D E V E L O P M E N T OF A S Y N A P T O S O M A L M O D E L TO D E T E R M I N E D R U G - I N D U C E D IN VIVO C H A N G E S
IN GABA-LEVELS OF N E R V E E N D I N G S IN 11 B R A I N REGIONS OF THE RAT
WOLFGANG L6SCHER*, MARTIN VETTER*§, FRANK MULLER*, GERHARD B(3HME~ and GISELA STOLTENBURG-DIDINGER~
*Laboratory of Pharmacology and Toxicology, tDepartment of Anatomy, Histology and Embryology, School of Veterinary Medicine, Free University Berlin and SDepartment of Neuropathology, Klinikum
Steglitz, Free University Berlin, West Germany
(Received 25 November 1983; accepted 20 December 1983)
Abstract--An experimental procedure was developed which allowed the simultaneous measurement of GABA in synaptosomes from 11 regions of one rat brain. Synaptosomal fractions were prepared by conventional subcellular fractionation procedures and characterized by electron microscopy. Post-mortem increases of GABA during removal and dissection of brain tissue, homogenization and fractionation procedures could be sufficiently minimized by rapid processing of the tissue at low temperatures and inclusion of 3-mercaptopropionic acid (1 mM) in the homogenizing medium. Experiments with addition of aminooxyacetic acid (AOOA, 1 mM) to the homogenizing medium indicated that GABA was not being degraded during synaptosome preparation. The presence of exogenous GABA (1 mM) did not alter the GABA levels in the organelles, indicating that no significant redistribution of GABA occurred during subcellular fractionation. On the basis of these findings, it was suggested that synaptosomal fractions could be used as a model to monitor indirectly the drug-induced changes in GABA levels of nerve endings in discrete brain areas of the intact animal. In vivo experiments with AOAA (30 mg/kg i.p.) and valproic acid (VPA, 200 mg/kg i.p.) showed that both drugs caused differential effects on synaptosmal GABA levels in different brain regions. Although AOAA was more potent than VPA in increasing GABA in whole tissue of most brain regions, significant increases of synaptosomal GABA levels after AOAA were only determined in olfactory bulbs and frontal cerebral cortex. In contrast, VPA induced significant synaptosomal GABA increases in olfactory bulbs, hypothalamus, superior and inferior colliculus, substantia nigra, and cerebellum. The data indicate that the synaptosomal model can provide useful information on the in vivo effects of drugs on GABA levels in nerve terminals and their ability to exert this effect in specific brain areas.
Understanding of the physiological role of v-aminobutyric acid (GABA), the principal in- hibitory substance in mammalian CNS, has been greately enhanced by using drugs which selectively alter GABA-media ted neurotransmission and study- ing associated functional effects. In this respect, drugs which decrease or increase G A B A concentrations in the brain have proved of special value. As a result of such studies, evidence has been accumulated to sug- gest that GABAergic systems are involved in a range of CNS-functions including control of brain ex-
Address correspondence to: Dr W. Lrscher, Laboratory of Pharmacology and Toxicology, School of Veterinary Medicine, Free University Berlin, Koserstrasse 20, D-1000 Berlin 33, West Germany.
§The results of this study are part of the Thesis of M. Vetter.
citability, feeding behaviour, cardiovascular function, hormonal secretion, sleep mechanisms, and ag- gressive behaviour (cf. Krogsgaard-Larsen et al.,
1979). However, several characteristics of the G A B A system complicate the interpretation of pharma- cological and neurochemical studies of GABAergic function. The most complicating characteristic is certainly the compartmentat ion of G A B A in brain tissue. Thus, synthesis, degradation and uptake of G A B A take place in a variety of cellular compart- ments, both neuronal and non-neuronal (Baxter, 1976). As a result, the degree to which drug-induced changes in endogenous G A B A can be expected to influence GABA-media ted synaptic transmission is difficult to ascertain. The most likely site for a change in G A B A to exert a functional impact would be in GABAergic nerve terminals. Unfortunately, the di- rect measurement of G A B A levels in nerve terminals
441
442 WOLF(iANG L(3SCHER et al.
is beyond the scope of current technology. On the other hand, recent data from several laboratories including our own indicate that the GABA content of synaptosomes, i.e. discrete particles which are derived from brain nerve endings, accurately reflect changes in nerve terminal GABA levels brought about in the intact animal by the administration of drugs (Abe and Matsuda, 1977; Abe, 1978: Wood et al., 1978b; Matsuda et al., 1979: Sarhan and Seiler, 1979: Wood et al., 1979 and 1980; L6scher, 1981a,b,c: L6scher, et
al., 198l). In fact, studies using this "'synaptosomal model" have provided important insights into the in
HLo effect of drugs on the compartmentation o f
GABA and could resolve some of the discrepancies regarding GABA levels versus brain excitability in the literature (e.g. Wood et al., 1979). However, so far all these studies employed synaptosomes isolated either from the whole brain or brain cortex of mice, thus disregarding regional differences in GABA changes of nerve terminals. In view of the variations in GABA content of different brain regions and the fact that GABA increases in different nuclei may play mutually opposite roles, it is evident that data ob- tained from measurement of GABA in the whole brain or synaptosomes isolated from the whole brain can only provide limited functionally interpretable information.
In order to improve the synaptosomal model, it thus seemed mandatory to extend the isolation of
synaptosomes and determination o1" GABA therein to various brain areas. For practical reasons, all pro- cedures necessary in this respect should be sensitive and reproducible enough to allow the simultaneous measurement of synaptosomal GABA levels in dis- crete brain areas of the same animal. During devel- opment of a suitable model, a number of possible problems required evaluation, including postmortem increases in the content of GABA during removal and dissection of tissue and preparation of hom- ogenates, redistribution of GABA during homogen- ization of the tissue and subcellular fractionation procedures, and metabolism of GABA in syn- aptosomes during their preparation. The present paper focuses on these issues and describes results with 2 GABA-e[evating agents, namely amino- oxyacetic acid (AOAA) and the antiepileptic valproic acid (VPA), on regional GABA levels in synaptosomes of rat brain.
