Neuroprotection by two polyphenols following excitotoxicity and experimental ischemia Miroslav Gottlieb, a,1,2 Rocı ´o Leal-Campanario, c,1 Marı ´a Rosario Campos-Esparza, a Marı ´a Victoria Sa ´nchez-Go ´ mez, a Elena Alberdi, a,b Amaia Arranz, a Jose ´ Marı ´a Delgado-Garcı ´a, c Agne `s Gruart, c and Carlos Matute a,b, * a Departamento de Neurociencias, Universidad del Paı ´s Vasco, E-48940 Leioa, Vizcaya, Spain b Neurotek, Parque Tecnolo ´gico de Bizkaia, 48170-Zamudio, Spain c Divisio ´n de Neurociencias, Universidad Pablo de Olavide, Carretera de Utrera Km. 1, E-41013 Sevilla, Spain Received 30 November 2005; revised 14 March 2006; accepted 31 March 2006 Available online 27 June 2006 Brain ischemia induces neuronal loss which is caused in part by excitotoxicity and free radical formation. Here, we report that mangiferin and morin, two antioxidant polyphenols, are neuroprotec- tive in both in vitro and in vivo models of ischemia. Cell death caused by glutamate in neuronal cultures was decreased in the presence of submicromolar concentrations of mangiferin or morin which in turn attenuated receptor-mediated calcium influx, oxidative stress as well as apoptosis. In addition, both antioxidants diminished the generation of free radicals and neuronal loss in the hippocampal CA1 region due to transient forebrain ischemia in rats when administered after the insult. Importantly, neuroprotection by these antioxidants was functionally relevant since treated-ischemic rats performed significantly better in three hippocampal-dependent behavioral tests. Together, these results indicate that mangiferin and morin have potent neuroprotectant activity which may be of therapeutic value for the treatment of acute neuronal damage and disability. D 2006 Elsevier Inc. All rights reserved. Keywords: Classical conditioning; Instrumental conditioning; Mangiferin; Morin; Neuronal death; Spatial orientation Introduction The principal pathophysiological processes in brain ischemia involve energy failure, loss of cell ion homeostasis, acidosis, increased intracellular calcium, excitotoxicity and free-radical- mediated toxicity. Transient forebrain ischemia, an animal model of cardiac arrest, induces molecular alterations which cause neuronal hyperexcitability and cell death in vulnerable regions of the brain such as the hippocampal CA1 area (Kirino, 1982; Kirino et al., 1984; Pulsinelli et al., 1982; Choi, 1996; Luhmann, 1996). Ischemia results in loss of ATP which impairs the function of glutamate transporters that normally remove released glutamate from the synaptic cleft (Conti and Weinberg, 1999). The resulting rise of glutamate in the extracellular space leads to excessive activation of glutamate receptors and pathological elevations in the levels of intracellular calcium which ultimately kill neurons and glial cells (Choi, 1996; Matute et al., 2002). However, glutamate receptor antagonists have not been effective in clinical trials of brain ischemia (Lee et al., 1999; Ikonomidou and Turski, 2002). Since both excitotoxicity and ischemia/reperfusion insults generate oxidative stress, it is conceivable that the administration of antioxidants may limit oxidative damage and ameliorate disease progression. Indeed, several exogenously administered antioxi- dants have been reported to be neuroprotective in experimental models of cerebral ischemia, but most of them did not show beneficial effects in clinical trials (Gilgun-Sherki et al., 2002). The failure to translate experimental results with antioxidants into efficient treatments for stroke may be due, at least in part, to the inadequate penetration of selected drugs into salvageable portions of the ischemic zone and hindered by an insufficient characteriza- tion of the alteration of cognitive functions in disease animal models. In addition, the therapeutic potential of new antioxidants, especially those of natural origin, needs to be assayed. In this regard, flavonoids and other polyphenol antioxidants present as bioactive molecules in vegetables, fruit and red wine have been shown to be potentially beneficial in neurodegenerative diseases associated with oxidative stress (Mandel et al., 2004). Here, we have assayed the neuroprotective efficacy of two natural polyphe- nolic antioxidants, mangiferin and morin, which ameliorate damage caused by experimental insults, including ischemia, to 0969-9961/$ - see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.nbd.2006.03.017 * Corresponding author. Fax: +34 94 6013400. E-mail address: [email protected](C. Matute). 1 Contributed equally to this work. 2 Permanent address: Institute of Neurobiology, Slovak Academy of Sciences, Soltesovej 6, 04001 Kosice, Slovak Republic. Available online on ScienceDirect (www.sciencedirect.com). www.elsevier.com/locate/ynbdi Neurobiology of Disease 23 (2006) 374 – 386
