J. Pineal Res. 2014; 56:213–223 © 2013 John Wiley & Sons A/S.

Published by John Wiley & Sons Ltd

Doi:10.1111/jpi.12114 Journal of Pineal Research

Melatonin improves neuroplasticity by upregulating the growth-

Molecular, Biological, Physiological and Clinical Aspects of Melatonin

associated protein-43 (GAP-43) and NMDAR postsynaptic density-

95 (PSD-95) proteins in cultured neurons exposed to glutamate
excitotoxicity and in rats subjected to transient focal cerebral
ischemia even during a long-term recovery period
Abstract: Recent evidence shows that the NMDAR postsynaptic density-95 Wei-Sheng Juan1,*, Sheng-Yang
(PSD-95), growth-associated protein-43 (GAP-43), and matrix Huang1,*, Che-Chao Chang1,
metalloproteinase-9 (MMP-9) protein enhance neuroplasticity at the subacute Yu-Chang Hung1, Yu-Wen Lin1,
stage of stroke. Here, we evaluated whether melatonin would modulate the Tsung-Ying Chen1,2, Ai-Hua Lee1,
Ai-Chiang Lee1, Tian-Shung Wu1,3
PSD-95, GAP-43, and MMP-9 proteins in cultured neurons exposed to
and E-Jian Lee1
glutamate excitotoxicity and in rats subjected to experimental stroke. Adult
1
male Sprague–Dawley rats were treated with melatonin (5 mg/kg) or vehicle at Neurophysiology Laboratory, Institute of
Biomedical Engineering & Department of
reperfusion onset after transient occlusion of the right middle cerebral artery Surgery, National Cheng Kung University
(tMCAO) for 90 min. Animals were euthanized for Western immunoblot Hospital, College of Medicine, National Cheng
analyses for the PSD-95 and GAP-43 proteins and gelatin zymography for the Kung University, Tainan, Taiwan; 2Department
of Anesthesiology, Buddhist Tzu-Chi University
MMP-9 activity at 7 days postinsult. Another set of animals was sacrificed for and Buddhist Tzu Chi General Hospital,
histologic and Golgi–Cox-impregnated sections at 28 days postinsult. In Hualien, Taiwan; 3Department of Chemistry,
cultured neurons exposed to glutamate excitotoxicity, melatonin significantly National Cheng Kung University, Tainan,
Taiwan
upregulated the GAP-43 and PSD-95 expressions and improved dendritic
aborizations (P < 0.05, respectively). Relative to controls, melatonin-treated
stroke animals caused a significant improvement in GAP-43 and PSD-95 Key words: glutamate excitotoxicity, melatonin,
neuroplasticity, neuroprotection, stroke
expressions as well as the MMP-9 activity in the ischemic brain (P < 0.05).
Address reprint requests to E-Jian Lee,
Consequently, melatonin also significantly promoted the dendritic spine
Department of Surgery, National Cheng Kung
density and reduced infarction in the ischemic brain, and improved University Medical center and Medical School,
neurobehaviors as well at 28 days postinsult (P < 0.05, respectively). Together, 138 Sheng-Li Road, Tainan 70428, Taiwan.
E-mail: ejian@mail.ncku.edu.tw
melatonin upregulates GAP-43, PSD-95, and MMP-9 proteins, which likely
*These two authors have made equal
accounts for its actions to improve neuroplasticity in cultured neurons exposed contributions to the study.
to glutamate excitotoxicity and to enhance long-term neuroprotection, Received November 4, 2013;
neuroplasticity, and brain remodeling in stroke rats. Accepted December 13, 2013.

Recent studies have highlighted the involvement of

Introduction
Current treatment modality for ischemic stroke is still con-
fined to thrombolysis administration that benefits only a
small proportion of stroke patients at the acute stage [1].
There is an urgent need to rectify the current thrombolytic
strategy so as to improve clinical outcome of stroke
patients at the subacute and chronic stages of recovery
[2–8]. Thus, functional deficits caused by ischemic stroke
may be restored either by enhancing the rewiring process
of the damaged brain tissues or by decreasing the remote
injury that occurs distal to the ischemic territory. More-
over, functional deficits may be replaced or compensated
for by de novo neurogenesis or using other intact brain
tissues. Strategies that have the potential to modulate neu-
roplasticity and/or neurogenesis, in addition to the neuro-
protective ability to reduce acute brain damage, have
therefore been suggested by us and others to improve
current stroke treatment [4–6, 8].
matrix metalloproteinase-9 (MMP-9) protein and NMDA-
type glutamate receptors (NMDAR) in synaptic plasticity
and brain remodeling following stroke [9–13]. At an early
stage of stroke, activated and overexpressed matrix metal-
loproteinases (MMPs) induce endothelial and extracellular
matrix (ECM) damage, leading to transmigration of
inflammatory cells and large, toxic molecules into the
brain parenchyma [14–16]. Conversely, MMPs enzymes
also played essential roles in the neuroplasticity and brain
remodeling at the subacute stage of stroke. In particular,
the MMP-mediated proteolysis is essential for regulating
synaptic plasticity and neurogenesis at the subacute and
late healing stages following stroke [11, 17–19].
The NMDA-type glutamate receptors (NMDAR) medi-
ate both trophic and excitotoxic signaling in CNS neurons
[12, 13, 20–22]. Inhibitors of the NMDAR postsynaptic
density-95 (PSD-95) within 3 hr postinsult reduced brain
damage following permanent or transient focal cerebral

