- Email: [email protected]
Differential expression of glutamate decarboxylase messenger RNA in cerebellar Purkinje cells and deep cerebellar nuclei of the genetically dystonic rat
Pergamon
PII:
Neuroscience Vol. 82, No. 4, pp. 1087–1094, 1998 Copyright ? 1997 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306–4522/98 $19.00+0.00 S0306-4522(97)00334-5
DIFFERENTIAL EXPRESSION OF GLUTAMATE DECARBOXYLASE MESSENGER RNA IN CEREBELLAR PURKINJE CELLS AND DEEP CEREBELLAR NUCLEI OF THE GENETICALLY DYSTONIC RAT L. NAUDON,* J. M. DELFS,* N. CLAVEL,* J. F. LORDEN,† M.-F. CHESSELET‡§ *Department of Pharmacology, University of Pennsylvania, Philadelphia, PA 19104, U.S.A. †Department of Psychology and Neurobiology Research Center, University of Alabama at Birmingham, Birmingham, AL 35294, U.S.A. Abstract––The genetically dystonic rat exhibits a motor syndrome that closely resembles the human disease, generalized idiopathic dystonia. Although in humans dystonia is often the result of pathology in the basal ganglia, previous studies have revealed electrophysiological abnormalities and alterations in glutamate decarboxylase, the synthetic enzyme for GABA, in the cerebellum of dystonic rats. In this study, we further characterized the alterations in cerebellar GABAergic transmission in these mutants by examining the expression of the messenger RNA encoding glutamate decarboxylase (67000 mol. wt) with in situ hybridization histochemistry at the single cell level in Purkinje cells and neurons of the deep cerebellar nuclei. Glutamate decarboxylase (67000 mol. wt) messenger RNA levels were increased in the Purkinje cells and decreased in the deep cerebellar nuclei of dystonic rats compared to control littermates, suggesting opposite changes in GABAergic transmission in Purkinje cells and in their target neurons in the deep cerebellar nuclei. In contrast, levels of glutamate decarboxylase (67000 mol. wt) messenger RNA in the pallidum, and of enkephalin messenger RNA in the striatum, were unaffected in dystonic rats. The data indicate that both the Purkinje cells and GABAergic neurons of the deep cerebellar nuclei are the site of significant functional abnormality in the dystonic rat. ? 1997 IBRO. Published by Elsevier Science Ltd. Key words: dystonia, GAD, genetic model, cerebellum, basal ganglia, rat.
The genetically dystonic (dt) rat exhibits a motor syndrome that closely resembles the human disease, generalized idiopathic dystonia.18 The rat disease follows an autosomal recessive pattern of inheritance. The dt rats are indistinguishable from phenotypically normal littermates on tests of motor function or spontaneous, open-field behaviour prior to postnatal days 9 to 10. Clinical signs first appear on postnatal day 10 and progress to include twisting of the neck and trunk, self-clasping of the paws, hyperflexion of the trunk, and frequent falls with sustained extension of the limbs. Motor function steadily deteriorates after the onset of the movement disorder. Even with supportive measures, dt rats die before reaching maturity. In humans, dystonias are generally considered disorders of the basal ganglia, because of frequent neuropathological findings in these regions in ‡To whom correspondence should be addressed. §Present address: M.-F. Chesselet, MD, PhD. Department of Neurology, UCLA School of Medicine, 710 Westwood Plaza, Los Angeles, CA 90095, 310-267-1781, U.S.A. 310-267-1786 (FAX) [email protected] Abbreviations: DCN, deep cerebellar nuclei; dt, genetically dystonic; GAD, glutamate decarboxylase; O.D., optical density; SSC, standard saline citrate.
patients with secondary dystonias.20 Similarly, in another model of dystonia, the genetically dystonic (dtsz) hamster, modifications of GABA concentration and dopamine receptor levels were found in the basal ganglia.19,21 However, in dt rats, previous studies have identified the cerebellum as a site of significant functional abnormality, whereas examination of the basal ganglia has been inconclusive.2,17,22,23,29 Electrophysiological studies have revealed a slower spontaneous firing in the Purkinje cells of dt rats than in control littermates, whereas the firing rates of neurons in the deep cerebellar nuclei (DCN) were increased in the dt rats.17,29 These alterations in electrophysiolgical activity are likely to result in alterations in GABAergic transmission in the cerebellum of the mutant rats, GABA being the neurotransmitter of the Purkinje cells and of a subset of DCN neurons7,10–13,30 Indeed, the activity of glutamate decarboxylase (GAD), the synthetic enzyme for GABA, is increased in the DCN of the dt rats compared to littermate controls.2,19,22 However, changes in GAD activity in the DCN are difficult to relate to changes in electrophysiological activity of various populations of neurons because GAD is not only present in axon terminals of the Purkinje cells but also in other axon terminals and in a
1087
1088
L. Naudon et al.
subpopulation of neurons of the DCN.10–13,30 Measurements of GAD activity cannot distinguish between effects occurring in these different neuronal populations. One way to circumvent this difficulty is to measure, instead of GAD, the mRNA encoding the enzyme, which is present in neuronal cell bodies, but is undetectable in axon terminals with routine in situ hybridization histochemistry.7 Therefore, changes in the level of expression of GAD mRNA can be measured in Purkinje cells and in GABAergic neurons of the DCN without interference from GABAergic terminals. In the present study, the levels of GAD (67000 mol. wt; GAD67) were measured both in the cerebellar Purkinje cells and in GABAergic neurons of the DCN with in situ hybridization histochemistry and emulsion autoradiography in dt rats and littermate controls. In addition, levels of GAD67 and enkephalin mRNA were examined in the globus pallidus and the striatum, respectively, to examine the possibility of a dysfunction of their expression in the basal ganglia of dt rats. EXPERIMENTAL PROCEDURES
Animals 22–25-day-old dystonic rats JFL:SD-dt and their littermates, as controls, were obtained from the colony maintained at the University of Alabama at Birmingham. The animals were killed by decapitation and the brains removed and frozen in isopentane cooled to "25)C with dry ice. All brains were maintained frozen ("70)C) until processed for determination of GAD67 and enkephalin mRNA levels. The animals were cared for in accordance with the National Institute of Health Guide for the Care and Use of Laboratory Animals, and the guidelines of the University of Alabama at Birmingham Institutional Animal Care and Use Committee. All efforts were made to minimize animal suffering, to use the smallest number of animals possible and to utilize alternatives to in vivo techniques. In situ hybridization histochemistry Coronal sections (10 µm) were cut at the level of the striatum, of the globus pallidus, of the entopeduncular nucleus and of the deep cerebellar nuclei (respectively, 0.70 mm, "0.80 mm, "2.8 mm and "11.60 mm from bregma).24 The levels of mRNA encoding GAD67 and enkephalin were determined by in situ hybridization with a probe specific for these mRNAs according to Chesselet et al.7 The GAD67 cDNA, which consists of a full-length 2.3 kilobase sequence isolated from a cat occipital cortex library and then inserted into the transcription vectors pSP64/65, was kindly provided by Dr A. J. Tobin (UCLA, Los Angeles CA). The cDNA for preproenkephalin, a 970 base pair sequence isolated from a rat striatal library and incorporated into pSP64/65, was kindly donated by Dr S. L. Sabol (National Institute of Mental Health, Bethesda, MD). The synthesis mixture consisted of 10 µM [S]UTP, 2.5 µM [35S]UTP (1000 Ci/mmol, NEN/DuPont, Boston, MA), ATP, CTP and GTP in excess, SP6 or T7 RNA polymerase, dithiothreitol, the ribonuclease inhibitor RNAsin and 2 µg of linearized DNA containing the insert. Because of its large size, the GAD67 RNA probe was partially hydrolysed into 100–200 base pair fragments by alkaline hydrolysis for increased tissue penetration.7 Following the synthesis, the probe was extracted in phenol/ chloroform/isoamyl alcohol and precipitated in ethanol at "70)C overnight.
