Food &Function

PAPER

Publ

ishe

d on

21

Febr

uary

201

4. D

ownl

oade

d by

Tuf

ts U

nive

rsity

on

27/1

0/20

14 2

1:19

:36.

View Article OnlineView Journal | View Issue

aDepartment of Neurosurgery, China Med

TaiwanbDepartment of Respiratory Therapy, China McDepartment of Health and Nutrition Biotec

TaiwandDepartment of Nutrition, China Medical U

Taichung City, Taiwan. E-mail: mcyin@m

Tel: +886-4-22053366 ext. 7510

Cite this: Food Funct., 2014, 5, 1125

Received 19th December 2013Accepted 19th February 2014

DOI: 10.1039/c3fo60689a

www.rsc.org/foodfunction

This journal is © The Royal Society of C

s-Methyl cysteine enhanced survival of nervegrowth factor differentiated PC12 cells underhypoxic conditions

Chun-lin Liu,a Te-chun Hsiab and Mei-chin Yin*cd

A nerve growth factor-differentiated PC12 cell line was used to investigate the protective effects of

s-methyl cysteine (SMC) at 1, 2, 4, and 8 mM under oxygen–glucose deprivation (OGD) conditions. OGD

decreased the cell viability. However, SMC pre-treatments at 2, 4 and 8 mM improved the cell viability,

decreased cleaved caspase-3 and Bax expression, and reserved Bcl-2 expression. Furthermore, SMC

maintained the mitochondrial membrane potential, lowered the intracellular Ca2+ concentration and

DNA fragmentation, and decreased the activity and expression of caspase-3 and caspase-8. OGD

increased the reactive oxygen species (ROS) and 3-nitrotyrosine production, decreased glutathione

peroxide (GPX) and glutathione reductase (GR) activities and the expression, enhanced nitric oxide

synthase (NOS) activity and inducible NOS (iNOS) expression. SMC pre-treatments at 2, 4 and 8 mM

lowered the ROS and 3-nitrotyrosine formation, maintained GPX and GR activities and expression, and

decreased NOS activity and iNOS expression. OGD up-regulated hypoxia-inducible factor (HIF)-1a,

nuclear transcription factor kappa (NF-k) B p50, NF-kB p65 and p-p38 expression. SMC pre-treatments

at 1–8 mM lowered HIF-1a expression and decreased p38 phosphorylation. SMC at 2, 4 and 8 mM

suppressed the protein expression of NF-kB p50 and NF-kB p65. When YC-1 (HIF-1a inhibitor),

pyrrolidine dithiocarbamate (NF-kB inhibitor) or SB203580 (p38MAPK inhibitor) were used to block

the activation of HIF-1a, NF-kB and p38, SMC pre-treatments did not affect the protein expression of

HIF-1a, NF-kB and p-p38. These results indicated that SMC was a potent neuro-protective agent.

Introduction

The obstruction of blood ow to the brain leads to irreversibledamage because insufficient blood supply results in oxygen–glucose deprivation (OGD) and causes neuronal apoptosis.1 Ithas been documented that OGD evokes excessive production ofreactive oxygen species (ROS) and reactive nitrogen species(RNS) in neuronal cells, which enhances oxidative stress andinitiates apoptotic insult, such as activating a caspasecascade.2,3 Furthermore, hypoxia or OGD induces mitochon-drial depolarization and dysfunction, increases intracellularCa2+ concentration and activates nuclear transcription factorkappa (NF-k) B and mitogen-activated protein kinase (MAPK)signaling pathways in neuronal cells.4–6 Under low oxygenconditions, hypoxia-inducible factor (HIF)-1a is overexpressed

ical University Hospital, Taichung City,

edical University, Taichung City, Taiwan

hnology, Asia University, Taichung City,

niversity, 16th Floor, 91, Hsueh-shih Rd,

ail.cmu.edu.tw; Fax: +886-4-22062891;

hemistry 2014

and regulates the transcription of several genes associated withcell survival or death.7 It has been reported that HIF-1a exhibitsneuroprotective or neurotoxic effects depending on the type ofcellular stress.8,9 Thus, any agent with the ability to attenuateoxidative injury, regulate HIF-1a, MAPK or NF-kB under OGDconditions may benet neural cell survival.

s-Methyl cysteine (SMC) is a hydrophilic cysteine-containingcompound naturally formed in Allium plants such as garlic andonion.10 Our previous study reported that the pre-intake of thisagent retarded glutathione and dopamine depletion, main-tained glutathione peroxidase activity, and decreased inam-matory cytokines in striatum of Parkinson's-like mice.11 Thestudy by Wassef et al.12 found that dietary SMC supplementa-tion could prevent Parkinson's-like syndromes in Drosophila viaits anti-oxidative effects. Ishiwata et al.13 reported that SMCcould increase the extracellular level of D-serine, a co-agonist ofN-methyl D-aspartate receptor (NMDAR), in the frontal cortex ofrats. D-Serine participates in regulating NMDAR-mediatedsynaptic transmission.14 These previous studies suggest thatSMC is a potent protective agent for the brain and neuralsystem. However, it remains unknown whether SMC couldenhance neural cell survival under hypoxic conditions. It is alsounclear whether this agent could attenuate oxidative stress,

Food Funct., 2014, 5, 1125–1133 | 1125

Food & Function Paper

Publ

ishe

d on

21

Febr

uary

201

4. D

ownl

oade

d by

Tuf

ts U

nive

rsity

on

27/1

0/20

14 2

1:19

:36.

View Article Online

stabilize mitochondrial membrane or regulate NF-kB andMAPKpathways for neural cells against OGD-induced damage.

