Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

b r a i n r e s e a r c h 1 5 4 9 ( 2 0 1 4 ) 5 2 – 6 2

0006-8993/$ - see frohttp://dx.doi.org/10.

Abbreviations: AT

protein homologous

kinase; eIF2α, euka

MTT, 3-[4,5-dimethnCorrespondence

Nanjing, Jiangsu 21E-mail addresses:

Research Report

Salubrinal protects against rotenone-inducedSH-SY5Y cell death via ATF4-parkin pathway

Liang Wua, Na Luoa, Hong-Rui Zhaoa, Qing Gaoa, Jie Lub, Yang Panb,Jing-Ping Shib, You-Yong Tiana,n, Ying-Dong Zhanga,n

aDepartment of Neurology, Nanjing First Hospital, Nanjing Medical University, Nanjing, Jiangsu 210006, PR ChinabDepartment of Neurology, Affiliated Nanjing Brain Hospital, Nanjing Medical University, Jiangsu 210029, PR China

a r t i c l e i n f o

Article history:

Accepted 4 January 2014

Parkinson s disease (PD) is a progressive neurodegenerative disorder, for which there are no

effective disease-modifying therapies. Growing evidence from studies in human PD brain,

Available online 10 January 2014

Keywords:

Salubrinal

Rotenone

ER stress

ATF4

Parkin

nt matter & 2014 Elsevie1016/j.brainres.2014.01.00

F4, activating transcrip

protein; ER, endoplasm

ryotic translation initiat

yl-thiazol-2-yl]-2,5-diphto: Department of Neuro0006, PR China. Fax: þ86tyy9956@163.com (Y.-Y.

a b s t r a c t

0

in addition to genetic and toxicological models, indicates that endoplasmic reticulum (ER)

stress is a common feature of the disease and contributes to neurodegeneration. We examine

whether salubrinal, a ER stress inhibitor, can protect the rotenone-induced SH-SY5Y cell death

and explore the mechanisms underlying this protection. Our results demonstrated that

rotenone induced a significant ER stress response and caused cell apoptosis, which was

inhibited by salubrinal. Activating transcription factor 4 (ATF4), a member of the ATF/CREB

family of basic leucine zipper transcription factors, has been implicated in the pathogenesis of

neurodegeneration. We showed that salubrinal increased the up-regulation of ATF4 expres-

sion. An ATF4 siRNA significantly increased the rotenone cytotoxicity and decreased the

salubrinal0s protection. Further, we showed that ATF4 siRNA inhibited the expression of

parkin, and parkin knockdown similarly aggravated the rotenone cytotoxicity and reduced the

salubrinal0s protection. Additionally, the protein level of parkin was declined after treatment

with rotenone, whereas this reduction was rescued by salubrinal. These findings indicate

ATF4-parkin pathway plays an important role in the salubrinal-mediated neuroprotection of

rotenone-induced dopaminergic cell death.

& 2014 Elsevier B.V. All rights reserved.

r B.V. All rights reserved.3

tion factor4; C/EBP, CCAAT-enhancer-binding protein; CHOP, CCAAT-enhancer-binding

ic reticulum; PERK, ER membrane proteins double-stranded RNA-activated kinase-like

ion factor 2 α; GRP78, glucose regulated protein 78; IRE1α, inositol requiring 1α enzyme;

enyl tetrazolium bromide; PD, Parkinson0s diseaselogy, Affiliated Nanjing First Hospital, Nanjing Medical University, No. 68 Changle Road,25 52269924.Tian), zhangyingdong@aliyun.com (Y.-D. Zhang).

1. Introduction

Parkinson0s disease (PD) is a progressive neurodegenerativedisorder characterized by the death of dopaminergic neurons

in the substantia nigra (Martin et al., 2011). The mechanismsof neuronal loss remain unclear, and current therapiesprincipally alleviate symptoms instead of targeting theunderlying neuronal loss (Levy et al., 2009). Growing evidence

Foid

b r a i n r e s e a r c h 1 5 4 9 ( 2 0 1 4 ) 5 2 – 6 2 53

from research in human PD brain, in addition to genetic andtoxicological models, indicates that endoplasmic reticulum(ER) stress is a common feature of PD and contributes todopaminergic neurodegeneration (Hoozemans et al., 2007;Colla et al., 2012; Chen et al., 2008). Therefore, therapeuticstrategies that mitigate ER stress may be beneficial to PDpatients.

ER stress is known to activate a series of signals thatcomprise the unfolded protein response (UPR). The UPR ismediated through three signaling pathways under the controlof the ER membrane proteins double-stranded RNA-activatedkinase-like kinase (PERK), activating transcription factor(ATF)-6 and inositol requiring-1α (IRE1α) (Ron and Walter,2007). These signals coordinate the cellular response tounfolded proteins, which includes (1) downregulation ofprotein translation via activation of PERK and subsequentphosphorylation of the eukaryotic translation initiation factor2 α (eIF2α) at Ser51, (2) enhanced expression of ER chaperoneproteins that promote protein refolding, and (3) activation ofproteases involved in the degradation of misfolded proteins(Walter and Ron, 2011). However, prolonged or severe ERstress can lead to the activation of apoptotic cell deathpathways. Rotenone is an environmental toxin used toinduce experimental Parkinsonism in animals and cell cul-tures. Many studies reveal that rotenone induces ER stress(Chen et al., 2008; Ryu et al., 2002), and that eIF2α siRNA candecrease rotenone cytotoxicity in SK-N-MC cells (Chen et al.,

ig. 1 – Effect of rotenone on ER stress and the death of SH-SY5Y cr thapsigargin (Tg, 5 μM) for the specified times. The expressionmmunoblotting. The relative amounts of GRP78 (B), CHOP (C) andensitometry and normalized to the expression of β-actin. npo0.