213
Juan et al.
ischemia [22]. A major pathway of NMDAR signaling to
cell death in cortical and hippocampal neurons requires
the scaffolding protein PSD-95 and activation of neuronal
nitric oxide synthase (nNOS) [23]. Neuronal protection
mediated by PSD-95, thus, correlates well with reduced
activation of nNOS. Accordingly, MMPs and PSD-95 are
thought to be essential proteins for the ECM remodeling
and synaptic plasticity cascades, although each protein vir-
tually has shown a Janus face of either toxic or neuronal
plasticity/trophic effects depending on different time
course and/or molecular basis of injury.
The 43-kDa growth-associated protein (GAP-43) is a
nervous tissue-specific protein that is synthesized at high
levels during axonal growth in neuronal development and
axonal regrowth in regeneration in the peripheral and
central nervous systems [14–26]. GAP-43, therefore, repre-
sents an important marker for axonal regeneration and
sprouting.
We and several other investigators have previously dem-
onstrated that melatonin is highly effective in reducing
acute ischemic brain damage [8, 14, 15, 27–29]. We used
melatonin, primarily because the agent and its metabolites
are potent, blood-brain barrier-permeable antioxidants,
and radical scavengers [30–34]. Thus, melatonin protects
the brain against acute ischemic damage to both gray and
white matter, ECM, blood-brain barrier permeability, and
functional neurovascular unit [14, 15, 27, 35–38]. More
recently, we have shown that melatonin improved the den-
dritic spine density and the somatosensory field potentials
of the ischemic brain at the subacute stage of insult, and
this was associated with an elevated level of SNAP-25, but
not synaptophysin (marker of synaptogenesis), protein
expression [8]. Whereas melatonin has been shown to con-
fer some presynaptic and dendritic neuroplasticity at a
subacute stage of an ischemic insult, there is still a need to
further clarify whether melatonin offers neuroplasticity at
the postsynaptic, axonal sprouting and ECM signaling
cascades. In particular, we presumed that melatonin and
its active metabolites function to inhibit nNOS activity
[38–41] and therefore might be suited to enhance the PSD-
95-mediated synaptic plasticity. It was also important to
determine whether melatonin-induced neuroplasticity
remains effectively during a long-lasting recovery period.
In the current study, we therefore examined whether
melatonin reduces brain infarct volumes and dendritic
spine damage as well as improves the postsynaptic plastic-
ity, axonal sprouting, and ECM remodeling during a long-
term recovery after transient focal cerebral ischemia in
rats. In addition, we explored possible contributions of the
melatonin-mediated neuroplasticity in cultured neurons
exposed to glutamate excitotoxicity.
Materials and methods
All procedures performed were approved by the Subcom-
mittee on Research Animal Care of the University Medical
Center. All chemicals were purchased from Sigma-Aldrich
Co. (St Louis, MO, USA) unless otherwise indicated. Mel-
atonin was dissolved in polyethylene glycol 400 (PEG 400,
Sigma-Aldrich) or dimethylsulfoxide (DMSO, Sigma-
Aldrich).
214
Neuronal cultures, glutamate-induced excitotoxicity,
and assay for dendritic aborizations
Based on a method described previously [36, 42], cultured
neurons were obtained from cerebral cortices of 1-day-old
Sprague–Dawley rats. Experiments were undertaken on
cultured neurons between 7 and 10 days in vitro (DIV).
Neurons were incubated melatonin (20–100 lM) or vehicle
(0.1% DMSO). Cultured neurons (n = 10–14) were pre-
treated with melatonin (20–100 lM) or vehicle (0. 1%
DMSO) for 30 min, and then, they were exposed to gluta-
mate (300 lM) for 48 hr. Thereafter, neurons were
co-incubated with mouse primary monoclonal antibodies
antimicrotubule-associated protein-2 (MAP-2; dilution
1:1000, Santa Cruz Biotechnology, Santa Cruz, CA, USA)
antibody. Anti-rabbit IgG secondary antibody conjugated
with biotin (1:100, Jackson ImmunoResearch, Inc., West
Grove, PA, USA) was then added, followed by fluorescein
(DTAF)-conjugated streptavidin (1:150, Jackson Immuno-
Research). Neurons were co-incubated with 4’,-6-
diamidino-2-phenylindole (DAPI; 2 lM, Sigma-Aldrich).
Fluorescence was assessed at excitation 365 nm and emis-
sion >420 nm for DAPI and was at excitation 450–490 nm
and emission >515 nm for DTAF detection. Dendritic
aborizations (length and branching) of the immunostained
neurons were determined using a microscope (Olympus
IX71, Olympus Optical Co. Ltd, Tokyo, Japan) equipped
with epifluorescence and a computerized image analyzer
(Image Pro plus 5.1, Media Cybernetics, Silver Spring,
MD, USA).
Animal preparation, anesthesia, and monitoring
Male Sprague–Dawley rats, weighing 240–290 g, were sup-
plied by the University Laboratory Animal Center and
were allowed free access to food and water. Animals were
anesthetized with 1–2% halothane in 70% N2O/30% O2.
During surgery, body temperature was maintained at
37.0  0.5°C, which corresponded to brain temperature of
37.2  0.5°C [43], using a thermostatically controlled
heating blanket and rectal probe (Harvard Apparatus,
South Natick, MA, USA). The right femoral artery was
cannulated for measuring arterial blood gases, glucose,
hematocrit, and blood pressure.
Experimental model and grouping
Focal cerebral ischemia was employed by intra-arterial
suture occlusion of the proximal right middle cerebral
artery (MCA) for 90 min, as described previously [8, 14,
27, 36, 43–45]. Laser Doppler flowmetry (LDF, Laserflo
BMP2, Vasamedics, St. Paul, MN, USA) was used for
local cortical blood perfusion (LCBF) measurements, and
the data of LCBF were expressed as a percentage of the
baseline values.
Melatonin (Sigma-Aldrich) was dissolved in a mixture
of PEG 400 (Sigma-Aldrich) and 0.9% normal saline (3:7,
vol./vol.). A dose of melatonin of 5 mg/kg was chosen
based on the pharmacokinetic study of exogenous melato-
nin in animal models [46] and the neuroprotective
dose–response studies of melatonin in experimental stoke