Fig. 1. Camera lucida drawing of a frontal section through the deep cerebellar nuclei illustrating the distribution of neurons labelled for GAD67 mRNA following in situ hybridization; the black dots represent the cells labelled for GAD67. DCN, deep cerebellar nuclei; mn, medialis nucleus; in, interpositus nucleus; ln, lateralis nucleus; PCL, Purkinje cell layer. Arrows identify the Purkinje cell layer of the cerebellar cortex. Scale bar=0.5 mm. For in situ hybridization histochemistry,7 the sections were brought to room temperature under a stream of cold air, postfixed in 3% paraformaldehyde containing 0.02% diethylpyrocarbonate (Sigma Chemical Co., St Louis, MO), acetylated and dehydrated. Sections were incubated with 3–5 ng of radiolabelled probe (400000 d.p.m./ng) in humid chambers at 50)C for 3.5 h. Post-hybridization treatments included three washes in 50% formamide/2#standard saline citrate (2#SSC; 0.3 M NaCl/0.003 M sodium citrate) at 52)C and a 30 min incubation in 100 µg/ml RNase A (Sigma Chemical Co., St Louis, MO) at 37)C after the second wash. After a overnight rinse in 2#SSC/0.05% Triton X-100, sections were dehydrated in graded ethanol, defatted in xylene and desiccated. Then, sections for GAD67 mRNA analysis were coated with Kodak NTB3 emulsion (diluted with an equivolume of 300 mM ammonium acetate) for single-cell analysis. Film autoradiography (Kodak X-OMAT) was performed for enkephalin mRNA analysis because the high density of striatal cells expressing enkephalin mRNA allows for a strong autoradiographic signal on films. Test slides and test films were developed at regular intervals to determine optimal exposure time, i.e. when specific labelling was robust but not saturating (two to three weeks). Emulsion and film autoradiograms were developed in Kodak D-19 developer and fixed in Kodak fixer. Emulsion-coated sections were lightly counterstained with Haematoxylin and Eosin and coverslipped with Eukitt mounting media (Calibrated Instrument, Hawthorne, NY). The anatomical specificity of the autoradiographic labelling has been previously characterized7,25 and was verified in each individual experiment. No specific labelling has been observed in control sections processed in a similar manner with sense RNA probes.7,25 Quantification In emulsion-coated sections, the level of labelling over individual neurons in the DCN, in Purkinje cells of the cerebellar cortex, and in the pallidum was measured with the Morphon Image Analysis System.6 This system consists of a MTI 65 video camera, a Leitz microscope, a Numonics Graphic-Master image digitizer and an IBM 212 computer. Only sections processed concurrently in the same experiment were compared quantitatively. One or two slides per rat were used for quantification in each experiment. The anatomical levels of sections examined in each group (dt
GAD in cerebellum of dystonic rats
1089
Fig. 2. Dark-field photomicrograph of cells in the cerebellar cortex labelled for GAD67 mRNA following in situ hybridization. Arrows identify individual Purkinje cells of the Purkinje cell layer. Brightness differences between littermates control (A) and dt rats (B) in this layer reflect the increased level of GAD67 mRNA found in dt rats. Scale bar=40 µm. rats and littermate controls) were carefully matched based on anatomical landmarks recognizable in the Haematoxylin and Eosin stain.
The structure of interest was outlined under low-power microscopy. Neurons were visualized under bright-field illumination with a 40# objective and projected onto a
1090
L. Naudon et al.
video monitor with a resulting magnification of 1200#. Each neuron of interest was outlined with the digitizing tablet and the threshold grey value was adjusted to distinguish grains from the cellular staining produced by the Haematoxylin and Eosin. The portion of the enclosed area which was above threshold illumination was determined and expressed as the number of pixels occupied by silver grains. Linear analysis has shown a direct correlation (+0.96 for 40# magnification) between the number of pixels measured with the Morphon system and visual grain counting.6 For the present experiments, specific labelling was arbitrarily defined as the presence of 10 or more grains per neuron (approximately 10 pixels). Autoradiographic background was very low in the experiments analysed, thus background labelling was not subtracted. The number of labelled cells was determined from camera lucida drawings performed under dark-field illumination. In the cerebellum, levels of labelling were only determined in Purkinje cells and in neurons of the DCN in this study. The cerebellar cortex contains other GABAergic neurons,7,11,12 but the level of labelling being different for each neuronal population, different exposure times, and therefore, separate experiments, would be necessary to examine all neuronal populations. In the case of the DCN, all the labelled cells were analysed on both sides (average number of labeled cells in the DCN of normal littermates, 69.4&9.6 (S.E.M.) cells per section, n=8). For the Purkinje cells and the pallidum, a random sample of 50 neurons per section, on both sides, was measured in each region analysed. This was accomplished by moving non-overlapping frames over the region, always starting at the same anatomical location, and measuring labelling in all neurons with at least 10 grains over the cell body. We have previously verified that this sample size provides a reliable estimate of the average level of labeling (J.-J. Soghomonian and M.-F. Chesselet, unpublished observations). A mean level of labelling per neuron was determined for each rat, and this value was used for statistical analysis. Additionally, frequency distributions of the level of labelling in the cerebellum were constructed for each group of rats. For striatal sections, in situ hybridization signal was measured on film autoradiograms with a Macintosh-based IMAGE analysis system (version 1.44, NIH, Bethesda, MD). Grey levels on the film were converted into optical density (O.D., arbitrary units) by constructing a curve based on Kodak optical density (O.D.) standards. Optical density measurements were taken for the left and right striatum, and the results from one to two sections per rat were averaged to determine an O.D. for each rat. Data analysis Statistical analyses were performed on absolute values using the Statview 512+ Interactive Statistics and Graphics Package (version 1.0, Abacus Concepts). Means were compared with unpaired two-tailed Student’s t-tests and medians were compaired with a two-tailed Mann–Whitney test, with P<0.05 considered significant. RESULTS
GAD67 messenger RNA expression in the cerebellum of dystonic rats As previously observed,7 in situ hybridization histochemistry with the radiolabelled RNA probe for GAD67 mRNA resulted in strong labeling in the Purkinje cell layer (Figs 1, 2A). Neurons labelled for GAD67 mRNA were present in all subdivisions of the DCN (Figs 1, 3A). The nucleus medialis, however, had a lower density than the interpositus
Table 1. Mean levels of labelling and medians of frequency distributions for GAD67 messenger RNA in Purkinje cells and neurons of the deep cerebellar nuclei Purkinje cells A) GAD67 mRNA levels Normal littermates 78.5&3.8 Dystonic rats 103.6&3.8* B) Medians Normal littermates 73.1&3.4 Dystonic rats 92.7&5.4*
DCN neurons 63.7&2 45.3&1* 58.9&2.1 42.6&0.9*
A. Mean levels of labelling for GAD67 mRNA were determined for each animal by averaging the number of pixels covered by grains over individual cells (pixels/cell). B. Median of frequency distribution of labelling for GAD67 mRNA in Purkinje cells and DCN. Frequency distributions of levels of labelling were constructed for individual animals and their medians were determined. Data are means&S.E.M. of values obtained in seven normal littermates and eight dystonic rats. *P<0.05, when compared to normal littermates with a two-tailed Student’s t-test (means) or Mann–Whitney U-test (medians).
