Intra-mitochondrial degradation of Tim23 curtails the survival of cells rescued from apoptosis by caspase inhibitors.

Caspase inhibition can extend the survival of cells undergoing apoptosis beyond the point of mitochondrial outer membrane permeabilisation (MOMP), but this does not confer long-term protection because caspase-independent death pathways emerge. Here, we describe a novel mechanism of mitochondrial self-destruction in caspase-inhibited cells, whose hallmark is the degradation of Tim23, the essential pore-forming component of the TIM23 inner membrane translocase. We show that Tim23 degradation occurs in cycling and post-mitotic cells, it is caspase-independent but Bax/Bak dependent, and it follows cytochrome c release. The proteolytic degradation of Tim23 is induced by MOMP and is mitochondrion-autonomous, as it also occurs in isolated mitochondria undergoing permeability transition. Degradation of Tim23 is selective, as expression of several other inner membrane proteins that regulate respiratory chain function is unaffected, and is not autophagic, as it occurs similarly in autophagy-proficient and -deficient (Atg-5 knockout) cells. Depleting Tim23 with siRNA is sufficient to inhibit cell proliferation and prevent long-term survival, while expression of degradation-resistant Tim23-GFP in mitochondria delays caspase-independent cell death. Thus, mitochondrial autodigestion of Tim23 joins the array of processes contributing to caspase-independent cell death. Because mitochondrial biogenesis requires a functional protein-import machinery, preventing Tim23 degradation might, therefore, be essential for repairing damaged mitochondria in chronic degenerative diseases.

Characterization of apoptosis in cultured rat sympathetic neurons after nerve growth factor withdrawal.

Sympathetic neurons depend on nerve growth factor (NGF) for their survival both in vivo and in vitro. In culture, the neurons die after NGF withdrawal by an autonomous cell death program but whether these neurons die by apoptosis is under debate. Using vital DNA stains and in situ nick translation, we show here that extensive chromatin condensation and DNA fragmentation occur before plasma membrane breakdown during the death of NGF-deprived rat sympathetic neurons in culture. Furthermore, kinetic analysis of chromatin condensation events within the cell population is consistent with a model which postulates that after NGF deprivation nearly all of the neurons die in this manner. Although the dying neurons display membrane blebbing, cell fragmentation into apoptotic bodies does not occur. Apoptotic events proceed rapidly at around the time neurons become committed to die, regardless of neuronal culture age. However the duration of NGF deprivation required to commit neurons to die, and the rate at which apoptosis occurs, increase with culture age. Thus, within the first week of culture, apoptosis is the predominant form of cell death in sympathetic neurons.

Death commitment point is advanced by axotomy in sympathetic neurons.

Axotomized neurons have several characteristics that are different from intact neurons. Here we show that, unlike established cultures, the axotomized sympathetic neurons deprived of NGF become committed to die before caspase activation, since the same proportion of NGF-deprived neurons are rescued by NGF regardless of whether caspases are inhibited by the pan-caspase inhibitor Boc-Asp(O-methyl)-CH(2)F (BAF). Despite prolonged Akt and ERK signaling induced by NGF after BAF treatment has prevented death, the neurons fail to increase protein synthesis, recover ATP levels, or grow. Within 3 d, all the mitochondria disappear without apparent removal of any other organelles or loss of membrane integrity. Although NGF does rescue intact BAF-treated 6-d cultures after NGF deprivation, rescue by NGF fails when these neurons are axotomized before NGF deprivation and BAF treatment. Moreover, cytosolic cytochrome c rapidly kills axotomized neurons. We propose that axotomy induces signals that make sympathetic neurons competent to die prematurely. NGF cannot repair these NGF-deprived, BAF-treated neurons because receptor signaling (which is normal) is uncoupled from protein renewal, and the mitochondria (which are damaged) go on to be eliminated. Hence, the order of steps underlying neuronal death commitment is mutable and open to regulation.

Ca2+ transients are not required as signals for long-term neurite outgrowth from cultured sympathetic neurons.

