Cultured neurons expressing phosphorylated tau are more resistant to apoptosis induced by NMDA or serum deprivation

Abstract Apoptosis is a programmed cell death that occurs during the development of the nervous system and in neurodegenerative disorders. Tau protein is a cytoskeletal component that promotes microtubule polymerization and stabilization. Apoptosis was induced in primary neuronal cultures by a prolonged exposure f 16 h) to the NMDA ( N-methyl-D-aspartate 20 PM) or by serum deprivation. The percentages of apoptotic neurons expressing phosphorylated tau (AT81 immunoreactivity are comparable in control and NMDA-exposed cultures (7.5 + 1.9 and 6.9 t_ 1.9%, respectivelly). At the opposite, the percentage of apoptotic neurons expressing de-phosphorylated tatt (taul ) immunolabellings is dramatically increased in NIMDA-treated cultures ( X 1.3 of controls). Similar results were also observed 48 h after serum deprivation. These results demonstrate in vitro that under these conditions. resistant and sensitive cortical neurons to apoptosia can be partly differentiated according to their phosphorylated tau immunoreuctivitics. K+ww&

Apoptosia; Excitotoxicity: Serum deprivation: Phosphorylated tau: Primary neuronal culture

Cellular death can occur during physiological conditions, such as development and aging, or at the occasion of disorders induced by hypoxia, toxins, infections or L..ring degenerative diseases [29]. For many cells, this process can be divided into two phenomenons: apoptosis which is a programmed cell death and necrosis which is often observed after a pathological stress [ 171. In the nervous system, both apoptosis and necrosis are described and it seems that sometimes one single mechanism can trigger this two kinds of neuronal degeneration. For example, excitotoxicity induced in vitro by NMDA can produce apoptosis or necrosis, depending upon the dose and the duration of intoxication [2]. Several signal transduction pathways were reported to produce apoptosis in various cell types [8,21]. These pathways include intracytosolic calcium signalling, regulation by proteine kinase C, CAMP, ceramide and oxygen radicals. Apoptosis is also controlled by genes, such as bcl-2, and by proteases like the cysteine protease interleukine- 1@converting enzyme [9,18,24]. Tau protein is a microtubule-associated protein which promotes microtubule stabilization and polymerization [ 161. Neu-

??

Corresponding author. Fax: + 33 (5) 5543-5893.

ronal tau protein is mainly localized into axons. In Alzheimer’s accumulates

disease. highly phosphorylated tau protein into paired helical filaments which arc the

main constituent of neurofibrillary tangles [lo]. Previous reports revealed that glutamate or NMDA toxicities can enhance in vitro tau immunoreactivity in a certain number of neurons [20,26,4]. The goal of the present study is to analyse tau immunoreactivity in apoptotic neurons in culture. Apoptosis was induced by a prolonged exposure to a mild concentration of the excirotoxin NMDA or by serum deprivation and an immunocytochemical study was carried out using AT8 (phosphorylated (au) or taul (de-phosphorylated tau) antibodies. Neuronal cultures. The method used for rat cortical cell cultures was previously described [ 141. Cortical cells were removed from l&day-old embryonic rats. Cortices were minced in PBS glucose free of Ca’+ and Mg’ ’ and mechanically dissociated. The cell suspension was centrifuged (300 X g for 10 min). The resulting pellets were re-suspended in MEM Earle’s salts. Culture dishes were coated with poly-t_-lysine (M, = 30000-70000) for 1 h at 37°C under humidified 95% air-S% CO?. Neural cells were added to MEM Earle’s salts containing 5% fetal calf serum, insulin (5 pg/ml), progesterone (2 x IO-’ MI,

0169-328X/97/$17.00 Copyright 0 1997 Elsevier Science B.V. All rights reserved. PII SOl69-328X(96)00284-7