[36–39]. In a randomized fashion, animals were given
intravenously either melatonin at 5 mg/kg (n = 39) or
vehicle (PEG-saline, n = 40) at the initiation time of reper-
fusion (i.e., 90 min after the onset of ischemia). In the first
series of experiments, animals treated either with melato-
nin (5 mg/kg, n = 10) or vehicle (PEG-saline, n = 10) were
assigned for evaluating brain infarction and neuronal
damage and were euthanized after 28 days following the
ischemic insult (Day 28). In the second set of experiments,
animals treated either with melatonin (5 mg/kg, n = 17) or
vehicle (n = 18) were assigned to assess the dendritic spine
density by the Golgi–Cox staining at Day 1 and Day 28
(n = 8 for both groups, respectively). The third series of
animals, received melatonin (5 mg/kg, n = 12) or vehicle
(n = 12) upon reperfusion, were assigned to evaluate the
synaptic integrity by Western blotting and the ECM
remodeling by zymography at Day 7. Additional 16
animals which received a sham-operated procedure served
as the nonischemic controls.
Animal sacrifice and histologic outcome measure
Sacrifice was performed under anesthesia by decapitation
after transcardial perfusion with normal saline at Day 28.
By the method described previously [8, 14, 36, 44, 45], the
brains were quickly removed and sectioned coronally,
followed by the staining of 0.5% cresyl violet. Brain
infarction volumes were measured using a computerized
image analyzer (MCID Elite, Imaging Research Inc.,
Ontario, Canada) and were expressed as a percentage of
the contralateral hemisphere volume [8, 14, 36]. In addi-
tion, individual cortical and subcortical (caudoputaminal
and hippocampal) infarct sizes and surviving neurons were
separately calculated.
Golgi–Cox analysis
At Day 1 or Day 28, animals were sacrificed, and the
brains were rapidly removed and immersed in a modified
Golgi–Cox solution for 14 days and, then, in 30%
sucrose solution for another 5 days [8]. Coronal brain
sections of 200 lm were harvested using a vibratome
(VSLM1 motorized Vibroslice, Campden Instrument,
Silbey, UK), mounted, and stained. The spine density of
the second- and third-order basilar dendrites of the pyra-
midal neurons at regions corresponding to the ischemic
core (layer II–III), the inner (layer III–IV), and the outer
(layer V–VI) boundary zones of the forepaw and the
hindpaw areas of the ischemic hemisphere, as well as at
the layer V–VI region of the contralateral, intact brain,
was quantified and traced at 1000 magnification by a
Zeiss Axioskop 2 Mot fluorescent microscope equipped
with an Axiocam high-resolution camera and AxioVision
image analysis system (Carl Zeiss, Oberkochen,
Germany).
Western blot analysis
Samples were obtained from cultured neurons and from
brain tissues of the ischemic core and penumbral regions,
which were quickly dissected on dry ice after the animal
Melatonin enhanced postinsult neuroplasticity
were sacrificed. Based on the changes measured in LCBF,
we defined an imaginary line, which was parallel to the
superior sagittal sinus (SSS) and 3.0 mm lateral to the
bregma, to divide the superior limb of the penumbral (left
side to the line) and the ischemic core (right side to the
line) cortices in the ischemic brain [8]. Cortical dissection
was continued between the rhinal fissure and the lateral
olfactory tract. The inferior margin of ischemic core was
defined as the cortical brain tissues located superiorly to
the level of the rhinal fissure.
Samples were homogenized in lysis buffer, containing
1% Triton X-100, 20 mM Tris-HCl (pH 7.5), 150 mM
NaCl, 0.5% deoxycholate, 1 mM EDTA, 0.1% SDS, and
were centrifuged at 18,000 g for 60 min at 4°C. Protein
concentrations of each sample were determined with a
BCA protein assay kit (Pierce, Rockford, IL, USA).
Cell lysates of 25 lg protein were separated by 10%
sodium dodecyl sulfate–polyacrylamide gels (SDS-PAGE)
and were transferred onto polyvinylidene difluoride
(PVDF) microporous membranes (IPVH00010, Millipore,
Billerica, MA, USA) [8, 45]. Membranes were blocked
with 5% milk, then probed with primary antibodies
against GAP-43 and PSD-95 (1:10,000 and 1:1000; Chem-
icon International, Temecula, CA, USA), and finally
incubated with horseradish peroxidase–conjugated immu-
noglobulin (1:5000; Chemicon International) [8, 45].
Bound antibody was visualized with the Amersham ECL
system (GE Healthcare Bio-Sciences Corp., Piscataway,
NJ, USA). Membranes were then probed for actin
(1:10,000; Chemicon International). Optical densities were
measured using a Luminescent Image Analyzer (Fujifilm
LAS-3000; Fuji Photo Film Co., Tokyo, Japan).
Gelatin zymography
Brain tissues were homogenized and centrifuged by the
method described previously [14–16]. Gelatinase activity in
the supernatants was extracted and purified with gelatin-
Sepharose 4B (GE Healthcare Bio-Sciences Corp., Piscata-
way, NJ, USA) [14, 15]. Samples normalized for protein
concentration were mixed with sample buffer and loaded
onto 7.5% sodium dodecyl sulfate (SDS)–polyacrylamide
gel containing 1.25 mg/mL gelatin. After electrophoresis,
the gel was washed with 2.5% Triton X-100 buffer for
three times. The gel was incubated in 100 mL of 50 mM
Tris-HCl (pH 7.4) containing 5 mM CaCl2, 200 mM NaCl,
and 0.2% Brij 35 at 37°C overnight. Gel was stained with
0.25% Coomassie blue R-250 (Sigma-Aldrich) and
destained appropriately.
Neurobehavioral testing and body weight
measurements
Body weight measurement was taken daily. A battery of
sensory-motor tests was conducted prior to and on a daily
basis after reperfusion by two observers unaware of treat-
ment protocol. Briefly, two neurologic grading systems
were used: (i) a sensorimotor grading scale modified from
previously published methods [8, 14, 29, 36, 44, 45, 47], (ii)
a grading scale of 0–28 developed by Clark et al. [44, 45,
48].
215
Juan et al.
Statistical analysis
All data were expressed as the mean standard error of
the mean (S.E.M.). Neurobehavioral scores and the den-
dritic spine densities were analyzed by Mann–Whitney
U-test. Paired Student’s t-test was used to evaluate the
response to a change in conditions, and unpaired Student’s
t-test/one-way ANOVA with Fisher’s protected least signifi-
cant difference (LSD) post hoc comparison was used to
evaluate differences between groups. P < 0.05 was selected
for statistical significance.
Results
At 10-DIV cultured neurons, the addition of glutamate
(300 lM) induced significant reductions in the dendritic
lengths and branches by 43.5% and 43.6%, respectively
(n = 10–14; P < 0.05; Fig. 1), relative to sham treatment,
cultured neurons. Melatonin (20–100 lM), however, exhib-
ited a neuroplasticity response by improving the dendritic
lengths and branches by 56.7–84.0 and 53.9–78.2%,
respectively, compared with the control-treated neurons
exposed to glutamate excitotoxicity (P < 0.05; Fig. 1).
Additionally, melatonin (n = 6) improved the expressions
of the GAP-43 and PSD-95 proteins by 116.4% and
216
45.0%, compared with controls (n = 6; P < 0.05, respec-
tively; Fig. 2A,B).
Seven animals (10.1%) died prior to completing the
recovery protocol and were excluded: Five were in the
vehicle-injected group, and two was in the melatonin-trea-
ted group. Relative to controls, melatonin did not signifi-
cantly affect the LCBF either at the ischemic core and the
penumbral regions, or at the contralateral cortical brain,
as assessed before and within 40 min after treatment
(P > 0.05; data not shown). The other physiologic param-
eters of the animals were kept within normal physiologic
limits during the course of experiments and did not differ
significantly between melatonin-treated animals and vehi-
cle-injected controls (P > 0.05; data not shown).
Brain infarction was reduced by 35.3% (P < 0.05) in
the melatonin-treated animals (n = 9; Fig. 3), when com-
pared with controls (n = 7), following a recovery period of
28 days. This reflected the melatonin-mediated reductions
in cortical and subcortical infarct sizes by 36.9 and 30.0%
(P < 0.05, respectively; Fig. 3) and increases in the num-
bers of the surviving neurons in the penumbral cortex and
caudoputamen by 246.7 and 289.2% (P < 0.001, respec-
tively; Fig. 3). Additionally, melatonin significantly
improved sensory, motor, and the 28-point neurologic
scores, but not the loss of body weight, taken at 1–28 days
Fig. 1. Melatonin improves dendritic
aborizations in cultured neurons expose to
glutamate excitotoxicity. In the microtu-
bule-associated protein-2 (MAP-2)-
immunostained neurons, treatment with
melatonin (20–100 lM) significantly
enhanced the numbers of total dendritic
lengths and branches, compared with
controls. *P < 0.05 versus vehicle data;
n = 10–14.
 Melatonin enhanced postinsult neuroplasticity

