- Email: [email protected]
Methylmercury-induced movement and postural disorders in developing rat: loss of somatostatin-immunoreactive interneurons in the striatum
Developmental Brain Research, 40 (1988) 11-23
11
Elsevier BRD50713
Methylmercury-induced movement and postural disorders in developing rat: loss of somatostatin-immunoreactive interneurons in the striatum John R. O'Kusky 1, James M. Radke 2 and Steven R. Vincent 2 1Department of Pathology, Division of Medical Microbiology, and 2Department of Psychiato', Kinsmen Laboratory of Neurological Research, University of British Columbia, Vancouver, B. C. (Canada) (Accepted 13 October 1987)
Key words: Methylmercury; Somatostatin; Glutamic acid decarboxylase (GAD); y-Aminobutyric acid (GABA); Cerebral cortex; Striatum; Postnatal development; Cerebral palsy
Tissue concentrations of the neuropeptide somatostatin and the specific activities of glutamic acid decarboxylase (GAD) were measured in several regions of the central nervous system in young rats, following chronic postnatal administration of methylmercuric chloride. By the beginning of the fourth postnatal week, these animals exhibited clinical signs of a mixed spastic/dyskinetic syndrome with visual deficits. At the onset of neurological impairment, a significant decrease in GAD activity was detected in the occipital cortex (48-49%) and striatum (45-50%) when compared to either normal or weight-matched controls. At one subclinical stage of toxicity, decreased GAD activity was detected only in the occipital cortex (29-30%). Tissue levels of somatostatin did not change significantly in the occipital cortex of methylmercury-treated animals at any stage of the experiment. However, somatostatin levels in the striatum were significantlyreduced at the onset of neurological impairment (55-57%) and at one subclinical stage of toxicity (49-54%). Immunohistochemistry for somatostatin- and neuropeptide Y-immunoreactive neurons confirmed a marked loss of cells in the dorsolateral region of the striatum with atrophy of the surviving neurons. In the cerebral cortex of methylmercury-treated animals the morphology and distribution of somatostatin-positive neurons appeared normal. In view of the reported co-localization of GAD and somatostatin in some non-pyramidal neurons of the cerebral cortex, these results indicate that methylmercury-induced lesions of the developing cerebral cortex involve a subpopulation of GABAergic neurons which are not co-localized with somatostatin. In the striatum, where GAD and somatostatin are not co-localized within the same neurons, methylmercury-induced lesions involve both GABAergic and somatostatin-positive neurons.
INTRODUCTION
motor signs included both spastic and flaccid paraly-
The toxicity of methylmercury and related alkylmercury compounds in the developing nervous system has been associated with a wide range of developmental disorders 4-6'~3. Following epidemic me-
sis, ataxia, athetosis, myoclonic jerking and generalized tonic convulsions. Visual disturbances varied from constriction of the visual field to total blindness. In animal models of methylmercury neurotoxicity during prenatal and early postnatal development,
thylmercury poisoning in Japan and Iraq, clinical studies of impairment infants and children have documented signs of neurological impairment resembling cerebral palsy syndrome 1"2'6,13"14,22. Clinical signs included severe psychomotor retardation, progressive microcephaly, persistence of primitive reflexes, hyperreflexia and hypersalivation. Additional
signs of neurological impairment have varied from subtle behavioral deficits 35'36 to the full spectrum of movement and postural disorders 29'3°. Recent studies have shown that the m o v e m e n t and postural disorders characteristic of methylmercuryinduced cerebral palsy syndrome can be produced in the developing rat following postnatal administration
Correspondence: J.R. O'Kusky, Kinsmen Laboratory of Neurological Research, 2255 Wesbrook Mall, Vancouver, B.C., V6T IW5, Canada. 0165-3806/88/$03.50© 1988 Elsevier Science Publishers B.V. (Biomedical Division)
12 of methylmercuric chloride 29"3°. At the onset of neurological impairment, lesions of the cerebral cortex were found to be more localized than those reported in human autopsy material 24. In the visual cortex, degenerating neurons were concentrated mainly in layer IV and the adjacent regions of layers III and V with a more diffuse distribution in the supra- and infragranular layers. Morphological features of degenerating neurons included a radial distribution of dendrites with few spines on degenerating dendritic profiles. A relatively selective degeneration of axon terminals forming symmetric contacts was observed in the cortical neuropil of all laminae. These morphological criteria suggest that aspinous or sparsely spinous stellate neurons, the inhibitory interneurons of the neocortex which utilize 7-aminobutyric acid (GABA) as a neurotransmitter, are most susceptible to methylmercury-induced lesions of the developing cerebral cortex. At the onset of neurological impairment, a significant decrease