The nerve growth factor (NGF)-differentiated PC12 cell linehas been widely used as an in vitro ischemic model to investi-gate the impact of OGD upon cell apoptosis.15 It is alsocommonly used to examine neural protective effects and theaction modes of certain compounds.16 The present study usedthis cell line to investigate the protection of SMC under OGDconditions. The effects of this agent at various doses upon cellviability, calcium release, mitochondrial membrane potentialand oxidative stress were examined. Furthermore, the regula-tion of this agent upon HIF-1a, NF-kB and MAPK was alsoevaluated in order to understand the possible action modes.

Materials and methodsChemicals

The medium, plates, antibiotics and chemicals used for the cellculture were purchased from Difco Laboratory (Detroit, MI,USA). NGF was purchased from Promega Co. (Madison, WI,USA). Fura-2 acetoxymethyl ester (Fura-2AM), Rhodamine 123(Rh123) and SMC were purchased from Sigma Chemical Co. (StLouis, MO, USA). YC-1 (HIF-1a inhibitor), pyrrolidine dithio-carbamate (PDTC, NF-kB inhibitor) and SB203580 (p38MAPKinhibitor) were purchased from Cell Signaling Technology(Boston, MA, USA). All the chemicals used in these measure-ments were of the highest purity commercially available.

Cell culture

PC12 cells were cultured in a 35 mm dish containing Dulbecco'smodied Eagle's medium (DMEM) supplemented with 10%heat-inactivated calf serum and 5% fetal bovine serum under95% air–5% CO2 at 37 �C. The PC12 cells were treated with NGF(50 ng ml�1) and allowed to differentiate for 5 days. The culturemedium was changed every three days and the cells were sub-cultured once a week. The medium was changed to a serum-deprived medium and the cells were washed with serum-freeDMEM 24 h before the experiments and replanted in 96 wellplates.

OGD model

Our preliminary data showed that 2, 3 or 4 h incubation resultedin 42, 71 and 98% incorporation of SMC into the cells and lowerSMC incorporation led to less protective effects. Thus, 4 hincubation was used for the present study. The PC12 cells (105

cells per ml) were treated with DMEM containing SMC (0, 1, 2, 4or 8 mM) for 4 h. Aer washing the cells twice with glucose-freeDMEM, the cells were incubated in this glucose-free DMEM inan oxygen-free incubator (95% N2 and 5% CO2) for 2 h. Then,the cells were returned to the normal culture medium andincubated under normal growth conditions for an additional 24h. PC12 cells without SMC treatment and incubated undernormal growth conditions were used as control groups. In orderto clarify the role of SMC on HIF-1a, NF-kB and p38, the cellswere pretreated with a 10 mM inhibitor for 60 min beforeexposure to OGD.

1126 | Food Funct., 2014, 5, 1125–1133

3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide(MTT) assay

An MTT assay was performed to examine the cell viability.Briey, the PC12 cells (105 cells per ml) were incubated with 0.25mg MTT per ml for 3 h at 37 �C. The amount of MTT formazanproduct was determined by measuring the absorbance at 570nm (630 nm as a reference) using a microplate reader (Bio-Rad,Hercules, CA, USA). The cell viability was expressed as a percentof the control groups.

Lactate dehydrogenase (LDH) assay

The plasma membrane damage of the PC12 cells (105 cells perml) was evaluated by measuring the amount of intracellularLDH in the medium. Fiy ml of the culture supernatants werecollected from each well. The LDH activity (U l�1) was deter-mined by a colorimetric LDH assay kit (Sigma Chemical Co., StLouis, MO, USA).

Measurement of themitochondrial membrane potential (MMP)

The MMP was monitored using the uorescent dye Rh123. Aerincubation with SMC under OGD conditions, the PC12 cells (105

cells per ml) were centrifuged at 1200 � g for 5 min andresuspended in DMEM. Rh123 (100 mg l�1) was added to thePC12 cells for 45 min at 37 �C. The cells were collected andwashed twice with PBS. The mean uorescence intensity (MFI)in the cells was analyzed by a ow cytometry (Beckman-FC500,Beckman Coulter, Fullerton, CA, USA).

Determination of the intracellular Ca2+ concentration

The intracellular Ca2+ concentration was determined by usingFura-2AM, a Ca2+-sensitive dye, to measure the uorescenceintensity according to the method of Lenart et al.17 Briey, thecells (105 cells per ml) were loaded with Fura-2AM (nalconcentration 5 mmol l�1), 0.1% DMSO and 1% BSA for 30 minat room temperature in dark conditions, which was followed byincubating at 37 �C for 30 min. The uorescence was deter-mined by a spectrouorimeter (Shimadzu, Model RF-5000,Kyoto, Japan) with excitation at 340 and 380 nm, and theemission at 510 nm. The calcium concentration was obtainedby converting the uorescence ratio according to the equation:[Ca2+] (nM) ¼ Kd � [(R � Rmin)/(Rmax � R)] � FD/FS, in which Kd

was 224 nM, R was 340 : 380 ratio, Rmax was determined bytreating the cells with triton X-100, Rmin was determined bytreating the cells with ethylene glycol tetraacetic acid, FD wasthe uorescence of the Ca2+-free form and FS was the uores-cence of the Ca2+-bound form at excitation wavelengths of 380and 340 nm, respectively.

Measurement of DNA fragmentation

A cell death detection ELISA plus kit (Roche MolecularBiochemicals, Mannheim, Germany) was used to quantify DNAfragmentation. The PC12 cells (105 cells per ml) were lysed in 50ml of a cold lysis buffer for 30 min at room temperature fol-lowed by centrifugation at 200 � g for 10 min. Then, 20 ml of thesupernatant was transferred onto the streptavidin-coated plate,

This journal is © The Royal Society of Chemistry 2014

Paper Food & Function

Publ

ishe

d on

21

Febr

uary

201

4. D

ownl

oade

d by

Tuf

ts U

nive

rsity

on

27/1

0/20

14 2

1:19

:36.