2008), suggesting that ER stress mediates rotenone-inducedcell death.

The effects of ATF4, a member of the ATF/CREB family ofbasic leucine zipper transcription factors, on neuronal survi-val or death are complex. Previous studies found that rote-none induces PERK phosphorylation, which results in ATF4and CHOP (CCAAT-enhancer-binding protein (C/EBP) homo-logous protein) protein induction in neuronal PC12 cells (Ryuet al., 2002). Indeed, in a previous study, we found that ATF4may be involved in dopaminergic cell death in the rotenonerat model (Wu et al., 2013). However, many target genes areregulated by ATF4 (Singleton and Harris, 2012), includingparkin, which is transcriptionally upregulated by ER stressand can protect cells from ER stress-induced cell death (Sunet al., 2013; Imai et al., 2000). Moreover, a dominant-negativeATF4 mutant prevents the ER stress-induced upregulation ofparkin (Bouman et al., 2011). And a recent report showed thatATF4-parkin pathway protects against neuronal deathinduced by 6-hydroxydopamine (6-OHDA) and 1-methyl-4-phenyl-pyridinium (MPPþ) (Sun et al., 2013). However, it isunknown whether ATF4-parkin pathway is beneficial orharmful in the rotenone model.

Salubrinal, a phosphatase inhibitor, selectively inhibitsdephosphorylation of the α subunit of eIF2 which is anupstream activator of ATF4. It has been shown that salubrinalcan modulate the cellular stress response to protect againstER stress-induced apoptosis (Boyce et al., 2005). In the present

ells. (A) SH-SY5Y cells were treated with rotenone (100 nM)levels of GRP78, CHOP and caspase-3 were determined withcleaved caspase-3 (D) imaged on the films was measured by05, nnpo0.01, vs. control group, n¼3.

F4Tc1DQ

b r a i n r e s e a r c h 1 5 4 9 ( 2 0 1 4 ) 5 2 – 6 254

study, we use rotenone-induced apoptosis in SH-SY5Y cellsas an in vitro model of dopaminergic neuronal degeneration.We investigate the effects and mechanisms of action ofsalubrinal on this model of neuronal cell death.

2. Results

2.1. Endoplasmic reticulum (ER) stress is involved inrotenone-induced death of SH-SY5Y cells

Previously, it was shown that ER stress plays a critical role inrotenone-induced death in SK-N-MC neuroblastoma cells(Chen et al., 2008). Here, we sought to confirm and extendthese observations by examining ER stress in rotenone-induced SH-SY5Y cell death. The levels of cleaved caspase-3increased significantly when SH-SY5Y cells were exposed torotenone for 6 h or longer (Fig. 1A and D). And GRP78 (glucoseregulated protein 78), a marker of ER stress, was upregulatedsignificantly with rotenone0s exposure periods of 6 h or longer(Fig. 1A and B). Moreover, rotenone had a modest effect onthe expression of CHOP (Fig. 1A and C). As a positive control,we treated cells with thapsigargin, a well-known ER stress

ig. 2 – Effect of salubrinal on rotenone-induced neuronal death.0 μM salubrinal (Sal) or Rot plus Sal for the indicated time. Cleavehe amounts of cleaved caspase-3 were measured by densitomeontrol group, ##po0.01, n¼3. (C) Apoptosis was investigated usi00 nM rotenone (Rot), 40 μM salubrinal (Sal) or Rot plus Sal. nnpo0ose-dependent protection by salubrinal of SH-SY5Y cells treateduantitative analysis of cell viability using MTT assay is shown.

inducer, and observed robust elevation of GRP78 and CHOPexpression. Similar to rotenone, thapsigargin induced deathof SH-SY5Y cells.

2.2. Salubrinal attenuates rotenone-induced SH-SY5Ycells death

To investigate whether salubrinal protects against rotenone-induced dopaminergic cell death, SH-SY5Y cells were treatedwith rotenone in the presence or absence of salubrinal. Wefound that with 6- and 12-h treatment periods, roten-one caused dramatic activation of caspase-3, while salubrinalsuppressed this rotenone-induced activation of caspase-3(Fig. 2A and B).

To further clarify the effect of salubrinal on rotenone-induced apoptosis, SH-SY5Y cells were double-labeled forannexin V and PI, and analyzed on a FACSCalibur system.Rotenone treatment for a 12-h period induced apoptosis in12.6771.74% of SH-SY5Y cells (Fig. 2C). In comparison, theapoptosis rate in SH-SY5Y cells treated with both rotenoneand salubrinal was 7.7271.4% (Fig. 2C), consistent with thefindings on caspase-3 activation. We also examined cellviability using the MTT assay. As shown in Fig. 2D, while

(A) SH-SY5Y cells were treated with 100 nM rotenone (Rot),d caspase-3 levels were determined by immunoblotting. (B)try and normalized to β-actin levels.

nnpo0.01, vs. matchedng flow cytometric analysis at 12 h after treatment with.01, vs. matched control group, #po0.05, ##po0.01, n¼3. (D)for 12 h with 0, 5, 10, 20, 40, 80 or 160 μM salubrinal (Sal).

n

po0.05, n¼3.

b r a i n r e s e a r c h 1 5 4 9 ( 2 0 1 4 ) 5 2 – 6 2 55

the viability of neurons decreased with 12 h of rotenonetreatment, salubrinal significantly inhibited rotenone-inducedneuronal cell death in a dose-dependent manner.