in the specific activity of glutamic acid decarboxylase (GAD), the GABA synthetic enzyme, was detected in both the cerebral cortex and striatum of methylmercury-treated animals 29. lmmunohistochemical studies of the localization of GAD have shown that GABAergic interneurons in the cerebral cortex comprise several morphological classes which are distributed throughout most cortical laminae 3~'34. Given the unusual laminar distribution of degenerating neurons in methylmercury-induced lesions of the cerebral cortex, it is possible that only particular subpopulations of GABAergic neurons are affected. In view of the reported co-localization of the neuropeptides somatostatin and neuropeptide Y with GAD in some interneurons of the cerebral cortex 16, it is possible that such co-localization defines such a subpopulation of GABAergic neurons. The present study was conducted to determine the extent to which somatostatin- and neuropeptide Yimmunoreactive neurons are involved in methylmercury-induced lesions of the developing cerebral cortex and striatum, The specific activity of GAD and tissue concentrations of somatostatin were determined in the cerebral cortex and striatum at the onset of neurological impairment and at two subclinical stages of toxicity. Immunohistochemistry for both somatostatin and neuropeptide Y was employed to in-
vestigate the pathology of peptidergic neurons during the course of methylmercury-induced movement and postural disorders. A histochemical method to detect reduced nicotinamide adenine dinucleotide phosphate (NADPH)-diaphorase activity was also employed, since this technique selectively stains striatal neurons which contain both somatostatin and neuropeptide y38. Preliminary results of these experiments have been published in abstract form 32. MATERIALS AND METHODS
Seven litters of Sprague-Dawley rats were culled to 9 pups per litter on postnatal day 2. Littermates of the same sex and similar body weight were organized into matched triplets. Individual members of these triplets were randomly assigned to one of 3 groups (methylmercury-treated, MeHg; weight-matched control, WMC; normal control, NC) and cross-fostered to separate dams. MeHg rats received subcutaneous injections of 0.01 M methylmercuric chloride in physiological saline (5 mg Hg/kg/day) beginning on postnatal day 5 and continuing until one of 3 stages of toxicity, defined as follows: stage I (day 18) when MeHg rats continued to gain weight although less rapidly than normal controls, stage II (days 20-22) when a given MeHg animal exhibited a persistent loss of body weight over 24 h, and stage III (days 22-25) at the onset of neurological impairment. Control animals were injected daily with an equivalent volume of saline. WMC animals were periodically isolated from their dam and placed in an incubator at 37 °C, maintaining their body weight within + 8% of MeHg animals. Nineteen matched triplets were employed to assay GAD activity and somatostatin immunoreactivity at the 3 stages of toxicity. The matched triplets were sacrificed by cervical fracture and the brains were immediately removed and dissected on ice. Tissue was sampled from the cerebral cortex (occipital region, including the primary visual cortex) and the striatum (caudate-putamen) at all 3 stages. From the 6 triplets at stage III, additional tissue was sampled from the thalamus, hypothalamus, hippocampus, substantia nigra, superior colliculus, cerebellum (vermal cortex, Iobules II-X) and spinal cord (cervical enlargement). For assays of the specific activity of GAD, each tis-
13 sue sample was homogenized in 10 vols. of cold 0.25 M sucrose. The specific activity of G A D was determined by measuring the UC02formed on incubation of 0.04 ml of tissue homogenate with L-[1-1aC]glutamic acid in the presence of excess pyridoxal phosphate .... . Samples were assayed in duplicate with the mean value expressed as either (1) ~mol CO 2 per g tissue wet weight per h or (2)/~mol CO 2 per mg protein per h. Protein was assayed according to Lowry et al. > with bovine serum albumin as standard. The tissue for somatostatin radioimmunoassay was weighed, boiled 10 rain in 1.0 ml of 1 N acetic acid and then sonicated. Samples were centrifuged at 4 °C for 20 rain at 10,000 rpm. The supernatant was lyophilized and stored at -70 °C until the day of assay. Somatostatin immunoreactivity was determined using a monoclonal antibody (Ab 3) 39 and Tyr-l-somatostatin labelled with 12sI via the chloramine T method. A standard curve was constructed with the antibody, labelled somatostatin and synthetic cyclic somatostatin (Peninsula Labs., Belmont, CA) in 0.4 ml of 24 mM sodium phosphate, 3.4 mM sodium acetate, 43.6 mM sodium chloride, 0.25 mM thimerosal, 0.5% bovine serum albumin and 500 KIU/ml aprotinin (pH 7.4). Tissue samples were diluted appropriately in assay buffer to obtain results within the sensitivity range of the standard curve (2.5-100 pg/mg). After a 3-day incubation at 4 °C, the free labelled somatostatin was separated with dextran-coated charcoal and counted in a gamma counter. Samples were assayed in duplicate with the