(A) (B)

Fig. 2. Western immunoblot analysis for

GAP-43 and PSD-95 in cultured neurons
expose to glutamate excitotoxicity.
Melatonin significantly improved the
expressions of GAP-43 (A) and PSD-95
(B), compared with vehicle-treated
controls. *P < 0.05 versus vehicle data;
n = 6 for each groups.

Fig. 3. Treatment with melatonin

reduced ischemic brain damage in rats
subjected to middle cerebral artery
(MCA) occlusion following a recovery
period of 28 days. The cresyl violet-
stained coronal sections were from
representative animals, which received an
intravenous injection of vehicle (PEG-
saline) or melatonin (5 mg/kg) at 90 min
after the ischemic onset. Six random and
nonoverlapping (500 9 400 lm²) regions
in the borders of the ischemic parietal
cortex and caudoputamen were selected
for counting the surviving neurons. Scale
bar=5 mm. melatonin-treated animals
(n = 9) had significantly reduced the
infarction volume and individual cortical
and striatal lesion sizes, and significantly
increased numbers of surviving neurons,
compared with controls (n = 7).
*P < 0.05 versus vehicle data.

after the ischemic onset (P < 0.05; Table 1), compared
with controls.
Intravenous injection of melatonin (n = 6) at 5 mg/kg
effectively improved the activity of the pro- and MMP-9
at penumbral cortical zones (Fig. 4A–C) by 24.0 and
15.0% (P < 0.05, respectively), compared with controls
(n = 6). Melatonin, however, did not affect the levels of
the pro- and MMP-9 activity at the ischemic core
(P > 0.05). Melatonin also did not affect the activity of
MMP-2 either at the ischemic core and the penumbral cor-
tical zones, but enhanced the MMP-2 activity by 22.6% at
the contralateral intact brain (Fig. 4A,D; P < 0.05).
Cerebral ischemia–reperfusion insult induced prolonged
reductions in the expressions of the PSD-95 and GAP-43
at the ischemic core and the penumbral cortical zones
(P < 0.05, respectively; Fig. 5A–D). Melatonin-treated
animals (n = 6), however, had a significantly improved
protein levels of the GAP-43 at the ischemic core and the
penumbral cortical zones (P < 0.05, respectively; Fig. 5A,
B), compared with vehicle-treated controls (n = 6). Mela-
tonin also significantly improved the expressions of PSD-
95 at the penumbral cortical zones (P < 0.05; Fig. 5D),
but did not affected the PSD-95 levels at the ischemic core
regions (P > 0.05; Fig. 5C).
At Day 1 and Day 28, vehicle-treated controls had a sig-
nificantly lower branches and density of the second- and
third-order basilar dendrites of the pyramidal neurons at
the ischemic core (layer II–III), the inner (layer III–IV)
and the outer (layer V–VI) boundary zones of the ischemic
cortex, compared to sham-operated, intact controls
(P < 0.01, respectively; Fig. 6A). Melatonin-treated
animals (n = 8 for both groups), however, had a signifi-
cantly improved numbers of mushroom spine density of
the second- and third-order basilar dendrites of the pyra-
midal neurons at the layers II–III, III–IV, and V–VI of
the ischemic hemisphere at Day 1 and 28 (P < 0.05,
respectively; Fig. 6A–C), relative to vehicle-injected con-
trols (n = 8 for both groups). Melatonin-treated animals

217
Juan et al.
Table 1. Melatonin improves sensorimotor behavioral scores and weight loss after cerebral ischemia–reperfusion

Neurologic behavioral score

Time Weight loss (g) Motor Sensory 28-point clinical scale

Day 1 Vehicle (n = 7) 29  3.1 2 (1.6–2.4) 4 (3.7–4.3) 17 (15.3–18.7)

