage in these cells (Fig. 5E). In another approach, treatment with the general Src kinase inhibitor, PP2, abrogated the anti-apoptotic effect of AcN1 in Jurkat cells (Fig. 5F). Furthermore, when tested after overnight treatment, at concentrations that abrogated the anti-apoptotic function of AcN1, PP2 also blocked AcN1-induced phosphorylation of p56 lck (Fig. 5F, inset). 14 h after transfection cells were continued in culture as such (open bars) or with 2.5 g/ml (black bars) or 5 g/ml (gray bars) etoposide. After 12 h GFP-positive cells were scored for apoptotic nuclei as described previously (23). E, J.CaM1.6 cells transfected as described in the legend to D with the addition of a group co-transfected with p56 lck . Cells were treated with etoposide for 8 h and apoptotic nuclear damage scored in GFP-positive cells in all transfection groups. The data have been plotted after normalizing to apoptotic damage in untreated cells in each condition. Inset, detection of the Myc-tagged p56 lck construct and p56 lck in cells transfected with AcN1-GFP and Myc-p56 lck (first lane). The second lane represents a lysate from cells transfected with AcN1-GFP alone. F, Jurkat cells were transiently transfected with GFP or AcN1-GFP at concentrations described in the legend to Fig. 2. 10 h after transfection, cells from each transfection group were either continued as such or treated with 5 M PP2 for 45 min. Finally each group was continued in the presence or absence of etoposide for another 8 -10 h before assessing apoptotic nuclear damage as described previously (23). Inset, phosphorylation of p56 lck was determined in GFP and AcN1-GFP cells treated with PP2 for 8 h. Notch-1 Associates with PI3K and p56 lck in AcN1-transfected and -activated T Cells-The transfected construct AcN1 is principally nuclear-localized (Fig. 4G), although PI3K is a membrane-localized signaling complex, as is p56 lck . We hypothesized that if indeed there is a direct interaction of the PI3K⅐p56 lck complex with Notch, then the association must involve Notch-1 localized outside the nucleus, most likely fulllength, endogenous Notch-1. The subsequent experiments were designed to test this possibility. When an antibody to GFP was used to immunoprecipitate AcN1-GFP from transfected cells, it failed to immunoprecipitate PI3K (Fig. 6A). The immunoprecipitated GFP-tagged AcN1 protein (Ͼ116 kDa) can be detected by the antibody (clone C17.9C6) raised to the Notch-1 intracellular region (NICD HB ). The AcN1-GFP protein is, however, not recognized by another antibody (NICD SC ) to the intracellular domain. The NICD SC clone did recognize endogenous protein (third lane, WCL) thereby providing a means to distinguish the transfected gene product (Ͼ116 kDa) from the endogenous protein (Ͼ83 kDa).
Immunoprecipitating endogenous p56 lck from AcN1-GFPtransfected Jurkat cells (Fig. 6B, second and fourth lanes) revealed a complex that contains the p85 regulatory subunit of PI3K but not AcN1-GFP or Akt (Fig. 6B, fourth lane), although these are detected in the whole cell lysate (second lane). The NICD SC antibody, which does not recognize AcN1-GFP, revealed that endogenous Notch-1 is in association with PI3K and p56 lck (Fig. 6B, fourth lane). Expectedly, there was no signal for PI3K or Notch-1 on immunoprecipitating of p56 lck in J.CaM1.6 cells, as these cells do not express p56 lck (Fig. 6B,  first and third lanes). In reverse immunoprecipitations we confirmed that immunoprecipitating endogenous Notch-1 using NICD SC brings down a complex that contains p56 lck and PI3K but not Akt from AcN1-GFP-transfected Jurkat cells (Fig. 6C). Thus, from these experiments we concluded that endogenous Notch-1 associates with PI3K and p56 lck . The complex does not stably associate with Akt or AcN1-GFP.
