Overexpression of melatonin membrane receptors increasescalcium-binding proteins and protects VSC4.1 motoneurons fromglutamate toxicity through multiple mechanisms

Introduction

Amyotrophic lateral sclerosis (ALS) and spinal cord injury(SCI) are devastating central nervous system (CNS) disor-ders that result in progressive and selective loss of

motoneurons in the ALS and both motor and sensoryneurons in the SCI [1]. Unfortunately, effective therapeu-tic interventions in these disorders are lacking. Well-

characterized mechanisms of motoneuron death in bothALS and SCI include excitotoxicity, free radical injury,inflammation, ischemia, and eventual activation of proteases(calpains and caspases) for mediation of cell death [2, 3].

These factors together create a hostile tissue environmentthat is not conducive to repair or regeneration after theinsults. Thus, development of new therapies targeting these

detrimental processes is essential to promote motoneuronsurvival and enhanceneurological outcomes inALSandSCI.

Melatonin (N-acetyl-5-methoxytryptamine) is of particu-

lar interest in the treatment of neurodegenerative disorders

[4–6]. Melatonin is known to exert neuroprotection through

its interrelated actions as an antioxidative, anti-inflamma-tory, and antiapoptotic agent [7–10]. Melatonin is a scav-enger of both reactive oxygen species (ROS) and reactive

nitrogen species (RNS) to prevent oxidative injury [10–12].It has also been shown to prevent nuclear translocation ofnuclear factor-kappa B (NF-jB) and binding to target DNA[13], thereby reducing upregulation of a variety of proin-

flammatory cytokines [14]. More recently, the neuroprotec-tive actions of melatonin have been linked to its ability toprevent increases in intracellular free Ca2+associated with

excitotoxicity. This may be due, at least in part, to elevatedexpression of calcium-binding proteins (CaBPs) such ascalbindin and calmodulin that are capable of buffering

intracellular free Ca2+ levels. Unfortunately, knowledgeabout the molecular mechanisms by which melatoninpromotes these neuroprotective actions is lacking.

An understanding of the signaling cascades initiated bymelatonin may shed new light on both the pharmacology

Abstract: Melatonin has shown particular promise as a neuroprotective

agent to prevent motoneuron death in animal models of both amyotrophic

lateral sclerosis (ALS) and spinal cord injuries (SCI). However, an

understanding of the roles of endogenous melatonin receptors including

MT1, MT2, and orphan G-protein receptor 50 (GPR50) in neuroprotection

is lacking. To address this deficiency, we utilized plasmids for transfection

and overexpression of individual melatonin receptors in the ventral spinal

cord 4.1 (VSC4.1) motoneuron cell line. Receptor-mediated cytoprotection

following exposure to glutamate at a toxic level (25 lm) was determined by

assessing cell viability, apoptosis, and intracellular free Ca2+ levels. Our

findings indicate a novel role for MT1 and MT2 for increasing expression of

the calcium-binding proteins calbindin D28K and parvalbumin. Increased

levels of calbindin D28K and parvalbumin in VSC4.1 cells overexpressing

MT1 and MT2 were associated with cytoprotective effects including

inhibition of proapoptotic signaling, downregulation of inflammatory

factors, and expression of prosurvival markers. Interestingly, the

neuroprotective effects conferred by overexpression of MT1 and/or MT2

were also associated with increases in the estrogen receptor b (ERb): estrogenreceptor a (ERa) ratio and upregulation of angiogenic factors. GPR50 did

not exhibit cytoprotective effects. To further confirm the involvement of the

melatonin receptors, we silenced both MT1 and MT2 in VSC4.1 cells using

RNA interference technology. Knockdown of MT1 and MT2 led to an

increase in glutamate toxicity, which was only partially reversed by

melatonin treatment. Taken together, our findings suggest that the

neuroprotection against glutamate toxicity exhibited by melatonin may

depend on MT1 and MT2 but not GPR50.

Arabinda Das1, Casey Holmes1,Joshua A. Smith1, Gerald WallaceIV1, Russel J. Reiter2, Abhay K.Varma1, Swapan K. Ray3 andNaren L. Banik1

1Division of Neurology, Department of

Neurosciences, Medical University of South

Carolina, Charleston, SC, USA; 2Department

of Cellular and Structural Biology, University of

Texas, San Antonio, TX, USA; 3Department of

Pathology, Microbiology, and Immunology,

University of South Carolina School of

Medicine, Columbia, SC, USA

Key words: apoptosis, calbindin D28K,

calpain, glutamate toxicity, G-protein receptor

50, melatonin receptors, parvalbumin, ventral

spinal cord 4.1

Address reprint requests to Naren L. Banik,

Division of Neurology, Department of

Neurosciences, Medical University of South

Carolina, Charleston, SC 29425, USA.

E-mail: baniknl@musc.edu

Received May 15, 2012;

Accepted June 8, 2012.

J. Pineal Res. 2012Doi:10.1111/j.1600-079X.2012.01022.x

� 2012 John Wiley & Sons A/S

Journal of Pineal Research

1

Mo

lecu

lar,

Bio

log

ical

,Ph

ysio

log

ical

an

d C

lin

ical

Asp

ects

of

Mel

ato

nin

and mechanisms of melatonin-mediated neuroprotection.Two distinct classes of melatonin membrane receptors(MTs) located in the plasma membrane have been identified

in humans: MT1 and MT2 (formerly designated, Mel1a andMel1b) [15, 16]. Both receptors are highly expressed in avariety of tissues including the CNS. Recently, G-proteinreceptor 50 (GPR50), an orphan melatonin-related recep-

tor, was also identified in humans [17]. This receptor sharesapproximately 45% homology with MT1 and MT2 at theamino acid level, but is incapable of binding melatonin.

Interestingly, GPR50 has been shown to dimerize withMT1 and antagonize MT1 function.Activation of melatonin membrane receptors results in a

number of physiological effects. MT1 activation inhibitsforskolin-stimulated cAMP formation, protein kinase A(PKA), and cAMP-responsive element binding protein(CREB) phosphorylation [18, 19]. Activation of MT1 also

increases the activity of mitogen-activated protein kinases(MAPKs) or extracellular signal-regulated kinase kinases 1and 2 (MEK1 and MEK2) and extracellular signal-

regulated kinases 1 and 2 (ERK1/2) via phosphorylation[18, 19]. MT1 enhances phosphoinositide turnover andregulates functional responses to melatonin in ion channels

[20]. Similar to MT1, activation of MT2 by melatonininhibits forskolin-stimulated cAMP formation. Addition-ally, activation of MT2 inhibits cGMP formation, which

differs from the typical response seen forMT1. In slices of thesuprachiasmatic nucleus (SCN), melatonin increases proteinkinaseC (PKC) activity through activation ofMT2. SelectiveMT antagonist application blocks these responses [18, 19].

