Embed Size (px)
Citation preview
Archives of Medical Research 33 (2002) 6–14
0188-4409/02 $–see front matter. Copyright © 2002 IMSS. Published by Elsevier Science Inc.PII S0188-4409(01)00347-2
ORIGINAL ARTICLE
Neuroprotective Effects of Progesterone on Damage Elicited by Acute Global Cerebral Ischemia in Neurons of the Caudate Nucleus
Miguel Cervantes,
a
María Dolores González-Vidal,
b
Rodrigo Ruelas,
b
Alfonso Escobar
c
and Gabriela Moralí
b
a
Laboratorio de Neurofarmacología, Centro de Investigación Biomédica de Michoacán, Instituto Mexicano del Seguro Social (IMSS),Morelia, Michoacán, Mexico
b
Unidad de Investigación Médica en Farmacología, Centro Médico Nacional Siglo XXI (CMN-SXXI), IMSS, Mexico City, Mexico
c
Departamento de Fisiología, Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México (UNAM), Mexico City, Mexico
Received for publication January 9, 2001; accepted May 25, 2001 (01/007).
Background.
In addition to the hippocampus, the dorsolateral caudate nucleus (CN) andthe pars reticularis of the substantia nigra (SNr) are among the most vulnerable brain areasto ischemia. A possible association of the neuronal injury in these two subcortical nucleihas been proposed, the primary damage affecting the CN GABAergic neurons innervatingthe SNr, and secondarily the SNr neurons as a result of an imbalance of GABAergic andglutamatergic input to the SNr. Progesterone (P
4
) exerts a GABAergic action on the cen-tral nervous system (CNS) and is known to protect neurons in the cat hippocampus fromthe damaging effect of acute global cerebral ischemia (AGCI). The effects of AGCI on theneuronal populations of the CN and SNr, in addition to the possible neuroprotective ef-fects of P
4
, were assessed in cats in the present study.
Methods.
Ovariectomized adult cats were treated subcutaneously (s.c.) with either P
4
(10 mg/kg/day) or corn oil during the 7 days before and 7 days after being subjected to aperiod of AGCI by 15 min of cardiorespiratory arrest followed by 4 min of reanimation.After 14 days of survival, animals were sacrificed and their brains perfused
in situ
withphosphate-buffered 10% formaldehyde for histologic examination.
Results.
ACGI resulted in an intense glial reaction in the CN and a significant loss (43%)of medium-sized neurons of the CN, but no difference was found in the densities of SNrneurons between controls and ischemic oil- and P
4
-treated cats. Progesterone treatmentcompletely prevented CN neuronal loss.
Conclusions.
The overall results point to the higher vulnerability of CN neurons to is-chemia as compared to neurons in the SNr and show the protective effects of P
4
upon CNneuronal damage after ischemia. © 2002 IMSS. Published by Elsevier Science Inc.
Key Words:
Global cerebral ischemia, Neuroprotection, Caudate nucleus, Progesterone, Cat.
Introduction
Acute global cerebral ischemia (AGCI) triggers a series ofpathophysiologic phenomena that result in acute, matura-tional, and delayed neuronal death in specific, highly vul-nerable brain structures (1–16) that include the following:
the pyramidal neurons of hippocampal CA subfields; pyra-midal neurons in cerebral cortex layers 3 and 5; Purkinjecells in the cerebellum, and middle- and small-sized neu-rons in the dorsolateral striatum (6,17–22). Several mecha-nisms have been associated with the high vulnerability ofthese neuronal types, including abundance of glutamatergicor dopaminergic innervation (4,6,8,23,24) and the contentof certain metallic compounds (25) among others. In partic-ular, the excessive glutamatergic and dopaminergic activityhave been shown to be involved in ischemia-induced neu-ronal damage in the striatum (26–28), where the cytotoxic
Address reprint requests to: Gabriela Moralí, Ph.D., Unidad de Investi-gación Médica en Farmacología, CMN-SXXI, IMSS, Apdo. Postal 73-032,03020, México, D.F., México. Tel.: (
�
52) (55) 5687-8606; FAX: (
�
52)(55) 5761-0952; E-mail: [email protected]
Progesterone Neuroprotection in Caudate Nucleus
7
effect of excessive dopamine (DA) seems to be mediated byD
2
receptors and free oxygen radicals resulting from themetabolism of the excessive amounts of DA released duringischemia and biotransformed during reperfusion (29–31).
Selective damage to the CA1 pyramidal neurons after is-chemia appears to result from an imbalance between excita-tory and inhibitory influences (32,33). Thus, agents reducingexcitatory aminoacid neurotransmission (34–37) or increas-ing GABAergic inhibitory neurotransmission (38–46) protectCA1 pyramidal neurons from the ischemia damaging effect.
Neuroprotective compounds that are effective in brainstructures such as the hippocampus, in which dopaminergicactivity is not a main component of pathophysiologic mecha-nisms of neuronal damage, may exhibit different neuropro-tective effects in the striatum where dopaminergic mecha-nisms of neuronal damage are important. Nevertheless, theGABAergic inhibitory influence directly exerted by CN inter-neurons on medium-sized caudate neurons (47–49) in addi-tion to the GABAergic influence on nigrostriatal dopaminer-gic neurons (50,51) may support a possible GABAergic-mediated neuroprotective effect such as that suggested forprogesterone (P
4
) (52–55).An association has been proposed to exist between the
neuronal damage in the dorsolateral part of the caudate nu-cleus (CN) and the pars reticularis of the substantia nigra(SNr), where the primary damage affects the CN GABAer-gic neurons innervating the SNr and secondarily the SNrneurons as a result of an imbalance of GABAergic andglutamatergic input to the SNr (56–59). Thus, ischemic neu-ronal damage in the SNr may be prevented or reduced by in-creasing the GABAergic activity (40,41,45).
In previously reported data (53), the neuroprotective ef-fects of P
4
on neuronal damage in the cat hippocampus wereexplained by an enhancement of the GABAergic inhibitoryinfluence exerted by this steroid either per se or through itsbiotransformation in the brain. Steroids derived from thebiotransformation of progesterone interact with specific rec-ognition sites in the GABA
A
receptor increasing the GABAer-gic neurotransmission (60–65). Furthermore, neurologic as-sessment of consciousness, sensory, motor, autonomic, andbehavioral conditions in ischemic, progesterone-treated catsshowed significantly lower damage as compared to that inischemic, vehicle-treated cats (53). These findings suggestthat the neuroprotective effect of progesterone is also ex-erted in other brain structures in addition to the hippocam-pus. Thus, an analysis of the neuronal populations of thesubstantia nigra and the dorsolateral striatum was consid-ered to be of interest, in view of their functional and neu-roanatomic relations.