Does p56 lck normally exist in association with endogenous Notch-1 and PI3K? Indeed, in untransfected Jurkat cells, p56 lck associates with PI3K and Notch-1, although expectedly the levels of Notch-1 are relatively low in these cells (Fig. 6D). Since the phosphorylation of both Akt and p56 lck is consistently enhanced on AcN1 transfection, and cells are protected from apoptosis, we hypothesized that AcN1 may stabilize the Notch-1-p56 lck -PI3K interaction by driving expression of full-length endogenous Notch-1. A similar feedback mechanism has also been reported in other systems (28,29). As already shown, the levels of p56 lck , PI3K, or Akt are not enhanced in AcN1 transfected cells, but the possibility that AcN1 triggers an increase in expression of endogenous Notch-1 was confirmed in subsequent experiments. Jurkat cells stably transfected with AcN1GFP express increased levels of a Ͼ250-kDa Notch-1 species that is detected using an antibody that recognizes an epitope in the Notch-1 extracellular domain (Fig. 6E). Furthermore, the Notch-1 ligand, Jagged, was also elevated in these cells (Fig. 6F), which suggests the possibility of increased ligand-mediated Notch signaling via the endogenous receptor in AcN1-transfected cells. We then tested whether the extracellular domain of Notch-1 is also present in the complex with p56 lck and PI3K. As shown in Fig. 6G, in experiments that replicate the protocol used for B and D, p56 lck immunoprecipitates a complex that contains the Notch-1 extracellular domain detected using two antibodies raised to the extracellular region of Notch-1: Notch-1 EC UB and Notch-1 EC SC . Expectedly, PI3K and NICD are also present in this complex.
Finally we asked whether Notch-1, PI3K, and p56 lck associate in non-transformed T cells. We tested for this association in polyclonally activated T cells sustained in culture with interleukin-2. The cytokine is important for T cell proliferation and survival and some of these effects are dependent on signaling through PI3K. In activated T cells we observed that endogenous Notch-1 (NICD SC ) immunoprecipitated p56 lck and PI3K (Fig. 6H) from untransfected, in vitro activated splenic T cells sustained in culture with the cytokine interleukin-2.

DISCUSSION
Mature T cells are known to express anti-apoptotic proteins to protect themselves from diverse apoptotic stimuli that they encounter in the course of their existence. Thus, when T cells are selected for survival rather than death they must also be equipped with proteins such as FLIP and members of the Bcl-2 family that serve to buffer them from various apoptotic stimuli FIG. 6. Endogenous Notch-1 associates with PI3K and p56 lck in T cells. A, Jurkat cell lysates prepared 8 h after transfection with AcN1-GFP were immunoprecipitated (IP) with mouse Ig (isotype control, first lane) or mouse anti-GFP (second lane) and analyzed by Western blot and sequential strip and re-probe for GFP, NICD SC , NICD HB , and PI3K. B, Jurkat (second and fourth lanes) and J CaM1.6 (first and third lanes) cells were transfected with 5 g of AcN1-GFP and after 10 h cell lysates were immunoprecipitated (IP) with an antibody to p56 lck . Immunoprecipitates were analyzed for the proteins indicated on the right by WBA. The first two lanes are whole cell lysates. C, lysates of Jurkat cells transfected with 5 g of AcN1GFP and cultured for 8 h were immunoprecipitated (IP) with beads coated with goat Ig (isotype control, first lane) or NICD SC , the Notch-1 antibody raised in goat (second lane). Immunoprecipitates were analyzed by WBA for expression of the proteins indicated in the panel on the right of the blots. WCL indicates the whole cell lysate. D, untransfected Jurkat cell lysates were immunoprecipitated (IP) using an antibody to p56 lck as described under "Experimental Procedures." Immunoprecipitates were analyzed for the presence of NICD SC and PI3K by Western blot analysis. E and F, whole cell lysates of Jurkat cells transiently transfected with GFP (first lane) or 5 g of AcN1GFP (second lane) were analyzed for the expression of full-length Notch-1 protein using the bTan20 antibody (E). Spectrin was used to assess parity of loading (LC). In the same experiment cells were also probed for the expression of Jagged (F). G, Jurkat cells transfected with GFP or AcN1-GFP were immunoprecipatated (IP) with the antibody to p56 lck . The immunoprecipitates were resolved on 8% SDS gels and probed for the expression of extracellular Notch-1 using the antibodies described under "Experimental Procedures," NICD SC and PI3K by Western blot analysis. WCL, whole cell lysate. H, day 5 T cell blasts were immunoprecipitated with anti-Notch-1 (second lane) and the immunoprecipitates probed for proteins shown in the figure. The asterisk indicates a nonspecific band that is seen in Western blots (WB) with these antibodies. (13). We propose that the Notch-1 receptor can mediate the induction of anti-apoptotic proteins. We show that ectopic expression of AcN1 resulted in the up-regulation of anti-apoptotic proteins and protected cells from diverse apoptotic stimuli. Our experiments suggest that the anti-apoptotic function of Notch-1 appears not to be restricted to specific pathways but probably functions to confer generalized protection to diverse apoptotic stimuli.
