J. Pineal Res.

2010; 48:185–193  2010 The Authors

Journal compilation  2010 John Wiley & Sons A/S
Doi:10.1111/j.1600-079X.2009.00740.x
Journal of Pineal Research

Melatonin protects periventricular white matter from damage due

to hypoxia
Molecular, Biological, Physiological and Clinical Aspects of Melatonin

Abstract: This study investigated the potential of melatonin in ameliorating C. Kaur, V. Sivakumar and
hypoxic damage to the periventricular white matter (PWM) in the neonatal E. A. Ling
brain. Vascular endothelial growth factor (VEGF), nitric oxide (NO), Department of Anatomy, Yong Loo Lin School
glutathione (GSH) and malondialdehyde (MDA) content in the PWM of 1- of Medicine, National University of Singapore,
day-old rats subjected to hypoxia for a period of 2 hr was examined. Vascular Singapore

endothelial growth factor, NO and MDA concentration was increased

whereas that of GSH was reduced after the hypoxic exposure. Additionally,
degenerating axons, apoptotic and necrotic cells and vacuolation of capillary
endothelial cells were observed in the PWM. The neighboring ependymal and
choroid plexus cells also appeared to undergo structural alterations. Increased
vascular permeability in the PWM of hypoxic rats was evidenced by the
leakage of rhodamine isothiocyanate (RhIC) which was taken up by the
amoeboid microglial cells. In vitro experiments showed increased apoptosis in Key words: glutathione, hypoxia, lipid
OLN-93 cells, an oligodendrocytic cell line, following hypoxic exposure. peroxidation, melatonin, neonatal brain,
periventricular white matter, vascular
Hypoxic rats treated with melatonin showed reduced VEGF, NO and MDA permeability
concentrations, increased GSH content and reduced RhIC leakage in the Address reprint requests to C. Kaur,
PWM. The ultrastructure of axons, endothelial, ependymal and choroid Department of Anatomy, Yong Loo Lin School
plexus epithelial cells appeared relatively normal in the hypoxic animals of Medicine, MD 10, 4 Medical Drive, National
University of Singapore, Singapore 117597.
treated with melatonin. The incidence of apoptotic OLN-93 cells was also E-mail: [email protected]
reduced with melatonin treatment. We suggest that the protective effects of
melatonin on various parameters in the PWM of hypoxic neonatal brains were Received August 21, 2009;
due to its antioxidant properties. accepted December 1, 2009.

vascular permeability [12, 13]. Melatonin reverses the

Introduction
inflammatory response by suppressing the production of
The periventricular white matter (PWM) peripheral to the inflammatory cytokines [14, 15]. Melatonin also suppressed
lateral ventricles is highly susceptible to damage during the production of NO in hypoxic brains [12, 16]. It is known
hypoxic–ischemic episodes in the developing brain. Oligo- to exert neuroprotection by preserving blood–brain barrier
dendrocytes die by necrosis or apoptosis resulting in delayed permeability and reducing oxidative stress in injuries of the
myelination. Reactive astrocytes and microglial cells, degen- central nervous system (CNS) [17]. It has also been reported
erating axons and subsequent diminution of PWM leading to have anti-apoptotic actions in many tissues including the
to ventriculomegaly have been reported [1–4]. Several brain [18–20]. Recent studies in the adult cerebellum have
factors including increased release of glutamate, inflamma- shown that melatonin preserved the ultrastructure of blood
tory cytokines, nitric oxide (NO) and other free radicals are vessels and dendrites in hypoxic injuries [12].
thought to be involved in such damage. Recently, we In view of the above, we examined whether melatonin
reported that increased production of vascular endothelial was beneficial in preserving the ultrastructure of axons and
growth factor (VEGF) may also add to the damage as it reducing VEGF and NO production in PWM of hypoxic
increases vascular permeability resulting in edema forma- neonatal rats. The antioxidant effect of melatonin was also
tion. Glutathione (GSH), a major cellular antioxidant evaluated by measuring lipid peroxidation (LPO) and GSH
protecting the cells against oxidative stress [5–7], has also levels in the PWM. Its effect on vascular permeability in the
been reported to be reduced significantly in the PWM in PWM was also examined.
hypoxic–ischemic injuries resulting in oligodendrocyte dam-
age [8, 9]. Besides the above, structural alterations in choroid
plexus epithelial cells [10] or the ependymal lining of lateral Materials and methods
ventricles resulting in seepage of cerebrospinal fluid in the
Animals and treatment
PWM may result in edema and damage.
Antioxidants such as melatonin have been reported to be Thirty-one 1-day-old rats were exposed to hypoxia by
effective in reducing cerebral edema formation in ischemic placing them in a chamber (Model: MCO 18M; Sanyo
animals [11] as it is known to reduce VEGF production and Biomedical Electrical Co, Ltd, Tokyo, Japan) filled with 5%

