Reduction of nerve growth factor level in the brain of genetically ataxic mice (weaver, reeler)

By a highly sensitive enzyme immunoassay we measured the level of nerve growth factor (NGF) in the cerebellum and cerebrum of the neurologically mutant mice, weaver, reeler and Purkinje cell degeneration (PCD). A significant decrease in NGF level was observed in both cerebellum and cerebrum of weaver and reeler mutants of either sex. However, there was no such difference between normals and mutants in the case of the PCD mice. These results show that weaver and reeler mice have abnormalities of NGF synthesis and/or degradation not only in the cerebellum but also in the cerebrum.     

Cerebellar Hypoplasia in the Gunn Rat with Hereditary Hyperbilirubinemia: Immunohistochemical and Neurochemical Studies

Immunohistochemical reactions were conducted, using the antibodies against GFA and S-100 proteins on sections of cerebellum from the homozygous (jj) and the heterozygous (Jj) Gunn rats. Hypertrophy of the fibrous astrocytes was observed but hyperplasia of the glial cells was not. Although the molecular layer was very thin, the Bergmann fibre appeared normal. Among the free amino acids in the cerebellum from the jj rat, glutamate concentration decreased to two-thirds of the control level. The protein profile of the cerebellum from the jj rat obtained by SDS-polyacrylamide gel electrophoresis revealed that the amount of P400 protein that is characteristic of Purkinje cells decreased considerably and there were also some changes of the other unidentified proteins. By two-dimensional electrophoresis, it was observed that in the supernatant from the jj rat cerebellum one protein spot diminished and in the particulate fraction from the jj rat one spot was enormously increased. The activity of 2',3'-cyclic nucleotide 3'-phosphohydrolase (CNPase) in the cerebellum from the jj rat did not differ significantly from that of the control; however, activities of choline acetyltransferase and acetylcholinesterase of the jj rat were about twice as high as those of the control. 2-Deoxyglucose incorporation was maximum in the granular layer from both the jj and the Jj rat cerebella. However, the incorporation in the jj cerebellum was not higher than in the Jj control and even lower in some parts of the jj cerebellum than in the control Jj cerebellum.     

Generation of a monoclonal antibody against the myelin protein CNP (2′,3′-cyclic nucleotide 3′-phosphodiesterase) suitable for biochemical and for immunohistochemical investigations of CNP

The functional role of CNP (2,3-cyclic nucleotide 3-phosphodiesterase), a minor component of central and peripheral myelin is still unclear. Here we describe preparation of a monoclonal antibody directed against CNP. The antibody, of the immunoglobulin IgG₁ type, raised with a basic 46 kDa membrane-associated protein solubilized from pig cerebellar membranes, can be used to detect immunoreactivity in solubilized brain homogenates from pig, mouse, rat, sheep, cow and man, in cerebrum and cerebellum, but not in other tissues such as liver, skeletal and heart muscle. The antibody recognizes the CNP doublet band and shows no cross-reactivity with any of the other brain proteins solubilized. In tissue sections from paraformaldehyde-fixed rat brain the antigen was localized in oligodendrocytes. In cultured glial cells from newborn mice the antibody stained cells which were identified as oligodendrocytes by co-localization of myelin basic protein. Even cells from a C6 rat glioma cell line, which contain very little of CNP, were labeled by the monoclonal antibody. Thus the monoclonal antibody recognizing CNP from several species is suitable for immunocytochemical investigations and also for biochemical studies of CNP, since the antibody has been employed for immunoprecipitation and immunopurification of CNP in crude brain homogenates.     

Comparative Study of Glial Marker Proteins in the Hypoplastic Cerebellum of Jaundiced Gunn Rats