mean value expressed as pg somatostatin immunoreactivity per mg tissue wet weight. The assay detects both somatostatin-14 and somatostatin-28 in equimolar amounts. The statistical significance of differences among the 3 groups at each stage of toxicity was determined using a 3 x 3 (stages x groups) analysis of variance 41. Two matched triplets at stages II and Ill were employed for immunohistochemistry (somatostatin, neuropeptide Y) and histochemistry (NADPH-diaphorase). Animals were anesthetized by an intraperitoneal injection of sodium pentobarbital (50 mg/kg) and perfused through the ascending aorta with a fixative solution containing 4% paraformaldehyde and 1 mM MgC12 in 0.1 M phosphate buffer (pH 7.4). The brains were removed, bisected in the midline and stored overnight in 0.1 M phosphate buffer containing 1 mM MgCI 2 and 15% sucrose. From the right
hemisphere of each brain, serial frozen sections (30 1~m) were cut in the frontal plane and organized into 4 series, with each series containing every fourth section through the cerebral hemisphere. Sections for somatostatin immunohistochemistry were processed using a monoclonal antibody (Ab 8) as previously described 39. Sections for neuropeptide Y immunohistochemistry were processed using a commercially available rabbit antiserum (RAS 7172, Peninsula Laboratories) as previously described 27. Sections for NADPH-diaphorase histochemistry were processed according to the indirect method of Vincent et al. 3s. Briefly, sections were incubated in 0.1 M Tris-C1 (pH 8.0) containing 1.0 mM NADP, 0.2 mM nitro blue tetrazolium, 15 mM sodium malate and 0.4% Tween80 at 37 °C for 60 rain. Sections were rinsed in phosphate-buffered saline (PBS) and coverslipped with glycerol-PBS. A fourth series of sections was stained for Nissl substance using 0.1% thionine in acetate buffer (pH 3.7). From the left hemisphere of each brain, blocks of tissue (0.5 mm thick) were sampled from the occipital cortex and striatum. These blocks were postfixed in 1% buffered osmium tetroxide, dehydrated in ascending grades of ethanol and embedded in Araldite. Semithin sections (0.5/2m) were cut and stained with 1% Toluidine blue in 0.4% sodium tetraborate. RESULTS The earliest sign of toxicity in MeHg animals was an abnormal rate of growth. The postnatal increase in body weight for MeHg animals was normal for the first 5 - 6 days of the experiment, followed by a period of decelerating growth where they continued to gain weight although less rapidly than NC animals. After attaining a maximum value between days 20 and 22, body weight gradually decreased, following within 2-4 days by the onset of neurological impairment. The average body weight for MeHg rats was significantly less than that of NC rats at all stages of toxicity (Table I). At no time during the experiment did the average body weight for MeHg rats differ significantly from that of WMC rats. At the time of sacrifice, the average brain weight for MeHg rats was also significantly less than that of NC rats at all stages (Table I). However, brain weights did not differ significantly between MeHg and WMC animals. It
14
days 22 and 25. These animals exhibited what appeared to be a mixed spastic/dyskinetic syndrome. All animals displayed a mild to marked hypertonia of flexor muscles in all 4 limbs, which was symmetrical and more pronounced in the hindlimbs. Flexion deformities were observed in a number of postures. For example, when NC and WMC animals were suspended by the tail, all 4 limbs were held in full extension. When MeHg animals were suspended by the tail, extension of the forelimbs was weak and incomplete while the hindlimbs were excessively adducted in full flexion. This adduction of the hindlimbs was similar to the phenomenon of hindlimb crossing reported in adult rats following exposure to methylmercury26. However, in these developing rats the hindlimbs were tightly flexed as opposed to the flaccid paralysis seen in adult animals. When MeHg animals were suspended by the scruff, the forelimbs were moderately flexed and abducted. When stimulated by manipulation of the hindlimbs in this position, extensor thrusts of the forelimbs were frequently observed with abnormal rotation of the limb and marked flexion of the digits. Hyperreflexia was observed in the auricular startle response and in the crossed extensor reflex. All MeHg animals exhibited some degree of visual impairment. The visual placing response was weak or absent, while the vibrissa placing and forelimb placing responses were present. Unlike control animals, they failed to avoid a visual cliff and
TABLE 1
Body weight and brain weightfor MeHg, WMC and NC animals at the three stages of toxicity Stage
Matched triplets, n =
1 6
Age of onset (days) Mean 18.0 Range 18
11
Ili
8
7
20.4 20- 22
23.8 22- 25
40.2+3.4* 30.4+1.8 31.8+2.6
50.6_+3.6* 32.3+2.8 33.7+3.2
68.2+3.9** 30.8+3.4 28.7+2.8
1.30+0.03" 1.23-+0.04 1.17+0.04
1.38_+0.04" 1.24_+0.06 1.19+0.05
1.51 _+ 0.04** 1.30_+0.06 1.25+0.05
Body weight (g) (mean +_ S.E.M. NC WMC MeHg Brain weight (g) (mean _+S~E.M. NC WMC MeHg
* P < 0.01, **P < 0.001, compared to MeHg animals.
