0306-4522/90 $3.00+ 0.00 PergamonPressplc 0 1990IBRO
Neuroscience Vol.39,NO.2, PP.419-029,1990 Printedin Great Britain
IMMUNOHISTOCHEMICAL DEMONSTRATION OF GLUTAMATE DEHYDROGENASE IN THE POSTNATALLY DEVELOPING RAT HIPPOCAMPAL FORMATION AND CEREBELLAR CORTEX: COMPARISON TO ACTIVITY STAINING F. ROTHE, G. WOLF* and G. SCH~NZEL Institute of Biology, Medical Academy of Magdeburg, Erich-Weinert-Strass 3, D/O-3014 Magdeburg, Germany
Abstract-Distribution patterns of activity and immunohistcchemical staining for glutamate dehydrogenase were compared during the postnatal development of rat hippocampal formation and eerebellar cortex. On postnatal day 5, dendritic layers of the hippocampal formation showed only a very weak enzyme activity. Similarly, when studied at the same age, the external granule cell layer and Purkinje cells of the cerebellar cortex exhibited a very faint and moderate staining, respectively. With advancing age, in both brain regions a marked postnatal increase in glutamate dehydrogenase activity occurred in neuropil area as glutamate@ structures matured. However, compared to activity staining, both brain regions of early postnatal stages showed a relatively high level of glutamate dehydrogenase-like immunoreactivity. In this case, the immunohistochemical staining of hippocampal dendritic layers and of the molecular layer of the cerebellar cortex was rather diffuse, being not very similar to parameters of the maturation of the respective glutamatergic structures. In contrast to the activity staining for the enzyme, the immunohistochemical labelling in adult rats revealed a selective predominance of immunoreactivity in astroglial cells from postnatal day 5 onwards. The Bergmann glia in the cerebellar cortex exhibited the strongest intensity of immunoreactivity. Generally, the patterns of immunoreactivity were found to depend on the fixation procedure adopted. Concluding from our results, glutamate dehydrogenase is demonstrable in glial and in neuronal cell elements as well. Therefore, it is recommended that activity staining and the immunohistochemical procedure be combined to study qualitative and quantitative aspects of glutamate dehydrogenase in nervous tissues.
In mammalian brain, substantial amounts of glutamate dehydrogenase [L-glutamate: NAD(P)+ oxidoreductase, deaminating; E.C. l4.1.3; GDH] activity were reported.6,10,18GDH is known to be associated with mitochondria of brain,26337in which it catalyses the interconversion of glutamate and 2-oxoglutarate.“JO In addition to its function in common metabolism, glutamate is established as one of the major excitatory neurotransmitters in the CNS.12 Therefore, GDH might play a key role in the synthesis or degradation of transmitter glutamate. Thus, in brain regions which are enriched with glutamatergic structures the concentration and activity of GDH might be expected at higher levels when compared to other brain regions. Our previous findings have given biochemical25 and histochemica13’,32,4’a43 support to the suggestion that in the maturing rat hippocampal formation and cerebellum GDH is a transmitter glutamate-metabolizing enzyme the postnatal activity increase of which seems
to be related more to the neurons than to glial cells.
Lately, immunocytochemical studies have shown, however, a more intense GDH-like immunoreactivity (GDH-LI) in glial cells than in neurons.2,3,14,1g,38 In order to compare our previous histochemical results and elucidate the role of GDH in the glutamate transmitter metabolism we studied the distribution of GDH-LI in the hippocampal formation and cerebellar cortex of the rat at different ages. Both brain regions are endowed with glutamatergic structures and, moreover, their architecture is well characterized. Tissue elements including neuronal contacts are organized within the first three weeks of postnatal life.‘s5 EXPERIMENTAL PROCEDURES Animals
Male rats of our breeding stock (derived from VELAZ, Prague) aged five, 10, 15, 20, 30, 35, 40 and 50 days were used for the immunohistochemical procedure and enzyme activity staining. The rats were housed under standardized conditions.25 Enzyme activity staining
*To whom all correspondence should be addressed. Abbreviations: GDH, glutamate dehydrogenase; GDH-LI, glutamate dehydrogenase-like immunoreactivity; GFAP, glial fibrillary acidic protein; LDH, lactate dehydrogenase; PAP, peroxidase-antiperoxidase.
The rats were killed by decapitation under ether anaesthesia. The rapidly removed brains were instantly frozen on a metal block cooled with liquid nitrogen. The protocol used for GDH-activity staining on sagittal cryostat sections
419
420
F. ROTHEei al.
(20 pm) was described previously in detail.‘* Control staining was obtained by omitting substrate (glutamate) or coenzyme (NAD).
The animals were anaesthetized with hexobarbital (100 mg/kg body weight) and the perfusion fixation was initiated by intracardial flow of 2% dextran (mol. wt 70,000) in 0.1 M sodium phosphate buffer, pH 7.4, followed by different fixation solutions containing (i) 4% paraformaldehyde and 0.1% glutaraldehyde (Merck-Schuchardt, Munich, F.R.G.), (ii) 2% alutaraldehvde. or (iii) 5% glutaraldehyde, each diluted-in the same buffer. The perfusion was performed as a function of advancing age of the animals for 5-15 min. Fiften minutes after termination of perfusion the hippocampal formations and cerebella were dissected out. The tissue was then postfixed in the particular fixative for 30 min, followed by cutting into 300~pm slices
using a tissue chopper. Stretched hippocampi were cut transversally from the septal to the temporal region. Parasagittal slices were prepared from the cerebellum. Slices were collected into buffer containing 0.1% [after fixation in (i)] and 0.5% glutaraldehyde [for (ii) and (iii)] and stored at 4°C prior to use. After cryoprotection with 30% sucrose in 0.1 M sodium phosphate buffer, pH 7.4, overnight, the tissue was resectioned on a freezing microtome to obtain 20-pm sections. The latter were used immediately for immunohistochemical localization of GDH or glial fibrillary acidic protein (GFAP). Rabbit anti-GFAP antiserum was generously supplied by the laboratory of Prof. Dr J. Storm-Math&en (Anatomical Institute, University of Oslo, Oslo, Norway). The antiserum to GDH was raised in white New Zealand rabbits. Rabbits received an initial injection of 3 mg bovine liver GDH (Boehringer, Mannheim, F.R.G.) in complete Freund’s adjuvant and three booster injections of 1 mg of antigen in incomplete Freund’s adjuvant. The injections were carried out into the popliteal lymph nodes subcutaneously and intracutaneously. The time intervals allowed between immunizations were four to five weeks. Animals were bled by heart puncture one week after the last immunization. The antiserum was characterized by microimmunoelectrophoresis on agarose plates.*’ The electrophoresis was carried out with (i) bovine liver GDH and (ii) concentrated homogenate supematant of rat hippocampal formation. The immunoprecipitates were developed by use of the antiserum to bovine liver GDH. The identification and specificity control of GDH-containing immunoprecipitate was carried out by activity staining as described above for the localization in tissue. For controls, substrate for NAD was omitted or preimmune serum was used (Fig. 1). The technique reported by Ottersen and StormMathisen24 was adopted for another specificity test. Cellulose ester filters containing spots of different antigens (Fig. 2) were processed with the antiserum and visualized immunohistochemically. Bach spot contained one of several antigens: GDH, lactate dehydrogenase, bovine serum albumin;rat serum, and GDH glutaraldehyde conjugates. To eliminate lactate dehvdrogenase (LDH) antibodies (which were found in our &de GDH antiserum, possibly because of contamination of the commercial GDH preparation used as antigen), the GDH antiserum was absorbed with an LDH complex under slow stirring at room temperature for 2 h. To prepare the LDH complex, 10 mg LDH from bovine heart (VEB Arzneimittelwerk, Dresden, G.D.R.) were precipitated by 0.1% glutaraldehyde (about IOmM) in 2ml 0.1 M sodium phosphate buffer, pH 7.4, under ‘stirring at room temperature for 3 h and kept overnight at 4°C. Following the resuspension of the pellet in sodium phosphate buffer the precipitate was centrifuged four times at 1OOOg for 10min. Free aldehyde groups
were neutralized with I M ethanolamine in 0.1 M sodium phosphate buffer, pH 8.4. The immunohistochemical procedure was based on the peroxidase-antipcroxidase (PAP) technique.34 All incubation steps were carried out with free-floating sectionc .___(20 pm) in-O.1 M Tris-HCl buffer, pH 7.4, containing 0.3 M NaCl. 0.5% Triton X-100. and 1% normal eoat serum. The 20-pm sections were pretreated with 1 M ithanolamine in 0.1 M sodium phosphate buffer, pH 8.4, and 1% H,O, in methanol to remove endogenous peroxidase activity.‘” Following washing with phosphate-buffered saline, pH 7.4, and preincubation with Tris-HCl buffer containing -3% .” normal goat serum, the sections were incubated with specific absorbed GDH antiserum (diluted 1:400) in Tris--HP1 buffer overnight. The sections were then washed and incubated with anti-rabbit globulin (purchased from Staatliches Institut fur Immunpriiparate und Nihrmedien, Berlin, G.D.R.) diluted 1: 100, for 1 h. After washing, the sections were incubated in PAP complex (obtained from Sektion Biowisenschaften der Karl-Marx-Universitat, Leipzig, G.D.R.; diluted 1:500) for 1 h under gentle shaking. The peroxidase reaction was visualized using diaminobenzidine (Fluka, F.R.G.; 7.5 mg/15 ml sodium phosphate buffer, pH 7.4) and 0.02% H20,.‘5 Thereafter, sections were rinsed in distilled water and mounted in glycerol gelatin. Controls were treated with rabbit preimmune serum, with GDH antiserum absorbed with a GDH complex conjugated by glutaraldehyde (procedure by analogy to preparation of LDH complex), or by omitting the linker globulin. To compare the distribution patterns of GDH with those of astrocytes, rabbit antiserum to GFAP diluted I:500 was used in the same way as for GDH immunostaining. 1
Il..