Materials and Methods
The experimental protocol was approved by the ScientificResearch Committee (March 29, 1996). Brain tissue sam-ples used for histologic analysis of the caudate nucleus and
substantia nigra were obtained from cats included in a pre-vious study (53). Assignment of cats to the three experimen-tal groups as well as the experimental procedures to whichthey were subjected was previously described (53). In brief,subjects included 18 adult ovariectomized female cats (2.5–3.2 kg body weight [b.w.]) randomly allotted to one of threegroups and given daily subcutaneous (s.c.) injections of ei-ther vehicle (corn oil, 0.5 mL/kg/day, groups 1 and 2) orprogesterone (10 mg/kg/day in corn oil, group 3) during 7days. This dose of progesterone allows the achievement andmaintenance of stable blood levels of this steroid abovephysiologic concentrations (53).
Blood samples were obtained under light anesthesia froma hind leg in all cats at days 0, 2, 4, and 7 of treatment tomeasure circulating P
4
by an automated immunoassay. Theserum was separated immediately and stored at
�
4
�
C untilassayed. Progesterone was measured by a chemilumines-cent enzyme immunoassay using commercial kits (IMMU-LITE Progesterone, Diagnostic Products Corporation, LosAngeles, CA, USA). The detection limit for P
4
was 0.09 ng/mL and the intra- and interassay coefficient of variation was6 and 8.5%, respectively.
On day 7, each cat of groups 2 and 3 was submitted to a15-min period of acute global cerebral ischemia as a resultof cardiorespiratory arrest (CRA), followed by reanimationwithin 4 min, according to a model of AGCI previously de-scribed (53,66–68). Animals in group 1 were subjected tosham procedures only. Experiments were carried out undercontrolled conditions including halothane anesthesia, as-sisted mechanical ventilation, blood pressure, blood pH,base excess, PaO
2
, PaCO
2
, glucose, and body temperature(53,68) as follows: cats were anesthetized for prearrest sur-gery with 4% halothane in oxygen; pancuronium bromide,0.3 mg/kg, was administered intravenously (i.v.) through asterile venous catheter placed in a hind leg, and endotra-cheal intubation was performed and assisted ventilationwith 1.5% halothane in oxygen (Bird MarkVIII ventilator)was begun to maintain anesthesia and PaCO
2
between 30and 35 mmHg. After previous skin infiltration with 1 mL of2% lidocaine, a small incision (1 cm) was made in the neckand in the groin. A sterile catheter was inserted through thejugular vein to guide the tip of a wire (0.75 mm in diameter)into the right atrium; its precise location was confirmed bycavitary electrocardiogram (EKG). Another sterile catheterwas inserted into the right femoral artery for continuousmonitoring of mean arterial pressure (MAP). On completionof surgery, neck and groin wounds were sutured and cov-ered. Halothane administration was interrupted and 1 minlater ventricular fibrillation was induced in cats of groups 2and 3 by passing alternating current (60 Hz, 20 V, 5–10 sec)from the tip of the atrial wire to a subcutaneous electrodeplaced at the apex until MAP showed a sudden decrease to
�
10 mmHg. Then, mechanical ventilation was stopped andthe tracheal cannula was occluded. Five min. after the be-ginning of the cardiac arrest, the atrial wire was removed.
8
Cervantes et al./ Archives of Medical Research 33 (2002) 6–14
Cardiac arrest and interruption of mechanical ventilationwere maintained for 15 min. Cardiopulmonary resuscitationwas initiated at the end of this period as follows: mechanicalventilation with FiO
2
�
1; external cardiac massage to in-crease and maintain MAP at 90 mmHg, and administrationof sodium bicarbonate, 1 mEq/kg i.v., and epinephrine hy-drochloride, 15
�
g/kg i.v. Defibrillation (Mennen Cardio-pak model 936, Mennen Greatbatch Electronics, Inc., Clear-ance, NY, USA) by means of a 20 J DC shock appliedbetween two chest paddles placed on the shaved lateralchest walls was first attempted 2 min after initiating car-diopulmonary resuscitation. When unsuccessful, additionalsodium bicarbonate and epinephrine hydrochloride were ad-ministered and defibrillation was repeated until successfulwhen MAP
�
90 mmHg was reached and maintained. Im-mediately after defibrillation, atropine sulfate, 50
�
g/kg i.v.and lidocaine hydrochloride, 1 mg/kg i.v. were administeredwhen needed to assist stabilization of cardiac sinusal rhythm.Cardiopulmonary resuscitation within a period no longerthan 4 min was a necessary condition for animals to be in-cluded in the study.
EKG (lead II) was continuously monitored through sub-cutaneous needle electrodes. Esophageal temperature waskept at 37.0–37.5
�
C. A sample (1 mL) of arterial blood wasdrawn 15 min prior to CRA, at 5 and 20 min, and at 1, 2,and 4 h following CRA or sham maneuvers to determinepH, PaO
2
, PaCO
2
, bicarbonate, base excess (pH-Blood Gas-CIBA Corning 2381 model, Ciba Corning de México, S.A.de C.V., México, D.F.), and glucose along the experiment.Changes in their pre-arrest values were promptly correctedthrough the administration of sodium bicarbonate or venti-lation adjustment. Values of PaCO
2
30–35 mmHg and pH7.30–7.35 were maintained until reversal of neuromuscularblockade. Assisted ventilation was maintained at FiO
2
�
1for 1 h after CRA or sham procedures and FiO
2
�
0.4 after-ward. Neuromuscular blockade (pancuronium bromide, 0.5mg/kg/h i.v.) was maintained until 6 h after CRA.
Cats in group 1 received no alternating current via the wireinserted through the jugular venous catheter to the atrium andmechanical ventilation was not stopped; thus, these cats weresubjected neither to cardiorespiratory arrest nor resuscitationmaneuvers but only to sham procedures including neuromus-cular blockade and assisted ventilation for 6 h.