A method for clamping cytosolic free Ca2+ ([Ca2+]i) in cultures of rat sympathetic neurons at or below resting levels for several days was devised to determine whether Ca2+ signals are required for neurite outgrowth from neurons that depend on Nerve Growth Factor (NGF) for their growth and survival. To control [Ca2+]i, normal Ca2+ influx was eliminated by titration of extracellular Ca2+ with EGTA and reinstated through voltage-sensitive Ca2+ channels. The rate of neurite outgrowth and the number of neurites thus became dependent on the extent of depolarization by KCl, and withdrawal of KCl caused an immediate cessation of growth. Neurite outgrowth was completely blocked by the L type Ca2+ channel antagonists nifedipine, nitrendipine, D600, or diltiazem at sub- or micromolar concentrations. Measurement of [Ca2+]i in cell bodies using the fluorescent Ca2+ indicator fura-2 established that optimal growth, similar to that seen in normal medium, was obtained when [Ca2+]i was clamped at resting levels. These levels of [Ca2+]i were set by serum, which elevated [Ca2+]i by integral of 30 nM, whereas the addition of NGF had no effect on [Ca2+]i. The reduction of [Ca2+]o prevented neurite fasciculation but this had no effect on the rate of neurite elongation or on the number of extending neurites. These results show that neurite outgrowth from NGF-dependent neurons occurs over long periods in the complete absence of Ca2+ signals, suggesting that Ca2+ signals are not necessary for operating the basic machinery of neurite outgrowth.

Phosphorylation of the pro-apoptotic protein BAD on serine 155, a novel site, contributes to cell survival.

Phosphorylation of BAD, a pro-apoptotic member of the Bcl-2 protein family, on either Ser112 or Ser136 is thought to be necessary and sufficient for growth factors to promote cell survival. Here we report that Ser155, a site phosphorylated by protein kinase A (PKA), also contributes to cell survival. Ser112 is thought to be the critical PKA target, but we found that BAD fusion proteins containing Ala at Ser112 (S112A) or Ser136 (S136A) or at both positions (S112/136A) were still heavily phosphorylated by PKA in an in vitro kinase assay. BAD became insensitive to phosphorylation by PKA only when both Ser112 and Ser136, or all three serines (S112/136/155) were mutated to alanine. In HEK293 cells, BAD fusion proteins mutated at Ser155 were refractory to phosphorylation induced by elevation of cyclic AMP(cAMP) levels. Phosphorylation of the S112/136A mutant was >90% inhibited by H89, a PKA inhibitor. The S155A mutant induced more apoptosis than the wild-type protein in serum-maintained CHO-K1 cells, and apoptosis induced by the S112/136A mutant was potentiated by serum withdrawal. These data suggest that Ser155 is a major site of phosphorylation by PKA and serum-induced kinases. Like Ser112 and Ser136, phosphorylation of Ser155 contributes to the cancellation of the pro-apoptotic function of BAD.

Mitochondria are selectively eliminated from eukaryotic cells after blockade of caspases during apoptosis.

Pan caspase inhibitors are potentially powerful cell-protective agents that block apoptosis in response to a wide variety of insults that cause tissue degeneration. In many conditions, however, the blockade of apoptosis by caspase inhibitors does not permit long-term cell survival, but the reasons are not entirely clear. Here we show that the blockade of apoptosis by Boc.Aspartyl(O-methyl)CH2F can result in the highly selective elimination of the entire cohort of mitochondria, including mitochondrial DNA, from both neurons and HeLa cells, irrespective of the stimulus used to trigger apoptosis. In cells that lose their mitochondria, the nuclear DNA, Golgi apparatus, endoplasmic reticulum, centrioles, and plasma membrane remain undamaged. The capacity to remove mitochondria is both specific and regulated since mitochondrial loss in neurons is completely prevented by the expression of the antiapoptotic protein Bcl-2 and partially suppressed by the autolysosomal inhibitor bafilomycin. Cells without mitochondria are more tolerant to an anaerobic environment but are essentially irreversibly committed to death. Prevention of mitochondrial loss may be crucial for the long-term regeneration of tissues emerging from an apoptotic episode in which death was prevented by caspase blockade.

Tubular epithelial cells in renal clear cell carcinoma express high RIPK1/3 and show increased susceptibility to TNF receptor 1-induced necroptosis.

We previously reported that renal clear cell carcinoma cells (RCC) express both tumor necrosis factor receptor (TNFR)-1 and -2, but that, in organ culture, a TNF mutein that only engages TNFR1, but not TNFR2, causes extensive cell death. Some RCC died by apoptosis based on detection of cleaved caspase 3 in a minority TUNEL-positive cells but the mechanism of death in the remaining cells was unexplained. Here, we underpin the mechanism of TNFR1-induced cell death in the majority of TUNEL-positive RCC cells, and show that they die by necroptosis. Malignant cells in high-grade tumors displayed threefold to four fold higher expression of both receptor-interacting protein kinase (RIPK)1 and RIPK3 compared with non-tumor kidney tubular epithelium and low-grade tumors, but expression of both enzymes was induced in lower grade tumors in organ culture in response to TNFR1 stimulation. Furthermore, TNFR1 activation induced significant MLKL(Ser358) and Drp1(Ser616) phosphorylation, physical interactions in RCC between RIPK1-RIPK3 and RIPK3-phospho-MLKL(Ser358), and coincidence of phospho-MLKL(ser358) and phospho-Drp1(Ser616) at mitochondria in TUNEL-positive RCC. A caspase inhibitor only partially reduced the extent of cell death following TNFR1 engagement in RCC cells, whereas three inhibitors, each targeting a different step in the necroptotic pathway, were much more protective. Combined inhibition of caspases and necroptosis provided additive protection, implying that different subsets of cells respond differently to TNF-α, the majority dying by necroptosis. We conclude that most high-grade RCC cells express increased amounts of RIPK1 and RIPK3 and are poised to undergo necroptosis in response to TNFR1 signaling.