transferrin (100 pg/ml), selenium (3 X lo-” mM) and putrescin (100 x 10e6 MI. Cortical neural cells were plated into dishes at a concentration of 1.25 X 10’ cells/ml and allowed to grow from 7-10 days before treatment. After this period, the percentage of astroglial cells (immunostained with a GFAP antibody) was = 5% of total cells. NMDA exposure. After 10 days of culture, cells were exposed to NMDA (20 PM) for 16 h. NMDA was dissolved in Hank’s balanced salts solution, free of Mg’+ and containing Ca?+ (1.8 mM), glycine (5 PM) and glucose (18 mM). Control cultures were only exposed to the vehicule alone. Serum deprivation. Serum deprivation in cortical cell cultures was carried out in 7-day-old cultures according to a previously determined procedure [151. Serum-deprived medium to which cultures were exposed for 48 h consisted in MEM Earle’s salts without fetal calf serum, hormones and trophic chemicals. Propidium iodide DNA stuining and Tunei method. For propidium iodide stainings, neuronal cultures were washed 3 x 5 min in a tyrode buffer (TB) (pH 7.4) containing 136.9 mM NaCl, 2.7 mM KCl, 1.8 mM CaC12, 0.5 mM MgCl,, 0.36 mM NaH?PO,, then fixed in methanol : water (80: 20) at 4°C for 40 min. Dishes were washed twice with TB and incubated at room temperature with propodium iodide 5 pg/ml for 5 min (Sigma). Then. dishes were coversliped using Vectashield medium (Vector) and cells were examined using fluorescence microscopy (Microphot FX, Nikon). Tune1 method (TdT-mediated dUTP nick-end-labelling) is performed according to the working procedures described in the in situ detection kit for immunohistochemical detection and quantification of apoptosis (Boehringer Mannheim). Neuronal cultures were fixed with a freshly prepared paraformaldehyde solution (4% in PBS, pH 7.4) for 30 min at room temperature; dishes were washed with PBS and cells were incubated in permeabilization solution (0.1% sodium citrate) for 2 min on ice. Endogenous peroxidases were inactivated by covering cells with 3% H,O, for 10 min at room temperature. Then, cells were incubated with the terminal deoxynucleotidyl transferase enzyme (TdT) and biotinylated dUTP in TdT buffer at 37°C for 60 min. Dishes were rinced, incubated with antifluorescein antibody conjugated with peroxidase for 30 min at 37°C. Labelling was revealed with the diaminobenzidine as a chromogene and dishes were coversliped and analysed under light microscope. Labelled DNA was not observed in dishes when TdT was omitted. The endonuclease inhibitor aurintricarboxylic acid (ATA) (50 PM) was added in some cultures 1 h before and during the time of NMDA exposure (16 h) or serum deprivation (24 h). Apoptosis was assessed in at least 3 culture dishes/condition (600 neurons) and experiments were done in duplicate. I~nmunojluorescence study. Indirect immunofluorescence was carried out with 2 different antibodies in

control and treated cultures: AT8 (Innogenetics) which is raised against phosphorylated tau and taul (Boehringer Mannheim) raised against de-phosphorylated tau. AT8 and taul cannot be taken as complementary antibodies but both of them recognized epitopes in the vicinity of serine 202. The second layer antibody was coupled to FlTC. Additionally, cell cultures were counterstained with propidium iodide for the evaluation of apoptotic nuclei with condensed or fragmented chromatin. The percentages of AT8 (concentration 1/ 100) and tau 1 (concentration 1/ 100) immunopositive and immunonegative apoptotic cells were evaluated in control, NMDA-exposed cultures and cultures deprived of serum. Neurons were considered to be immunopositive when the cell body around the apoptotic nucleus was immunolabelled. At least 3 culture dishes were assessed per condition (600 neurons). Experiments were done in duplicate under a fluorescence microscope. Statistical analyis. Statistical analysis was carried using comparison of means by analysis of variance (ANOVA). Results are expressed as mean f S.D. Fig. 1 shows the results of apoptosis evaluation in control cultures, in cultures exposed to NMDA or cultured in serum-deficient medium. The percentages of apoptotic neurons determined by propidium iodide stainings and the Tune1 method were 37.8 &-4.2 and 29.9 of:4.5%, respectively, in cultures treated with NMDA 20 PM for 16 h. Both findings are rather comparable. In serum-deprived cultures, the rates of apoptotic neurons were 30.8 + 5.0 and 41.5 + 6.2%, respectively. Neuronal apoptosis induced by NMDA or serum deprivation was markedly reduced by the non-specific endonuclease inhibitor aurintricarboxylic acid as shown in Table I. The immunocytochemical analysis reveals that 41.1 + 5.7 and 40.7 _+5.1%, respectively, of the total number of neurons are AT8- or tau l-positive in control cultures. Fig. 2 displays the results of the evaluation of apoptotic neurons immunostained with the two differents antibodies: AT8 and taul. No statistical differences were noted between the percentages of apoptotic neurons immunostained with AT8 antibody in control cultures and in cultures exposed to NMDA (7.5 + 1.9 and 6.9 3- 1.9%, respectively). At the opposite, an increased percentage of apoptotic neurons immunolabelled by taul was noted in cultures exposed to NMDA compared to control cultures (X 2.3 of controls) (Fig. 2A). These findings suggest that apoptosis induced by NMDA toxicity can occur more readily in neurons concurrently strongly expressing de-phosphorylated tau (taul! as compared to neurons clearly expressing phosphorylated tau. Fig. 2B shows the results of AT8 and taul immunostainings in control cultures and in serum-deprived cultures for 48 h. The percentage of apoptotic neurons immunolabelled by AT8 was comparable to the one determined in controls (13.5 5 1.7 and 13.4 f 3.1%, respectively) whereas the percentage immunostained by tau 1 was 2.3 X as big compared to controls. These findings are similar to those observed in cortical cell cultures exposed to 20 PM NMDA. Fig. 3