At this time, the cats were allowed to recover from neu-romuscular blockade until normal spontaneous respiratoryactivity was resumed. Neostigmine methylsulfate 0.06 mg/kg was administered i.v. to reverse neuromuscular blockadeand atropine sulfate, 0.04 mg/kg i.v. was administered toprevent bradycardia. Arterial and atrial cannulas were re-moved under halothane anesthesia and the animals were ex-tubated. After extubation, each cat was placed in a cage at atemperature of 25
�
C until 24 h after resuscitation. On thefollowing days, cats were allowed to drink milk and waterand eat tuna fish paste. If needed, 50 mL/kg/day mainte-nance fluid was injected s.c.
Progesterone or vehicle treatment was continued for anadditional 7 days following CRA or sham procedures. Eachcat was subjected to daily neurologic evaluations in a blindmanner, assessing a number of neurologic parameters suchas level of consciousness, respiration, cranial nerves, andspinal reflexes as well as postural, locomotor, and behav-ioral reactions according to the procedure designed by Toddet al. (66). Points were assigned to each neurologic alter-ation and added together to obtain a neurologic deficit scoreranging from 0 to 100 (score 0, normal neurologic condi-tion; score 100, maximal neurologic deficit).
On survival day 14, the cats were deeply anesthetizedwith pentobarbital (35 mg/kg i.p.), the chest was opened,the right auricle incised, and a 14-gauge needle inserted intothe left ventricle for perfusion. Transcardiac perfusion be-gan with 400 mL saline, followed by 800 mL 10% phos-phate-buffered formaldehyde and 300 mL Clarke fixative(ethanol-acetic acid 3:1 v/v) (69). Following perfusion, thebrains were removed and immersed in the same fixative forat least 7 days prior to histologic processing. Brains werethen cut into 3-mm coronal slices, dehydrated, and embed-ded in paraffin. Semiserial 10-
�
m sections were sampledfrom the caudate nucleus and the substantia nigra, andstained with Klüver-Barrera technique with Luxol fast blueand cresyl violet (70).
For each brain, five sections through the central portionof the caudate nucleus, located between A15.0 and A17.0,i.e., 15–17 mm anterior to the interaural line according tothe atlas of Snider and Niemer (71), and five sections of thesubstantia nigra at the level of the emergence of the oculo-motor nerve (cranial nerve III) at the ventral mesencephalonbetween A4.5 through A5.5, i.e., 4.5–5.5 mm anterior to theinteraural line, were analyzed for cell counting. The numberof surviving medium-sized (17–25
�
m in diameter) neuronsin at least six microscopy fields measuring 450
�
m in diam-eter, located in the dorsolateral part of the caudate nucleusin each section, was counted and averaged. Only neuronsshowing normal morphology and visible nucleolus werecounted. The number of surviving large (15–42
�
m thelongest diameter) and small (10–15
�
m in diameter) neu-rons of the reticulata and globular types (72), in at least sixmicroscopy fields of 450
�
m in diameter located in the parsreticularis of the substantia nigra in each section, wascounted and averaged. Only neurons showing normal mor-phology, abundant Nissl material, and visible nucleoluswere counted. Neurons that had shrunken cell bodies withsurrounding empty spaces were excluded. Sections were ex-amined in a blind fashion under light microscopy at a mag-nification of
400. The average numbers of neurons in thevarious microscopy fields were evaluated in each animal.
Analysis of variance and Duncan tests were used to com-pare body weight, progesterone levels, MAP, blood glu-cose, pH, blood gases, and base excess values under the dif-ferent experimental conditions. Mann-Whitney
U
test wasused to compare neurologic deficit scores between proges-
Progesterone Neuroprotection in Caudate Nucleus
9
terone- and vehicle-treated cats. Values of number of neu-rons per microscopy field in each cat in each experimentalgroup were expressed as median and range. Statistical anal-ysis was done using Kruskal-Wallis test followed by Mann-Whitney
U
tests for comparison of number of each type ofneurons counted in the caudate nucleus and in the substantianigra, among the groups (73,74).
Results
Values of body weight were not significantly differentamong the three groups of cats, i.e., sham (group 1), vehicle(group 2)-, and P
4
(group 3)-treated groups submitted to is-chemia (Table 1). Data on the different variables relevantfor the experimental model of acute global cerebral is-chemia and on neurologic outcome under vehicle or P
4
treatment were previously reported in the same experimen-tal subjects included in the present study (53) and are sum-marized in Tables 1 and 2. Serum levels of P
4
in female catson day zero, prior to initiating vehicle or progesterone treat-ment, were similarly low in groups 1, 2, and 3; serum levelsof P
4
in cats receiving vehicle remained low throughout the7 days of treatment, and daily s.c. administration of progest-erone resulted in a gradual increase in serum levels of P
4
that reached 193.5
70.2 ng/mL at day 7 after onset of P
4
treatment.Blood gases, pH, base excess, and MAP were within
physiologic ranges in all cats immediately before CRA orsham procedures and were similar among the sham, the ve-hicle-treated, and the P
4
-treated groups.A transient but significant decrease in pH and base ex-
cess values (
p
�
0.01) and a significant increase in PaCO
2
values (
p
�
0.05) were found 5 and 20 min after the end ofresuscitation in both vehicle- and P
4
-treated cats submittedto ischemia as compared to their pre-CRA values, and to thesham group. However, these values were corrected, so thatmean values of the blood components obtained in subse-quent determinations from 1–4 h after CRA were not signif-icantly different from those obtained prior to CRA. An in-crement in PaO
2
values (
p
�
0.05) due to assisted ventilationwith FiO
2
�
1 during the first hour after CRA was found inboth experimental groups (233.1
90.5 and 219.5
111.0mmHg, respectively). Blood gases, pH, and base excesswere within the physiologic range in intact cats submitted tosham procedures.
In both groups submitted to ischemia, plasma glucoseconcentrations significantly increased after resuscitation(Table 1) and remained high throughout the post-CRA pe-
Table 1.