nucleus and the lateralis nucleus, as noted in previous immunohistochemical studies.1,13 Taking into account (i) the difficulty of distinguishing the individual nuclei of the DCN during the microscopic analysis, and (ii) the small number of GAD67 mRNA labelled neurons per nuclei, the totality of the neurons in the DCN was analysed in each section. Quantification of the autoradiographic signal over individual neurons in the Purkinje cell layer revealed a significant increase in the level of labelling for GAD67 mRNA in dt rats compared to wild-type littermates (+32%, Table 1A, Fig. 2). Analysis of the frequency distributions of the level of labelling per neuron confirmed the increase in level of labelling by showing a shift to the right of the histogram (Fig. 4B), and an increase in the medians of the distribution for dt rats (Table 1B), compared to the wild-type littermates (Fig. 4A, Table 1B). In contrast, the level of labelling for GAD67 mRNA was significantly lower in the DCN of dt rats compared to littermate controls ("29%, Table 1A, Fig. 3), without modification of the number of labelled cells (wild-type littermates, 69.4&9.6 [n=8]; dt rats, 65.6&3.8 [n=7], mean number of labelled neurons in the DCN per section&S.E.M.). Confirming the decrease in labelling, the frequency distribution of labelling per neuron was shifted to the left and the medians significantly decreased (Fig. 4D, Table 1B). Expression of enkephalin and GAD67 messenger RNA in the basal ganglia of dt rats As previously shown using a similar technique, a high level of expression of enkephalin mRNA was observed in the striatum.25 Similarly, an intense signal for GAD67 mRNA was present in a majority
GAD in cerebellum of dystonic rats
1091
Fig. 3. Photomicrographs of neurons labelled for GAD67 mRNA in the deep cerebellar nuclei following in situ hybridization. Dark-field photomicrograph (A, B); arrows identify individual neurons. Brightness differences between littermates control (A) and dt rats (B) in this layer reflect the decreased level of GAD67 mRNA found in dt rats. Scale bar=40 µm. High power bright-field photomicrograph in control littermates (C); labelling was observed over a neuron (arrow) but not over glial cells (arrowheads). Note that only the Haematoxylin-stained nuclei are visible. The pale Eosin-stained cytoplasm cannot be distinguished on this black and white photomicrograph. Scale bar=10 µm
of neurons of the globus pallidus and of the entopeduncular nucleus.7 The levels of labelling for these mRNAs were examined in the dt rats because previous studies have revealed changes in expression of these mRNAs in animal models of movement disorders.5 However, no significant differences in the levels of labelling for GAD67 mRNA in the globus pallidus and entopeduncular nucleus or enkephalin mRNA in the striatum were found in dt rats compared to wild-type littermates (Table 2).
DISCUSSION
This study revealed changes in the level of expression of GAD67 mRNA both in Purkinje cells and in GABAergic neurons of the DCN, confirming that the cerebellum is a major site of abnormalities in dt rats. In contrast, no changes were found in the levels of enkephalin mRNA in the striatum or GAD67 mRNA in the pallidum. This argues against major abnormalities in the basal ganglia of dt rats, although
1092
L. Naudon et al.
Fig. 4. Histograms of frequency distributions of labelling for GAD67 in the Purkinje cells and the DCN of normal littermates control rats (A, C) or dystonic (dt) rats (B, D). Quantification of silver grains over individual neurons was done with the Morphon Image Analysis System on sections processed for emulsion autoradiography. Data includes values from 50 Purkinje cells analysed or the totality of the labelled neurons of the DCN in two sections on both sides per rat for seven to eight rats. Note the shift to the right in the histogram in (B) and to the left in (D), illustrating, respectively, an increase in the level of labelling in Purkinje cells of dt rats, and a decrease in the level of labelling in the neurons of the DCN of dt rats. The medians of the frequency distribution (arrows) are presented in Table 1B. Table 2. Enkephalin and GAD67 messenger RNA expression in basal ganglia mRNA Enkephalin (striatum)a GAD67 (globus pallidus)b GAD67 (entopeduncular nucleus)b
Normal littermates
Dystonic rats
0.036&0.004 85.96&3.01 23.90&3.14
0.041&0.007 81.61&2.12 24.14&3.34
a
Optical density measured from autoradiograms (see Experimental Procedures). Data are means&S.E.M. values for 10 normal littermates and nine dystonic rats. P=0.51, two-tailed Student’s t-test. b Level of labelling (pixels per neuron) as measured by single-cell analysis (see Experimental Procedures). Data are means&S.E.M. values for eight normal littermates and 10 dystonic rats (striatum and globus pallidus) or four normal and five dystonic rats (entopeduncular nucleus). NS, two-tailed Student’s t-test.
the functional integrity of other neuronal circuits in this region remains to be tested. Glutamate decarboxylase messenger RNA in rat cerebellum In dt rats, the levels of GAD67 mRNA were modified compared to their littermate controls in two interconnected neuronal populations of the cerebellum, the Purkinje cells and neurons of the DCN. The
Purkinje cells form the only output pathway from the cerebellar cortex, therefore alterations in GABAergic transmission in these cells is likely to have major functional consequences. It cannot be excluded, however, that other GABAergic neruons in the cerebellar cortex7,11,12 are also affected in the mutant rats. An exhaustive study of all neuronal types in the cerebellum in dt rats was beyond the scope of this work, which focused on alterations in the main output pathway of the cerebellum for which electrophysiological studies had already provided evidence of alteration in the dt rat.17,29 Similarly, we have concentrated on the expression of the mRNA encoding GAD67, because, though mRNA encoding GAD (65000 mol. wt) is also expressed in the cerebellum,11,12 previous studies, both in the cerebellum and in other brain regions, have shown that the levels of mRNA encoding GAD67 vary in relation to the electrophysiological activity of GABAergic neurons,3,5,16,26–28 whereas alterations in GAD65 mRNA levels are less commonly observed, and usually are of a smaller amplitude in the rat.9,27,28 Although the bulk of GAD activity measured biochemically in the DCN is usually considered to originate from Purkinje cell afferents, immunohistochemical studies have revealed that the DCN also contains a population of GABAergic neurons.10,13 In situ hybridization histochemistry experiments with cellular resolution have confirmed the presence of GAD mRNA in neurons of the DCN. (Refs 11,12 and present report) The situation, however, is more complex in the DCN than in the cerebellar cortex, because, in contrast to the Purkinje cells which are homogeneously GABAergic, neurons of the DCN are heterogeneous with regard to their neurotransmitter content. In addition to small GABAergic neurons, the DCN contains small neurons containing glutamate1 and large neurons using aspartate and/or glutamate.1,10,13 A relatively high proportion of small neurons contains both GABA and glycine (50–70%), and the remainders are either GABAergic or glycine-containing neurons.4 Both GABAergic neurons and large neurons containing other neurotransmitters receive afferents from the Purkinje cells.10,13 Glutamate decarboxylase messenger RNA levels and firing patterns of cerebellar neurons in the dt rat Previous findings of an increase in GAD activity coupled to a decease in the density of GABAA receptors in the DCN suggest an increase in GABA release in the DCN resulting in a compensatory down-regulation of GABA receptors.2 The present data support the hypothesis that GABAergic transmission is increased in Purkinje cells, which express higher levels of GAD67 mRNA, and decreased in GABAergic neurons of the DCN, which express lower levels of GAD67 mRNA in dt rats than in control littermates. Such changes in GABAergic
GAD in cerebellum of dystonic rats
transmission, however, are difficult to explain based on the changes in firing rates, slower in Purkinje cells and increased in DCN neurons, which have been reported in dt rats.17,29 In contrast, changes in GAD67 mRNA could be better explained by modifications of the firing patterns reported in the mutant rats. GAD67 mRNA levels were increased in neurons showing a more irregular pattern of firing (i.e. Purkinje cells29) and decreased in a region where the most consistent finding was an increase in the regularity of firing (i.e. neurons of the DCN17). A similar relationship between firing pattern and GAD67 mRNA has been reported in the globus pallidus.5,8 Recent work has provided increasing evidence that patterns rather than rates of spontaneous firing may be critical for neurotransmitter release and synaptic efficiency in a variety of neuronal systems.15 The present data support this hypothesis by showing a correlation between levels of the mRNA encoding the
1093
enzyme of GABA synthesis and pattern of activity in cerebellar GABAergic neurons. CONCLUSION
In conclusion, the results of this study confirm a complex alteration in the functioning of the cerebellar output in the dt rat, which is consistent with abnormal electrophysiological activity in target regions from the DCN.14,17 Furthermore, this study indicates that the GABAergic subpopulation of DCN neurons may contribute significantly to the generation of this abnormal output. Acknowledgements—We are grateful to Drs A. Tobin (UCLA) and S. Sabol (NIH) for the gift of cDNAs. This work was supported by PHS grant R37 MH-44894 (M.F.C.) and the Dystonia Medical Research Foundation (J.F.L.)
REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.
Batini C., Compoint C., Buisseret-Delmas C., Daniel H. and Guegan M. (1992) Cerebellar nuclei and the nucleocortical projections in the rat: retrograde tracing coupled to GABA and glutamate immunohistochemistry. J. comp. Neurol. 315, 74–84. Beales M., Lorden J. F., Walz E. and Oltmans G. A. (1990) Quantitative autoradiography reveals selective changes in cerebellar GABA receptors of the rat mutant dystonic. J. Neurosci. 10, 1874–1885. Benson D. L., Huntsman M. M. and Jones E. G. (1994) Activity-dependent changes in GAD and preprotachykinin mRNAs in visual cortex of adult monkeys. Cerebr. Cortex 4, 40–51. Chen S. and Hillman D. E. (1993) Colocalization of neurotransmitters in the deep cerebellar nuclei. J. Neurocytol. 22, 81–91. Chesselet M.-F. and Delfs J. M. (1996) Basal ganglia and movement disorders: an update. Trends Neurosci. 19, 417–422. Chesselet M.-F. and Weiss-Wunder L. T. (1994) Quantification of in situ hybridization histochemistry. In In situ Hybridization in Neurobiology, Advances in Methodology (eds Eberwine J. H., Valentino K. L. and Barchas J. D.), pp. 114–123. Oxford University Press, New York. Chesselet M.-F., Weiss L., Wuenschell C., Tobin A. J. and Affolter H. U. (1987) Comparative distribution of mRNAs for glutamic acid decarboxylase, tyrosine hydroxylase, and tachykinins in the basal ganglia: an in situ hybridization study in the rodent brain. J. comp. Neurol. 262, 125–140. Delfs J. M., Ciaramitaro V. M., Parry T. J. and Chesselet M.-F. (1995) Subthalamic nucleus lesions: widespread effects on changes in gene expression induced by nigrostriatal dopamine depletion in rats. J. Neurosci. 15, 6562–6575. Delfs J. M., Ciaramitaro V. M., Soghomonian J.-J. and Chesselet M.-F. (1996) Unilateral nigrostriatal lesions induce a bilateral increase in glutamate decarboxylase messenger RNA in the reticular thalamic nucleus. Neuroscience 71, 383–395. De Zeeuw C. I. and Barrebi A. S. (1995) Postsynaptic targets of Purkinje cell terminals in the cerebellar and vestibular nuclei of the rat. Eur. J. Neurosci. 7, 2322–2333. Esclapez M., Tillakaratne N. J., Kaufman D. L., Tobin A. J. and Houser C. R. (1994) Comparative localization of two forms of glutamic acid decarboxylase and their mRNAs in rat brain supports the concept of functional differences between the forms. J. Neurosci. 14, 1834–1855. Feldblum S., Erlander M. G. and Tobin A. J. (1993) Different distributions of GAD65 and GAD67 mRNAs suggest that the two glutamate decarboxylases play distinctive functional roles. J. Neurosci. Res. 34, 689–706. Kumoi K., Saito N., Kuno T. and Tanaka C. (1988) Immunohistochemical localization of ã-aminobutyric acid- and aspartate-containing neurons in the rat deep cerebellar nuclei. Brain Res. 439, 302–310. LeDoux M. S., Lorden J. F. and Meinzen-Derr J. (1995) Selective elimination of cerebellar output in the genetically dystonic rat. Brain Res. 697, 91–103. Lisman J. E. (1997) Burst as a unit of neural information: making unreliable synpases reliable. Trends Neurosci. 20, 38–43. Litwak J., Mercugliano M., Chesselet M.-F. and Oltmans G. A. (1990) Increased glutamic acid decarboxylase (GAD) mRNA and GAD activity in cerebellar Purkinje cells following lesion-induced increases in cell firing. Neurosci. Lett. 116, 179–183. Lorden J. F., Lutes J., Michela V. L. and Ervin J. (1992) Abnormal cerebellar output in rats with an inherited movement disorder. Expl Neurol. 118, 95–104. Lorden J. F., McKeon T. W., Baker H. J., Cox N. and Walkley S. U. (1984) Characterization of the rat mutant dystonic (dt): a new animal model of dystonia musculorum deformans. J. Neurosci. 4, 1925–1932. Loscher W. and Horstermann D. (1992) Abnormalities in amino acid neurotransmitters in discrete brain regions of genetically dystonic hamsters. J. Neurochem. 59, 689–694. Marsden C. D. (1992) Motor function and movement disorders. In Advances in Neurology: Dystonia 2 (eds Asbury A. K., McKhan G. M. and McDonald W. I.), pp. 309–318. W. B. Saunders Co, Philadelphia.
1094 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.
L. Naudon et al. Nobrega J. N., Richter A., McIntyre Burnham W. and Loscher W. (1995) Alterations in the brain GABAA/ benzodiazepine receptor-chloride ionophore complex in a genetic model of paroxysmal dystonia: a quantitative autoradiographic analysis. Neuroscience 64, 229–239. Oltmans G. A., Beales M. and Lorden J. F. (1986) Glutamic acid decarboxylase activity in micropunches of the deep cerebellar nuclei of the genetically dystonic (dt) rat. Brain Res. 385, 148–151. Oltmans G. A., Beales M., Lorden J. F. and Gordon J. H. (1984) Alterations in cerebellar glutamic acid decarboxylase (GAD) activity in a genetic model of torsion dystonia (rat). Expl Neurol. 85, 216–222. Paxinos G. and Watson C. (1986) In The Rat Brain in Stereotaxic Coordinates. Academic Press, San Diego. Salin P. and Chesselet M.-F. (1992) Paradoxical increase in striatal neuropeptide gene expression following ischemic lesions of the cerebral cortex. Proc. natn. Acad. Sci. U.S.A. 89, 9954–9958. Segovia J., Tillakaratne N. J. K., Whelan K., Tobin A. J. and Gale K. (1990) Parallel increases in striatal glutamic acid decarboxylase activity and mRNA levels in rats with lesion of the nigrostriatal pathway. Brain Res. 529, 345–348. Soghomonian J. J. and Chesselet M.-F. (1992) Effects of nigrostriatal lesions on the levels of messenger RNAs encoding two isoforms of glutamate decarboxylase in the globus pallidus and entopeduncular nucleus of the rat. Synapse 11, 124–133. Soghomonian J. J., Gonzales C. and Chesselet M.-F. (1992) Messenger RNAs encoding glutamate-decarboxylases are differentially affected by nigrostriatal lesions in subpopulations of striatal neurons. Brain Res. 576, 68–79. Stratton S. E., Lorden J. F., Mays L. E. and Oltmans G. A. (1988) Spontaneous and harmaline-stimulated Purkinje cell activity in rats with a genetic movement disorder. J. Neurosci. 8, 3327–3336. Wassef M., Simons J., Tappaz M. L. and Sotelo C. (1986) Non-Purkinje cell GABAergic innervation of the deep cerebellar nuclei: a quantitative immunocytochemical study in C57BL and in Purkinje cell degeneration mutant mice. Brain Res. 399, 125–135. (Accepted 13 June 1997)
PII:
Neuroscience Vol. 82, No. 4, pp. 1087–1094, 1998 Copyright ? 1997 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306–4522/98 $19.00+0.00 S0306-4522(97)00334-5
DIFFERENTIAL EXPRESSION OF GLUTAMATE DECARBOXYLASE MESSENGER RNA IN CEREBELLAR PURKINJE CELLS AND DEEP CEREBELLAR NUCLEI OF THE GENETICALLY DYSTONIC RAT L. NAUDON,* J. M. DELFS,* N. CLAVEL,* J. F. LORDEN,† M.-F. CHESSELET‡§ *Department of Pharmacology, University of Pennsylvania, Philadelphia, PA 19104, U.S.A. †Department of Psychology and Neurobiology Research Center, University of Alabama at Birmingham, Birmingham, AL 35294, U.S.A. Abstract––The genetically dystonic rat exhibits a motor syndrome that closely resembles the human disease, generalized idiopathic dystonia. Although in humans dystonia is often the result of pathology in the basal ganglia, previous studies have revealed electrophysiological abnormalities and alterations in glutamate decarboxylase, the synthetic enzyme for GABA, in the cerebellum of dystonic rats. In this study, we further characterized the alterations in cerebellar GABAergic transmission in these mutants by examining the expression of the messenger RNA encoding glutamate decarboxylase (67000 mol. wt) with in situ hybridization histochemistry at the single cell level in Purkinje cells and neurons of the deep cerebellar nuclei. Glutamate decarboxylase (67000 mol. wt) messenger RNA levels were increased in the Purkinje cells and decreased in the deep cerebellar nuclei of dystonic rats compared to control littermates, suggesting opposite changes in GABAergic transmission in Purkinje cells and in their target neurons in the deep cerebellar nuclei. In contrast, levels of glutamate decarboxylase (67000 mol. wt) messenger RNA in the pallidum, and of enkephalin messenger RNA in the striatum, were unaffected in dystonic rats. The data indicate that both the Purkinje cells and GABAergic neurons of the deep cerebellar nuclei are the site of significant functional abnormality in the dystonic rat. ? 1997 IBRO. Published by Elsevier Science Ltd. Key words: dystonia, GAD, genetic model, cerebellum, basal ganglia, rat.