0 CONTROLS Fig.

I.

SERUM DEPRIVATION

This (ipure shows the percrntu~e~ ot’apoptotic neurons tn c~rn~n~l

cultures. in cultures treated for I6 h with JO ~Lhl NMDA

(A)

and tn

Erg

2. Ftg. 2 di\pldys

the re~h\

serum-deprtved cultures after 48 h (B). A marked increase of the percent-

neurons in control cultures.

ages of apoptotic

NMDA

serum-deprived

neurons

is observed

cultures compared

in NMDA-treated

cultures or

to control cultures using propidium

iodide stainings or the Tunei method.

’ P 5 0.05; ’ ’ P s 0.005: ’’* P

I 0.ooo1.

displays the results of the immunocytochemical study using AT8 or taul antibodies associated with propidium iodide stainings. In control cultures, several neuronal cell bodies are immunostained by AT8 (Fig. 3A) or taut (Fig. 3D) antibodies. In NMDA-exposed cultures immunolabelled with AT8 antibody, no apoptotic neurons are ATSpositive (Fig. 3B,C). At the opposite, in NMDA-treated cultures immunostained with tau 1 antibody, apoptotic nuTdbie

(A) or in serum-deprived

ATB-positive

of’ ATX-

or taul -positive

in cultures exposed for

I6

aptq~~rttc

h to 20

apoptotic neurons is comparable in NMDA-treated

cultures

or serum-deprived cultures and control cultures whereas the percentage of taui-positive

apoptotic neurons is clearly augmented in NMDA-treated

cuhures

serum-deprived

or

NMDA-treated

cultures

compared

to

control

cultures.

cultures were exposed after IO days whereas serum-de-

prived cultures were utilized after 7 days. This variation could explain the \iight difference between the percentages of AT8-positive two control groups. NS, non-significant.

neurons in the

’ ’ P s 0.005; * * * P 5 0.0001.

clei with condensed or fragmented chromatin are surrounded by a taul-immunopositive reaction (Fig. 3E,F). Similar findings were observed in serum-deprived cultures (data not shown).

i

Evaluation of the percentage of apoptotic neurons after propidium iodide stainings in controls cultures, in cu1turr.s treated with NMDA

(20 /.LM. I6 h) or

submitted to serum deprivation (24 h) in the absence (2nd column) or in the presence (3rd column) of the endonuciease inhibitor aurintricarboxylic

