Regulation of both apoptosis and cell survival by the v-Src oncoprotein

A number of oncogenes alter the regulation of the cell cycle and cell death, contributing to the altered growth of tumours. Expression of the v-Src oncoprotein in Rat-1 fibroblasts prevented cell cycle exit in response to growth factor withdrawal. Here we investigated whether survival of v-Src transformed cells in low serum is dependent on v-Src activity. We used a temperature sensitive v-Src to study the effect inactivating v-Src on transformed cells growing under low serum conditions. We found when we switched off v-Src the cells died by apoptosis characterised by activation of caspases and the stress-activated kinases, JNK (Jun N-terminal kinase) and p38 MAP (mitogen activated protein) kinase. We were able to prevent cell death by addition of serum or overexpression of Bcl-2. Thus v-Src transformed Rat-1 cells can be protected from apoptosis by serum, v-Src, or Bcl-2. We investigated how v-Src protects from apoptosis under these conditions. Amongst other effects, v-Src activates two kinases which have been shown to protect cells from apoptosis, phosphatidylinositol 3-kinase (PI3-K) and extracellular signal-regulated kinase (ERK1/2). We found that switching off v-Src led to a decrease in the activity of both PI3-K and ERK1/2, however, we found that adding a specific inhibitor of PI3-K (LY294002) to v-Src transformed Rat-1 cells grown in low serum induced apoptosis while a specific ERK kinase (MEK1) inhibitor (PD98059) had no effect. This suggests that v-Src protects from apoptosis under low serum conditions by activating PI3-K.     

Kinetics of the apoptotic response induced by anti-IgM engagement of the B cell receptor is dependent on the density of cell surface immunoglobulin M expression

The effect of B cell receptor (BCR) density on anti-BCR-induced apoptosis was assessed in Ramos cell lines, expressing low, medium, or high levels of surface IgM (sIgM(LO), sIgM(MED), sIgM(HI)). All cells required a 6-mug/ml threshold of anti-IgM to elicit apoptosis. Anti-IgM treatment of sIgM(LO) cells induced growth inhibition and limited dose-independent apoptosis. Anti-IgM treatment of sIgM(MED) cells induced dose-independent death with a 32-h lag. Ligation of the BCR in the sIgM(HI) cells induced rapid apoptosis beginning by 6 h, which was dose-dependent. Secondary crosslinking reagents did not affect apoptosis, and this effect was independent of anti-IgM concentration, time, or sIgM density. These results suggest that the response to BCR engagement strongly depends on the cell surface receptor density.     

High cell density provides potent survival signals for resting T-cells.

Activated T-cells are susceptible to apoptosis through two particularly important pathways: ligation of CD95 (Fas/Apo-1) or cytokine deprivation. Resting T-cells have until recently been considered to be relatively resistant to apoptosis. In this report we show that resting T-cells die rapidly by apoptosis when deprived of serum or cell contact. Primed CD45RO+ cells were more susceptible than naive CD45RA+ cells, consistent with their relative expression of Bcl-2. CD4+, CD8+ and gammadelta T-cells were equally prone to apoptosis under all studied conditions. A linear relationship between cell survival and serum concentration was observed for cells cultured between 0.5-2x10(6)/ml. T-cells cultured at low density died even in high concentrations of serum. However, resting T-cells cultured at high cell density (4x10(6)/ml) survived for extended periods in the absence of serum or other survival factors. This effect was mediated by the production of soluble factors and independent of integrin mediated signals. These results suggest that T-cells at sites of high density such as the lymph node paracortex are independent of external survival factors, while those trafficking through the peripheral circulation are highly dependent on serum derived factors for survival.     

Caspases Mediate 6-Hydroxydopamine-Induced Apoptosis but Not Necrosis in PC12 Cells

