Effect of γ-hydroxybutyrate in two rat models of focal cerebral damage

Brain Research 986 (2003) 181–190www.elsevier.com/ locate/brainres

Research report

E ffect of g-hydroxybutyrate in two rat models of focal cerebraldamage

a , a b c*Alessandra Ottani , Sabrina Saltini , Marta Bartiromo , Davide Zaffe , Annibale Renzod e e aBotticelli , Anna Ferrari , Alfio Bertolini , Susanna Genedani

aSection of Pharmacology, Department of Biomedical Sciences, University of Modena and Reggio Emilia, via G. Campi 287, 41100Modena, ItalybDepartment of Medical Biotechnology, University of Napoli, Napoli, Italy

cDepartment of Morphological Sciences and Forensic Medicine, University of Modena and Reggio Emilia, Largo del Pozzo 71, 41100Modena, ItalydDepartment of Human Pathology, University of Pavia, Via Forlanini 14, 27100Pavia, Italy

eSection of Clinical Pharmacology and Toxicology, Department of Medicine, University of Modena and Reggio Emilia, Largo del Pozzo 71,41100 Modena, Italy

Accepted 2 July 2003

Abstract

g-Hydroxybutyrate (GHB) and its lactone,g-butyrolactone (GBL) have been previously shown to produce a protective effect in animalmodels of cerebral ischaemia/hypoxia, as well as in human conditions of head injury-induced coma. The aim of the present research wasto study the effect of GHB in experimental conditions of focal cerebral damage, either induced by ischaemia or excitotoxicity. Undergeneral anaesthesia, rats were injected into the right striatum with either endothelin-1 (ET-1, 0.43 nmol) or kainic acid (7.5 nmol) in avolume of 1ml. Sham-lesioned rats received 1ml of the solvent. Both ET-1- and kainic acid-lesioned rats were randomly assigned to one

21of the following intraperitoneal (i.p.) treatments: (i) and (ii) GHB, 100 or 300 mg kg 2 h after the lesion, followed by 50 or 100 mg21 21 21kg , respectively, every 12 h; (iii) saline, 2 ml kg , same schedule. Sham animals were treated with saline, 2 ml kg , same schedule.

Treatments lasted for 10 days. The higher dose of GHB produced a significant protection against the ET-1-induced impairments insensory-motor orientation and coordinated limb use (evaluated 24 and 42 days after the lesion) and in place learning and memory (Morristest, performed 19 and 39 days after the lesion). The same dose regimen reduced the circling behaviour induced by apomorphine inkainate-lesioned rats (10 days after the lesion), and limited or prevented at all the histological damage produced either by ET-1 or bykainic acid (evaluated 43 or 10 days after the lesion, respectively). These results show that GHB limits both histological and functionalconsequences of a focal ischaemic or excitotoxic insult of the brain, in rats, even if the treatment is started 2 h after the lesion. 2003 Elsevier B.V. All rights reserved.

Theme: Disorders of the nervous system

Topic: Ischemia

Keywords: g-Hydroxybutyrate; Cerebral ischemia; Cerebral excitotoxicity; Endothelin-1; Kainic acid; Spatial learning; Sensory-motor test; Circlingbehavior; Immunohistochemistry

1 . Introduction for release and uptake[7,27,39,40,44,47].GHB is uneven-ly distributed in the brain, the highest concentrations being

There is a general agreement on the physiological role found in the hippocampus and in the thalamus, principallyof g-hydroxybutyrate (GHB) in the neuronal transmission in the synaptosomal compartment[7,27,39,40,44].Highin mammals. Specific mechanisms for synthesis have been and low affinity specific binding sites are alike heteroge-thoroughly described in neurons, as well as mechanisms neously distributed, and are most densely located in the

hippocampus and fronto-parietal cortex[21].A protective effect of GHB and its lactone,g-butyrolac-*Corresponding author. Tel.:139-59-205-5382/205-5378; fax:139-

tone (GBL), in animal models of cerebral ischaemia/59-205-5376.E-mail address: [email protected](A. Ottani). hypoxia as well as in human conditions of head injury-

0006-8993/03/$ – see front matter 2003 Elsevier B.V. All rights reserved.doi:10.1016/S0006-8993(03)03252-9

mailto:[email protected]

182 A. Ottani et al. / Brain Research 986 (2003) 181–190

induced coma, has been repeatedly described quite rapidly and is submaximal at the 3rd hour after[10,14,23,48,51].In particular, we have recently shown treatment; 1 week after ET-1 injection the volume of the[48] that in a rat model of transient (30 min) global lesion is rather stabilised around 70% of the maximumcerebral ischaemia (four-vessel occlusion model[34]), the [17]. The microinjection of kainic acid into the brain tissue

21intraperitoneal (i.p.) administration of GHB (300 mg kg produces a biphasic effect. In a first phase, lasting less than30 min before or 10 min after artery occlusion, followed 24 h, kainic acid potently stimulates neuronal activity; in a

21by 100 mg kg twice daily for the following 10 days) second phase there is degeneration of perikarya, whereasproduced a significant protection against the loss of passing axons and ending of extrinsic neurons are leftneurons in the hippocampal CA1 subfield, as well as intact: this histological picture is well evident 10 days afteragainst the ischaemia-induced impairment in a sensory- kainic acid injection[32].motor test battery and in a spatial learning and memorytest. 2 .3. Endothelin-1-induced focal ischaemia and

The aim of our present research was to study the behavioural testspossible effect of GHB in experimental conditions of focalcerebral damage, either induced by ischaemia[17,36]or by 2 .3.1. Sensory-motor orientation and coordinated limbexcitotoxicity [6,20,35]. use