185
Kaur et al.

oxygen and 95% nitrogen for 2 hr. They were then allowed stable reaction product of NO. Homogenates from the
to recover under normoxic conditions for 3, 24 hr, 3 or PWM were prepared and nitrite colorimetric assays per-
7 days before killing. A second group of 31 rats was given formed according to the manufacturerÕs protocol.
melatonin before and after subjecting the rats to hypoxia.
A third group of 31 rats was kept outside the chamber
Lipid peroxidation
and the rats were used as age matched controls. All
experiments were approved by the Institutional Animal The content of LPO in the PWM tissue supernatant from
Care and Use Committee, National University of Singa- control and hypoxic and hypoxic + melatonin groups was
pore. All efforts were made to reduce the number of rats determined with Thiobarbituric Acid Reactive Substances
used and their suffering. assay kit (Cayman Chemical Company, Ann Arbor, MI,
In vitro experiments were performed to examine cell USA) according to the manufacturerÕs instructions. The
death in response to hypoxia in OLN-93 cells which show optical density was measured at 540 nm with precision
characteristics of immature oligodendrocytes. microplate reader (Molecular Devices Corporation,
Sunnyvale, CA, USA). The amount of LPO in the tissue
supernatant was calculated using Malondialdehyde (MDA)
Melatonin administration
standard curve.
To assess the effect of melatonin on ultrastructural changes,
LPO, GSH content, VEGF and NO production as well as
Glutathione assay
vascular permeability in the PWM, 31 rats were given
intraperitoneal injections of melatonin (Sigma-Aldrich, The content of GSH in the PWM tissue supernatant from
St Louis, MO, USA) dissolved in normal saline (10 mg/kg control and hypoxic and hypoxic + melatonin groups was
body weight) as described previously [21]. Each rat received determined with QuantiChrom Glutathione assay kit
three injections, the first immediately before exposure to (DIGT-250; BioAssay Systems, Hayward, CA, USA)
hypoxia, the second immediately after exposure to hypoxia according to the manufacturerÕs instructions. The optical
and the third injection at 1 hr after exposure to hypoxia. density was measured at 540 nm with precision microplate
Ultrastructure, LPO, GSH content, VEGF concentration, reader (Molecular Devices Corporation). The amount of
NO production and vascular permeability was determined GSH in the tissue supernatant was calculated using GSH
at 3, 24 hr, 3, and 7 days after hypoxic exposure. standard curve.

Measurement of VEGF, NO, lipid peroxidation and Apoptosis analysis with caspase-3 labeling
glutathione
OLN-93 cells, an oligodendrocytic cell line that was derived
The PWM above the lateral ventricles was removed from at a late stage of differentiation from spontaneously
the rats at 3, 24 hr, 3, and 7 days (n = 5 at each time point) transformed cells in a primary rat brain glial culture [22],
after hypoxic exposure, their corresponding controls were used for caspase-3 labeling. These cells show charac-
(n = 5 at each time point) and hypoxic rats treated with teristics of immature oligodendrocytes, and their antigenic
melatonin (n = 5 at 3, 24 hr, 3, and 7 days and each after and morphological properties resemble a highly pure
hypoxic exposure). They were homogenized with tissue culture of primary oligodendrocytes.
protein extraction reagent (Pierce Biotechnology, Inc., IL, OLN-93 cells were cultured at a density of 2 · 105 cells/
Rockford, USA) containing protease inhibitors, centri- well onto 24-well culture plates in DulbeccoÕs Modified
fuged at 14,000 g for 15 min and the supernatant was EagleÕs Medium, supplemented with 10% fetal bovine
collected to measure LPO and GSH content by Enzyme- serum mixture at 37C in a humidified incubator under 5%
Linked ImmunoSorbent Assay (ELISA), VEGF by CO2 and were divided into three groups. Cells in Group I
Enzyme Immunoassay (EIA) and NO by colorimetric assay were exposed to hypoxia in a chamber (Model: MCO-18M;
as described below. SANYO Electrical Co. Ltd, Tokyo, Japan) filled with a gas
mixture of 3% O2, 5% CO2 and 92% N2 at 37C for 4 hr.
Group II cells were incubated with melatonin at a concen-
VEGF concentration by EIA
tration of 1 lg/mL in the medium for 3 hr after hypoxic
The amount of VEGF (ng/mL) released in the PWM from exposure and Group III cells incubated in a chamber 95%
control, hypoxic and hypoxic + melatonin rats were air/5% CO2 were used as controls.
determined with a Chemikine VEGF EIA kit (Chemicon The OLN-93 cells were fixed in 4% paraformaldehyde in
International Inc., Temecula, CA, USA). Homogenates 0.1 m phosphate buffer (pH 7.4) for 20 min and blocked in
were prepared and EIA measurements were performed 3% normal goat serum in phosphate buffered saline (PBS)
according to the manufacturerÕs protocol. for 1 hr. They were then incubated at room temperature
with a cocktail mix of two primary antibodies, caspase-3
(1:200; Cell Signalling Technology, Inc., Danvers, MA,
Nitric oxide: colorimetric assay
USA) and O4 (1:100; Chemicon International, Inc), a
The total amount of NO in the PWM of control, hypoxic marker for immature oligodendrocytes. Subsequent anti-
and hypoxia + melatonin rats was assessed by the Griess body detection was carried out with a cocktail mix of two
reaction using a colorimetric assay kit (US Biological, secondary antibodies: Cy3-conjugated goat anti-rabbit
Swampscott, MA, USA) that detects nitrite (NO2)), a IgG and FITC-conjugated sheep anti-mouse IgG (1:100;

