Available online at www.sciencedirect.com

www.elsevier.com/locate/npep

Neuropeptides 42 (2008) 653–661

Neuropeptides

CART protects brain from damage through ERK activation inischemic stroke

Jia Jia a,b, Xuemei Chen a, Wenjing Zhu a, Yun Luo a, Zichun Hua b, Yun Xu a,b,c,*

a Department of Neurology, The Affiliated Drum Tower Hospital of Nanjing University Medical School, PR Chinab The State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University, PR China

c Jiangsu Key Laboratory for Molecular Medicine, PR China

Received 31 January 2008; accepted 29 May 2008Available online 21 July 2008

Abstract

Cocaine and amphetamine-regulated transcript (CART) is a neuropeptide that protects brains against ischemic injury in vivo andin vitro. By using small interference RNA against CART(CARTi), this study shows that CART knockdown by CARTi downreg-ulated exogenous and endogenous CART mRNA and protein expression in vivo and in vitro. Consequently, CART knockdownexacerbated neuronal cell death induced by oxygen and glucose deprivation (OGD). It also showed that CART knockdownincreased infarct size in a mouse middle cerebral artery occlusion model. CART’s protective effects are most likely mediated throughthe ERK 1/2 pathway, since ERK 1/2 phosphorylation, not that of p38 or JNK is activated in CART-treated neurons after OGD.Furthermore, neuroprotection of CART is abolished by CART knockdown and by pretreatment with ERK antagonist PD98059and U0126, but not with p38 or JNK antagonists SB203580 or SP600125. These results provide further evidence that CART isan endogenous neuroprotective peptide against cerebral ischemia and it does so through the MAPK/ERK signaling pathway.Therefore, CART may be developed into a therapeutic agent for stroke-related brain injury.� 2008 Elsevier Ltd. All rights reserved.

Keywords: CART; Stroke; siRNA; MAPK/ERK

1. Introduction

Cocaine and amphetamine-regulated transcript(CART) is a neuropeptide widely expressed in brainand neuroendocrine tissue. It has been implicated in avariety of brain functions including food intake, drugaddiction, the body’s response to stress, neuroprotectiveproperties and promoting the survival and differentia-tion of neurons in vitro. Research on CART peptides

0143-4179/$ - see front matter � 2008 Elsevier Ltd. All rights reserved.

doi:10.1016/j.npep.2008.05.006

* Corresponding author. Address: Department of Neurology, TheAffiliated Drum Tower Hospital of Nanjing University MedicalSchool, 321 Zhong Shan Road, Nanjing 210008, PR China. Tel.:+86 25 83105208.

E-mail address: xuyun20042001@yahoo.com.cn (Y. Xu).

has mostly focused on feeding behavior, reinforcementand reward, and meditating the locomotor effects of psy-chostimulants. Although evidence also exists in litera-ture to suggest that CART may exert neuroprotectiveproperties (Mao et al., 2007; Wu et al., 2006), researchin the field is limited. Previously studies showed thatexogenous CART protects primary cultured corticalneurons from cell death induced by oxygen–glucosedeprivation (OGD) and decreases infarct size after ische-mic stroke in the mouse middle cerebral artery occlusionmodel (MCAO). Furthermore, post-ischemic inductionof CART by estrogen contributed to estrogen-mediatedneuroprotection in ischemic stroke (Xu et al., 2006).

The present study, further confirmed the neuropro-tective role of endogenous CART in experimental strokeusing small interfering RNA (siRNA). It was also found

654 J. Jia et al. / Neuropeptides 42 (2008) 653–661

that neuroprotection by CART is in part mediated bythe activation of mitogen-activated protein kinase(MAPK)/extracellular signal-regulated kinase (ERK)after OGD, but not p38 or JNK activation.