would appear that the reduced brain weights in MeHg rats resulted from the associated malnutrition rather than from a direct neurotoxic effect of methylmercury. Clinical signs of neurological impairment in the 7 MeHg animals at stage III were observed between TABLE 1I
Tiss'ue concentrations of somatostatin and specific activities of GAD in the occipital cortex and striatum of MeHg, WMC and NC animals' at the three stages of toxicity ~Concentrations in pg/mg tissue wet weight as the mean + S.E.M. ; t'Enzyme activities in/~mol/g tissue/h as the mean _+ S.E.M.
Stage
Somatostatin"
GAD b
I
11
111
1
11
III
Matched triplets ~l-
6
7
6
6
7
6
Occipital cortex NC WM( McHg
113.9 ~ 12.3 128.6,+ 1 4 . 2 124.6± 7.9
96.2_+20.2 102.(1± 9.9 115.0,+ 15.4
182.1 ,+31.1 191.6+24.5 189.0_+26.3
12.1 + 0 . 7 9 . 2 + 1.1 9.3+0.8
15.4+0.8' 15.6+0.6" 10.9+ 1.1
16.9+0.7"* 17.3+0.9" 8 . 8 ± 1.2
146.9± 13.4 :+ 162.~_+_20.5 + 75.3~: 13.5
125.6+ 14.2 ~ 131.8±21.1 +: 56.4+_ t,~7
11.3 _+ 0.~; 10.2± 1.0 9,9-+ 1.1
13.2_+0.5 14.0_+0.7
14.3_+ 1.0'* 15.6+_0.4 :::
12.0_+ 1.1
7.8-+ 1.1
SlrJattml ~,{ ~M( ~lelle
7 7 . 9 + 14.! 6 1 . 6 + 4.~ 75.1~ ~ (,.I
i' < (> i)l, / ' < {)t)0l, compared to Mcflg animals.
15 demonstrated difficulty in negotiating obstacles in a novel environment. Myoclonic jerking of the hindlimbs was observed in 2 MeHg animals immediately prior to sacrifice.
nificant differences were detected among MeHg and control animals in the specific activity of G A D , tissue concentrations of somatostatin or concentrations of total protein.