186
 Melatonin protects PWM against hypoxia

Sigma-Aldrich). After several washes with PBS, the sections brain was removed and coronal slices (approximately 1 mm
were mounted with fluorescent mounting medium (Dako- thickness) were cut. Blocks of PWM including the lateral
Cytomation, Glostrup, Denmark). The number of caspase- ventricle and choroid plexus were further cut from these
3 and O4 positive cells was counted in six randomly selected slices. Vibratome sections of 80–100 lm thickness were
microscopic fields from three wells in each group at ·60 prepared from these blocks and rinsed overnight in 0.1 m
magnification. The cells with O4 labeling overlapping with phosphate buffer. They were then postfixed for 2 hr in 1%
caspase-3 were counted as O4/caspase-3 positive cells, while osmium tetroxide, dehydrated and embedded in Araldite
those cells emitting only FITC fluorescence were counted as mixture. Ultrathin sections were cut and viewed in a Philips
O4/caspase-3 negative cells. The percentages of cells with CM 120 electron microscope (FEI, Eindhoven, The
positive expression for caspase-3 were calculated and Netherlands).
averaged.
Statistical analysis
Vascular permeability
For ELISA, EIA and NO colorimetric assays, data are
Three control rats, three hypoxic rats at 24 hr, and three reported as mean ± S.D. and analyzed by One-way
hypoxic rats treated with melatonin were each given an ANOVA followed by post hoc analysis using DunnetÕs test
intraperitoneal injection of rhodamine isothiocyanate (SPSS version 15.0 software, Chicago, IL, USA) to deter-
(RhIC) (5 lL of 1% RhIC/g body weight; Sigma, MO, mine the statistical significance of differences in values
USA) dissolved in normal saline. The rats were killed by between the control and hypoxic and hypoxic and
perfusion with 2% paraformaldehyde at 6 hr after RhIC hypoxia + melatonin groups. A value of P < 0.01 (*)
injection according to our earlier study [23]. The brains was considered significant.
were removed, frozen coronal sections of 40 lm thickness
at the level of optic chiasma were cut and rinsed in PBS.
Results
They were then incubated with FITC-conjugated lectin
(Lycopersicon esculentum, 1:100; Sigma-Aldrich), a marker We have shown in our earlier study [4] that VEGF and NO
for blood vessels and microglia. The sections were then production are increased in the PWM in a hypoxic insult to
washed in PBS and mounted with a fluorescent mounting the developing brain. The present study confirmed that
medium (DakoCytomation). Colocalization of RhIC and VEGF concentration increased significantly in the PWM at
lectin in amoeboid microglial cells (AMC) in the PWM was 3, 24 hr, 3 and 7 days in hypoxic rats when compared with
then examined by confocal microscopy (FV1000; Olympus the controls (Fig. 1A). It was suppressed significantly in
Company Pte Ltd, Tokyo, Japan). rats given melatonin treatment (Fig. 1A).
Nitric oxide levels in the PWM were significantly
increased following hypoxic exposure at 3, 24 hr, 3 and
Ultrastructural studies
7 days when compared with corresponding controls
The hypoxic rats (n = 4 at 3 and 24 hr) and their (Fig. 1B). With melatonin treatment, NO levels were
corresponding controls (n = 4 at each time point) along reduced significantly when compared with the hypoxic rats
with melatonin treated hypoxic rats (n = 4 at each time without melatonin treatment (Fig. 1B).
point) were perfused with a mixed aldehyde fixative The amount of LPO as manifested by MDA concentra-
composed of 2% paraformaldehyde and 3% glutaralde- tion in the PWM was increased significantly at 3 and 24 hr
hyde in 0.1 m phosphate buffer, pH 7.2. After perfusion, the or 3 days followed by a decrease at 7 days after the hypoxic

(A) Control
(B)
3.5 5 Control
Hypoxia
*# *#
No concentration (µ M)

4.5 Hypoxia
3 Hypoxia + Melatonin *#
released (ng/mL)
Amount of VEGF

4 Hypoxia + Melatonin
2.5 *# 3.5
*# *#
2 *# 3
2.5
* #
1.5 2
1 1.5
1
0.5
0.5
0 0
3 hr 24 hr 3d 7d 3 hr 24 hr 3d 7d

Fig. 1. Significant increase in vascular endothelial growth factor (VEGF) concentration and nitric oxide (NO) production between control and
hypoxic rats (*P < 0.01) or reductions in these parameters (#P < 0.01) after melatonin administration are evident in A (VEGF) and B (NO).

(A) (B)
2.5
MDA concentration (µ M)

Control 2.5 Control

MDA concentration (µ M)

Hypoxia
2
Hypoxia * 2 Hypoxia + Melatonin *
Fig. 2. Significant differences in malondi- * * # #
aldehyde (MDA) concentration between 1.5 * * 1.5
*# *
control and hypoxic rats are indicated by 1 1
*P < 0.01 in A. B shows significant
reductions (#P < 0.01) in levels of MDA 0.5 0.5
at 3, 24 hr and 3 days after melatonin 0 0
3 hr 24 hr 3d 7d
administration in hypoxic rats. 3 hr 24 hr 3d 7d

187
Kaur et al.

(A) (B) Control

Hypoxia
2 2 Hypoxia + Melatonin #

GSH concentration (µ M)
Control
GSH concentration (µ M)

Hypoxia #
* 1.5 # * Fig. 3. Significant differences in glutathi-
1.5 * * *
* one (GSH) concentration between control
1 1
and hypoxic rats are indicated by
*P < 0.01 in A. B shows a significant
0.5 0.5
increase (#P < 0.01) in levels of GSH at
0 0 3, 24 hr and 3 days after melatonin
3 hr 24 hr 3d 7d 3 hr 24 hr 3d 7d administration in hypoxic rats.