To test sensory-motor orientation and coordinated limbuse on each side of the body, rats were subjected to the test

2 . Materials and methods battery of Marshall and Teitelbaum[28], modified by¨Bjorklund et al. [8], and further slightly modified by

2 .1. Animals ourselves, 24 and 42 days after ET-1-induced focal is-chaemia (that is 14 and 32 days after the last administra-

Adult male rats of an SPF Wistar strain (Harlan Italy, tion of GHB, respectively). In order to test orientation toCorrezzana, Milano), weighing 200–230 g upon arrival, sensory stimuli, the rat was placed on the surface of awere used. They were housed four per cage (40321315 bench. Head orientation towards (with or without biting of)cm), on a 12 h light /dark cycle (light phase:07.00 h–19.00 the stimulus probe was recorded first on one side of theh), with food in pellets (TRM, Harlan Teklab) and tap body and then on the other side for each of the followingwater freely available, under the temperature (2261 8C), stimuli: (a) somaesthesis—a pin prick was applied to sixhumidity (60%) and ventilation conditions advised by the sites, involving combinations of dorsal and ventral place-European Community ethical regulations on the care of ments at rostral, middle and caudal levels on the lateralanimals for scientific research. They were acclimatised to surface of the body; (b) whisker touch—a toothpick wasour housing facilities for at least 1 week before use. lightly brushed against the vibrissae, approaching from the

lower rear of the animal so as to avoid the visual field; (c)2 .2. Induction of focal brain damage snout probe—the toothpick was gently rubbed against the

snout of the rat; (d) olfaction—a small cotton swab dipped21Under ketamine (115 mg kg i.p.) plus xylazine (2 mg in ammonia solution was gently approached in a lateral-

21kg i.p.) anaesthesia, rats were placed in a stereotaxic medial direction towards the rat’s nose.apparatus and randomly injected into the right neostriatum Limb reflexes and coordinated limb use were assessed in(0.9 mm anterior to bregma, 2.5 mm lateral to bregma, 4.3 the following tests: (a) forelimb placement—the rat wasmm ventral to the dura[33]) with either endothelin-1 grasped around the abdomen and slowly lowered in a(ET-1, 0.43 nmol)[17] or kainic acid (7.5 nmol)[20,35], head-down orientation towards the surface of the bench.in a volume of 1ml. Injections were made over 2 min; the The accuracy and coordination of reflex placement of theneedle was left in place for 2 min more and then slowly forelimbs was noted; (b) forelimb suspension—the rat waswithdrawn. ET-1 (Tocris Cookson, Bristol, UK) was grasped by the forepaw ipsilateral to the lesion anddissolved in saline. Kainic acid (Sigma Chemical Co., St suspended. Normal animals rapidly grasp the hand with theLouis, MO, USA) was dissolved in 0.1 M phosphate- free paw and use this to pull themselves up onto the hand;buffered saline (PBS) and adjusted to pH 7.4 with 0.1 N latency to achieve successful pull-up was recorded, with aNaOH. Sham-lesioned animals were injected with 1ml of failure criterion of 10 s; (c) climbing grid—the rat wasPBS (pH 7.4). placed on a vertically clamped wire grid and its climbing

Body temperature was maintained at 37.560.58C with a reflex was observed and rated (immediate50; slow51;rectal thermostar connected to a heating lamp until rats absent52); (d) mouth probe—the rat was held verticallyrecovered thermal homeostasis, that is, at least for 3 h after around the body with head upwards, and a toothpickthe induction of anaesthesia. inserted into the side of the mouth. Normal rats show a

The lesion induced by microinjection of ET-1 into the licking to this stimulus; grasping of the probe with theneostriatum at the dose of 0.43 nmol reaches its maximum ipsilateral forepaw, or attempts to bite the probe werevolume 24 h after the microinjection. However, it develops recorded. The deficit in each orientation and limb-use test

A. Ottani et al. / Brain Research 986 (2003) 181–190 183

was rated on a 3-point scale (absent50; weak51; strong5 side were totally recorded during 1 h. Rotations were2) and a total index score for each rat was obtained. considered as 3608 turns. Results are expressed as the total

number of ipsilateral turns in a 1 hperiod.2 .3.2. Spatial learning and memory

The Morris water maze test[29,30], was used with 2 .5. Histologyminor modifications [48]. This test measures the rat’sability to learn, remember and go to a place in space Kainic acid- or ET-1-microinjected rats (six rats perdefined only by its position relative to distal extramaze group; three groups: sham-operated; lesioned and treatedcues [29]. In our experimental conditions, the apparatus with saline; lesioned and treated with the higher doseconsisted of a circular white pool (1.60 m diameter) filled regimen of GHB) were decapitated under ether anaes-to a depth of 30 cm with water at 2861 8C, rendered thesia, 10 or 43 days, respectively, after the lesion (that isopaque with milk powder. Rats were trained to find the 2 h or 33 days after last administration of GHB). Thespatial location of a platform, made of clear perspex, brains were dissected out and immediately immersed inhidden by arranging for its top surface, 11 cm in diameter, 10% buffered (pH 7.0) formalin solution for 24 to 48 h.to be 1 cm below the water level. Four points on the pool According to the ‘‘Rat Atlas Image Database’’[43],rim (North, South, East, West) defined four 908 quadrants striatum in rat is assessable in section performed at 7 mmon the pool surface (NE, NW, SE, SW) and the platform from the frontal pole. Brains were cut after fixation. Tooccupied a position in the middle of a cardinal quadrant, verify the correct location of rat striatum, a preliminary35 cm from the rim of the pool. The tank was virtually study on an additional six untreated-rat brains was per-divided into four quadrants by two wires intersecting at formed confirming the Atlas indications. So, the brainsright angles. Conspicuous cues (racking, wall plates, door, were sliced at 7 mm from frontal pole and each slicethe observer himself) were placed in a fixed position (thickness 2 mm) was then embedded in a paraffin block,around the pool. On the day before starting training, each serially sectioned (6mm thickness) and collected on glassrat was given 120 s of adaptation to the pool (i.e. the rat slides coated with aminoalkylsilane (Dako, Glostrup, Den-was placed in the pool—without platform—and allowed to mark). Histological sections were stained with Haematox-