The genetically dystonic (dt) rat exhibits a motor syndrome that closely resembles the human disease, generalized idiopathic dystonia.18 The rat disease follows an autosomal recessive pattern of inheritance. The dt rats are indistinguishable from phenotypically normal littermates on tests of motor function or spontaneous, open-field behaviour prior to postnatal days 9 to 10. Clinical signs first appear on postnatal day 10 and progress to include twisting of the neck and trunk, self-clasping of the paws, hyperflexion of the trunk, and frequent falls with sustained extension of the limbs. Motor function steadily deteriorates after the onset of the movement disorder. Even with supportive measures, dt rats die before reaching maturity. In humans, dystonias are generally considered disorders of the basal ganglia, because of frequent neuropathological findings in these regions in ‡To whom correspondence should be addressed. §Present address: M.-F. Chesselet, MD, PhD. Department of Neurology, UCLA School of Medicine, 710 Westwood Plaza, Los Angeles, CA 90095, 310-267-1781, U.S.A. 310-267-1786 (FAX) [email protected] Abbreviations: DCN, deep cerebellar nuclei; dt, genetically dystonic; GAD, glutamate decarboxylase; O.D., optical density; SSC, standard saline citrate.
patients with secondary dystonias.20 Similarly, in another model of dystonia, the genetically dystonic (dtsz) hamster, modifications of GABA concentration and dopamine receptor levels were found in the basal ganglia.19,21 However, in dt rats, previous studies have identified the cerebellum as a site of significant functional abnormality, whereas examination of the basal ganglia has been inconclusive.2,17,22,23,29 Electrophysiological studies have revealed a slower spontaneous firing in the Purkinje cells of dt rats than in control littermates, whereas the firing rates of neurons in the deep cerebellar nuclei (DCN) were increased in the dt rats.17,29 These alterations in electrophysiolgical activity are likely to result in alterations in GABAergic transmission in the cerebellum of the mutant rats, GABA being the neurotransmitter of the Purkinje cells and of a subset of DCN neurons7,10–13,30 Indeed, the activity of glutamate decarboxylase (GAD), the synthetic enzyme for GABA, is increased in the DCN of the dt rats compared to littermate controls.2,19,22 However, changes in GAD activity in the DCN are difficult to relate to changes in electrophysiological activity of various populations of neurons because GAD is not only present in axon terminals of the Purkinje cells but also in other axon terminals and in a
1087
1088
L. Naudon et al.
subpopulation of neurons of the DCN.10–13,30 Measurements of GAD activity cannot distinguish between effects occurring in these different neuronal populations. One way to circumvent this difficulty is to measure, instead of GAD, the mRNA encoding the enzyme, which is present in neuronal cell bodies, but is undetectable in axon terminals with routine in situ hybridization histochemistry.7 Therefore, changes in the level of expression of GAD mRNA can be measured in Purkinje cells and in GABAergic neurons of the DCN without interference from GABAergic terminals. In the present study, the levels of GAD (67000 mol. wt; GAD67) were measured both in the cerebellar Purkinje cells and in GABAergic neurons of the DCN with in situ hybridization histochemistry and emulsion autoradiography in dt rats and littermate controls. In addition, levels of GAD67 and enkephalin mRNA were examined in the globus pallidus and the striatum, respectively, to examine the possibility of a dysfunction of their expression in the basal ganglia of dt rats. EXPERIMENTAL PROCEDURES
Animals 22–25-day-old dystonic rats JFL:SD-dt and their littermates, as controls, were obtained from the colony maintained at the University of Alabama at Birmingham. The animals were killed by decapitation and the brains removed and frozen in isopentane cooled to "25)C with dry ice. All brains were maintained frozen ("70)C) until processed for determination of GAD67 and enkephalin mRNA levels. The animals were cared for in accordance with the National Institute of Health Guide for the Care and Use of Laboratory Animals, and the guidelines of the University of Alabama at Birmingham Institutional Animal Care and Use Committee. All efforts were made to minimize animal suffering, to use the smallest number of animals possible and to utilize alternatives to in vivo techniques. In situ hybridization histochemistry Coronal sections (10 µm) were cut at the level of the striatum, of the globus pallidus, of the entopeduncular nucleus and of the deep cerebellar nuclei (respectively, 0.70 mm, "0.80 mm, "2.8 mm and "11.60 mm from bregma).24 The levels of mRNA encoding GAD67 and enkephalin were determined by in situ hybridization with a probe specific for these mRNAs according to Chesselet et al.7 The GAD67 cDNA, which consists of a full-length 2.3 kilobase sequence isolated from a cat occipital cortex library and then inserted into the transcription vectors pSP64/65, was kindly provided by Dr A. J. Tobin (UCLA, Los Angeles CA). The cDNA for preproenkephalin, a 970 base pair sequence isolated from a rat striatal library and incorporated into pSP64/65, was kindly donated by Dr S. L. Sabol (National Institute of Mental Health, Bethesda, MD). The synthesis mixture consisted of 10 µM [S]UTP, 2.5 µM [35S]UTP (1000 Ci/mmol, NEN/DuPont, Boston, MA), ATP, CTP and GTP in excess, SP6 or T7 RNA polymerase, dithiothreitol, the ribonuclease inhibitor RNAsin and 2 µg of linearized DNA containing the insert. Because of its large size, the GAD67 RNA probe was partially hydrolysed into 100–200 base pair fragments by alkaline hydrolysis for increased tissue penetration.7 Following the synthesis, the probe was extracted in phenol/ chloroform/isoamyl alcohol and precipitated in ethanol at "70)C overnight.