RESULTS Glutamate dehydrogenase activity staining Hippocampal formation. In the hippocampal formation of five-day-old rats pyramidal and granular cell perikarya were predominantly stained (Fig. 3). Cell nuclei were nearly free of the reaction product, formazan. The dendritic layers generally revealed a weak staining. The distribution of GDH activity extended increasingly with advancing age of the rats, especially in neuropil layers. From day 30 to day 50 of life, an intense activity staining for GDH was found in all dendritic layers, particularly in the stratum lacunosum-moleculare and in the molecular layer of the dentate gyrus, whereas neuronal perikarya showed only a moderate GDH activity. Maximum staining intensity was seen when the concentration of sodium ions in the incubation medium ranged from 130 to 150 mM. No selective staining of glial cells was observed under these circumstances. Total sodium concentration diminishing to 60 mM or less, the neuropil areas of the hippocampal formation appeared to be almost free of GDH activity. In this case staining was restricted to perikarya of the strata pyramidale and granulosum, and to scattered cells in neuropil areas (Fig. 5). Cerebeilar cortex. In the cerebellar cortex (Fig. 4) of rats aged five and 10 days, histochemically demonstrable GDH activity revealed fairly low formazan deposits. The granular cell bodies of the external layer exhibited a very faint staining, whereas most of the perikarya of Purkinje cells were noted to be
Glutamate dehydrogenase in rat brain
421
Fig. la-c and 2. Fig. 1. Characterization of GDH antiserum by microimnmnoelectrophoresis. The antigen wells were tilled with concentrated homogenate supematant of rat hippocampal formation. After electrophoresis the trough between the antigen walls was filled with the antiserum against bovine liver GDH. (a) Dark-field micrograph shows precipitation arcs in the agarose gel. (b) Specimen (a) after GDH-activity staining. (c) GDH-activity stained arc (lower part of microslide) compared to control reaction (without glutamate;
upper part). Fig. 2. Specificity test by use of cellulose ester filters. Spots (1 pl) contained the following antigens: (1) GDH (20 mg protein/ml); (a) 1: 10; (b) 1: 100; (c) 1: 1000. (2) Complexed GDH (conjugated by glutaraldehyde). (3) Bovine serum albumin (20 mg/rnl). (4) Rat serum, non-diluted. (5) LDH (2.5 mg protein/ml); (a) non-diluted, (b) 1: 10; (c) 1: 100. The filters were processed by the immunohistochemical technique using the crude GDH-antiserum (B) and GDH-antiserum absorbed in LDH complexes (A, C). The crude antiserum reacted strongly with GDH and GDH complex (A) and weakly with LDH (B). Following absorption in LDH complex the antiserum did not react with LDH (C), nor did crude and absorbed antiserum with bovine serum albumin and rat serum (A).
markedly stained. With advancing age of rats, a dramatic increase in staining intensity was observed, being pronounced between days 15 and 40. In SO-dayold rats, GDH staining occurred throughout the molecular layer. Granule cell bodies, too, exhibited an intense staining. Spotted staining of the internal granule cell layer was a noticeable finding. Stained
patches are possibly identical to the cerebellar glomeruli. The enzyme activity in Purkinje cell perikarya was enhanced to a lesser extent. No selective staining of glial cells was observed. Generally, control experiments omitting substrate or coenzyme resulted in a very faint staining and poor contrast.
422
F. ROTHEet ul
Fig. 3a-c and 4a-c. Fig. 3 and 4. Hist~he~cal demonstration of glutamate dehydrogenase activity in parasagittal sections of rat hippocampal formation (Fig. 3) and cerebellar cortex (Fig. 4) on postnatal days 5 (a), 20 (b) and 50 (c), using the vitro-mono~tra~olium salt p-DhTT. Scale bars = 500 pm (Fig. 3): IOOpm (Fig. 4).
Immunohistochemistry Following the electrophoresis of bovine liver GDH and homogenate supernatant of the rat hippocampal formation, the anti-GDH serum revealed one pr~ipitation arc in each case. The pr~~pitation arc was stained by the histochemical reaction for
GDH. No staining was seen when the reaction specificity was tested by omitting glutamate as substrate (Fig. 1). The use of preimmune serum did not produce precipitation arcs. When cellulose ester filters were employed for testing (Fig. 2) the non-absor~d antiserum reacted very strongly with both the commercially obtained
Glutamate dehydrogenase in rat brain GDH preparation and the GDH glutaraldehyde complex, while the reaction with purified LDH was poor. The antiserum did not react with bovine serum albumin and rat serum. Following absorption of the antiserum in the LDH complex, no staining of LDH spots was observed. In contrast to the activity staining, nerve cell perikarya showed light, if any, GDH-LI. In neuronal elements and preferentially in glial cells, the intensity and localization of GDH-LI proved to be dependent on the fixation procedure (Fig. 6). Still, when subjected to an identical fixation procedure glial cells exhibited a different reaction as a function of the brain region studied. Thus, after fixation with paraformaldehyde+ttaraldehyde (i; see Experimental Procedures), only a very weak labelling of glial cells was observed in the hippocampal formation irrespective of the developmental stage (Fig. 6a), whereas fixation with 2 or 5% glutaraldehyde (ii and iii) resulted in a distinct visualization of glial cells (Figs 6c and 7). Bergmann glia of the cerebellar cortex was markedly visualized after fixation with paraformaldehyde-glutaraldehyde or glutaraldehyde (Figs 6d and 8).
423
Using preimmune serum for control sections, fixation with 5% glutaraldehyde resulted in a slight background staining, whereas fixation with 2% glutaraldehyde produced a completely negative result (Fig. 6b). Therefore, our development study of GDHLI patterns mainly refers to material fixed with 2% glutaraldehyde. Paraformaldehydeformation. Hippocampal glutaraldehyde being adopted, a weak but clear GDH-LI was found throughout the dendritic layers of the developmental stages studied. The intensity of staining exhibited by the hilus and the inner third of the molecular layer of the dentate gyrus was slightly higher than that of other areas of the hippocampal formation (Fig. 6a). In rats up to day 20, the cytoplasm of pyramidal and granule cells was nearly free of GDH-LI. The nuclei were unstained. With increasing age the cytoplasm of pyramidal cells showed a moderate GDH-LI, predominantly in the CA3 region. Following fixation with 2% glutaraldehyde, numerous GDH-labelled cells were seen from postnatal days 5 to 50. The cells were distributed more densely in the stratum lacunosum-moleculare and in the hilus of dentate gyrus. Other hippocampal regions
Fig. 5a and b. Fig. 5. GDH-activity staining in the hippocampal formation (a) and cerebellar cortex (b) at a concentration of 60 mM sodium ions in the incubation medium. Scale bars = 500 pm (a); 100 pm (b).