¨swim freely with no opportunity for escape). During ylin–Eosin (morphology), Nissl (neurons) and Kluver–training, a trial began when the rat, held facing the side Barrera Luxol Fast Blue (myelin). Immunohistochemicalwall, was immersed in the water. Latency to escape onto methods (antibodies were from Zymed Laboratories, Santhe hidden platform and time spent outside the target Francisco, CA, USA) were applied to highlight the neuro-quadrant were recorded with a stopwatch. If the rat failed nal filaments of neurons [antibody: monoclonal mouseto locate the platform within 60 s it was placed on. The rat anti-neurofilament protein (NPT), clone ZCN37, isotypewas left on the platform for 20 s, then it was removed. IgG1, kappa] and the fibrillary acid protein of normal andEach rat received four consecutive trials on each day, reactive astrocytic cells [antibody: monoclonal mouse anti-starting each time from a different cardinal point in a glial fibrillary acid protein (GFAP), clone ZCG29 isotyperandom succession. In the first 5-day training sequence, IgG1, kappa]. The slides were incubated overnight at 48Cwhich started 19 days after ET-1 microinjection (that is 9 in a moist and darkened chamber. All markers weredays after last administration of GHB), the platform incubated with 1:200 streptavidin biotinylated complexremained in its allocated position. In the second 3-day (Dako, Glostrup, Denmark) for 60 min and developed intraining sequence, which started 39 days after ET-1 diaminobenzidine.microinjection (that is 29 days after last administration of Histomorphological analyses were performed using anGHB), the position of the platform was changed, while the Axiophot photomicroscope (Carl Zeiss Jena, Germany).extramaze cues were maintained in the same position. Morphometrical data were collected using image analysisTests were performed between 10.00 and 13.00 in a by means of a digital camera (DC100) Leica (Leicasoundproof room by an observer unaware of the treat- Microsystems, Heerbrugg, Switzerland), applied to a per-ments. The pool was drained and cleaned each day at the sonal computer connected with a Leica Photomicroscopeend of testing. Laborlux 20 (Germany).

Histometry [1] was evaluated on injured brain areas at2 .4. Kainate-induced neurotoxicity and behavioural test 350 magnification factor on a TV screen. The area of

injured tissue was evaluated using a Sony CCD colour2 .4.1. Rotation behaviour camera on a Leica microscope connected with a personal

Circling behaviour was evaluated by apomorphine chal- computer. The injured tissue area was evaluated on thelenge [35]. Ten days after the intrastriatal injection of serial sections using an image system (SC Image DB,kainic acid, 1 h after last administration of GHB, animals Casti, Italy) and automatically processed by means of an

21were treated with apomorphine (2 mg kg , subcutaneous- image analysis program (Microimage, Casti, Italy).3ly, s.c.; Sigma), placed into individual glass boxes (603 The formula for the calculation of the volume (in mm )

40340 cm), and then ipsilateral rotations to the lesioned was:S [(A 1 A ):2]3 d, were A5area of lesion,2 n i i2n


T able 1n5number of sections andd5interlevel distance along theInfluence ofg-hydroxybutyrate (GHB) on sensory-motor orientation androstrocaudal axis[17].coordinated limb use after focal ischaemic lesion induced by ET-1

Group Cumulative score

2 .6. Drugs and treatment 24 days after ET-1 42 days after ET-1

1. Sham-operated 12 12Either ET-1-lesioned or kainic acid-lesioned rats were 2. ET-11saline 82* 74*

3. ET-11GHB 100 69* 63*randomly assigned to one of the following i.p. treatments:† †21 4. ET-11GHB 300 36* 32*(i) saline, 2 ml kg , 2 h after the brain lesion, and then

21every 12 h for the subsequent 10 days; (ii)g-hydroxy- Saline, 2 ml kg , was i.p. administered 2 h after the brain lesion, andthen every 12 h for the subsequent 10 days;g-hydroxybutyrate (GHB)butyrate (GHB, Laboratorio Farmaceutico C.T., Sanremo,

2121 was i.p. administered at a dose of 100 or 300 mg kg , 2 h after the brainItaly), 100 mg kg , 2 h after the brain lesion, and then 50 21lesion or sham-lesion, and then at a dose of 50 or 100 mg kg ,21mg kg every 12 h for the subsequent 10 days; (iii) GHB, respectively, every 12 h for the subsequent 10 days. Data are presented as

21300 mg kg 2 h after the brain lesion, and then 100 mg cumulative scores obtained from 12 rats per group. *P,0.05 compared21 †kg every 12 h for the subsequent 10 days. GHB was with sham-operated group;P,0.05 compared with ET-11saline group

(Friedman’s test followed by the Mann–WhitneyU-test).dissolved in saline immediately before treatment, and21injected in a volume of 2 ml kg . Sham brain-lesioned