2. Materials and methods

2.1. Plasmids

Full-length CART cDNA was amplified from TAcloning plasmid containing CART mRNA sequence(from our Lab, GenBank NM_013732) using primers:forward primer, 50-CGGAATTCTGATGGA-GAGCTCCCGCCT-30 and reverse primer, 50-CAG-

GGATCCCGCAAGCACTTCAAGAGGAA-30. Theamplicon was digested with EcoR I and BamH I andinserted into the two enzyme sites of pEGFPN1 (Clon-tech, USA). We used small RNA (siRNA) Target Fin-der software (from Genescript Corporation, http://www.genscript.com) to design siRNA and constructsmall hairpin RNA against CART (shRNA). shRNAswere synthesized and subsequently cloned into pRNA-U6.1/Neo vector using BamH I and Hind III (Gen-script, USA). Five CART siRNAs sequences (siA–siE)were designed to target the 390 bp mouse CART openreading frame – ORF (GenBank NM_013732). Theplasmid expressing siRNA against luciferase (pSiLuc)was used as a negative control. shRNA sequences wereas follows:

J. Jia et al. / Neuropeptides 42 (2008) 653–661 655

The oligonucleotides were synthesized by BiocolorBioScience & Technology Company (Shanghai, China),and control siRNA(pRNA-U6.1/Neo/siLuc) was pur-chased from Genscript Corporation.

2.2. Cell culture and transient transfection

Primary cortical neurons were prepared from E16-17mouse embryos. Cortices were dissected, treated withtrypsin, and plated at 4 � 105 cell/ml on poly-D-lysine-coated 24-well plates or 96-well plates. Cells were grownin Neurobasal media supplemented with B27 supple-ment (Invitrogen, Carlsbad, CA) and 25 nM glutamineat 37 �C in a humidified 5% CO2 incubator (Xu et al.,2006).

HEK293 cells were grown in 24-well plates inDMEM (Invitrogen) supplemented with 10% fetalbovine serum (Gibco), 2 mM glutamine, and 1% penicil-lin/streptomycin at 37 �C, as described previously (Xuet al., 2004a).

Cells were kept in medium without fetal bovine serumand antibiotics the day before transfection, and grownto reach 90–95% confluence at the time of transfection.293 Cells were co-transfected with pCART-EGFP orpEGFPN1 expression plasmids and plasmid containingshRNA duplexes (siA, siB, siC, siD or SiE), or siLucusing Lipofectamine 2000 (Invitrogen, CA, USA)according to the manufacturer’s instructions. Neurons

were only transfected with shRNA plasmids. At 48 hpost transfection, the cells were examined for EGFPexpression by fluorescence microscopy and total RNAwas extract from cells for RT-PCR assay according tostandard methods.

2.3. OGD (oxygen–glucose deprivation)

After transfection with CART siRNA plasmids for48 h, neurons were subjected to combined OGD for2 h to simulate ischemia in vitro (Xu et al., 2006). Briefly,cultures were switched from the normal feeding mediumto the oxygen-depleted, glucose-free medium. Cells wereincubated in a hypoxia chamber (Billups-Rothenberg,Del Mar, CA) previously flushed for 15 min with5%CO2/95%N2 at 2 psi (1 psi = 6.89 kPA). Valves wereclosed, and chambers were incubated at 37 �C for 2 h.At the end of OGD, cells were returned to the normalfeeding medium and incubated under normal conditionsat 37 �C for 24 h for later experiments.

2.4. RT-PCR

Total RNA of the cells was extracted with a Trizolcommercial kit (Invitrogen, USA). The cDNA was syn-thesized using a reverse transcriptase kit (Fermen-tas,USA) The oligonucleotide primer sequences wereas follows:

656 J. Jia et al. / Neuropeptides 42 (2008) 653–661

CART sense: 50-ATGGAGAGCTCCCGCCTG-30,anti-sense: 50-CAAGCACTTCAAGAGGAA-30;b-actin sense: 50-CTACGTCGCCCTGGACTTC-

GAGC-30,anti-sense: 50-GATGGAGCCGCCGATCCACAC-

GG-30;GAPDH sense: 50-AACGACCCCTTCATTGAC-30,anti-sense: 50-TCCACGACATACTCAGCAC-30.Reaction products were separated on an agarose gel

with ethidium-bromide. The bands were visualizedunder UV illumination and photographed. The intensityof bands was quantified with densitometry, and finalresults were normalized to b-actin levels in 293 cells orGAPDH levels in neurons.

2.5. CART peptides assayed by RIA

After neurons were transfected with CART siRNAplasmids and treated with OGD, the concentration ofCART in culture medium was measured with aCART55-102 RIA kit (RK-003-62) from Phoenix Phar-maceuticals according to the manufacturer’s protocol.