Abstract: The neurotoxin 6-hydroxydopamine (6-OHDA) induces apoptosis in the rat phaeochromocytoma cell line PC12. 6-OHDA-induced apoptosis is morphologically indistinguishable from serum deprivation-induced apoptosis. Exposure of PC12 cells to a low concentration of 6-OHDA (25 µ M ) results in apoptosis, whereas an increased concentration (50 µ M ) results in a mixture of apoptosis and necrosis. We investigated the involvement of caspases in the apoptotic death of PC12 cells induced by 6-OHDA, using a general caspase inhibitor, benzyloxycarbonyl-Val-Ala-Asp-fluoromethyl ketone (zVAD-fmk), and compared this with serum deprivation-induced apoptosis, which is known to involve caspases. We show that zVAD-fmk (100 µ M ) completely prevented the apoptotic morphology of chromatin condensation induced by exposure to either 6-OHDA (25 and 50 µ M ) or serum deprivation. Furthermore, cell lysates from 6-OHDA-treated cultures showed cleavage of a fluorogenic substrate for caspase-3-like proteases (caspase-2, 3, and 7), acetyl-Asp-Glu-Val-Asp-aminomethylcoumarin, and this was inhibited by zVAD-fmk. However, although zVAD-fmk restored total cell viability to serum-deprived cells or cells exposed to 25 µ M 6-OHDA, the inhibitor did not restore viability to cells exposed to 50 µ M 6-OHDA. These data show the involvement of a caspase-3-like protease in 6-OHDA-induced apoptosis and that caspase inhibition is sufficient to rescue PC12 cells from the apoptotic but not the necrotic component of 6-OHDA neurotoxicity.     

Expression of gamma-glutamyltransferase 1 in glioblastoma cells confers resistance to cystine deprivation–induced ferroptosis

Ferroptosis is an iron-dependent mode of cell death caused by excessive oxidative damage to lipids. Lipid peroxidation is normally suppressed by glutathione peroxidase 4, which requires reduced glutathione. Cystine is a major resource for glutathione synthesis, especially in cancer cells. Therefore, cystine deprivation or inhibition of cystine uptake promotes ferroptosis in cancer cells. However, the roles of other molecules involved in cysteine deprivation–induced ferroptosis are unexplored. We report here that the expression of gamma-glutamyltransferase 1 (GGT1), an enzyme that cleaves extracellular glutathione, determines the sensitivity of glioblastoma cells to cystine deprivation–induced ferroptosis at high cell density (HD). In glioblastoma cells expressing GGT1, pharmacological inhibition or deletion of GGT1 suppressed the cell density–induced increase in intracellular glutathione levels and cell viability under cystine deprivation, which were restored by the addition of cysteinylglycine, the GGT product of glutathione cleavage. On the other hand, cystine deprivation induced glutathione depletion and ferroptosis in GGT1-deficient glioblastoma cells even at an HD. Exogenous expression of GGT1 in GGT1-deficient glioblastoma cells inhibited cystine deprivation–induced glutathione depletion and ferroptosis at an HD. This suggests that GGT1 plays an important role in glioblastoma cell survival under cystine-limited and HD conditions. We conclude that combining GGT inhibitors with ferroptosis inducers may provide an effective therapeutic approach for treating glioblastoma.     

Different subsets of T cells in the adult mouse bone marrow and spleen induce or suppress acute graft-versus-host disease.

Fractionation of normal adult mouse spleen and bone marrow cells (C57BL/Ka) was performed by discontinuous Percoll density gradients. The fractionated low density (1.050-1.060 g/ml) C57BL/Ka spleen cells completely suppressed acute lethal graft vs host disease (GVHD) when coinjected with unfractionated C57BL/Ka spleen cells into sublethally irradiated (400 rad) BALB/c mice. In dose response experiments, as few as 0.5 x 10(6) low density cells from the spleen fractions suppressed acute GVHD induced by 2.5 x 10(6) unfractionated allogeneic spleen cells. Although the low density spleen fractions inhibited acute GVHD, the high density (1.075-1.090 g/ml) spleen fractions induced acute GVHD in sublethally irradiated BALB/c recipients. Fractionation of C57BL/Ka bone marrow cells showed that none of the high or low density fractions or unfractionated cells induced lethal GVHD. When these fractions were tested for their capacity to suppress GVHD by coinjection with C57BL/Ka unfractionated spleen cells, all fractions protected the BALB/c recipients. Unfractionated bone marrow cells showed modest protection. Evaluation of the dose response characteristics of the suppressive activity of the low and middle density (1.060-1.068 g/ml) bone marrow cell fraction showed that reproducible protection could be achieved at a 5:1 ratio of inducing to suppressing cells. The low density fractions of both bone marrow and spleen cells had a marked depletion of typical TCR(+)-alpha beta CD4+ or CD8+ T cells, and a predominant population of TCR(+)-alpha beta CD4- CD8- T cells. Purified populations of the latter cells suppressed GVHD. Recipients given unfractionated C57BL/Ka spleen cells and protected with low-density bone marrow or spleen cells were chimeras.     