GAD activity/somatostatin immunoreactivity Postnatal changes in the specific activity of G A D are presented in Table II. Treatment with methylmercury resulted in significant decreases in G A D activity in the cerebral cortex at stage II (29-30%) and stage III (48-49%). In the striatum a significant decrease was detected in MeHg animals only at stage III (45-50%). The differences between NC and WMC animals were not significant at any stage, indicating that the degree of malnutrition occurring during the experiment was not sufficient to alter the normal postnatal changes in G A D activity. An analysis of G A D kinetics with respect to glutamic acid was performed using combined tissue homogenates from the cerebral cortex and striatum at stage III of toxicity. No significant differences in the affinity for substrate were detected among MeHg, WMC and NC animals, with values of K m ranging from 0.9-1.2 x 10 -3 M. The value of Vmax for MeHg animals was found to be only 69% of the value for control animals. Concentrations of total protein in the cerebral cortex and striatum did not differ significantly among MeHg, WMC and NC animals at any stage of the experiment. Consequently, expressing the specific activity of G A D as/~mol CO2/mg protein/h did not alter the experimental findings. The results of radioimmunoassay for tissue concentrations of somatostatin in the cerebral cortex and striatum are presented in Table II. Tissue concentrations of somatostatin in the cerebral cortex did not differ significantly among MeHg, WMC and NC animals at any stage of the experiment. In the striatum a significant decrease in somatostatin levels was detected in MeHg animals at stage II (49-54%) and stage III (55-57%) when compared to either WMC or NC animals. Assays for G A D activity and somatostatin immunoreactivity were performed in tissue samples from several additional regions of the central nervous system at the onset of neurological impairment (stage III). In all regions examined, including the thalamus, hypothalamus, hippocampus, substantia nigra, superior colliculus, cerebellum and spinal cord, no sig-
Histopathology in the cerebral cortex and striatum Somatostatin immunohistochemistry in the cerebral cortex confirmed the results of radioimmunoassay. The morphology and distribution of somatostatin-immunoreactive neurons in the cerebral cortex of MeHg animals appeared to be normal (Fig. 1). Neurons were distributed throughout layers I I - V I and were occassionally observed in the subcortical white matter. In layer IV the cell bodies were somewhat smaller and less intensely stained than neurons in the supra- and infragranular layers. Immunoreactive dendritic processes exhibited no systematic pattern of branching and examples of multipolar, bipolar and bitufted neurons could be identified. In MeHg animals the length and branching pattern of dendritic processes appeared normal. Neuropeptide Y-immunoreactive neurons in the cerebral cortex were similar in size, morphology and distribution to somatostatin-immunoreactive neurons, although fewer in number. In MeHg animals these neurons appeared normal. Cortical pathology was evident on Toluidine bluestained sections of the cerebral cortex at both stage II and stage III. Degenerating neurons were observed in concentrated mainly in layer IV with a more diffuse distribution in layers II, III, V and VI. The cell bodies of degenerating neurons exhibited an increased staining of the cytoplasm with an accumulation of clear vacuoles. Their nuclei were condensed and pyknotic. This pattern of pathology was similar to that reported in an earlier study 3°. Somatostatin immunohistochemistry of the striaturn revealed extensive pathology in MeHg animals (Fig. 2). There was a marked loss of neurons in the dorsolateral half of the caudate-putamen (Fig. 2A,B). The loss of somatostatin-immunoreactive neurons along the dorsolateral margin of the nucleus was virtually complete. The cell bodies of surviving neurons in the ventromedial region were smaller with a reduction in the number and length of immunoreactive dendritic processes (Fig. 2C,D). The moderately dense network of somatostatin-immunoreactive fibers, normally seen throughout the nucleus in NC
1~
and WMC animals, was markedly reduced in MeHg animals: Neuropeptide Y immunohistochemistry revealed the same pattern of pathology (Fig. 3) with a loss of neurons in the dorsolateral region of the striatum and atrophy of surviving neurons. The dense network of neuropeptide Y-immunoreactive fibers seen in control animals was virtually absent in the MeHg animals (Fig. 3C,D). NADPH-diaphorase histochemistry confirmed the loss of neuropeptideimmunoreactive neurons in the striatum (Fig. 4). Surviving NADPH-diaphorase-positive neurons were atrophic with a reduction in the number and
length of labeled dendritic processes. Labeled fibers were distinctly beaded in appearance. The presence of degenerating neurons in the striatum of MeHg animals was determined on Toluidine blue-stained sections (Fig. 5). At subclinical stage II (Fig. 5A) the nuclei of degenerating neurons were condensed and pyknotic. Increased intensity of staining was evident in the cytoplasm, which contained numerous clear vacuoles. Degenerating neurons were scattered along the dorsolateral margin of the caudate-putamen, representing less than 2% of the total number of neuronal profiles. Many adjacent
200
.~:ig. 1. Somatostatin-immunoreactive neurons in the primary visual cortex of (A) weight-matched control and (B) MeHg-treated anili~.qs at lhc onset of neurological impairment (stage III). Calibration bar in iml.
b
8
~J
3 ~
p~
~
r~
3
i ~
~
0
~Z
E~
2{}
Fig. 5. Aratdite-embedded sections (0.5 ~m, stained with Toluidine blue) from the striatum of MeHg-treated animals at (A) stage II and (B) stage III. Degenerating neurons were pyknotic with an accumulation of clear vacuoles in the cytoplasm. At stage II, isolated degenerating neurons were observed among neurons with apparently normal morphology (n). Calibration bar in/~m. neurons and the surrounding neuropil a p p e a r e d normal. At the onset of neurological i m p a i r m e n t (Fig. 5B), numerous degenerating neurons were o b s e r v e d in the dorsolateral region of the striatum. A l t h o u g h the m o r p h o l o g y of these neurons was similar to that seen at stage II, the surrounding neuropil had a spongy a p p e a r a n c e resulting from m a r k e d astrocyte cuffing and swelling of dendritic processes.