13
Embed
Neuroprotection by two polyphenols following excitotoxicity and experimental ischemia … · Neuroprotection by two polyphenols following excitotoxicity and experimental ischemia
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
www.elsevier.com/locate/ynbdi
Neurobiology of Disease 23 (2006) 374 – 386
Neuroprotection by two polyphenols following excitotoxicity and
experimental ischemia
Miroslav Gottlieb,a,1,2 Rocıo Leal-Campanario,c,1 Marıa Rosario Campos-Esparza,a
Marıa Victoria Sanchez-Gomez,a Elena Alberdi,a,b Amaia Arranz,a Jose Marıa Delgado-Garcıa,c
Agnes Gruart,c and Carlos Matutea,b,*
aDepartamento de Neurociencias, Universidad del Paıs Vasco, E-48940 Leioa, Vizcaya, SpainbNeurotek, Parque Tecnologico de Bizkaia, 48170-Zamudio, SpaincDivision de Neurociencias, Universidad Pablo de Olavide, Carretera de Utrera Km. 1, E-41013 Sevilla, Spain
Received 30 November 2005; revised 14 March 2006; accepted 31 March 2006
Available online 27 June 2006
Brain ischemia induces neuronal loss which is caused in part by
excitotoxicity and free radical formation. Here, we report that
mangiferin and morin, two antioxidant polyphenols, are neuroprotec-
tive in both in vitro and in vivo models of ischemia. Cell death caused
by glutamate in neuronal cultures was decreased in the presence of
submicromolar concentrations of mangiferin or morin which in turn
attenuated receptor-mediated calcium influx, oxidative stress as well as
apoptosis. In addition, both antioxidants diminished the generation of
free radicals and neuronal loss in the hippocampal CA1 region due to
transient forebrain ischemia in rats when administered after the insult.
Importantly, neuroprotection by these antioxidants was functionally
relevant since treated-ischemic rats performed significantly better in
three hippocampal-dependent behavioral tests. Together, these results
indicate that mangiferin and morin have potent neuroprotectant
activity which may be of therapeutic value for the treatment of acute
neuronal damage and disability.D 2006 Elsevier Inc. All rights reserved.
0.001 as compared to neurons treated with agonist alone). Antioxidants
were added during agonist exposure and left in the medium for 3 h until
cell viability was measured. Cell death in panel B is plotted vs. control,
glutamate-treated cultures. Values in panels A and B are illustrated as
mean T SEM of quadruplicates from 3 to 4 different experiments.
Results
Mangiferin and morin attenuate glutamate-receptor-mediated
excitotoxicity in cortical neurons in vitro
We initially elaborated the dose–response curve of glutamate
excitotoxicity in cultures of neurons derived from the cerebral cortex
of 18-day-old rat embryos. Cells were exposed to glutamate or
NMDA (1–1000 AM) for 10 min, and viability was assayed 3 h later
with the MTT method. Cell death was concentration-dependent,
with an EC50 for glutamate and NMDA of 35 AMand 53 AM (n = 4),
respectively, and reached a maximum at 1 mM (Fig. 1A). Toxicity
was prevented when glutamate was applied in the presence of MK-
801 (20 AM), an antagonist of NMDA receptors (Fig. 1A). In
contrast, CNQX andGYKI 53655 (both at 100 AM), two antagonists
of non-NMDA ionotropic receptors, did not have any protective
effect. These data confirm that NMDA receptor stimulation accounts
for the majority of glutamate excitotoxicity in these neurons, as
described previously (Sattler and Tymianski, 2001).
To test the efficacy of mangiferin and morin to attenuate
excitotoxicity in cortical neurons, we added them at 1–1000 nM to
the culture medium during glutamate (50 AM) exposure and left
them there for 3 h until the end of the experiment. Both
antioxidants substantially reduced cell death as compared with
cultures exposed to glutamate in the presence of vehicle (Fig. 1B).