exposure as compared to the controls (Fig. 2A). At 3, 24 hr increased significantly at 3, 24 hr and 3 days when com-
and 3 days in hypoxic rats given melatonin treatment, the pared to those without melatonin treatment (Fig. 3B).
MDA concentration was reduced but remained unchanged There was no change in the GSH content with melatonin
at 7 days when compared to hypoxic rats not treated with treatment at 7 days in hypoxic rats.
melatonin (Fig. 2B). O4-immunoreactive OLN-93 cells in the control group
The GSH content in the PWM was significantly were not labeled by caspase-3 (Fig. 4A–C). Following
decreased at 3, 24 hr and 3 days in rats subjected to hypoxic exposure, however, many of the O4-immunoreac-
hypoxia as compared to the controls (Fig. 3A) but it tive cells showed caspase-3 labeling (Fig. 4D–F). O4
remained similar to the control levels at 7 days. With immunoreactive cells subjected to hypoxia and treated with
melatonin treatment of hypoxic rats, the GSH content melatonin, showed markedly reduced caspase-3 labeling

Control Hypoxia Hyp + Metn

(A) (D) (G)

20.0 µm 20.0 µm 20.0 µm

(B) (E) (H)

20.0 µm 20.0 µm 20.0 µm

(C) (F) (I)

20.0 µm 20.0 µm 20.0 µm Fig. 4. Confocal images show O4-labeled

OLN-93 cells (arrows) (A, D, G; green)
and caspase-3 positive (B, E, H; red)
(J) 70
(arrows) control, hypoxic and hypoxia +
60
* melatonin cells. The colocalized expres-
% of caspase-3
positive cells

50 sion of O4 and caspase-3 can be seen in C,

40 # F and I. Bar graphs in J show a significant
30 increase in percentage of caspase-3 posi-
20 tive cells (J) following hypoxic exposure
10 which decreased when the hypoxic cells
0 were treated with melatonin (*P < 0.01).
Control Hypoxia Hyp + Metn Scale bars: A–I = 20 lm.

188
 Melatonin protects PWM against hypoxia

(Fig. 4G–I). The percentage of O4/caspase-3 positive cells identified as the oligodendrocytes by us earlier [24]. The
following hypoxic exposure was significantly higher than necrotic and apoptotic cells were phagocytosed by the
the corresponding controls. It was however, reduced in AMC (Fig. 6B). Swollen and degenerating axons were also
hypoxia + melatonin cells (Fig. 4J). observed in the PWM after the hypoxic insult (Fig. 6C). In
Lectin labeled blood vessels and macrophagic cells called addition, intraventricular hemorrhage, changes in the
the AMC were observed in the PWM in the control, ependymal lining of the lateral ventricles such as vacuola-
hypoxic and hypoxia + melatonin rats (Fig. 5A, D, G). tion of the ependymal cells and widening of intercellular
Leakage of RhIC was evidenced in the PWM of control rats spaces were observed (Fig. 6E). The choroid plexus epithe-
as some of the AMC emitted weak fluorescence for RhIC lial cells showed vacuolation and large accumulation of
(Fig. 5A–C). Our earlier studies have shown that some of glycogen in their cytoplasm (Fig. 6G). Occurrence of edema
the AMC were fluorescent within a few hours after the in the PWM was evidenced by widened perivascular spaces
administration of the dye intraperitoneally in neonatal rats. and swollen astrocytes (Fig. 7A).
In hypoxic rats, increased leakage of RhIC was evidenced The endothelial cells of the blood vessels often showed
by bright immunofluorescence emitted by the blood vessels; the presence of multivesicular aggregations and vacuoles in
the leaked RhIC was internalized by the lectin labeled the cytoplasm in hypoxic rats (Fig 7B,C). With melatonin
AMC (Fig. 5D–F). Melatonin administration reduced treatment, the swelling of axons subsided (Fig 6D). The
RhIC leakage as fewer blood vessels were labeled by RhIC occurrence of apoptotic and necrotic cells was reduced. The
in melatonin administered rats (Fig. 5G–I). Lectin labeled vacuolation of ependymal, choroid plexus epithelial
AMC showed a weak RhIC labeling as compared to the (Fig. 6F,H) and endothelial cells (Fig. 7D) was not
hypoxic rats (Fig. 5G–I). observed. The glycogen content of the choroid plexus
Cells undergoing necrosis or apoptosis were a common epithelial cells was also reduced (Fig. 6H).
feature in the PWM of hypoxic rats (Fig. 6A); these were
Discussion
Control Hypoxia Hypoxia + Metn Hypoxia–ischemia is an important factor affecting the
(A) (D) (G) normal development and maturation of the CNS. The white
matter peripheral to the lateral ventricles called PWM, is
selectively vulnerable to damage in premature infants
[25–29]. The pathogenesis of PWM damage (PWMD) is
complex and multifactorial. Damage to axons and imma-
ture oligodendrocytes before the onset of myelination has
20.0 µm 20.0 µm 20.0 µm
been described as a characteristic feature of PWMD [4, 30,
(B) (E) (H) 31]. Excessive glutamate release, generation of free radicals
and inflammatory reactions are some of the factors which
have been reported to be involved in this damage.
The antioxidant properties of melatonin and its metab-
olites are due to their ability to scavenge free radicals and
their ability to induce expression of antioxidant enzymes
20.0 µm 20.0 µm 20.0 µm
[32–34]. Melatonin and its by products remove reactive
(C) (F) (I) oxygen and reactive nitrogen species including the hydroxyl
radical, hydrogen peroxide, singlet oxygen, hypochlorous
acid, peroxynitrite anion and/or peroxynitrous acid [35, 36].
The activity and expression of antioxidant enzymes, such as
superoxide dismutase, catalase, GSH peroxidase and GSH
20.0 µm 20.0 µm 20.0 µm
reductase has been shown to be increased by melatonin [37–
41], supporting its indirect antioxidant action. Further
Fig. 5. Confocal images showing the distribution of lectin (A, evidence of the antioxidant effect of melatonin is provided
green) and rhodamine isothiocyanate (RhIC; B, red) labeled blood by its ability to reduce LPO [42], a degradative phenom-
vessels and amoeboid microglia (arrows) in the periventricular enon involved in the pathogenesis of many diseases. Lipid
white matter (PWM) of a 2-day-old control rat at 6 hr after RhIC peroxidation contributes significantly to PWM injury [29].
administration. Amoeboid microglia (arrows) showing weak RhIC
Aldehydes such as MDA and 4-hydroxy-2-nonenal (4-
labeling are seen in B. Co-localized labeling of lectin stained
amoeboid microglia with RhIC (arrows) is detected in C. D, E and HNE) are formed during LPO and cause tissue damage by
F show lectin (D, green), RhIC (E, red) and co-localized labeling of augmenting free radical events [43]. Increased levels of
lectin and RhIC (F) in the PWM of a rat at 24 hr after hypoxic MDA and 4-HNE were reported in the white matter
exposure. Note lectin and RhIC co-labeled blood vessels (arrow- following a hypoxic–ischemic injury [44] and were reported
heads) and amoeboid microglia (arrows). G, H and I show lectin to be toxic to axons and oligodendrocytes [45]. Melatonin
(G, green), RhIC (H, red) and co-localized labeling of lectin and has been reported to directly interact with MDA and
RhIC (I) in blood vessels (arrowheads) and amoeboid microglia
protect against cellular damage induced by oxidative stress
(arrows) in hypoxic rats after melatonin administration. Note the
reduced RhIC labeling of blood vessels and amoeboid microglia [46]. The present study showed that MDA content in the
after melatonin administration in hypoxic rats. Scale bars: A– developing white matter was increased significantly up to
I = 20 lm. 3 days followed by a decrease at 7 days following the