2.6. Propidium iodide (PI) and calcein AM staining

Neuronal cell death was assayed by staining cells withPI (6. 25 lg/ml; Sigma) in physiological salt solution for10 min at 37 �C. Viable cells were determined by0.25 ng/ml calcein AM (Molecular Probes, Eugene,OR) as described previously (Xu and Tao, 2004).Fluo-rescent neuronal images were captured with a digitalcamera on a microscope (ZEISS, Germany) with imagesoftware.

2.7. MTT assay

Following the treatment described above, neuronalviability was determined by MTT assay (Hansen et al.,1989). Briefly, Cell culture medium in 96-well plateswas aspirated and replaced with fresh neuron feedingmedium containing 0.5 mg/ml MTT. After 3 h incuba-tion at 37 �C, the supernatant was then discarded and100 ll DMSO was added. The optical density wasassessed at 570 nm using an ELISA plate reader(TECAN, Switzerland). Cell survival rates wereexpressed in percentage of the value of cells withoutany treatment (original medium group).

2.8. Western blot analysis

The cultured neurons were harvested according toprocedures described previously (Xu et al., 2004b). AnEqual amount of protein samples were separated bySDS–PAGE and blotted onto polyvinylidene fluoride(PVDF) membranes. Membranes were probed with pri-mary antibodies against phosphorated ERK1/2, p38,

JNK (Cell Signaling, 1:1000) or total ERK1/2, p38,JNK (Cell signaling, 1:1000). The proteins were detectedusing horseradish peroxidase-conjugated anti-rabbit oranti-mouse secondary antibodies and visualized usingchemiluminescence reagents provided with the ECL kit(Amersham Pharmacia Biotech, Piscataway, NJ) andexposure to film. The intensity of blots was quantifiedwith densitometry.

2.9. Middle cerebral artery occlusion (MCAO) model

and gene delivery in vivo

The animal study was conducted in accordance withNational Regulations of Experimental Animal Adminis-tration and all the animal experimental protocols per-formed on animals were approved by the Committeeof Experimental Animal Administration of NanjingUniversity. Mice were randomly divided into threegroups (n = 10): 1) animals receiving ICV injection of0.5 lg CART (55–102) peptide, immediately followedby 2 h MCAO model, 2) animals receiving ICV injectionof 15 ll M-PEI complexed psiCART (5 lg plasmid perinjection) 24 h before MCAO, 3) animals receivingICV injection of M-PEI comlexed psiLuc 24 h beforeMCAO. In vivo effectiveness of siRNA was performedon additional twelve mice. Six mice received ICV injec-tion of 15 ll M-PEI complexed psiCART, and six micereceived ICV injection of 15 ll M-PEI complexed psi-Luc. At 24 h post injection, cortices were dissected formeasurement of CART peptide by RIA. This M-PEImediated transfection method has been established inour lab (Dong et al., 2006; Li et al., 2007).

The animals were anesthetized with an intraperito-neal injection of Sodium Pentobarbital (1%) at a doseof 45 mg/kg. Body temperature was maintained at37 ± 0.5 �C by a heating lamp and heating pad. Middlecerebral artery occlusion (MCAO) was achieved by theintraluminal filament methods as previously described(Xu et al., 2006). Briefly, a 6–0 monofilament nylonsuture with heat-rounded tip was inserted through theexternal carotid artery and advanced into the internalcarotid artery to occlude the origin of the middlecerebral artery. After 2 h of occlusion, the filamentwas withdrawn to allow for reperfusion for 24 h, andthen the brains were quickly removed for TTCstaining.

2.10. Assessment of infarct size

Tissue infarction was identified by TTC staining inthick (2-mm) coronal sections. Slices were photographedand images were studied with image-analysis software(OSIRIS 4.19, Switzerland). Infarct volume in all sliceswas expressed as a percentage of the contralateralhemisphere.

siA siB siC siD siE siLuc N

CART

GAPDH

293T Vec siA siB siC siD siE siLuc

CART

β-actin

A

B

Fig. 1. CART mRNA suppression by siRNA. (A) CART siRNAswere co-transfected with pCART-EGFP into 293 cell, respectively. (B)siRNAs were transfected into Neurons. Vector and siLeu are negativecontrol. b-Actin or GAPDH is as loading control. siRNA A and E cansignificantly inhibit ectopic and endogenous CART, siRNA B and Cslightly inhibit CART RNA, while siRNA D had no effect on CARTlevel. N is as normal neurons.