The behavior of marker proteins of glial cells [alpha-enolase, beta-S100 protein, and glial fibrillary acidic protein (GFAP)] was investigated quantitatively by using enzyme immunoassay systems during the development of cerebellar hypoplasia in jaundiced Gunn rats. A neuronal marker protein, gamma-enolase, was also measured as a reference. At postnatal day 8 corresponding to the early stage of cerebellar damage, the amount of beta-S100 on a protein basis was significantly higher in jaundiced homozygotes (jj) than in control nonjaundiced heterozygotes (j+), whereas no differences in alpha- and gamma-enolases and GFAP were observed between the two groups of rats. At days 15 and 30, which correspond, respectively, to the advanced and late stages of cerebellar damage, the three glial proteins, especially GFAP, were higher and the neuronal protein was lower in the jj rat cerebellum than in the control. These results are consistent with the reported histological observations that neuronal cells are vulnerable and damaged by bilirubin, whereas glial cells seem to be less sensitive. On the other hand, the amounts of beta-S100 and alpha-enolase per cerebellum were significantly lower in jj rats at days 15 and 30, as in the case of gamma-enolase, whereas that of GFAP remained at the same level as the control at day 15 and showed a slight but significant decrease at day 30. The possibility is suggested that beta-S100 and GFAP may be available as biochemical indicators of glial cells, especially in the early and advanced stages of cerebellar damage, respectively, but that alpha-enolase is less available.     

Galactosyltransferase Defects in Reeler Mouse Brains

Galactosyltransferase activities were examined in the cerebellum, cerebral cortex, and brain stem of reeler and wild-type mice. Galactosyltransferase assays were optimal for all required substrates, linear with incubation time, and proportional to protein concentration. In brain areas affected by the reeler mutation (i.e., cerebral cortex and cerebellum), galactosylation of both endogenous and exogenous glycoprotein acceptors was greatly reduced in reeler relative to controls. On the other hand, glycosylation of endogenous glycolipids was low, and equal between reeler and wild-type. Galactosyltransferase activities were similar, though not identical, in reeler and wild-type brain stems, which are phenotypically normal in reeler mice. Glucosyltransferase, beta-galactosidase, beta-N-acetylglucosaminidase, acid phosphatase, and lactate dehydrogenase specific activities were all unaffected in reeler cerebella, while galactosyltransferase activity was 52% of control. Inhibition of either UDPgalactose hydrolysis or beta-galactosidase had no effect on galactosyltransferase activity. The spectrum or galactosyltransferase deficiencies in reeler suggests that this enzyme is associated with the development of young granule cells.     

Induction of 2′,3′-cyclic nucleotide 3′-phosphohydrolase and morphological alterations in C6 glioma cells by dexamethasone, (3-butoxy-4-methoxybenzyl)-2-imidazolinone and prostaglandin E1

Summary Dexamethasone, R020-1724 and prostaglandin E₁ all induced morphological alterations and increased the glial specific enzyme 2′,3′-cyclic nucleotide 3′-phosphohydrolase (CNP) in rat C6 glioma cells in culture. Morphological alterations consisted mainly in the development of astrocytelike changes. Increases in dexamethasone-induced CNP activity was time dependent. Dexamethasone reduced cell growth rate, depending on the concentration employed. This paper is supported in part by N.I.M.H. Research Grant AA02372. Dr. S. Kamath participated initially in this study.     

Early and late passage C-6 glial cell growth: Similarities with primary glial cells in culture

Earlier studies in our laboratory have shown that C-6 glial cells in culture exhibit astrocytic properties with increasing cell passage. In this study, we tested the responsiveness of early and late passage C-6 glial cells to various cultures conditions: culture substrata (collagen, poly-L-lysine, plastic), or supplements for the culture medium, DMEM, [fetal calf, or heat inactivated (HI) serum, or media conditioned from mouse neuroblastoma cells (NBCM) or primary chick embryo cultured neurons (NCM)]. Glutamine synthetase (GS) and cyclic nucleotide phosphohydrolase (CNP), astrocytic and oligodendrocytic glial markers, were used. Cell numer and protein content increased exponentially with days in culture regardless of the type of the substratum or cell passage. Differences in cell morphology among the three types of substratum were also reflected on GS activity, which rose by three-fold on culture day 3 for cells grown on collagen; thereafter, GS profiles were similar for all substrata. This early rise in GS is interpreted to reflect differential cell adhesion processes on the substrata; specifically, cell adhesion on the collagen stimulated differentiation into astrocytic phenotype.Analogous to immature glia cells in primary cultures, early passage C-6 glial cells responded to neuronal factors supplied either from NCM or NBCM by expressing reduced GS activity, the astrocytic marker and enhanced CNP activity, the oligodendrocytic marker. Thus, early passage cells can be induced to express either astrocytic or oligodendrocytic phenotype. In accordance with our previous reports on primary glial cells, late passage C-6 cells exhibit their usual astrocytic behavior, responding to serum factors with GS activity. Moreover, whereas NCM or NBCM alone markedly lowered GS activity, a combination with serum restored activity. The present findings confirm our previous observations and further establish the C-6 glial cells as a reliable model to study immature glia.Special issue dedicated to Dr. Paola S. Timiras.     