DISCUSSION The results of the present study d e m o n s t r a t e that a specific subpopulation of G A B A e r g i c interneurons is involved in m e t h y l m e r c u r y - i n d u c e d lesions of the developing cerebral cortex. Many neurons in the cerebral corlex a p p e a r to use G A B A as a neurotransmittcr ~ Neurons displaying G A D immunoreactivity have been identified in all layers of the rat visual corto,: q ~. l'hcse neurons arc non-pyramidal in their morphology and have been divided into several types ~ccnrding to the size and laminar distribution of their
cell bodies and the m o r p h o l o g y and sites of termination of their axon terminals. Some G A D - i m m u n o reactive neurons in the cerebral cortex of the cat and monkey have been shown to contain the n e u r o p e p tides somatostatin and n e u r o p e p t i d e y16. All somatostatin- and n e u r o p e p t i d e Y-immunoreactive neurons in the cerebral cortex of the cat and most of these neurons in the cerebral cortex of the m o n k e y are also G A D - i m m u n o r e a c t i v e . Although they occur in all cortical laminae, they have a preferential distribution in layers II, I l i and VI of the cerebral cortex. Like other cortical interneurons they are small aspiny non-pyramidal neurons whose axons form predominantly symmetric synapses, although occasional asymmetric p e p t i d e - i m m u n o r e a c t i v e contacts have been r e p o r t e d jr. Some G A D - i m m u n o r e a c t i v e neurons do not a p p e a r to contain these neuropeptides. These neurons occur in all laminae, but they have a preferential distribution in layers IV and V. In the present study the neurotoxicity of methylmercury resulted in a significant decrease in the specific activity t}f G A D in the cerebral cortex at the on-
21 set of neurological impairment (stage III) and at subclinical stage II. Cortical concentrations of somatostatin did not change during any stage of the experiment. In the cerebral cortex of MeHg animals, degenerating neurons were observed concentrated mainly in layer IV and the adjacent regions of layers III and V with a more diffuse distribution in the supra- and infragranular layers. This laminar distribution of degenerating neurons agrees with a previous study, where degenerating axons forming predominantly symmetric synapses were observed in all cortical laminae of MeHg animal 3°. These morphological criteria are consistent with a selective degeneration of GABAergic interneurons. The normal morphology and distribution of somatostatin- and neuropeptide Y-immunoreactive neurons in the cerebral cortex of MeHg animals indicates that these neurons were spared in methylmercury-induced lesions. Thus, the neurotoxicity of methylmercury in the developing cerebral cortex appears to result in a relatively selective degeneration of those GABAergic neurons which are not co-localized with somatostatin and neuropeptide Y. This selective pathology would explain the unusual laminar distribution of degenerating GABAergic neurons. GAD-immunoreactive neurons which are not co-localized with somatostatin and neuropeptide Y are preferentially distributed in laminae IV and V with a more diffuse distribution in the supra- and infragranular layers 16. One morphologically distinct class of GAD-immunoreactive neuron in the cerebral cortex, the basket cell, appears to be among those GABAergic intemeurons which are not co-localized with somatosatin or neuropeptide y16. Basket cells have a large cell body and are distributed mainly in layers II1-V118. Their axons form pericellular nests around the cell bodies and proximal dendrites of pyramidal neurons in the deep cortical layers 12'19'21. In a previous ultrastructural study of synaptic degeneration in methylmercury-induced lesions 3°, it was found that pyramidal neurons in layers III, V and VI had multiple degenerating terminals forming symmetric synapses on their somata. These morphological criteria, combined with the neurochemical results of the present study, suggest that cortical basket cells comprise one type of GABAergic interneuron involved in methylmercury-induced lesions of the cerebral cortex.