424
F. ROTIE et al.
,,
b
Fig. 6a-d. Fig. 6. Imrn~ohist~he~~l localization of GDH in the hipp~ampa1 formation (a, b, c) and cerebellar cortex (d) of SO-day-old rats as a function of the fixation procedure (a) paraformaidehyde-glutaraldehyde; (b) 2% glutaraldehyde; (c, d) 5% glutaraldehyde. Micrograph (b) shows a control section treated with preimmune serum. Scale bars = 5OOym (a, b); 250,um (c); IOO~rn (d).
exhibited a more uniform distribution of GDHpositive cells. In the mature hippocampal formation, the GDH-LI of the round or spindle-shaped cells and their proximal parts of processes was intense. Their distribution pattern (Fig. 7) was largely identical with that of GFAP-labelled cells; hence, these structures were considered to belong to an identical cell type, i.e. astroglial cells. Using GFAP staining, branched processes of such cells were visible to a greater extent, though. Cerebelhr cortex. The intensity of GRH labelling of Bergmann glial cell bodies and fibres was most pronounced in adult rats irrespective of the fixation medium used. Processes of Bergmann glia showed a punctate staining. Similarly heavily stained puncta apparently belonging to glial cells were observed in the white matter of the cerebellum. A rather diffuse labelling, moderate in intensity, was found throughout the molecular layer foliowing fixation with 5% glutaraldehyde (Fig. 6d). Fixation with 2% glutaraldehyde produced such a diffuse labelhng at immature stages (up to day 20) only. In adult animals
the neuropil of the molecular layer was found to be nearly free of labelling, except for Bergmann glial cells which contrasted distinctly from the background. Our developmentai studies (2% glutaraldehydefixed material, Fig. 8) showed that, at all stages, Purkinje and granule cell bodies were only lightly labelled with nuclei free of GDH-LI. The granule cell layer of adult rats exhibited a more patchy labelling which, most likely, reflected glomeruli being typical of this stratum. Staining the sections for GFAP visualized astroglial cells in the granule cell layer and in the white matter, in addition to fibres of Bergmann glia. Unlike in GDH immunostaining, GFAP-positive cell bodies of Bergmann glia were, if at all, seen at advanced stages only. The developmental patterns of immunola~lling for GDH were quite different from that seen in GDH-activity staining. In rats aged five and 10 days, the cerebellar cortex was more intensely stained by immunohistochemi~l iabeliing than by activity staining, a finding similar to that obtained for the
a,
Fig. 7. Immunohistochemical localization of GDH and GFAP in the rat hippocampal formation on postnatal days 5 (a, b), 10 (c, d), 20 (e, f) and 50 (g, h) after perfusion fixation with 2% alutaraldehyde. Micrographs in a, c, e, g show s&ions Gth intenseGDH labelling of glial cells. The neuripil was weakly stained. For comparison, sections b, d. f, h show GFAP-positive astrocytes. Scale bar = SOOpm
426
F.
ROTHE et
hippocampal formation. Generally, the external granule cell layer showed a low GDH-LI, whereas the formation of the inner granule cell layer was accompanied by an advancing degree of labelling. On days 15 and 20, the growing molecular layer displayed a diffuse staining of increasing density. On the same days, cell bodies of Bergmann glia, interspersed between Purkinje cell bodies, were markedly labelled. From day 35 onwards, the distribution of GDH-LI was identical to that of adult rats. DJSCUSSION
Recently, we reported on methodological aspects of the tetrazolium salt technique for the histochemical assay of GDH as applied in the present study. This technique is characterized by negative control experiments, sharply circumscribed distribution pattern of the reaction product (formazan), low formazan production in the incubation medium and the possibility of quantification and of controlling the reaction kinetics as shown by calibration curves.30-32With the aid of this technique, in both the hippocampal formation and cerebellar cortex of the rat the postnatal development of GDH activity revealed a distribution pattern that was remarkably similar to that of presumed aminoacidergic structures. Thus, the postnatal increase in GDH activity in the hippocampal formation3’,32 occurred particularly in the dendritic layers which are essentially provided with the glutamatergic and/or aspartatergic nerve endings. 7-gThe possible role of GDH as a transmitter glutamate-metabolizing enzyme was further strengthened by the similarity of developmental patterns of the enzyme activity and the functional maturation of hippocampal glutamatergic structures which are characterized by the postnatal increase in glutamate high-affinity uptake,*’ glutamate binding capacity,4 (6sZn) uptake capacity4 and kainate vulnerability39 as well as other glutamaterelated enzymes such as glutaminase@ or aspartate aminotransferase. 28Like the hippocampal formation, in the adult and developing rat cerebellar cortex the distribution of histochemically demonstrable GDH activity as reported previously4’ was shown to be consistent with functionally maturing structures which use dicarboxylic amino acids as neurotransmitter.*’ The histochemical results are in line with biochemical findings indicating the rapid postnatal
al
increase in GDH activity in homogenates of the brain regions mentioned,‘8,25 particularly during the critical phase of maturing glutamatergic or aspartatergic structures. No such increase in GDH activity was observed in nervous tissues in which glutamatergic and/or aspartatergic transmission processes are obviously absent, underlying the validity of the findings obtained for brain regions enriched with aminoacidergic structures.‘5 Controversially, the postnatal changes in GDH concentration revealed by immunohistochemical techniques and referred to as GDH-LI, differed in some respects from those in GDH-activity staining. In contrast to the very low level of GDH activity in the first postnatal week of life indicated by biochemical and histochemical findings, a relatively heavy GDH-LI was observed in the hippocampus and cerebellar cortex of five-day-old rats. Notably the molecular layer of cerebellar cortex of rats up to day 20 showed a more intense, diffuse GDH-LI than observed for adult rats. The changes in the distribution patterns of GDH-LI during postnatal development, which differed from that in GDH activity, may have been due to varied and/or insufficient penetration of antibodies and the PAP complex into specific tissue elements, irrespective of the addition of Triton to the incubation medium used. Differences between activity and immunohistochemical staining were not restricted to neuronal elements. In contrast to the activity staining of GDH, immunohistochemical findings suggest that GDH is primarily an astrocytic enzyme.2~3~‘4Js~‘g~38 As shown in the present study. during the postnatal development of both hippocampal formation and cerebellar cortex there was immunohistochemical evidence of GDH-positive glial cells occurring at an early stage. After birth, shape and distribution patterns of glial cells containing GDH-LI were found to be consistent with those of GFAP-positive cells. In the hippocampal formation of adult rats, a similar shape and distribution pattern of GDH-positive cells was histochemically demonstrable.3.‘4 Recently, Kugler” has reported a modification of activity staining for GDH, implying that the localization of GDH activity in the hippocampus was restricted to cells that appeared to be astroglial cells. However, the histochemical demonstrability of GDH activity depends on the Na+ concentration of the incubation medium. 33,42Following a reduction of Na+ to below 60 mM, we found that in the dendritic layer of the hippocampal formation the formazan
Fig. 8. Immunohistochemical localization of GDH (a, c, e, g) and GFAP (b, d, f, h) in the rat cerebellar cortex on postnatal days 5 (a, b), 15 (c, d), 20 (e, f) and 50 (g, h) after perfusion fixation with 2% glutaraldehyde. Micrographs a, c, e show a moderate to intense labelling in the maturing molecular layer and the inner granule cell layer. On day 50 (g), a marked reduction in GDH-LI was seen in the molecular layer, whereas the labelling of the granule cell layer was increased. In (c), (e) and (g) developing Bergmann elia show medominatine GDH-nositive cell bodies and processes. With GFAP-antiserum only shafts of iergmann’glia and astricytes in-the granule cell layer w&e stained (b, d, f, h). Scale bars = 100 pm (a-f) and (g, h).
Fig. 8a-h. 427
428
F. ROTHEef al.
production was restricted to cells which may be classed as astrocytes. In the cerebellar cortex, the entire molecular layer was almost free of formazan deposit when low Na+ concentrations were used. As distinct from the findings obtained by the immunohistochemical identification procedure, Bergmann glia were not visible. When the Na+ was elevated (optimum 130-150 mM), GDH activity was demonstrable to the full extent as shown in Figs 3 and 4. Immunohistochemically, bodies and fibres of Bergmann glia exhibited the most intense staining, thus supporting other findings.2,‘4~‘9~38 GFAP-LI present was mainly localized in the large shafts of Bergmann glia processes, being in agreement with previous results. l6 In the granular layer, we found a patch-like distribution of GDH-LI that was attributed to synaptic areas (glomeruli?) rather than to GFAP-positive astroglial cells.
CONCLUSION
From our results, activity and immunohistochemical staining methods have demonstrated GDH in glial and in neuronal cell elements, too. In spite of its ubiquitous distribution, there are indications of a functional relation of the enzyme to glutamatergic structures. In such structures, both cell types are considered to be quasi-symbiontically interrelated.13~22~23 In view of the conflicting results depending on the mode of the identification technique used, it is recommended that activity and immunohistochemical procedures be combined to study qualitative and quantitative aspects of GDH in nervous tissues. Acknowledgements-We would like to thank Mrs A. Tiedge, S. Tschom and I. Biick for their excellent technical assistance. The help of R. Giebson in revising the English manuscript is appreciated.