21rats were i.p. treated with saline, 2 ml kg, 2 h after 3 .2. Spatial learning and memorysuture of the skin, and then every 12 h for the subsequent10 days. The doses of GHB were chosen on the basis of The focal, unilateral ischaemic lesion produced in theprevious experiments performed in our laboratories using striatum by the microinjection of ET-1 caused a severerats of the same strain, in a different condition of brain impairment in place learning and memory. The effect wasischaemia[48]. In preliminary experiments, GHB, at the particularly evident in the first 5-day training session (Fig.doses used in this study, had no significant influence on 1), which started 19 days after ET-1 injection (that is 9sensory-motor performance and learning and memory of days after the last administration of GHB), but was stillsham-operated rats. significant in the second 3-day training session (Fig. 2),

which started 39 days after ET-1 injection (that is 29 daysafter last administration of GHB). Both the latency to

2 .7. Statistical analysis escape onto the hidden platform, and the time spent outsidethe quadrant where the platform was located, did not

Data for spatial learning and memory and rotation improve during the first session (except for the 4th day),behaviour were analysed using analysis of variance and only slightly improved during the second session.(ANOVA) followed by Bonferroni’s test. Data for sensory- On the other hand, while treatment with GHB at themotor test were analysed using Friedman’s test followed lower dose regimen had no significant effect, treatment

21by the Mann–WhitneyU-test. Data concerning the volume with GHB at the higher dose regimen (300 mg kg , i.p., 221of the injured tissue were analysed using the Mann– h after the lesion, followed by 100 mg kg , i.p., twice

Whitney U-test. daily for 10 days) caused a progressive, significant reduc-tion of both latency to escape and time spent outside thequadrant. The effect was particularly evident during thefirst session (latency to escape and outside time: day 3:

3 . Results F 515.76,P50.000 andF 519.68,P50.000; day 4:3,44 3,44

F 514.99,P50.000 andF 510.76,P50.000; day 5:3,44 3,44

3 .1. Sensory-motor orientation and coordinated limb use F 516.68,P50.000 andF 514.07,P50.000) (Fig.3,44 3,44

1); during the second session the effect was impressive onAs shown inTable 1,the focal ischaemic lesion induced day 2 (latency to escape and outside time:F 56.23,3,44

by ET-1 produced a significant impairment in sensory- P50.001 andF 58.35, P50.000, respectively), while3,44

motor orientation and coordinated limb use, evaluated 24 it was no longer seen on day 3 (latency to escape and2(rx 530.6 P50.000; T54.192, P50.000) and 42 days outside time:F 58.00, P50.000 andF 55.80, P53,44 3,442(rx 523.3, P50.000; T54.192, P50.000) after the 0.002, respectively) (Fig. 2).

microinjection into the right striatum (that is 14 and 32days after last administration of GHB, respectively). 3 .3. Rotation behaviourTreatment with GHB at the lower dose regimen wasineffective, whereas the higher dose regimen afforded a Following apomorphine challenge, 10 days after thesignificant protection at either interval after the ischaemic intrastriatal injection of kainic acid (1 h after last treatmentlesion (T54.07,P50.000 andT53.26,P50.001, respec- with GHB), lesioned rats made 275.67659.86 ipsilateraltively). turns/h. Treatment with GHB at the lower dose regimen


Fig. 1. Effect of g-hydroxybutyrate (GHB) in the first 5-day trainingsession of Morris water maze, 19 days after endothelin-1 (ET-1) Fig. 2. Effect ofg-hydroxybutyrate (GHB) in the second 3-day training

21injection. SHAM5sham-operated; SAL5saline, 2 ml kg 2 h after session of Morris water maze, 39 days after endothelin-1 (ET-1)21ET-1, and then every 12 h for the subsequent 10 days. GHB5100 or 300 injection. SHAM5sham-operated; SAL5saline, 2 ml kg 2 h after

21 21mg kg 2 h after ET-1, and then 50 or 100 mg kg , respectively, every ET-1, and then every 12 h for the subsequent 10 days. GHB5100 or 30021 2112 h for the subsequent 10 days. Values are the mean6S.E.M. of the mg kg 2 h after ET-1, and then 50 or 100 mg kg , respectively, every

escape latencies and the outside of the quadrant times; 12 rats per group.12 h for the subsequent 10 days. Values are the mean6S.E.M. of the[*P,0.05 compared with sham-operated group;P,0.05 compared with escape latencies and the outside of the quadrant times; 12 rats per group.

[ET-1-lesioned group (ANOVA followed by Bonferroni’s test). *P,0.05 compared with sham-operated group;P,0.05 compared withET-1-lesioned group (ANOVA followed by Bonferroni’s test).

produced a non-significant reduction of the circling be-haviour (241645.5 turns/h), whereas the reduction was T able 2highly significant with the higher dose regimen of GHB Influence ofg-hydroxybutyrate (GHB) on apomorphine-induced rotation

behaviour, 10 days after intrastriatal injection of kainic acid(28.1768.65 turns/h) (F 514.11,P50.000) (Table 2).3,44

Group Ipsilateral turns/60 min3 .4. Histology

1. Sham-operated1saline 0.1760.112. Kainic acid1saline 275.67659.86

In all rats with brain ischaemic injury induced by the 3. Kainic acid1GHB 100 241.0645.504. Kainic acid1GHB 300 28.1768.65*microinjection of ET-1, there was focal coagulative necro-

21sis of putamen associated with a loss of a number of Saline, 2 ml kg , was i.p. administered 2 h after the brain lesion, andneurons and demyelination of white matter at the level of then every 12 h for the subsequent 10 days;g-hydroxybutyrate (GHB)

21was i.p. administered at the dose of 100 or 300 mg kg , 2 h after thethe of the right striatum zone. Several neurons showed an21brain lesion, and then at the dose of 50 or 100 mg kg , respectively,intensely eosinophilic cytoplasm with numerous large

every 12 h for the subsequent 10 days. Data are presented asvacuoles and pale or unstained nucleus without nucleolus.mean6S.E.M. of the number of ipsilateral turns in a 1 h period; 12 ratsSome neurons were without nucleus and showed paleper group. *P,0.05 compared with kainic acid1saline and kainic acid1cytoplasm (ghost cells).The white matter showed irregular GHB 100 groups (ANOVA followed by Bonferroni’s test).