Several studies also indicate that neuronal populationsexpressing CaBPs, including the calbindins, parvalbumin,and calmodulin help neurons to survive ischemic and

excitotoxic insults when compared with neurons lackingCaBPs [21–23]. Induction of calbindin D28K by tumornecrosis factors a and b (TNF-a and TNF-b) renders

different neuronal populations resistant to excitotoxicityand glucose deprivation, indicating a potential role forcalcium buffering by CaBPs in neuroprotection [23, 24].

Motoneurons exhibit relatively low levels of CaBP expres-sion and, thus, may exhibit increased susceptibility toneurodegenerative processes triggered by secondary SCI.Regardless of the processes initiating motoneuron injury,

induction of CaBP expression may offer a novel therapeuticstrategy to protect these cells following SCI. Recent evidencesuggests that melatonin may upregulate the expression of

CaBPs to aid in buffering intracellular free Ca2+ levels.However, the role of MT1 and MT2 in this process remainsunknown.

Recent findings indicate that MT1 and MT2, under someconditions, play a central role in melatonin�s cytoprotectiveeffects in motoneurons [25, 26]. Our current study expandsprevious understanding of neuroprotective signaling mech-

anisms of melatonin receptors by examining cell death,expression of calcium associated factors (e.g. calpain andcalbindin), inflammatory responses, and angiogenic mark-

ers after overexpression of a melatonin receptor (eitherMT1, MT2, or GPR50) in ventral spinal cord 4.1 (VSC4.1)motoneurons and exposure to glutamate. We also provide

evidence that silencing MT1 and MT2 by RNA interference(RNAi) technology enhances glutamate toxicity in

motoneurons and that GPR50 does not have a role inneuroprotective actions of melatonin.

Materials and methods

Cell culture and treatments

Ventral spinal cord 4.1 motoneurons were seeded at adensity of 1 · 106 cells/well in 35-cm2 plates for 24 h prior

to transfection [26]. After washing with serum-free standardmedium (DMEM/F12), motoneurons were transfected in1% low serum medium [26] over the course of 24 h at 37�Cwith 35 lL lipofectamine (Life Technologies, Grand Island,NY, USA) and 10 lg of GFP-tagged ORF clone ofhomosapiens MT1 (MTNR1A: RG210385), MT2

(MTNR1B: RG216445), or GPR50 (RG217839) plasmid(Origene, Rockville, MD, USA). We also transfected acontrol shuttle mammalian vector with C-terminal tGFPtag: (PS100010) in VSC4.1 cells as a negative control.

Medium was then replaced with standard medium contain-ing 2% fetal bovine serum (FBS) (HyClone, Logan, UT,USA) for 24 h. To check transfection efficiency, we

measured the fluorescence of pCMV6GFP reporter plasmidat 482 nm excitation and 502 nm emission in a SpectramaxGemini XS spectrofluorimeter (Molecular Devices, Sunny-

vale, CA, USA). Then, cells were treated with L-glutamate(LGA) once the transfection efficiency reached 50 ± 10%.Optimal doses of glutamate (Sigma, St. Louis, MO, USA)and melatonin (Sigma) were determined for VSC4.1 cells.

To compare the above-mentioned experiments, VSC4.1 cellswere treated with melatonin (100 nm) for 24 h followed by25 lm LGA. Cells were grown in an incubator at 37�C with

5% CO2 and full humidity. Cells were collected aftertreatments for 24 h. The proliferation rate of VSC4.1 cellsoverexpressing the MT receptor was lower than in vector-

transfected and parental cells, although it was statisticallynot significant in comparison with parental cells.

Trypan blue dye exclusion test for residual cellviability

Following LGA and melatonin treatments of the MT

overexpressing or MT silenced VSC4.1 cells, the viability ofattached and detached cell populations was estimated bytrypan blue dye exclusion test. Viable cells do not take up

trypan blue and maintain membrane integrity [19]. Deadcells with compromised cell membranes took up trypanblue. At least 500 cells were counted in four different fields.

The percentage of residual cell viability was calculatedusing the following formula: percentage of residual cellviability = [number of trypan blue negative cells/(numberof trypan blue positive cells + number of trypan blue

negative cells)] · 100.

ApopTag assay

ApopTag Peroxidase kit (Intergen, Purchase, NY, USA)was used to assess the extent of cell death following

different treatments. Following treatments, cells werewashed with phosphate-buffered saline (PBS), pH 7.4, andthen centrifuged to sediment onto the microscopic slides.

Das et al.

2

Residual PBS was removed and cells were fixed using 95%(v/v) ethanol and allowed to dry overnight. Slides werepretreated with a protein-digesting enzyme for 15 min and

then washed with distilled water for 2 min. Cells werequenched with 3% (v/v) hydrogen peroxide for 5 minfollowed by washing with PBS. Terminal deoxynucleotidyltransferase (TdT) enzyme was added to the pre-equilibrated

cells and incubated for 1 h at 37�C. Stop-buffer was addedto the slide and agitated for 15 s followed by 10 minincubation at room temperature. After washing three times

with PBS for 1 min each, anti-digoxigenin peroxidaseconjugate was added to the slides and incubated for30 min. After slides were washed twice with PBS, freshly

prepared peroxidase substrate 3,3¢-diaminobenzidine wasadded to the slides and kept for 6 min and slides were thenwashed twice with water. Slides were counterstained with0.5% (w/v) methyl green for 10 min followed by washing

with water and then 100% n-butanol. After 10 min, cellswere dehydrated in xylene for 2 min and then mounted withglass coverslips. Experiments were conducted in triplicates

and the percentage of ApopTag-positive cells was deter-mined by counting cells under the light microscopy.

Fura-2 assay

Intracellular free Ca2+ levels were determined by Fura-2

assay. Fura-2/AM, a fluorescent Ca2+ indicator, was usedas described previously [26]. The cell-specific constants weredetermined using the standards of Calcium CalibrationBuffer Kit with Magnesium (Molecular Probes, Eugene,

OR, USA).