Biochemical Correlates of Epilepsy in the El Mouse: Analysis of Glial Fibrillary Acidic Protein and Gangliosides

The E1 (epileptic) mouse is considered a model for complex partial seizures in humans. Seizures in E1 mice begin around 7-8 weeks of age and persist throughout life. To determine if astrocytic gliosis was present in adult seizing E1 mice, the distribution of glial fibrillary acidic protein (GFAP) was studied in the hippocampus using an antibody to GFAP. The mean number of GFAP-positive cells per square millimeter of hippocampus was approximately 15- to 40-fold higher in adult E1 mice than in nonseizing control C57BL/6J (B6) mice or in young nonseizing E1 mice. Relative GFAP concentration (expressed per milligram of total tissue protein) in hippocampus and cerebellum was estimated by densitometric scanning of peroxidase-stained western blots. GFAP concentration was 2.7-fold greater in hippocampus of adult seizing E1 mice than in the control B6 mice. No differences in GFAP content were detected between the strains in the cerebellum. Because gangliosides can serve as cell surface markers for changes in neuronal cytoarchitecture, they were analyzed to determine if the gliotic response in E1 mice was associated with changes in neural composition. Although the total ganglioside concentration of hippocampus, cerebral cortex, and cerebellum was similar in adult E1 and control B6 mice, a synaptic membrane enriched ganglioside, GD1a, was elevated in the adult E1 cerebral cortex and hippocampus. The findings indicate that E1 mice express a type of gliosis that is not accompanied by obvious neuronal loss.     

EFFECTS OF CORTICOSTEROIDS ON THE BIOCHEMICAL MATURATION OF RAT BRAIN: POSTNATAL CELL FORMATION

(1) Treatment with cortisol acetate (0.2 mg daily during the first 4 days after birth) reduced the rate of growth in the rat: at 35 days of age the body weight was reduced by 50 per cent and the brain weight, depending on the region, by up to 30 per cent. (2) In the brain the normal increase in cell number was severely inhibited during the period of cortisol treatment; this resulted in a final deficit in cell number of about 20 per cent in the cerebrum and 30 per cent in the cerebellum. (3) To determine whether cortisol affected primarily cell formation or cell destruction the labelling of brain DNA was studied 1 h after a subcutaneous injection of 20 Ci/100 g [2-¹⁴C]thymidine. In the controls the amount of labelled DNA increased by a factor of two in the cerebrum and seven in the cerebellum during the period 2-13 days, and it decreased to 40 and 27 per cent of the peak values in the cerebrum and cerebellum respectively in the following 7 days. The results indicated that mitotic activity is higher in the cerebellum than in the cerebrum in the 2nd week of life. It would appear that in the cerebrum appreciable cell death accompanies new cell formation, especially during the period 13-35 days of age. (4) Cortisol treatment affected cell division rather than cell destruction in the brain since it strongly inhibited the incorporation of [2-¹⁴C]thymidine into DNA. The inhibition was severe during the period of treatment but it did not result in a lasting fall in mitotic activity. At the age of 13 days the amount of labelled DNA formed approached the normal level and it was twice that in controls at 20 days, indicating a tendency for compensating cell deficit by an accelerated mitotic activity. Nevertheless, massive cell proliferation ceased at about the same age as in normals; the labelling of DNA decreased markedly between 13 and 20 days after birth, and the DNA content did not increase after the age of 20 days. (5) In contrast to the marked effect on cell number, cortisol treatment did not influence significantly the maturational changes related to average cell size (DNA concentration) or the chemical composition of cells (RNA/DNA and protein/DNA).     