2.3. Salubrinal-mediated cytoprotection of SH-SY5Y cellsis dependent on ATF4

Salubrinal is known to protect against ER stress-inducedcytotoxicity by selectively inhibiting the dephosphorylationof eIF2α (Boyce et al., 2005). ATF4 is a downstream effector ofphosphorylated eIF2α. We sought to determine whethersalubrinal-mediated cytoprotection of SH-SY5Y cells isdependent on ATF4. Consistent with previous studies, salu-brinal significantly elevated the levels of phosphorylatedeIF2α, but did not affect the levels of total eIF2α (Fig. 3A,C and D). As shown in Fig. 4A and B, ATF4 levels modestlyincreased after treatment with rotenone for 6 or 12 h. Whencells were incubated simultaneously with rotenone andsalubrinal for 6 or 12 h, salubrinal enhanced the upregulationof ATF4 expression by 29.4% and 58.4%, respectively. Theseresults were confirmed by immunofluorescence assay usingan anti-ATF4 antibody; incubation with salubrinal and rote-none for 12 h markedly elevated ATF4 immunoreactivity in

Fig. 3 – Effect of salubrinal on the protein expression of eIF2α, p(A) SH-SY5Y cells were treated with 100 nM rotenone (Rot), 40 μexpression levels of eIF2α, p-eIF2α and CHOP were determined wp-eIF2α (C) and CHOP (D) imaged on the films were measured bnpo0.05, nnpo0.01, vs. matched control group, ##po0.01, n¼6.

SH-SY5Y cells, although rotenone alone also increased immu-noreactivity modestly (Fig. 4C).

We further examined the effect of ATF4 on salubrinal-mediated cytoprotection of SH-SY5Y cells. ATF4 siRNA wasused to knock down the expression of ATF4 (Fig. 5C). ATF4siRNA significantly decreased cell viability and enhancedrotenone-induced apoptosis (Fig. 5A and B), indicating thatATF4 protects against ER stress-induced cell death. The ATF4siRNA also diminished salubrinal0s protective effect againstthe rotenone-induced death of SH-SY5Y cells, suggesting thatATF4 is necessary for salubrinal to protect against ER stress-induced cell death. In addition, we found that ATF4 siRNAsignificantly decreased the expression of CHOP, which wasnot upregulated after treatment with rotenone or salubrinal(Fig. 5D and E).

2.4. Parkin mediates the cytoprotection of ATF4

To explore how ATF4 exerts its cytoprotective effect in therotenone-induced death of SH-SY5Y cells, we examined thetarget genes of ATF4. Parkin is a component of the UPR in thePERK/ATF4 pathway, and protects cells from ER stress-induced cell death (Bouman et al., 2011). In our study, ATF4siRNA significantly diminished the levels of parkin, but

-eIF2α and CHOP in SH-SY5Y cells treated with rotenone.M salubrinal (Sal) or Rot plus Sal for the indicated time. Theith immunoblotting. The relative protein levels of eIF2α (B),

y densitometry and normalized to the expression of β-actin.

Fig. 4 – Effect of salubrinal on rotenone-mediated induction of ATF4 expression in SH-SY5Y cells. (A) SH-SY5Y cells weretreated with 100 nM rotenone (Rot), 40 μM salubrinal (Sal) or Rot plus Sal for the indicated time. The expression of ATF4 wasdetermined with immunoblotting. (B) The relative amount of ATF4 protein imaged on the films was measured bydensitometry and normalized to the expression of β-actin. npo0.05, nnpo0.01, vs. control group, #po0.05, ##po0.01, n¼3.(C) Immunocytochemistry showing the effect of salubrinal on rotenone-induced ATF4 expression in SH-SY5Y cells. Aftertreatment with rotenone (100 nM) for 12 h, SH-SY5Y cells were fixed and immunolabeled with anti-ATF4 antibodies,followed by FITC-conjugated secondary antibodies. The nucleus was counterstained with DAPI. Fluorescence images(40� magnification): green, ATF4; blue, DAPI. Scale bar, 50 μm.

b r a i n r e s e a r c h 1 5 4 9 ( 2 0 1 4 ) 5 2 – 6 256

scrambled siRNA did not change the expression of parkin(Fig. 6A). Similar to the findings in a previous study on6-OHDA and MPPþ-induced death of neuronal PC12 cells(Sun et al., 2013), rotenone decreased the expression ofparkin, but salubrinal inhibited this downregulation of parkinexpression (Fig. 6A).

To further investigate that the effect of parkin onsalubrinal-mediated cytoprotection of SH-SY5Y cells, we kno-cked down the expression of parkin using parkin siRNA. Asindicated in Fig. 6C and E, parkin siRNA significantlyincreased the apoptosis percent and inhibited the cell viabi-lity, and salubrinal did not reduce the apoptosis percent andincrease the cell viability in parkin siRNA cells. This suggeststhat salubrinal may protect cells through the ATF4-parkinpathway.

2.5. Salubrinal reduces the expression of LIP and inhibitsthe CHOP levels in nucleus

ATF4 induces expression of the transcription factorCHOP in many ER stress paradigms (Harding et al., 2000;

Averous et al., 2004). CHOP is considered a major mediator ofapoptotic cell death (Silva et al., 2005). As shown in Figs. 3Aand 5D, CHOP expression is upregulated significantly inrotenone-treated SH-SY5Y cells, but salubrinal did notdecrease CHOP expression. Because the nuclear transloca-tion of CHOP is a necessary step for the induction ofapoptosis, we wondered the effect of salubrinal on CHOPlevels in cell nucleus. We checked the CHOP expression inthe nuclear extracts and detected that the increase in CHOPexpression caused by rotenone was abolished by salubrinal(Fig. 7A). It has shown that the bZIP protein C/EBP isoformLIP is required for nuclear translocation of CHOP during ERstress, and the CHOP–LIP interaction plays a critical role inthe induction of apoptosis (Chiribau et al., 2010). Therefore,we examined changes in LIP expression in rotenone-treatedSH-SY5Y cells co-treated with salubrinal. As indicated inFig. 7B, rotenone did not change the expression of LIP in SH-SY5Y cells, but the levels of LIP decreased significantly upontreatment with salubrinal. This suggests that, althoughsalubrinal increases CHOP levels in SH-SY5Y cells, CHOPmay not exacerbate apoptosis.