189
Kaur et al.

(A) (B)

(C) (D)

(E) (F)

Fig. 6. Electron micrographs showing an

apoptotic cell (asterisk, A), an amoeboid
microglial cell (AMC) which has phagocy-
tosed an apoptotic cell (asterisk, B) and
swollen axons (asterisks, C) in the peri-
ventricular white matter of a rat at 24 hr
after hypoxic exposure. E shows dilated
intercellular spaces (asterisks) between the
(G) (H)
ependymal cells (EP) lining the lateral
ventricle (Lv). The choroid plexus epithelial
cells (EC) with large glycogen accumula-
tions (asterisks) and vacuoles (V) are seen in
G. At 24 hr after melatonin administration
in a hypoxic rat, the swelling of axons
(asterisks, D), dilation of intercellular
spaces between the EP lining the lateral
ventricle (asterisks, F) and glycogen accu-
mulation in the choroid plexus epithelial
cells (H) has subsided. Scale bars: A,
B = 1 lm; C = 0.5 lm; D = 0.2 lm; E,
F, G = 2 lm; H = 5 lm.

hypoxic exposure. The decrease in the MDA content at sis by blocking a glutamate–cysteine anti-porter in the
7 days could be due to the antioxidant action of GSH, the plasma membrane [48]. Glutathione depletion in oligoden-
levels of which were comparable to the controls at this drocytes leading to their death by free radical mediated
delayed time interval. Melatonin administration was effec- mechanisms has been described to be caused by glutamate
tive in reducing the hypoxia induced MDA content. [49]. Developing oligodendrocytes were reported to be
Hypoxia significantly reduced the GSH content in the vulnerable to intracellular GSH depletion [8, 9] in hypoxic–
PWM. Glutathione constitutes an important defense ischemic conditions, the mechanism for this vulnerability
against oxidative stress in the brain and is reported to be being intracellular accumulation of reactive oxygen species
used clinically for neuropsychiatric conditions and excito- (ROS) [8, 49, 50]. We have shown earlier that melatonin
toxic cell damage [47]. It protects the cells against oxidative treatment reduced glutamate concentration in the develop-
stress [5–7]. Hypoxia is known to cause accumulation of ing white matter [51]. The present study showed that
extracellular glutamate in the white matter and it has been melatonin increased GSH content in the white matter
reported that extracellular glutamate inhibits GSH synthe- following the hypoxic exposure. This is supported by results