EXPERIMENTAL PROCEDURES
Animal,~
Female rats of the Wistar strain (Winkelmann Ver- suchstierzucht, Borchen. F.R.G.), weighing 180-230 g, were used. They were kept in groups of 10 in Makrolon" cages at an ambient temperature of 24-.26 C and ted on Altromin" 1324 standard food (Altromin, Lage, F.R.G.).
Drug.s used
Valproic acid (VPA), used as the sodium salt, was kindly provided by Desitin-Werk Carl Klinke (Hamburg, F.R.G.).
A29
Fig. 1. Sections lines used for dissection of rat brain; dorsal view on left, ventral on right. The location and designation of the respective lines was taken from Zeman and lnnes (1963).
Drug-induced changes in synaptosomal GABA content 443
Aminooxyacetic acid hemihydrochloride (AOAA) was purchased from Sigma (Munich, F.R.G.) and 3-mercaptopropionic acid (3-MP) from Merck (Darmstadt , F.R.G.). For animal experiments, the drugs were freshly dissolved in water and administered intraperitoneally in a volume of 2 ml/kg. Controls received 2 ml/kg saline.
Dissection o f brain regions
Rats were killed by decapitation and the brains were rapidly removed and dissected on a cold plate at - 18°C into 11 regions within 4 min after decapitation. In detail, opening of the cranium started from the foramen magnum. After removal of the calotte, the brain was taken out in toto and placed on a cold plate at - 18'C (Leitz Kryomat , Wetzlar, F.R.G.) within 30 s after decapitation. The olfactory bulbs, which remained in the ethmoid fossae, were scooped out with a sharp spoon. For further dissection of the brain at
- 18C, we used the section lines described by Zeman and Innes (1963) as shown in Fig. 1. By a transverse incision at A22 the brain was separated into the rostrally located part A and the caudally located part B. Part A contained the telencephalon and portions of the diencephalon; part B consisted largely of midbrain, rhombencephalon and cere- bellum. Intermedian penetration with a sharp spoon 0 3 mm) approx 3 mm into part A was effected from the cut surface about 2 m m above the base of the brain. The paramedian tissue thus contained the main portions of the hypothalamus, the 3rd ventricle being located in the middle. At A29, part A, an incision was then made on the dorsal surface over the entire width, proceeding rostro-ventrally in such a way that only the cortex was resected. Thus the frontal cortex was obtained. The remaining part A was now divided into 2 halves by sagittal section. In connection with the first incision at A, 2 and the removal of the hypo- thalamus, this sagittal section resulted in caudo-medial corner zones which contained the thalamus. The corner regions were bilaterally isolated by an oblique section. The corpora striata, which were now clearly visible, were medi- ally scooped out from the remaining halves of part A with a sharp spoon (0 4 mm). Then, an oblique section was made at At7 on part B to obtain a tissue slice of 1-2 mm thickness. The rostral portions of the hippocampus, which were clearly recognizable on the newly cut surface, were removed from this slice. The remaining portions of the neocortex and the hippocampus were then removed from part B, thereby exposing the mesencephalon with the tectum, the teg- mentum and the crura cerebri. The substantia nigra located above the crura cerebri was dissected by setting the scalpel to the lower third of the transverse cut surface Al7 for bilateral excision of a tissue wedge in a caudo-ventral direction. A further transversal incision at A 9 separated the rostrally located mesencephalon, and the tectum (superior and inferior colliculus) was obtained by a horizontal section. Finally, the medulla oblongata was dissected by a trans- versal incision at As, and the cerebellum was separated from the pons by a horizontal incision into the remaining tissue at the level of the 4th ventricle.
During elaboration of the dissection technique, the ex- cised tissue samples were cut into 5-10 # sections, stained with hematoxylin-eosin, and used for microscopic confirmation of the different brain areas.
Determination o f regional GABA levels in whole tissue
Following dissection, the individual regions were rapidly weighed and homogenized by an Ultra-Turrax in 2 ml of
80~o ethanol (tubes immersed in a bath of methanol at - 30°C). The max imum time from decapitation to homoge- nization of a region in ethanol was 4 min. The tubes were then vigorously agitated by a Reax I shaker (Heidolph, Kelheim, F.R.G.) for 2 min and centrifuged. After centrifu- gation at 4000rpm for 10min at - 5 ° C , the pellet was resuspended in 1 ml 80~ ethanol and again centrifuged. The combined supernatants were evaporated to dryness by a stream of nitrogen in a water bath at 37°C. The residue was dissolved in 0.5ml pyrophosphate buffer (0.1 M, pH 8.3) and GABA was measured in aliquots of 0 .2ml by the enzymatic "GABAase" method as described by Baxter (1972). Protein content of each brain region was determined in 100/~1 aliquots of the initial homogenates by the method of Lowry et al. (1951 ) as modified by Markwell et al. (1978). To examine if postmortem increase of GABA occurred during the dissection, 5 rats were killed by decapitation 2.5 min after i.p. injection of 3-MP (100 mg/kg), an inhibitor of G A B A synthesis (cf. van der Heyden and Korf, 1978: Carmona et al., 1980), and regional GABA levels were compared with those determined in a control group decap- itated at the same time without 3-MP pretreatment.