REFERENCES
1. Altman J. (1972) Postnatal development of the cerebellar cortex in the rat. II. Phase in the maturation of Purkinje cells and the molecular layer. J. camp. Neurol. 145, 399464. 2. Aoki C., Milner T. A., Berger S. B., Sheu K-F. R., Blass J. P. and Pickel V. M. (1987) Glial glutamate dehydrogenase: ultrastructural localization and regional distribution in relation to the mitochondrial enzyme, cytochrome oxidase. J. Neurosci. Res. 18. 305-318. 3. Aoki C., Milner T. A., Sheu K-F. R., Blass J. P. and Pickel V. M. (1987) Regional distribution of astrocytes with intense immunoreactivity for glutamate dehydrogenase in rat brain: implications for neuron-glia interactions in glutamate transmission. J. Neurosci. 7, 2214223 1. 4. Baudry M. and Lynch G. (1981) Hippocampal glutamate receptors. Molec. cell. Biochem. 38, 5-18. 5. Bayer S. A. (1985) Hippocampal region. In The Rut Nervous System (ed. Paxinos G.), Vol. 1, pp. 335-352. Academic Press, Sydney. 6. Colon A. D., Plaitakis A., Perakis A., Berl S. and Clarke D. D. (1986) Purification and characterization of a soluble and a particulate glutamate dehydrogenase from rat brain. J. Neurochem. 46, 1811-1819. 7. Cotman C. W., Monaghan D. T., Ottersen 0. P. and Storm-Mathisen J. (1987) Anatomical organization of excitatory amino acid receptors and their pathways. Trends Neurosci. 10, 273-280. 8. Cotman C. W. and Nadler J. V. (1981) Glutamate and aspartate as hippocampal transmitters: biochemical and pharmacological evidence. In Glutamate: Trunsmitter in the Central Nervous System (eds Roberts P. J., Storm-Mathisen J. and Johnston G. A. R.), pp. 117-154. John Wiley and Sons, Chichester. 9. Fagg G. E. and Foster A. C. (1983) Amino acid neurotransmitters and their pathways in the mammalian central nervous system. Neuroscience 9, 701-719. 10. Filla A., DeMichele G., Brescia Morra V., Palma V., Di Lauro A., Di Geronimo G. and Campanella G. (1986) Glutamate dehydrogenase in human brain: regional distribution and properties. J. Neurochem. 46, 422424. 11. Fischer H. F. (1985) L-Glutamate dehydrogenase from bovine liver. In Methods in Enzymology (ed. Meister A.), Vol. 113, pp. 16-27. Academic Press, Orlando, FL. 12. Fonnum F. (1984) Glutamate: a neurotransmitter in mammalian brain. J. Neurochem. 42, l-l 1. 13. Hertz L. and Schousboe A. (1980) Interactions between neurons and astrocytes in the turnover of GABA and glutamate. Brain Res. Bull. 5 (Suppl. 2), 389-395. 14. Kaneko T., Akiyama H. and Mizuno N. (1987) Immunohistochemical demonstration of glutamate dehydrogenase in astrocytes. Neurosci. Lett. 77, 171-175. 15. Kaneko T., Shigemoto R. and Mizuno N. (1988) Metabolism of glutamate and ammonia in astrocyte: an immunocytochemical study. Bruin Res. 457, 160-164. 16. Kosaka T. and Hama K. (1986) Three-dimensional structure of astrocytes in the rat dentate gyrus. J. camp. Neural. 249, 242-260. 17. Kugler P. (1988) The enzyme histochemistry of neurotransmitter metablism. In Neurotrunsmission in the Hippocampus (eds Frotscher M., Kugler P., Misgeld U. and Zilles K.). Springer, Berlin. Adv. Anat. Embryol. Cell Biol. 111, 40-60. 18. Leong S. F. and Clark J. B. (1984) Regional development of glutamate dehydrogenase in the rat brain. J. Neurochem. 43, 106-111. 19. Mad1 J. E., Clements J. R., Beitz A. J., Wenthold R. J. and Larson A. A. (1988) Immunocytochemical localization of glutamate dehydrogenase in mitochondria of the cerebellum: an ultrastructural study using a monoclonal antibody. Briin Res. 452, j9&2. 20. McCarthy A. D. and Tipton K. F. (1983) Glutamate dehydrogenase. In Glutamine, Glutamate and GABA in the Central Nervous System (eds Hertz L., Kvamme E., McGeer E. G. and Schousboe A.), pp. 19-32. Alan R. Liss, New York. 21 Michelson H. B. and Lothman E. W. (1989) An in vivo electrophysiological study of the ontogeny of excitatory and inhibitory processes in the rat hippocampus. Dev/ Brain Res. 47, 113-122.
Glutamate dehydrogenase in rat brain
429
22. Nicklas W. J. (1986) Glia-neuronal interrelationships in the metabolism of excitatory amino acids. In EXcirotorY Amino Acids (eds Roberts P. J., Storm-Mathisen J. and Bradford H. F.), pp. 574. Macmillan, London. 23. Nicklas W. J., Zeevalk G. and Hyndman A. (1987) Interactions between neurons and gha in ghttamate/~utamine compartmentation. Biochem. Sot. Trans. 15, 208-210. 24. Ottersen 0. P. and Storm-Math&n J. (1987) Localization of amino acid neurotransmitters by immunocytcchemistry. Trends Neurosci. 10, 250-255. 25. Rothe F., Schmidt W. and Wolf G. (1983) Postnatal changes in the activity of glutamate dehydrogenase and aspartate aminotransferase in the rat nervous system with special reference to the glutamate transmitter metabolism. De01 Brain Res. 11, 67-74. 26. Salganicoff L. and DeRobertis E. (1965) S&cellular distribution of the enzymes of the glutamic acid, glutamine and
y-aminobutyric
acid cycles in rat brain. J. Neurochem. 12, 287-309.
27. Scheidegger J. J. (1955) Une micro-method de l’immunoelectrophorese. Int. Archs Allergy appl. Zmmun., Base1 7, 103-l 10. 28. Schmidt W. and Wolf G. (1987) Activity of aspartate aminotransferase in the hippocampal formation of the rat during
postnatal development and after lesion of the hippocampal Schaffer’s collaterals: a quantitative histochernical study. Neurochem. Int. 11, 39-47. 29. Schmidt W. and Wolf G. (1988) High-affinity 30. 31. 32.
33. 34.
35. 36.
37. 38. 39.
40. 41.
42. 43.
44.
uptake of L-(3H)glutamate and D-(‘H)aspartate during postnatal development of the hippocampal formation: a quantitative autoradiographic study. Expl Brain Res. 70, 50-54. Schiinzel G. und Seidler E. (1984) Verbesserter topochemischer Dehydrogenase-Nachweis im ZNS mit einem Nitro-Monotetraxeliumsalx. Acta Histochem. 30 (Suppl.), 319-325. Schiinzel G. and Wolf G. (1982) Topographic and quantitative characteristics of glutamate dehydrogenase of the hippocampus formation during the postnatal development of the rat brain. Acta Histochem. 71, 145-151. Schiinzel G., Wolf G., Rothe F. and Seidler E. (1986) Histophotometric evaluation of glutamate dehydrogenase activity of the rat hippocampal formation during postnatal development, with special reference to the glutamate transmitter metabolism. Cell. molec. Neurobioi. 6, 3142. Schiinzel G., Wolf G., Rothe F. and Seidler E. (1986) Enzyme histochemical demonstration of sodium-dependent glutamate uptake in glutamatergic brain structures of the rat (in German). Acta Histochem. 79, 61-66. Stemberger L. A. (1979) Immunocytochemistry. John Wiley, New York. Storm-Mathisen J., Ottersen 0. P. and Fu-Long T. (1986) Antibodies for the localization of excitatory amino acids. In Excitatory Amino Acids (eds Roberts P. J., Storm-Math&r J. and Bradford H. F.), pp. 101-116. Macmillan, London. Svenneby G. and Storm-Math&n J. (1983) Immunological studies on phosphate activated glutaminase. In Glutamine, Glutamate and GABA in the Central Nervous System (eds Hertz L., Kvamme E., McGcer E. G. and Schousboe A.), pp. 69-76. Alan R. Liss, New York. Van den Berg C. J. (1974) Enzymes in the developing brain. In Biochemistry of the Developing Bruin (ed. Hinwich, W. A.), pp. 149-198. Marcel Dekker, New York. Wenthold R. J., Altschuler R. A., Skaggs K. K. and Reeks K. A. (1987) Immunocytochemical characterization of glutamate dehydrogenase in the cerebellum of the rat. J. Neurochem. 48, 636643. Wolf G. and Keilhoff G. (1984) Kainate and glutamate neurotoxicity in dependence of the postnatal development with special reference to hippocampal neurons. De01 Brain Res. 11, 15-21. Wolf G., Richter K., Schmidt W., Svenneby G. and Storm-Mathisen J. (1989) Glutaminase in the postnatally developing rat cerebellum: a comparison of activity staining and immunocytochemistry. Neurochem. Res. 14, 583-588. Wolf G. and Schiinxel G. (1987) Glutamate dehydrogenase in aminoacidergic structures of the postnatally developing rat cerebellum. Neurosci. Mt. 78, 7-l 1. Wolf G., Schiinxel G. and Rothe F. (1986) Histochemical demonstration of sodium-dependent glutamate uptake in brain tissues by glutamate dehydrogenase reaction. Expi Brain Res. 62, 659-662. Wolf G., Schiinzel G. and Storm-Mathisen J. (1984) Lesions of Schaffer’s collaterals in the rat hippocampusaffecting glutamate dehydrogenase and succinate dehydrogenase activity in the stratum radiatum of CA 1. A study with special reference to the glutamate transmitter metabolism. J. Himforsch. 25, 249-253. Wolf G., Schutte M. and Riimhild W. (1984) Uptake and subcellular distribution of 6’Zn in brain structures during the postnatal development of the rat. Neurosci. Lett. 51, 277-280. (Accepted 14 May 1990)
Neuroscience Vol.39,NO.2, PP.419-029,1990 Printedin Great Britain
IMMUNOHISTOCHEMICAL DEMONSTRATION OF GLUTAMATE DEHYDROGENASE IN THE POSTNATALLY DEVELOPING RAT HIPPOCAMPAL FORMATION AND CEREBELLAR CORTEX: COMPARISON TO ACTIVITY STAINING F. ROTHE, G. WOLF* and G. SCH~NZEL Institute of Biology, Medical Academy of Magdeburg, Erich-Weinert-Strass 3, D/O-3014 Magdeburg, Germany
Abstract-Distribution patterns of activity and immunohistcchemical staining for glutamate dehydrogenase were compared during the postnatal development of rat hippocampal formation and eerebellar cortex. On postnatal day 5, dendritic layers of the hippocampal formation showed only a very weak enzyme activity. Similarly, when studied at the same age, the external granule cell layer and Purkinje cells of the cerebellar cortex exhibited a very faint and moderate staining, respectively. With advancing age, in both brain regions a marked postnatal increase in glutamate dehydrogenase activity occurred in neuropil area as glutamate@ structures matured. However, compared to activity staining, both brain regions of early postnatal stages showed a relatively high level of glutamate dehydrogenase-like immunoreactivity. In this case, the immunohistochemical staining of hippocampal dendritic layers and of the molecular layer of the cerebellar cortex was rather diffuse, being not very similar to parameters of the maturation of the respective glutamatergic structures. In contrast to the activity staining for the enzyme, the immunohistochemical labelling in adult rats revealed a selective predominance of immunoreactivity in astroglial cells from postnatal day 5 onwards. The Bergmann glia in the cerebellar cortex exhibited the strongest intensity of immunoreactivity. Generally, the patterns of immunoreactivity were found to depend on the fixation procedure adopted. Concluding from our results, glutamate dehydrogenase is demonstrable in glial and in neuronal cell elements as well. Therefore, it is recommended that activity staining and the immunohistochemical procedure be combined to study qualitative and quantitative aspects of glutamate dehydrogenase in nervous tissues.