Fig. 2. Ca2+ overload, oxidative stress and apoptosis caused by excitotoxic insults were reduced by mangiferin and morin. (A) Traces (left) illustrate the time
course of the [Ca2+]i increase (mean T SEM) upon incubation with agonist. (B and C) Histogram values depict the decrease in the [Ca2+]i peak response to
NMDA and glutamate induced by morin and mangiferin (**P < 0.01; ***P < 0.001, n = 19–26 neurons). (D) A 10 min exposure of cortical neurons in vitro to
glutamate (50 AM) induced the generation of radical oxygen species (ROS) which progressively increased during the period examined (15–60 min) up to two-
fold as compared to control, vehicle-treated cultures (*P < 0.001). (E) Co-application of glutamate (50 AM) together with mangiferin or morin (100 nM)
significantly reduced the levels of ROS observed when glutamate was applied alone (100%, control) (*P < 0.01, **P < 0.001). (F) Quantification of caspase-3+
neurons in cultures exposed to glutamate in the absence or presence of antioxidants (*P < 0.05 as compared to neurons treated with glutamate alone). (G)
Representative panels showing cleaved caspase-3+ neurons (green). Inserts at the right top are fields shown at higher magnification to illustrate that cleaved
caspase-3+ neurons also display chromatin condensation (blue staining). Bars in all instances represent the mean T SEM of triplicates from at least 3 different
experiments. Scale bar, 60 and 20 Am in the low power magnification photographs and in inserts, respectively.
M. Gottlieb et al. / Neurobiology of Disease 23 (2006) 374–386378
Fig. 3. Superoxide production in the CA1 region after ischemia was
reduced by mangiferin and morin. (A–D) A panoramic view showing the
increase in the fluorescence emitted by oxidized hydroethidine in the
pyramidal layer of the CA1 region after ischemia and its reduction by
mangiferin and morin. (a–d) A detailed view of the levels of fluorescence
detected within the cytoplasm of cells within the pyramidal layer of the
CA1 region. (E) Quantification of fluorescence intensity shows that
mangiferin and morin reduced superoxide levels after 1 and 2 days
reperfusion (*P < 0.001 as compared with control, n = 4; **P < 0.001 as
compared with rats treated with vehicle, n = 3). Scale bar, 50 and 10 Am in
the left and right columns respectively.
M. Gottlieb et al. / Neurobiology of Disease 23 (2006) 374–386 379
Peak protection was at 100 nM, and higher concentrations of these
polyphenols did not increase further the viability of neurons after
exposure to glutamate.
Since NMDA-receptor-induced toxicity is triggered by Ca2+
influx, we studied whether mangiferin and morin would attenuate
[Ca2+]i overload under these experimental conditions. To that end,
we analyzed by microfluorimetry the changes of [Ca2+]i in
individual neurons after a brief exposure (30 s) to NMDA alone
or together with these polyphenols. Activation of NMDA
receptors with 50 AM NMDA in the presence of 10 AM Gly
increased the [Ca2+]i by 864 T 44 nM (n = 26) (Fig 2A, B).
Interestingly, we found that this increase was reduced around
40% (504 T 25 nM, n = 19) and 60% (320 T 29 nM, n = 20) by
co-application of 100 nM morin and mangiferin, respectively
(Figs 2A, B). Similarly, morin and mangiferin diminished Ca2+
overload induced by glutamate (Fig. 2C). Overall, these results
indicate that Ca2+ influx during NMDA receptor activation in
neurons is attenuated by antioxidants morin and mangiferin.
We next assayed whether mangiferin and morin were capable of
reducing the levels of reactive oxygen species (ROS) generated by
excitotoxic insults. To this end, we initially evaluated the levels of
ROS at 15–60 min after glutamate exposure (50 AM; 10 min) and
observed that they increased with time by up to two-fold (Fig. 2D).
In contrast, the addition of mangiferin and morin at the time of
glutamate application resulted in a clear reduction in the levels of
ROS to about 75% of that observed in the absence of these
polyphenols (Fig. 2E), indicating that both antioxidants efficiently
alleviate oxidative stress caused by excitotoxicity.