J. Jia et al. / Neuropeptides 42 (2008) 653–661 657

2.11. Statistical analysis

Statistical tests to determine the differences betweengroups were performed with one-way ANOVA followedby a Newman–Keuls test using SPSS 11.5. Data wasassessed as mean ± SD. Significance was set at p < 0.05.

3. Results

3.1. CART mRNA Suppression by siRNA

To demonstrate the effectiveness of siRNA con-structs, plasmids expressing CART as well as CART

Fig. 2. CART Protein expression suppressed by siRNA. siRNA and CART-protein expression(middle lane). Left lane, no siRNA, right lane is negative

Fig. 3. Effect of CARTi on CART peptide release and neuronal cell death. Cendogenous CART peptide(A) and enhanced neuronal death(B). *p < 0.05 v

siRNAs were co-transfected into 293 cells, respectively.CART mRNA levels were measured by RT-PCR (Fig.1A). The result shows that siRNA A and E remarkablyinhibited ectopic CART, siRNA B and C, slightly inhib-ited CART mRNA, while siRNA D had no effect onCART. Vector and siLuc were used as negative controls.b-Actin was used as a loading control. We also testedthe interfering effect of CART shRNA constructs in pri-mary neurons. In agreement with the 293 cells’ data, theCART siRNA inhibited endogenous CART mRNAexpression in neurons (Fig. 1B). The experiments weredone in triplicate.

3.2. Suppression of ectopically expressed CART proteinby siRNA in 293 cells

To determine if CART siRNA suppresses CARTprotein expression, siRNA and CART-GFP plasmidwere co-transfected into 293 cells (Fig. 2). The resultindicated that CARTi reduced CART protein expres-sion. All experiments were performed in triplicate.

3.3. Effect of CARTi on CART peptide release and

neuronal cell death

To further demonstrate that CART is an endogenousneuroprotection peptide and protects neurons fromdeath after OGD, CARTi was transfected into neurons(Fig. 3). As expected, the results indicated that CART

GFP plasmid were co-transfected into 293 cells. siRNA reduces CARTcontrol.

ARTi was pretransfected into neurons before OGD. CARTi inhibitedersus vec-transfected neurons exposed to OGD.

CARTi Control CARTp

0

0.2

0.4

0.6 * p<0.05

<0.01

# p<0.01

CARTiCARTpControl

Vector RNAi0

0.2

0.4

0.6

0.8

1

1.2

Rel

ativ

e C

AR

T p

rote

in pA

B

Fig. 4. Effect of CART peptide and CARTi on infarct size after MCAO in mice. Mice were pretreated with CART peptide or CARTi before MCAO,(A) shows siE decreased CART peptides by 39% compared to the control (with vector, p < 0.01). Then, infarct size was measured at 24 h afterMCAO by TTC staining. CARTi increased infarct size by 34% compared to control group, and CARTp decreased by 68% compared to controlgroup. # p < 0.01 versus the control. *p < 0.05 versus the control (B).

Total p44/42

P-p44/42

10’ 30’ 60’ 60’Treatment Time

CART(nM) - - 0.4 0.4 0.4 2 PD PD+C

OGD - + + + + + + +

P-P44/42

0.4nM 2nM siLuc siA siE

CART peptide siCART

OGD - + + + + + +

A

B

658 J. Jia et al. / Neuropeptides 42 (2008) 653–661

siRNA inhibited endogenous CART peptide andenhanced neuronal death after OGD. Furthermore, lessCART peptide resulted in more neuronal cell death. siAdecreased CART peptides by 23% compared to the con-trol, while the percentage of neuronal death after OGDwas increased by 45%. siE or combined siA and siEreduced CART peptide by 37% and 48%, respectively,compared to the control, while the percentage of neuro-nal death after OGD was increased by 2-times the con-trol level (p < 0.01). Both siLuc and vector were used ascontrols. All assays were performed in three indepen-dent experiments.

P44/42

Fig. 5. CART Increases ERK Phosphorylation in Primary Neurons.(A) 0.4 nM CART active p44/42 ERK phosphorylation in a dose- andtime-dependent manner after OGD. Total ERK was used as a loadingcontrol. MAPK/ERK inhibitor PD98095 inhibits ERK activation, andso does CARTi (B). siLuc is as control.