The cellular apoptosis of murine BW5147 thymoma during cold shock]

The mode of T-lymphoma cell death induced by cold shock was studied. The rewarming of cells at 37 degrees C following a brief period of cold (0 degrees C) resulted in internucleosomal DNA fragmentation. The cells underwent cold shock-mediated apoptosis only at a reduced (2%) serum concentration. The apoptosis was not blocked by macromolecular synthesis inhibitors such as cycloheximide and antinomycin D, or by Quin-2. EGTA per se was responsible for the initiation of cell death. Colchicine also induced internucleosomal fragmentation of DNA. Our findings suggest that cold shock induced apoptosis is associated with low temperature mediated disruption of microtubules. The role of Ca2+ and growth factors in cold shock induced cell death is discussed.     

Potassium and sodium balance in U937 cells during apoptosis with and without cell shrinkage.

Staurosporine (STS) and etoposide (Eto) induced apoptosis of the human histiocytic lymphoma cells U937 were studied to determine the role of monovalent ions in apoptotic cell shrinkage. Cell shrinkage, defined as cell dehydration, was assayed by measurement of buoyant density of cells in continuous Percoll gradient. The K+ and Na+ content in cells of different density fractions was estimated by flame emission analysis. Apoptosis was evaluated by confocal microscopy and flow cytometry of acridine orange stained cells, by flow DNA cytometry and by effector caspase activity. Apoptosis of U937 cells induced by 1 muM STS for 4 h was found to be paralleled by an increase in buoyant density indicating cell shrinkage. An increase in density was accompanied by a decrease in K+ content (from 1.1 to 0.78 mmol/g protein), which exceeded the increase in Na+ content (from 0.30 to 0.34 mmol/g) and resulted in a significant decrease of the total K+ and Na+ content (from 1.4 to 1.1 mmol/g). In contrast to STS, 50 microM Eto for 4 h or 0.8-8 microM Eto for 18-24 h induced apoptosis without triggering cell shrinkage. During apoptosis of U937 cells induced by Eto the intracellular K(+)/Na+ ratio decreased like in the cells treated with STS, but the total K+ and Na+ content remained virtually the same due to a decrease in K+ content being nearly the same as an increase in Na+ content. Apoptotic cell dehydration correlated with the shift of the total cellular K+ and Na+ content. There was no statistically significant decrease in K+ concentration per cell water during apoptosis induced by either Eto (by 13.5%) or STS (by 8%), whereas increase in Na+ concentration per cell water was statistically significant (by 27% and 47%, respectively). The data show that apoptosis can occur without cell shrinkage-dehydration, that apoptosis with shrinkage is mostly due to a decrease in cellular K+ content, and that this decrease is not accompanied by a significant decrease of K+ concentration in cell water.     

A Novel Protein Expressed in Mammalian Cells Undergoing Apoptosis

Human and rodent cells undergoing apoptosis were observed to express high levels of a novel 45,000 M_r protein. The protein, which we have termed apoptosis specific protein (ASP), was found in Burkitt lymphoma (BL) cells and in adenovirus-transformed human and rat embryo cells induced into apoptosis by a variety of stimuli, including serum deprivation, exposure to the Ca²⁺ ionophore, ionomycin, treatment with inhibitors of macromolecular synthesis (cycloheximide and actinomycin D), and cold shock. In BL cells treated with apoptotic stimuli, expression of the oncoprotein Bcl-2 was found to both protect from apoptosis and prevent expression of ASP. ASP was not detected either in viable cells or in cells dying passively by necrosis. Laser scanning confocal microscopy showed high levels of ASP in the cytoplasm of cells displaying the chromatin condensation and fragmentation patterns typical of apoptosis. Retention of ASP was observed even when DNA was no longer detectable, and two-color immunofluorescence staining indicated that the protein primarily colocalized with, but was clearly distinct from, nonmuscle actin. These findings, together with the observation that biochemical extraction of ASP was only possible under conditions which caused solubilization of the cytoskeleton, lead us to conclude that ASP forms part of, or at least strongly associates with, a modified cytoskeleton unique to cells undergoing apoptosis. While elucidation of its function will require further work, ASP constitutes a powerful marker for the diagnosis and quantitation of apoptosis in vivo and in vitro.     