Values (mean
SD) of the physiologic variables recorded in cats under different experimental conditions
Group SHAM ISQ
�
VEH ISQ
�
P
4
Body weight (kg) 2.9
0.4 2.9
0.3 3.0
0.6Serum progesterone (ng/mL)
Day 0 2.1
0.7 2.3
0.6 1.9
0.8Day 2 1.5
0.3 1.5
0.7 122.0
27.3
b
Day 4 2.5
0.7 1.8
0.6 146.3
35.2
b
Day 7 1.9
0.5 2.8
0.7 193.5
70.2
b
pHBasal 7.36
0.02 7.38
0.08 7.36
0.0820 min 7.36
0.15 7.15
0.30
b
7.17
0.10
b
1
�
4 h 7.37
0.20 7.36
0.20 7.35
0.13Base excess (mEq/lt)
Basal 3.0
0.1
�
11.0
2.0
�
12.0
3.020 min
�
11.0
1.6
�
22.0
4.0
b
�
17.0
3.0
b
1
�
4 h
�
9.1
0.6
�11.2 4.1 �9.0 2.2PaCO2
Basal 26.0 2.0 21.0 6.0 21.0 8.520 min 15.0 8.0 37.0 13.0a 30.0 10.0a
1�4 h 21.0 8.0 23.0 11.0 24.0 8.0PaO2
Basal 84.0 13.0 120.0 45.0 124.0 64.01 h 90.0 15.0 233.1 90.5a 219.5 111.0a
2�4 h 102.5 20.2 129.7 39.0 139.2 56.4Glucose (mg/dL)
Basal 130 21 136 31 127 5410�30 min 150 25 208 48a 268 67a
1�4 h 165 21�215 50 185 92�227 36 189 33�204 61MAP
Basal 95 5 110 22 107 2230 min 110 28 160 28a 134 24a
1�4 h 120 14�140 14 119 18�136 20 109 22�126 22
ap �0.05; bp �0.01 as compared to the sham group. Duncan test.
10 Cervantes et al./ Archives of Medical Research 33 (2002) 6–14
riod; however, differences were not significant betweengroups when compared at specific times after CRA. In thegroup submitted to sham procedures, plasma glucose con-centrations ranged from 130 21 mg/mL to 215 50 mg/dL during the entire experimental period, values signifi-cantly lower than those of groups 2 and 3 at 10 and 30 minafter CRA, but not later.
Values of MAP were elevated during the first minute af-ter CRA in the vehicle-treated and P4-treated groups and de-creased to values similar to those of the sham group duringthe remaining experimental period. There were no signifi-cant differences in these values when comparisons weremade at 1-h intervals after CRA between vehicle- and P4-treated groups. Esophageal temperature was maintained at37.0–37.5�C in all cats throughout the experimental phase.
Neurologic deficit scores (Table 2) ranged from 56 to 81points on day 1 post-CRA in the vehicle-treated cats andfrom 23 to 42 points in the progesterone-treated animals.These scores clearly tended toward a reduction in both vehi-cle- and P4-treated groups, showing neurologic deficit scoresfrom 12 to 62 and from 8 to 12, respectively, on day 4 andfrom 3 to 42 and 1 to 6, respectively, on day 7; no furtherchanges were observed on days 8–14 after CRA. Thus, neu-rologic deficit scores in this period are not shown in Table2. High neurologic deficit scores in vehicle-treated cats re-sulted from persistence of abnormal pupil size and light re-flex, diminution of facial pain perception, flexor reflex topain, and orienting reflex to loud clap, as well as from lackof placing paw reflex, on the days following CRA. Neuro-logic deficit scores were significantly lower (p �0.05) inP4-treated cats on the days following CRA than in vehicle-treated cats.
Figure 1 shows representative images of the neuronalpopulation in the caudate nucleus of cats subjected to shamprocedures (upper image) or subjected to ischemia andtreated either with the vehicle (middle image) or with P4
(lower image). In the cats subjected to 17–19 min of is-chemia and treated with vehicle, there was a clear reductionin the number of medium-sized neurons in the caudate nu-cleus as compared to that of cats in the sham group. On theother hand, in progesterone-treated cats subjected to is-
chemia, the neuronal population was equal to that in shamcats. An intense glial reaction was associated with the re-duction of the neuronal population in the caudate nucleus ofvehicle-treated cats.
Numerical data of medium-sized neurons in the caudatenucleus of the different groups of cats are shown in Table 3.In the group treated with vehicle and subjected to ischemia,the number of neurons (Md: 37.6/field; range: 29.5–68.4)was significantly lower (p �0.01) than that of the sham
Table 2. Neurologic deficit scores shown on the first 7 days following ischemia, by cats treated either with vehicle or with progesterone (P4, 10 mg/kg/day)
Ischemia � vehicle Md range Ischemia � P4 Md range
Day 1 69.5 56�81 28.0 23�42a
Day 2 49.5 26�71 17.0 14�30a
Day 3 47.0 16�66 16.0 10�20a
Day 4 40.0 12�62 11.5 8�12a
Day 5 31.0 8�57 8.0 4�11a
Day 6 25.0 6�57 5.0 1�7Day 7 15.5 3�42 3.0 1�6
ap �0.05 as compared to vehicle-treated cats. Mann-Whitney U test.
Figure 1. Representative photomicrographs of the neuronal populationfound in the dorsolateral caudate nucleus of cats subjected to sham proce-dures (SHAM), and cats subjected to acute global cerebral ischemia undereither vehicle (ISCH � VEH) or progesterone (ISCH � P4) treatment.Note the smaller amount of medium-sized neurons in the caudate nucleusof ischemic, vehicle-treated cats as compared to sham cats, and the preser-vation of neurons in P4-treated cats. Luxol fast blue and cresyl violet. Scalebar, 100 �m.
Progesterone Neuroprotection in Caudate Nucleus 11
group (Md: 66.4/field; range: 51.8–80.3), amounting to only57% of the total number of neurons in sham cats taken as100% (Table 3). In contrast, surviving neurons in the proges-terone-treated cats subjected to ischemia (Md: 63.0/field;range: 48.6–73.0) amounted to 95% of the total neuronalpopulation of the sham group (100%) without significantdifferences between these two groups; numbers of neuronsin the caudate nucleus of P4-treated cats were significantlyhigher (p �0.01) than those of vehicle-treated animals (Ta-ble 3). On the other hand, neuronal population in the parsreticularis of the substantia nigra did not differ amonggroups, as seen in Figure 2 and Table 4.
Discussion
In the present study, a carefully controlled experimentalmodel of acute global cerebral ischemia provoked by car-diorespiratory arrest was used. Control of variables (MAP,blood gases, pH, etc.) able to influence the magnitude of is-chemic-induced neuronal damage was carried out in a simi-lar manner in each cat from the different experimental groups,as we and other authors have done in similar studies (53,66–68). In particular, blood glucose increases were similar be-tween P4-treated and vehicle-treated cats. Hence, it can beassumed that these factors did not contribute to neurologicor histologic differences between groups.