In mammalian brain, substantial amounts of glutamate dehydrogenase [L-glutamate: NAD(P)+ oxidoreductase, deaminating; E.C. l4.1.3; GDH] activity were reported.6,10,18GDH is known to be associated with mitochondria of brain,26337in which it catalyses the interconversion of glutamate and 2-oxoglutarate.“JO In addition to its function in common metabolism, glutamate is established as one of the major excitatory neurotransmitters in the CNS.12 Therefore, GDH might play a key role in the synthesis or degradation of transmitter glutamate. Thus, in brain regions which are enriched with glutamatergic structures the concentration and activity of GDH might be expected at higher levels when compared to other brain regions. Our previous findings have given biochemical25 and histochemica13’,32,4’a43 support to the suggestion that in the maturing rat hippocampal formation and cerebellum GDH is a transmitter glutamate-metabolizing enzyme the postnatal activity increase of which seems
to be related more to the neurons than to glial cells.
Lately, immunocytochemical studies have shown, however, a more intense GDH-like immunoreactivity (GDH-LI) in glial cells than in neurons.2,3,14,1g,38 In order to compare our previous histochemical results and elucidate the role of GDH in the glutamate transmitter metabolism we studied the distribution of GDH-LI in the hippocampal formation and cerebellar cortex of the rat at different ages. Both brain regions are endowed with glutamatergic structures and, moreover, their architecture is well characterized. Tissue elements including neuronal contacts are organized within the first three weeks of postnatal life.‘s5 EXPERIMENTAL PROCEDURES Animals
Male rats of our breeding stock (derived from VELAZ, Prague) aged five, 10, 15, 20, 30, 35, 40 and 50 days were used for the immunohistochemical procedure and enzyme activity staining. The rats were housed under standardized conditions.25 Enzyme activity staining
*To whom all correspondence should be addressed. Abbreviations: GDH, glutamate dehydrogenase; GDH-LI, glutamate dehydrogenase-like immunoreactivity; GFAP, glial fibrillary acidic protein; LDH, lactate dehydrogenase; PAP, peroxidase-antiperoxidase.
The rats were killed by decapitation under ether anaesthesia. The rapidly removed brains were instantly frozen on a metal block cooled with liquid nitrogen. The protocol used for GDH-activity staining on sagittal cryostat sections
419
420
F. ROTHEei al.
(20 pm) was described previously in detail.‘* Control staining was obtained by omitting substrate (glutamate) or coenzyme (NAD).
The animals were anaesthetized with hexobarbital (100 mg/kg body weight) and the perfusion fixation was initiated by intracardial flow of 2% dextran (mol. wt 70,000) in 0.1 M sodium phosphate buffer, pH 7.4, followed by different fixation solutions containing (i) 4% paraformaldehyde and 0.1% glutaraldehyde (Merck-Schuchardt, Munich, F.R.G.), (ii) 2% alutaraldehvde. or (iii) 5% glutaraldehyde, each diluted-in the same buffer. The perfusion was performed as a function of advancing age of the animals for 5-15 min. Fiften minutes after termination of perfusion the hippocampal formations and cerebella were dissected out. The tissue was then postfixed in the particular fixative for 30 min, followed by cutting into 300~pm slices
using a tissue chopper. Stretched hippocampi were cut transversally from the septal to the temporal region. Parasagittal slices were prepared from the cerebellum. Slices were collected into buffer containing 0.1% [after fixation in (i)] and 0.5% glutaraldehyde [for (ii) and (iii)] and stored at 4°C prior to use. After cryoprotection with 30% sucrose in 0.1 M sodium phosphate buffer, pH 7.4, overnight, the tissue was resectioned on a freezing microtome to obtain 20-pm sections. The latter were used immediately for immunohistochemical localization of GDH or glial fibrillary acidic protein (GFAP). Rabbit anti-GFAP antiserum was generously supplied by the laboratory of Prof. Dr J. Storm-Math&en (Anatomical Institute, University of Oslo, Oslo, Norway). The antiserum to GDH was raised in white New Zealand rabbits. Rabbits received an initial injection of 3 mg bovine liver GDH (Boehringer, Mannheim, F.R.G.) in complete Freund’s adjuvant and three booster injections of 1 mg of antigen in incomplete Freund’s adjuvant. The injections were carried out into the popliteal lymph nodes subcutaneously and intracutaneously. The time intervals allowed between immunizations were four to five weeks. Animals were bled by heart puncture one week after the last immunization. The antiserum was characterized by microimmunoelectrophoresis on agarose plates.*’ The electrophoresis was carried out with (i) bovine liver GDH and (ii) concentrated homogenate supematant of rat hippocampal formation. The immunoprecipitates were developed by use of the antiserum to bovine liver GDH. The identification and specificity control of GDH-containing immunoprecipitate was carried out by activity staining as described above for the localization in tissue. For controls, substrate for NAD was omitted or preimmune serum was used (Fig. 1). The technique reported by Ottersen and StormMathisen24 was adopted for another specificity test. Cellulose ester filters containing spots of different antigens (Fig. 2) were processed with the antiserum and visualized immunohistochemically. Bach spot contained one of several antigens: GDH, lactate dehydrogenase, bovine serum albumin;rat serum, and GDH glutaraldehyde conjugates. To eliminate lactate dehvdrogenase (LDH) antibodies (which were found in our &de GDH antiserum, possibly because of contamination of the commercial GDH preparation used as antigen), the GDH antiserum was absorbed with an LDH complex under slow stirring at room temperature for 2 h. To prepare the LDH complex, 10 mg LDH from bovine heart (VEB Arzneimittelwerk, Dresden, G.D.R.) were precipitated by 0.1% glutaraldehyde (about IOmM) in 2ml 0.1 M sodium phosphate buffer, pH 7.4, under ‘stirring at room temperature for 3 h and kept overnight at 4°C. Following the resuspension of the pellet in sodium phosphate buffer the precipitate was centrifuged four times at 1OOOg for 10min. Free aldehyde groups
were neutralized with I M ethanolamine in 0.1 M sodium phosphate buffer, pH 8.4. The immunohistochemical procedure was based on the peroxidase-antipcroxidase (PAP) technique.34 All incubation steps were carried out with free-floating sectionc .___(20 pm) in-O.1 M Tris-HCl buffer, pH 7.4, containing 0.3 M NaCl. 0.5% Triton X-100. and 1% normal eoat serum. The 20-pm sections were pretreated with 1 M ithanolamine in 0.1 M sodium phosphate buffer, pH 8.4, and 1% H,O, in methanol to remove endogenous peroxidase activity.‘” Following washing with phosphate-buffered saline, pH 7.4, and preincubation with Tris-HCl buffer containing -3% .” normal goat serum, the sections were incubated with specific absorbed GDH antiserum (diluted 1:400) in Tris--HP1 buffer overnight. The sections were then washed and incubated with anti-rabbit globulin (purchased from Staatliches Institut fur Immunpriiparate und Nihrmedien, Berlin, G.D.R.) diluted 1: 100, for 1 h. After washing, the sections were incubated in PAP complex (obtained from Sektion Biowisenschaften der Karl-Marx-Universitat, Leipzig, G.D.R.; diluted 1:500) for 1 h under gentle shaking. The peroxidase reaction was visualized using diaminobenzidine (Fluka, F.R.G.; 7.5 mg/15 ml sodium phosphate buffer, pH 7.4) and 0.02% H20,.‘5 Thereafter, sections were rinsed in distilled water and mounted in glycerol gelatin. Controls were treated with rabbit preimmune serum, with GDH antiserum absorbed with a GDH complex conjugated by glutaraldehyde (procedure by analogy to preparation of LDH complex), or by omitting the linker globulin. To compare the distribution patterns of GDH with those of astrocytes, rabbit antiserum to GFAP diluted I:500 was used in the same way as for GDH immunostaining. 1
Il..