Fig. 1. Camera lucida drawing of a frontal section through the deep cerebellar nuclei illustrating the distribution of neurons labelled for GAD67 mRNA following in situ hybridization; the black dots represent the cells labelled for GAD67. DCN, deep cerebellar nuclei; mn, medialis nucleus; in, interpositus nucleus; ln, lateralis nucleus; PCL, Purkinje cell layer. Arrows identify the Purkinje cell layer of the cerebellar cortex. Scale bar=0.5 mm. For in situ hybridization histochemistry,7 the sections were brought to room temperature under a stream of cold air, postfixed in 3% paraformaldehyde containing 0.02% diethylpyrocarbonate (Sigma Chemical Co., St Louis, MO), acetylated and dehydrated. Sections were incubated with 3–5 ng of radiolabelled probe (400000 d.p.m./ng) in humid chambers at 50)C for 3.5 h. Post-hybridization treatments included three washes in 50% formamide/2#standard saline citrate (2#SSC; 0.3 M NaCl/0.003 M sodium citrate) at 52)C and a 30 min incubation in 100 µg/ml RNase A (Sigma Chemical Co., St Louis, MO) at 37)C after the second wash. After a overnight rinse in 2#SSC/0.05% Triton X-100, sections were dehydrated in graded ethanol, defatted in xylene and desiccated. Then, sections for GAD67 mRNA analysis were coated with Kodak NTB3 emulsion (diluted with an equivolume of 300 mM ammonium acetate) for single-cell analysis. Film autoradiography (Kodak X-OMAT) was performed for enkephalin mRNA analysis because the high density of striatal cells expressing enkephalin mRNA allows for a strong autoradiographic signal on films. Test slides and test films were developed at regular intervals to determine optimal exposure time, i.e. when specific labelling was robust but not saturating (two to three weeks). Emulsion and film autoradiograms were developed in Kodak D-19 developer and fixed in Kodak fixer. Emulsion-coated sections were lightly counterstained with Haematoxylin and Eosin and coverslipped with Eukitt mounting media (Calibrated Instrument, Hawthorne, NY). The anatomical specificity of the autoradiographic labelling has been previously characterized7,25 and was verified in each individual experiment. No specific labelling has been observed in control sections processed in a similar manner with sense RNA probes.7,25 Quantification In emulsion-coated sections, the level of labelling over individual neurons in the DCN, in Purkinje cells of the cerebellar cortex, and in the pallidum was measured with the Morphon Image Analysis System.6 This system consists of a MTI 65 video camera, a Leitz microscope, a Numonics Graphic-Master image digitizer and an IBM 212 computer. Only sections processed concurrently in the same experiment were compared quantitatively. One or two slides per rat were used for quantification in each experiment. The anatomical levels of sections examined in each group (dt
GAD in cerebellum of dystonic rats
1089
Fig. 2. Dark-field photomicrograph of cells in the cerebellar cortex labelled for GAD67 mRNA following in situ hybridization. Arrows identify individual Purkinje cells of the Purkinje cell layer. Brightness differences between littermates control (A) and dt rats (B) in this layer reflect the increased level of GAD67 mRNA found in dt rats. Scale bar=40 µm. rats and littermate controls) were carefully matched based on anatomical landmarks recognizable in the Haematoxylin and Eosin stain.
The structure of interest was outlined under low-power microscopy. Neurons were visualized under bright-field illumination with a 40# objective and projected onto a
1090
L. Naudon et al.
video monitor with a resulting magnification of 1200#. Each neuron of interest was outlined with the digitizing tablet and the threshold grey value was adjusted to distinguish grains from the cellular staining produced by the Haematoxylin and Eosin. The portion of the enclosed area which was above threshold illumination was determined and expressed as the number of pixels occupied by silver grains. Linear analysis has shown a direct correlation (+0.96 for 40# magnification) between the number of pixels measured with the Morphon system and visual grain counting.6 For the present experiments, specific labelling was arbitrarily defined as the presence of 10 or more grains per neuron (approximately 10 pixels). Autoradiographic background was very low in the experiments analysed, thus background labelling was not subtracted. The number of labelled cells was determined from camera lucida drawings performed under dark-field illumination. In the cerebellum, levels of labelling were only determined in Purkinje cells and in neurons of the DCN in this study. The cerebellar cortex contains other GABAergic neurons,7,11,12 but the level of labelling being different for each neuronal population, different exposure times, and therefore, separate experiments, would be necessary to examine all neuronal populations. In the case of the DCN, all the labelled cells were analysed on both sides (average number of labeled cells in the DCN of normal littermates, 69.4&9.6 (S.E.M.) cells per section, n=8). For the Purkinje cells and the pallidum, a random sample of 50 neurons per section, on both sides, was measured in each region analysed. This was accomplished by moving non-overlapping frames over the region, always starting at the same anatomical location, and measuring labelling in all neurons with at least 10 grains over the cell body. We have previously verified that this sample size provides a reliable estimate of the average level of labeling (J.-J. Soghomonian and M.-F. Chesselet, unpublished observations). A mean level of labelling per neuron was determined for each rat, and this value was used for statistical analysis. Additionally, frequency distributions of the level of labelling in the cerebellum were constructed for each group of rats. For striatal sections, in situ hybridization signal was measured on film autoradiograms with a Macintosh-based IMAGE analysis system (version 1.44, NIH, Bethesda, MD). Grey levels on the film were converted into optical density (O.D., arbitrary units) by constructing a curve based on Kodak optical density (O.D.) standards. Optical density measurements were taken for the left and right striatum, and the results from one to two sections per rat were averaged to determine an O.D. for each rat. Data analysis Statistical analyses were performed on absolute values using the Statview 512+ Interactive Statistics and Graphics Package (version 1.0, Abacus Concepts). Means were compared with unpaired two-tailed Student’s t-tests and medians were compaired with a two-tailed Mann–Whitney test, with P<0.05 considered significant. RESULTS
GAD67 messenger RNA expression in the cerebellum of dystonic rats As previously observed,7 in situ hybridization histochemistry with the radiolabelled RNA probe for GAD67 mRNA resulted in strong labeling in the Purkinje cell layer (Figs 1, 2A). Neurons labelled for GAD67 mRNA were present in all subdivisions of the DCN (Figs 1, 3A). The nucleus medialis, however, had a lower density than the interpositus
Table 1. Mean levels of labelling and medians of frequency distributions for GAD67 messenger RNA in Purkinje cells and neurons of the deep cerebellar nuclei Purkinje cells A) GAD67 mRNA levels Normal littermates 78.5&3.8 Dystonic rats 103.6&3.8* B) Medians Normal littermates 73.1&3.4 Dystonic rats 92.7&5.4*
DCN neurons 63.7&2 45.3&1* 58.9&2.1 42.6&0.9*
A. Mean levels of labelling for GAD67 mRNA were determined for each animal by averaging the number of pixels covered by grains over individual cells (pixels/cell). B. Median of frequency distribution of labelling for GAD67 mRNA in Purkinje cells and DCN. Frequency distributions of levels of labelling were constructed for individual animals and their medians were determined. Data are means&S.E.M. of values obtained in seven normal littermates and eight dystonic rats. *P<0.05, when compared to normal littermates with a two-tailed Student’s t-test (means) or Mann–Whitney U-test (medians).