Loss of medium-sized neurons in the dorsolateral cau-date nucleus has been a consistent finding following globalcerebral ischemia, although the magnitude of neuronal dam-age varies depending on some experimental variables, mainlythe species and the duration of ischemia (19–22,29,35). In thepresent study, a loss of 43% of the population of medium-sized caudate neurons was observed following 17–19 minof acute global cerebral ischemia.
An imbalance between excitatory and inhibitory neu-rotransmission has been proposed as a main factor leading toneuronal damage in the striatum following a period of cere-bral ischemia. In particular, abnormally augmented gluta-matergic and dopaminergic excitatory activity has beenshown to be involved in neuronal damage affecting medium-sized neurons in the dorsolateral striatum, including an im-portant proportion of GABAergic neurons (26–31).
A consequence of the lesion of medium-sized neurons ofthe caudate nucleus is the reduction of the neuronal popula-tion of the SNr, as observed in other experimental models inwhich single or repetitive episodes of global cerebral is-chemia were induced (57–59). This has been interpreted asdue to transneuronal degeneration of neurons in the SNr re-sulting from the lack of normal GABAergic innervationfrom the medium-sized GABAergic neurons of the caudatenucleus and the globus pallidus projecting to the reticulata-type neurons of the SNr and the lateral SN leading to animbalance between excitatory and inhibitory inputs, thus en-hancing glutamate-mediated excitotoxicity (40,41,57–59).
Table 3. Number of medium-sized neurons (17�25 �m diameter) found in microscopy fields of 450 �m of diameter in the dorsolateral caudate nucleus of cats under the various experimental conditions
Number of neurons/field(Md, range)
Sham 66.4 51.8�80.3Ischemia � vehicle 37.6 29.5�68.4a
Ischemia � P4 63.0 48.6�73.0b
a p �0.01 as compared to the sham group; bp �0.01 as compared to the ve-hicle-treated group. Mann-Whitney U test.
Figure 2. Representative photomicrographs of the neuronal populationfound in the substantia nigra pars reticularis of cats subjected to sham pro-cedures (SHAM), and cats subjected to acute global cerebral ischemiaunder either vehicle (ISCH � VEH), or progesterone (ISCH � P4) treat-ment. No significant differences were found in the densities of neuronsamong cats under the various experimental conditions. Luxol fast blue andcresyl violet. Scale bar, 100 �m.
12 Cervantes et al./ Archives of Medical Research 33 (2002) 6–14
It is known that neurotoxic damage in the striatum destroys95% of caudate neurons resulting in secondary neuronaldamage in the SNr (56), and that extensive damage of thestriatum is a necessary condition for transneuronal degener-ation of SNr neurons following ischemia (57). Nonetheless,in this case detailed quantitative data concerning the magni-tude of striatal damage in terms of its neuronal populationand the resulting loss of SNr neurons have not been de-scribed. In the present study it appears that the severity ofthe caudate damage, with 43% loss of medium-sized neu-rons, was not of sufficient magnitude to produce a second-ary, significant reduction of SNr neuronal population 14days after ischemia.
Present results support the neuroprotective effect of P4
on the medium-sized neurons of the dorsolateral caudatenucleus. The possibility exists that some of the effects foundmay be mediated by P4 biotransformation to some of its3�,5�- and 3�,5�-reduced neuroactive metabolites, as sug-gested for other effects of P4 on the brain (75–77). In fact,two of the key steroid-metabolizing enzymes, 5�-reductaseand 3�-hydroxysteroid oxidoreductase, are widely distrib-uted in the brain (78) and although lower than in other brainareas, their activity has been demonstrated both in the hip-pocampus and in the striatum (79).
A neuroprotective effect of P4 treatment as shown bypreservation of pyramidal neurons of the hippocampus andbetter neurologic outcome following acute global ischemiain cats (53) has been explained as due to an enhancement ofGABAergic activity in the central nervous system. It hasbeen demonstrated that the neuroactive metabolites of P4
may interact with the GABAA receptor, increasing the in-ward Cl� currents at both presynaptic and postsynaptic lev-els (60–65). Thus, inhibition of neuronal excitability (75)and a reduction in the release of excitatory neurotransmit-ters (80) may account for the neuroprotective effect of thesesteroids under different cerebral injury conditions (81–88). AGABAergic-mediated neuroprotective mechanism couldalso be involved in the preservation of the medium-sizedcaudate neurons in the present study under P4 treatment be-cause these neurons receive GABAergic innervation, thusrendering them suitable targets for neuroprotective drugsenhancing GABAergic neurotransmission. In fact, both di-rect and indirect GABA agonists exert a neuroprotective ef-
fect on caudate and SNr neurons following forebrain is-chemia in gerbils and rats (40,41,44–46).
In addition, enhancement of GABAergic inhibitory activityinduced by P4 treatment may counteract the ischemia-inducedexcitotoxic phenomena associated with excessive glutamater-gic and dopaminergic activity, thus resulting in neuroprotec-tion of the vulnerable neurons of the caudate nucleus.
Other mechanisms in addition to the increase in GABAer-gic activity may contribute to the neuroprotective effects ofprogesterone on the caudate nucleus neuronal loss observedin the present study. Progesterone has been shown to attenu-ate lipid peroxidation induced by FeSO4 and amyloid �-peptide,protect neuronal cultures against glutamate toxicity and glu-cose deprivation (89), and reduce lipid peroxidation aftercortical contusion in rats (83). Because an oxidative damagehas been proposed to contribute to striatal damage after is-chemia (29–31), a possible reduction of lipid peroxidationby P4 may contribute to its neuroprotective effects found inthe present study. Further studies should be undertaken toconfirm these possibilities.
Overall data support the conclusion that the neuroprotec-tive effects of P4 are not limited to the highly vulnerable py-ramidal neurons of the hippocampus but may also be ex-erted in the neuronal components of other brain structuressuch as the caudate nucleus. This idea is consistent with thepossibility that alterations in neurologic phenomena whoseintegration depends on the functioning of these brain struc-tures, aside from the hippocampus, may also be reduced un-der progesterone treatment.
AcknowledgmentsThis work was partially supported by a research grant from theConsejo Nacional de Ciencia y Tecnología (CONACYT 3400P-M0896), Mexico.
References1. Raichle ME. The pathophysiology of brain ischemia. Ann Neurol
1983;13:2–10.2. Dearden NM. Ischaemic brain. Lancet 1985;2:255–259.3. Graham DI. The pathology of brain ischaemia and possibilities for
therapeutic intervention. Br J Anaesth 1985;57:3–17.4. Wieloch T. Neurochemical correlates to selective neuronal vulnerabil-
ity. In: Kogure K, Hossman KA, Siesjö BK, Welsh FA, editors. Mo-lecular mechanisms of ischemic brain damage. Progress in brain re-search. Vol. 63. New York: Elsevier;1985. pp. 69–82.