T able 3¨pallor after staining with Kluver–Barrera for myelinEffect of g-hydroxybutyrate (GHB) on the volume of the endothelin-1- orbecause of disintegration of myelin sheaths and swollenkainate-induced lesion in the neostriatum

axons. The peripheral margins of the necrotic area pre-3Volume (mm )sented irregular demarcation and were well stained with

¨Kluver–Barrera for myelin and GFAP for astrocytes. The Saline GHB 300

neuropil appeared spongy as a consequence of astrocytosisEndothelin-1 2.9960.10 1.5160.006*and swelling of protoplasmic processes. Several Kainic acid 2.3060.16 1.1460.13*phagocytes were also present around necrotic areas. 21Saline, 2 ml kg , was i.p. administered 2 h after endothelin-1 or kainic

Fig. 3 shows an area of demyelinated tissue, consequentacid injection into the right neostriatum, and then every 12 h for theto ET-1-induced ischaemia. Forty three days after ET-1 subsequent 10 days;g-hydroxybutyrate (GHB) was i.p. administered at

21the dose of 300 mg kg , 2 h after endothelin-1 or kainic acid injection,microinjection (that is 33 days after last administration of21followed by 100 mg kg every 12 h for the subsequent 10 days; six ratsGHB), necrotic areas of variable size, demyelination and

per group. Brains were removed 43 or 10 days after endothelin-1- orgliosis were observed in the right striatum of all saline- kainate-induced lesion, respectively. *P,0.05 compared with salinetreated rats. The mean volume of the lesion was sig- group (Mann–WhitneyU-test).nificantly reduced in rats treated with the higher doseregimen of GHB compared with rats treated with saline days later (that is 2 h after last administration of GHB)

3(1.5160.06 and 2.9960. 1 mm , respectively;T521) (Fig. 4), were represented by severe lesions in all animals(Table 3). The consequences of the microinjection of (neuronal degeneration, astrogliosis, high degree of de-kainic acid in saline-treated rats, histologically verified 10 myelination with presence of myelin debris). The mean

Fig. 3. Endothelin-1 (ET-1)-induced ischaemic damage in the neostriatum. Brains were removed 43 days after the lesion. The image shows the injured(demyelinated) tissue of the right striatum (light grey area surrounded by arrows), due to ET-1-induced ischaemia, in rats treated with only saline.Hematoxylin–Eosin stain. Field width52.5 mm.


Fig. 4. Kainic acid-induced damage in the neostriatum. Brains were removed 10 days after the lesion. The squared areas correspond to the field of the highmagnification inserts on the bottom of each image. Demyelinated tissue (arrows) is detectable in A, and presence of myelin debris is highlighted in theAainsert. In B, a large extent of the injured tissue is shown. Neuronal ghosts are shown in the Ba insert and perinecrotic gliosis (separated from the injured

¨tissue by a superimposed dotted line) is shown in the Bb insert; Kluver–Barrera (A and Aa inserts); Hematoxylin–Eosin stain (B insert); NPTimmunoreaction (Ba insert); GFAP immunoreaction (Bb insert). Scale bars: A5500 mm; B5100 mm; a and b550 mm.

volume of the lesion in brain from GHB-treated rats was 4 . Discussionsignificantly reduced in comparison with rats treated with

3saline (1.1460.13 and 2.360.16 mm , respectively;T5 Our present results show thatg-hydroxybutyrate (GHB)21) (Table 3). produces a significant reduction of both histological and


functional damage that are induced in rats by an episode of GBL prevented the development of cerebral hyperaemiafocal ischaemic or excitotoxic insult. and, later, the prolonged cerebral hypoperfusion that was

An important detail is that such protective effect is experienced by the saline-treated animals; moreover, theobtained starting the treatment 2 h after the focal damage; neuronal loss in striatum and hippocampus was signifi-that is, we are not dealing with a preventive effect but cantly reduced in GBL-treated rats compared to controlsprobably with a curative activity of this drug in such [23]. In a rat model of embolic stroke, the development ofexperimental conditions. brain oedema 12 h later was inhibited by GBL treatment

In rats where the damage was the consequence of the[10]. Finally, in comatose head-injured patients GBL wasischaemia produced by the unilateral intrastriatal microin- more effective than barbiturates in lowering the cerebraljection of ET-1, the histological lesions found 43 days later metabolic rate of oxygen, without depressing cerebralwere slight or absent completely in GHB-treated animals, perfusion pressure (a life-threatening side-effect of barbitu-whereas they were well evident and severe in all saline- rates[14]).treated ones. Quite similar was the picture in rats where the We previously showed[48] that GHB affords a signifi-damage had been produced by the unilateral intrastriatal cant protection against both histological and behaviouralmicroinjection of kainic acid; in this case, too, the his- consequences of a transient (30 min) global cerebraltological lesions found 10 days later were severe in all ischaemia in rats, produced by means of the four-vesselsaline-treated animals, while they were much reduced or occlusion model. In such a previous study, as in presentabsent completely in GHB-treated ones. experiments, the protective effect of GHB was observed