Melatonin (n = 9) 25  16.3 1 (0.6–1.4)* 2 (1.5–2.5)* 9 (6.6–11.4)*
Day 7 Vehicle (n = 7) 20  17.4 2 (1.5–2.4) 3 (2.7–3.3) 15.5 (14.6–16.5)
Melatonin (n = 9) 15  50.0 1 (0.6–1.4)* 2 (1.7–2.3)* 8 (6.7–9.3)*
Day 14 Vehicle (n = 7) 28  19.0 2.5 (2.1–2.9) 3.5 (2.9–4.1) 16 (14.4–17.6)
Melatonin (n = 9) 20  56.4 1 (0.7–1.3)* 2 (1.6–2.4)* 8 (6.6–9.4)*
Day 21 Vehicle (n = 7) 72  5.8 2 (1.6–2.4) 3 (2.5–3.5) 16 (14.3–17.7)
Melatonin(n = 9) 51  56.0 1 (0.6–1.4)* 2 (1.6–2.4)* 8 (7.2–8.8)*
Day 28 Vehicle (n = 7) 82  31 2 (15–2.5) 3 (2.4–3.6) 14 (12.9–15.1)
Melatonin (n = 9) 87  24.8 1 (0.8–1.2)* 2 (1.7–2.3)* 8 (7.0–9.0)*

Weight data and neurologic behavioral scores are expressed by mean  S.E.M. and by median (95% CI), respectively. Intravenous injec-
tion of melatonin (5 mg/kg) 90 min after the onset of MCA occlusion significantly improved sensorimotor neurologic scores, but not
weight loss, compared with vehicle-injected control values.
*P < 0.05 versus vehicle data.

(B)
(A)

(C) (D)

Fig. 4. Zymography of MMP-2 and

MMP-9 in the ischemic brain. (A) The
photographs show typical pattern of
changes in MMP-2 and MMP-9 at the
ischemic core (IC), penumbra (P), and the
contralateral intact, homotopic (L) areas.
Melatonin not only significantly improved
the pro- and MMP-9 activity (B and C,
respectively) at the ischemic penumbral,
but not ischemic core, areas, but also
enhanced the MMP-2 activity (D) at the
contralateral intact brain, compared with
controls. *P < 0.05 versus vehicle data;
n = 6, respectively.

also had a significant improvement in the density of den-
dritic spines of pyramidal neurons at the layer V–VI
regions of the contralateral, intact brain at Day 1
(P < 0.05, respectively; Fig. 6D,E), but did not affect the
contralateral dendritic spine density at Day 28 (P > 0.05;
Fig. 6D,E), compared with vehicle-treated controls.
Discussion
This study showed that melatonin increased the
expressions of the PSD-95 and GAP-43 proteins, and
thus, improved dendritic spine lengths and branches
(aborizations) in cultured neurons exposed to glutamate
excitotoxicity. We also demonstrated that melatonin
not only improved neuronal dendritic spine density,
but also enhanced the PSD-95 and GAP-43 protein
expressions and the MMP-9 activity at a subacute
stage of recovery following cerebral ischemia–reperfu-
sion [9, 20, 24]. Another novel finding was that treat-
ment with melatonin reduced neuronal damage,
improved neuronal survival and plasticity [8, 35], and,
thus, enhanced neurobehavioral outcomes following a
long-term recovery period after transient focal cerebral
ischemia in rats.