NMDA receptor activation can induce both apoptosis and
necrosis (Bonfoco et al., 1995), the former through the activation
of caspase-3 (Lalitha and Stuart, 2001). To examine if mangiferin
and morin inhibit apoptosis in excitotoxic insults, we incubated
neurons with glutamate (50 AM) together with glycine (10 AM) for
10 min in the presence or absence of antioxidants (at 100 nM) and
stained the cultures with antibodies to activated caspase-3. Indeed,
we found that both polyphenols substantially reduced by about
30% the number of caspase-3+ cells (Figs. 2F, G) which indicates
that mangiferin and morin attenuate excitotoxic neuronal death by
apoptosis.
Mangiferin and morin reduce ROS and increase the number of
surviving pyramidal neurons in the CA1 after ischemia
We next measured oxidized HEt fluorescence at 1 and 2 days
of reperfusion to evaluate the levels of superoxide anion
generated as a consequence of 10 min transient global ischemia.
We observed a drastic increase in fluorescence in neuronal cell
bodies of the pyramidal layer of the CA1 region as compared to
controls, sham-operated rats (Figs. 3A,a, B,b and E). Treatment
with mangiferin and morin greatly attenuated the increase in
oxidized HEt at 1 day reperfusion (Figs. 3C,c, D,d, E). In turn,
the levels of fluorescence measured in animals treated with these
antioxidants returned to control after 2 days of ischemia (Fig.
3E). These results demonstrate that mangiferin and morin
efficiently reduce reactive oxygen species at the reperfusion time
points examined.
Toluidine blue staining of the hippocampal CA1 region of the
brains used in our experiments revealed that, at 7 and 70 days after
10 min of ischemia, the vast majority of neurons in the pyramidal
cell layer were damaged or lost (Figs. S1A–C). In addition, intense
microgliosis was observed in the CA1 region after 7 days of
recirculation; this was attenuated by 70 days postischemia, as
revealed by immunohistochemistry with an OX42 antibody, which
labels microglia (Figs. S1D–F). In contrast, microglia in less
vulnerable regions such as the dentate gyrus and the CA3 region
had an appearance which was similar to that observed in control
M. Gottlieb et al. / Neurobiology of Disease 23 (2006) 374–386380
animals (data not shown). Overall, these histological findings are
consistent with those described earlier after longer (20–30 min)
ischemic insults (Pulsinelli et al., 1982; Morioka et al., 1991;
Gottlieb and Matute, 1997).
To quantify the number of surviving neurons, we used
immunohistochemistry with NeuN and anti-MAP2 antibodies,
two markers of neurons (Figs. 4 and 5). A dramatic decrease in
the number of NeuN+ cells was noted in the CA1 pyramidal cell
layer of vehicle-treated animals at 7 and 70 days after ischemia
(Fig. 4A), corroborating the results observed using toluidine blue
staining (Figs. S1A–C). Thus, the number of NeuN+ cells at 7
and 70 days postlesion was reduced to an average of 13% and
11% (499 T 91 cells/mm2 and 375 T 65 cells/mm2) respectively
of those present in sham-operated rats (Fig. 4B). In contrast, rats
subjected to ischemia and subsequently treated with polyphenols
displayed a higher number of NeuN+ cells in the pyramidal layer
of the CA1 region both at 7 and at 70 days of recirculation (Fig.
4). Thus, the number of NeuN+ cells in rats treated with
mangiferin 7 and 70 days after ischemia was 31 and 25%
(1157 T 190 cells/mm2 and 901 T 127 cells/mm2), respectively, of
those present in sham-operated animals, with these percentages
being 30 and 31% (1061 T 104 cells/mm2 and 1068 T 214 cells/
mm2) in the case of morin-treated rats. Consistently, the drastic
reduction in MAP-2 immunostaining observed after ischemia was
attenuated in animals treated with mangiferin and morin (Fig. 5).
Moreover, the number of apoptotic neurons in the CA1 pyramidal
layer was reduced by these antioxidants (Fig. 6). Together, these
data indicate that both polyphenols are neuroprotective after
global ischemia in the conditions examined.
Morin improves spatial memory
Animals included in this and subsequent tests did not present
significant differences in their spontaneous motor and behavioral
activities, but ISCH and I + MNG animals were observed to be
more hyperactive than those included in the C and I + MOR
groups.