21 21Functional data are completely consistent with, and with a dose of 300 mg kg i.p., followed by 100 mg kgconfirm, the histological picture. Treatment with GHB i.p. twice daily for the following 10 days. But in that study

21significantly limited sensory-motor and learning and mem- the first dose (300 mg kg ) was injected either 30 minory deficiencies caused by the ET-1-induced ischaemic before or 10 min after artery occlusion; the degree oflesion, as well as the ipsilateral circling behaviour precipi- protection being similar in either case.tated by apomorphine in kainate-lesioned rats: both the The present data confirm and extend some previousrole of striatum in learning and memory (particularly ones, particularly those obtained by Artru et al.[2] and byworking and reference memory), and the importance of Lavyne et al.[23], as well as those produced by ourselvesstriatal lesions in cognitive deficits are well established, [48]. In addition, our present data show that GHB iseither in experimental animals or in humans effective in different models of cerebral insult, and that it[3,11,16,24,31,37,42,50].On the basis of these results, it is able to significantly attenuate the neuronal damage,would seem that GHB ameliorates both learning and whatever produced. Most important, our present data showmemory; however, memory seems to be more positively that the treatment with GHB is effective even if startedaffected, especially on the basis of the results of the 2nd after a rather long time-lag.training sequence. At present we can only speculate about the possible

In rats, following i.v. administration, the plasma con- mechanisms of such activity of GHB. The followingcentrations of GHB undergo a first phase of rapid decline, established features of the pharmacodynamics of GHBfollowed by a slower second phase. After doses of 150– may be involved: (i) GHB reduces glucose and high

21300 mg kg pharmacokinetic parameters are not sig- energy phosphate utilisation rates in the cerebral tissuenificantly different. Maximum concentrations, immediately [51], and increases glycogen and glucose—while decreas-after i.v. injection, are comprised between 700 to 1000mg ing pyruvate and lactate—contents in the brain[19,25],

21ml . The average pooled values are forV 206861400 this lowers oxygen demand and consumption by the brain;max21 21 21

mg min kg , K 58616 mg ml andV 476612 ml (ii) GHB has oxygen radical scavenging properties[9,12];m d21kg [46]. So, GHB has a small apparent volume of (iii) in the range of doses used in our experiments, GHB

distribution and a relatively fast distribution phase. Mich- reduces the firing rate of neurons and causes hypothermiaaelis–Menten elimination kinetics for GHB have also been [18,22]; (iv) at these dose levels, GHB binds GABAB

reported in the dog[38], the cat [41] and in man at receptors besides its own receptors[13,26]: activation ofnon-anaesthetic doses (used in the treatment of alcohol both GHB and GABA receptors may reduce the releaseB

dependence)[15]. The V in rats is approximately 10- of excitatory amino acids[5,7]; (v) GHB causes hyperpo-max21 21times greater than in man (70–200mg min kg ) [45]. larisation of neurons by activating GABA receptors andB

1Finally, GHB easily crosses the blood–brain barrier after by increasing K conductance[52]: the effectiveness inperipheral administration. stroke of drugs that increase GABAergic transmission and

1Only a few previous studies concerning the effect of of K channel openers is under clinical scrutiny[4,49].GHB or of its lactone GBL in experimental conditions of In view of the effectiveness of GHB treatment even ifcerebral ischaemia, or in clinical conditions of cerebral started 2 h after the brain insult, as shown by our presentinjury, are available. In a mouse model of cerebral data, it will be necessary to investigate possible influenceshypoxia, GHB and GBL afforded a moderate protection of this drug on the late consequences of tissue damage.[2]. In a rat model of transient global cerebral ischaemia, In conclusion, our results confirm that GHB is able to


action on presynaptic GABA receptors, Neurosci. Lett. 216 (1996)significantly limit the histological and functional outcomes B

121–124.of focal, permanent brain lesion, be they produced by[14] E . Escuret, B. Roquefeuil, P. Frerebean, M. Baldy-Moulinier, Effect

ischaemia or excitotoxicity. And this can be achieved of hyperventilation associated with administration of central nervousstarting the treatment 2 h after the brain insult. Particularly depressant in brain injuries, Acta Neurol. Scand. 56 (1977) 154–in view of this last datum, as well as of the quite low 155.

[15] S .D. Ferrara, S. Zotti, L. Tedeschi, G. Frison, F. Castagna, L.toxicity of this drug, it would be advisable to extend theseGallimberti, G.L. Gessa, P. Palatini, Pharmacokinetics of gamma-studies to other species, and possibly to patients.hydroxybutyric acid in alcohol dependent patients after single andrepeated oral doses, Br. J. Clin. Pharmacol. 34 (1992) 231–235.

[16] S .B. Floresco, D.N. Braaksma, A.G. Phillips, Thalamic-cortical-striatal circuitry subserves working memory during delayed respond-A cknowledgementsing on a radial arm maze, J. Neurosci. 19 (1999) 11061–11071.

[17] K . Fuxe, N. Kurosawa, A. Cintra, A. Hallstrom, M. Goiny, L.This work was supported in part by grants from Minis-Rosen, L.F. Agnati, U. Ungerstedt, Involvement of local ischemia in

`tero dell’Istruzione dell’Universita e della Ricerca endothelin-1 induced lesions on the neostriatum of the anaesthetized(MIUR), Roma. rat, Exp. Brain Res. 88 (1992) 131–139.

[18] S .R. Godbout, P. Jelenic, C. Labrie, M. Schmitt, J.J. Bourguignon,Effect of gamma-hydroxybutyrate and its antagonist NCS-382 onspontaneous cell firing in the prefrontal cortex of the rat, Brain Res.