3.4. Effect of CART peptide and CARTi on infarct size

after MCAO in mice

To confirm the effectiveness of RNAi in vivo, micecortex CART peptides were measured by RIA with orwithout siE (Fig. 4a).The results indicated that siEdecreased CART peptides by 39% compared to the con-trol (with vector, p < 0.01) and suggest siE can knockdown CART peptide in vivo (six mice from each groupwere used). To further investigate the role of CARTagainst ischemic stroke, we treated mice with CARTpeptide or CARTi, using interference vector as a con-trol. Mice were then subjected to MCAO, and infarctsize was measured at 24 h after reperfusion by TTCstaining (Fig. 4b). The result showed that CARTiincreased infarct size by 34% compared to the control,and CARTp decreased the infarct size by 68% comparedto the control, strongly suggesting that CART is an

endogenous neuroprotective peptide. Ten mice fromeach group were used.

3.5. CART increases ERK phosphorylation in primaryneurons

To study the mechanism of neuroprotection byCART, phosphorylated p44/42 ERK, phosphorylatedp38 and phosphorylated JNK were measured in neuronsby Western blot. The results showed no significant dif-ference in phosphorylated p38 and phosphorylatedJNK between the CART-treated group and the

0

0.2

0.4

0.6

0.8

1

Cel

l via

bil

ity

ll

P<0.001P<0.001P<0.05

P<0.01

CON U0126 PD98059 SB203580 SP600125

CON OGD OGD+CART

Fig. 6. Neuroprotection by CART through activating MAPK/ERKphosphorylation, not by MAPK/P38 or MAPK/JNK phosphoryla-tion. Neurons had been pretreated by CART, CART with MAPK/ERK inhibitor U0126 , PD98095, or P38 inhibitor SB203580, or JNKinhibitor SP600125, respectively, before OGD, neuroprotection ofCART was blocked by MAPK/ERK inhibitor U0126 and PD98095(p < 0.01), not by P38 inhibitor SB203580, or JNK inhibitor SP600125.

J. Jia et al. / Neuropeptides 42 (2008) 653–661 659

non-treated group (data not shown). Results suggestedthat the activation of the MAPK/ ERK pathway mayplay a role in neuroprotection of CART, in a dose andtime-dependent manner. Total ERK was used as a load-ing control (Fig. 5A and 5B). After treating neuronswith 0.4 nM CART peptide for 10, 30 and 60 min postOGD, the ERK phosphorylation was increased by72.84%, 28.9%, 44.6%, respectively, compared to thecontrol (5 A). ERK phosphorylation was inhibited byCARTi (5B), and also by MAPK/ERK inhibitorPD98059 (Fig. 5A). All experiments were performed atleast in triplicate.

3.6. MAPK/ERK inhibitor suppress neuroprotection of

CART in neurons

To further elucidate the mechanism of neuroprotec-tion by CART through the phosphorylated p44/42ERK pathway, not the phosphorylated p38 and phos-phorylated JNK pathway, four antagonists (p44/42ERK antagonist-PD98059 and U0126, p38 antagonist-SB 203580, JNK antagonist-SP600125) were appliedon neurons before CART treatment and OGD. Thenneuronal viability in primary cortical neuronal cultureswas determined by MTT. At concentrations of0.4 nM, CART enhanced neuronal viability afterOGD by 41.8% compared to the control neurons. Thepercentage of neuronal viability pretreated with 10 lMPD98059 and 10 lM U0126 was reduced by44.62 ± 0.0669%, 66.44 ± 0.0471%, respectively, com-pared to neurons treated with CART after OGD(87.6 ± 0.0427%, p < 0.001) (Fig. 6). In contrast, theneuronal viability of neurons treated with SB203580(10 lM) or SP600125 (10 lM) alone was not signifi-cantly changed compared to the control levels. Theresults suggest that neuroprotection by CART afterOGD is blocked by MAKP/ERK antagonists, not by

MAPK/p38 or MAPK/JNK antagonists. All experi-ments were performed in triplicate.

4. Discussion

The present study shows that: 1) siRNA can be usedagainst CART to effectively inhibit CART mRNA andprotein expression in neurons; 2) Inhibition of CARTexpression by siRNA exacerbates OGD-induced neuro-nal cell death and increases infarct size after MCAO; 3)Neuroprotection by CART against stroke is partiallymediated by the activation of the MAPK/ERKpathway.