An unexpected finding of the present study was the significant decrease in concentrations of somatostatin in the striatum at the onset of neurological impairment (stage III) and at subclinical stage II. This decrease is specific to the striatum as tissue levels of somatostatin were normal in the other 8 regions of the central nervous system examined. Decreased somatostatin levels appear to result from a loss of somatostatin-immunoreactive neurons, particularly in the dorsolateral margin of the nucleus. Similar cell losses were noted using neuropeptide Y immunohistochemistry and NADPH-diaphorase histochemistry. Given that these 3 substances have been shown to be co-localized in the same neurons in the stria° tum 3s, there would appear to be a loss of neurons rather than an inhibited synthesis of somatostatin. The presence of degenerating neurons in the striatum at stages II and III, as seen on Toluidine blue-stained sections, confirmed a neuron loss in the appropriate region of the nucleus. In the striatum of the rat, somatostatin-immunoreactive neurons comprise a relatively sparse and homogeneous population of medium-sized aspiny neurons 37'4°. These are striatal interneurons whose axons are distributed within the caudate-putamen and form symmetric synapses mainly on dendritic shafts 9'37. Although their function is largely unknown, there is preliminary evidence to suggest that somatostatin-immunoreactive neurons in the striatum have some influence on the extrapyramidal control of movement. In experimental animals, treatment with somatostatin modulates the release of dopamine 7 and alters levels of spontaneous motor activity 3'x5'33. Decreased levels of somatostatin have been detected in the cerebrospinal fluid of patients with Parkinson's disease l° and Huntington's chorea s. Autopsy material from the neostriatum of patients with Huntington's chorea has been shown to contain both altered tissue concentrations of somatostatin 2s and altered morphology of somatostatin-immunoreactive neurons and fiber networks 23. In the present study the relatively early loss of somatostatin-immunoreactive neurons in the striatum suggests that these neurons may be involved in the pathogenesis of pediatric movement and postural disorders. The results of the present study indicate that discrete populations of interneurons in the cerebral cortex and striatum are selectively impaired in methyl-
m e r c u r y - i n d u c e d lesions of the d e v e l o p i n g central
genesis of m e t h y l m e r c u r y - i n d u c e d
n e r v o u s system. In t h e c e r e b r a l c o r t e x G A B A e r g t c i n t e r n e u r o n s , which are p r o b a b l y not c o - l o c a l i z e d
postural d i s o r d e r s d u r i n g d e v e l o p m e n t .
movement
w~th the n e u r o p e p t i d e s o m a t o s t a t i n , d e g e n e r a t e at a
ACKNOWLEDGEMENTS
and
subclinical stage o f toxicity. In the s t r i a t u m i n t e r n e u rons which are s o m a t o s t a t i n - and n e u r o p e p t i d e Y-
This r e s e a r c h was s u p p o r t e d by grants f r o m the
i m m u n o r e a c t i v e d e g e n e r a t e at this subclinical stage.
U n i t e d C e r e b r a l Palsy R e s e a r c h a n d E d u c a t i o n a l
A t the o n s e t of n e u r o l o g i c a l i m p a i r m e n t , d e g e n e r a t ing striatal n e u r o n s a p p e a r to include b o t h G A B A -
F o u n d a t i o n , Inc. and the M e d i c a l R e s e a r c h C o u n c i l of C a n a d a . J . R . O . and S . R . V . are R e s e a r c h Schol-
ergic and p e p t i d e r g i c n e u r o n s . F u r t h e r r e s e a r c h is
ars of the M R C . W e t h a n k D r . J . C . B r o w n , M R C
n e e d e d to d o c u m e n t the m e t a b o l i c c h a n g e s in these
Regulatory
neurotransmitter-specific
C o l u m b i a , for the m o n o c l o n a l a n t i b o d i e s to s o m a t o -
neurons,
which
precede
Peptide
Group,
U n i v e r s i t y of British
n e u r o n a l d e g e n e r a t i o n and c o n t r i b u t e to the p a t h o -
statin.