Antibodies

Monoclonal primary IgG antibody against b-actin (Sigma)was used to standardize cytosolic protein loading on the

SDS-PAGE. Primary IgG antibodies against MT1, MT2,andGPR50 were purchased from Santa Cruz Biotech (SantaCruz, CA, USA). All other primary IgG antibodies werepurchased from either Santa Cruz Biotech or Calbiochem.

The secondary IgG antibodies were horseradish peroxidase-conjugated goat anti-mouse IgG (ICNBiomedicals, Aurora,OH, USA) and horseradish peroxidase-conjugated goat

anti-rabbit IgG (ICN Biomedicals, Irvine, CA, USA).

Western blotting

Western blotting was performed as described previously[25, 26]. Autoradiograms were scanned using Photoshop

software (Adobe Systems, Seattle, WA, USA), and opticaldensity (OD) of each band was determined using QuantityOne software (Bio-Rad, Hercules, CA, USA).

Colorimetric assay for the measurement of caspase-9and caspase-3 activities

Measurements of caspase activities in cells were per-formed with the commercially available caspase-9 (Abcam,Cambridge, MA, USA) and caspase-3 (Sigma) assay kits.

The colorimetric assays were based on the hydrolysis ofthe Ac-LEHD-pNA by caspase-9 and Ac-DEVD-pNA by

caspase-3, resulting in the release of the p-nitroaniline(pNA) moiety. Proteolytic reactions were carried out inextraction buffer containing 200 lg of cytosolic protein

extract and 40 lm Ac-LEHD-pNA or 40 lm Ac-DEVD-pNA. The reaction mixtures were incubated at roomtemperature for 2 h, and the formation of pNA wasmeasured at 405 nm in a colorimeter. The concentration

of the pNA released from the substrate was calculated fromthe absorbance values. Comparison of the absorbance ofpNA from an apoptotic sample with an uninduced control

allowed determination of the increases in caspase activities.Experiments were performed in triplicate.

Knockdown of endogeneous melatonin receptors

We employed small interfering RNA (siRNA) to knock-down expression of endogenous melatonin receptors.

Briefly, VSC4.1 cells were seeded at a density of6 · 105 cells/well overnight in 75-cm2 cell culture plates,washed with DMEM/F12, and then treated for 24 h with

Opti-MEM (Invitrogen, Carlsbad, CA). MT1 stealth SelectRNAi 3 siRNA set (MTNR1A Rat), MT2 stealth SelectRNAi 3 siRNA set (MTNR1B Rat), and GPR50 stealth

Select RNAi 3 siRNA set (GPR50 Rat) from a commercialsource (Invitrogen) were used to knockdown expression ofendogenous MT1, MT2, and GPR50. RNAi solutions were

prepared for transfection by mixing 30 lL RNAi (20 nm)with 1 mL Opti-Mem II media (Invitrogen) and 35 lLLipofectamine RNAiMAX (Invitrogen) with 1 mL Opti-Mem II. RNAi and transfection solutions were mixed

together and allowed to incubate at room temperature for10–15 min before being added to 10 mL cell culture media.Cells were co-transfected with pGFP to provide a normal-

ization control and a measure of transfection efficiency.This medium was replaced with growth medium at 4–6 hafter transfection. Transfection efficiencies in cells

were thereby evaluated using a fluorescence microscope.After 24 h of transfection, cells were treated with 25 lm

LGA or 100 nm melatonin (15-min post-treatment) +25 lm LGA.

Statistical analysis

Results were analyzed using StatView software (AbacusConcepts, Berkeley, CA, USA). Data are expressed asmean ± standard error of mean (SEM) of separate exper-

iments (n > 3) and compared by one-way analysis ofvariance (ANOVA) followed by Fisher�s post hoc test.Significant difference between control (Con) and LGA,

LGA-treated GPR50-overexpressed cells, or silenced MT1/MT2+ LGA cells, was indicated by *(P £ 0.05) or**(P £ 0.01). Significant differences between LGA treat-ment and melatonin treatment + LGA- or MT1- and/or

MT2-overexpressed cells +LGA were indicated by#(P < 0.05) or ##(P < 0.01).

Results

To assess the role MT1, MT2, and GPR50 in neuro-

protection following glutamate excitotoxicity, exogenousoverexpression (50 ± 10%) of MT1, MT2, and GPR50

Melatonin receptors promote calcium-binding protein expression

3

was performed in the VSC4.1 motoneurons. Western blotanalysis was used to confirm overexpression of melatoninreceptors (MT1, MT2, and GPR50) following plasmid

transfection. Results indicated successful overexpression ofMT1, MT2, and GPR50 in VSC4.1 motoneurons (Fig. 1).Viability (trypan blue dye exclusion test) and apoptosis(ApoTag assay) of VSC4.1 motoneurons were evaluated

under the light microscope. Cells treated with melatoninalone and MT overexpressing cells showed no significant

difference in viability or apoptotic death. Con and GPR50overexpressing cells treated with 25 lm LGA showed asignificant decrease in viability (>50%) and an increase in

the number of apoptotic cells (P < 0.05), whereas therewas no significant difference in cell viability and apoptoticdeath between Con cells and MT (MT1 and/or MT2)overexpressing cells exposed to LGA (P = 0.768), further

suggesting a role for the melatonin receptors MT1 andMT2, but not GPR50, in motoneuron death (Fig. 1).

(A) (B)

(C)

(D)

Fig. 1. Melatonin receptor (MT1 and MT2) overexpression prevented ventral spinal cord 4.1 (VSC4.1) motoneuron death. Plasmid-mediated increase (P›) in expression was shown. Treatment groups: control cells (Con), 100 nm melatonin (24 h pretreatment), MT1-overexpressed cells, MT2-overexpressed cells, MT1 + MT2–overexpressed cells, G-protein receptor 50 (GPR50)-overexpressed cells, 25 lm

L-glutamate (LGA) (24 h exposure), 100 nm melatonin (24 h pretreatment) + 25 lm LGA (24 h), MT1-overexpressed cells + 25 lm LGA(24 h), MT2-overexpressed cells + 25 lm LGA (24 h), MT1 + MT2–overexpressed cells + 25 lm LGA (24 h), GPR50-overexpressedcells + 25 lm LGA (24 h). (A) ApopTag assay showing representative images from each treatment group. Arrows indicate apoptotic cells.(B) Bar graphs indicating the percentage of apoptotic cells (ApopTag assay) and viability (trypan blue dye exclusion test) counted from eachgroup. (C) Representative pictures show levels of MT1, MT2, GPR50, and b-actin levels (Western blotting). (D) Bar graphs indicating thechanges in expression of MT1, MT2, and GPR50 over Con.