Morphological and Biochemical Studies on the Cerebral Cortex from Reeler Mutant Mice: Development of Cortical Layers and Metabolic Mapping by the Deoxyglucose Method

Abstract: Cerebral cortex from reeler mutant mice was examined morphologically and biochemically. The sequential process of postnatal cell migration in the cerebral cortex of reeler (rl/rl) was examined morphologically. The dense cellular cortical plate lies below the molecular layer near the cerebral surface just after birth in normal mice while in reeler most of the cells are concentrated in the center of the cortex. In the cortex of adult reeler, the broad laminar structure of the neurons could be seen to form inverted positions in the cortical layers. The total wet weight, and the concentration of DNA and RNA in the pallium cerebri from reeler did not differ significantly from those in the control. As to the protein profiles of the pallium cerebri detected by SDS- polyacrylamide gel electrophoresis, no significant differences were observed. Activities of CNPase (2',3'-cyclic nucleotide 3'-phosphohydrolase), which is a myelin enzyme of CNS, and choline acetyltransferase were at the same level in both the reeler and the control. Therefore, reeler mutation does not appear to affect the genetically determined cell numbers, number of cholinergic fibers, and myelination. By autoradiographic observation of the cerebral cortex after intraperitoneal injection of [¹⁴C]2-deoxyglucose, it was revealed that 2-deoxyglucose was incorporated intensively into the fourth layer (granular layer) of the cerebrum from the control. In reeler it was also incorporated into the granular layer but in a more widespread distribution. We conclude that terminals to the granular layer make metabolically active synapse, perhaps even in a manner inverted from normal.     

Brain DNA metabolism in myelin deficient mutant jp,jpmsd and qk mice

We studied metabolism of brain DNA in three myelin deficient mutants qk, jp and jp^msd mice. The DNA content, the in vivo incorporation of [¹⁴C]thymidine in DNA and the activity of acid DNase in tissues (cerebellum and cerebrum) from normal littermates and affected mice were compared. The results showed that neither the DNA content, the incorporation of [¹⁴C]thymidine in DNA nor the activity of acid DNase in brain were altered in qk affected mice. In jp^msd mice, however, the DNA content as well as the incorpation of thymidine in DNA were reduced in both cerebellum and cerebrum, but the activity of acid DNase was reduced in cerebrum only. In jp mice, although the DNA content was reduced in both cerebellum and cerebrum, the incorporation of thymidine in DNA and the activity of acid DNase were reduced in cerebrum only. The data suggest a) that in qk mutants DNA metabolism and hence cell (glial) proliferation is not affected; b) that in jp^msd mutants DNA synthesis, and thus the cell proliferation is reduced in cerebellum as well as in cerebrum of the affected mice and c) that in jp mutants the synthesis of DNA and the cell proliferation is reduced in cerebrum but not in cerebellum.     

Selective Increase in S-100β Protein by Aging in Rat Cerebral Cortex

Changes in the concentrations of nervous tissue-related proteins and their isoproteins, such as S-100 proteins (S-100 alpha and S-100 beta), enolase isozymes (alpha-enolase and gamma-enolase), and GTP-binding proteins (Go alpha, Gi2 alpha, and beta-subunits), were determined in the CNS of male rats of various ages (from 2 to 30 months old) by means of enzyme immunoassay. The weights of brains and the concentrations of soluble proteins in the cerebral cortex, cerebellum, and brainstem were constant during the observation period. The concentration of S-100 beta protein, which is predominantly localized in glial cells, increased gradually in the cerebral cortex with age; levels in the 25-month-old rats increased to approximately 150% of the levels in the young (2-month-old) rats. However, the S-100 beta concentrations in the cerebellum and brainstem were relatively constant, showing similar values in rats 2-30 months old. Levels of other proteins, including both neuronal (gamma-enolase and Go alpha) and glial (alpha-enolase and S-100 alpha) marker proteins, did not change significantly with age in the cerebral cortex, cerebellum, and brainstem. These results suggest that there is a close relation between the age-dependent changes of the CNS function and S-100 beta protein levels in the cerebral cortex.     

Differential immunocytochemical staining for glial fibrillary acidic (GFA) protein,S-100 protein and glutamine synthetase in the rat subcommissural organ,nonspecialized ventricular ependyma and adjacent neuropil

Summary Antibodies raised against glial fibrillary acidic protein (GFA), S-100 protein (S100) and glutamine synthetase (GS) are currently used as glial markers. The distribution of GFA, S100 and GS in the ependyma of the rat subcommissural organ (SCO), as well as in the adjacent nonspecialized ventricular ependyma and neuropil of the periaqueductal grey matter, was studied by use of the immunocytochemical peroxidase-antiperoxidase technique. In the neuropil, GFA, S100 and GS were found in glial elements, i.e., in fibrous (GFA, S100) and protoplasmic astrocytes (S100, GS). The presence of S100 in the majority of the ventricular ependymal cells and tanycytes, and the presence of GFA in a limited number of ventricular ependymal cells and tanycytes confirm the glial nature of these cells. The absence of S100, GFA and GS from the ependymocytes of the SCO, which are considered to be modified ependymal cells, suggests either a non-astrocytic lineage of these cells or an extreme specialization of the SCO-cells as glycoprotein-synthesizing and secreting elements, a process that may have led to the disappearance of the glial markers.     