218

(A)
(C)
Fig. 5. Western immunoblot analysis for
GAP-43 and PSD-95 in the ischemic
brain. The photographs showed typical
changes in the GAP-43 and PSD-95
protein expressions in the ischemic core
(A, C) and the penumbral (B, D) cortices
at Day 7 after ischemia. Densitometric
analyses showed that melatonin
significantly improved the expression in
the GAP-43 protein in the ischemic core
(A) and penumbral (B) cortices.
Melatonin also improved the PSD-95
protein at the penumbral (D), but not the
ischemic core (C), brain tissues.
*P < 0.05 versus control; n = 6 (each
column).
The neuroprotection and neuroplasticity observed here
at the levels of the neuronal stoma, dendritic spine density,
postsynaptic spines, axonal sprouting, and ECM remodel-
ing cannot be accounted for by changes in hemodilution (as
measured by blood hematocrit), arterial blood pressure,
heart rate, glucose, core temperature, or differences in
LCBF, because these values were not significantly different
when compared between vehicle-injected and melatonin-
treated animals [44, 48]. The finding that improvements for
the dendritic aborizations, postdendritic spine, and axonal
sprouting were also be observed in the cultured neurons
exposed to glutamate further justified the melatonin-medi-
ated long-lasting neuroplasticity, as shown here.
We have observed two specific patterns of reductions of
the dendritic spine density in macroscopically intact brain
tissues at Day 1 after transient MCA occlusion: (i) changes
in the peri-infarct penumbral regions and (ii) remote
changes contralateral to the side of the infarct [8, 27, 48,
49]. Consistent with the Day-7 data reported previously
[8], we found that at Day 1, both the two patterns of
reduced dendritic spine density could be improved with
melatonin either in the peri-infarct penumbral or the
remote contralateral, intact brain regions. At Day 28, mel-
atonin, however, only significantly enhanced the reduced
dendritic spine density in the penumbral brain, because
Melatonin enhanced postinsult neuroplasticity
(B)
(D)
contralateral diaschisis in dendritic spine density returned
to normal levels after a recovery period of 28 days in con-
trols and melatonin-treated animals. Interestingly, melato-
nin not only improved the MMP-9 activity and the
expressions of PSD-95 and GAP-43 proteins in the pen-
umbral brain at Day 7, but also enhanced the MMP-2
levels at the contralateral homotopic intact brain [9, 20,
25]. This melatonin-mediated increase in the MMP-2 levels
at Day 7 was correlated with the improved contralateral
neuroplasticity observed in melatonin-treated animals dur-
ing the early recovery period (Day 1 and Day 7), as
observed here and reported previously [8]. Accordingly,
the data supported that melatonin improved functional
deficits after stroke by providing direct neuroplasticity to
damaged brain tissues, and also by reducing remote brain
injury, or, using other intact brain to compensate for the
lost function during a course of healing.
Optimal sensorimotor function requires both anatomical
and functional linkages among the neuronal soma, axons,
dendritic spines, synapses as well as the ECM integrity [8,
26]. These structures and their axonal synaptodendritic
connections as well as the ECM integrity are involved in
the processing and integration of sensorimotor informa-
tion [8, 9, 26]. The marked and prolonged reductions of
the dendritic spine density in the peri-infarct penumbral
219
Juan et al.
(A)
(B) (C)
(D) (E)
region indicate that there may be relatively widespread and
persisting functional disturbances in this region [8, 48, 49].
The findings with improved neurobehavioral outcomes
observed with melatonin following a long-term recovery
further justify its ability to improve neuroplasticity at the
postsynaptic, axonal sprouting, dendritic spine density, and
the ECM levels following ischemic stroke, as observed here.
Previous findings that melatonin and its main brain metab-
olites worked well to inhibit nNOS activity in the ischemic
brain also validated its ability to enhance the PSD-95-medi-
ated synaptic plasticity observed here [36–39].
A variety of mechanisms including free radical forma-
tion, disordered neurotransmitter storage, uptake and
release, or a general stress response, may underlie the
220
Fig. 6. Photomicrographs of representative
dendritic segments from pyramidal neurons
sampled at regions corresponding to the
inner (layer III–IV) boundary zones of the
ischemic damage in sham-operated controls
(A, right upper panel), vehicle-injected
ischemic controls (A, left lower panel),
and melatonin-treated animals (A, right
lower panel) at Day 28 following stroke.
Scale bar = 40 lm; = 15 lm in the
magnification inset. The left upper panel
in A indicates the 4 nonoverlapping
sampling sites selected for dendritic spine
counting in the study. Vehicle-injected
ischemic controls had a decreased
dendritic spine density in the ischemic
core, inner and outer boundary zones of
ischemia (Fig. A), compared with sham-
operated controls. At Day 1 and Day 28,
melatonin-treated animals, however, had
a significantly improved density of
mushroom spines of the second- (A, B)
and third-order (A, C) dendrites at the
layers II–III, III–IV, and V–VI of the
ischemic hemisphere, relative to vehicle-
injected controls. In addition, we have
observed that, at Day 1 but not Day 28,
vehicle-injected ischemic controls had a
significantly lower density of the second-
and third-order basilar dendritic spines of
pyramidal neuron at the layer V–VI
region of the contralateral, intact brain
(D, E), compared with sham-operated
controls. Melatonin treatment could
effectively reverse the reduction in the
dendritic spines at Day 1 in the
contralateral, intact brain (D, E). n = 8
for each group; *P < 0.05 versus vehicle
data.
ipsilateral and remote excitability changes observed after
stroke; this needs further evaluation [8, 50]. One of the
most obvious explanations for the diaschisis changes in
dendritic spine density observed at Day 1 and Day 7 at
the contralateral homotopic brain is transhemispheric/
transcallosal deafferentation [51]. Whether the changes in
the MMP-2 level of the contralateral brain observed here
account for the severity of the contralateral diaschisis
needs further evaluation [9, 10]. Both the peri-infarct and
the remote changes in the dendritic spine density and other
molecular signaling cascades may turn out to be valuable
indicators of functional disturbances, which could become
targets of therapeutic interventions at the early and
subacute stage after stroke [8].