Das et al.

4

Cells treated for 24 h with LGA also demonstratedsignificant increase (P = 0.008) in intracellular [Ca2+] levelwhen compared with untreated Con cells as measured by

the Fura-2 assay. Increases in intracellular free Ca2+ levelsfollowing LGA exposure were attenuated by approximately50% (P = 0.019) in melatonin-treated cells and melatoninreceptor (MT1 and/or MT2)–overexpressed cells (Fig. 2).

Furthermore, there were no significant differences betweenintracellular free Ca2+ concentrations in untreated Concells and those treated with LGA plus melatonin or those

overexpressing the MT1 and MT2 receptors.Because the upregulation of CaBPs has been suggested as

a mechanism by which melatonin receptor activation may

reduce intracellular free Ca2+ levels, we have also assessedexpression of calbindin D28K and parvalbumin by Westernblotting. Our findings showed a significant reduction inexpression of calbindin D28K and parvalbumin proteins in

cells treated with LGA when compared with untreated cells.Melatonin treatment or overexpression of MT1 or MT2receptor restored expression of both calbindin D28K and

parvalbumin (Fig. 2). No changes in expression of calbin-din D28K and parvalbumin were seen following over-expression of GPR50 in cells exposed to LGA.

Increases in intracellular free Ca2+ levels could lead toactivation of the Ca2+ -dependent protease, calpain, fordegradation of calpastatin, the endogenous inhibitor of

calpain, and cell death [26]. Therefore, expression of calpain

and calpastatin were measured via Western blotting. Thecalpain/calpastatin ratio was significantly increased in cellsexposed to LGA alone, indicating a shift toward calpain

activation in these cells (P =0.009). Results obtained fromthe Western blotting showed a significant decrease(P =0.019) in the calpain/calpastatin ratio in cells treatedwith LGA + melatonin or in melatonin receptor (MT1

and MT2) overexpressing cells (Fig. 2).Previously reported results demonstrated that melatonin

may modulate expression and signaling activity of the

estrogen receptors (ERa and ERb). To ascertain whetherERa and ERb contribute to protection against LGA inmelatonin receptor overexpressing VSC4.1 motoneurons,

expression of ERa and ERb was measured at the proteinlevel via Western blotting (Fig. 3). MT1 and/or MT2overexpressing cells and normal VCS4.1 cells exposed tomelatonin showed increase in ERb and decrease in ERa at

protein levels (Fig. 3). Densitometric analysis of theWestern blots showed a significant increase in ERb/ERaratio in the cells treated with LGA + melatonin or in

melatonin receptor (MT1 and/or MT2) overexpressing cells,when compared with LGA-treated VSC4.1 cells. Significantchanges (P = 0.024) were seen in protein expression of ERaand ERb in cells treated with melatonin or in melatoninreceptor (MT1 and/or MT2) overexpressing cells, whencompared with Con cells. No changes in ERb/ERa ratio

were seen in GPR50 overexpressing cells (Fig. 3).

(A) (C)

(D)(B)

Fig. 2. Overexpression of MT1 and MT2 increases calbindin D28K and parvalbumin for suppressing Ca2+ rise and calpain:calpastatinratio in ventral spinal cord 4.1 (VSC4.1) cells. Plasmid-mediated increase (P›) in expression was shown. Treatment groups: control cells(Con), 100 nm melatonin (24 h pretreatment), MT1-overexpressed cells, MT2-overexpressed cells, MT1 + MT2–overexpressed cells, G-protein receptor 50 (GPR50)-overexpressed cells, 25 lm L-glutamate (LGA) (24 h), 100 nm melatonin (24 h pretreatment) + 25 lm LGA(24 h exposure), MT1-overexpressed cells + 25 lm LGA (24 h exposure), MT2-overexpressed cells + 25 lm LGA (24 h exposure),MT1 + MT2–overexpressed cells + 25 lm LGA (24 h exposure), GPR50-overexpressed cells + 25 lm LGA (24 h exposure). (A)Determination of intracellular free Ca2+ levels at 24 h. (B) Western blot analysis to show levels of calbindin D28K, parvalbumin, calpain,calpastatin, and b-actin. (C) Bar graphs indicating the changes in expression of calbindin D28K and parvalbumin over Con. (D) Densi-tometric analysis showing the calpain:calpastatin ratio.

Melatonin receptors promote calcium-binding protein expression

5

To determine whether overexpression of melatoninreceptor may suppress proinflammatory changes in moto-

neurons following LGA treatments, expression of nuclearfactor-jB (NF-jB) and cyclooxygenase (COX-2) weremeasured. Western blotting results showed increased

expression of both NF-jB and COX-2 in VSC4.1 cellstreated with LGA and in GPR50 overexpressing cells,whereas melatonin-treated cells or melatonin receptor

(MT1 and/or MT2) overexpressing cells reversed changesin and NF-jB and COX-2 at protein levels following LGAtreatments (Fig. 3). These results suggested that overex-

pression of MT1 and MT2 played an important role incontrolling the inflammatory response in motoneurons.Recently, it has been shown that upregulation of

the number of antiapoptotic or survival proteins (e.g.

Bcl-2, Akt) is associated with expression of VEGF and itsreceptors [Flk-1/KDR (VEGF-R2), and Flt-1 (VEGF-R1)].These factors may play a central role in preventing

programmed cell death, and their reduced expression hasbeen associated with various neurodegenerative disorders.Therefore, we examined expression of known survival

factors, VEGF, and VEGF receptors in VSC4.1 cells afterLGA treatment (Fig. 4). Western blotting demonstratedsignificant increases in phosphorylated Akt (p-Akt, anactivated survival protein), Bcl-2 (antiapoptotic protein),

p-Bad (an inactivated proapoptotic protein), and angio-genesis factors (VEGF, Flk-1, and Flt-1) in MT1 and/orMT2 overexpressing cells treated with or without LGA

(Fig. 4). Changes in these markers were not observed incells that overexpressed GPR50, as compared with Concells. Taken together, these data demonstrated that the

overexpression of MT1 and/or MT2 or melatonin treat-ment alone provided neuroprotective effects that might bemediated via VEGF-related pathways.