Relation of Cholesterol to Oligodendroglial Differentiation in C-6 Glial Cells

Abstract: The relation of cellular cholesterol content to a biochemical expression of oligodendroglial differentiation was studied in cultured C-6 glial cells. Induction of the oligodendroglial marker enzyme 2′: 3′-cyclic nucleotide 3′-phosphohydrolase (CNP) was determined after alteration of the sterol content of cellular membranes by exposure to compactin, a specific inhibitor of 3-hydroxy-3-methylglutaryl coenzyme A reductase and cholesterol synthesis. The sterol content and as a consequence, the sterol/phospholipid molar ratio of C-6 glial cells were decreased by treating the cells, in 10% lipoprotein-poor serum, with various concentrations of compactin for 24 h. The degrees of sterol depletion thus produced were maintained for 48 h after removal of the compactin if the cells were maintained in serum-free medium, the culture conditions necessary for induction of CNP in untreated cells. Forty-eight hours after removal of serum, no induction of CNP occurred in cells previously treated with 0.5 μg/ml of compactin, whereas untreated cells exhibited a three- to fourfold increase in CNP activity. Intermediate degrees of sterol depletion resulted in intermediate degrees of inhibition of the CNP induction. Moreover, the morphological expressions of glial differentiation observed in the untreated cells did not occur in the sterol-depleted cells. That the effect of compactin on the induction of CNP relates to depletion of sterol was indicated by the finding that when low-density lipoprotein was added to the compactin-treated cells, the induction of CNP, the morphological expressions of differentiation and the sterol/phospholipid molar ratios were preserved. The degree of sterol depletion that totally prevented the induction of CNP had no effect on (Na⁺+ K⁺)-activated ATPase activity, total protein synthesis and cell viability. The data define a critical role for sterol in oligodendroglial differentiation in this model system.     

Microcephalic cerebrum with hypomyelination in the growth hormone-deficient mouse (lit)

To determine whether GH has an independent action on cerebral development, we examined the central nervous system of thelittle mouse (lit), a promissing model of isolated growth hormone deficiency. Our findings are (A); the weights of two parts of thelit brain were significantly less than those of the normal controls, 81.5% less for the cerebrum, and 81.6% for the cerebellum, (B): the total DNA content was reduced to approximately 80% in the cerebrum and 84% in the cerebellum compared to those of the normal controls, (C); the total RNA content was also reduced in the cerebrum and cerebellum, proportional to the reduction in DNA, (D); CNPase activity was reduced selectively in the cerebrum of thelit mouse (74.4% of the normal control), and (E); thelit mice exhibited a strikingly reduced level of activity with an indistinct diurnal periodicity. These results indicate that GH has independent actions on cerebral development, especially on glial cell proliferation as a precondition of myelin formation.Dedicated to Professor Yasuzo Tsukada.     

Immunohistochemical localization of 2′,3′-cyclic nucleotide 3′-phosphodiesterase in adult bovine cerebrum and cerebellum

We have previously shown that 2,3-cyclic nucleotide 3-phosphodiesterase (CNP; EC 3.1.4.37) in rat central nervous tissues can be immunohistochemically stained with anti-bovine CNP serum. However, the anti-bovine CNP serum prepared in our laboratory has only weak cross-reactivity with rat CNP. Sections of bovine nervous tissues were found to be stained effectively with the serum, and the localization of CNP has been revealed in greater detail. We describe here the immunohistochemical localization of CNP in adult bovine cerebrum and cerebellum. CNP stained was localized in myelin sheaths, oligodendrocytes, and the processes of oligodendrocytes; astrocytes and neurons were negative. All myelinated nerve fibers appeared to be stained with the anti-CNP serum. Perineuronal and perivascular oligodendrocytes, and oligodendrocytes extending their processes to isolated myelin fibers were stained. Interfascicular oligodendrocytes, however, did not react or reacted faintly to the anti-CNP serum; only their processes were reactive. Comparison with the stain for S-100 protein was helpful to distinguish oligodendrocytes from astrocytes particularly when both glial cells were situated together at the perineuronal and perivascular positions.Dedicated to Professor Yasuzo Tsukada.     