Preparation o f synaptosomal fractions Immediately after dissection, the individual regions of
each rat brain were rapidly weighed and homogenized in 1 ml 0.32 M sucrose (pH 7, 4°C) by means of a motor-driven Potter-Elvehjem glass homogenizer with Teflon pestle. As already described above for homogenization of regions in ethanol, the max imum time from decapitation to homo- genization was 4min . The homogenization and all sub- sequent steps were carried out at 0 ~ ° C . Unless otherwise indicated, the 0.32 M sucrose used as homogenizing medium contained 1 m M 3-MP to inhibit G A B A synthesis during synaptosome preparation. Since 3-MP addition altered the pH of the sucrose medium, it had to be readjusted to pH 7.0 with dilute NaOH. The procedure used for preparation of the synaptosomal fractions was modified from that de- scribed by Cotman (1974) for rat cortex. For separation into primary fractions, the homogenates were centrifuged at 1000g for 5min. The resultant pellet (P~, nuclei, large myelin fragments, tissue debris) was discarded and the supernatant fluid (S 0 was centrifuged in a Beckman Spinco L 50 ultracentrifuge with a 50 Ti rotor at 15,000 g for 12 min to precipitate the P2 fraction (mitochondria, synaptosomes, small myelin fragments, some microsomes). This fraction was resuspended in 1 ml 0.32 M sucrose and applied to a Ficoll-sucrose gradient consisting of 1 ml 4~ , 2 ml 60/0, and 2ml 13~o Ficol in 0.32 M sucrose (adjusted to pH 6.5-7.5 with dilute NaOH) in 10 ml polycarbonate tubes. The tubes were centrifuged at 64,000g for 45 min in a Beckmann L 8 ultracentrifuge with a fixed angle 30.2 rotor using the slow acceleration facility. After centrifugation, the synaptosomal fraction was collected from the 6/13~o interface. The frac- tion thus obtained was diluted to a final volume of 4 ml with 0.32 M sucrose and centrifuged at 50,000g for 20 min with the 50 Ti rotor to yield a pellet of the synaptosomal fraction.
For morphological examination of the synaptosomal fractions, the respective synaptosomal pellets were pro- cessed essentially as described by Wood et al. (1979). Briefly, pellets were fixed for 18 h in 5~o glutaraldehyde (in 0.2 M S6rensen phosphate buffer, pH 7.4, 320 mosmol) at 0 C in situ in the tubes. After a rinse with buffer, the pellets were dislodged and postfixed for 5 h in buffered 2.5~o osmium
444 WOLFGANG LrSCHE~ et al.
tetrodoxide. The pellets were rinsed twice with buffer for 10 min, dehydrated with alcohol and cleared with propylene oxide. The pellets were then embedded in Epon B and sections were cut with a Reichert U2 Ultramicrotome. Sections were viewed on a Siemens 1A or 101 electron microscope and electron micrographs were taken at various levels of the pellets and printed at 660~20,000.
Determination o f GABA in synaptosomal J?actions
For GABA analysis, the pellets of the synaptosomal fractions were resuspended in 1 ml 0.05 M Tris-citrate buffer (pH 7.1 at 4"C). Protein concentrations were mea- sured in this suspension by the method of Lowry et al. (1951) as modified by Markwell et al. (1978). In view of the low levels of G A B A in synaptosomal fractions and the high number of samples to be analysed, we chose a simple, sensitive and specific radioreceptor assay (Enna and Snyder, 1976) for GABA analysis as described in detail previously (Lrscher, 1979). The assay is based upon the principle that the amount of [3H]GABA bound to rat brain synaptic membranes is inversely related to the amount of GABA present in the incubation medium. For this assay it has been shown that the only substance in normal brain which will interfere with the bound [3H]GABA under the conditions of the assay is GABA itself (Enna and Snyder, 1976). Hence. GABA concentrations in brain or CSF determined with this method are virtually identical to those obtained with other analytical techniques, such as ion exchange fluorometry (Wood et al., 1978a), gas chromatography mass spec- trometry (Enna et al., 1977) and the enzymatic "GABAase" assay (Enna and Snyder, 1976). For analysis of syn- aptosomal GABA levels by the radioreceptor assay, 5(~100pl of the synaptosomal fraction were diluted to 2.5 ml with distilled water and GABA was measured in 1 ml aliquots. All samples were analysed in duplicate. None of the drugs used for tissue homogenization (3-MP) or in the animal experiments (VPA, AOAA) caused an effect on [3H]GABA binding in the radioreceptor assay (drugs tested up to 0.5-1 mM).
Experiments with VPA and AOAA
Regional GABA concentrations in rat brain were deter- mined in groups of 10 rats 1 h after i.p. injection of AOAA (30 mg/kg) and 0.5 h after VPA (200 mg/kg of the sodium salt i.p.), respectively. Five rats of each group were used for GABA determination in whole tissue of the respective brain regions and 5 animals were used for GAB A analysis in synaptosomal fractions. GAB A concentrations in each
treated group of rats were compared with those in a control group of 10 rats which were killed immediately prior to the treated animals.
Statistics
Arithmetical means and SE are given for the biochemical determinations. Significance of differences was calculated by comparing each treated group with a concurrent control group of the same day, using Student 's t-test.
RESULTS
Evaluat ion o f s ynap tosomal ./?actions c~f rat brain
regions by electron microscopy
A c o m p a r a t i v e s e m i q u a n t i t a t i v e e x a m i n a t i o n o f
e lec t ron m i c r o g r a p h s f r o m 33 s y n a p t o s o m a l pellets
der ived f r o m 11 reg ions o f ra t b ra in , n a m e l y ol fac-
to ry bulb , f ron ta l cerebra l cor tex , c o r p u s s t r i a t u m ,
h i p p o c a m p u s , t h a l a m u s , h y p o t h a l a m u s , s u p e r i o r a n d
infer ior col l iculus , s u b s t a n t i a n igra , pons , m e d u l l a
a n d ce rebe l lum, s h o w e d no o b v i o u s qua l i t a t ive or
q u a n t i t a t i v e d i f ferences in c o m p o s i t i o n o f subce l lu la r
par t ic les . In all s amp le s , s y n a p t o s o m e s a c c o u n t e d for
the m a j o r i t y o f s t r u c t u r e s wh ich cou ld be identif ied.