To understand the MT1/MT2 cascade responsible forBcl-2 relocalization to mitochondria, the intrinsic pathwayof apoptosis and the subsequent activation of caspases wereexamined. An increase in the Bax:Bcl-2 ratio is an impor-

tant initiator in the intrinsic or mitochondrial pathway ofapoptosis and a shift in the Bax:Bcl-2 ratio can activatecaspase-9 and caspase-3, the well-known apoptotic indica-

tors. Melatonin treatment or MT1 and/or MT2 overex-pressing cells treated with or without LGA exhibitedsignificant decreases in Bax:Bcl-2 ratio (P = 0.013),

37 kD active caspase-9 fragment, and caspase-9 activity,when compared with cells treated with LGA (Fig. 5).Uniform expression of b-actin served as a loading control

for cytosolic proteins. Melatonin-treated cells or MT1 and/or MT2 overexpressing cells did not demonstrate anysignificant change in the Bax:Bcl-2 ratio and caspase-9expression and activity when exposed to LGA. Caspase-3

activation and activity were measured by Western blottingand colorimetric assay, respectively (Fig. 5). There was asignificant decrease (P = 0.001) in the production of active

20 kD caspase-3 fragment and caspase-3 activity in mela-tonin-treated cells and melatonin receptor (MT1 and MT2)overexpressing cells, when compared with cells exposed to

LGA. GPR50 overexpression had no effect on Bax:Bcl-2ratio as well as on caspase-9 and caspase-3 expression andactivity.Finally, to ensure the involvement of MT1 and MT2 in

cytoprotection, endogenous MT1 and/or MT2 receptorswere silenced by RNAi and treated with LGA and/ormelatonin. Western blot analysis indicated effective knock-

down (�50% of control values) of MT1, MT2, andGPR50. Significantly decreased cell viability and increasedcaspase-3 activity were observed in LGA-treated and

MT1- and MT2-silenced cells when compared with LGA-treated cells (Fig. 6). These results strongly support the

(A)

(B)

(C)

Fig. 3. Overexpression of MT1 and MT2 increase estrogen recep-tor (ERb:ERa) ratio and suppression inflammatory factors inventral spinal cord 4.1 (VSC4.1) cells. Plasmid-mediated increase(P›) in expression was shown. Treatment groups: control cells(Con), 100 nm melatonin (24 h pretreatment), MT1-overexpressedcells, MT2-overexpressed cells, MT1 + MT2–overexpressed cells,G-protein receptor 50 (GPR50)-overexpressed cells, 25 lm L-glu-tamate (LGA) (24 h), 100 nm melatonin (24 h pretreat-ment) + 25 lm LGA (24 h), MT1-overexpressed cells + 25 lm

LGA (24 h), MT2-overexpressed cells + 25 lm LGA (24 h),MT1 + MT2–overexpressed cells + 25 lm LGA (24 h), GPR50-overexpressed cells + 25 lm LGA (24 h). (A) Western blot anal-ysis to show levels of ERb, ERa, NF-jB, COX-2, and b-actin.(B) Densitometric analysis showing the ERb: ERa ratio. (C) Bargraphs indicating the changes in the expression of NF-jB andCOX-2 over Con.

Das et al.

6

hypothesis that the endogenous MT1 and MT2 modulatecellular responses to excitotoxic injury. When the tests wererepeated following silencing of GPR50, no changes were

observed in cell viability and caspase-3 activity followingLGA exposure. Also, to determine whether the mechanismby which melatonin protects VSC4.1 motoneurons isprimarily receptor mediated, the melatonin receptor–

silenced cells were exposed to LGA + melatonin. Resultsobtained from viability and caspase-3 activity showedpartial inhibition of LGA toxicity in melatonin receptor–

silenced cells treated with LGA + melatonin, when com-pared with melatonin receptor–silenced cells and treatedwith LGA (Fig. 6). The reason may be that melatonin has

actions, which are not receptor mediated, particularly as aregulator of antioxidant and pro-oxidant enzymes. Simi-larly, GPR50-silenced cells treated with LGA + melatoninshowed slightly higher viability, when compared with

GPR50-silenced cells exposed to LGA alone. Based onthese findings, it appears that the absence of MT1 and MT2does reduce viability following glutamate excitotoxicity and

may reduce melatonin responsiveness. Thus, our resultsshowed the involvement of MT1 and MT2 in protection ofmotoneurons from glutamate excitotoxicity.

Discussion

Although melatonin membrane receptors are expressednearly ubiquitously, the functional role of these receptors inneuroprotection remains unclear. Previous findings fromour laboratory demonstrated the involvement of MT1 and

MT2 in neuroprotection in VSC4.1 cells exposed to avariety of insults such as LGA, oxidative stress, and TNF-a[26]. In the present investigation, we demonstrate that

overexpression of melatonin receptors (MT1 and/or MT2)leads to a significant reduction in cell death induced byglutamate excitotoxicity. The results obtained suggest that

overexpression of the melatonin receptor (MT1 and/orMT2) enhances the expression of a subset of CaBPs(calbindin D28K and parvalbumin), protects against celldeath by increasing antiapoptotic and angiogenic factors,

suppresses intracellular free Ca2+ levels, and therebyreduces the calpain:calpastain ratio. The present resultssupport a direct relationship between cytoprotection and

decreases in Bax:Bcl-2 ratio and protease activities (calpain,caspases-9, and caspase-3). These data demonstrate thatmelatonin receptor (MT1 and MT2)–silenced (50%) cells

exposed to LGA + melatonin are slightly more viable thanmelatonin receptor (MT1 and MT2)–silenced (50%) cellsexposed to LGA. Thus, the present study indicates the

involvement of the melatonin membrane receptors incytoprotective responses. Similar experiments evaluatingthe orphan melatonin receptor GPR50 by both overex-pression and RNAi knockdown suggest that GPR50 has no

role in cytoprotection of neurons.Motoneurons are particularly vulnerable to aberrant

intracellular Ca2+ influx owing to overactivation of gluta-

mate receptors, partly because expression of CaBPs, such ascalbindin D28K and parvalbumin are lowered [27, 28].Therefore, it is possible that increasing the ability of

motoneurons to buffer intracellular free Ca2+ may pro-tect them from cell death, thus preventing subsequent

(A)

(B)

(C)

Fig. 4 Overexpression of MT1 and MT2 increases survival andangiogenesic factors in ventral spinal cord 4.1 (VSC4.1) cells. Plas-mid-mediated increase (P›) in expression was shown. Treatmentgroups: control cells (Con), 100 nm melatonin (24 h pretreatment),MT1-overexpressed cells, MT2-overexpressed cells, MT1 + MT2–overexpressed cells, G-protein receptor 50 (GPR50)-overexpressedcells, 25 lm L-glutamate (LGA) (24 h exposure), 100 nm melatonin(24 h pretreatment) + 25 lm LGA (24 h exposure), MT1-overex-pressed cells + 25 lm LGA (24 h), MT2-overexpressed cells +25 lm LGA (24 h exposure), MT1 + MT2–overexpressed cells +25 lm LGA (24 h exposure), GPR50-overexpressed cells + 25 lm

LGA (24 h exposure). (A) Western blot analysis to show levels ofp-Akt, Bcl-2, p-Bad, Flk-1, Flt-1, VEGF, and b-actin. (B) Bargraphs indicating the changes in expression of p-Akt, Bcl-2, andp-Bad over Con. (C) Bar graphs indicating changes in the expressionof Flk-1, Flt-1, and VEGF over Con.