Fig. 5 – Effect of ATF4 on rotenone-induced neuronal death in SH-SY5Y cells. (A) SH-SY5Y cells were transfected with ATF4siRNA or a scrambled siRNA for 48 h, and then treated with 100 nM rotenone (Rot), 40 μM salubrinal (Sal) or Rot plus Sal for12 h. Cell viability was determined by MTT assay and expressed as a percentage of the control. npo0.05, nnpo0.01, vs.matched group, &po0.05, vs. control group, #po0.05, ##po0.01, n¼6. (B) SH-SY5Y cells were transfected with a scrambledsiRNA or ATF4 siRNA for 48 h, and then treated with 100 nM rotenone (Rot), 40 μM salubrinal (Sal) or Rot plus Sal for 12 h.Apoptosis was investigated using flow cytometric analysis. npo0.05, nnpo0.01, vs. matched group, &&po0.01, vs. controlgroup, #po0.05, ##po0.01, n¼3. (C) SH-SY5Y cells were transfected with a scrambled siRNA or ATF4 siRNA for 48 h. Theexpression of ATF4 was determined with immunoblotting. (D) SH-SY5Y cells were transfected with ATF4 siRNA or ascrambled siRNA for 48 h. The expression of CHOP was determined with immunoblotting. (E) The relative amount of CHOPprotein imaged on the films was measured by densitometry and normalized to the expression of β-actin.nnpo0.01, vs.matched group, ##po0.01, n¼3.

b r a i n r e s e a r c h 1 5 4 9 ( 2 0 1 4 ) 5 2 – 6 2 57

3. Discussion

In the present study, we provide data showing that treatmentwith salubrinal attenuates rotenone-induced neuronal death.We also demonstrate that ATF4 is necessary for thesalubrinal-mediated cytoprotection. Moreover, we find thatsalubrinal protects against rotenone neurotoxicity throughthe ATF4-parkin pathway.

Rotenone, a mitochondrial complex I inhibitor, possesseshighly selective toxicity towards dopaminergic neurons in vitro(Hartley et al., 1994) and in vivo (Betarbet et al., 2000). Androtenone has been extensively used to experimentally modelPD pathogenesis (Bove and Perier, 2012; Xiong et al., 2012).A recent study found that ER stress mediates rotenone-induceddeath of SK-N-MC neuroblastoma cells, and that ER stress-mediated induction of the apoptotic process may precedeoxidative stress within cells (Chen et al., 2008). In fact, accu-mulating evidence suggests that ER stress is a common patho-logical feature of PD, and is found in models of both the familialand sporadic forms of the disease (Colla et al., 2012; Holtz and

O0Malley, 2003). ER stress markers have been reported in thesubstantia nigra pars compacta of human post-mortem spora-dic PD cases (Hoozemans et al., 2007; Slodzinski et al., 2009).Numerous studies suggest that ER stress is an early componentin PD pathogenesis, even preceding oxidative stress (Chen et al.,2008). Thus, potential therapeutic strategies that target ER stressmay be beneficial for PD patients. Salubrinal is a selectiveinhibitor of eIF2α dephosphorylation that was recently foundto protect against ER stress-mediated apoptosis (Boyce et al.,2005). Undifferentiated SH-SY5Y cells are identified as a suitablein vitro model for studying the molecular and cellular mechan-isms underlying the pathophysiology of PD as well as forevaluating pharmacological interventions (Xie et al., 2012;Yong-Kee et al., 2012; Pan et al., 2009; Cheung et al., 2009).In our studies, salubrinal protected against rotenone-inducedSH-SY5Y cell death. This finding is in line with that of a recentstudy showing that salubrinal attenuates β-amyloid-inducedneuronal death (Huang et al., 2012). However, that study foundthat inhibition of the NF-κB pathway may be associated withsalubrinal0s protective effect.

Fig. 6 – Effect of Salubrinal on the expression of parkin in SH-SY5Y cells treated with rotenone. (A) SH-SY5Y cells weretransfected with a scrambled siRNA or ATF4 siRNA for 48 h, and then treated with 100 nM rotenone (Rot), 40 μM salubrinal(Sal) or Rot plus Sal for 12 h. The expression of cleaved parkin was determined with immunoblotting. (B) The relative amountof parkin protein imaged on the films was measured by densitometry and normalized to the expression of β-actin. npo0.05,nnpo0.01, vs. matched group; &po0.05, vs. control group, ##po0.01, vs. rotenone group, n¼3. (C) SH-SY5Y cells weretransfected with a scrambled siRNA or parkin siRNA for 48 h, and then treated with 100 nM rotenone (Rot), 40 μM salubrinal(Sal) or Rot plus Sal for 12 h. Apoptosis was investigated using flow cytometric analysis. npo0.05, nnpo0.01, vs. matchedgroup, &&po0.01, vs. control group, #po0.05, ##po0.01, n¼3. (D) SH-SY5Y cells were transfected with a scrambled siRNA orparkin siRNA for 48 h. The expression of parkin was determined with immunoblotting. (E) SH-SY5Y cells were transfectedwith a scrambled siRNA or parkin siRNA for 48 h. Cell viability was determined by MTT assay and expressed as a percentageof the control. npo0.05, nnpo0.01, vs. matched group, &po0.05, vs. control group, #po0.05, ##po0.01, n¼6.

b r a i n r e s e a r c h 1 5 4 9 ( 2 0 1 4 ) 5 2 – 6 258

Salubrinal blocks eIF2α dephosphorylation, thereby main-taining phosphorylation of the protein. Although globalprotein synthesis can be suppressed by phosphorylated eIF2α,