(ATA 50 jdw Treated cultures

Treated cultures + ATA

12.5 * 2.3%

37.8 +_ 4.3%

20.6 rt 2.9%

10.6 + 3.1%

27.7 + 4.2%

12.9 ) 5.4%

Controls NMDA

PM

cultures for 48 h (B). The percentage of

exposure

Serum deprivation

acid

130

M. ixsort et al./ Molentlw

BruCr Resmrch

45 f 1997) 127-132

8Ur results collfh preVi0Us reports showing that excitotoxicity or serum eprivation [I] can be ~~s~~~~~~~~fo,n inducing neuronal apoptosis in vitro [?I. A recent study revealed that necrosis or apoptosis can be produced by NMDA exposure in cortical cuhures depending upon the dose and the duration of the intoxication. Apoptosis in our cell culture system, is triggered by Pow concentrations of NMDA for a prolonged period of intoxication. Hnaddition, another in vivo study showed that excitotoxim injection into the rat striatlrm induces both types of cellular death that are observed using morphological and biochemical methods [7]. It is now well-demonstrated that NMDA toxicity is associated with an increased cellular influx of calcium, accompanied by the production of free radicals and the liberation of nitric oxyde [3]. Previous reports have shown that increased cellular calcium or free radicals are also responsible in vitro for the induction of apoptosis in different cell culture systems 127,253. In the present series of experiments using NMDA toxicity, one can assume that th~sc com.“,i~ed toxic iiiS;i:t5 XC as:;ocidtrd tu iriggcr the signal transduction pathways leading to apoptosis. Recently, Koh et al. also reported that cortical cell cultures exposed to serum-deprived medium undergo a widespread neuronal apoptosis that was markedly attenuated by the protein synthesis inhibitor cycloheximide [II?]. The production of free radicals and excitotoxicity could also be involved in the mechanisms leading to neuronal apoptosis in serum-deprived cultures [ll. The principal finding revealed by our observations consist in the fact that cortical neurons can be separated into two groups according to their ability to resist to, or delay the onset of apoptosis induced by NMDA or by serum deprivation. Resistant neurons without morphological signs of nuclear apoptosis are mostly immunostained by AT8 which recognizes tau phosphorylated epitopes at serine 202 and threonine 205 [l 11. At the opposite, sensitive neurons to apoptosis are rarely AT&positive but can be immunolabelled by taul raised against tau de-phosphorylated epitopes, suggesting that a process of tau de-phosphorylation occurs in a subgroup of apoptotic neurons chronically exposed to NMDA or serum deprivation. These observatians can explain the contradicting results showing that NMDA toxicity can enhance AT8 immunostaining in certain resistant neurons [4] and that glutamate toxicity increases taul inununostainings [5]. The presence of various types of neurons in cortical cell cultures and exposed to a toxic insult can account for these different modifications of tau immuno-

The

reactivity. itgi

10

gP0upS

but

exposed htion

nature

apoptosis

to

is

One

of nOt

can

~~Optotic

could

take

part

the

propose triggers in

relative

elucidated

the

that to

resistziace in

the

prevent

biochemical

these

or

ability tau

bensiti\-

neuronal f0r

subneurOns

de-phOsphOry-

ccPluPar

response,

with other biochemical pathways. allowing neuPOns t0 stop or delay apoptosis. Another recent study has shown that neuronal apoptosis induced by serum deprivation is accompanied by tau de-phosphorylation marked by an enhanced tau 1 immunostaining in PC I2 neuroblastoma cells deprived of NGF [22]. Tau de-phosphorylation at least in the vicinity of AT8 and taul epitopes seems to be a partial marker for neuronal apoptosis induced by excitotoxicity or serum deprivation in these two cellular systems. The phosphorylation and the de-phosphorylation of tau protein involves a certain number of kinases and phosphatases including MAP kinases, glycogene synthase kinascs, tau protein kinase!: [ I6j and phosphatase 2A and phosphatase 2E [28.12,13]. A recent study has reven!cd that calciiirur iii oi- plluqharnse 25 and phosphatase 2A can de-phosphorylate tau protein at serine 199 and 202, the AT8 epitopes [23]. In addition, the calcineurin inhibitors FK506 and cyclosporin A can reduce glutamate toxicity in neuronal cultures [19.6] and cyclosporin A can blocked apoptosis in immature thymocytes [30]. Further works will have to determine in neuronal cell cultures exposed to signals inducing apoptosis if the lnodifications of phosphatase activity can produce changes in apoptosis sensitivassociated

ity associated tau protein.

with changes

in the phosphorylation

state of

M.L. Autef and 1. Teissandier for typing the manuscript, F. Forestier for technical assistance. This work was supported by PHRC and ‘Conseil Regional du Limousin’. The authors

thank

References [I] Atabay.

C., Cagnoli, C.M.. Kharlamov. E.. Ikonomovic. M.D. and Manev. H.. Removal of serum from primary cuhres of cerebehr granule neurons induces oxidative stress and DNA

fragmentation:

protection with antioxidants and glutamate receptor antagonists. h’eu,osci.

RET.,

J.

43 (1996) 465-475.