Inhibition of Drug-Induced Apoptosis by Survival Factors in PC12 Cells

Abstract: Pheochromocytoma (PC12) cells have been shown to undergo apoptosis (programmed cell death) when deprived of serum and to be rescued by nerve growth factor, fibroblast growth factor, dibutyryl cyclic AMP, aurintricarboxylic acid, or exogenous expression of bcl-2 . We show here that the cytotoxic drugs cycloheximide, actinomycin D, colchicine, and EGTA also induce apoptosis in PC12 cells. These findings prompted us to investigate whether apoptosis induced by these drugs involves similar pathways in each case, and whether the factors preventing the apoptotic death of serum-deprived PC12 cells can also protect the cells from apoptosis induced by the cytotoxic drugs. Nerve growth factor, dibutyryl cyclic AMP, and expression of bcl-2 inhibited apoptosis induced by all four cytotoxic drugs. Fibroblast growth factor inhibited apoptosis induced by EGTA or colchicine. Aurintricarboxylic acid inhibited apoptosis induced by EGTA. These results suggest that apoptosis induced by treatments with the various drugs is mediated by different initiating pathways, all of which converge into a final, common pathway. Nerve growth factor, dibutyryl cyclic AMP, and bcl-2 appear to affect the final common pathway, whereas fibroblast growth factor and aurincarboxylic acid appear to be more specific and affect only some of the pathways.     

HDL3 reduces the association and modulates the metabolism of oxidized LDL by osteoblastic cells: a protection against cell death

Oxidized low density lipoproteins (OxLDL) are known to promote atherosclerosis, but it is only recently that OxLDL have been associated with alterations of the functions of bone-forming osteoblasts and osteoporosis. Although high density lipoproteins (HDL) are recognized for their anti-atherogenic action, there is less information about their ability to protect against osteoporosis. Therefore, we investigated the capacity of HDL3 to prevent the cell death induced by OxLDL in human osteoblastic cells. Simultaneous exposure of the cells to HDL3 and OxLDL abolished the reduction of cell viability monitored by MTT activity measurement and the induction of apoptosis determined by annexin V staining indicating that HDL3 prevent the apoptosis of osteoblasts induced by OxLDL. This protection correlated with the displacement by HDL3 of OxLDL association to osteoblasts, signifying that OxLDL binding and/or internalization are/is necessary for their cytotoxic effects. We also found that exposition of osteoblastic cells to HDL3 prior to incubation with OxLDL reduced cell death and preserved the lysosomal integrity. This protection was correlated with an increase of SR-BI expression, a modification of OxLDL metabolism with less global uptake of OxLDL and greater selective uptake of cholesterol from OxLDL. These results strongly suggest that, as for atherosclerosis, HDL may exert beneficial actions on bone metabolism.     

Anti-apoptotic activity of low levels of wild-type p53.

Induction of apoptosis is a function of both an external stimulus and the physiology of the cell, which includes the expression of multiple oncogenes and tumor suppressors. Here we have studied the apoptotic response of immortalized mouse fibroblasts to serum withdrawal. We show that, in addition to the p53-independent apoptosis observed in p53- cells, overexpression of wild-type p53 tumor suppressor results in a high rate of programmed cell death. However, physiological range, low levels of the p53 protein protect fibroblasts from induction of apoptosis. Our results indicate that, as a function of its dose, the wild-type p53 can either protect from death or promote apoptosis. This new, anti-apoptotic, activity of p53 may have implications for the understanding of the role played by p53 in embryonic development as well as in initial stages of oncogenesis.     

Expression of gamma-glutamyl transpeptidase protects ramos B cells from oxidation-induced cell death

The ectoenzyme, gamma-glutamyl transpeptidase (GGT, EC ) cleaves glutathione (GSH) to facilitate the recapture of cysteine for synthesis of intracellular GSH. The impact of GGT expression on cell survival during oxidative stress was investigated using the human B cell lymphoblastoid cell line, Ramos. Ramos cells did not express surface GGT and exhibited no GGT enzyme activity. In contrast, Ramos cells stably transfected with the human GGT cDNA expressed high levels of surface GGT and enzymatic activity. GGT-transfected Ramos cells were protected from apoptosis when cultured in cyst(e)ine-deficient medium. The GGT-expressing cells also had lower levels of intracellular reactive oxygen species (ROS). Homocysteic acid and alanine, inhibitors of cystine and cysteine uptake, respectively, caused increased ROS content and diminished viability of GGT expressing cells. Exogenous GSH increased the viability of the GGT-transfected cells more effectively than that of control cells, whereas the products of GSH metabolism prevented death of both the control and GGT-transfected cells comparably. These data indicate that GGT cleavage of GSH and the subsequent recapture of cysteine and cystine allow cells to maintain low levels of cellular ROS and thereby avoid apoptosis induced by oxidative stress.     