190
 Melatonin protects PWM against hypoxia

(A) (B) could be production of large amounts of NO generated

from the inducible nitric oxide synthase (iNOS). We have
shown earlier that AMC, the active phagocytes in the
PWM, express intense iNOS immunoreactivity following a
hypoxic injury [4]. Nitric oxide interacts with ROS such as
O2) leading to the generation of peroxynitrite (ONOO)),
which is believed to be even more destructive than its
precursor NO [56, 57]. As mentioned earlier, melatonin can
2 µm 2 µm remove reactive oxygen and reactive nitrogen species
including ONOO) [35, 36]. It has been shown to down
(C) (D) regulate prooxidant enzymes such as NO synthases [58].
The antioxidant actions of melatonin may have helped to
preserve the ultrastructure of axons, blood vessels, epen-
dymal cells and choroid plexus epithelial cells in the present
study.
The present study has shown that VEGF, NO and MDA
concentration were increased whereas GSH content was
0.5 µm 2 µm
reduced in the PWM of neonatal rats following a hypoxic
exposure. Degenerating axons, apoptotic and necrotic cells,
Fig. 7. Electron micrographs showing a swollen astrocyte (As) vacuolation of capillary endothelial, ependymal and cho-
with watery cytoplasm in A and a blood vessel (BV) with a large roid plexus cells were also observed. Increased vascular
vacuole (arrow) in the endothelial cell in B in the periventricular permeability was evidenced by the leakage of RhIC in the
white matter of a rat at 3 hr after the hypoxic exposure. C shows
enlarged view of the endothelial cell (En) with the vacuole (V). At
PWM of hypoxic rats and the leaked RhIC was taken up by
24 hr after melatonin administration in a hypoxic rat, vacuolation the AMC. Melatonin administration decreased VEGF, NO
of the vascular endothelial cells (arrow, D) could not be seen. Scale and MDA concentration, increased GSH content and
bars: A,B,D = 2 lm; C = 0.5 lm. reduced RhIC leakage in the PWM of hypoxic rats. The
ultrastructure of axons, endothelial, ependymal and cho-
of earlier studies which reported that physiological con- roid plexus cells was also preserved in the hypoxic animals
centrations of melatonin induced GSH peroxidase and treated with melatonin. In view of the above, it is suggested
superoxide dismutases at the mRNA level in neuronal cell that melatonin may be of therapeutic value in ameliorating
lines [52]. It also reduced the expression of caspase-3 by O4- hypoxic damage to the developing white matter.
immunoreactive OLN-93 cells suggesting that it is effective
in reducing cell death by apoptosis.
Acknowledgements
Hypoxia is known to increase vascular permeability by
increasing the production of VEGF. Increased levels of This study was supported by research grants from the
VEGF in the cerebrospinal fluid and serum of newborns National University of Singapore (R181–000–065–112 and
with hypoxic ischemic encephalopathy have been reported R-181–000–098–112) and National Medical Research
[53]. Increased tissue concentration of VEGF in PWM was Council of Singapore (R-181–000–120–213). The technical
reported recently by us in the neonatal brain [4]. Increased assistance Ms Y.G. Chan is gratefully acknowledged. There
leakage of intraperitoneally administered fluorescent tracer, is no conflict of interest among the authors.
RhIC was detected in the PWM in the neonatal brain
following hypoxic injury.
References
Along with VEGF, NO production in the PWM was also
increased following the hypoxic exposure. Increased expres- 1. Takashima S, Armstrong DL, Becker LE. Subcortical
sion of endothelial nitric oxide synthase (eNOS) in the leukomalacia. Relationship to development of the cerebral
blood vessels was observed in hypoxic injury [4]. Vascular sulcus and its vascular supply. Arch Neurol 1978; 35:470–472.
endothelial growth factor has been shown to induce 2. Inder TE, Huppi PS, Warfield S et al. Periventricular white
increased vascular permeability via synthesis or release of matter injury in the premature infant is followed by reduced
NO [54] predominantly derived from eNOS [55]. Melatonin cerebral cortical gray matter volume at term. Ann Neurol
suppressed VEGF and NO production in the PWM and 1999; 46:755–760.
also ameliorated the leakage of RhIC suggesting its 3. Volpe JJ. Neurobiology of periventricular leukomalacia in the
effectiveness in reducing vascular permeability. premature infant. Pediatr Res 2001; 50:553–562.
4. Kaur C, Sivakumar V, Ang LS, Sundaresan A. Hypoxic
A large number of degenerating axons were observed in
damage to the periventricular white matter in neonatal brain:
the vicinity of the blood vessels in the PWM of neonatal
role of vascular endothelial growth factor, nitric oxide and
rats subjected to hypoxia suggesting that increased leakage
excitotoxicity. J Neurochem 2006; 98:1200–1216.
of serum derived substances may be involved in such
5. Meister A, Anderson ME Glutathione. Annu Rev Biochem
degeneration [4]. Endothelial cells of the blood vessels, 1983; 52:711–760.
ependymal and choroid plexus epithelial cells also showed 6. Mizui T, Kinouchi H, Chan PH. Depletion of brain gluta-
degenerative changes such as the presence of large vacuoles thione by buthionine sulfoximine enhances cerebral ischemic
in the cytoplasm. Besides increased vascular permeability, injury in rats. Am J Physiol 1992; 262:H313–H317.
another factor responsible for the degenerative changes