30 PM NMDA for I6 h (B.C.E.F). Nuclei were Fig. 3. This figure shows the immunonuorescencestudy of control cultures (A-D) and cultures exposed to _ stained with propidium iodide and tau immunocytochemistry was carried out with AT8 antibody (A-C) or taul antibody (D--F): no apoprotic neurons Ue seen in control cultures (A-D). In B,C, apoptotic neurons with condensednucleus are not immunostained with AT8 antibody. Apoptotic neurons display round shape and condensed chromatin. The small nuclear debris also observed in C are not included in the cell count. At the opposite in E.F. apoptolic neurons with condensed nuc!eus (E) or fragmented nucleus (F) are also immunolabelled with taul antibody associated with an increasedimmunoreactivity. This figure demonstrate that neuronal apoptosisinduced by NMDA can be associated with a tau de-phosphorylation at the tau I epitope. A.B ( X 4300): C.D (X2400):E(x4200);F(x4400).

132

M. Lesort et nl. / Moiecukrr

Brcrirr Rescwrch -15

[2] Bonfoco. E.. Krainc. D.. Ankarcrona. M.. Nicotera. P. and Lipton. S.A., Apoptosis and necrosis: two distinct events induced. respectively. by mild and intense insults with N-methyl-o-aspartate or nitric oxide/superoxide in cortical cell cultures. Proc. Nd Ad. Sci. USA, 92 (1995) 7162-7166. [3] Choi, D.W.. Excitotoxic cell death, J. Neltrobiol.. 23 (1992) 12611276. [4] Couratier, P., Lesort, M., Sindou, P., Esciaire. F.. Yardin. C. and F?,lon, J.. Modifications of neuronal phosphorylated tau immunoreactivity induced by NMDA toxicity, Mol. Chern. Neuropathol., 27 (1996) 259-273. [S] Davis, R.D.. Brion, J.P., Couck. A.M., Gallo. J.M., Hanger, D.P.. Ladhani. K., Lewis, C., Miller, C.C.J., Rupniak, T.. Smith, C. and Anderton, B.H.. The phosphorylation state of the microtubule-associated protein tau as affected by glutamate, colchicine and fi-amyloid in primary rat cortical neuronal cultures, Biocken~. J.. 309 (1995) 94 I -949. [6] Dawson, T.M.. Steiner, J.P., Dawson, V.L.. Dinerman, J.L.. Uhl. G.R. and Snyder, S.H.. lmmunosuppressant FK506 enhances phosphorylation of nitric oxide synthase and protects against glutamate neurotoxicity. Proc. Nd Acud. Sci. USA, 90 ( 1993) 9808-9812. [7] Ferrer, I., Martin, F.. Serrano. T.. Reiriz, J.. Perez-Navarro, E., Alberch, J.. Macaya, I.. and Planas. A.M., Both apoptosis and necrosis occur following intrastriatal administration of excitotoxins. Actn Netrropathol., Qfl i 1995) 504-5 10. 18) Fraser, A., McCarthy, N. and Evan. G.I., Biochemistry of cell death. Curr. Opin. Neurobiol., 6 ( 1996) 7 l-80. [9] Garcia, I., Martinou. I., Tsujimoto, Y. and Martinou. J.C.. Prevention of programmed cell death of sympathetic neurons by the 6~1-2 proto-oncogene. Scietrce 258 (1992) 302-304. 1101 Goedert, M., Spillantini, M.G., Cairns. N.J. and Crowther, R.A., Tau proteins of Alzheimer paired helical filaments: abnormal phosphorylation of all six brain isoform,. Neuron. 8 (1992) I59- 168. [I I] Goeden. M., J&s, R. and Vanmechelen, E.. The monoclonal

antibody AT8 recognises tau protein phosphorylated at both serine 202 and threonine 205. Neurosci. L&t.. I89 ( I995) 167- 169. 1121