Ferroptosis: an iron-dependent form of nonapoptotic cell death

Nonapoptotic forms of cell death may facilitate the selective elimination of some tumor cells or be activated in specific pathological states. The oncogenic RAS-selective lethal small molecule erastin triggers a unique iron-dependent form of nonapoptotic cell death that we term ferroptosis. Ferroptosis is dependent upon intracellular iron, but not other metals, and is morphologically, biochemically, and genetically distinct from apoptosis, necrosis, and autophagy. We identify the small molecule ferrostatin-1 as a potent inhibitor of ferroptosis in cancer cells and glutamate-induced cell death in organotypic rat brain slices, suggesting similarities between these two processes. Indeed, erastin, like glutamate, inhibits cystine uptake by the cystine/glutamate antiporter (system x(c)(-)), creating a void in the antioxidant defenses of the cell and ultimately leading to iron-dependent, oxidative death. Thus, activation of ferroptosis results in the nonapoptotic destruction of certain cancer cells, whereas inhibition of this process may protect organisms from neurodegeneration.     

Expression of insulin-like growth factor binding protein-5 by ovine granulosa cells is regulated by cell density and programmed cell death in vitro

In vivo, in the sheep ovary, the expression of insulin-like growth factor binding protein (IGFBP)-2 and particularly IGFBP-5 has been shown to increase dramatically in apoptotic granulosa cells from atretic follicles. The aim of this work was to study the relationship between apoptosis induced by serum starvation in vitro and expression of IGFBP-2 and -5 by ovine granulosa cells. For this purpose, granulosa cells from follicles 1–3 mm in diameter were cultured in the presence of serum for 2 days, then cultured in the presence or absence of serum for 24, 48, or 72 hr. At the end of the culture, cells were counted, cell viability was assessed by studying DNA fragmentation, and IGFBPs expression was studied by quantitative autoradiography, Western-ligand blotting, immunoblotting, and quantitative in situ hybridization. In vitro, IGFBP-2 and particularly IGFBP-5 were the main IGFBPs secreted by ovine granulosa cells. Serum starvation provoked (i) apoptosis of granulosa cells within 48 hr, (ii) a marked decrease in cell density, and (iii) a marked increase in the amount of IGFBP-5 associated with cell membranes and with the walls of culture wells, but no change in culture medium. The increase in the amount of cell- and wall-associated IGFBP-5 after serum starvation was essentially due to the consecutive decrease in cell density rather than to an increase in cell apoptosis. Indeed, irrespective of the presence or absence of serum, the amount of IGFBP-5 associated to cell membranes was inversely correlated to cell density. In contrast, the amount of IGFBP-5 present in culture medium was positively correlated to cell density. Furthermore, expression of IGFBP-5 mRNA was shown to increase with both cell density and cell death. Indeed, the expression of IGFBP-5 mRNA dramatically increased with cell density, irrespective of the presence or absence of serum, but at a similar cell density, expression was higher in serum-free than in serum conditions. Overall, these results indicate that, in vitro, the localization of IGFBP-5 on ovine granulosa cell membranes and in culture medium, respectively, was mainly dependent on cell density, whereas expression of IGFBP-5 mRNA was related to both cell density and cell death. These data suggest that IGFBP-5 is involved in both growth arrest and apoptosis of granulosa cells in the sheep. J. Cell. Physiol. 177:13–25, 1998. © 1998 Wiley-Liss, Inc.     