ATF4 expression was reported to be increased by eIF2αphosphorylation (Dever, 2002). In our present study, salu-brinal enhanced the upregulation of ATF4 expression. Using

Fig. 7 – Effect of Salubrinal on the nuclear protein levels of CHOP and the expression of LIP in SH-SY5Y cells treated withrotenone. SH-SY5Y cells were treated with 100 nM rotenone (Rot), 40 μM salubrinal (Sal) or Rot plus Sal for 12 h. (A) The proteinlevels of CHOP from nuclear extracts were determined with immunoblotting. (B) The relative level of CHOP imaged on thefilms was measured by densitometry and normalized to the expression of histone H3. nnpo0.01, vs. control group, ##po0.01,vs. rotenone group, n¼3. (C) The protein levels of C/EBPβ were determined with immunoblotting. (D) The relative level of LIPimaged on the films was measured by densitometry and normalized to the expression of β-actin. npo0.05, vs. control group;#po0.05, vs. rotenone group, n¼3.

b r a i n r e s e a r c h 1 5 4 9 ( 2 0 1 4 ) 5 2 – 6 2 59

siRNA-mediated knockdown, we demonstrate that ATF4 isessential to salubrinal0s protective effect. Our study is inagreement with a recent report showing that ATF4 protectsagainst neuronal cell death induced by 6-OHDA and MPPþ

(Sun et al., 2013). In addition, it has been reported that ATF4-null neurons are more sensitive to DNA-damaging agents(Galehdar et al., 2010), and that activating mutations in ATF4reduce glutamate toxicity (Lewerenz et al., 2012). Collectively,these studies and the present study demonstrate thatATF4 has a cytoprotective function. However, a couple ofstudies have suggested that ATF4 may exacerbate cell death(Galehdar et al., 2010; Lange et al., 2008; Han et al., 2013).ATF4-null mice exhibit significantly smaller infarcts andimproved behavioral recovery compared with wild-type micein a rodent model of ischemic stroke (Lange et al., 2008), andcortical neurons from ATF4-null mice are resistant to oxida-tive stress-induced cell death (Lange et al., 2008). Addition-ally, Galehdar et al. (2010) has found that the ATF4–CHOP–PUMA pathway is a key signaling pathway in ER stress-induced neuronal death, and that ATF4-deficient neuronsare more resistant to ER stress. These contrasting findingsindicate that the effects of ATF4 on neuronal survival arecomplex. ATF4 can promote either cell survival or deathdepending on the specific cell type or stressor. A factor likely

contributing to the differential effects of ATF4 on neuronalsurvival is the set of transcriptional targets regulated byATF4. Specially, ATF4 regulates the expression of parkin,which is transcriptionally upregulated by ER stress and canprotect cells from ER stress-induced cell death (Imai et al.,2000). In fact, a dominant-negative ATF4 mutant prevents ERstress-induced upregulation of parkin (Imai et al., 2000).However, ATF4 overexpression had little effect on parkinmRNA. These findings indicated that ATF4 is necessary butnot sufficient for transactivation of parkin (Sun et al., 2013).Consistent with the previous study, ATF4 knockdown led to adecrease in parkin protein and increase rotenone inducedapoptosis in our study.

In our studies, rotenone led to a substantial fall in parkinprotein levels. Sun et al. (2013) observed this effect in bothdifferentiated PC12 cells and primary cortical neurons andsuggested parkin levels decreased via enhanced proteasomaldegradation. The parkin has a protective role in PD cellularand animal models (Sun et al., 2013; Lewerenz et al., 2012;Yasuda et al., 2011), and recent study demonstrated that anincrease in parkin expression could slow aging both at thelevel of biochemical and molecular marker of aging (Ranaet al., 2013). Of note, enhancing ATF4-parkin pathway repre-sents a potential neuroprotective strategy in PD (Sun et al.,

b r a i n r e s e a r c h 1 5 4 9 ( 2 0 1 4 ) 5 2 – 6 260

2013). The same opinion has been confirmed in our presentstudy. Apoptosis percent was enhanced in parkin-siRNA SH-SY5Y cells. Similar to ATF4, the ER stress inhibitor salubrinalincreased the protein levels of parkin, which may be causedby inhibiting proteasomal degradation.

Interestingly, ATF4 also regulates the expression of CHOP,which was found to mediate apoptotic death in substantianigra dopamine neurons in an in vivo neurotoxin modelof Parkinsonism (Shin et al., 2011). However, it has alsobeen reported that CHOP possesses neuroprotective effects(Halterman et al., 2010). In present study, we observed thatsalubrinal increases protein expression of CHOP, but salurinalabolished the increase in CHOP expression of the cell nucleus.LIP, which is one of the three functionally relevant isoformsof C/EBP, interacts with CHOP during ER stress and transportsit into the nucleus (Chiribau et al., 2010). CHOP is then able tomodulate the levels of its target genes by repressing theexpression of Bcl-2 and other antiapoptotic genes (Huanget al., 2009; Su and Kilberg, 2008). We found that LIP levels,which did not change in the rotenone-treated neurons, weredecreased significantly by salubrinal. Taken together, ourfindings suggest that CHOP does not translocate to thenucleus during early ER stress, and therefore cannot induceapoptosis, in our model.

In summary, we demonstrate that salubrinal protects againstrotenone-induced dopaminergic neuron death, and that therole of the cytoprotection may depend on the activation ofATF4. Although the effect of ATF4 activation on neuronalsurvival is complex, our study provides an insight intothe potential involvement of ATF4-parkin pathway in thesalubrinal-mediated neuroprotection of rotenone-induced dopa-minergiccell death.