View Article Online

and 80 ml a freshly prepared immunoreagent was added to eachwell and incubated for 2 h at room temperature. Aer washingwith PBS, the substrate solution was added and incubated for 15min. The absorbance at 405 nm (reference wavelength 490 nm)was measured using a microplate reader. The DNA fragmenta-tion was expressed as the enrichment factor using the followingequation: enrichment factor ¼ (absorbance of the sample)/(absorbance of the control).

Measurement of caspases activity

The activity of caspase-3 and -8 was detected using uoro-metric assay kits (Upstate, Lake Placid, NY, USA) according tothe manufacturer's protocol. The intra-assay CV was 3.3–4.2%,and the inter-assay CV was 5.4–6.5%. In brief, control ortreated cells (105 cells per ml) were lysed and incubated in icefor 10 min. Fiy ml of the cell lysate was mixed with 50 ml of areaction buffer and 5 ml of uorogenic substrates specic forcaspase-3 or -8 in a 96-well microplate. Aer incubation at37 �C for 1 h, the uorescent activity was measured using auorophotometer with excitation at 400 nm and emission at505 nm. The data were expressed as a percentage of the controlgroups.

ROS and DNA oxidation assay

The cells (105 cells per ml) were washed and homogenized. Thedye DCFH2-DA was used to measure the ROS level (nmol mg�1

protein) according to the method of Fu et al.18 Aer incubatingwith 50 mmol l�1 of the dye for 30 min and washing with PBS,the cell suspension was centrifuged at 412 � g for 10 min.Then, the medium was removed and the cells were dissolvedwith 1% Triton X-100. Fluorescence changes were measured atan excitation wavelength of 485 nm and an emission wave-length of 530 nm using a uorescence microplate reader. DNAfractions were obtained using a DNA Extractor WB kit (WakoPure Chemical Industries Ltd., Tokyo, Japan), and oxidativedamage was determined using an ELISA kit (OXIS HealthProducts Inc, Portland, OR, USA) for 8-OHdG (ng mg�1

protein).

Fig. 1 Effects of SMC upon cell viability determined by an MTT assayand plasma membrane damage determined by an LDH assay. NGFdifferentiated-PC12 cells were pretreated with SMC at 0, 1, 2, 4 or 8 mMfor 4 h and followed by incubation under normal or OGD conditions.The control groups were cells containing no SMC and incubated undernormal conditions. The data are mean � SD (n ¼ 10). (a–e) Meansamong bars without a common letter differ, p < 0.05.

Analyses for glutathione (GSH), oxidized glutathione (GSSG)and activity of glutathione peroxidase (GPX), glutathionereductase (GR)

The cells (105 cells per ml) were washed twice with PBS, thenwere scraped from the plates followed by homogenizing in20 mM PBS containing 0.5 mM butylated hydroxytoluene toprevent further oxidation. The homogenate was centrifugedat 3000 � g for 20 min at 4 �C, and the supernatant wasused for these assays according to the manufacturer'sinstructions. GSH and GSSG concentrations (ng mg�1

protein) were determined by commercial colorimetric GSHand GSSG assay kits (OxisResearch, Portland, OR, USA). Theactivity (U mg�1 protein) of GPX and GR in the PC12 cellswas determined using assay kits (EMD Biosciences, SanDiego, CA, USA).

This journal is © The Royal Society of Chemistry 2014

Nitrite assay, 3-nitrotyrosine level and nitric oxide synthase(NOS) activity

The production of nitric oxide was determined by measuring theformation of nitrite. Briey, 100 ml of the supernatant was treatedwith nitrate reductase, NADPH and FAD, followed by incubatingfor 1 h at 37 �C in the dark. Aer centrifuging at 6000 � g, thesupernatant was mixed with Griess reagent for color develop-ment. The absorbance at 540 nm was measured and comparedwith a sodium nitrite standard curve. The 3-nitrotyrosine level(nmol mg�1 protein) was measured by a commercial assay kitNorthwest Life Science Specialties (Vancouver, WA, USA). Themethod described by Sutherland et al.19 was used to measure thetotal NOS activity (pmol min�1 mg�1 protein) by incubating 30 mlof the homogenate with 10 mM NADP, 10 mM L-valine, 3000 Uml�1 calmodulin, 5mM tetrahydrobiopterin, 10mMCaCl2, and amixture of 100 mM L-arginine containing L-[3H]arginine.

Preparation of the cytosolic and nuclear fractions

The cells (105 cells per ml) were suspended in ice-cold buffer Acontaining 10 mM HEPES, pH 7.9, 1.5 mM MgCl2, 10 mM KCl,0.5 mM DTT, 0.5 mM phenylmethylsulfonyl uoride (PMSF), 1mg ml�1 of leupeptin, and 1 mg m�1 of aprotinin for 15 min.Cytosolic fractions were collected aer centrifugation at 14 000� g for 1 min at 4 �C. The remaining nuclear pellets wereresuspended in buffer B containing 20 mM HEPES, pH 7.9, 1.5mMMgCl2, 450 mM NaCl, 25% glycerol, 0.2 mM EDTA, 0.5 mM

Food Funct., 2014, 5, 1125–1133 | 1127

Fig. 2 Effects of SMC upon Bcl-2, Bax and cleaved caspase-3expression determined by western blot analyses. NGF differentiated-PC12 cells were pretreated with SMC at 0, 1, 2, 4 or 8 mM for 4 h fol-lowed by incubation under OGD conditions. The control groups werecells containing no SMC and incubated under normal conditions. Thedata are mean � SD (n ¼ 10). (a–e) Means among bars without acommon letter differ, p < 0.05.