R eferences 673 (1995) 157–160.[19] Y . Godin, J. Mark, P. Mandel, The effects of 4-hydroxybutyric acid

on the biosynthesis of amino acid in the central nervous system, J.[1] J .M. Anderson, J. Lowe, Histometry and image analysis, in: J.D.Neurochem. 15 (1968) 1085–1091.Bancroft, A. Stevens (Eds.), Theory and Practice of Histological

[20] E . Gratacos, E. Rerez-Navarro, E. Tolosa, E. Arenas, J. Alberch,Techniques, Churchill Livingstone, Edinburgh, UK, 1990, pp. 597–Neuroprotection of striatal neurons against kainate excitotoxicity by618.neurotrophins and GDNF family members, J. Neurochem. 78 (2001)[2] A . Artru, P. Steen, J. Michenfelder, Gamma-hydroxybutyrate: cere-1287–1296.bral, metabolic, vascular and protective effects, J. Neurochem. 35

(1980) 1114–1119. [21] V . Hechler, D. Weissmann, E. Mach, J.F. Pujol, M. Maitre, Regional3[3] L . Backman, N. Ginovart, R.A. Dixon, T.B. Wahlin, A. Wahlin, C. distribution of high affinityg-[ H]hydroxybutyrate binding site as

Halldin, L. Farde, Age-related cognitive deficits mediated by determined by quantitative autoradiography, J. Neurochem. 49changes in the striatal dopamine system, Am. J. Psychiatry 157 (1987) 1025–1032.(2000) 635–637. [22] E .E. Kaufman, L.J. Porrino, T. Nelson, Pyretic action of low doses

[4] A .M. Basile, P. Nencini, D. Inzitari, Meccanismi molecolari de- of gamma-hydroxybutyrate in rats, Biochem. Pharmacol. 40 (1990)ll’ischemia cerebrale: possibili interventi terapeutici, Ann. Ital. Med. 2637–2640.Int. 14 (1999) 94–105. [23] M . Lavyne, R. Hariri, T. Tankosic, T. Babiak, Effect of low dose

[5] C . Behl, L. Hovey, S. Krajewsky, D. Schubert, J.C. Reed, BCL-2 g-butyrolactone therapy on forebrain neuronal ischemia in unre-prevents killing of neuronal cells by glutamate but not by amyloid strained, awake rat, Neurosurgery 12 (1983) 430–434.beta protein, Biochem. Biophys. Res. Commun. 197 (1993) 949– [24] O . Lehmann, H. Jeltsch, O. Lehnardt, L. Pain, C. Lazarus, J.C.956. Cassel, Combined lesions of cholinergic and serotonergic neurons in

[6] Y . Ben-Ari, E. Tremblay, D. Riche, G. Ghilini, R. Naquet, Elec- the rat brain using 192 IgG-saporin and 5,7-dihydroxytryptamine:trographic, clinical and pathological alterations following systemic neurochemical and behavioural characterization, Eur. J. Neurosci. 12administration of kainic acid, bicuculline or pentetrazole: metabolic (2000) 67–79.mapping using the deoxyglucose method with special reference to [25] V . Macmillan, The effects of gamma-hydroxybutyrate and gamma-the pathology of epilepsy, Neuroscience 6 (1981) 1361–1391. butyrolactone upon the energy metabolism of the normoxic and

[7] R . Bernasconi, P. Mathivet, S. Bischoff, C. Marescaux, Gamma- hypoxic rat brain, Brain Res. 146 (1978) 177–187.hydroxybutyric acid: an endogenous neuromodulator with abuse [26] M . Maitre, The gamma-hydroxybutyrate signalling system in brain:potential?, Trends Pharm. Sci. 20 (1999) 135–141. organization and functional implications, Prog. Neurobiol. 51 (1997)

[8] A . Bjorklund, S.B. Dunnett, U. Stenevi, M.E. Lewis, S.D. Iversen, 337–361.Reinnervation of the denervate striatum by substantia nigra trans- [27] M . Maitre, J.F. Rumigny, C.D. Cash, P. Mandel, Subcellularplants: functional consequences as revealed by pharmacological and distribution ofg-hydroxybutyrate binding sites in rat brain. Principalsensory-motor testing, Brain Res. 199 (1980) 307–333. localization in the synaptosomal fraction, Biochem. Biophys. Res.

[9] A .J. Boyd, I.A. Sherman, F.G. Saibil, M. Mamelak, The protective Commun. 110 (1983) 262–265.effect of g-hydroxybutyrate in regional intestinal ischemia in the [28] F . Marshall, P. Teitelbaum, Further analysis of sensory inattentionhamster, Gastroenterology 99 (1990) 860–862. following lateral hypothalamic damage in rats, J. Comp. Physiol.

[10] J . Bralet, P. Beley, A.M. Bralet, A. Beliz, Influence of various agents Psychol. 86 (1974) 375–395.on the development of brain edema in rat following microembolism, [29] J .R.K. Mcnamara, R.W. Skelton, The neuropharmacological andStroke 10 (1979) 653–656. neurochemical basis of place learning in the Morris water maze,

[11] A . Csillag, Striato-telencephalic and striato-tegmental circuits: rele- Brain Res. Rev. 18 (1993) 33–49.vance to learning in domestic chicks, Behav. Brain Res. 98 (1999) [30] R .G.M. Morris, Development of a water-maze procedure for study-227–236. ing spatial learning in the rat, J. Neurosci. Methods 11 (1984)

[12] R .S. Dosmugambetova, Prevention of stress-related disorders in the 47–60.contractile function of non-ischemic areas of the heart during [31] U . Muller, T. Wachter, H. Barthel, M. Reuter, D.Y. Von Cramon,myocardial infarction using gamma-hydroxybutyric acid, Bull. Esp. Striatal [123I]beta-CIT SPECT and prefrontal cognitive functions inBiol. Med. 96 (1983) 28–30. Parkinson’s disease, J. Neural Transm. 107 (2000) 303–319.