We identify PI3K as one intracellular mediator of AcN1/ Notch-dependent anti-apoptotic activity. Genetic and pharmacological inhibitors of PI3K and/or Akt signaling disrupted AcN1 function identifying Akt as an intermediate critical to this process (30). In T cells, PI3K can be activated by costimulatory molecules like CD28 (31), cell adhesion molecules such as ILK (32), the intracellular adhesion molecule ICAM-2 that activates Src kinases (33) or via cytokines (34). In this study we exploit a mutant cell line to demonstrate that the T cell-specific non-receptor tyrosine kinase p56 lck is one of the mechanisms by which Notch-1 activates PI3K signaling. We confirmed this observation by showing that p56 lck binds Notch-1 and PI3K in T cell lines and non-transformed T cells. We also show that the anti-apoptotic effect of Notch-1 is blocked by the Src kinase inhibitor PP2, indicating that Notch-1 may also interact with Src kinases other than p56 lck .
Since Akt does not co-immunoprecipitate with the Notch-1⅐p56 lck ⅐PI3K complex, how does Akt integrate with the complex? The generation of PIP3 by the Notch-p56 lck -PI3K interaction would be one mechanism that activates Akt. Also possible is the recruitment of adapter molecules, which associate with the Notch-1⅐p56 lck ⅐PI3K complex and indirectly interact with Akt to initiate signaling. A detailed understanding of how the components of signaling pathways intersect to regulate apoptosis requires analysis of the subcellular organization of these macromolecular assemblies.
We demonstrate that PI3K-p56 lck and Notch-1 are also present as a complex in non-transformed T cells. This is consistent with the observation that disruption of any one of these pathways such as Akt or p56 lck results in increased susceptibility to apoptotic signaling (35,36). p56 lck is expressed in T cells and with p59 fyn and ZAP-70 has been implicated in early activation resulting from T cell receptor ligation. p56 lck activity is downregulated in anergic T cells, and this kinase is essential for the proliferation of naive T cells (37). p56 lck phosphorylates PI3K and the tyrosine phosphatase SHP-1 (38). Furthermore, p56 lck and p59 fyn are implicated in regulating the association of CD28 with PI3K and GRB-2 (39), which is critical to stabilizing interleukin-2-dependent signaling following T cell activation. Cytokines regulate many aspects of T cells, functioning not only to modulate proliferative responses but also as regulators of T cell survival.
In view of the current understanding of Notch-1 signaling in T cell survival, further studies are required to address the physiological significance of Notch-1 interaction with PI3K and p56 lck in T cell survival. A possibility consistent with the data is that Notch functions as part of cytokine signaling relays to regulate cell survival in specific spatio-temporal contexts in T cell compartments. Following an antigen response, effector T cells are eliminated by active mechanisms (40). Linked to the loss of the effector pool of T cells is generation of the memory subset (41). While initial events resulting in the generation of memory may be stochastic, long term survival of memory T cells is likely dependent on instructional signals delivered by cytokines (42). Anti-apoptotic proteins like Bcl-2 and Bcl-x L are critical for immune memory, and their expression is also linked to cytokine signaling in T cells (43). Our experiments reveal a functional relationship between Notch-1, p56 lck , and PI3K-dependent activation of Akt, molecules that are key intermediates in T cell survival. Elucidation of the mechanism by which Notch and PI3K influence signaling events that link T cell stimulation to the regulation of apoptotic responses and long term survival of antigen-stimulated T cells would yield further insights into molecular mechanisms underlying T cell homeostasis.