f /997) 127- 132

1161 Kosik,

KS.. The molecular and cellular biology of tau. Swirl 3 (1993) 39-43. 1171 Lo, A.C.. Houenou. L.J. and Oppenheim, R.W., Apopto\is in the nervous system: morphological features. methods. pathology and prevention. Arch. Him/. C~fol.. 58 ( 1995) IN- 119. [18] Mah, S.P.. Zhong. L.T., Liu. Y.. Roghoni. A.. Edwards. R.H. and Bredescen. D.E.. The protooncogene hcl- 2 inhibit\ apoptosis in PC12 cells. 1. .Neurochtm., 60 (1993) 1183-I I&h [19] Marcaida, G., Kosenko, E.. Minana. M.D., Gisolia, S. and Felipo, V., Glutamate induces a calcineurin-mediated dephosphorylation of Na’,K+-ATPase that results in its activation in cerebellar neurons in culture, J. Neurochem. 66 ( 19%) 99- 104. [20] Mattson, M.P.. Effects of microtubule stabilization and destabilization on tau immunoreactivity in cultured hippocampal neurons. Braitz Res., 582 (1992) 107-l 18. [21] McConkey. D.J. and Orrenius. S., Signal transduction h:.*thways to apoptosis. Trends Cell Bid., 4 ( 1994) 370-375. [22] Miles. J.C. and Pittman. R.N.. Tau protein is Jepho! :h(>ry!:!ted by an okadaic acid inhibitahle phnsphatase durin p neural apoprusis. Sot. Neurosci. Abstr.. 2 I ( 1995) 1304. 1231 Onn. T.. Yamamoto, H.. Tashima. K.. Nakashima. H.. Okumura, E.. Yamada, K.. Hisanaga. S.. Kishimotn. T.. Miyakawa. T. and Miyamoto, E., Dephosphorylation of abnormal sites of tau factor by protein phosphatases and its implication for Alzheimer’s disease. Neuruchem hzt., 26 ( 1995) 205-2 I5. [24] Rubin, L.L.. Gatchalian. C.L., Rimon, G. and Brooks, S.F., The molecular mechanisms of neuronal apoptosis. Clorr. 0pi1~. Neurobiof.. 4 ( I994) 696-702. Pm/d..

[25]

Sandstrom. P.A., Mannie. M.D. and Buttke. T.M.. Inhibition of activation-induced death in T cell hybridomas by thiol antioxidants: Bid., 55 oxidative stress as a mediator of apoptosis. J. Letrkxyr (1994)

[26]

X1-226.

Sindou. P., Lesort. M., Couratier. P.. Yardin. C.. Esclaire. F. and Hugon. J., Glutamate increases tau phosphorylation in primary neuronal cultures from fetal mt cerebral cortex, Bruirl Rrs., 646 (1994) 124-128.

Gong, C.X.. Singh. T.J.. Grundke-Iqbal, I. and Iqbal. K.. Alzheimer’s

[27] Takei,

disease abnormally phosphoylated tau is dephosphorylated by protein phosphatase-2B (calcineurin). J. Nmrwhertt.. 62 (I 994) 80.7806.

[I%] Wang, 3.2. Gong. C.X.. Grundke-lqbul.

1131 Goto, S.. Yamamoto. H.. Fukunaga. K.. Iwasa. T.. Matsukudo. Y. and Miyamoto, E., Dephosphorylation of microtubule-associated protein 2, tau factor tmd tubulin hy calcineurin, J. Netwochettr., 45 (1995) 276-283. 1141 Hu@xt. J., Hugon, F., Esclaire, F., Lesort, M. and Diop, A.G., The presence of calbindin in neuron in cultures protects from oxydative stress, Brain Res.. 707 (1996) 288-292. [15] Koh. J.Y.. Gwag, B.J.. Lobner. D. and Choi, D.W.. Potentiated necrosis of cultured cortical neurons by neurotrophins, Science, 268 (1995) 573-57s.

N. and Endo. Y., Ca” innophore-induced apoptosis on cultured embryonic rat cortical neurons. Bruits Rcs.. 6.52 (1994) 65 .-70.

I. and Iqbal. K., Dephos-

phorylntion of Alzheimer paired helical filaments by protein phosphatase 2A and 2B. Newobiol. Aging ADstr.. I5 (8~&.) (1994) S23. 1291 Wyllie. A.H., Kerr, J.F.R. and Currie, A.R., Cell death: the significance of apoptosis, ItIt. Rev. Cytol., 68 (1980)25 I-307. [30] Yazdanbakbsh, Cyclosporin A activity of the USA, 92 (1995)

K.. Choi, J.W., Li, Y.. Lau. L.F. and Choi, Y., blocks apoptosis by inhibiting the DNA binding transcription factor Nur77. Proc. Nutl. Acad. Sci. 437-441.