Down-regulation of G alpha s in human myometrium in term and preterm labor: a mechanism for parturition.

We have previously reported that G alpha s is expressed at considerably higher levels in myometrium taken from pregnant than from nonpregnant women. In the present study we have determined adenylyl cyclase activity in myometrial membranes by measuring the conversion of [alpha-32P]ATP to [32P]cAMP and have measured guanosine triphosphate-binding protein expression by immunoblotting with specific antibodies. Here we report that the increase in G alpha s expression in pregnant myometrium is associated with a significant increase in G alpha s-coupled adenylyl cyclase activity, as estimated by incubating myometrial membranes in the presence of 5'-guanylyl-imidodiphosphate with and without prostaglandin E2. Moreover, in myometrium from women in spontaneous labor G alpha s levels and G alpha s-coupled adenylyl cyclase activity are reduced to the levels observed in nonpregnant tissue. There was no apparent change in forskolin-stimulated adenylyl cyclase activity in nonpregnant, pregnant, and laboring tissue. The increase in G alpha s expression in pregnant myometrium may facilitate agonist-induced cAMP formation, resulting in prolonged relaxation of the uterus during gestation. Down-regulation of G alpha s would decrease the relaxing effect exerted by cAMP and may be a triggering mechanism for the initiation of labor.

Deprivation of nerve growth factor rapidly increases purine efflux from cultured sympathetic neurons.

The efflux of [3H]purines from cultured sympathetic neurons prelabelled with [3H]adenine is accelerated 2-3-fold within hours of nerve growth factor (NGF) withdrawal and is reduced by readdition of NGF. Addition of 8-(4-chlorphenyl-thio) cAMP, which delays neurite degeneration, reduced the enhanced efflux of purines, as did the addition of cycloheximide, MgCl2 and the protease inhibitor tosyl-L-lysine chloromethyl ketone. Colchicine accelerated purine efflux and neurite degeneration but 2-deoxyglucose increased purine efflux without inducing degeneration, suggesting that ATP reduction itself is not the cause of neurite degeneration. The increase in purine efflux is thus an early biochemical event that has diagnostic value for the study of NGF action since deprivation is detected well before irreversible changes become established.

Na+/H+ exchange is the major mechanism of pH regulation in cultured sympathetic neurons: measurements in single cell bodies and neurites using a fluorescent pH indicator.

The regulation of intracellular pH in single cell bodies and in neurites of cultured neurons from rat superior cervical ganglion was studied by continuous monitoring of pH transients using the fluorescent indicator bis(carboxyethyl)carboxy-fluorescein. Intracellular pH was 7.03 +/- 0.05 (n = 8) in bicarbonate-free media at pH 7.4 and was not affected by depolarization with high potassium. Brief exposure to NH4Cl caused rapid cytoplasmic acidification followed by an exponential return of intracellular pH to the resting value. The apparent first order rate constant for recovery from an NH4Cl-induced acid load was 0.2 +/- 0.03 min-1 (37 degrees C) and was similar in media at pH 6.5 or 7.8. Recovery from an acid load was blocked by removal of extracellular Na+ or by amiloride but was not dependent on extracellular Cl- or phosphate or blocked by inhibitors of anion transport, in the presence or absence of bicarbonate. Addition of 5-10 mM bicarbonate at pH 7.4 resulted in a slight alkalinization of the cytoplasm and enhanced complete restoration of pHi after an NH4Cl-induced acid load. Nerve growth factor did not affect intracellular pH of either growing cells deprived of nerve growth factor up to 6 days or of newly isolated neurons left at 4 degrees C for a week before exposure to nerve growth factor. Phorbol 12-myristate 13-acetate had no effect on the pH of cell bodies of growing cells and increased pH of cells deprived of nerve growth factor by less than 0.05 pH units. It is concluded that: pH regulation in cultured sympathetic neurons is largely achieved by Na+/H+ exchange; Bicarbonate may also participate in pH regulation, but not by its exchange with Cl-.