The spatial learning and memory task was designed to
determine whether the experimental animal was able to remember
the explored (familiar) arms of the Y-maze when the third (novel)
arm was available for exploration (Fig. 7A). The C group
selected the novel arm as the first choice during the second trial
in 100% of cases, while the ISCH and I + MNG groups did so in
only 20% of cases (v2 test; *P < 0.001), i.e., even less frequently
than by random choice (33% of the cases; Fig. 7B). In contrast,
the I + MOR group performed the task similarly to the C group
(80% with novel arm as first choice). As shown in Fig. 6B,
during the third trial, the C and I + MOR groups moved to
random values (40%), while the ISCH and I + MNG groups
selected the novel arm with values exceeding random ones
(60%). In conclusion, animals from the C and I + MOR groups
performed similarly, whereas animals belonging to the ISCH and
I + MNG groups presented impaired recognition of the novel
arm; it is possible that the hyperactivity represented a handicap to
solve this spatial memory test.
Fig. 4. The number of NeuN-antibody-labeled neurons in the postischemic
CA1 region increased after treatment with polyphenols. (A) Left and right
columns depict NeuN+ cells at 7 and 70 days postoperation. Top to bottom
rows illustrate the appearance of staining in sham-operated animals
(control) and in animals treated with vehicle, mangiferin or morin after
ischemia. Note that the area and number of neurons in the pyramidal layer
observed in sham-operated rats (top row) are drastically diminished in
animals subjected to ischemia and subsequently treated with vehicle at the
reperfusion times studied. In contrast, ischemic insults followed by
treatment with either mangiferin or morin presented a higher number of
neurons in the pyramidal layer. Calibration bar is 100 Am. (B) The number
of NeuN+ cells in vehicle-treated animals as compared to sham-operated
rats (100%). The number of NeuN+ cells was higher in animals treated with
mangiferin or with morin as compared to vehicle-treated animals (*P <
0.05, **P < 0.01). Number of NeuN+ cells was referred to as 100%. Each
bar represents the mean T SEM of counts obtained from two sections of the
right and left hippocampi obtained from 4 to 5 animals.
Fig. 6. Treatment with mangiferin and morin reduces the number of
cleaved caspase-3+ cells in the CA1 pyramidal layer. Values are referred
to the number of apoptotic cells present in animals treated with vehicle.
Each bar represents the mean T SEM of counts obtained from two
sections of the right and left hippocampi obtained from 3 to 5 animals
(*P < 0.05, ***P < 0.001 as compared to animals subjected to ischemia
and then vehicle-treated).
Fig. 5. The level of MAP2+ immunostaining in the postischemic CA1
region at 7 days reperfusion increased after treatment with polyphenols.
(A–D) A panoramic view of hippocampus and (a–d) an enlargement of the
CA1 region indicate the loss of neurons in this area at 7 days reperfusion
after ischemia. (E) The quantification of MAP2 staining shows a significant
attenuation of neuronal damage in the animals treated with mangiferin and
morin (*P < 0.05 as compared to controls, n = 3; **P < 0.001 as compared
to animals subjected to ischemia and then vehicle-treated). The calibration
bar in panels A–D and in panels a–d is 500 and 100 Am respectively.
M. Gottlieb et al. / Neurobiology of Disease 23 (2006) 374–386 381
Improved instrumental conditioning following mangiferin and
morin administration
The cumulative records from representative animals collected
from the tenth conditioning session indicate that animals from the
different groups pressed the lever at different rates, suggesting that
they were unable to learn an appropriate performance in response
to this instrumental conditioning test (Fig. 8A). Control rats
presented a lower mean number of lever presses than the other
groups, although without any statistical difference (Fig. 8B).
However, a better indication of the proper acquisition of this
instrumental learning is the time (with respect to pellet delivery) at
which the lever presses were concentrated. In order to quantify the
appropriated timing of lever presses in relation to the presentation
of the reinforcement, we defined a performance index (see Fig.