D a t a o b t a i n e d f rom q u a n t i t a t i v e d e t e r m i n a t i o n o f
the p r o p o r t i o n o f s y n a p t o s o m e s p r e sen t in the pellets
f r om one r a n d o m l y c h o s e n b ra in reg ion , i.e. the
h i p p o c a m p u s , are s h o w n in Tab le 1. U s i n g the str ict
c r i te r ion t ha t s y n a p t o s m e s cons i s t o f m e m b r a n e -
b o u n d s t r u c t u r e s c o n t a i n i n g c y t o p l a s m a n d clearly
def ined vesicles, the f igures are b iased t o w a r d a low
va lue for the s y n a p t o s o m e c o n t e n t because o f the
difficulty o f pos i t ive ly iden t i fy ing the s t r u c t u r e s in
p l anes o f sec t ion o t h e r t h a n ideal. T h e p r o p o r t i o n o f
free m i t o c h o n d r i a p r e sen t in the s y n a p t o s o m a l pellets
was very low (1.1'Jo) .... a n d so was the c o n t a m i n a t i o n
wi th mye l in (4.5~,o). A b o u t 60~';, o f s t r u c t u r e s revealed
by e lec t ron m i c r o s c o p y cou ld no t be u n e q u i v o c a l l y
identif ied, bu t m a n y o f these un iden t i f i ab le e l emen t s
were m o s t p r o b a b l y also o f s y n a p t o s o m a l or igin . T h e
Table 1. Morphological assessment of synaptosomal preparations
Free Synaptosomes Mitochondria Axons U nidemilied Reference
Rat hippocampus 30 + 2.0 1. I + 0.2 4.5 + 0.6 64 + 2.11 Present work Rat cerebral cortex 48 + 2.5 4.3 + 1.2 2.8 + (I.6 45 + 2.7 Dodd et al. 1981
Semiquantitative examination of electron micrographs from preparations of all 11 brain regions showed no obvious differences m composition of subcellular particles. Thus, one region was randomly chosen for quantitative determination of the proportion of synaptosomes present in the samples. Thirteen electron micrographs were taken at various levels of pellets derived from the hippocampus and structures on the respective micrographs were assigned to one of the 4 groups listed above. Synaptosomes were defined as membrane-bound, vesicle containing structures. Free mitochondria were identified by the presence of cristae, while axons were detected by their characteristic lamellar appearance. For comparison, values reported in the literature tbr rat and mouse brain preparations are also given. Values are means + SE expressed as a percentage of all structures present.
Drug-induced changes in
values are in good agreement with figures obtained by other studies with density gradient centrifugation procedures (Table 1).
G A B A concentrations in whole tissue and syn- aptosoma/ fractions o f rat brain regions
GABA levels in whole tissue of 11 brain regions from 135 rats are shown in Table 2. The respective regional GABA concentrations are essentially consis- tent with those determined by Balcom et al. 0975) in the same rat brain regions after microwave fixation, except that GABA levels measured in the substantia nigra are 30~o lower in the present study. This difference may be attributed to differences in the dissection technique.
The similarity of the regional GABA levels shown in Table 2 with those reported after microwave fixation (Balcom et al., 1975) indicate that post- mortem increases of GABA did not occur during rapid dissection of the brain at -18~C. This was confirmed by an experiment in which rats obtained 3-MP (100 mg/kg i.p.) 2.5 min prior to decapitation, which previously had been shown to completely prevent postmortem increase of GABA (van der Heyden and Korf, 1978). GABA levels obtained in this way did not differ from those determined without 3-MP pretreatment (data not illustrated).
Whereas during sample preparation for whole tis- sue GABA determinations, all enzymes which could alter GABA levels in vitro were destroyed by homoge- nization in 80~o ethanol immediately following dis- section, preparation of synaptosomal fractions has to be carried out in aqueous media. Thus, although care was taken to maintain all material and tissue at 0-4~'C throughout all the fractionations, it was entirely possible that GABA metabolism remained active in synaptosomes during their isolation despite the use of low temperatures. This was evaluated by inclusion of 3-MP (1 mM) in the homogenizing medium. As shown in Fig. 2, preparation of synaptosomal frac- tions with 3-MP resulted in significantly lower GABA levels compared to concurrent control preparations without 3-MP, indicating that GABA synthesis took place in 3-MP-free medium. About the same de- creases of synaptosomal GABA levels were observed when the experiment was repeated with 1 mM AOAA instead of 3-MP (not illustrated) AOAA at this concentration has been shown to cause almost total inhibition of both GABA synthesis and degradation in brain homogenates (Wood et aL, 1979), thus again indicating that GABA synthesis took place during preparation of control synaptosomal fractions, whereas GABA degradation was apparently not ac-
synaptosomal GABA content 4 4 5
0
o
i
~ l ~ l ~ l ~ l ~ l ~ l ~ l ~ 1 ~ 1 ~ 1 ~ 1
" ~ ' ~ ] + 1 + 1 + 1 + 1 + 1 + 1 + 1
o o ~ _ ~
.o
o o o o o o o
-H -H -H 4-1 ÷1 -H -H ÷1 -H ÷1 -H ~ I~
~ o +~+~+~+' ~ o
_*A
i . i
ell :'~
d d d d d d o d d d o +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 ~ '~'
~ ~.~
+t ~-I 4-1 ÷1 ~1 ÷1 ÷1 4-1 ÷1 ÷1 ÷1 ~ o ..~
c~
.~aa o - ~ ~.,.~.