Melatonin receptors promote calcium-binding protein expression

7

decrements in motor function. This study tested thispossibility by exogenously overexpressing MT1 and/orMT2 to modulate motoneuron survival. We have demon-strated in this study, for the first time, that MT1 and MT2

overexpression upregulates the CaBPs, calbindin D28K andparvalbumin, to control intracellular free Ca2+ levels.Several experimental studies have shown that blocking

Ca2+ entry, either via voltage-gated Ca2+ channels orglutamate receptor antagonists protects against ischemiaand ischemia-like insults [29]. Limiting intracellular Ca2+

accumulation by administering various chelators has alsobeen shown to protect neurons against excitotoxicity andthe effects of experimental stroke. The current data show

that by overexpressing MT1 and/or MT2, VSC4.1 moto-neurons can be made more resistant to glutamate excito-toxicity, partly by upregulation of calbindin D28K andparvalbumin, arguing for a neuroprotective role for these

receptors. These results support previous findings whereinneurons induced to overexpress CaBP are more resistant toexcitotoxic and ischemia-related injury than neurons lack-

ing CaBP [29]. There are also reports that CaBP overex-pression protects neurons in other models ofneurodegeneration. Recently, parvalbumin has been sug-

gested as a marker of motoneurons that are resistant toALS. Interestingly, our work also showed that calbinidinD28K and parvalbumin overexpression protected againstLGA excitotoxicity and mitochondrial insults.

Previous results from our and other laboratories dem-onstrated that melatonin decreases intracellular free Ca2+

by inhibiting the influx through voltage-dependent

Ca2+ channels and inositol (1,4,5)-trisphosphate-mediatedCa2+ release from intracellular stores [30]. The present invitro study partially confirmed a unique role for MT1 and

MT2 in motoneuron protection via suppression of intra-cellular free Ca2+ dynamics in MT1- and/or MT2-overex-pressed motoneurons. As the calpain–calpastatin system is

activated during Ca2+ influx in a number of CNS diseases,calpain:calpastatin ratio was measured in melatonin recep-tor overexpressing motoneurons. Results obtained follow-ing overexpression of MT1 and/or MT2 suggest an

important role for them in modulation of neuronal defensemechanisms in response to excitotoxic injuries.Melatonin modulates expression and activity of the ERa

and ERb, which also have a role in neuroprotection [31].Whether MT1 or MT2 mediate this regulation of ERexpression remains unknown. In this study, we investigated

(A)(C)

(D)

(B)

Fig. 5. Overexpression of MT1 and MT2 suppresses apoptotic pathways in ventral spinal cord 4.1 (VSC4.1) cells. Plasmid-mediatedincrease (P›) in expression was shown. Treatment groups: control cells (Con), 100 nm melatonin (24 h pretreatment), MT1-overexpressedcells, MT2-overexpressed cells, MT1 + MT2–overexpressed cells, G-protein receptor 50 (GPR50)-overexpressed cells, 25 lm L-glutamate(LGA) (24 h), 100 nm melatonin (24 h pretreatment) + 25 lm LGA (24 h), MT1-overexpressed cells + 25 lm LGA (24 h), MT2-over-expressed cells + 25 lm LGA (24 h), MT1 + MT2–overexpressed cells + 25 lm LGA (24 h), GPR50-overexpressed cells + 25 lm LGA(24 h). (A) Western blot analysis to show levels of Bax, Bcl-2, active caspase-9, active caspase-3, and b-actin. (B) Densitometric analysisshowing the Bax:Bcl-2 ratio. (C) Bar graphs indicating the changes in expression of active caspase-9 and active caspase-3 over Con. (D)Colorimetric determination of caspase-9 and caspase-3 activities.

Das et al.

8

whether the ERa and ERb expression were regulated byMT1 and MT2 during LGA-induced cell death. Cellsurvival increased by approximately 27–36% in

LGA + melatonin or in LGA-treated MT1 and MT2overexpressing cells as compared with LGA-treated cells,and ERb:ERa ratio was augmented by treatment withLGA + melatonin or in LGA-treated MT1- and/or MT2-

overexpressed cells. These results suggest that overexpres-sion of MT1 and/or MT2 in VSC4.1 cells increaseERb:ERa ratio that may play a role in protecting cells

against LGA-mediated cell death.NF-jB is a widespread transcription factor that regulates

a number of genes involved in immune and inflammatory

responses (TNF-a, IL-2, IL-6, adhesion and chemotaxismolecules, and iNOS) and a NF-jB binding site in theCaBP-9k promoter region has been identified (nucleotides

)848 to )834 from the transcriptional start site) [32]. Thesignaling of NF-jB pathway is associated with the regula-tion of COX-2 expression. Thus, the inhibition of NF-jBand COX-2 in the melatonin receptor overexpressing cellsrepresents a general mechanism of protection againstinflammatory mediators. Melatonin was described asproducing MT1-dependent control of other immunoin-

flammatory pathways [33]. These include the COX-2-dependent prostaglandin E2-mediated inhibition of IL-2production and the control of IL-2 receptor function in

human lymphocytes [34]. Taken together, these resultssuggest that overexpression of MT1 and/or MT2 in VSC4.1cells attenuate inflammatory factors.