8C). The performance index favors lever presses concentrated in
the second half of the time period (30 s) designed for pellet
delivery. Optimum performance would be a single lever press
carried out in the 15 s preceding pellet delivery. Upon applying the
performance index to the collected data (Fig. 8D), it became
evident that the best performance corresponded to the C group,
whose values from the 5th to the tenth sessions were significantly
different to those of the I + MNG and I + MOR groups, (P < 0.05,
two-way ANOVA). The performance index corresponding to the
(I) group was significantly lower than that of the C group for the 10
performance indices corresponding to the I + MNG and I + MOR
groups were significantly different to those of the ISCH group,
with the exception of those obtained for the ninth session (P <
0.05, two-way ANOVA). By way of example, the performance
indices of the four experimental groups during the tenth training
session were: C, 0.55 T 0.06; ISCH, 0.36 T 0.04; I + MNG, 0.48 T0.09; and I + MOR, 0.45 T 0.07. Accordingly, animals from the C
group performed significantly better than those from the ISCH
group, whereas the I + MNG and I + MOR groups reached
intermediate values, but the I + MOR animals obtained a similar
number of pellets with less lever presses than those included in the
I + MNG group. Here, again, the larger amount of spontaneous
activity presented by I + MNG animals could prevent then to
perform the learning task in a more efficient way.
Mangiferin and morin administration are associated with
improved classical conditioning
We compared the associative learning capabilities of the four
experimental groups, using a trace conditioning paradigm (Fig.
9A). Animals included in groups C, I + MNG and I + MOR
presented normal learning curves, which consisted of a steady rise
Fig. 7. Spatial memory test in the Y-maze for control (C) and ischemic
(ISCH) rats and for ischemic rats treated with mangiferin (I +MNG) or morin
(I + MOR). (A) Experimental design. In the first trial (left drawing), the
animal was placed at the START arm and allowed to visit both the left
(FAMILIAR) and the STARTarms for 15 min, while the right (NOVEL) arm
was closed (arrow). For trials 2 and 3 (right drawing), the animal was allowed
to explore the three arms of themaze for 5min. Trial 2was performed 5 h after
trial 1, and trial 3 took place 7 days later. (B) Percentage of times that the
novel armwas visited the first by control and experimental rats during trials 2
(white bars) and 3 (black bars). Significant differences between trial 2 and 3
for the four groups and between the C group and the other experimental
groups are indicated (v2 test; *P < 0.001). Note that the I + MOR group
performed this task very similarly to the C group.
M. Gottlieb et al. / Neurobiology of Disease 23 (2006) 374–386382
in the number of conditioned responses (CRs) during the first four
conditioning sessions and asymptotic values (>75% of CRs per
session) by the 5th–6th conditioning sessions (Fig. 9B). No
significant differences were observed between mean CR values for
these three experimental groups across conditioning. Moreover, the
electromyographic (EMG) profiles of CRs were similar for these
three groups (Fig. 9A). In contrast, the ISCH group presented a
mean of 27.5 T 8.5 CRs by the fifth conditioning session
(compared with 80.4 T 4.1 CRs performed by the C group) and
did not manage to exceed 50% of CRs by the ninth and tenth
conditioning sessions. With the exception of values collected
during the first conditioning session, CR values in the ISCH group
across conditioning were significantly lower than those
corresponding to the C group (P < 0.05, two-way ANOVA).
Moreover, the EMG profiles corresponding to CRs recorded from
ischemic animals were smaller in amplitude than those collected
from the other three groups (Fig. 9A). The lower learning
performance of the ischemic group was compared quantitatively
with values pertaining to the C group. Thus, the integrated EMG
(rectified records, expressed in AV s) activity of CRs recorded
during the conditioned stimulus–unconditioned stimulus (CS–US)
interval was significantly smaller for the ischemia group compared
with values pertaining to the C group, during the ninth
conditioning session (14.3 T 6.2 AV s for the ISCH group against
26.7 T 4.3 AV s for the C group; P < 0.05, Scheffe test).
Discussion
We provide evidence here that two antioxidant polyphenols,
mangiferin and morin, attenuate calcium overload, oxidative stress
and cell death caused by excitotoxicity in culture. In addition, we
show that both antioxidants also reduce oxidative stress and are
neuroprotective after ischemia and, more importantly, that they
attenuate the associated learning deficits.
Mangiferin and morin are xanthone and flavonoid polyphenols,
respectively, which are present in plants. Mangiferin has anti-
inflammatory and immunomodulatory activities (Middleton et al.,
2000; Bremner and Heinrich, 2002; Sarkar et al., 2004), whereas
morin has been shown to reduce damage to peripheral organs
caused by ischemia (Wu et al., 1995; Ahlenstiel et al., 2003).