. . . . .~ o N
~ ea 0
N o N ~ ~ ,,., ~ ~
446 WOLFGAN(i LGS('HER et al.
[ nmol/mg protelni
Z5
20
lr~ '
0]
@ABA n syngptosomal t rachons
prepclrahon without } Mr;
I I preparohon w4th 3 NP 1ram
I
q
r]
i i
E ~ c r
_ w n ' J
Fig. 2. GABA levels in synaptosomal fractions prepared from I1 rat addition of 1 mM 3-mercaptopropionic acid (3-MP) in the homogenizing
of 10 determinations: significance of differences (P < 0.05) is
brain regions with and without medium. Values are means + SE indicated by asterisks.
rive. Therefore, to avoid artifactual GABA increases during the isolation procedures, 3-MP (1 mM) was included in the homogenizing medium for all further determinations of synaptosomal GABA concen- trations.
The possibility that a redistribution of GABA occurred among subcellular organelles during the homogenization or the subsequent centrifugations was tested by comparing the GABA concentration in synaptosomes of the respective regions from 10 rat brains homogenized in 0.32 M sucrose without an additive (except 3-MP) or containing 1 mM GABA, respectively. The presence of exogenous GABA did not significantly alter the GABA levels in syn- aptosomal fractions of any region (data not shown), thereby indicating that redistribution of GABA was not a complicating factor in the present investigation.
Control GABA levels in synaptosmal fractions isolated from brain regions of 50 rats are shown in
Table 2. With respect to absolute GABA concen- trations, about 2 10°~, of GABA in whole tissue of the respective regions were found in the syn- aptosomal fraction, which corresponds rather well to figures reported previously for synaptosomes pre- pared from whole brain and brain cortex of mice (Abe, 1978; Sarhan and Seiler, 1979: L6scher, 1981a). These data obviously could not provide an absolute measure of GABA concentrations present in nerve endings of the respective regions in the intact animal, since there is a large loss of material during fraction- ation. However, Wood e t a l . (1979) have demon- strated with mouse brain homogenates that the loss of GABA within the synaptosomal fraction parallels that in organelle protein content, so that the GABA concentration on a mg protein basis remains un- changed.
As shown in Table 2, synaptosomal GABA concentrations expressed in nmol/mg protein were
Drug-induced changes in synaptosomal GABA content 447
highest in superior and inferior colliculus, corpus striatum, olfactory bulb, thalamus and hypothalamus and lowest in pons and medulla. On the mg protein basis, a significant relationship was found between GABA concentrations in synaptosomes and whole tissue of the respective 11 brain regions (r = 0.679, P < 0.05).
Drug-induced changes in GABA content of whole tissue and synaptosomal .fractions .from rat brain regions
The effect of AOAA (30 mg/kg i.p., 1 h) and VPA (200mg/kg i.p., 0.5 h) on GABA levels in whole tissue and synaptosomal fractions of 11 brain regions in rats is shown in Fig. 3. Both drugs caused significant increases in GABA concentrations when compared with concurrent controls, but the degree, compartmentation and profile of the increases across areas generated by AOAA were quite distinct from those obtained with VPA. After AOAA, the most marked GABA elevation in both whole tissue and synaptosomal fraction was seen in frontal cortex and olfactory bulbs. In both regions, the relative increase in whole tissue GABA after AOAA was considerably larger than that obtained in synaptosomes. Significant whole tissue GABA increases after AOAA were also determined in hippocampus, hypo- thalamus, substantia nigra, and cerebellum, however, in these regions AOAA exerted no significant effect on synaptosomal GABA levels.
After VPA, GABA increases in whole tissue of most regions were lower than after AOAA, whereas the opposite was the case in respect to synaptosomal GABA increases. Thus, VPA induced significant el- evation of GABA in synaptosomal fractions from olfactory bulb, hypothalamus, superior and inferior colliculus, substantia nigra, and cerebellum, the most marked effect (80~o over control) being determined in the hypothalamus. Both AOAA and VPA exerted no significant effects on protein content of synaptosomal fractions.
D I S C U S S I O N
During the last few years, measurement of GABA in synaptosomes derived from the whole brain or brain cortex of mice has been repeatedly used to monitor indirectly the in vivo effects of drugs on GABA levels in brain nerve endings (Abe and Mat- suda, 1977; Abe, 1978; Wood et al., 1978; Matsuda et al., 1979; Sarhan and Seiler, 1979; Wood et al., 1979 and 1980; L6scher, 1981a,b,c; L6scher et al., 1981; Wood et al., 1981). Although such experiments
are fraught with the danger that the data obtained do not represent the true situation in the intact animal, reports of different groups strongly suggested that the GABA content of synaptosomal fractions is stable during the preparation and isolation of the organelles and that changes in this parameter should accurately reflect changes in nerve ending GABA brought about by the administration of drugs (Wood et al., 1978b; Matsuda et al., 1979; Sarhan and Seiler, 1979; Wood et al., 1979). In view of the marked regional differences in GABA content and function in the brain, the purpose of the present study was to extend the synaptosomal model to determination of drug- induced in vivo changes in GABA levels of discrete brain areas.
First of all, well-established density gradient cen- trifugation procedures for isolating synaptosomes were adapted to the small tissue samples of rat brain regions and the high number of sample preparations necessary per experiment. By the use of 2 ultra- centrifuges, the described method allows the prepara- tion of synaptosomes from 11 brain regions of 10 rats (i.e. 110 samples) within 1 day. The synaptosomal fractions thus obtained were characterized by elec- tron microscopy which indicated that the fractions were sufficiently pure for our purpose. To circumvent the long preparation times inherent to density gra- dient centrifugation procedures, several recent studies on GABA in "synaptosomes" (e.g. Sarhan and Seiler, 1979; Pagliusi et al., 1983) have used the crude P2 fraction of Gray and Whittaker (1962) which can be obtained in 35~0min . However, this preparation contains substantially higher amounts of enzyme-rich contaminants including free mitochondria, myelin fragments and lysosomes, all of which may interfere with the study of drugs which affect GABA metabo- lism (Gray and Whittaker, 1962).