nucleus and the lateralis nucleus, as noted in previous immunohistochemical studies.1,13 Taking into account (i) the difficulty of distinguishing the individual nuclei of the DCN during the microscopic analysis, and (ii) the small number of GAD67 mRNA labelled neurons per nuclei, the totality of the neurons in the DCN was analysed in each section. Quantification of the autoradiographic signal over individual neurons in the Purkinje cell layer revealed a significant increase in the level of labelling for GAD67 mRNA in dt rats compared to wild-type littermates (+32%, Table 1A, Fig. 2). Analysis of the frequency distributions of the level of labelling per neuron confirmed the increase in level of labelling by showing a shift to the right of the histogram (Fig. 4B), and an increase in the medians of the distribution for dt rats (Table 1B), compared to the wild-type littermates (Fig. 4A, Table 1B). In contrast, the level of labelling for GAD67 mRNA was significantly lower in the DCN of dt rats compared to littermate controls ("29%, Table 1A, Fig. 3), without modification of the number of labelled cells (wild-type littermates, 69.4&9.6 [n=8]; dt rats, 65.6&3.8 [n=7], mean number of labelled neurons in the DCN per section&S.E.M.). Confirming the decrease in labelling, the frequency distribution of labelling per neuron was shifted to the left and the medians significantly decreased (Fig. 4D, Table 1B). Expression of enkephalin and GAD67 messenger RNA in the basal ganglia of dt rats As previously shown using a similar technique, a high level of expression of enkephalin mRNA was observed in the striatum.25 Similarly, an intense signal for GAD67 mRNA was present in a majority
GAD in cerebellum of dystonic rats
1091
Fig. 3. Photomicrographs of neurons labelled for GAD67 mRNA in the deep cerebellar nuclei following in situ hybridization. Dark-field photomicrograph (A, B); arrows identify individual neurons. Brightness differences between littermates control (A) and dt rats (B) in this layer reflect the decreased level of GAD67 mRNA found in dt rats. Scale bar=40 µm. High power bright-field photomicrograph in control littermates (C); labelling was observed over a neuron (arrow) but not over glial cells (arrowheads). Note that only the Haematoxylin-stained nuclei are visible. The pale Eosin-stained cytoplasm cannot be distinguished on this black and white photomicrograph. Scale bar=10 µm
of neurons of the globus pallidus and of the entopeduncular nucleus.7 The levels of labelling for these mRNAs were examined in the dt rats because previous studies have revealed changes in expression of these mRNAs in animal models of movement disorders.5 However, no significant differences in the levels of labelling for GAD67 mRNA in the globus pallidus and entopeduncular nucleus or enkephalin mRNA in the striatum were found in dt rats compared to wild-type littermates (Table 2).
DISCUSSION
This study revealed changes in the level of expression of GAD67 mRNA both in Purkinje cells and in GABAergic neurons of the DCN, confirming that the cerebellum is a major site of abnormalities in dt rats. In contrast, no changes were found in the levels of enkephalin mRNA in the striatum or GAD67 mRNA in the pallidum. This argues against major abnormalities in the basal ganglia of dt rats, although
1092
L. Naudon et al.
Fig. 4. Histograms of frequency distributions of labelling for GAD67 in the Purkinje cells and the DCN of normal littermates control rats (A, C) or dystonic (dt) rats (B, D). Quantification of silver grains over individual neurons was done with the Morphon Image Analysis System on sections processed for emulsion autoradiography. Data includes values from 50 Purkinje cells analysed or the totality of the labelled neurons of the DCN in two sections on both sides per rat for seven to eight rats. Note the shift to the right in the histogram in (B) and to the left in (D), illustrating, respectively, an increase in the level of labelling in Purkinje cells of dt rats, and a decrease in the level of labelling in the neurons of the DCN of dt rats. The medians of the frequency distribution (arrows) are presented in Table 1B. Table 2. Enkephalin and GAD67 messenger RNA expression in basal ganglia mRNA Enkephalin (striatum)a GAD67 (globus pallidus)b GAD67 (entopeduncular nucleus)b
Normal littermates
Dystonic rats
0.036&0.004 85.96&3.01 23.90&3.14
0.041&0.007 81.61&2.12 24.14&3.34
a
Optical density measured from autoradiograms (see Experimental Procedures). Data are means&S.E.M. values for 10 normal littermates and nine dystonic rats. P=0.51, two-tailed Student’s t-test. b Level of labelling (pixels per neuron) as measured by single-cell analysis (see Experimental Procedures). Data are means&S.E.M. values for eight normal littermates and 10 dystonic rats (striatum and globus pallidus) or four normal and five dystonic rats (entopeduncular nucleus). NS, two-tailed Student’s t-test.
the functional integrity of other neuronal circuits in this region remains to be tested. Glutamate decarboxylase messenger RNA in rat cerebellum In dt rats, the levels of GAD67 mRNA were modified compared to their littermate controls in two interconnected neuronal populations of the cerebellum, the Purkinje cells and neurons of the DCN. The
Purkinje cells form the only output pathway from the cerebellar cortex, therefore alterations in GABAergic transmission in these cells is likely to have major functional consequences. It cannot be excluded, however, that other GABAergic neruons in the cerebellar cortex7,11,12 are also affected in the mutant rats. An exhaustive study of all neuronal types in the cerebellum in dt rats was beyond the scope of this work, which focused on alterations in the main output pathway of the cerebellum for which electrophysiological studies had already provided evidence of alteration in the dt rat.17,29 Similarly, we have concentrated on the expression of the mRNA encoding GAD67, because, though mRNA encoding GAD (65000 mol. wt) is also expressed in the cerebellum,11,12 previous studies, both in the cerebellum and in other brain regions, have shown that the levels of mRNA encoding GAD67 vary in relation to the electrophysiological activity of GABAergic neurons,3,5,16,26–28 whereas alterations in GAD65 mRNA levels are less commonly observed, and usually are of a smaller amplitude in the rat.9,27,28 Although the bulk of GAD activity measured biochemically in the DCN is usually considered to originate from Purkinje cell afferents, immunohistochemical studies have revealed that the DCN also contains a population of GABAergic neurons.10,13 In situ hybridization histochemistry experiments with cellular resolution have confirmed the presence of GAD mRNA in neurons of the DCN. (Refs 11,12 and present report) The situation, however, is more complex in the DCN than in the cerebellar cortex, because, in contrast to the Purkinje cells which are homogeneously GABAergic, neurons of the DCN are heterogeneous with regard to their neurotransmitter content. In addition to small GABAergic neurons, the DCN contains small neurons containing glutamate1 and large neurons using aspartate and/or glutamate.1,10,13 A relatively high proportion of small neurons contains both GABA and glycine (50–70%), and the remainders are either GABAergic or glycine-containing neurons.4 Both GABAergic neurons and large neurons containing other neurotransmitters receive afferents from the Purkinje cells.10,13 Glutamate decarboxylase messenger RNA levels and firing patterns of cerebellar neurons in the dt rat Previous findings of an increase in GAD activity coupled to a decease in the density of GABAA receptors in the DCN suggest an increase in GABA release in the DCN resulting in a compensatory down-regulation of GABA receptors.2 The present data support the hypothesis that GABAergic transmission is increased in Purkinje cells, which express higher levels of GAD67 mRNA, and decreased in GABAergic neurons of the DCN, which express lower levels of GAD67 mRNA in dt rats than in control littermates. Such changes in GABAergic
GAD in cerebellum of dystonic rats
transmission, however, are difficult to explain based on the changes in firing rates, slower in Purkinje cells and increased in DCN neurons, which have been reported in dt rats.17,29 In contrast, changes in GAD67 mRNA could be better explained by modifications of the firing patterns reported in the mutant rats. GAD67 mRNA levels were increased in neurons showing a more irregular pattern of firing (i.e. Purkinje cells29) and decreased in a region where the most consistent finding was an increase in the regularity of firing (i.e. neurons of the DCN17). A similar relationship between firing pattern and GAD67 mRNA has been reported in the globus pallidus.5,8 Recent work has provided increasing evidence that patterns rather than rates of spontaneous firing may be critical for neurotransmitter release and synaptic efficiency in a variety of neuronal systems.15 The present data support this hypothesis by showing a correlation between levels of the mRNA encoding the
1093
enzyme of GABA synthesis and pattern of activity in cerebellar GABAergic neurons. CONCLUSION
In conclusion, the results of this study confirm a complex alteration in the functioning of the cerebellar output in the dt rat, which is consistent with abnormal electrophysiological activity in target regions from the DCN.14,17 Furthermore, this study indicates that the GABAergic subpopulation of DCN neurons may contribute significantly to the generation of this abnormal output. Acknowledgements—We are grateful to Drs A. Tobin (UCLA) and S. Sabol (NIH) for the gift of cDNAs. This work was supported by PHS grant R37 MH-44894 (M.F.C.) and the Dystonia Medical Research Foundation (J.F.L.)
REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.
Batini C., Compoint C., Buisseret-Delmas C., Daniel H. and Guegan M. (1992) Cerebellar nuclei and the nucleocortical projections in the rat: retrograde tracing coupled to GABA and glutamate immunohistochemistry. J. comp. Neurol. 315, 74–84. Beales M., Lorden J. F., Walz E. and Oltmans G. A. (1990) Quantitative autoradiography reveals selective changes in cerebellar GABA receptors of the rat mutant dystonic. J. Neurosci. 10, 1874–1885. Benson D. L., Huntsman M. M. and Jones E. G. (1994) Activity-dependent changes in GAD and preprotachykinin mRNAs in visual cortex of adult monkeys. Cerebr. Cortex 4, 40–51. Chen S. and Hillman D. E. (1993) Colocalization of neurotransmitters in the deep cerebellar nuclei. J. Neurocytol. 22, 81–91. Chesselet M.-F. and Delfs J. M. (1996) Basal ganglia and movement disorders: an update. Trends Neurosci. 19, 417–422. Chesselet M.-F. and Weiss-Wunder L. T. (1994) Quantification of in situ hybridization histochemistry. In In situ Hybridization in Neurobiology, Advances in Methodology (eds Eberwine J. H., Valentino K. L. and Barchas J. D.), pp. 114–123. Oxford University Press, New York. Chesselet M.-F., Weiss L., Wuenschell C., Tobin A. J. and Affolter H. U. (1987) Comparative distribution of mRNAs for glutamic acid decarboxylase, tyrosine hydroxylase, and tachykinins in the basal ganglia: an in situ hybridization study in the rodent brain. J. comp. Neurol. 262, 125–140. Delfs J. M., Ciaramitaro V. M., Parry T. J. and Chesselet M.-F. (1995) Subthalamic nucleus lesions: widespread effects on changes in gene expression induced by nigrostriatal dopamine depletion in rats. J. Neurosci. 15, 6562–6575. Delfs J. M., Ciaramitaro V. M., Soghomonian J.-J. and Chesselet M.-F. (1996) Unilateral nigrostriatal lesions induce a bilateral increase in glutamate decarboxylase messenger RNA in the reticular thalamic nucleus. Neuroscience 71, 383–395. De Zeeuw C. I. and Barrebi A. S. (1995) Postsynaptic targets of Purkinje cell terminals in the cerebellar and vestibular nuclei of the rat. Eur. J. Neurosci. 7, 2322–2333. Esclapez M., Tillakaratne N. J., Kaufman D. L., Tobin A. J. and Houser C. R. (1994) Comparative localization of two forms of glutamic acid decarboxylase and their mRNAs in rat brain supports the concept of functional differences between the forms. J. Neurosci. 14, 1834–1855. Feldblum S., Erlander M. G. and Tobin A. J. (1993) Different distributions of GAD65 and GAD67 mRNAs suggest that the two glutamate decarboxylases play distinctive functional roles. J. Neurosci. Res. 34, 689–706. Kumoi K., Saito N., Kuno T. and Tanaka C. (1988) Immunohistochemical localization of ã-aminobutyric acid- and aspartate-containing neurons in the rat deep cerebellar nuclei. Brain Res. 439, 302–310. LeDoux M. S., Lorden J. F. and Meinzen-Derr J. (1995) Selective elimination of cerebellar output in the genetically dystonic rat. Brain Res. 697, 91–103. Lisman J. E. (1997) Burst as a unit of neural information: making unreliable synpases reliable. Trends Neurosci. 20, 38–43. Litwak J., Mercugliano M., Chesselet M.-F. and Oltmans G. A. (1990) Increased glutamic acid decarboxylase (GAD) mRNA and GAD activity in cerebellar Purkinje cells following lesion-induced increases in cell firing. Neurosci. Lett. 116, 179–183. Lorden J. F., Lutes J., Michela V. L. and Ervin J. (1992) Abnormal cerebellar output in rats with an inherited movement disorder. Expl Neurol. 118, 95–104. Lorden J. F., McKeon T. W., Baker H. J., Cox N. and Walkley S. U. (1984) Characterization of the rat mutant dystonic (dt): a new animal model of dystonia musculorum deformans. J. Neurosci. 4, 1925–1932. Loscher W. and Horstermann D. (1992) Abnormalities in amino acid neurotransmitters in discrete brain regions of genetically dystonic hamsters. J. Neurochem. 59, 689–694. Marsden C. D. (1992) Motor function and movement disorders. In Advances in Neurology: Dystonia 2 (eds Asbury A. K., McKhan G. M. and McDonald W. I.), pp. 309–318. W. B. Saunders Co, Philadelphia.
1094 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.
L. Naudon et al. Nobrega J. N., Richter A., McIntyre Burnham W. and Loscher W. (1995) Alterations in the brain GABAA/ benzodiazepine receptor-chloride ionophore complex in a genetic model of paroxysmal dystonia: a quantitative autoradiographic analysis. Neuroscience 64, 229–239. Oltmans G. A., Beales M. and Lorden J. F. (1986) Glutamic acid decarboxylase activity in micropunches of the deep cerebellar nuclei of the genetically dystonic (dt) rat. Brain Res. 385, 148–151. Oltmans G. A., Beales M., Lorden J. F. and Gordon J. H. (1984) Alterations in cerebellar glutamic acid decarboxylase (GAD) activity in a genetic model of torsion dystonia (rat). Expl Neurol. 85, 216–222. Paxinos G. and Watson C. (1986) In The Rat Brain in Stereotaxic Coordinates. Academic Press, San Diego. Salin P. and Chesselet M.-F. (1992) Paradoxical increase in striatal neuropeptide gene expression following ischemic lesions of the cerebral cortex. Proc. natn. Acad. Sci. U.S.A. 89, 9954–9958. Segovia J., Tillakaratne N. J. K., Whelan K., Tobin A. J. and Gale K. (1990) Parallel increases in striatal glutamic acid decarboxylase activity and mRNA levels in rats with lesion of the nigrostriatal pathway. Brain Res. 529, 345–348. Soghomonian J. J. and Chesselet M.-F. (1992) Effects of nigrostriatal lesions on the levels of messenger RNAs encoding two isoforms of glutamate decarboxylase in the globus pallidus and entopeduncular nucleus of the rat. Synapse 11, 124–133. Soghomonian J. J., Gonzales C. and Chesselet M.-F. (1992) Messenger RNAs encoding glutamate-decarboxylases are differentially affected by nigrostriatal lesions in subpopulations of striatal neurons. Brain Res. 576, 68–79. Stratton S. E., Lorden J. F., Mays L. E. and Oltmans G. A. (1988) Spontaneous and harmaline-stimulated Purkinje cell activity in rats with a genetic movement disorder. J. Neurosci. 8, 3327–3336. Wassef M., Simons J., Tappaz M. L. and Sotelo C. (1986) Non-Purkinje cell GABAergic innervation of the deep cerebellar nuclei: a quantitative immunocytochemical study in C57BL and in Purkinje cell degeneration mutant mice. Brain Res. 399, 125–135. (Accepted 13 June 1997)