Studies on a Cerebellar 50,000-Dalton Protein Associated with Cerebellar Hypoplasia in Jaundiced Gunn Rats: Its Identity with Glial Fibrillary Acidic Protein as Evidenced by the Improved Immunoblotting Method

On the basis of our previous findings that a 50,000-dalton protein (GR-50) shows a marked increase in the hypoplastic cerebellum of jaundiced homozygous Gunn rats and its electrophoretic behavior is similar to that of glial fibrillary acidic protein (GFAP), we tried to identify GR-50 as GFAP by two-dimensional electrophoresis of rat cerebellar membrane proteins using an improved immunoblotting method. In this method a blot immunostained for a specific antigen was also visualized for other proteins, thereby enabling us to determine the location of the antigen on the blot more precisely. With the methodology it was found that GFAP antigen occupied exactly the same position as GR-50 on the blot, suggesting strongly the identity of both proteins. Immunohistochemical studies revealed that in the cerebellum of homozygotes compared with that of heterozygotes GFAP antigen was greatly increased and that it was especially rich in the homozygous rat cerebellar lobules with a high degree of hypoplasia. Further, it was shown that not only the fibers of the Bergmann glial cells but also their somata were intensely immunostained in the affected lobules. A 47,000-dalton protein (SG-47), which has been reported to be increased in staggerer mutant mice with cerebellar hypoplasia, also immunoreacted with the antiserum to GFAP.     

猫嗅球外丛层和僧帽细胞层年龄相关形态学变化

分别用Nissl法及免疫组织化学ABC法标记青、老年猫嗅球中嗅觉二级神经元和外丛层胶质细胞,显微镜下观察其分布并计数,对嗅觉二级神经元胞体直径和外丛层厚度进行测量,比较其年龄相关性变化,研究神经元与胶质细胞之间的关系,探讨老年性嗅觉功能衰退的相关神经机理。结果显示,老年猫嗅觉二级神经元胞体直径和分布密度均有不同程度的显著性下降(P<0.05);外丛层厚度变化不明显(P>0.05);外丛层胶质细胞特别是星形胶质细胞显著性增生(P<0.05)。表明在衰老过程中嗅觉二级神经元有丢失,并呈现功能下降,可能是老年性嗅觉功能衰退的原因之一。同时外丛层胶质细胞增生以进一步保护神经元,延缓其衰老。     

Immunocytochemical Characterization of Glutamate Dehydrogenase in the Cerebellum of the Rat

The immunocytochemical distribution of glutamate dehydrogenase was studied in the cerebellum of the rat using antibodies made in rabbit and guinea pig against antigen purified from bovine liver. Antiserum was found to block partially enzymatic activity both of the purified enzyme and of extracts of the rat cerebellum. Using immunoblots of proteins of rat cerebellum, a major immunoreactive protein and several minor immunoreactive proteins were detected with antiserum. Only a single immunoreactive protein was detected using affinity-purified antibody preparations. This protein migrates with a molecular weight identical to that of the subunit of glutamate dehydrogenase. Further evidence that the antibodies were selective for glutamate dehydrogenase in rat cerebellum was obtained through peptide mapping. Purified glutamate dehydrogenase and the immunoreactive protein from rat cerebellum generated similar patterns of immunoreactive peptides. No significant cross-reaction was observed with glutamine synthetase. Immunocytochemistry was done on cryostat- and Vibratome-cut sections of the cerebellum of rats that had been perfused with cold 4% paraformaldehyde. Glial cells were found to be the most immunoreactive structures throughout the cerebellum. Most apparent was the intense labeling of Bergmann glial cell bodies and fibers. In the granule cell layer, heavy labeling of astrocytes was seen. Purkinje and granule cell bodies were only lightly immunoreactive, whereas stellate, basket, and Golgi cells were unlabeled. Labeling of presynaptic terminals was not apparent. These findings suggest that glutamate dehydrogenase, like glutamine synthetase, is enriched in glia relative to neurons.