We have observed that, perhaps via different mecha-
nisms of action (e.g., MMP-9 and MMP-2, respectively),
melatonin overcame poststroke functional deficits by
improving in situ and remote neuroplasticity at both the
penumbral brain and those intact brain tissues distal from
the area of injury [8, 49, 51]. The present study and our
previous work also indicate that melatonin can modulate
MMPs and therefore exhibit neuroprotective or plasticity/
trophic actions depending on different time course and/or
molecular basis of ischemic stroke [14–16]. Thus, melato-
nin is a potent regulator of MMPs at various stages of
injury [14–16, 52]. Recent evidence also strongly support
that tight binding of melatonin in the active site is
involved in modulating the catalytic activity of MMP-9
[52]. Additional studies are, however, required to decipher
more molecular events through which melatonin regulates
the activity of the MMPs and leads to the improved long-
term neuroplasticity observed here.
It has been shown that neuroprotective effects of mela-
tonin against acute brain damage do not depend on mel-
atonin receptors in a mouse model of ischemic stroke
[53]. Further studies are needed to clarify whether mela-
tonin receptors are involved in the melatonin-mediated
neuronal growth/maturation, neuroplasticity, and neuro-
genesis during the subacute and long-term stages of
ischemic stroke as observed here and reported previously
[8, 53–55]. Additional studies are also needed to clarify
the melatonin’s beneficial role in the field of human
stroke, although a lack of economic benefits for pharma-
cological companies may actually hinder its further
clinical trials.
In summary, we demonstrate that administration with
melatonin induces in vitro and in vivo neuroprotection and
neuroplasticity through improving the PSD-95 and GAP-
43 protein expressions. In addition, we show that melato-
nin decreases poststroke diaschisis and enhances long-term
neuroplasticity probably via upregulating the MMP-2 and
MMP-9 in the nonischemic and ischemic brain, respec-
tively. As melatonin offers the advantage for improving
postischemic neuroplasticity for both the penumbral and
the contralateral, intact brain, it may be worthwhile fur-
ther investigating in the field of ischemic stroke in human
beings.
Acknowledgements
This research was supported by a grant from the National
Science Council of Taiwan (NSC No. 99-2314-B-006-022-
MY3). A.H.L. was the recipient of the summer student-
ship of Department of Occupational Safety and Heath,
Chang Jung University, Tainan, Taiwan. A.C.L. was the
recipient of the summer studentship of Chemistry and
Biochemistry Department, University of California Los
Angeles, CA, USA.
References
1. DEMAERSCHALK BM, YIP TR. Economic benefit of increasing
utilization of intravenous tissue plasminogen activator for
acute ischemic stroke in the United States. Stroke 2005;
36:2500–2503.
Melatonin enhanced postinsult neuroplasticity
2. ROMERO JR, BABIKIAN VL, KATZ DI et al. Neuroprotection
and stroke rehabilitation: modulation and enhancement of
recovery. Behav Neurol 2006; 17:17–24.
3. RAMIC M, EMERICK AJ, BOLLNOW MR et al. Axonal plasticity
is associated with motor recovery following amphetamine
treatment combined with rehabilitation after brain injury in
the adult rat. Brain Res 2006; 1111:176–186.
4. SIZONENKO SV, BEDNAREK N, GRESSENS P. Growth factors
and plasticity. Semin Fetal Neonatal Med 2007; 12:241–249.
5. LEGOS JJ, BARONE FC. Update on pharmacological strategies
for stroke: prevention, acute intervention and regeneration.
Curr Opin Investig Drugs 2003; 4:847–858.
6. HAAS S, WEIDNER N, WINKLER J. Adult stem cell therapy in
stroke. Curr Opin Neurol 2005; 18:59–64.
7. HAGGT T. From neurotransmitters to neurotrophic factors to
neurogenesis. Neuroscientist 2009; 15:20–27.
8. CHEN HY, HUNG YC, CHEN TY et al. Melatonin improves
presynaptic protein, SNAP-25, expression and dendritic spine
density and enhances functional and electrophysiological
recovery following transient focal cerebral ischemia in rats.
J Pineal Res 2009; 47:260–270.
9. MICHALUK P, WAWRZYNIAK M, ALOT P et al. Influence of
matrix metalloproteinase MMP-9 on dendritic spine morphol-
ogy. J Cell Sci 2011; 124:3369–3380.
10. FRAGKOULI A, PAPATHEODOROPOULOS C, GEORGOPOULOS S
et al. Enhanced neuronal plasticity and elevated endogenous
sAPPa levels in mice over-expressing MMP9. J Neurochem
2012; 121:239–251.
11. OLIVEIRA-SILVA P, JURGILAS PB, TRINDADE P et al. Matrix me-
talloproteinase-9 is involved in the development and plasticity
of retinotectal projections in rats. NeuroImmunoModulation
2007; 14:144–149.
12. HAYAKAWA K, NAKANO T, IRIE K et al. Inhibition of reactive
astrocytes with fluorocitrate retards neurovascular remodeling
and recovery after focal cerebral ischemia in mice. J Cereb
Blood Flow Metab 2010; 30:871–882.
13. BELL KF, BENT RJ, MEESE-TAMURI S et al. Calmodulin
kinase IV-dependent CREB activation is required for neuro-
protection via NMDA receptor-PSD95 disruption. J Neuro-
chem 2013; 126:274–287.
14. HUNG YC, CHEN TY, LEE EJ et al. Melatonin decreases
matrix metalloproteinase-9 activation and expression and
attenuates reperfusion-induced hemorrhage following tran-
sient focal cerebral ischemia in rats. J Pineal Res 2008;
45:459–461.
15. TAI SH, CHEN HY, LEE EJ et al. Melatonin inhibits postis-
chemic matrix metalloproteinase-9 (MMP-9) activation via
dual modulation of plasminogen/plasmin system and endoge-
nous MMP inhibitor in mice subjected to transient focal cere-
bral ischemia. J Pineal Res 2010; 49:332–341.
16. CHANG CC, TIEN CH, LEE EJ et al. Melatonin inhibits matrix
metalloproteinase-9 (MMP-9) activation in the lipopolysac-
charide (LPS)-stimulated RAW 264.7 and BV2 cells and a
mouse model of meningitis. J Pineal Res 2012; 53:188–197.
17. BOZDAGI O, NAGY V, KWEI KT et al. In vivo roles for matrix
metalloproteinase-9 in mature hippocampal synaptic physiol-
ogy and plasticity. J Neurophysiol 2007; 98:334–344.
18. BARKHO BZ, MUNOZ AE, LI X et al. Endogenous matrix
metalloproteinase (MMP)-3 and MMP-9 promote the differ-
entiation and migration of adult neural progenitor cells in
response to chemokines. Stem Cells 2008; 26:3139–3149.
19. WOJCIK L, SAWICKA A, RIVERA S et al. Neurogenesis in
gerbil hippocampus following brain ischemia: focus on the
221
Juan et al.
involvement of metalloproteinases. Acta Neurobiol Exp
(Wars) 2009; 69:52–61.
20. FAN J, VASUTA OC, ZHANG LY et al. N-methyl-D-aspartate
receptor subunit- and neuronal-type dependence of excitotox-
ic signaling through post-synaptic density 95. J Neurochem
2010; 115:1045–1056.
21. YAN BC, PARK JH, AHN JH et al. Postsynaptic density pro-