Fig. 3 Effects of SMC upon caspase-3 and caspase-8 activity (a) andprotein expression (b). NGF-differentiated PC12 cells were pretreated

Food & Function Paper

Publ

ishe

d on

21

Febr

uary

201

4. D

ownl

oade

d by

Tuf

ts U

nive

rsity

on

27/1

0/20

14 2

1:19

:36.

View Article Online

DTT, 0.5 mM PMSF, 1 mg ml�1 of leupeptin and 1 mg ml�1 ofaprotinin for 30 min. The nal nuclear fractions were collectedaer centrifugation at 14 000 � g for 15 min at 4 �C.

with SMC at 0, 1, 2, 4 or 8 mM for 4 h followed by incubation underOGD conditions. The control groups were cells containing no SMCand incubated under normal conditions. The data are mean � SD (n ¼10). (a–f) Means among bars without a common letter differ, p < 0.05.

Western blot analysis

A sample at 40 mg protein was applied to 10% SDS-poly-acrylamide gel electrophoresis, and transferred to a

Table 1 Effects of SMC upon MMP determined as MFI, Ca2+ releaseand DNA fragmentation determined as the enrichment factor. NGF-differentiated PC12 cells were pretreated with SMC at 0, 1, 2, 4 or 8 mMfor 4 h followed by incubation under OGD conditions. The controlgroups were cells containing no SMC and incubated under normalconditions. The data are mean � SD (n ¼ 10)

MFI [Ca2+], nM Enrichment factor

Control 100e 484 � 63a 1.00a

OGD 52 � 2a 1877 � 154e 2.23 � 0.15e

SMC 1 + OGD 59 � 3b 1652 � 100d 1.97 � 0.10d

SMC 2 + OGD 71 � 4c 1273 � 122c 1.70 � 0.08c

SMC 4 + OGD 83 � 3d 925 � 89b 1.43 � 0.12b

SMC 8 + OGD 85 � 4d 818 � 57b 1.34 � 0.06b

a Means in a column without a common letter differ, p < 0.05. b Meansin a column without a common letter differ, p < 0.05. c Means in acolumn without a common letter differ, p < 0.05. d Means in a columnwithout a common letter differ, p < 0.05. e Means in a column withouta common letter differ, p < 0.05.

1128 | Food Funct., 2014, 5, 1125–1133

nitrocellulose membrane (Millipore, Bedford, MA, USA) for 1h. Aer blocking with a solution containing 5% nonfat milkfor 1 h to prevent non-specic binding of antibodies, themembrane was incubated with anti-caspase-3, anti-caspase-8,anti-cleaved caspase-3 (1 : 1000), anti-Bcl-2 (1 : 2000), anti-Bax(1 : 1000), anti-GPX, anti-GR, anti-iNOS (1 : 500), anti-HIF-1a,anti-NF-kB p65, anti-NF-kB p50 (1 : 1000), anti-p38, anti-p-p38, anti-JNK, anti-p-JNK, anti-ERK1/2 and anti-p-ERK1/2(1 : 2000) monoclonal antibody (Boehringer-Mannheim,Indianapolis, IN, USA) at 4 �C overnight, followed by reactingwith horseradish peroxidase-conjugated antibody for 3.5 h atroom temperature. The detected bands were quantied by animage analyzer (ATTO, Tokyo, Japan) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as a loadingcontrol. The blot was quantied by densitometric analysis.The results were normalized to GAPDH, and given as arbitraryunits (AU).

This journal is © The Royal Society of Chemistry 2014

Table 2 Effects of SMC upon the level of ROS (nmol mg�1 protein), GSSG (ng mg�1 protein), GSH (ng mg�1 protein), 8-OHdG (ng mg�1 protein),and the activity of GPX and GR (U mg�1 protein). NGF-differentiated PC12 cells were pretreated with SMC at 0, 1, 2, 4 or 8 mM for 4 h followed byincubation under OGD conditions. The control groups were cells containing no SMC and incubated under normal conditions. The data are mean� SD (n ¼ 10)

ROS GSSG GSH 8-OHdG GPX GR

Control 0.19 � 0.04a 0.88 � 0.09a 93 � 5f 0.51 � 0.11a 69.1 � 1.9e 65.2 � 2.1e

OGD 1.38 � 0.21e 2.40 � 0.19f 41 � 2a 2.02 � 0.18e 38.5 � 0.8a 34.9 � 0.7a

SMC 1 + OGD 1.13 � 0.15d 2.05 � 0.12e 48 � 3b 1.95 � 0.07e 40.2 � 1.3a 36.0 � 0.9a

SMC 2 + OGD 1.03 � 0.07d 1.79 � 0.08d 59 � 4c 1.62 � 0.12d 48.3 � 1.2b 40.6 � 1.1b

SMC 4 + OGD 0.75 � 0.10c 1.51 � 0.13c 70 � 4d 1.28 � 0.09c 56.4 � 1.5c 47.5 � 1.4c

SMC 8 + OGD 0.46 � 0.08b 1.19 � 0.10b 82 � 3e 0.97 � 0.06b 58.7 � 1.2d 54.8 � 1.3d

a Means in a column without a common letter differ, p < 0.05. b Means in a column without a common letter differ, p < 0.05. c Means in a columnwithout a common letter differ, p < 0.05. d Means in a column without a common letter differ, p < 0.05. e Means in a column without a commonletter differ, p < 0.05. f Means in a column without a common letter differ, p < 0.05.