Postmortem increases in GABA content of brain regions can occur during removal and dissection of the brain tissue and during preparation of the ho- mogenate when the GABA-synthesizing enzyme glu- tamic decarboxylase (GAD; EC 4.1.1.15) is not inac- tivated (cf. Balcom et al., 1975; van der Heyden and Korf, 1978). Microwave irradiation totally prevents postmortem changes of GABA levels, but un- fortunately this procedure cannot be used when subcellular fractionation is required. Injection of the potent GAD-inhibitor 3-MP prior to decapitation has also been used to prevent postmortem GABA increase, but this may interfere with study of in vivo effects of drugs on GABA levels. The results presented here indicate, that GABA changes during removal and dissection of brain tissue can also be
NCI 6/4~C
448 WOLFGANG L/JSCHER et al.
100
80. o "6
S 60
c
o
0.5 hr offer VPA, 200mglk9 ~.p
[
I synaptosomal frachon
1 t
100
~ 8O o
o 60 t)
c &0
~ 20
S
1 hr ofter AOAA 30 mg/k ~ i.p.
, _ l &
i >" E
-•" T i
' T ~
Fig. 3. Effect of VPA and AOAA on GABA levels in whole tissue and synaptosomal fractions of 11 rat brain regions. Preparation of synaptosomes was carried out using I mM 3-MP in the homogenizing medium. Results are expressed as percent increase (means + SE of 5 rats) over concurrent control determinations (see Table 2 for absolute control values). Significance of differences (P < 0.05) to the individual controls is indicated by asertisks. All drug effects were calculated from experiments in which
GABA concentrations were expressed in nmol/mg protein.
Drug-induced changes in synaptosomal GABA content 449
sufficiently minimized by rapid processing and by the use of low temperature. In the case of GABA deter- minations in whole tissue samples, GABA changes which might occur after the dissection were prevented by rapid homogenization in ethanol. Regional GABA levels thus obtained corresponded well to those found after 3-MP pretreatment as well as to values determined previously with microwave irra- diation (Balcom et al., 1975). However, the out- standing problem with isolation of synaptosomes was that GABA metabolism could remain operative after dissection during fractionation of tissue in aqueous media. Previous experiments of Matsuda et al. (1979) and Wood et al. (1979) have indicated that GABA is not increased during preparation of synaptosomal fractions from whole mouse brain, provided the temperature of the fractionation media is kept at 0-2"C. In contrast, the present experiments with and without inclusion of 3-MP in the homogenizing me- dium strongly suggest that GABA synthesis remains active during the fractionating procedures in 3-M P-free media despite the use of low temperatures. At the concentration of 3-MP employed (1 mM), inhibition of GABA synthesis should be complete since a K~ of 1.8/~m has been reported for GAD purified from mouse brain (Wu, 1976). Interestingly, AOAA (1 mM), which in vi tro is as potent as 3-MP to inhibit GAD (Wu, 1976) but more potently in- hibits the GABA degrading enzyme GABA amino- transferase (GABA-T; EC 2.6.1.19), caused the same decrease in regional synaptosomal GABA levels as 3-MP when added to the homogenizing medium. These results suggest that, in contrast to GABA synthesis, GABA degradation is not of major im- portance during synaptosome preparation. We there- fore decided to use 3-MP treated homogenates for all further experiments with synaptosomal fractions.
Another technical problem in the current type of investigation arises from the possibility that syn- aptosomes both take up GABA from and release GABA into the surrounding medium during the fractionation procedures. However, in agreement with previous studies on synaptosomes isolated from the whole brain or brain cortex (Wood et al., 1978; Matsuda et al., 1979; Sarhan and Seiler, 1979) we found that high concentrations of GABA in the homogenizing medium do not influence the GABA content in synaptosomes isolated from discrete brain areas. Furthermore, although there is obviously a significant loss of total GABA content during prepa- ration of synaptosomes, recent experiments of Wood et al. (1979) demonstrated that this loss is not due to release or leakage of GABA from synaptosomes but
can be attributed to a significant loss of organelles. It was therefore concluded that the GABA content of the synaptosomes is not subject to change during the fractionation procedures provided GABA levels are expressed per mg organelle protein.
Despite these considerations, the present data on synaptosomal GABA levels in various brain regions of the rat cannot provide an absolute measure of the GABA concentration in nerve endings, since only a portion of the isolated synaptosomes arises from GABAergic terminals. In any event, the levels of GABA determined under the conditions used in the present study provide a minimum estimate of the levels of the amino acid in the intact animal, and drug-induced changes in synaptosomal GABA con- tent should parallel changes in the GABA levels of the terminals. Accordingly, we used the syn- aptosomal model to evaluate the effects of two GABA-elevating agents, namely AOAA and VPA, on synaptosmal GABA concentrations in different brain regions. Both drugs were examined at doses and times of pretreatment previously determined to result in about the same anticonvulsant effect on the electroconvulsive threshold in mice (L6scher, 1980).
Although AOAA and VPA have been widely used to correlated elevation of brain GABA levels and neuropharmacological effects, especially anti- convulsant action, they differ in a number of aspects. AOAA is a potent inhibitor of GABA-T, whereas GAD activity in vivo is only decreased by very high doses of AOAA (e.g. Wood and Peesker, 1973). In contrast, VPA is a weak inhibitor of GABA-T (Fowler et al., 1975) but seems to increase GAD activity when administered to rodents in anti- convulsant doses (L6scher and Frey, 1977). The degree of increase in brain GABA that occurs with anticonvulsant doses of VPA and AOAA strikingly differs. Anticonvulsant activity of VPA is seen with whole brain GABA increases in the range of 30-60~ over control, whereas antiseizure actions of AOAA only occur with GABA increases between 200~00~o (Wood and Peesker, 1973; Simler et al., 1973). This discrepancy might relate to the possibility that AOAA and VPA exert their predominant effects on different compartments of GABA. Iadarola and Gale (1979) examined the effect of both drugs on GABA concentrations occurring in the presence and absence of GABA-containing nerve terminals by using rats in which the GABAergic afferent projection to the substantia nigra was surgically destroyed on one site of the brain. Indeed, the results indicated that VPA preferentially increases the GABA concentration in nerve terminals while AOAA primarily elevates