Several studies have indicated that Akt signaling path-ways participate in neuronal survival and regulation ofother downstream neuronal survival markers, such as Bcl-2

(A)

(B) (C)

Fig. 6. Silencing MT1 and MT2 enhanced glutamate toxicity in ventral spinal cord 4.1 (VSC4.1) cells. We indicated the use of RNAinterference (RNAi) to cause a decrease in expression. Treatment groups: control cells (Con), 25 lm L-glutamate (LGA) (24 h exposure),MT1-silenced cells + 25 lm LGA (24 h exposure); MT2-silenced cells + 25 lm LGA (24 h), MT1 + MT2-silenced cells + 25 lm LGA(24 h exposure); G-protein receptor 50 (GPR50) silenced cells + 25 lm LGA (24 h), 25 lm LGA (24 h), 25 lm LGA + 100 nm melatonin(24 h exposure), MT1-silenced cells + 25 lm LGA+ 100 nm melatonin (24h exposure), MT2 silenced cells + 25 lm LGA + 100 nm

melatonin (24 h exposure), MT1 + MT2-silenced cells + 25 lm LGA + 100 nm melatonin (24 h exposure), GPR50 silencedcells + 25 lm LGA + 100 nm melatonin (24 h exposure). (A) Representative pictures showing levels of MT1, MT2, GPR50, and b-actinlevels (Western blotting). (B) Bar graphs indicating the percentage of viability (trypan blue dye exclusion test) counted from each group. (C)Colorimetric determination of caspase-3 activity.

Melatonin receptors promote calcium-binding protein expression

9

[35, 36]. VEGF has an effect on hippocampal neurons toprotect them against glutamate excitotoxicity. This effect ismediated primarily by the Flk-1 receptor through Akt

signaling pathways. Our results also supported the generalhypothesis that Akt phosphorylates Bad, thus preventingits binding to and inhibition of Bcl-2 [35]. Furthermore, thecurrent results demonstrate that both MT1 and MT2 have

neuroprotective effect through activation of Akt. Thesefindings imply that MT1 and/or MT2 agonist-like agentsmay be useful for the treatment of neurodegenerative

disorders.These aspects of melatonin biology help explain why this

indoleamine or its receptor may deserve being classified

as neuroprotective factors. Previously published resultsdemonstrate that melatonin can prevent mitochondrialdamage induced by ROS, and this can contribute to controlthe intrinsic pathway of apoptotic death signaling through

MT1 and MT2 [26]. A decrease in Bax:Bcl-2 ratio in theMT1 and/or MT2 overexpressing cells confirmed antia-poptotic cascade activation. Since excitotoxic cell death

follows a classical apoptotic model and is dependent oncaspase activation and activity, we have analyzed thedownstream events in the MT1- and/or MT2-overexpressed

VSC4.1 cells after different treatments. Processing ofprocaspase-3 to its active form is considered to be a pointof no return in the death-signaling cascade. Treatment with

LGA alone leads to an increase in caspase-9 and caspase-3activities and is in agreement with the results of previousstudies [25, 26]. The current observations of a reduction incaspase-9 and caspase-3 expression and activity correlate

well with the observed reduction in the Bax:Bcl-2 ratio inMT1 and MT2 overexpressing cells treated with LGA.Taken together, these data support earlier evidence that

MT1 and MT2 mediate cytoprotection, in part, owing toinhibition of protease activity [25, 26].In this study, we also investigated the endogenous role

of melatonin receptors in motoneuron protection byRNAi silencing of MT1, MT2, and GPR50 in VSC4.1cells. Receptor silencing affected increased sensitivity to

LGA toxicity. Interestingly, when LGA and melatoninwere used to treat these silenced cells, a modest neuro-protective effect was maintained. Such an effect couldrepresent a receptor-independent role for melatonin-med-

iated neuroprotection, possibly due to the radical scav-enging activity of the indoleamine [37]. Conversely, noneuroprotection was evidenced by overexpressing GPR50

in VSC4.1 cells.In conclusion, the present study demonstrates that

overexpression of the melatonin receptors MT1 and

MT2 increases calbindin D28K and parvalbumin expres-sion, while also attenuates excitotoxic cell death. Theseprotective effects appear to be mediated by both receptor-dependent and receptor-independent mechanisms. To elu-

cidate the receptor-dependent neuroprotective efficacy ofmelatonin, further studies should be performed usingselective melatonin receptor agonists and antagonists. The

existence of MT1- and MT2-deficient mice may alsoprovide an excellent tool to study the mechanisms ofMT-mediated neuroprotection in vivo. In summary, this

study establishes the importance of MT1 and MT2 inmotoneuron protection from gluatamate excitotoxicity.

Acknowledgements

Completion of this project was made possible by fundingfrom the National Institute of Neurological Disordersand Stroke (NS-31622, NS-38146, and NS-41088) andthe South Carolina Spinal Cord Injury Research Fund

(SCSCIRF).

References

1. Lepore AC. Intraspinal cell transplantation for targeting

cervical ventral horn in amyotrophic lateral sclerosis and

traumatic spinal cord injury. J Vis Exp 2011; 18:55.

2. Dong XX, Wang Y, Qin ZH. Molecular mechanisms of

excitotoxicity and their relevance to pathogenesis of neurode-

generative diseases. Acta Pharmacol Sin 2009; 30:379–387.

3. Caccamo D, Campisi A, Curro M. Excitotoxic and post-

ischemic neurodegeneration: Involvement of transglutaminas-

es. Amino Acids 2004; 27:373–379.

4. Kwon KJ, Kim JN, Kim MK et al. Melatonin synergistically

increases resveratrol-induced heme oxygenase-1 expression

through the inhibition of ubiquitin-dependent proteasome

pathway: a possible role in neuroprotection. J Pineal Res 2011;

50:110–123.

5. Tsai MC, Chen WJ, Tsai MS et al. Melatonin attenuates

brain contusion-induced oxidative insult, inactivation of signal

transducers and activators of transcription 1, and upregulation

of suppressor of cytokine signaling-3 in rats. J Pineal Res 2011;

51:233–245.

6. Rosales-Corral SA, Acuna-Castroviejo D, Coto-Montes

A et al. Alzheimer�s disease: pathological mechanisms and

the beneficial role of melatonin. J Pineal Res 2012; 52:167–202.

7. Bonnefont-Rousselot D, Collin F, Jore D, Gardes-

Albert M. Reaction mechanism of melatonin oxidation by

reactive oxygen species in vitro. J Pineal Res 2011; 50:328–335.

8. Wu UI, Mai FD, Sheu JN et al. Melatonin inhibits microglial

activation, reduces pro-inflammatory cytokine levels, and res-

cues hippocampal neurons of adult rats with acute Klebsiella

pneumoniae meningitis. J Pineal Res 2011; 50:159–170.