Table 3 Effects of SMC upon the NO level (mmol mg�1 protein), 3-nitrotyrosine level (nmol mg�1 protein) and NOS activity (pmol min�1

mg�1 protein). NGF-differentiated PC12 cells were pretreated withSMC at 0, 1, 2, 4 or 8 mM for 4 h followed by incubation under OGDconditions. The control groups were cells containing no SMC andincubated under normal conditions. The data are mean � SD (n ¼ 10)

NO 3-Nitrotyrosine NOS

Control 3.41 � 0.28a 0.11 � 0.03a 2.11 � 0.19a

OGD 24.05 � 1.59e 0.86 � 0.08e 11.58 � 0.87e

SMC 1 + OGD 22.90 � 1.07e 0.80 � 0.05e 10.84 � 0.96e

SMC 2 + OGD 19.37 � 0.55d 0.65 � 0.04d 8.82 � 0.45d

SMC 4 + OGD 15.08 � 0.61c 0.50 � 0.06c 6.90 � 0.32c

SMC 8 + OGD 11.85 � 0.43b 0.32 � 0.05b 5.14 � 0.26b

a Means in a column without a common letter differ, p < 0.05. b Meansin a column without a common letter differ, p < 0.05. c Means in acolumn without a common letter differ, p < 0.05. d Means in a columnwithout a common letter differ, p < 0.05. e Means in a column withouta common letter differ, p < 0.05.

Fig. 4 Effects of SMC upon GPX, GR and iNOS expression deter-

Paper Food & Function

Publ

ishe

d on

21

Febr

uary

201

4. D

ownl

oade

d by

Tuf

ts U

nive

rsity

on

27/1

0/20

14 2

1:19

:36.

View Article Online

Statistical analysis

The effect of each treatment was analyzed from ten differentpreparations (n ¼ 10). The data were reported as means �standard deviation (SD), and subjected to analysis of variance.The differences among the means were determined by the leastsignicance difference test with signicance dened at p < 0.05.

mined by western blot analyses. NGF-differentiated PC12 cells werepretreated with SMC at 0, 1, 2, 4 or 8 mM for 4 h and followed byincubation under OGD conditions. The control groups were cellscontaining no SMC and incubated under normal conditions. Thedata are mean � SD (n ¼ 10). (a–e) Means among bars without acommon letter differ, p < 0.05.

Results

Under normal incubation conditions, SMC did not affect theviability and LDH release (Fig. 1, p > 0.05). Thus, SMC-treatedPC12 cells followed by normal incubation conditions were notused for further analyses. OGD led to 44% cell viability andincreased LDH release 3.5 fold. Compared with OGD treat-ment alone, SMC pre-treatments at 1–8 mM led to 54–83% cellviability (p < 0.05), and dose-dependently decreased LDHactivity (p < 0.05). OGD down-regulated Bcl-2 expression, andup-regulated protein production of Bax and cleaved caspase-3(Fig. 2, p < 0.05). SMC pre-treatments at 2, 4 and 8 mMdecreased 16–67% cleaved caspase-3 expression and 28–63%Bax expression (p < 0.05), SMC only at 4 and 8 mM raised 27–47% Bcl-2 expression (p < 0.05).

This journal is © The Royal Society of Chemistry 2014

OGD reduced MMP, and increased Ca2+ release and DNAfragmentation (Table 1, p < 0.05). SMC pre-treatments at 1–8 mMmaintained MMP, and lowered intracellular Ca2+ release andDNA fragmentation (p < 0.05). OGD raised the activity andexpression of caspase-3 and caspase-8 (Fig. 3, p < 0.05). SMC pre-treatments at 2, 4 and 8 mM decreased the activity and expres-sion of caspase-3 (p < 0.05) by 26–71% and 35–69% respectively,but dose-dependently reduced the activity and expression ofcaspase-8 (p < 0.05) by 24–83% and 32–80% respectively.

Food Funct., 2014, 5, 1125–1133 | 1129

Fig. 5 Effects of SMC upon HIF-1a, NF-kB p50, NF-kB p65 and MAPK expression determined by western blot analyses. NGF-differentiated PC12cells were pretreated with SMC at 0, 1, 2, 4 or 8 mM for 4 h and followed by incubation under OGD conditions. The control groups were cellscontaining no SMC and incubated under normal conditions. The data are mean � SD (n ¼ 10). (a–f) Means among bars without a common letterdiffer, p < 0.05.

Food & Function Paper

Publ

ishe

d on

21

Febr

uary

201

4. D

ownl

oade

d by

Tuf

ts U

nive

rsity

on

27/1

0/20

14 2

1:19

:36.

View Article Online

OGD increased ROS, GSSG and 8-OHdG production, low-ered the GSH level, and decreased GPX and GR activities(Table 2, p < 0.05). SMC pre-treatments at 1–8 mM retained theGSH level, but reduced ROS and GSSG formation (p < 0.05).SMC at 2, 4 and 8 mM lowered 8-OHdG generation, andmaintained GPX and GR activities (p < 0.05). OGD enhancedNOS activity, and increased NO and 3-nitrotyrosine produc-tion (Table 3, p < 0.05). SMC pre-treatments at 2, 4 and 8 mMdiminished NOS activity, and decreased NO and 3-nitro-tyrosine levels (p < 0.05). OGD suppressed GPX and GRexpression, and raised iNOS expression (Fig. 4, p < 0.05). SMCpre-treatments at 2, 4 and 8 mM maintained 23–59% GPXexpression and 26–74% GR expression, as well as decreasediNOS expression by 31–58%.