tein (PSD)-95 expression is markedly decreased in the hippo-
campal CA1 region after experimental ischemia-reperfusion
injury. J Neurol Sci 2013; 330:111–116.
22. SUN HS, DOUCETTE TA, LIU Y et al. Effectiveness of PSD95
inhibitors in permanent and transient focal ischemia in the
rat. Stroke 2008; 39:2544–2553.
23. ZHOU L, LI F, XU HB et al. Treatment of cerebral ischemia
by disrupting ischemia-induced interaction of nNOS with
PSD-95. Nat Med 2010; 16:1439–1443.
24. STROEMER RP, KENT TA, HULSEBOSCH CE. Neocortical
neural sprouting, synaptogenesis, and behavioral recovery
after neocortical infarction in rats. Stroke 1995; 26:2135–
2144.
25. BUFFO A, HOLTMAAT AJ, SAVIO T et al. Targeted overexpres-
sion of the neurite growth-associated protein B-50/GAP-43 in
cerebellar Purkinje cells induces sprouting after axotomy but
not axon regeneration into growth-permissive transplants.
J Neurosci 1997; 17:8778–8791.
26. STROEMER RP, KENT TA, HULSEBOSCH CE. Enhanced neocor-
tical neural sprouting, synaptogenesis, and behavioral recov-
ery with D-amphetamine therapy after neocortical infarction
in rats. Stroke 1998; 29:2381–2395.
27. LEE EJ, WU TS, LEE MY et al. Delayed treatment with
melatonin enhances electrophysiological recovery following
transient focal cerebral ischemia in rats. J Pineal Res 2004;
36:33–42.
28. REITER RJ, TAN DX, LEON J et al. When melatonin gets on
your nerves: its beneficial actions in experimental models of
stroke. Exp Biol Med 2005; 230:104–117.
29. KOH PO. Melatonin attenuates decrease of protein phospha-
tase 2A subunit B in ischemic brain injury. J Pineal Res 2012;
52:57–61.
30. TAN DX, CHEN LD, POEGGELER B et al. Melatonin: a potent,
endogenous hydroxyl radical scavenger. Endocr J 1993; 1:57–
60.
31. TAN DX, MANCHESTER LC, TERRON MP et al. One molecule,
many derivatives: a never-ending interaction of melatonin
with reactive oxygen and nitrogen species? J Pineal Res 2007;
42:28–42.
32. PEYROT F, DUCROCQ C. Potential role of tryptophan deriva-
tives in stress responses characterized by the generation of
reactive oxygen and nitrogen species. J Pineal Res 2008;
45:235–246.
33. GALANO A, TAN DX, REITER RJ. On the free radical scaveng-
ing activities of melatonin’s metabolites, AFMK and AMK.
J Pineal Res 2013; 54:245–257.
34. TAN DX, REITER RJ, MANCHESTER LC et al. Chemical and
physical properties and potential mechanisms: melatonin as a
broad spectrum antioxidant and free radical scavenger. Curr
Top Med Chem 2002; 2:181–197.
35. LEE EJ, LEE MY, CHEN HY et al. Melatonin attenuates gray
and white matter damage in a mouse model of transient focal
cerebral ischemia. J Pineal Res 2005; 38:42–52.
36. TAI SH, HUNG YC, LEE EJ et al. Melatonin protects against
transient focal cerebral ischemia in both reproductively active
and estrogen-deficient female rats: the impact of circulating
222
estrogen on its hormetic dose-response. J Pineal Res 2011;
50:292–303.
37. CHEN HY, CHEN TY, LEE MY et al. Melatonin decreases
neurovascular oxidative/nitrosative damage and protects
against early increases in the blood-brain barrier permeability
after transient focal cerebral ischemia in mice. J Pineal Res
2006; 41:175–182.
38. PEI Z, CHEUNG RT. Pretreatment with melatonin exerts anti-
inflammatory effects against ischemia/reperfusion injury in a
rat middle cerebral artery occlusion stroke model. J Pineal
Res 2004; 37:85–91.
39. TJONG YW, LI MF, HUNG MW et al. Melatonin ameliorates
hippocampal nitric oxide production and large conductance
calcium-activated potassium channel activity in chronic inter-
mittent hypoxia. J Pineal Res 2008; 44:234–243.
40. BETTAHI I, POZO D, OSUNA C et al. Melatonin reduces nitric
oxide synthase activity in rat hypothalamus. J Pineal Res
1996; 20:205–210.
41.  J, ESCAMES G, RODRIGUEZ MI et al. Inhibition of neuro-
LEON
nal nitric oxide synthase activity by N1-acetyl-5-methoxyky-
nuramine, a brain metabolite of melatonin. J Neurochem
2006; 98:2023–2033.
42. LEE WT, LIN MH, LEE EJ et al. Magnolol reduces gluta-
mate-induced neuronal excitotoxicity and protects against
permanent focal cerebral ischemia up to 4 hr. PLoS ONE
2012; 7:e39952.
43. LEE EJ, LEE MY, CHANG GL et al. Delayed treatment with
magnesium: reduction of brain infarction and improvement
of electrophysiological recovery following transient focal cere-
bral ischemia in rats. J Neurosurg 2005; 102:1085–1093.
44. CHEN TY, TAI SH, LEE EJ et al. Cinnamophilin offers pro-
longed neuroprotection against gray and white matter dam-
age and improves functional and electrophysiological
outcomes after transient focal cerebral ischemia. Crit Care
Med 2011; 39:1130–1137.
45. CHEN TY, LIN MH, LEE WT et al. Nicotinamide inhibits
nuclear factor-kappa B translocation after transient focal
cerebral ischemia. Crit Care Med 2012; 40:532–537.
46. YELESWARAM K, MCLAUGHLIN LG, KNIPE JO et al. Pharma-
cokinetics and oral bioavailability of exogenous melatonin in
preclinical animal models and clinical implications. J Pineal
Res 1997; 22:45–51.
47. LEE EJ, WU TS, CHANG GL et al. Delayed treatment with
nicotinamide inhibits brain energy depletion, improves cere-
bral microperfusion, reduces brain infarct volume, but does
not alter neurobehavioral outcome following permanent focal
cerebral ischemia in Sprague Dawley rats. Curr Neurovasc
Res 2006; 3:203–213.
48. CLARK WM, RINKER LG, LESSOV NS et al. Lack of interleu-
kin-6 expression is not protective against focal central
nervous system ischemia. Stroke 2000; 31:1715–1720.
49. HUNG YC, CHOU YS, CHANG CH et al. Early reperfusion
improves the recovery of contralateral electrophysiological di-
aschisis following focal cerebral ischemia in rats. Neurol Res
2010; 32:828–834.
50. NEUMANN-HAEFELIN T, WITTE OW. Periinfarct and remote
excitability changes after transient middle cerebral artery
occlusion. J Cereb Blood Flow Metab 2000; 20:45–52.
51. REDECKER C, WANG W, FRITSCHY JM et al. Widespread and
long-lasting alterations in GABA(A)-receptor subtypes after
focal cortical infarcts in rats: mediation by NMDA-depen-
dent processes. J Cereb Blood Flow Metab 2002; 22:1463–
1475.

52. RUDRA DS, PAL U, MAITI NC et al. Melatonin inhibits
matrix metalloproteinase-9 activity by binding to its active
site. J Pineal Res 2013; 54:398–405.
53. 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 enhanced postinsult neuroplasticity
54. YIP HK, CHANG YC, WALLACE CG et al. Melatonin treat-
ment improves adipose-derived mesenchymal stem cell ther-
apy for acute lung ischemia-reperfusion injury. J Pineal Res
2013; 54:207–221.
55. LIU J, SOMERA-MOLINA KC, HUDSON RL et al. Melatonin
potentiates running wheel-induced neurogenesis in the dentate
gyrus of adult C3H/HeN mice hippocampus. J Pineal Res
2013; 54:222–231.
223