450 WOLFGANG L()SCHI?R el al.
GABA in non-nerve terminal compartments (neu- ronal perikarya and glia). Although these data seemed to explain the anomalous results regarding whole brain G A B A levels versus anticonvulsant ac- tivity of A O A A and VPA, subsequent experiments using the synaptosomal model have shown that A O A A exerts a significantly greater effect on GABA in whole brain synaptosomes than does VPA when both drugs are administered in equally effective anti- convulsant doses (L6scher, 1981a). These contro- versial results obtained by different models for estimation of drug-induced nerve terminal GABA increases are readily explained by the present data. In the substantia nigra, VPA significantly increased G A B A in synaptosomes while A O A A was withoul effect, confirming the results obtained by ladarola and Gale (1979) who used the technique of the GABA-denervated substantia nigra. Significant elevation of synaptosomal GABA by VPA was also found in olfactory bulb, hypothalamus, superior and inferior colliculus, and cerebellum, whereas AOAA- induced increases achieved significance only in olfac- tory bulb and cerebral cortex. In fact, the cortex was the only region in which A O A A was significantly more potent than VPA in increasing synaptosomal G A B A levels. Therefore, one might suggest that the previously reported effect of A O A A on G A B A levels in synaptosomes from whole brain (Abe, 1978: Wood et al., 1978: L6scher, 1981a) primarily relates to the increase m cortical concentrations. The present data with VPA and A O A A thus demonstrate that mea- surement of G A B A in whole brain and synaptosomes isolated therefrom would be highly insensitive to those changes in G A B A taking place only in a few selected nuclei, changes which may be sufficient to influence markedly specific behavioural and phys- iological functions. In terms of anticonvulsant ac- tion, recent experiments of ladarola and Gale (1982) strongly suggest that the substantia nigra is the anatomical site that mediates GABA-related anti- convulsant activity. With respect to the differential effects of A O A A and VPA on synaptosomal GABA levels in this region, it is thus interesting to note that A O A A (30 mg/kg i.p.) 1 h after injection is without effect in rats on seizures induced by 3-MP (50 mg/kg i.p.) whereas VPA (200 mg/kg i.p.) after 0.5 h blocks these seizures almost completely (L6scher and Vetter, unpublished results).
In conclusion, the results presented here indicate that determination of regional G A B A levels in syn- aptosomes can provide useful information on the m vivo effects of drugs on G A B A levels in nerve termi- nals and their ability to exert this effect in specific
brain regions, provided the data are analysed care- fully with regard to the possible occurrence of changes taking place during dissection, homoge- nization and fractionation. In view of the theoretical and practical importance of information about the changes in transmitter pools in nerve endings during pathological processes or elicited by drugs, further application of this synaptosomal model should be warranted.
Acknowledgements We thank Mrs Rfickauer for skillful technical assistance. The study was supported by grants from the Deutsche Forschungsgemeinschaft, Research into epilepsy program and Lo 274/2-3).
R E F E R E N C E S
Abe M. (1978) Relationship between 7-aminobutyric acid metabolism and antivitamin B~-induced convulsions. J. Nutr. Sci. Vitam. 24, 419--427.
Abe M. and Matsuda M. (1977) 7-Aminobutyric acid metabolism in subcellular particles of mouse brain and its relationship to convulsions. J. Biochem. (Tokyo) 82, 195-200.
Balcom G. J., Lenox R. H. and Meyerhoff J. L. 11975) Regional 7-aminobutyric acid levels in rat brain deter- mined after microwave fixation. J. Neurochem. 24, 609 613.
Baxter C. F. (1972) Assay of ;'-aminobutyric acid and enzymes involved in its metabolism. In: Methods o/ Neurochemistrv (Fried R., edt, Vol. 3. pp. 1 73. Marcel Dekker, New York.
Baxter C. F. (1976) Some recent advances in studies of GABA metabolism and compartmentation, in GABA in Nervous System Function (Roberts E., Chase T. N. and Tower D. B., eds), pp. 61 87. Raven Press, New York.
Carmona E., Gomes C. and Trolin G. (1980) On the importance of GABA-ergic neurons for the AOAA in- duced accumulation of GABA in the rat brain. Naunyn- Schmiedeber£s Arch. exp. Path. Pharmak. 313, 221-224.
Cotman C. W. (1974) Isolation of synaptosomal and syn- aptic plasma tractions. In: Methods in Enzymology (Fleischer S. and Packer L., eds), Vol. 31, pp. 445452. Academic Press. New York.
Dodd P. R., Hardy J. A., Oakley A. E., Edwardson J. A., Perry E. K. and Delaunoy J.-P. ( 1981 ) A rapid method for preparing synaptosomes: Comparison with alternative procedures. Brain Res. 226, 107-I18.
Enna S. J. and Snyder S. H. (1976) A simple, sensitive and specific radio-receptor assay lbr endogenous GABA in brain tissue. J. Neurochem. 26, 221 224.
Enna S. J., Stern L. Z., Wastek G. J. and Yamamura H. I. (1977) Cerebrospinal fluid 7-aminobutyric acid variations in neurological disorders. Arch. Neurol. 34, 683-685.
Fowler L. J., Beckford J. and John R. A. (1975) An analysis of the kinetics of the inhibition of rabbit brain 7-aminobutyrate amino-transferase by sodium n- dipropylacetate and some other simple carboxylic acids. Biochem. Pharmac. 24, 1267 1270.