RESULTS Glutamate dehydrogenase activity staining Hippocampal formation. In the hippocampal formation of five-day-old rats pyramidal and granular cell perikarya were predominantly stained (Fig. 3). Cell nuclei were nearly free of the reaction product, formazan. The dendritic layers generally revealed a weak staining. The distribution of GDH activity extended increasingly with advancing age of the rats, especially in neuropil layers. From day 30 to day 50 of life, an intense activity staining for GDH was found in all dendritic layers, particularly in the stratum lacunosum-moleculare and in the molecular layer of the dentate gyrus, whereas neuronal perikarya showed only a moderate GDH activity. Maximum staining intensity was seen when the concentration of sodium ions in the incubation medium ranged from 130 to 150 mM. No selective staining of glial cells was observed under these circumstances. Total sodium concentration diminishing to 60 mM or less, the neuropil areas of the hippocampal formation appeared to be almost free of GDH activity. In this case staining was restricted to perikarya of the strata pyramidale and granulosum, and to scattered cells in neuropil areas (Fig. 5). Cerebeilar cortex. In the cerebellar cortex (Fig. 4) of rats aged five and 10 days, histochemically demonstrable GDH activity revealed fairly low formazan deposits. The granular cell bodies of the external layer exhibited a very faint staining, whereas most of the perikarya of Purkinje cells were noted to be
Glutamate dehydrogenase in rat brain
421
Fig. la-c and 2. Fig. 1. Characterization of GDH antiserum by microimnmnoelectrophoresis. The antigen wells were tilled with concentrated homogenate supematant of rat hippocampal formation. After electrophoresis the trough between the antigen walls was filled with the antiserum against bovine liver GDH. (a) Dark-field micrograph shows precipitation arcs in the agarose gel. (b) Specimen (a) after GDH-activity staining. (c) GDH-activity stained arc (lower part of microslide) compared to control reaction (without glutamate;
upper part). Fig. 2. Specificity test by use of cellulose ester filters. Spots (1 pl) contained the following antigens: (1) GDH (20 mg protein/ml); (a) 1: 10; (b) 1: 100; (c) 1: 1000. (2) Complexed GDH (conjugated by glutaraldehyde). (3) Bovine serum albumin (20 mg/rnl). (4) Rat serum, non-diluted. (5) LDH (2.5 mg protein/ml); (a) non-diluted, (b) 1: 10; (c) 1: 100. The filters were processed by the immunohistochemical technique using the crude GDH-antiserum (B) and GDH-antiserum absorbed in LDH complexes (A, C). The crude antiserum reacted strongly with GDH and GDH complex (A) and weakly with LDH (B). Following absorption in LDH complex the antiserum did not react with LDH (C), nor did crude and absorbed antiserum with bovine serum albumin and rat serum (A).
markedly stained. With advancing age of rats, a dramatic increase in staining intensity was observed, being pronounced between days 15 and 40. In SO-dayold rats, GDH staining occurred throughout the molecular layer. Granule cell bodies, too, exhibited an intense staining. Spotted staining of the internal granule cell layer was a noticeable finding. Stained
patches are possibly identical to the cerebellar glomeruli. The enzyme activity in Purkinje cell perikarya was enhanced to a lesser extent. No selective staining of glial cells was observed. Generally, control experiments omitting substrate or coenzyme resulted in a very faint staining and poor contrast.
422
F. ROTHEet ul
Fig. 3a-c and 4a-c. Fig. 3 and 4. Hist~he~cal demonstration of glutamate dehydrogenase activity in parasagittal sections of rat hippocampal formation (Fig. 3) and cerebellar cortex (Fig. 4) on postnatal days 5 (a), 20 (b) and 50 (c), using the vitro-mono~tra~olium salt p-DhTT. Scale bars = 500 pm (Fig. 3): IOOpm (Fig. 4).
Immunohistochemistry Following the electrophoresis of bovine liver GDH and homogenate supernatant of the rat hippocampal formation, the anti-GDH serum revealed one pr~ipitation arc in each case. The pr~~pitation arc was stained by the histochemical reaction for
GDH. No staining was seen when the reaction specificity was tested by omitting glutamate as substrate (Fig. 1). The use of preimmune serum did not produce precipitation arcs. When cellulose ester filters were employed for testing (Fig. 2) the non-absor~d antiserum reacted very strongly with both the commercially obtained
Glutamate dehydrogenase in rat brain GDH preparation and the GDH glutaraldehyde complex, while the reaction with purified LDH was poor. The antiserum did not react with bovine serum albumin and rat serum. Following absorption of the antiserum in the LDH complex, no staining of LDH spots was observed. In contrast to the activity staining, nerve cell perikarya showed light, if any, GDH-LI. In neuronal elements and preferentially in glial cells, the intensity and localization of GDH-LI proved to be dependent on the fixation procedure (Fig. 6). Still, when subjected to an identical fixation procedure glial cells exhibited a different reaction as a function of the brain region studied. Thus, after fixation with paraformaldehyde+ttaraldehyde (i; see Experimental Procedures), only a very weak labelling of glial cells was observed in the hippocampal formation irrespective of the developmental stage (Fig. 6a), whereas fixation with 2 or 5% glutaraldehyde (ii and iii) resulted in a distinct visualization of glial cells (Figs 6c and 7). Bergmann glia of the cerebellar cortex was markedly visualized after fixation with paraformaldehyde-glutaraldehyde or glutaraldehyde (Figs 6d and 8).
423
Using preimmune serum for control sections, fixation with 5% glutaraldehyde resulted in a slight background staining, whereas fixation with 2% glutaraldehyde produced a completely negative result (Fig. 6b). Therefore, our development study of GDHLI patterns mainly refers to material fixed with 2% glutaraldehyde. Paraformaldehydeformation. Hippocampal glutaraldehyde being adopted, a weak but clear GDH-LI was found throughout the dendritic layers of the developmental stages studied. The intensity of staining exhibited by the hilus and the inner third of the molecular layer of the dentate gyrus was slightly higher than that of other areas of the hippocampal formation (Fig. 6a). In rats up to day 20, the cytoplasm of pyramidal and granule cells was nearly free of GDH-LI. The nuclei were unstained. With increasing age the cytoplasm of pyramidal cells showed a moderate GDH-LI, predominantly in the CA3 region. Following fixation with 2% glutaraldehyde, numerous GDH-labelled cells were seen from postnatal days 5 to 50. The cells were distributed more densely in the stratum lacunosum-moleculare and in the hilus of dentate gyrus. Other hippocampal regions
Fig. 5a and b. Fig. 5. GDH-activity staining in the hippocampal formation (a) and cerebellar cortex (b) at a concentration of 60 mM sodium ions in the incubation medium. Scale bars = 500 pm (a); 100 pm (b).