191
Kaur et al.

7. Bobyn PJ, Franklin JL, Wall CM, Thornhill JA, Juur- 24. Deng Y, Lu J, Sivakumar V, Ling EA, Kaur C. Amoeboid
link BH, Paterson PG. The effects of dietary sulfur amino microglia in the periventricular white matter induces oligo-
acid deficiency on rat brain glutathione concentration and dendrocyte damage through expression of proinflammatory
neural damage in global hemispheric hypoxia–ischemia. Nutr cytokines via MAP kinase signalling pathway in hypoxic
Neurosci 2002; 5:407–416. neonatal rats. Brain Pathol 2008; 18:387–400.
8. Back SA, Gan X, Li Y, Rosenberg PA, Volpe JJ. Matura- 25. Johnston MV. Hypoxic and ischemic disorders of infants and
tion-dependent vulnerability of oligodendrocytes to oxidative children. Lecture for 38th meeting of Japanese Society of Child
stress-induced death caused by glutathione depletion. J Neu- Neurology, Tokyo, Japan, July 1996. Brain Dev 1997; 19:235–
rosci 1998; 18:6241–6253. 239.
9. Back SA, Han BH, Luo NL et al. Selective vulnerability of 26. Mcquillen PS, Ferriero DM. Selective vulnerability in the
late oligodendrocyte progenitors to hypoxia–ischemia. J Neu- developing central nervous system. Pediatr Neurol 2004;
rosci 2002; 22:455–463. 30:227–235.
10. Sivakumar V, Lu J, Ling EA, Kaur C. Vascular endothelial 27. Rezaie P, Dean A. Periventricular leukomalacia, inflamma-
growth factor and nitric oxide production in response to hy- tion and white matter lesions within the developing nervous
poxia in the choroid plexus in neonatal brain. Brain Pathol system. Neuropathology 2002; 22:106–132.
2008; 18:71–85. 28. Follett PL, Deng W, Dai W et al. Glutamate receptor-
11. Kondoh T, Uneyama H, Nishino H, Torii K. Melatonin mediated oligodendrocyte toxicity in periventricular leuko-
reduces cerebral edema formation caused by transient fore- malacia: a protective role for topiramate. J Neurosci 2004;
brain ischemia in rats. Life Sci 2002; 72:583–590. 24:4412–4420.
12. Kaur C, Sivakumar V, Zhang Y, Ling EA. Hypoxia-in- 29. Folkerth RD. Periventricular leukomalacia: overview and
duced astrocytic reaction and increased vascular permeability recent findings. Pediatr Dev Pathol 2006; 9:3–13.
in the rat cerebellum. Glia 2006; 54:826–839. 30. Dammann O, Hagberg H, Leviton A. Is periventricular
13. Kaur C, Ling EA. Antioxidants and neuroprotection in the leukomalacia an axonopathy as well as an oligopathy? Pediatr
adult and developing central nervous system. Curr Med Chem Res 2001; 49:453–457.
2008; 15:3068–3080. 31. Ness JK, Romanko MJ, Rothstein RP, Wood TL, Levison
14. Pei Z, Cheung RT. Pretreatment with melatonin exerts anti- SW. Perinatal hypoxia–ischemia induces apoptotic and ex-
inflammatory effects against ischemia/reperfusion injury in a citotoxic death of periventricular white matter oligodendrocyte
rat middle cerebral artery occlusion stroke model. J Pineal Res progenitors. Dev Neurosci 2001; 23:203–208.
2004; 37:85–91. 32. Tan DX, Chen LD, Poeggler B, Manchester LC, Reiter
15. Lee MY, Kuan YH, Chen HY et al. Intravenous adminis- RJ. Melatonin: a potent endogenous hydroxyl radical
tration of melatonin reduces the intracerebral cellular inflam- scavenger. Endocrine J 1993; 1:57–60.
matory response following transient focal cerebral ischemia in 33. Peyrot F, Ducrocq C. Potential role of tryptophan deriva-
rats. J Pineal Res 2007; 42:297–309. tives in stress responses characterized by the generation of
16. Cuzzocrea S, Costantino G, Gitto E et al. Protective effects reactive oxygen and nitrogen species. J Pineal Res 2008;
of melatonin in ischemic brain injury. J Pineal Res 2000; 45:235–246.
29:217–227. 34. Manda K, Ueno M, Anzai K. Space radiation-induced
17. Erşahin M, Sehirli O, Toklu HZ et al. Melatonin improves inhibition of neurogenesis in the hippocampal dentate gyrus
cardiovascular function and ameliorates renal, cardiac and and memory impairment in mice: ameliorative potential of the
cerebral damage in rats with renovascular hypertension. melatonin metabolite, AFMK. J Pineal Res 2008; 45:430–438.
J Pineal Res 2009; 47:97–106. 35. Reiter RJ, Tan DX, Manchester LC, Qi W. Biochemical
18. Caballero B, Vega-Naredo I, Sierra V et al. Melatonin reactivity of melatonin with reactive oxygen and nitrogen
alters cell death processes in response to age-related oxidative species: a review of the evidence. Cell Biochem Biophys 2001;
stress in the brain of senescence-accelerated mice. J Pineal Res 34:237–256.
2009; 46:106–114. 36. Reiter RJ, Paredes SD, Manchester LC, Tan DX.
19. Luchetti F, Betti M, Canonico B et al. ERK MAPK acti- Reducing oxidative/nitrosative stress: a newly-discovered gene
vation mediates the antiapoptotic signaling of melatonin in for melatonin. Crit Rev Biochem Mol Biol 2009; 44:175–200.
UVB-stressed U937 cells. Free Radic Biol Med 2009; 46:339– 37. Pablos MI, Agapito MT, Gutierrez R et al. Melatonin
351. stimulates the activity of the detoxifying enzyme glutathione
20. Maity P, Bindu S, Dey S et al. Melatonin reduces indometh- peroxidase in several tissues of chicks. J Pineal Res 1995;
acin-induced gastric mucosal cell apoptosis by preventing 19:111–115.
mitochondrial oxidative stress and the activation of mitochon- 38. Ozturk G, Coskun S, Erbas D, Hasanoglu E. The effect of
drial pathway of apoptosis. J Pineal Res 2009; 46:314–323. melatonin on liver superoxide dismutase activity, serum nitrate
21. Kaur C, Sivakumar V, Ling EA. Expression of tranferrin and thyroid hormone levels. Jpn J Physiol 2000; 50:149–153.
receptors in the pineal gland of postnatal and adult rats and its 39. Rodriguez C, Mayo JC, Sainz RM et al. Regulation of
alteration in hypoxia and melatonin treatment. Glia 2007; antioxidant enzymes: a significant role for melatonin. J Pineal
55:263–273. Res 2004; 36:1–9.
22. Richter-Landsberg C, Heinrich M. OLN-93: a new per- 40. Reiter RJ, Tan DX, Gitto E et al. Pharmacological utility of
manent oligodendroglia cell line derived from primary rat melatonin in reducing oxidative cellular and molecular dam-
brain glial cultures. J Neurosci Res 1996; 45:161–173. age. Pol J Pharmacol 2004; 56:159–170.
23. Kaur C, Sivakumar V, Lu J, Ling EA. Melatonin attenuates 41. Subramanian P, Mirunalini S, Pandi-Perumal SR,
hypoxia-induced ultrastructural changes and increased vascu- Trakht I, Cardinali DP. Melatonin treatment improves the
lar permeability in the developing hippocampus. Brain Pathol antioxidant status and decreases lipid content in brain and liver
2008; 18:533–547. of rats. Eur J Pharmacol 2007; 571:116–119.