[13] Z . Emri, K. Antal, V. Crunelli, Gamma-hydroxybutyric acid de- [32] M .C. Olianas, G.M. De Montis, G. Mulas, A. Tagliamonte, Thecreases thalamic sensory excitatory postsynaptic potentials by an striatal dopaminergic function is mediated by the inhibition of a


nigral non-dopaminergic neuronal system via a strio-nigral [43] A .W. Toga, E.M. Santori, R. Hazani, K. Ambach, A 3D digital mapGABAergic pathway, Eur. J. Pharmacol. 49 (1978) 233–241. of rat brain, Brain Res. Bull. 38 (1995) 77–85.

[33] G . Paxinos, G. Watson, The Rat Brain in Stereotaxic Coordinates, [44] G . Tunnicliff, Significance ofg-hydroxybutyric acid in the brain,Academic Press, New York, 1984. Gen. Pharmacol. 23 (1992) 1027–1034.

[34] W .A. Pulsinelli, J.B. Brierly, A new model of bilateral hemispheric [45] C .A.M. Van Ginneken, J.M. Van Rossum, H.L.J.M. Fleuren, Linearischemia in the unanesthetized rat, Stroke 10 (1979) 267–272. and nonlinear kinetics of drug elimination. I. Kinetics on the basis of

[35] A . Santamaria, J. Ordaz-Moreno, M. Rubio-Osornio, F. Solis-Her- a single capacity-limited pathway of elimination with or withoutnandez, C. Rios, Neuroprotective effect of dapsone against quino- simultaneous supply-limited elimination, J. Parmacokinet. Bio-late- and kainate-induced striatal neurotoxicity in rats, Pharmacol. pharm. 2 (1974) 395–415.Toxicol. 81 (1997) 271–275. [46] D .K. Van Sassenbroeck, P. De Paepe, F.M. Belpaire, M.T. Rosseel,

[36] J . Sharkey, I.M. Ritchie, P.A.T. Kelly, Perivascular microapplication P. Martens, P.A. Boon, W.A. Buylaert, Relationship between gamma-of endothelin-1: a new model of focal cerebral ischaemia in the rat, hydroxybutyrate plasma concentrations and its electroencephalog-J. Cereb. Blood Flow Metab. 13 (1993) 865–871. raphic effects in the rat, J. Pharm. Pharmacol. 53 (2001) 1687–

[37] S .Y. Shu, X. Bao, S. Li, D. Niu, Z. Xu, Y. Li, A new subdivision of 1696.mammalian neostriatum with functional implication to learning and [47] P . Vayer, P. Mandel, M. Maitre, Gamma-hydroxybutyrate, a possiblememory, J. Neurosci. Res. 58 (1999) 242–253. neurotransmitter, Life Sci. 41 (1987) 1547–1557.

[38] J .S. Shumate, O.C. Snead, Plasma and central nervous system [48] A .V. Vergoni, A. Ottani, A.R. Botticelli, D. Zaffe, L. Guano, A.kinetics of gamma-hydroxybutyrate, Res. Commun. Mol. Pathol. 25 Loche, S. Genedani, G.L. Gessa, A. Bertolini, Neuroprotective(1979) 241–256. effect of gamma-hydroxybutyrate in transient global cerebral is-

[39] O .C. Snead, The gamma-hydroxybutyrate model of absence seizure: chemia in the rat, Eur. J. Pharmacol. 397 (2000) 75–84.correlation of regional brain levels of gamma-hydroxybutyric acid [49] N .G. Wahlgren, The Clomethiazole Acute Stroke Study (CLASS):and gamma-butyrolactone with spike and wave discharges, Neuro- safety results in 1356 patients with acute hemispheric stroke,pharmacology 30 (1991) 161–167. Neurology 50 (1998) 212.

[40] O .C. Snead, C.C. Liu, Gamma-hydroxybutyric acid binding sites in [50] D .S. Wolf, M. Gearing, D.A. Snowdon, H. Mori, W.R. Markesbery,rat and human brain synaptosomal membranes, Biochem. Phar- S.S. Mirra, Progression of regional neuropathology in Alzheimermacol. 16 (1984) 2587–2590. disease and normal elderly: findings from the Nun study, Alzheimer

[41] O .C. Snead, R.K. Yu, P.R. Huttenlocher, Gamma-hydroxybutyrate: Dis. Assoc. Disord. 13 (1999) 226–231.correlation of serum and cerebrospinal fluid levels with electroence- [51] L .I. Wolfson, O. Sakurada, L. Sokoloff, Effects of gamma butyrolac-phalographic and behavioural effects, Neurology 26 (1976) 51–56. tone on local cerebral glucose utilization in the rat, J. Neurochem.

[42] G .T. Stebbins, J.D. Gabrielli, K.M. Shannon, R.D. Penn, C.G. 29 (1977) 777–783.Goetz, Impaired frontostriatal cognitive functioning following post- [52] X . Xie, T.G. Smart,g-Hydroxybutyrate hyperpolarizes hippocampaleroventral pallidotomy in advanced Parkinson’s disease, Brain neurons by activating GABA receptors, Eur. J. Pharmacol. 212B

Cogn. 42 (2000) 348–363. (1992) 291–294.

Documents

Effect of γ-hydroxybutyrate in two rat models of focal cerebral damage