424
F. ROTIE et al.
,,
b
Fig. 6a-d. Fig. 6. Imrn~ohist~he~~l localization of GDH in the hipp~ampa1 formation (a, b, c) and cerebellar cortex (d) of SO-day-old rats as a function of the fixation procedure (a) paraformaidehyde-glutaraldehyde; (b) 2% glutaraldehyde; (c, d) 5% glutaraldehyde. Micrograph (b) shows a control section treated with preimmune serum. Scale bars = 5OOym (a, b); 250,um (c); IOO~rn (d).
exhibited a more uniform distribution of GDHpositive cells. In the mature hippocampal formation, the GDH-LI of the round or spindle-shaped cells and their proximal parts of processes was intense. Their distribution pattern (Fig. 7) was largely identical with that of GFAP-labelled cells; hence, these structures were considered to belong to an identical cell type, i.e. astroglial cells. Using GFAP staining, branched processes of such cells were visible to a greater extent, though. Cerebelhr cortex. The intensity of GRH labelling of Bergmann glial cell bodies and fibres was most pronounced in adult rats irrespective of the fixation medium used. Processes of Bergmann glia showed a punctate staining. Similarly heavily stained puncta apparently belonging to glial cells were observed in the white matter of the cerebellum. A rather diffuse labelling, moderate in intensity, was found throughout the molecular layer foliowing fixation with 5% glutaraldehyde (Fig. 6d). Fixation with 2% glutaraldehyde produced such a diffuse labelhng at immature stages (up to day 20) only. In adult animals
the neuropil of the molecular layer was found to be nearly free of labelling, except for Bergmann glial cells which contrasted distinctly from the background. Our developmentai studies (2% glutaraldehydefixed material, Fig. 8) showed that, at all stages, Purkinje and granule cell bodies were only lightly labelled with nuclei free of GDH-LI. The granule cell layer of adult rats exhibited a more patchy labelling which, most likely, reflected glomeruli being typical of this stratum. Staining the sections for GFAP visualized astroglial cells in the granule cell layer and in the white matter, in addition to fibres of Bergmann glia. Unlike in GDH immunostaining, GFAP-positive cell bodies of Bergmann glia were, if at all, seen at advanced stages only. The developmental patterns of immunola~lling for GDH were quite different from that seen in GDH-activity staining. In rats aged five and 10 days, the cerebellar cortex was more intensely stained by immunohistochemi~l iabeliing than by activity staining, a finding similar to that obtained for the
a,
Fig. 7. Immunohistochemical localization of GDH and GFAP in the rat hippocampal formation on postnatal days 5 (a, b), 10 (c, d), 20 (e, f) and 50 (g, h) after perfusion fixation with 2% alutaraldehyde. Micrographs in a, c, e, g show s&ions Gth intenseGDH labelling of glial cells. The neuripil was weakly stained. For comparison, sections b, d. f, h show GFAP-positive astrocytes. Scale bar = SOOpm
426
F.
ROTHE et
hippocampal formation. Generally, the external granule cell layer showed a low GDH-LI, whereas the formation of the inner granule cell layer was accompanied by an advancing degree of labelling. On days 15 and 20, the growing molecular layer displayed a diffuse staining of increasing density. On the same days, cell bodies of Bergmann glia, interspersed between Purkinje cell bodies, were markedly labelled. From day 35 onwards, the distribution of GDH-LI was identical to that of adult rats. DJSCUSSION
Recently, we reported on methodological aspects of the tetrazolium salt technique for the histochemical assay of GDH as applied in the present study. This technique is characterized by negative control experiments, sharply circumscribed distribution pattern of the reaction product (formazan), low formazan production in the incubation medium and the possibility of quantification and of controlling the reaction kinetics as shown by calibration curves.30-32With the aid of this technique, in both the hippocampal formation and cerebellar cortex of the rat the postnatal development of GDH activity revealed a distribution pattern that was remarkably similar to that of presumed aminoacidergic structures. Thus, the postnatal increase in GDH activity in the hippocampal formation3’,32 occurred particularly in the dendritic layers which are essentially provided with the glutamatergic and/or aspartatergic nerve endings. 7-gThe possible role of GDH as a transmitter glutamate-metabolizing enzyme was further strengthened by the similarity of developmental patterns of the enzyme activity and the functional maturation of hippocampal glutamatergic structures which are characterized by the postnatal increase in glutamate high-affinity uptake,*’ glutamate binding capacity,4 (6sZn) uptake capacity4 and kainate vulnerability39 as well as other glutamaterelated enzymes such as glutaminase@ or aspartate aminotransferase. 28Like the hippocampal formation, in the adult and developing rat cerebellar cortex the distribution of histochemically demonstrable GDH activity as reported previously4’ was shown to be consistent with functionally maturing structures which use dicarboxylic amino acids as neurotransmitter.*’ The histochemical results are in line with biochemical findings indicating the rapid postnatal
al
increase in GDH activity in homogenates of the brain regions mentioned,‘8,25 particularly during the critical phase of maturing glutamatergic or aspartatergic structures. No such increase in GDH activity was observed in nervous tissues in which glutamatergic and/or aspartatergic transmission processes are obviously absent, underlying the validity of the findings obtained for brain regions enriched with aminoacidergic structures.‘5 Controversially, the postnatal changes in GDH concentration revealed by immunohistochemical techniques and referred to as GDH-LI, differed in some respects from those in GDH-activity staining. In contrast to the very low level of GDH activity in the first postnatal week of life indicated by biochemical and histochemical findings, a relatively heavy GDH-LI was observed in the hippocampus and cerebellar cortex of five-day-old rats. Notably the molecular layer of cerebellar cortex of rats up to day 20 showed a more intense, diffuse GDH-LI than observed for adult rats. The changes in the distribution patterns of GDH-LI during postnatal development, which differed from that in GDH activity, may have been due to varied and/or insufficient penetration of antibodies and the PAP complex into specific tissue elements, irrespective of the addition of Triton to the incubation medium used. Differences between activity and immunohistochemical staining were not restricted to neuronal elements. In contrast to the activity staining of GDH, immunohistochemical findings suggest that GDH is primarily an astrocytic enzyme.2~3~‘4Js~‘g~38 As shown in the present study. during the postnatal development of both hippocampal formation and cerebellar cortex there was immunohistochemical evidence of GDH-positive glial cells occurring at an early stage. After birth, shape and distribution patterns of glial cells containing GDH-LI were found to be consistent with those of GFAP-positive cells. In the hippocampal formation of adult rats, a similar shape and distribution pattern of GDH-positive cells was histochemically demonstrable.3.‘4 Recently, Kugler” has reported a modification of activity staining for GDH, implying that the localization of GDH activity in the hippocampus was restricted to cells that appeared to be astroglial cells. However, the histochemical demonstrability of GDH activity depends on the Na+ concentration of the incubation medium. 33,42Following a reduction of Na+ to below 60 mM, we found that in the dendritic layer of the hippocampal formation the formazan
Fig. 8. Immunohistochemical localization of GDH (a, c, e, g) and GFAP (b, d, f, h) in the rat cerebellar cortex on postnatal days 5 (a, b), 15 (c, d), 20 (e, f) and 50 (g, h) after perfusion fixation with 2% glutaraldehyde. Micrographs a, c, e show a moderate to intense labelling in the maturing molecular layer and the inner granule cell layer. On day 50 (g), a marked reduction in GDH-LI was seen in the molecular layer, whereas the labelling of the granule cell layer was increased. In (c), (e) and (g) developing Bergmann elia show medominatine GDH-nositive cell bodies and processes. With GFAP-antiserum only shafts of iergmann’glia and astricytes in-the granule cell layer w&e stained (b, d, f, h). Scale bars = 100 pm (a-f) and (g, h).
Fig. 8a-h. 427
428
F. ROTHEef al.
production was restricted to cells which may be classed as astrocytes. In the cerebellar cortex, the entire molecular layer was almost free of formazan deposit when low Na+ concentrations were used. As distinct from the findings obtained by the immunohistochemical identification procedure, Bergmann glia were not visible. When the Na+ was elevated (optimum 130-150 mM), GDH activity was demonstrable to the full extent as shown in Figs 3 and 4. Immunohistochemically, bodies and fibres of Bergmann glia exhibited the most intense staining, thus supporting other findings.2,‘4~‘9~38 GFAP-LI present was mainly localized in the large shafts of Bergmann glia processes, being in agreement with previous results. l6 In the granular layer, we found a patch-like distribution of GDH-LI that was attributed to synaptic areas (glomeruli?) rather than to GFAP-positive astroglial cells.
CONCLUSION
From our results, activity and immunohistochemical staining methods have demonstrated GDH in glial and in neuronal cell elements, too. In spite of its ubiquitous distribution, there are indications of a functional relation of the enzyme to glutamatergic structures. In such structures, both cell types are considered to be quasi-symbiontically interrelated.13~22~23 In view of the conflicting results depending on the mode of the identification technique used, it is recommended that activity and immunohistochemical procedures be combined to study qualitative and quantitative aspects of GDH in nervous tissues. Acknowledgements-We would like to thank Mrs A. Tiedge, S. Tschom and I. Biick for their excellent technical assistance. The help of R. Giebson in revising the English manuscript is appreciated.