OGD enhanced cytosolic HIF-1a, NF-kB p50 and NF-kB p65expression (Fig. 5, p < 0.05) and the SMC pre-treatments sup-pressed these expressions (p < 0.05). OGD also up-regulatednuclear HIF-1a, NF-kB p50, NF-kB p65 and p-p38 expression

1130 | Food Funct., 2014, 5, 1125–1133

(Fig. 5, p < 0.05). SMC pre-treatments at 1–8 mM lowered HIF-1aexpression by 21–76%, and p38 phosphorylation by 22–68%(p < 0.05). SMC at 2, 4 and 8 mM attenuated protein expression ofNF-kB p50 and NF-kB p65 (p < 0.05). SMC at test doses failed toaffect the phosphorylation of JNK and ERK1/2 (p > 0.05). As shownin Fig. 6, YC-1, PDTC or SB203580 blocked the activation of HIF-1a, NF-kB and p38, and the SMC pre-treatments did not affect theprotein expression of these factors (p > 0.05). These inhibitorssignicantly restored cell viability, and decreased ROS and NOproduction when compared with OGD treatment alone (Table 4,p < 0.05). The SMC treatments, under the presence of inhibitors,did not affect the cell viability, and ROS or NO formation (p > 0.05).

Discussion

OGD activated signaling pathways, evoked oxidative stress, andcaused apoptosis in neural cells.20,21 Our present study foundthat SMC pre-treatments markedly protected the NGF-treated

This journal is © The Royal Society of Chemistry 2014

Fig. 6 Effects of SMC upon the nuclear expression of HIF-1a (a), NF-kB(b) and p38 (c) with the presence of an inhibitor. NGF-differentiatedPC12 cells were pretreated with SMC at various doses for 4 h, and 10 mMYC-1 (HIF-1a inhibitor), PDTC (NF-kB inhibitor) or SB203580 (p38inhibitor) for 1 h, and incubated under OGD conditions. The controlgroups were cells containing no SMC and incubated under normalconditions. The data are mean � SD (n ¼ 10).

Table 4 Effects of SMC upon the cell viability (%), ROS (nmol mg�1

protein) and NO (mmol mg�1 protein) levels under the presence of aninhibitor. NGF-differentiated PC12 cells were pretreated with SMC atvarious doses for 4 h, and 10 mM YC-1 (HIF-1a inhibitor), PDTC (NF-kBinhibitor) or SB203580 (p38 inhibitor) for 1 h, and incubatedunder OGD conditions. The control groups were cells containing noSMC and incubated under normal conditions. The data are mean � SD(n ¼ 10)

Viability ROS NO

Control 100 � 2c 0.17 � 0.03a 3.27 � 0.18a

OGD 43 � 4a 1.42 � 0.15c 24.11 � 1.34c

YC-1 + OGD 87 � 3b 0.35 � 0.08b 7.62 � 0.42b

SMC, 1 85 � 5b 0.31 � 0.02b 7.33 � 0.36b

SMC, 2 88 � 4b 0.29 � 0.05b 7.21 � 0.38b

SMC, 4 90 � 4b 0.30 � 0.06b 7.17 � 0.40b

SMC, 8 87 � 3b 0.28 � 0.03b 7.10 � 0.35b

PDTC + OGD 92 � 4b 0.43 � 0.07b 7.58 � 0.43b

SMC, 1 93 � 5b 0.33 � 0.04b 7.05 � 0.39b

SMC, 2 91 � 5b 0.30 � 0.02b 6.89 � 0.29b

SMC, 4 91 � 3b 0.32 � 0.06b 6.94 � 0.32b

SMC, 8 93 � 2b 0.31 � 0.05b 6.75 � 0.24b

SB203580 + OGD 85 � 4b 0.32 � 0.03b 7.03 � 0.37b

SMC, 1 86 � 4b 0.30 � 0.02b 7.13 � 0.30b

SMC, 2 88 � 3b 0.29 � 0.04b 6.83 � 0.27b

SMC, 4 86 � 5b 0.28 � 0.05b 6.72 � 0.32b

SMC, 8 88 � 2b 0.29 � 0.03b 6.95 � 0.36b

a Means in a column without a common letter differ, p < 0.05. b Meansin a column without a common letter differ, p < 0.05. c Means in acolumn without a common letter differ, p < 0.05.

Paper Food & Function

Publ

ishe

d on

21

Febr

uary

201

4. D

ownl

oade

d by

Tuf

ts U

nive

rsity

on

27/1

0/20

14 2

1:19

:36.

View Article Online

PC12 cells against subsequent OGD-induced injury andenhanced cell survival. These ndings implied that SMC was apreventive agent to protect neural cells against hypoxic injury.

Bax and caspase-3 are pro-apoptotic molecules and Bcl-2 isan anti-apoptotic molecule. In our present study, SMC at 2, 4and 8 mM down-regulated Bax, cleaved caspase-3 production,and retained Bcl-2 expression in PC12 cells under OGD condi-tions, which in turn diminished apoptotic injury and benetedcell survival. These results suggested that this compound couldpenetrate into the NGF-treated PC12 cells, and mitigateapoptotic stress under OGD conditions by regulating both anti-apoptotic and pro-apoptotic molecules. OGD disrupted mito-chondrial membrane permeability and increased Ca2+ releasein neural cells, which consequently triggered the apoptoticprocess.22,23 White et al.24 reported that excessive Ca2+ activatedCa2+-dependent catabolic enzymes, such as caspase cascades,and caused neuronal injury. The loss of mitochondrialmembrane potential could activate caspases, including caspase-3 and caspase-8.25 These two caspases could act as apoptoticexecutors, responsible for cell death because they directly affectcell morphological changes and the cleavage of nuclearproteins.26 In our present study, SMC pre-treatments main-tained the mitochondrial membrane potential, decreased the

This journal is © The Royal Society of Chemistry 2014

intracellular Ca2+ concentration, and repressed the activity andprotein expression of caspase-3 and caspase-8 in the NGF-treated PC12 cells. Since SMC retarded those adverse eventscaused by OGD, the lower LDH activity and greater viability inSMC-treated PC12 cells under OGD conditions could beexplained.