5. Kaplan J, Dimlich RVW, Biros MH, Hedges J. Mechanisms of is-chemic cerebral injury. Resuscitation 1987;15:149–169.
6. Kogure K, Tanaka J, Araki T. The mechanism of ischemia-inducedbrain cell injury. Neurochem Pathol 1988;9:145–170.
7. Krause GS, White BC, Aust SD, Nayini NR, Kumar K. Brain celldeath following ischemia and reperfusion: a proposed biochemical se-quence. Crit Care Med 1988;16:714–726.
8. Haddad GG, Jiang C. O2 deprivation in the central nervous system: onmechanisms of neuronal response, differential sensitivity and injury.Prog Neurobiol 1993;40:277–318.
9. Hara H, Sukamoto T, Kogure K. Mechanism and pathogenesis of is-chemia-induced neuronal damage. Prog Neurobiol 1993;40:645–670.
Table 4. Number of large (15�42 �m longest diameter) and small (10�15 �m diameter) neurons found in microscopy fields of 450 �m of diameter in the substantia nigra pars reticularis of cats under the various experimental conditions
Number of largeneurons/field(Md, range)
Number of smallneurons/field(Md, range)
Sham 8.3 3.5�8.6 2.9 0.7�5.1Ischemia � vehicle 7.8 2.7�10.1 3.5 1.4�5.0Ischemia � P4 5.8 0.5�11.3 3.6 1.0�8.2
Progesterone Neuroprotection in Caudate Nucleus 13
10. Kogure K, Kato H. Altered gene expression in cerebral ischemia.Stroke 1993;24:2121–2127.
11. White BC, Grossman LI, Krause GS. Brain injury by global ischemiaand reperfusion: a theoretical perspective on membrane damage andrepair. Neurology 1993;43:1656–1665.
12. Siesjö BK, Katsura K, Kristián T. The biochemical basis of cerebral is-chemic damage. J Neurosurg Anesthesiol 1995;7:47–52.
13. Chan PH. Role of oxidants in ischemic brain damage. Stroke 1996;27:1124–1129.
14. MacMannus JP, Linnik MD. Gene expression induced by cerebral is-chemia: an apoptotic perspective. J Cereb Blood Flow Metab 1997;17:815–832.
15. Roine RO. Global cerebral ischemia. In: Fisher M, Bougouslavsky J,editors. Current reviews in cerebrovascular disease. Philadelphia, PA,USA: Current Medicine;1997. pp. 159–164.
16. Sarraf-Yazdi S, Lascowitz D, Warner DS. Pathophysiology of is-chemic brain damage. In: Shuaib A, Goldstein LB, editors. Manage-ment of acute stroke. New York: Marcel Dekker;1999. pp. 243–277.
17. Hossmann KA. Post-ischemic resuscitation of the brain: selective vul-nerability versus global resistance. In: Kogure K, Hossman KA, SiesjöBK, Welsh FA, editors. Molecular mechanisms of ischemic brain dam-age. Progress in brain research. Vol. 63. New York: Elsevier;1985. pp.3–17.
18. Kirino T, Tamura A, Sano K. Selective vulnerability of the hippocam-pus to ischemia-reversible and irreversible types of ischemic cell dam-age. In: Kogure K, Hossman KA, Siesjö BK, Welsh FA, editors. Mo-lecular mechanisms of ischemic brain damage. Progress in brainresearch. Vol. 63. New York: Elsevier;1985. pp. 39–58.
19. Pulsinelli WA, Brierley JB, Plum F. Temporal profile of neuronaldamage in a model of transient forebrain ischemia. Ann Neurol 1982;11:491–498.
20. Pulsinelli WA. Selective neuronal vulnerability: morphological and mo-lecular characteristics. In: Kogure K, Hossman KA, Siesjö BK, WelshFA, editors. Molecular mechanisms of ischemic brain damage. Progressin brain research. Vol. 63. New York: Elsevier;1985. pp. 29–37.
21. Crain BJ, Westerkam WD, Harrison AH, Nadler JV. Selective neu-ronal death after transient forebrain ischemia in the Mongolian gerbil:a silver impregnation study. Neuroscience 1988;27:387–402.
22. Kawai K, Nitecka L, Ruetzler CA, Nagashima G, Joó F, Mies G,Nowak TS, Saito N, Lohr JM, Klatzo I. Global cerebral ischemia asso-ciated with cardiac arrest in the rat. I. Dynamics of early neuronalchanges. J Cereb Blood Flow Metab 1992;12:238–249.
23. Rothman S. Synaptic release of excitatory aminoacid neurotransmittermediates anoxic neuronal death. J Neurochem 1984;4:1884–1891.
24. Meldrum B, Garthwaite J. Excitatory amino acid neurotoxicity andneurodegenerative disease. Trends Pharmacol Sci 1990;11:379–387.
25. Choi DW, Koh JY. Zinc and brain injury. Annu Rev Neurosci 1998;21:347–376.
26. Globus MYT, Busto P, Dietrich WD, Martínez E, Valdés I, GinsbergMD. Effect of ischemia on the in vivo release of striatal dopamine,glutamate, and -aminobutyric acid studied by intracerebral microdial-ysis. J Neurochem 1988;51:1455–1464.
27. Slivka A, Brannan TS, Weinberger J, Knott PJ, Cohen G. Increase inextracellular dopamine in the striatum during cerebral ischemia: astudy utilizing cerebral microdialysis. J Neurochem 1988;50:1714–1718.
28. Le Peillet E, Arvin B, Moncada C, Meldrum BS. The non-NMDA an-tagonists, NBQX and GYKI 52466, protect against cortical and striatalcell loss following transient global ischaemia in the rat. Brain Res1992;571:115–120.
29. Hashimoto N, Matsumoto T, Mabe H, Hashitani T, Nishino H. Dopa-mine has inhibitory and accelerating effects on ischemia-induced neu-ronal cell damage in the rat striatum. Brain Res Bull 1994; 33:281–288.
30. Maker HS, Weiss C, Brannan TS. Amine-mediated toxicity. The effectof dopamine, norepinephrine, 5-hydroxytriptamine, 6-hydroxydopa-mine, ascorbate, glutathion and peroxide on the in vitro activities of cre-
atinine and adenylate kinase in the brain of the rat. Neuropharmacol-ogy 1986;25:25–31.