4. Experimental procedures

4.1. Chemicals and antibodies

Salubrinal was obtained from Tocris BioScience (R&D, Ellis-ville, MO, USA). Rotenone was purchased from Sigma-Aldrich(St. Louis, MO, USA). Antibodies used in this study were anti-cleaved caspase-3, anti-eIF2α, anti-phospho-eIF2α (Ser51), anti-ATF4, anti-GRP78, anti-CHOP, anti-parkin from Cell SignalingTechnology (Danvers, MA, USA), anti-C/EBPβ from Abcam(Cambridge, MA, USA), anti-histone H3 and anti-β-actin fromSanta Cruz (Dallas, TX, USA), and. FITC-conjugated secondaryantibody was purchased from Zhongshan Inc. (Beijing, China).

4.2. Cell cultures

Human neuroblastoma SH-SY5Y cells (ATCC) were culturedin RPMI 1640 medium supplemented with 10% heat-inactivated fetal calf serum containing 2 mM glutamine,100 units/ml penicillin and 100 μg/ml streptomycin under ahumidified atmosphere of 95% air–5% CO2 at 37 1C.

4.3. MTT assay for cell viability

Cell viability was assessed using MTT (3-[4,5-dimethyl-thia-zol-2-yl]-2,5-diphenyl tetrazolium bromide). Briefly, SH-SY5Ycells were collected and seeded in 96-well plates at a densityof 1�105 cells/well. After the indicated treatment, 20 μL ofMTT tetrazolium salt (Sigma-Aldrich) was added to each wellfor 3 h at 37 1C. Afterwards, the growth medium was replacedwith dimethyl sulfoxide, and the absorbance of each well wasmeasured with a plate reader using a test wavelength of490 nm and a reference wavelength of 630 nm.

4.4. Apoptosis assay

Quantitative evaluation of apoptosis was performed usingflow cytometry after double labeling with an FITC Annexin VApoptosis Detection kit (BD Biosciences, NJ, USA) to discrimi-nate between early apoptotic (annexin V-positive) and necro-tic (annexin V-propidium iodide (PI) double-positive) cells.SH-SY5Y cells were cultured in 6-well plates, and were thentreated with 100 nM rotenone, 40 μM salubrinal, or rotenoneplus salubrinal for different periods of time. The cells wereharvested after treatment with 0.25% trypsin, washed withphosphate-buffered saline (PBS) and incubated in 1� BindingBuffer containing FITC annexin V and PI at 37 1C in the darkfor 15 min. The apoptosis rate¼ [annexin V(þ)PI (� )cellsþannexin V(þ)PI (þ) cells]/total cells�100%. The specificfluorescence of 10,000 cells was analyzed with a FACSCalibursystem (BD Biosciences) within 1 h after the antigen–antibodyreaction. Data were analyzed using FSC express version 3.0software (De Novo Software, Los Angeles, CA, USA).

4.5. Immunofluorescence staining

After the indicated experimental treatment, SH-SY5Y cellswere fixed with 4% paraformaldehyde and permeabilized with0.5% Triton X-100. The cells were then incubated in blockingsolution followed by overnight incubation with monoclonalrabbit anti-ATF4 (1:200). After washing with PBS, the cells wereincubated with FITC-conjugated goat-anti-rabbit IgG (1:100).The nuclei were counterstained with DAPI using VectashieldMounting Mediumwith DAPI (Vector Laboratories, Burlingame,CA, USA). Immunostaining was visualized at 40� magnifica-tion under a fluorescence microscope.

4.6. Small interfering RNA transfection

ATF4 or parkin small interfering RNA (siRNA) and its scr-ambled siRNA (Santa Cruz Biotechnology, Santa Cruz, TX,USA) were transfected into SH-SY5Y cells with Lipofectamine2000 reagent (Invitrogen, Life Technologies, USA) according tothe manufacturer0s protocol. At 48 h after transfection, thecells were subjected to immunoblotting analysis for ATF4expression.

4.7. Western blot analysis

Western blot analysis was performed as described previously(Wu et al., 2013). Briefly, total protein extracts (20 μg) ornuclear protein extracts (10 μg) were separated by

b r a i n r e s e a r c h 1 5 4 9 ( 2 0 1 4 ) 5 2 – 6 2 61

electrophoresis on 10–15% SDS–polyacrylamide gel electro-phoresis (SDS–PAGE) gels, transferred to polyvinylidene fluor-ide (PVDF) membranes, and blocked in 5% bovine serumalbumin in 1� Tris-buffered saline containing 0.1% Tween20 (1� TBST). The membranes were sequentially incubatedwith primary antibodies and horseradish peroxidase (HRP)-conjugated secondary antibodies. After washing with 1�TBST, protein bands were detected with chemiluminescenceHRP substrate (SuperSignal West Pico; Thermo Scientific Inc.,Pittsburgh, USA) and exposed to X-ray film (Fujifilm Inc.,Tokyo, Japan). Band intensities were analyzed using Image J1.44 (NIH, Bethesda, MD, USA).

4.8. Statistical analysis

All data are presented as mean7standard error of the mean(SEM). The alterations were analyzed using analysis of var-iance (ANOVA) with SPSS software (version 13.0, SPSS, Chi-cago, IL, USA). P-valueso0.05 were considered statisticallysignificant.

Acknowledgments

This study was supported by grants from National NaturalScience of China (No. 81271418), Natural Science Foundationof Jiangsu Province (No. BK2012524), Six talent Summit ofJiangsu Province (No. 2012-WS-086), and the Research andInnovation Project for College Graduates of Jiangsu Province(No. CXZZ12_0576).

r e f e r e n c e s

Averous, J., Bruhat, A., Jousse, C., Carraro, V., Thiel, G.,Fafournoux, P., 2004. Induction of CHOP expression by aminoacid limitation requires both ATF4 expression and ATF2phosphorylation. J. Biol. Chem. 279, 5288–5297.