Oxidative stress is a crucial factor contributing to OGD-induced neuronal cell death.2,3 We found that SMC pre-treat-ments effectively decreased ROS and GSSG formation, retainedthe GSH content, and preserved the activity and expression ofGPX and GR in PC12 cells under OGD conditions. Apparently,SMC was able to diminish OGD-evoked oxidative stress in theNGF-treated PC12 cells by enhancing the glutathione redoxcycle. 8-OHdG is a marker of DNA oxidative damage. IncreasedDNA fragmentation and 8-OHdG generation in OGD-treatedPC12 cells, as we observed, indicates that the nuclear compo-nents of these cells were impaired. However, SMC pre-treat-ments at 2–8 mM attenuated OGD-induced DNA fragmentationand 8-OHdG production. These data suggest that SMC couldpenetrate into cells and protect the DNA and nuclear compo-nents in these cells. Esposito et al.27 reported that the expressionof iNOS was up-regulated under cerebral ischemic conditions,which promoted oxidative and inammatory injury, evenleading to cell death. In our present study, SMC pre-treatmentsat 2–8 mM lowered NO and 3-nitrotyrosine overproduction, andreduced the NOS activity and iNOS expression, which conse-quently mitigated RNS-related oxidative stress. In addition,increased ROS are responsible for Ca2+ release in neural cells.28

Since the ROS and RNS levels have been reduced, the decreased

Food Funct., 2014, 5, 1125–1133 | 1131

Food & Function Paper

Publ

ishe

d on

21

Febr

uary

201

4. D

ownl

oade

d by

Tuf

ts U

nive

rsity

on

27/1

0/20

14 2

1:19

:36.

View Article Online

Ca2+ release and greater viability in SMC-treated PC12 cellscould be explained. Thus, the observed protection from SMC inNGF-treated PC12 cells against OGD could be ascribed to itsanti-oxidative activities.

OGD enhances the expression of HIF-1a and NF-kB, andpromotes their translocation from cytosol to the nucleus.29,30Wefound that SMC pre-treatments lowered OGD-induced expres-sions of HIF-1a and NF-kB in both cytosolic and nuclear frac-tions. Apparently, SMC was an effective inhibitory agent uponthe activation of HIF-1a and NF-kB, two key transcriptionfactors. It is reported that the activation of HIF-1a, NF-kB andMAPK pathways from OGD elicits the generation of oxidativeand apoptotic factors, and nally facilitates cell death.31,32 Luet al.28 indicated that attenuating p38 signaling could reduceOGD-associated neuronal cell death in rat hippocampalneurons. In our present study, SMC dose-dependently sup-pressed nuclear HIF-1a expression and at 2–8 mM, limited p38phosphorylation and nuclear NF-kB expression in OGD-treatedPC12 cells. Therefore, the observed improvement from SMCupon cell survival against OGD could be explained as itssuppressive effects upon the activation of HIF-1a, p38 andNF-kB. On the other hand, the presence of YC-1, PDTC andSB203580 blocked the regulation of SMC upon HIF-1a, p38 andNF-kB expression, counteracted the OGD-induced cytotoxicity,and decreased ROS and NO production. These ndings onceagain supported that these pathways were essential for SMC toexecute its protective actions. Under the presence of inhibitors,the slight decrease in ROS or NO levels from SMC treatmentsmight be simply due to SMC's anti-oxidative activity. In addi-tion, HIF-1a and NF-kB are key factors responsible for theproduction of iNOS and NO.29,33 Since SMC down-regulated theexpression of HIF-1a and NF-kB, it was reasonable to observelower iNOS expression and NO formation.

SMC is a cysteine derivative, and naturally occurs in manyplant foods, such as garlic and onion. Our previous studyreported that dietary SMC intake increased GSH content inmicestriatum.11 Thus, the consumption of SMC or foods rich in thiscompound may be safe and benecial for neuronal protection.Although we found SMC exhibited substantial protective activ-ities against OGD in NGF-differentiated PC12 cells, this presentstudy was based on a cell line model. So far, informationregarding the availability and effects of this compound in thebrain from dietary intake is limited. Thus, further animalhypoxic study is necessary to examine the deposit, protectiveeffects, actionmodes and dosage safety of this compound in thebrain before it is applied to humans.

Conclusion

In conclusion, our present study found that SMC pre-treatmentsat 2–8 mM markedly enhanced NGF-differentiated PC12 cellsurvival under OGD conditions. SMC decreased OGD inducedoxidative and apoptotic stress via suppressing HIF-1a, NF-kBand p38 activation, decreasing ROS and RNS production, andstabilizing the mitochondrial membrane. These results suggestthat SMC might be a potent neuro-protective agent againsthypoxic injury.

1132 | Food Funct., 2014, 5, 1125–1133

Acknowledgements

This study was partially supported by a grant from the ChinaMedical University, Taichung City, Taiwan (CMU102-ASIA-01).

References

1 K. Kitagawa, J. Neurosci. Res., 2012, 90, 1043–1054.2 T. Sugawara and P. H. Chan, Antioxid. Redox Signaling, 2003,5, 597–607.