192
 Melatonin protects PWM against hypoxia

42. Manda K, Ueno M, Anzai K. Melatonin mitigates oxidative free radical scavengers and by a diffusible glial factor. J Neu-
damage and apoptosis in mouse cerebellum induced by high- rochem 1996; 67:566–573.
LET 35Fe particle irradiation. J Pineal Res 2008; 44:189–196. 51. Sivakumar V, Ling EA, Lu J, Kaur C. Role of glutamate
43. Esterbauer H, Schaur RJ, Zollner H. Chemistry and and its receptors and insulin-like growth factors in hypoxia
biochemistry of 4-hydroxynonenal, malonaldehyde and related induced periventricular white matter injury. Glia 2009 Doi:
aldehydes. Free Radic Biol Med 1991; 11:81–128. 10.1002/glia.20940.
44. Lin S, Rhodes PG, Lei M, Zhang F, Cai Z. alpha-Phenyl-n- 52. Mayo JC, Sainz RM, Antolin I, Herrera F, Martin V,
tert-butyl-nitrone attenuates hypoxic–ischemic white matter Rodriguez C. Melatonin regulation of antioxidant enzyme
injury in the neonatal rat brain. Brain Res 2004; 1007:132–141. gene expression. Cell Mol Life Sci 2002; 59:1706–1713.
45. Mccracken E, Valeriani V, Simpson C, Jover T, McCul- 53. Ergenekon E, Gücüyener K, Erbas D, Aral S, Koç E,
loch J, Dewar D. The lipid peroxidation by-product 4-hy- Atalay Y. Cerebrospinal fluid and serum vascular endothelial
droxynonenal is toxic to axons and oligodendrocytes. J Cereb growth factor and nitric oxide levels in newborns with hypoxic
Blood Flow Metab 2000; 20:1529–1536. ischemic encephalopathy. Brain Dev 2004; 26:283–286.
46. Li J, Baud O, Vartanian T, Volpe JJ, Rosenberg PA. 54. Mayhan WG. VEGF increases permeability of the blood–
Peroxynitrite generated by inducible nitric oxide synthase and brain barrier via a nitric oxide synthase/cGMP-dependent
NADPH oxidase mediates microglial toxicity to oligodendro- pathway. Am J Physiol 1999; 76:C1148–C1153.
cytes. Proc Natl Acad Sci USA 2005; 102:9936–9941. 55. Fukumura D, Gohongi T, Kadambi A et al. Predominant
47. Jana¢Ky R, Cruz-Aguado R, Oja SS, Shaw CA. Glutathi- role of endothelial nitric oxide synthase in vascular endothelial
one in the nervous system: roles in neural function in health growth factor-induced angiogenesis and vascular permeability.
and implicationsfor neurological disease. In: Handbook of Proc Natl Acad Sci USA 2001; 98:2604–2609.
Neurochemistry and Molecular Neurolobiology, Oja SS, 56. Hoog N, Darley-Usmar VN, Wilson MT, Moncada S.
Schousboe A, Saransaari P, eds., 3rd edn, vol 6. Springer, New Production of hydroxyl radicals from the simultaneous gen-
York, pp. 347–399. eration of superoxide and nitric oxide. Biochemistry 1992;
48. Tan S, Sagara Y, Liu Y, Maher P, Schubert D. The reg- 281:419.
ulation of reactive oxygen species production during pro- 57. Radi R, Beckmann JS, Bush KM, Freeman BA.
grammed cell death. J Cell Biol 1998; 141:1423–1432. Peroxynitrite oxidation of sulfhydryls. The cytotoxic
49. Oka A, Belliveau MJ, Rosenberg PA, Volpe JJ. Vulnera- potential of superoxide and nitric oxide. J Biol Chem 1991;
bility of oligodendroglia to glutamate: pharmacology, mecha- 266:4244–4250.
nisms, and prevention. J Neurosci 1993; 13:1441–1453. 58. Reiter RJ, Garcia JJ, Pie J. Oxidative toxicity in models of
50. Yonezawa M, Back SA, Gan X, Rosenberg PA, Volpe JJ. neurodegeneration: responses to melatonin. Restor Neurol
Cystine deprivation induces oligodendroglial death: rescue by Neurosci 1998; 12:135–142.

193