Gray E. G. and Whittaker V. P. (1962) The isolation of nerve endings from brain: An electron microscopic study
Drug-induced changes in synaptosomal GABA content 451
of cell fragments divided by homogenization and centrif- ugation. J. Anat. (Lond.) 96, 79-88.
Hardy J. A., Dodd P. R., Oakley A. E., Perry R. H., Edwardson J. A. and Kidd A. M. (1983) Metabolically active synaptosomes can be prepared from frozen rat brain and human brain. J. Neurochem. 40, 608-614.
Iadarola M. J. and Gale K. (1979) Dissociation between drug-induced increases in nerve terminal and non-nerve terminal pools of GABA in vivo. Eur. J. Pharmac. 59, 125-129.
Iadarola M. J. and Gale K. (1982) Substantia nigra: Site of anticonvulsant activity mediated by 7-aminobutyric acid. Science 218, 1237-1240.
Krogsgaard-Larsen P., Scheel-Kr/iger J. and Kofod H. (1979) GA BA -Neurotransmitters. PharmacochemicaL Biochemical and Pharmacological Aspects. Munksgaard, Cogenpagen.
L6scher W. (1979) GABA in plasma and cerebrospinal fluid of different species. Effects of 7-acetylenic GABA, y-vinyl GABA and sodium valproate. J. Neurochem. 32, 1587-1591.
L6scher W. (1980) A comparative study of the pharma- cology of inhibitors of GABA-metabolism. Naunyn- Schmiedehergs Arch. exp. Path. Pharmak. 315, 119-128.
L6scher W. (1981a) Effect of inhibitors of GABA amino- transferase on the metabolism of GABA in brain tissue and synaptosomal fractions. J. Neurochem. 36, 1521-1527.
L6scher W. (1981b) Valproate induced changes in GABA metabolism at the subcellular level. Biochem. Pharmac. 30, 1364-1366.
L6scher W. (1981c) Relationship between drug-induced changes in seizure thresholds and the GABA content of brain and brain nerve endings. Naunyn-Schmiedebergs Arch. exp. Path. Pharmak. 317, 131-134.
L6scher W., B6hme G., Schfifer H. and Kochen W. (1981) Effect of metabolites of valproic acid on the metabolism of GABA in brain and brain nerve endings. Neu- ropharmacology 20, 1187 1192.
L6scher W. and Frey H.-H. (1977) Zum Wir- kungsmechanismus von Valproinsfiure. Arzneim.- Forsch./Drug Res. 27, 1081-1082.
Lowry O. H., Rosebrough N. J., Farr A. L. and Randall R. J. (1951) Protein measurement with the Folin phenol reagent. J. biol. Chem. 193, 265-275.
Markwell M. A. K., Haas S. M., Bieber L. L. and Tolbert N. E. (1978) A modification of the Lowry procedure to simplify protein determination in membrane and tipo- protein samples. Analyt. Biochem. 87, 206-210.
Matsuda M., Abe M., Hoshino M. and Sakurai T. (1979) 7-Aminobutyric acid in subcellutar fractions of mouse brain and its relation to convulsions. Biochem. Pharmac. 28, 2785-2789.
Pagliusi S. R., Gomes C., Leite J. R. and Trolin G. (1983) Aminooxyacetic acid induced accumulation of GABA in the rat brain. Interaction with GABA receptors and distribution in compartments. Naunyn-Schmiedebergs Arch. exp. Path. Pharmak. 322, 210-215.
Sarhan S. and Seiler N. (1979) Metabolic inhibitors and subcellular distribution of GABA. J. Neurosci. Res. 4, 399-421.
Simler S., Ciesielski L., Maitre M., Randrianarisoa H. and Mandel P. (1973) Effect of sodium n-dipropylacetate on audiogenic seizures and brain 7-aminobutyric acid level. Biochem. Pharmac. 22, 1701-1708.
Van der Heyden J. A. M. and Korf J. (1978) Regional levels of GABA in the brain: Rapid semiautomated assay and prevention of postmortem increase by 3-mercapto-propionic acid. J. Neurochem. 31, 197-203.
Wood J. D. and Peesker S. J. (1973) The role of GABA metabolism in the convulsant and anticonvulsant actions of aminooxyacetic acid. J. Neurochem. 20, 379 387.
Wood J. H., Glaeser B. S., Enna S. J. and Hare T. A. (1978a) Verification and quantification of GABA in human cerebrospinal fluid. J. Neurochem. 30, 291-293.
Wood J. D., Kurylo E. and Newstead J. D, (1978b) Aminooxyacetic acid induced changes in ~,-amino- butyrate metabolism at the subcellular level. Can. J. Biochem. 56, 667-672.
Wood J. D., Russell M. P., Kurylo E. and Newstead J. D. (1979) Stability of synaptosomal GABA levels and their use in determining the in L, iL'o effects of drugs: Convulsant agents. J. Neurochem. 33, 61-68.
Wood J. D., Russell M. P. and Kurylo E. (1980) The -aminobutyrate content of nerve endings (syn-
aptosomes) in mice after the intramuscular injection of ~-aminobutyrate-elevating agents: A possible role in anti- convulsant activity. J. Neurochem. 35, 125-130.
Wood J. D., Kurylo E. and Tsui S.-K. (1981) Interactions of di-n-propylacetate, gabaculine, and aminooxyacetic acid: Anticonvulsant acitvity and the ],-aminobutyrate system. J. Neurochem. 37, 144(~1447.
Wu J.-Y. (1976) Purification, characterization, and kinetic studies of GAD and GABA-T from mouse brain, in GABA. In: Nert,ous System Function (Roberts E., Chase T. N. and Tower D. B., eds), pp. 7-55. Raven Press, New York.
Zeman W. and lnnes J. R. M. (1963) Craigie's Neuro- anatomy o f the Rat. p. 183. Academic Press, New York.