TNF inhibited the apoptosis by activation of Akt serine/threonine kinase in the human head and neck squamous cell carcinoma

Tumour necrosis factor (TNF) is known to induce apoptosis, but recently, TNF was shown to promote cell survival, a process regulated by phosphatidylinositol-3-OH kinase (PI3K) and the NFkappaB pathway. In this study, we investigated the relationship between the molecules implicated in regulating TNF-induced cell survival and apoptosis induced by TNF in a human head and neck squamous cell carcinoma cell line (SAS), with special reference to the Akt pathway, one of the pathways related to cell survival. In SAS cells, TNF induced the phosphorylation of Akt at both Ser473 and Thr308, causing the activation of Akt, and also induced the phosphorylation and degradation of IkappaB (inhibitor of NFkappaB). This phosphorylation and degradation was inhibited by pretreating the cells with the PI3K inhibitors, wortmannin or LY294002. The apoptosis of SAS cells induced by TNF was dependent on the concentration: a high concentration of TNF, but not a low concentration, induced apoptosis within 30 h. However, a low concentration of TNF in the presence of wortmannin or LY294002 induced apoptosis. Furthermore, expression of the kinase-negative form of Akt, IKKalpha or IKKbeta, and the undegradable mutant of IkappaB, also induced apoptosis at low concentrations of TNF. When the SAS cells expressed constitutively activated Akt, apoptosis was not induced, even by high concentrations of TNF. These observations suggest that, in the SAS cell line, the PI3K-NFkappaB pathway contributes to TNF-induced cell survival and that inhibition of this pathway accelerates apoptosis.     

Effect of neural stem cells on apoptosis of PC12 cells induced by serum deprivation

Neural stem cells (NSCs) have a bright application prospect to be used to treat neurodegenerative diseases due to their capacity to give rise to the appropriate cell types when they are grafted. At present, however, the function of NSCs after transplantation is not quite ensured, whether to replace the degenerative cells or to secrete nutrient factors. On the other hand, pheochromocytoma cell line 12 (PC12) cells have been widely used for investigating Parkinson's disease (PD) since their apoptosis is similar to that of dopaminergic neuron cells. Therefore, the possible cytoprotective effects of NSCs on the apoptosis of PC12 cells induced by serum deprivation were investigated in this paper. PC12 cells were cocultured with NSCs in DMEM/F12 medium free of serum, and their morphologies, viabilities, and survival were observed with an inverted microscope and assessed with a CCK-8 assay. In addition, the concentrations of glial derived neurotrophic factor (GDNF) in different medium were detected with a GDNF Elisa kit, and the mechanism of NSC's protective effect on PC12 cell apoptosis induced by serum deprivation was analyzed. The results showed that (1) PC12 cell apoptosis induced by serum deprivation increased with time, and only about 44.25% PC12 cells survived after 72 h; (2) NSCs culture medium protected against PC12 cell apoptosis insignificantly; (3) NSCs' supernatant and NSCs mildly prevented PC12 cells from apoptosis; (4) the amount of GDNF secreted by NSCs increased after the coculture with the apoptotic PC12 cells induced by serum deprivation. It can be concluded that there exists clear interaction between NSCs and apoptotic PC12 cells, and that GDNF secretion from NSCs is one of the important mechanisms to prevent the apoptosis of PC12 cells.     

Insulin rescues retinal neurons from apoptosis by a phosphatidylinositol 3-kinase/Akt-mediated mechanism that reduces the activation of caspase-3.

The ability of insulin to protect neurons from apoptosis was examined in differentiated R28 cells, a neural cell line derived from the neonatal rat retina. Apoptosis was induced by serum deprivation, and the number of pyknotic cells was counted. p53 and Akt were examined by immunoblotting after serum deprivation and insulin treatment, and caspase-3 activation was examined by immunocytochemistry. Serum deprivation for 24 h caused approximately 20% of R28 cells to undergo apoptosis, detected by both pyknosis and activation of caspase-3. 10 nm insulin maximally reduced the amount of apoptosis with a similar potency as 1.3 nm (10 ng/ml) insulin-like growth factor 1, which acted as a positive control. Insulin induced serine phosphorylation of Akt, through the phosphatidylinositol (PI) 3-kinase pathway. Inhibition of PI 3-kinase with wortmannin or LY294002 blocked the ability of insulin to rescue the cells from apoptosis. SN50, a peptide inhibitor of NF-kappaB nuclear translocation, blocked the rescue effect of insulin, but neither insulin or serum deprivation induced phosphorylation of IkappaB. These results suggest that insulin is a survival factor for retinal neurons by activating the PI 3-kinase/Akt pathway and by reducing caspase-3 activation. The rescue effect of insulin does not appear to be mediated by NF-kappaB or p53. These data suggest that insulin provides trophic support for retinal neurons through a PI 3-kinase/Akt-dependent pathway.