Adenosine 5'-triphosphate synthesis and metabolism localized in neurites of cultured sympathetic neurons.

Adenosine triphosphate synthesis and metabolism in cultured sympathetic neurons was studied after the incorporation of [2-3H]adenine into intact or microdissected neurites to determine whether ATP is provided locally during neurite outgrowth, when and where it is synthesized and how its levels are regulated at rest and following depolarization. Neurites maintained an independent capability for synthesis of ATP at any stage of growth: [3H]ATP levels increased in neurites in direct proportion to neurite length and equivalent amounts of [3H]ATP were synthesized by intact neurites, by neurites separated from cell bodies or by neurites further segmented into sections. Thus, metabolic labelling of cultured neurons with [3H]adenine provides a simple method to measure relative neurite outgrowth. Neurite ATP was maintained mainly by respiration but also by glycolysis and [3H]ATP levels were stable for at least 14 h after adenine withdrawal when cells were at rest. Depolarization overcame respiratory control, causing a quantitative conversion of ATP to adenosine monophosphate (AMP) and inosine monophosphate (IMP) and the release of nucleosides (adenosine and inosine) and nucleotides [adenosine diphosphate (ADP) and adenosine monophosphate (AMP)]. Release of nucleosides, but not of nucleotides or [3H]noradrenaline, was enhanced by NaN3 or 2-deoxyglucose under nondepolarizing conditions and was prevented by the adenosine transport inhibitor p-nitrobenzyl-6-thioinosine. It is concluded that neurites can use local mechanisms for ATP synthesis that do not depend on a functional connection to the cell body. Any metabolic stress which causes ATP breakdown causes these cells to express a transient purinergic phenotype involving release of adenosine and inosine by facilitated diffusion. To promote the release of purine nucleotides, however, more specific stimuli are required.

Alternative adrenal chromaffin cell fates induced by basic fibroblast growth factor or cyclic AMP in vitro depend on a collaboration with the growth substrate.

Basic fibroblast growth factor stimulates cultured adrenal chromaffin cells to divide and to transform into sympathetic neurons, but the efficacies reported for these actions of basic fibroblast growth factor have varied widely. We have examined the effect of various growth substrates on basic fibroblast growth factor responses and here we show that the ability of basic fibroblast growth factor to transform neonatal rat chromaffin cells into sympathetic neurons depends on laminin as the culture substrate. On collagen, less than 5% of the cells were transformed into neurons by basic fibroblast growth factor, even when the culture was supplemented with heparin or heparan sulphate, but 65% of cells entered the S phase in the presence of basic fibroblast growth factor compared to 15% in its absence, showing that the basic fibroblast growth factor receptor is still active. On laminin, by contrast, over 60% of the cells transformed into neurons in response to the same concentrations of basic fibroblast growth factor, suggesting that an overlapping pool of cells change their phenotype depending on growth substrate. The cyclic AMP analogue 8-(4-chlorophenylthio)cyclic AMP apparently mimicked the actions of basic fibroblast growth factor, promoting neuronal differentiation on laminin, but mitogenic stimulation on collagen. These data support the notion that collagen and laminin promote different instructions in chromaffin cells that can collaborate with the signals induced by basic fibroblast growth factor and 8-(4-chlorophenylthio)cyclic AMP to determine chromaffin or neuronal cell fates.

Active p21Ras is sufficient for rescue of NGF-dependent rat sympathetic neurons.

We have examined whether p21Ras proteins can rescue nerve growth factor-deprived rat sympathetic neurons from death, to test further our hypothesis that p21Ras is a central mediator in the nerve growth factor-to-survival signalling pathway. After crosslinking [125I]nerve growth factor to live neurons, two forms of Trk (molecular weight approximately 140,000 and 115,000) were immunoprecipitated with anti-Trk antibodies. Nerve growth factor induced tyrosine phosphorylation of both Trk forms and at least two additional proteins. When these phosphorylations were prevented by staurosporine (in a protein kinase C-independent manner) the neurons died. However, neurons were rescued from death due to staurosporine treatment by intracellular loading of oncogenic Ha-Ras(val12) protein. Both Ha-Ras(val12) and cellular Ha-Ras proteins maintained survival for several days in the absence of nerve growth factor and mimicked other actions of nerve growth factor, inducing rapid c-Fos protein expression and robust neurite outgrowth. Conversely, Fab fragments of neutralizing antibodies to p21Ras which blocked the capacity of nerve growth factor to promote neuron survival were also found to inhibit the early expression of c-Fos protein in these neurons. The close correspondence observed between the timing of onset of c-Fos responsiveness and acquisition of nerve growth factor-dependence in embryonic day 17 sympathetic neurons, and the coordinate increase found in both parameters until embryonic day 19 indicates that c-Fos protein expression is a good biochemical indicator of the presence of a functional nerve growth factor-to-survival signal transduction pathway. Nevertheless, expression of c-Fos is not sufficient for survival since phorbol esters induce c-Fos with no effect on survival. These data strengthen our proposal that p21Ras proteins are crucial anti-apoptotic mediators of survival in rat sympathetic neurons by demonstrating that p21Ras is both necessary and sufficient to rescue neurons which are disabled from signalling through Trk receptors.