31. Hall ED, Andrus PK, Oostveen JA, Althaus JS, Von Voigtlander PF.Neuroprotective effects of the dopamine D2/D3 agonist pramipexoleagainst postischemic or methamphetamine-induced degeneration of ni-grostriatal neurons. Brain Res 1996;742:80–88.
32. Mordecai Y, Globus T, Busto R, Martínez E, Valdés I, Ginsberg MD.Excitotoxic index—a biochemical marker of selective vulnerability.Stroke 1991;22:128 (Abstract).
33. Lyden PD. GABA and neuroprotection. In: Green AR, Cross AJ, edi-tors. Neuroprotective agents and cerebral ischemia. International re-view of neurobiology. Vol. 40. San Diego, CA, USA: Academic Press;1997. pp. 233–258.
34. Simon RP, Swan JH, Griffiths T, Meldrum BS. Blockade of N-methyl-D-aspartate receptors may protect against ischemic damage in thebrain. Science 1984;226:850–852.
35. Von Lubitz DKEJ, Dambrosia JM, Redmond DJ. Protective effect ofcyclohexyl adenosine in treatment of cerebral ischemia in gerbils.Neuroscience 1989;30:451–462.
36. Li H, Buchan AM. Treatment with an AMPA antagonist 12 hours fol-lowing severe normothermic forebrain ischemia prevents CA1 neu-ronal injury. J Cereb Blood Flow Metab 1993;13:933–939.
37. Crumrine RC, Bergstrand K, Cooper AT, Faison WL, Cooper BR.Lamotrigine protects hippocampal CA1 neurons from ischemic dam-age after cardiac arrest. Stroke 1997;28:2230–2237.
38. Sternau LL, Lust WD, Ricci AJ, Ratcheson R. Role for -aminobutyricacid in selective vulnerability in gerbils. Stroke 1989;20:281–287.
39. Johansen FF, Diemer NH. Enhancement of GABA neurotransmissionafter cerebral ischemia in the rat reduces loss of hippocampal CA1 py-ramidal cells. Acta Neurol Scand 1991;84:1–6.
40. Shuaib A, Ijaz S, Hasan S, Kalra J. Gamma-vinyl GABA prevents hip-pocampal and substantia nigra reticulata damage in repetitive transientforebrain ischemia. Brain Res 1992;590:13–17.
41. Shuaib A, Mazagri R, Ijaz S. GABA agonist “muscimol” is neuropro-tective in repetitive transient forebrain ischemia in gerbils. Exp Neurol1993;123:284–288.
42. Lyden PD, Lonzo L. Combination therapy protects ischemic brain inrats. A glutamate antagonist plus a -aminobutyric acid agonist. Stroke1994;25:189–196.
43. Shuaib A, Ijaz S, Kanthan R. Clomethiazole protects the brain in tran-sient forebrain ischemia when used up to 4 h after the insult. NeurosciLett 1995;197:109–112.
44. Schwartz RD, Yu X, Katzman MR, Hayden-Hixson DN, Perry JM. Di-azepam given postischemia protects selectively vulnerable neurons inthe rat hippocampus and striatum. J Neurosci 1995;15:529–539.
45. Hall ED, Andrus PK, Fleck TJ, Oostven JA, Carter DB, Jacobsen EJ.Neuroprotective properties of the benzodiazepine receptor, partial ago-nist PNU-101017 in the gerbil forebrain ischemia model. J CerebBlood Flow Metab 1997;17:875–883.
46. Ito H, Watanabe Y, Isshiki A, Uchino H. Neuroprotective properties ofpropofol and midazolam, but not pentobarbital on neuronal damage in-duced by forebrain ischemia, based on the GABAA receptors. Acta An-aesthesiol Scand 1999;43:153–162.
47. McGeer PL, McGeer EG. Evidence for glutamic acid decarboxylase-containing interneurons in the neostriatum. Brain Res 1975;91:331–337.
48. Rivak CE, Vaughn JE, Roberts E. The GABA neurons and their axonterminals in rat corpus striatum as demonstrated by GAD immunocy-tochemistry. J Comp Neurol 1979;187:261–266.
49. Gonzales C, Lin RCS, Chesselet MF. Relative sparing of GABAergicinterneurons in the striatum of gerbils with ischemia-induced lesions.Neurosci Lett 1992;135:53–58.
50. Fisher RS, Buchwald NA, Hull CD, Levine MS. The GABAergic stri-atonigral neurons of the cat: demonstration by double peroxidase la-beling. Brain Res 1986;398:148–156.
51. Nitsch C, Riesemberg R. Immunocytochemical demonstration of
14 Cervantes et al./ Archives of Medical Research 33 (2002) 6–14
GABAergic synaptic connections in rat substantia nigra after differentlesions of the striatonigral projections. Brain Res 1988;461:127–142.
52. Jiang N, Chopp M, Stein D, Feit H. Progesterone is neuroprotective af-ter transient middle cerebral artery occlusion in male rats. Brain Res1996;735:101–107.
53. González-Vidal MD, Cervera-Gaviria M, Ruelas R, Escobar A, MoralíG, Cervantes M. Progesterone: protective effects on the cat hippocam-pal neuronal damage due to acute global cerebral ischemia. Arch MedRes 1998;29:117–124.
54. Alkayed NJ, Murphy SJ, Traystman RJ, Hurn PD. Neuroprotective ef-fects of female gonadal steroids in reproductively senescent femalerats. Stroke 2000;31:161–168.
55. Chen J, Chopp M, Li Y. Neuroprotective effects of progesterone aftertransient middle cerebral artery occlusion in rat. J Neurol Sci 1999;171:24–30.
56. Saji M, Reis DJ. Delayed transneuronal death of substantia nigraneurons prevented by -aminobutyric acid agonist. Science 1987;235:66–69.
57. Saji M, Volpe BT. Delayed histologic damage and neuron death in thesubstantia nigra reticulata following transient forebrain ischemia de-pends on the extent of initial striatal injury. Neurosci Lett 1993;155:47–51.
58. Saji M, Cohen M, Blau AD, Wessel TC, Volpe BT. Transient fore-brain ischemia induces delayed injury in the substantia nigra reticulata:degeneration of GABA neurons, compensatory expression of GADmRNA. Brain Res 1994;643:234–244.