3 I. Olmez and H. Ozyurt, Neurochem. Int., 2012, 60, 208–212.4 A. Sanchez, D. Tripathy, X. Yin, K. Desobry, J. Martinez,J. Riley, D. Gay, J. Luo and P. Grammas, J. Alzheimer's Dis.,2013, 32, 587–597.

5 C. C. Chio, J. W. Lin, H. A. Cheng, W. T. Chiu, Y. H. Wang,J. J. Wang, C. H. Hsing and R. M. Chen, Arch. Toxicol.,2013, 87, 459–468.

6 W. Yin, L. Ma, J. Zhang, K. Huang, Q. Yang, Y. Y. Guo,S. B. Liu, Y. H. Liu and Y. M. Wu, CNS Neurosci. Ther.,2013, 19, 145–153.

7 N. Singh, G. Sharma and V. Mishra, Cell. Mol. Neurobiol.,2012, 32, 491–507.

8 X. Fan, C. J. Heijnen, M. A. van der Kooij, F. Groenendaal andF. van Bel, Brain Res. Rev., 2009, 62, 99–108.

9 S. H. Yeh, L. C. Ou, P. W. Gean, J. J. Hung and W. C. Chang,Brain Pathol., 2011, 21, 249–262.

10 M. G. Jones, J. Hughes, A. Tregova, J. Milne, A. B. Tomsettand H. A. Collin, J. Exp. Bot., 2004, 55, 1903–1918.

11 C. M. Chen, M. C. Yin, C. C. Hsu and T. C. Liu, Nutrition,2007, 23, 589–597.

12 R. Wassef, R. Haenold, A. Hansel, N. Brot, S. H. Heinemannand T. Hoshi, J. Neurosci., 2007, 27, 12808–12816.

13 S. Ishiwata, S. Ogata, A. Umino, H. Shiraku, Y. Ohashi,Y. Kajii and T. Nishikawa, Amino Acids, 2013, 44, 1391–1395.

14 S. A. Fuchs, C. M. Peeters-Scholte, M. M. de Barse,M. W. Roeleveld, L. W. Klomp, R. Berger and T. J. deKoning, Amino Acids, 2012, 43, 355–363.

15 B. J. McCranor, R. A. Bozym, M. I. Vitolo, C. A. Fierke,L. Bambrick, B. M. Polster, G. Fiskum andR. B. Thompson, J. Bioenerg. Biomembr., 2012, 44, 253–263.

16 S. J. Tsai, C. Y. Lin, M. C. Mong, M. W. Ho and M. C. Yin,J. Agric. Food Chem., 2010, 58, 7104–7108.

17 B. Lenart, D. B. Kintner, G. E. Shull and D. Sun, J. Neurosci.,2004, 24, 9585–9597.

18 W. Fu, H. Luo, S. Parthasarathy and M. P. Mattson,Neurobiol. Dis., 1998, 5, 229–243.

19 B. A. Sutherland, O. M. Shaw, A. N. Clarkson,D. N. Jackson, I. A. Sammut and I. Appleton, FASEB J.,2005, 19, 258–260.

20 M. T. Rizzo and H. A. Leaver, Mol. Neurobiol., 2010, 42, 52–63.

21 M. D. Martın-de-Saavedra, L. del Barrio, N. Canas, J. Egea,S. Lorrio, E. Montell, J. Verges, A. G. Garcıa andM. G. Lopez, Neurochem. Int., 2011, 58, 676–683.

22 Y. M. Leung, BioMedicine, 2011, 1, 16–20.23 Z. Mo, Y. Fang, Y. He and X. Ke, Cell Biol. Int., 2012, 36, 1043–

1048.

This journal is © The Royal Society of Chemistry 2014

Paper Food & Function

Publ

ishe

d on

21

Febr

uary

201

4. D

ownl

oade

d by

Tuf

ts U

nive

rsity

on

27/1

0/20

14 2

1:19

:36.

View Article Online

24 B. C. White, J. M. Sullivan, D. J. DeGracia, B. J. O'Neil,R. W. Neumar, L. J. Grossman, J. A. Rafols andG. S. Krause, J. Neurol. Sci., 2000, 179, 1–33.

25 S. M. Cardoso, A. C. Rego, N. Penacho and C. R. Oliveira,Neurochem. Int., 2004, 45, 693–698.

26 A. M. Shari, H. Eslami and B. Larijani, Neurosci. Lett., 2009,459, 47–51.

27 G. Esposito, D. De Filippis, M. C. Maiuri, D. De Stefano,R. Carnuccio and T. Iuvone, Neurosci. Lett., 2006, 399, 91–95.

28 Q. Lu, T. F. Rau, V. Harris, M. Johnson, D. J. Poulsen andS. M. Black, Eur. J. Neurosci., 2011, 34, 1093–1101.

This journal is © The Royal Society of Chemistry 2014

29 V. Sivakumar, J. Lu, E. A. Ling and C. Kaur, Brain Pathol.,2008, 18, 71–85.

30 S. C. Correia and P. I. Moreira, J. Neurochem., 2010, 112, 1–12.

31 W. T. Lee, S. Hong, S. H. Yoon, J. H. Kim, K. A. Park,G. J. Seong and J. E. Lee, Brain Res., 2009, 1281, 64–70.

32 C. Nito, H. Kamada, H. Endo, P. Narasimhan, Y. S. Lee andP. H. Chan, J. Neurotrauma, 2012, 29, 2404–2412.

33 H. Ihara, H. Yamamoto, T. Ida, H. Tsutsuki, T. Sakamoto,T. Fujita, T. Okada and S. Kozaki, Biosci., Biotechnol.,Biochem., 2012, 76, 1843–1848.

Food Funct., 2014, 5, 1125–1133 | 1133