Herbimycin A induces sympathetic neuron survival and protects against hypoxia.

We have examined how herbimycin affects the survival and neuritogenesis of avian sympathetic neurons. Herbimycin promoted sympathetic neuron survival and neuritogenesis. At higher concentrations (> or = 100 ng/ml), herbimycin still enhanced neuron survival but blocked neuritogenesis. Addition of herbimycin (10-30 ng/ml) to neurons cultured in the presence of NGF or retinal conditioned medium altered neuronal morphology, with an increase in the number of neurites. Addition of NGF during hypoxia rescued 52% of the neurons compared to 14% survival in control conditions. Herbimycin alone rescued about 50% of the neurons. In the presence of NGF and 100 ng/ml herbimycin, 81% of the neurons survived hypoxia. Our results show that herbimycin promotes survival of chick sympathetic neurons and potentiates the effects of NGF.

Nerve growth factor-induced PKB/Akt activity is sustained by phosphoinositide 3-kinase dependent and independent signals in sympathetic neurons.

Phosphoinositide 3-kinase and its downstream effector kinase PKB/Akt have been suggested to have crucial roles in suppressing apoptosis in several classes of neurons. However, few studies have conducted a long-term investigation of either kinase activity, many studies relying instead on use of the phosphoinositide 3-kinase inhibitors wortmannin and LY294002. When we added LY294002 or wortmannin to sympathetic neurons, apoptosis in the presence of nerve growth factor (NGF) was very slow compared to that obtained by NGF deprivation. However, expression of a kinase-inactive mutant of PKB/Akt in the presence of NGF induced apoptosis in a significant proportion of the neurons. To understand this discrepancy, we investigated more closely the regulation of PKB/Akt activity by NGF. NGF stimulation induced a rapid increase in PKB/Akt activity which was sustained at approximately 6-fold up to 24 h. Phosphoinositide 3-kinase was also rapidly activated by NGF. However, concentrations of wortmannin which completely blocked phosphoinositide 3-kinase activity in the neurons inhibited no more than 50-70% of cellular PKB/Akt activity. Similarly, approximately 50% of maximal NGF-stimulated PKB/Akt activity remained elevated at concentrations of LY294002 which completely blocked neurite outgrowth, a process known to be phosphoinositide 3-kinase dependent. We suggest that a proportion of the sustained PKB/Akt activity induced by NGF is mediated by phosphoinositide 3-kinase-independent pathways. These results raise a cautionary note as to the usefulness of LY294002 or wortmannin as tools to dissect the role of PKB/Akt in neuronal survival.

Phosphorylation of Puma modulates its apoptotic function by regulating protein stability.

Puma is a potent BH3-only protein that antagonises anti-apoptotic Bcl-2 proteins, promotes Bax/Bak activation and has an essential role in multiple apoptotic models. Puma expression is normally kept very low, but can be induced by several transcription factors including p53, p73, E2F1 and FOXO3a, whereby it can induce an apoptotic response. As Puma can to bind and inactivate all anti-apoptotic members of the Bcl-2 family, its activity must be tightly controlled. We report here, for the first time, evidence that Puma is subject to post-translational control through phosphorylation. We show that Puma is phosphorylated at multiple sites, with the major site of phosphorylation being serine 10. Replacing serine 10 with alanine causes reduced Puma turnover and enhanced cell death. Interestingly, Puma turnover occurs through the proteasome, and substitution of serine 10 causes elevated Puma levels independently of macroautophagy, Bcl-2 family member binding, caspase activity and apoptotic death. We conclude, therefore, that phosphorylation of Puma at serine 10 promotes Puma turnover, represses Puma's cell death potential and promotes cell survival. Owing to the highly pro-apoptotic nature of Puma, these studies highlight an important additional regulatory step in the determination of cellular life or death.