However, their properties as neuroprotectants have not been
previously assayed.
Excessive activation of glutamate receptors generates calcium
overload in the cytosol, oxidative stress and excitotoxicity which
underlies cell damage in ischemia (Mattson, 2003). Accordingly,
excitotoxicity in neuronal cultures has been widely used as a model
to investigate the molecular mechanisms leading to acute and
chronic neurodegeneration (Nicholls, 2004). In the present study,
we employed cultures of cortical neurons to assay the ability of
polyphenols to protect cells from moderate excitotoxic insults.
Consistent with previous studies, we found that excitotoxicity in
these cultures is mediated by activation of NMDA receptors
(Sattler and Tymianski, 2001) and associated with the generation of
ROS (Reynolds and Hastings, 1995). Interestingly, mangiferin and
morin attenuated excitotoxic neuronal death as well as the
generation of ROS due to the sustained activation of glutamate
receptors. In addition to reducing oxidative stress, both mangiferin
and morin lowered [Ca2+]i induced by excitotoxic insults and thus
contribute to diminish neuron demise. This feature is consistent
with previous findings demonstrating that ROS promote the
reverse-mode function of the Na+/Ca2+ exchanger (Eigel et al.,
2004) and that estrogen and estrogen-like antioxidants with a
phenolic structure inhibit the rise of [Ca2+]i by improving the
correct functioning of that exchanger (Sugishita et al., 2003).
Overall, our findings are in agreement with the ability of other
polyphenols to attenuate excitotoxicity (Lee et al., 2004; Cho and
Lee, 2004) and point out to an important role of mangiferin and
morin in regulating [Ca2+]i under oxidative stress. In addition,
these antioxidant polyphenols may also target directly various
intracellular pathways unrelated to oxidative stress including those
involved in apoptosis and inflammation (Mandel et al., 2004).
Epidemiological studies have shown that dietary polyphenols
protect against stroke (Gilgun-Sherki et al., 2002). In particular,
regular intake of the flavonoid quercetin or beta-carotene was
inversely associated with stroke incidence, but this was not the case
for vitamins C and E (Keli et al., 1996). In addition, pretreatment
Fig. 8. Quantitative analysis of instrumental conditioning (a fixed interval schedule, 30 s) for control (C, and dots) and ischemic (ISCH, and circles) rats and for
ischemic rats treated with mangiferin (I + MNG, and squares) or morin (I + MOR, and triangles). (A) After a three-day period of shaping to the Skinner box,
using a fixed ratio (1:1) schedule, animals were presented with the fixed interval (FI30VV) schedule in which they could obtain a food pellet when pressing the
lever just after every 30-s period. The graph illustrates cumulative records from representative animals collected from the tenth conditioning session. The y axes
indicate lever press (responses). The angled bars indicate reinforcements. (B) A representation of the mean (TSEM) number of lever presses across the 10
conditioning sessions. (C) In order to determine the performance index (PI), i.e., the optimum performance of lever pressures for the fixed interval schedule
(i.e., one lever press every 30 s), we developed the equation shown in C, in which Ry represents lever presses during the second part of the 30-s period and Rx
represents presses during the first part. (D) Performance index (PI) for the four experimental groups. Note that although animals with ischemia presented a high
rate of lever presses they obtained a low rate of reinforcements.
M. Gottlieb et al. / Neurobiology of Disease 23 (2006) 374–386 383
with plant extracts enriched in the polyphenol mangiferin and
flavonoids reduces postischemic hippocampal neuronal death
(Martınez-Sanchez et al., 2001; Zhang et al., 2004). In the current
study, we first assayed if administration of mangiferin and its
flavonoid analog morin after a transient forebrain ischemic insult
reduces oxidative stress. Indeed, we observed that both antiox-
idants induced a dramatic reduction in ROS in the pyramidal layer
of the CA1 region at 1 and 2 days of reperfusion which indicates
that these polyphenols are bioavalaible in vulnerable neurons and
efficiently support the antioxidant cellular defenses after injury.
Subsequently, we quantified the number of surviving neurons with
neuronal markers and observed a significant increase in the number
of surviving neurons in the pyramidal cell layer of the CA1 region
of the hippocampus.