Betarbet, R., Sherer, T.B., MacKenzie, G., Garcia-Osuna, M.,Panov, A.V., Greenamyre, J.T., 2000. Chronic systemic pesticideexposure reproduces features of Parkinson0s disease. Nat.Neurosci. 3, 1301–1306.

Bouman, L., Schlierf, A., Lutz, A.K., Shan, J., Deinlein, A., Kast, J.,Galehdar, Z., Palmisano, V., Patenge, N., Berg, D., Gasser, T.,Augustin, R., Trumbach, D., Irrcher, I., Park, D.S., Wurst, W.,Kilberg, M.S., Tatzelt, J., Winklhofer, K.F., 2011. Parkin istranscriptionally regulated by ATF4: evidence for aninterconnection between mitochondrial stress and ER stress.Cell Death Differ. 18, 769–782.

Bove, J., Perier, C., 2012. Neurotoxin-based models of Parkinson0sdisease. Neuroscience 211, 51–76.

Boyce, M., Bryant, K.F., Jousse, C., Long, K., Harding, H.P., Scheuner, D.,Kaufman, R.J., Ma, D., Coen, D.M., Ron, D., Yuan, J., 2005. Aselective inhibitor of eIF2alpha dephosphorylation protects cellsfrom ER stress. Science 307, 935–939.

Chen, Y.Y., Chen, G., Fan, Z., Luo, J., Ke, Z.J., 2008. GSK3beta andendoplasmic reticulum stress mediate rotenone-induceddeath of SK-N-MC neuroblastoma cells. Biochem. Pharmacol.76, 128–138.

Cheung, Y.T., Lau, W.K., Yu, M.S., Lai, C.S., Yeung, S.C., So, K.F.,Chang, R.C., 2009. Effects of all-trans-retinoic acid on humanSH-SY5Y neuroblastoma as in vitro model in neurotoxicityresearch. Neurotoxicology 30, 127–135.

Chiribau, C.B., Gaccioli, F., Huang, C.C., Yuan, C.L., Hatzoglou, M.,

2010. Molecular symbiosis of CHOP and C/EBP beta isoform LIP

contributes to endoplasmic reticulum stress-induced

apoptosis. Mol. Cell. Biol. 30, 3722–3731.Colla, E., Coune, P., Liu, Y., Pletnikova, O., Troncoso, J.C., Iwatsubo, T.,

Schneider, B.L., Lee, M.K., 2012. Endoplasmic reticulum stress is

important for the manifestations of alpha-synucleinopathy

in vivo. J. Neurosci. 32, 3306–3320.Dever, T.E., 2002. Gene-specific regulation by general translation

factors. Cell 108, 545–556.Galehdar, Z., Swan, P., Fuerth, B., Callaghan, S.M., Park, D.S.,

Cregan, S.P., 2010. Neuronal apoptosis induced by

endoplasmic reticulum stress is regulated by ATF4–CHOP-

mediated induction of the Bcl-2 homology 3-only member

PUMA. J. Neurosci. 30, 16938–16948.Halterman, M.W., Gill, M., DeJesus, C., Ogihara, M., Schor, N.F.,

Federoff, H.J., 2010. The endoplasmic reticulum stress

response factor CHOP-10 protects against hypoxia-induced

neuronal death. J. Biol. Chem. 285, 21329–21340.Han, J., Back, S.H., Hur, J., Lin, Y.H., Gildersleeve, R., Shan, J., Yuan,

C.L., Krokowski, D., Wang, S., Hatzoglou, M., Kilberg, M.S.,

Sartor, M.A., Kaufman, R.J., 2013. ER-stress-induced

transcriptional regulation increases protein synthesis leading

to cell death. Nat. Cell Biol. 15, 481–490.Harding, H.P., Novoa, I., Zhang, Y., Zeng, H., Wek, R., Schapira, M.,

Ron, D., 2000. Regulated translation initiation controls stress-

induced gene expression in mammalian cells. Mol. Cell 6,

1099–1108.Hartley, A., Stone, J.M., Heron, C., Cooper, J.M., Schapira, A.H.,

1994. Complex I inhibitors induce dose-dependent apoptosis

in PC12 cells: relevance to Parkinson0s disease. J. Neurochem.

63, 1987–1990.Holtz, W.A., O0Malley, K.L., 2003. Parkinsonian mimetics induce

aspects of unfolded protein response in death of

dopaminergic neurons. J. Biol. Chem. 278, 19367–19377.Hoozemans, J.J., van Haastert, E.S., Eikelenboom, P., de Vos, R.A.,

Rozemuller, J.M., Scheper, W., 2007. Activation of the unfolded

protein response in Parkinson0s disease. Biochem. Biophys.

Res. Commun. 354, 707–711.Huang, C.C., Chiribau, C.B., Majumder, M., Chiang, C.M., Wek, R.C.,

Kelm, R.J., Khalili, K., Snider, M.D., Hatzoglou, M., 2009. A

bifunctional intronic element regulates the expression of the

arginine/lysine transporter Cat-1 via mechanisms involving the

purine-rich element binding protein A (Pur alpha). J. Biol. Chem.

284, 32312–32320.Huang, X., Chen, Y., Zhang, H., Ma, Q., Zhang, Y.W., Xu, H., 2012.