9. Han F, Chen YX, Lu YM et al. Regulation of the ischemia-

induced autophagy-lysosome processes by nitrosative stress in

endothelial cells. J Pineal Res 2011; 51:124–135.

10. Galano A, Tan DX, Reiter RJ. Melatonin as a natural

ally against oxidative stress: a physicochemical examination.

J Pineal Res 2011; 51:1–16.

11. Poeggeler B, Saarela S, Reiter RJ, Manchester LC,

Barlow-Walden LR. Melatonin – a highly potent endoge-

nous radical scavenger and electron donor: new aspects of the

oxidation chemistry of this indole accessed in vitro. Ann N Y

Acad Sci 1994; 738:419–420.

12. Reiter RJ, Paredes SD, Manchester LC, Tan DX. Reduc-

ing oxidative/nitrosative stress: a newly-discovered genre for

melatonin. Crit Rev Biochem Mol Biol 2009; 44:175–200.

13. Natarajan M, Sadeghi K, Reiter RJ et al. The neurohor-

mone melatonin inhibits cytokine, mitogen and ionizing radi-

ation induced NF-KappaB. Biochem Mol Biol Int 1995;

37:1063–1070.

14. Samantaray S, Das A, Thakore NP et al. Therapeutic

potential of melatonin in traumatic central nervous system

injury. J Pineal Res 2009; 47:134–142.

15. Godson C, Reppert SM. The Mel1a melatonin receptor is

coupled to parallel signal transduction pathways. Endo-

crinology 1997; 138:397–404.

Das et al.

10

16. Dubocovich ML, Markowska M. Functional MT1 and

MT2 melatonin receptors in mammals. Endocrine 2005;

27:101–110.

17. Levoye A, Dam J, Ayoub MA et al. The orphan GPR50

receptor specifically inhibits MT1 melatonin receptor function

through heterodimerization. EMBO J 2006; 25:3012–3023.

18. Radio NM, Doctor JS, Witt-Enderby PA. Melatonin

enhances alkaline phosphatase activity in differentiating

human adult mesenchymal stem cells grown in osteogenic

medium via MT2 melatonin receptors and the MEK/ERK

(1/2) signaling cascade. J Pineal Res 2006; 40:332–342.

19. Hardeland R. Melatonin: signaling mechanisms of a pleio-

tropic agent. BioFactors 2009; 35:183–192.

20. Jockers R, Maurice P, Boutin JA, Delagrange P. Melato-

nin receptors, heterodimerization, signal transduction and

binding sites: what�s new? Br J Pharmacol 2008; 154:1182–1195.

21. Baimbridge KG, Celio MR, Rogers JH. Calcium-binding

proteins in the nervous system. Trends Neurosci 1992; 15:303–

308.

22. Goodman JH, Wasterlain CG, Massarweh WF et al.

Calbindin-D28k immunoreactivity and selective vulnerability

to ischemia in the dentate gyrus of the developing rat. Brain

Res 1993; 606:309–314.

23. Mattson MP, Rychlik B, Chu C, Christakos S. Evidence

for calcium reducing and excito-protective roles for the cal-

cium-binding protein calbindin-D28k in cultured hippocampal

neurons. Neuron 1991; 6:41–51.

24. Mattson MP, Cheng B, Baldwin SA et al. Brain injury and

tumor necrosis factors induce calbindin D-28k in astrocytes:

evidence for a cytoprotective response. J Neurosci Res 1995;

42:357–370.

25. Das A, Belagodu A, Reiter RJ et al. Cytoprotective effects

of melatonin on C6 astroglial cells exposed to glutamate

excitotoxicity and oxidative stress. J Pineal Res 2008; 45:

117–124.

26. Das A, Mcdowell M, Pava MJ et al. The inhibition of

apoptosis by melatonin in VSC4.1 motoneurons exposed to

oxidative stress, glutamate excitotoxicity, or TNF-alpha tox-

icity involves membrane melatonin receptors. J Pineal Res

2010; 48:157–169.

27. Elliott JL, Snider WD. Parvalbumin is a marker of

ALS-resistant motor neurons. NeuroReport 1995; 6:449–452.

28. Dekkers J, Bayley P, Dick JR et al. Over-expression of

parvalbumin in transgenic mice rescues motoneurons from

injury-induced cell death. Neuroscience 2004; 123:459–466.

29. Yenari MA, Minami M, Sun GH et al. Calbindin d28k

overexpression protects striatal neurons from transient focal

cerebral ischemia. Stroke 2001; 32:1028–1035.

30. Balık A, Kretschmannova K, Mazna P et al. Melatonin

action in neonatal gonadotrophs. Physiol Res 2004; 53(Suppl

1):S153–S166.

31. Del Rıo B, Garcıa Pedrero JM, Martınez-Campa C et al.

Melatonin, an endogenous-specific inhibitor of estrogen recep-

tor alpha via calmodulin. J Biol Chem 2004; 279:38294–38302.

32. Lee GS, Choi KC, Han HJ, Jeung EB. The classical and a

non-classical pathways associated with NF-kappaB are

involved in estrogen-mediated regulation of calbindin-D9k

gene in rat pituitary cells. Mol Cell Endocrinol 2007; 15:277.

33. Carrillo-Vico A, Reiter RJ, Lardone PJ et al. The mod-

ulatory role of melatonin on immune responsiveness. Curr

Opin Investig Drugs 2006; 7:423–431.

34. Carrillo-Vico A, Lardone PJ, Fernandez-Santos JM

et al. Human lymphocyte-synthesized melatonin is involved in

the regulation of the interleukin-2/interleukin-2 receptor

system. J Clin Endocrinol Metab 2005; 90:992–1000.

35. Zeng KW, Wang XM, Ko H et al. Hyperoside protects pri-

mary rat cortical neurons from neurotoxicity induced by

amyloid b-protein via the PI3K/Akt/Bad/Bcl(XL)-regulated

mitochondrial apoptotic pathway. Eur J Pharmacol 2011;

672:45–55.

36. Kilic U, Kilic E, Reiter RJ, Bassetti CL, Hermann DM.

Signal transduction pathways involved in melatonin-induced

neuroprotection after focal cerebral ischemia in mice. J Pineal

Res 2005; 38:67–71.

37. Kilic U, Yilmaz B, Ugur M et al. Evidence that membrane-

bound G protein-coupled melatonin receptors MT1 and MT2

are not involved in the neuroprotective effects of melatonin in

focal cerebral ischemia. J Pineal Res 2012; 52:228–235.

Melatonin receptors promote calcium-binding protein expression

11