To date, few antioxidants have provided effective neuroprotec-
tion when administered after ischemia (Gilgun-Sherki et al., 2002).
These include the yellow pigment curcumin (Ghoneim et al., 2002)
and pyrrolidine dithiocarbamate, an inhibitor of nuclear factor
kappa-B (Nurmi et al., 2004). In addition, two naturally occurring
polyphenols, epigallocatechine gallate (Lee et al., 2004) and
Crataegus flavonoid (Zhang et al., 2004), have joined the list of
antioxidants with therapeutic potential for acute CNS damage.
However, the long-term functional relevance of neuronal preserva-
tion after treatment with antioxidants has not been carefully
evaluated. Notably, in the current study, we found that the neuro-
protection produced in the hippocampus by mangiferin and morin
alleviates functional deficits caused by brain ischemia, as assessed
by three key behavioral assays relevant to the integrity of this
structure (Gerlai, 2001).
The hippocampus is considered to be one of the sites for the
acquisition and/or storage of different learning and memory
abilities (O’Keefe and Dostrovsky, 1971; Zola-Morgan and
Squire, 1986; Squire, 1992; Weiss et al., 1999; Corbit and
Balleine, 2000; Alvarez et al., 2001). In addition, hippocampal
Fig. 9. A quantitative analysis of classically conditioned eyelid responses
from control (C, and dots) and ischemic (ISCH, and circles) rats and from
ischemic rats treated with mangiferin (I + MNG, and squares) or morin (I +
MOR, and triangles). (A) Electromyographic (EMG, in mV) recordings from
representative animals of each of the indicated experimental groups collected
during the ninth conditioning session. For trace conditioning, a tone (600 Hz,
90 dB) was presented for 20 ms as a conditioned stimulus (CS). The tone was
followed 270ms later by an electrical shock (500 As, 2� threshold) presented
to the supraorbitary nerve as an unconditioned stimulus (US). Bent arrows
indicate the presence of conditioned responses (CRs). Arrowheads indicate
the appearance of unconditioned eyelid responses. (B) Graphs of mean
(TSEM) percent conditioned responses across the 10 conditioning sessions
for the four experimental groups. Results collected from conditioned groups
are indicated by continuous lines, whereas results corresponding to
pseudoconditioned groups are indicated by discontinuous lines. Note the
low learning curve corresponding to ischemic animals.
M. Gottlieb et al. / Neurobiology of Disease 23 (2006) 374–386384
neuronal circuits are involved in many different behavioral and
cognitive functions including spatial navigation (O’Keefe and
Dostrovsky, 1971; Rosenzweig et al., 2003), non-spatial relational
memory (Alvarez et al., 2001), associative strength and/or CS
predictive value in trace classical conditioning (Weiss et al.,
1999; Munera et al., 2001) and related learning activities (Zola-
Morgan and Squire, 1986; Squire, 1992). It can thus be expected
that the death of hippocampal neurons induced by experimental
ischemia will give rise to selective cognitive deficits (Auer et al.,
1989). Consistently, we found that spatial orientation related to
the exploration of novel territories was affected by ischemia but
is preserved by the administration of morin. However, the
administration of mangiferin was ineffective for the proper
performance of this task, suggesting that the neuronal population
spared by these two antioxidants may subserve diverse functional
properties. Indeed, our behavioral data indicate that ischemia
produces hyperactivity in the animals, which was better counter-
balanced by morin.
In addition, animals which underwent ischemia and were not
treated with polyphenols presented difficulties when carrying out
the proposed instrumental conditioning task. In fact, they presented
a number of lever presses which were excessively higher than the
rate demanded by the fixed interval schedule which we had
selected (30:1). It has already been reported that hippocampal-
lesioned rats are unable to inhibit their behavior to the same degree
as controls (Corbit and Balleine, 2000). Thus, ischemic rats
presented duration-related impairments and/or were unable to
generate a proper timing of lever pressing (Nelson et al., 1997). In
the present case, treatment with mangiferin or morin improved
significantly the performance index of treated animals in the
Skinner box. Nevertheless, morin-treated animals obtained the
same number of pellets using less lever pressings, an indication
that this antioxidant compensates in a more efficacious way than