Salubrinal attenuates beta-amyloid-induced neuronal death

and microglial activation by inhibition of the NF-kappaB

pathway. Neurobiol. Aging 33 (1007.e9–17).Imai, Y., Soda, M., Takahashi, R., 2000. Parkin suppresses

unfolded protein stress-induced cell death through its E3

ubiquitin-protein ligase activity. J. Biol. Chem. 275,

35661–35664.Lange, P.S., Chavez, J.C., Pinto, J.T., Coppola, G., Sun, C.W.,

Townes, T.M., Geschwind, D.H., Ratan, R.R., 2008. ATF4 is an

oxidative stress-inducible, prodeath transcription factor in

neurons in vitro and in vivo. J. Exp. Med. 205, 1227–1242.Levy, O.A., Malagelada, C., Greene, L.A., 2009. Cell death pathways

in Parkinson0s disease: proximal triggers, distal effectors, and

final steps. Apoptosis 14, 478–500.Lewerenz, J., Sato, H., Albrecht, P., Henke, N., Noack, R., Methner, A.,

Maher, P., 2012. Mutation of ATF4 mediates resistance of

neuronal cell lines against oxidative stress by inducing xCT

expression. Cell Death Differ. 19, 847–858.Martin, I., Dawson, V.L., Dawson, T.M., 2011. Recent advances in

the genetics of Parkinson0s disease. Annu. Rev. Genomics

Hum. Genet. 12, 301–325.

b r a i n r e s e a r c h 1 5 4 9 ( 2 0 1 4 ) 5 2 – 6 262

Pan, T., Rawal, P., Wu, Y., Xie, W., Jankovic, J., Le, W, 2009.Rapamycin protects against rotenone-induced apoptosisthrough autophagy induction. Neuroscience 164, 541–551.

Rana, A., Rera, M., Walker, D.W., 2013. Parkin overexpressionduring aging reduces proteotoxicity, alters mitochondrialdynamics, and extends lifespan. Proc. Natl. Acad. Sci. USA110, 8638–8643.

Ron, D., Walter, P., 2007. Signal integration in the endoplasmicreticulum unfolded protein response. Nat. Rev. Mol. Cell Biol.8, 519–529.

Ryu, E.J., Harding, H.P., Angelastro, J.M., Vitolo, O.V., Ron, D.,Greene, L.A., 2002. Endoplasmic reticulum stress and theunfolded protein response in cellular models of Parkinson0sdisease. J. Neurosci. 22, 10690–10698.

Shin, J.H., Ko, H.S., Kang, H., Lee, Y., Lee, Y.I., Pletinkova, O.,Troconso, J.C., Dawson, V.L., Dawson, T.M., 2011. PARIS(ZNF746) repression of PGC-1alpha contributes toneurodegeneration in Parkinson0s disease. Cell 144, 689–702.

Silva, R.M., Ries, V., Oo, T.F., Yarygina, O., Jackson-Lewis, V., Ryu, E.J.,Lu, P.D., Marciniak, S.J., Ron, D., Przedborski, S., Kholodilov, N.,Greene, L.A., Burke, R.E., 2005. CHOP/GADD153 is a mediator ofapoptotic death in substantia nigra dopamine neurons in anin vivo neurotoxin model of Parkinsonism. J. Neurochem. 95,974–986.

Singleton, D.C., Harris, A.L., 2012. Targeting the ATF4 pathway incancer therapy. Expert Opin. Ther. Targets 16, 1189–1202.

Slodzinski, H., Moran, L.B., Michael, G.J., Wang, B., Novoselov, S.,Cheetham, M.E., Pearce, R.K., Graeber, M.B., 2009.Homocysteine-induced endoplasmic reticulum protein (herp)is up-regulated in Parkinsonian substantia nigra and presentin the core of Lewy bodies. Clin. Neuropathol. 28, 333–343.

Su, N., Kilberg, M.S., 2008. C/EBP homology protein (CHOP)interacts with activating transcription factor 4 (ATF4) and

negatively regulates the stress-dependent induction of the

asparagine synthetase gene. J. Biol. Chem. 283, 35106–35117.Sun, X., Liu, J., Crary, J.F., Malagelada, C., Sulzer, D., Greene, L.A.,

Levy, O.A., 2013. ATF4 protects against neuronal death in

cellular Parkinson0s disease models by maintaining levels of

parkin. J. Neurosci. 33, 2398–2407.Walter, P., Ron, D., 2011. The unfolded protein response: from

stress pathway to homeostatic regulation. Science 334,

1081–1086.Wu, L., Tian, Y.Y., Shi, J.P., Xie, W., Shi, J.Q., Lu, J., Zhang, Y.D., 2013.

Inhibition of endoplasmic reticulum stress is involved in the

neuroprotective effects of candesartan cilexitil in the

rotenone rat model of Parkinson0s disease. Neurosci. Lett. 548,

50–55.Xie, L., Tiong, C.X., Bian, J.S., 2012. Hydrogen sulfide protects SH-

SY5Y cells against 6-hydroxydopamine-induced endoplasmic

reticulum stress. Am. J. Physiol. Cell Physiol. 303, C81–C91.Xiong, N., Long, X., Xiong, J., Jia, M., Chen, C., Huang, J., Ghoorah, D.,

Kong, X., Lin, Z., Wang, T., 2012. Mitochondrial complex I

inhibitor rotenone-induced toxicity and its potential

mechanisms in Parkinson0s disease models. Crit. Rev. Toxicol.

42, 613–632.Yasuda, T., Hayakawa, H., Nihira, T., Ren, Y.R., Nakata, Y.,

Nagai, M., Hattori, N., Miyake, K., Takada, M., Shimada, T.,

Mizuno, Y., Mochizuki, H., 2011. Parkin-mediated protection of

dopaminergic neurons in a chronic MPTP-minipump mouse

model of Parkinson disease. J. Neuropathol. Exp. Neurol. 70,

686–697.Yong-Kee, C.J., Warre, R., Monnier, P.P., Lozano, A.M., Nash, J.E.,

2012. Evidence for synergism between cell death mechanisms

in a cellular model of neurodegeneration in Parkinson0sdisease. Neurotox Res. 22, 355–364.