59. Yamada K, Goto S, Yoshikawa M, Okamura A, Ushio Y. Involvementof N-methyl-D-aspartate receptor in the delayed transneuronal regres-sion of substantia nigra neurons in rats. Brain Res 1996;743:233–239.
60. Majewska MD, Harison NL, Shwartz RD, Barker JL, Paul SM. Steroidhormone metabolites are barbiturate-like modulators of the GABA re-ceptors. Science 1986;232:1004–1007.
61. Gee KW, Bolger MB, Brinton RE, Coirini H, McEwen BS. Steroidmodulation of the chloride ionophore in rat brain: structure-activity re-quirements, regional dependence and mechanism of action. J Pharma-col Exp Ther 1988;246:803–812.
62. Kirkness EF. Steroid modulation reveals further complexity ofGABAA receptors. Trends Pharmacol Sci 1989;101:6–7.
63. Turner DM, Ransom RW, Yang JSJ, Olsen RW. Steroid anestheticsand naturally occurring analogs modulate the -aminobutyric acid re-ceptor complex at a site distinct from barbiturates. J Pharmacol ExpTher 1989;248:960–966.
64. Hawkinson JE, Kimbrough CL, McCauley LD, Bolger MB, Lan NC,Gee KW. The neuroactive steroid 3�-hydroxy-5�-pregnane-20-one isa two-component modulator of ligand binding to the GABAA receptor.Eur J Pharmacol 1994;269:157–163.
65. Lambert JJ, Belelli D, Hill-Venning C, Peters JA. Neurosteroids andGABAA receptor function. Trends Pharmacol Sci 1995;16:295–303.
66. Todd MM, Chadwick HS, Shapiro HM, Dunlop BJ, Marshall LF,Dueck R. The neurologic effects of thiopental therapy following ex-perimental cardiac arrest in cats. Anesthesiology 1982;57:76–87.
67. Tateishi A, Fleischer JE, Drummond JC, Scheller MS, Zornow MH,Grafe MR, Shapiro HM. Nimodipine does not improve neurologic out-come after 14 minutes of cardiac arrest in cats. Stroke 1989;20:1044–1050.
68. Cervantes M, Ruelas R, Chávez-Carrillo I, Contreras-Gómez A, Anto-nio-Ocampo A. Effects of propofol on alterations of multineuronal ac-tivity of limbic and mesencephalic structures and neurological deficitelicited by acute global cerebral ischemia. Arch Med Res 1995;26:385–395.
69. Barron KD, Schreiber SS, Cova JL, Scheibly ME. Quantitative cy-
tochemistry of RNA in axotomized feline rubral neurons. Brain Res1977;130:469–481.
70. Luna LG. Manual of histologic staining methods of the Armed ForcesInstitute of Pathology. 3rd ed. New York: McGraw Hill;1960. pp.185–300.
71. Snider RS, Niemer WT. A stereotaxic atlas of the cat brain. Chicago,IL, USA: The University of Chicago Press;1961.
72. Poirier LJ, Giguère M, Marchand R. Comparative morphology of thesubstantia nigra and ventral tegmental area in the monkey, cat and rat.Brain Res Bull 1983;11:371–397.
73. Siegel S. Nonparametric statistics for the behavioral sciences. NewYork: McGraw-Hill;1956.
74. Downie MM, Heath RW. Basic statistical methods. New York: Harperand Row;1983.
75. Kubli-Garfias C, Cervantes M, Beyer C. Changes in multiunit activityand EEG induced by administration of natural progestins to flaxedilimmobilized cats. Brain Res 1976;114:71–81.
76. Bitran D, Purdy RH, Kellogg CK. Anxiolytic effect of progesterone isassociated with increases in cortical allopregnanolone and GABAA re-ceptor function. Pharmacol Biochem Behav 1993;45:423–428.
77. Friess E, Tagaya H, Trachsel L, Holsboer F, Rupprecht R. Progester-one-induced changes in sleep in male subjects. Am J Physiol 1997;272(Endocrinol Metab 35:E885–E891).
78. Barnea A, Hajibeigi A, Trant JM, Mason JI. Expression of steroid me-tabolizing enzymes by aggregating fetal brain cells in culture: a modelfor developmental regulation of the progesterone 5-alpha-reductasepathways. Endocrinology 1990;127:500–502.
79. Korneyev A, Guidotti A, Costa E. Regional and interspecies differ-ences in brain progesterone metabolism. J Neurochem 1993;61:2041–2067.
80. Smith SS. Progesterone administration attenuates excitatory aminoacid responses of cerebellar Purkinje cells. Neuroscience 1991;42:309–320.
81. Roof RL, Duvdevani R, Braswell L, Stein DG. Progesterone facilitatescognitive recovery and reduces secondary neuronal loss caused by cor-tical contusion injury in male rats. Exp Neurol 1994;129:64–69.
82. Frye CA. The neurosteroid 3�,5�-THP has antiseizure and possibleneuroprotective effects in an animal model of epilepsy. Brain Res1995;696:113–120.
83. Roof RL, Hoffmann SW, Stein DG. Progesterone protects against lipidperoxidation following traumatic brain injury in rats. Mol Chem Neu-ropathol 1997;31:1–11.
84. Stein DG, Fulop ZL. Progesterone and recovery after traumatic braininjury: an overview. Neuroscientist 1998;4:435–441.
85. Vongher JM, Frye CA. Progesterone in conjunction with estradiol hasneuroprotective effects in an animal model of neurodegeneration.Pharmacol Biochem Behav 1999;64:777–785.
86. Frye CA, Scalise TJ. Anti-seizure effects of progesterone and 3�,5�-THP in kainic acid and perforant pathway models of epilepsy. Psycho-neuroendocrinology 2000;25:407–420.
87. Roof RL, Hall ED. Gender differences in acute CNS trauma and stroke:neuroprotective effects of estrogen and progesterone. J Neurotraum2000;17:367–388.
88. Kumon Y, Kim SC, Pompkins P, Stevens A, Sakaki S, Loftus CM.Neuroprotective effect of postischemic administration of progesteronein spontaneously hypertensive rats with focal cerebral ischemia. J Neuro-surg 2000;92:848–852.
89. Goodman Y, Bruce AJ, Cheng B, Mattson MP. Estrogens attenuateand corticosterone exacerbates excitotoxicity, oxidative injury, andamyloid beta-peptide toxicity in hippocampal neurons. J Neurochem1996;66:1836–1844.