CART is a neuropeptide that is specifically distrib-uted throughout the brain, gut and other parts of thebody. Since Douglass (Douglass et al., 1995) showedthat CART mRNA was increased in the nucleus accum-bens after psychostimulant administration, most studieshave focused on CART effects on feeding and obesity,drug abuse, and stress response. Recently the role ofthe CART peptide in neurotrophic action and neuro-protection in neurological disease has received muchattention. These studies and others indicate that CARTis expressed only in neurons, not in the other cell typesin the brain (Xu et al., 2006; Kuhar et al., 2000), colocal-ized with the dopamine D1 receptor in the ventral teg-mental area (Hubert and Kuhar, 2006) and adrenergicC1 neurons in the rostral ventrolateral medulla (Dunet al., 2002).These findings are compatible with the viewthat CART may act as neurotransmitters to modulatesome neuron activities (Hubert and Kuhar, 2005; Dunet al., 2002). CART peptides can be released from neu-rons into extracellular space including dense core vesi-cles of axon terminals (Smith et al., 1997, 1999; Dallet al., 2000; Xu et al., 2006). CART may also exhibit neu-rotrophic properties (Risold et al., 2006). Small amountsof rsCART 1-89(0.01 ng/ml) increase the survival ofmotor neurons and enhance the number of dopaminetransporters in primary cell cultures (Louis, 1996).CART might also affect development in mammals(Adams et al., 1999). A recent study shows that CARTpincreases the number of surviving neurons and main-tains the integrity in primary hippocampal neurons byupregulating the brain-derived neurotrophic factor(Wu et al., 2006). CART expression is increased in thebrain after stroke, which was enhanced by estrogenand contributes to estrogen’s neuroprotective property.Exogenous CARTp provides neuroprotection afterstroke in vivo and in vitro. (Xu et al., 2006).

To further investigate the role of CART in braindamage after stroke, endogenous CART was knockeddown by small interference RNA. CART siRNA inhib-ited endogenous CART peptide in primary neurons andenhanced OGD-induced neuronal death. The CARTpeptide level is inversely correlated with cell death rate.

660 J. Jia et al. / Neuropeptides 42 (2008) 653–661

In the in vivo model, as well as in vitro, CARTi increasedinfarct sizes after MCAO, while exogenous rat CARTpeptide (55–102) significantly decreases infarct size.These results support previous observations and indicatethat CART is an endogenous neuroprotective peptideagainst stroke.

So far, little is known about CART peptide-respon-sive cells, receptor, or intracellular signaling mechanismsby which CART peptides exert their effects (Stein et al.,2006; Vicentic et al., 2006). A growing number of exper-imental results suggest that CART neuroprotectionmight involve several mechanisms. CART inhibitsL-type voltage-gated Ca2+ channel activity via a G-pro-tein-dependent pathway in primary cell cultures of hip-pocampus (Yermolaieva et al., 2001). CART alsoincreases CREB phosphorylation in the ventral parvo-cellular subdivision of PVN (Sarkar et al., 2004), andCART up-regulation by estradiol is likely mediatedthrough CREB activation in neurons (Xu et al., 2006).CART increases c-Fos levels in a variety of neurons(Vrang et al., 2000). CART might enhance neuronal sur-vival after OGD through mitochondrial mechanism(Mao et al., 2007). These and other studies showed thatCART activates the MAPK /ERK pathway in AtT20cells and neurons (Lakatos et al., 2005; Xu et al., 2006).

In this study, rat CART 55–102 induced the activa-tion of ERK1/2 in a time- and dose-dependent mannerin neurons after OGD. The activation of ERK1/2 byCART was blocked by PD98059 (5 nM), a MEK1inhibitor. It was also found that there was slight activa-tion of ERK1/2 in OGD neurons, which was partly con-tributed by endogenous CART since CARTi candecrease the phosphorylated ERK1/2 (Fig. 5B). TheERK1/2 is serine–threonine kinases that play an impor-tant role in cell survival, proliferation, and differentia-tion and may be involved in early development andformation of the brain (Belcheva and Coscia, 2002;Werry et al., 2006). There are at least three welldescribed components of MAKP signaling cascadesincluding ERK1/2, p38 MAPK and SAPK/JNK. How-ever, the studies found that CART only activatedERK1/2, but not MAPK/p38 or JNK. The specificityof CART on the MAPK pathway was established byinvestigating CART’s neuroprotection on neuronal celldeath induced by OGD using a panel of MAPK antag-onists. We found that neuroprotection of CART wasblocked by ERK1/2 antagonists (U0126 andPD98059), but not by p38 and JNK antagonists(SB203580 and SP600125), suggesting that the neuro-protection by CART is mediated, at least partially, bythe MAPK/ERK1/2 signaling pathway.