REFERENCES
1. Altman J. (1972) Postnatal development of the cerebellar cortex in the rat. II. Phase in the maturation of Purkinje cells and the molecular layer. J. camp. Neurol. 145, 399464. 2. Aoki C., Milner T. A., Berger S. B., Sheu K-F. R., Blass J. P. and Pickel V. M. (1987) Glial glutamate dehydrogenase: ultrastructural localization and regional distribution in relation to the mitochondrial enzyme, cytochrome oxidase. J. Neurosci. Res. 18. 305-318. 3. Aoki C., Milner T. A., Sheu K-F. R., Blass J. P. and Pickel V. M. (1987) Regional distribution of astrocytes with intense immunoreactivity for glutamate dehydrogenase in rat brain: implications for neuron-glia interactions in glutamate transmission. J. Neurosci. 7, 2214223 1. 4. Baudry M. and Lynch G. (1981) Hippocampal glutamate receptors. Molec. cell. Biochem. 38, 5-18. 5. Bayer S. A. (1985) Hippocampal region. In The Rut Nervous System (ed. Paxinos G.), Vol. 1, pp. 335-352. Academic Press, Sydney. 6. Colon A. D., Plaitakis A., Perakis A., Berl S. and Clarke D. D. (1986) Purification and characterization of a soluble and a particulate glutamate dehydrogenase from rat brain. J. Neurochem. 46, 1811-1819. 7. Cotman C. W., Monaghan D. T., Ottersen 0. P. and Storm-Mathisen J. (1987) Anatomical organization of excitatory amino acid receptors and their pathways. Trends Neurosci. 10, 273-280. 8. Cotman C. W. and Nadler J. V. (1981) Glutamate and aspartate as hippocampal transmitters: biochemical and pharmacological evidence. In Glutamate: Trunsmitter in the Central Nervous System (eds Roberts P. J., Storm-Mathisen J. and Johnston G. A. R.), pp. 117-154. John Wiley and Sons, Chichester. 9. Fagg G. E. and Foster A. C. (1983) Amino acid neurotransmitters and their pathways in the mammalian central nervous system. Neuroscience 9, 701-719. 10. Filla A., DeMichele G., Brescia Morra V., Palma V., Di Lauro A., Di Geronimo G. and Campanella G. (1986) Glutamate dehydrogenase in human brain: regional distribution and properties. J. Neurochem. 46, 422424. 11. Fischer H. F. (1985) L-Glutamate dehydrogenase from bovine liver. In Methods in Enzymology (ed. Meister A.), Vol. 113, pp. 16-27. Academic Press, Orlando, FL. 12. Fonnum F. (1984) Glutamate: a neurotransmitter in mammalian brain. J. Neurochem. 42, l-l 1. 13. Hertz L. and Schousboe A. (1980) Interactions between neurons and astrocytes in the turnover of GABA and glutamate. Brain Res. Bull. 5 (Suppl. 2), 389-395. 14. Kaneko T., Akiyama H. and Mizuno N. (1987) Immunohistochemical demonstration of glutamate dehydrogenase in astrocytes. Neurosci. Lett. 77, 171-175. 15. Kaneko T., Shigemoto R. and Mizuno N. (1988) Metabolism of glutamate and ammonia in astrocyte: an immunocytochemical study. Bruin Res. 457, 160-164. 16. Kosaka T. and Hama K. (1986) Three-dimensional structure of astrocytes in the rat dentate gyrus. J. camp. Neural. 249, 242-260. 17. Kugler P. (1988) The enzyme histochemistry of neurotransmitter metablism. In Neurotrunsmission in the Hippocampus (eds Frotscher M., Kugler P., Misgeld U. and Zilles K.). Springer, Berlin. Adv. Anat. Embryol. Cell Biol. 111, 40-60. 18. Leong S. F. and Clark J. B. (1984) Regional development of glutamate dehydrogenase in the rat brain. J. Neurochem. 43, 106-111. 19. Mad1 J. E., Clements J. R., Beitz A. J., Wenthold R. J. and Larson A. A. (1988) Immunocytochemical localization of glutamate dehydrogenase in mitochondria of the cerebellum: an ultrastructural study using a monoclonal antibody. Briin Res. 452, j9&2. 20. McCarthy A. D. and Tipton K. F. (1983) Glutamate dehydrogenase. In Glutamine, Glutamate and GABA in the Central Nervous System (eds Hertz L., Kvamme E., McGeer E. G. and Schousboe A.), pp. 19-32. Alan R. Liss, New York. 21 Michelson H. B. and Lothman E. W. (1989) An in vivo electrophysiological study of the ontogeny of excitatory and inhibitory processes in the rat hippocampus. Dev/ Brain Res. 47, 113-122.
Glutamate dehydrogenase in rat brain
429
22. Nicklas W. J. (1986) Glia-neuronal interrelationships in the metabolism of excitatory amino acids. In EXcirotorY Amino Acids (eds Roberts P. J., Storm-Mathisen J. and Bradford H. F.), pp. 574. Macmillan, London. 23. Nicklas W. J., Zeevalk G. and Hyndman A. (1987) Interactions between neurons and gha in ghttamate/~utamine compartmentation. Biochem. Sot. Trans. 15, 208-210. 24. Ottersen 0. P. and Storm-Math&n J. (1987) Localization of amino acid neurotransmitters by immunocytcchemistry. Trends Neurosci. 10, 250-255. 25. Rothe F., Schmidt W. and Wolf G. (1983) Postnatal changes in the activity of glutamate dehydrogenase and aspartate aminotransferase in the rat nervous system with special reference to the glutamate transmitter metabolism. De01 Brain Res. 11, 67-74. 26. Salganicoff L. and DeRobertis E. (1965) S&cellular distribution of the enzymes of the glutamic acid, glutamine and
y-aminobutyric
acid cycles in rat brain. J. Neurochem. 12, 287-309.
27. Scheidegger J. J. (1955) Une micro-method de l’immunoelectrophorese. Int. Archs Allergy appl. Zmmun., Base1 7, 103-l 10. 28. Schmidt W. and Wolf G. (1987) Activity of aspartate aminotransferase in the hippocampal formation of the rat during
postnatal development and after lesion of the hippocampal Schaffer’s collaterals: a quantitative histochernical study. Neurochem. Int. 11, 39-47. 29. Schmidt W. and Wolf G. (1988) High-affinity 30. 31. 32.
33. 34.
35. 36.
37. 38. 39.
40. 41.
42. 43.
44.
uptake of L-(3H)glutamate and D-(‘H)aspartate during postnatal development of the hippocampal formation: a quantitative autoradiographic study. Expl Brain Res. 70, 50-54. Schiinzel G. und Seidler E. (1984) Verbesserter topochemischer Dehydrogenase-Nachweis im ZNS mit einem Nitro-Monotetraxeliumsalx. Acta Histochem. 30 (Suppl.), 319-325. Schiinzel G. and Wolf G. (1982) Topographic and quantitative characteristics of glutamate dehydrogenase of the hippocampus formation during the postnatal development of the rat brain. Acta Histochem. 71, 145-151. Schiinzel G., Wolf G., Rothe F. and Seidler E. (1986) Histophotometric evaluation of glutamate dehydrogenase activity of the rat hippocampal formation during postnatal development, with special reference to the glutamate transmitter metabolism. Cell. molec. Neurobioi. 6, 3142. Schiinzel G., Wolf G., Rothe F. and Seidler E. (1986) Enzyme histochemical demonstration of sodium-dependent glutamate uptake in glutamatergic brain structures of the rat (in German). Acta Histochem. 79, 61-66. Stemberger L. A. (1979) Immunocytochemistry. John Wiley, New York. Storm-Mathisen J., Ottersen 0. P. and Fu-Long T. (1986) Antibodies for the localization of excitatory amino acids. In Excitatory Amino Acids (eds Roberts P. J., Storm-Math&r J. and Bradford H. F.), pp. 101-116. Macmillan, London. Svenneby G. and Storm-Math&n J. (1983) Immunological studies on phosphate activated glutaminase. In Glutamine, Glutamate and GABA in the Central Nervous System (eds Hertz L., Kvamme E., McGcer E. G. and Schousboe A.), pp. 69-76. Alan R. Liss, New York. Van den Berg C. J. (1974) Enzymes in the developing brain. In Biochemistry of the Developing Bruin (ed. Hinwich, W. A.), pp. 149-198. Marcel Dekker, New York. Wenthold R. J., Altschuler R. A., Skaggs K. K. and Reeks K. A. (1987) Immunocytochemical characterization of glutamate dehydrogenase in the cerebellum of the rat. J. Neurochem. 48, 636643. Wolf G. and Keilhoff G. (1984) Kainate and glutamate neurotoxicity in dependence of the postnatal development with special reference to hippocampal neurons. De01 Brain Res. 11, 15-21. Wolf G., Richter K., Schmidt W., Svenneby G. and Storm-Mathisen J. (1989) Glutaminase in the postnatally developing rat cerebellum: a comparison of activity staining and immunocytochemistry. Neurochem. Res. 14, 583-588. Wolf G. and Schiinxel G. (1987) Glutamate dehydrogenase in aminoacidergic structures of the postnatally developing rat cerebellum. Neurosci. Mt. 78, 7-l 1. Wolf G., Schiinxel G. and Rothe F. (1986) Histochemical demonstration of sodium-dependent glutamate uptake in brain tissues by glutamate dehydrogenase reaction. Expi Brain Res. 62, 659-662. Wolf G., Schiinzel G. and Storm-Mathisen J. (1984) Lesions of Schaffer’s collaterals in the rat hippocampusaffecting glutamate dehydrogenase and succinate dehydrogenase activity in the stratum radiatum of CA 1. A study with special reference to the glutamate transmitter metabolism. J. Himforsch. 25, 249-253. Wolf G., Schutte M. and Riimhild W. (1984) Uptake and subcellular distribution of 6’Zn in brain structures during the postnatal development of the rat. Neurosci. Lett. 51, 277-280. (Accepted 14 May 1990)