In conclusion, CART is an endogenous neuroprotec-tant against ischemic neuronal cell death in vitro and in

vivo; The neuroprotective mechanism of CART is inpart linked to ERK activation. Therefore, CART may

serve as a therapeutic agent against stroke-related braindamage.

Acknowledgments

This work was supported by the National NatureScience Foundation (30470612, 30670739) of China,the Doctoral Program Foundation (20060284044) ofthe Ministry of Education of China, and the Interna-tional Cooperation Program and talented man pro-gram(BZ2006045, 06-B-002) of Jiangsu Province ofChina. We thank Dr. Wangsen Cao for helpful discus-sions, and Marilyn White for writing modification.

References

Adams, L.D., Gong, W., Vechia, S.D., Hunter, R.G., Kuhar, M.J.,1999. CART: from gene to function. Brain Res. 848, 137–140.

Belcheva, M.M., Coscia, C.J., 2002. Diversity of G protein-coupledreceptor signaling pathways to ERK/MAP kinase. Neurosignals11, 34–44.

Dall, V.S., Lambert, P.D., Couceyro, P.C., Kuhar, M.J., Smith, Y.,2000. CART peptide immunoreactivity in the hypothalamus andpituitary in monkeys: analysis of ultrastructural features andsynaptic connections in the paraventricular nucleus. J. Comp.Neurol. 416, 291–308.

Dong, W., Jin, G.H., Li, S.F., Sun, Q.M., Ma, D.Y., Hua, Z.C., 2006.Cross-linked polyethylenimine as potential DNA vector for genedelivery with high efficiency and low cytotoxicity. Acta Biochim.Biophys. Sin. (Shanghai) 38, 780–787.

Douglass, J., McKinzie, A.A., Couceyro, P., 1995. PCR differentialdisplay identifies a rat brain mRNA that is transcriptionallyregulated by cocaine and amphetamine. J. Neurosci. 15, 2471–2481.

Dun, S.L., Ng, Y.K., Brailoiu, G.C., Ling, E.A., Dun, N.J., 2002.Cocaine- and amphetamine-regulated transcript peptide-immuno-reactivity in adrenergic C1 neurons projecting to the intermedio-lateral cell column of the rat. J. Chem. Neuroanat. 23, 123–132.

Hansen, M.B., Nielsen, S.E., Berg, K., 1989. Re-examination andfurther development of a precise and rapid dye method formeasuring cell growth/cell kill. J. Immunol. Methods 119, 203–210.

Hubert, G.W., Kuhar, M.J., 2005. Colocalization of CART withsubstance P but not enkephalin in the rat nucleus accumbens. BrainRes. 1050, 8–14.

Hubert, G.W., Kuhar, M.J., 2006. Colocalization of CART peptidewith prodynorphin and dopamine D1 receptors in the rat nucleusaccumbens. Neuropeptides 40, 409–415.

Kuhar, M.J., Adams, L.D., Hunter, R.G., Vechia, S.D., Smith, Y.,2000. CART peptides. Regul. Peptides 89, 1–6.

Lakatos, A., Prinster, S., Vicentic, A., Hall, R.A., Kuhar, M.J., 2005.Cocaine- and amphetamine-regulated transcript (CART) peptideactivates the extracellular signal-regulated kinase (ERK) pathwayin AtT20 cells via putative G-protein coupled receptors. Neurosci.Lett. 384, 198–202.

Li, S., Dong, W., Zong, Y., Yin, W., Jin, G., Hu, Q., Huang, X., Jiang,W., Hua, Z.C., 2007. Polyethylenimine-complexed plasmid parti-cles targeting focal adhesion kinase function as melanoma tumortherapeutics. Mol. Ther. 15, 515–523.

Louis JCM, 1996. Methods of preventing neuron degeneration andpromoting neuron regeneration. Amgen, International patentapplication, publication number WO96/34619.

J. Jia et al. / Neuropeptides 42 (2008) 653–661 661

Mao, P., Ardeshiri, A., Jacks, R., Yang, S., Hurn, P.D., Alkayed, N.J.,2007. Mitochondrial mechanism of neuroprotection by CART.Eur. J. Neurosci. 26, 624–632.

Risold, P.Y., Bernard-Franchi, G., Collard, C., Jacquemard, C., La,R.A., Griffond, B., 2006. Ontogenetic expression of CART-peptides in the central nervous system and the periphery: a possibleneurotrophic role. Peptides 27, 1938–1941.

Sarkar, S., Wittmann, G., Fekete, C., Lechan, R.M., 2004. Centraladministration of cocaine- and amphetamine-regulated transcriptincreases phosphorylation of cAMP response element bindingprotein in corticotropin-releasing hormone-producing neurons butnot in prothyrotropin-releasing hormone-producing neurons inthe hypothalamic paraventricular nucleus. Brain Res. 999, 181–192.

Smith, Y., Kieval, J., Couceyro, P.R., Kuhar, M.J., 1999. CARTpeptide-immunoreactive neurones in the nucleus accumbens inmonkeys: ultrastructural analysis, colocalization studies, andsynaptic interactions with dopaminergic afferents. J. Comp. Neurol.407, 491–511.

Smith, Y., Koylu, E.O., Couceyro, P., Kuhar, M.J., 1997. Ultrastruc-tural localization of CART (cocaine- and amphetamine-regulatedtranscript) peptides in the nucleus accumbens of monkeys. Synapse27, 90–94.

Stein, J., Steiner, D.F., Dey, A., 2006. Processing of cocaine- andamphetamine-regulated transcript (CART) precursor proteins byprohormone convertases (PCs) and its implications. Peptides 27,1919–1925.

Vicentic, A., Lakatos, A., Jones, D., 2006. The CART recep-tors: background and recent advances. Peptides 27, 1934–1937.

Vrang, N., Larsen, P.J., Kristensen, P., Tang-Christensen, M., 2000.Central administration of cocaine-amphetamine-regulated tran-script activates hypothalamic neuroendocrine neurons in the rat.Endocrinology 141, 794–801.

Werry, T.D., Christopoulos, A., Sexton, P.M., 2006. Mechanisms ofERK1/2 regulation by seven-transmembrane-domain receptors.Curr. Pharm. Des. 12, 1683–1702.

Wu, B., Hu, S., Yang, M., Pan, H., Zhu, S., 2006. CART peptidepromotes the survival of hippocampal neurons by upregulatingbrain-derived neurotrophic factor. Biochem. Biophys. Res. Com-mun. 347, 656–661.

Xu, Y., Tao, Y.X., 2004. Involvement of the NMDA receptor/nitricoxide signal pathway in platelet-activating factor-induced neuro-toxicity. Neuroreport 15, 263–266.

Xu, Y., Traystman, R.J., Hurn, P.D., Wang, M.M., 2004a. Membranerestraint of estrogen receptor alpha enhances estrogen-dependentnuclear localization and genomic function. Mol. Endocrinol. 18,86–96.

Xu, Y., Zhang, B., Hua, Z., Johns, R.A., Bredt, D.S., Tao, Y.X.,2004b. Targeted disruption of PSD-93 gene reduces platelet-activating factor-induced neurotoxicity in cultured cortical neu-rons. Exp. Neurol. 189, 16–24.

Xu, Y., Zhang, W., Klaus, J., Young, J., Koerner, I., Sheldahl, L.C.,Hurn, P.D., Martinez-Murillo, F., Alkayed, N.J., 2006. Role ofcocaine- and amphetamine-regulated transcript in estradiol-medi-ated neuroprotection. Proc. Natl. Acad. Sci.USA 103, 14489–14494.

Yermolaieva, O., Chen, J., Couceyro, P.R., Hoshi, T., 2001. Cocaine-and amphetamine-regulated transcript peptide modulation ofvoltage-gated Ca2+ signaling in hippocampal neurons. J. Neurosci.21, 7474–7480.