REFERENCES
13 Harada, H., Methyl mercury poisoning due to environmental contamination (Minamata disease). In F.W. Oehme (Ed.), Toxicity of Heavy Metals in the Environment, Marcel Dekker, New York, 1978, pp. 261-302. 14 Harada, Y., Congenital (or fetal) Minamata disease. In M. Katsuma (Ed.), Minamata Disease, Kumamoto University, Kumamoto, Japan, 1968, pp. 93-117. 15 Havlicek, V., Rezek, M. and Friesen, H., Somatostatin and thyrotropin-releasing hormone: central effect on sleep and motor system, Pharmacol. Biochem. Behav., 4 (1976) 455-459. 16 Hendry, S.H.C., Jones, E.G., DeFelipe, J., Schmechel, D., Brandon, C. and Emson, P.C., Neuropeptide-containing neurons of the cerebral cortex are also GABAergic, Proc. Natl. Acad. Sci. U.S.A., 81 (1984) 6526-6530. 17 Hendry, S.H.C., Jones, E.G. and Emson, P.C., Morphology, distribution, and synaptic relations of somatostatin- and neuropeptide Y-immunoreactive neurons in rat and monkey neocortex, J. Neurosci., 4 (1984) 2497-2517. 18 Houser, C.R., Hendry, S.H.C., Jones, E.G. and Vaughn, J.E., Morphological diversity of immunocytochemically identified GABA neurons in the monkey sensory-motor cortex, J. Neurocytol., 12 (1983) 617-638. 19 Jones, E.G., Varieties and distribution of non-pyramidal cells in the somatic sensory cortex of the squirrel monkey, J. Comp. Neurol., 160 (1975) 205-268. 20 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, E.J., Protein measurement with the Folin phenol reagent, J. Biol. Chem., 193 (1951) 265-275. 21 Marin-Padilla, M., Origin of the pericellular baskets of the pyramidal cells of the human motor cortex: a Golgi study, Brain Res., 14 (1969) 633-646. 22 Marsh, D.O., Myers, G.J., Clarkson, T.W., Amin-Zaki, L., Tikriti. S. and Majeed, M.A., Fetal methylmercury poisoning: clinical and toxicological data on 29 cases, Ann. Neurol., 7 (1980) 348-353. 23 Marshall, P.E. and Landis, D.M.D., Huntington's disease is accompanied by changes in the distribution of somatostatin-containing neuronal processes, Brain Res., 329 (1985) 71-82. 24 Matsumoto. H., Koya, G. and Takeuchi, T.. Fetal Minamata disease, J. Neuropathol. Exp. Neurol., 24 (1965) 563- 574.
1 Amin-Zaki, L.. Elhassani, S., Majeed, M.A., Clarkson,
2
3
4
5 6
7
8
9
10
li
T.W., Doherty, R.A. and Greenwood, M., Intrauterine methylmercury poisoning in Iraq, Pediatrics, 54 (1974) 587--595. Amin-Zaki, k., Majeed, M.A., Elhassani, S.B., Clarkson, T.W., Greenwood, M.R. and Doherty, R.A., Prenatal methylmercury poisoning, Am. J. Dis. Child.. 133 (1979) 172- 177. Brown, M. and Vale, W., Central nervous system effects of hypothalamic peptides. Endocrinology, 96 (1975) 1333-1336. Cassidy. D.R., Toxicity of inorganic and organic mercury compounds in animals. In F.W. Oehme (Ed.), Toxicity of Heavy Metal~ in the Environment, Marcel Dekker, New York, 1978, pp. 303-330. Chang, L.W., Neurotoxic effects of mercury - - a review, Emiron. Res., 14 (1977) 329-373. ('hang, L.W. and Annau, Z., Developmental neuropathoIogy and behavioral teratology of methylmercury. In J. Yanai (Ed.), Neurobehavioral Teratology, Elsevier, New York, 1984, pp. 405-432. Chesselet. M.F, and Reisine, T.D., Somatostatin regulates dopamine release in rat striatal slices and cat caudate nuclei, J. Neurosci.. 3 (1983) 232-236. Cramcr. H.J., Kohler, G., Oepen, G., Schomburg, G. and Schratcr. E.. Huntington's chorea. Measurements of somatostatin, substance P and cyclic nucleotides in the cerebrospinal fluid, J. Neurol., 225 (1981) 183-187. DiFiglia, M. and Aronin, N.. Quantitative electron microscopic study of immunoreactive somatostatin axons in the ra~t neostriatum, Neurosci. Lett., 50 (1984) 325-331. Dupont, E., Christensen, S.E., Hansen, A.P., Olivarius, B.F. and Orskov, H., Low cerebrospinal fluid somatostatin m Parkinson's disease: an irreversible abnormality, Neurol~g), 32 (1982) 312-314. Emsan, P.C. and Lindvall, O., Distribution of putative ~etm~transmitters in the neocortex, Neuroscience, 4 (1979)
12 Fairctt, A., 1)cFelipc, J. and Regidor, J.. Non-pyramidal llcur~ms In A. Peters and E.G. Jones (Eds.). Cerebral Cor~,'~, Plenum, New York, 1984. pp. 201-253.
23 25 McGeer, P.L., McGeer, E.G., Wada, J.A. and Jung, E., Effects of globus pallidus lesions and Parkinson's disease on brain glutamic acid decarboxylase, Brain Res., 32 (1971) 425-431. 26 Miyakawa, T. and Deshimaru, M., Electron microscopical study of experimentally induced poisoning due to organic mercury compound, Acta Neuropathol.. 14 (1969) 126-133. 27 Nakamura, S. and Vincent, S.R., Somatostatin- and neuropeptide Y-immunoreactive neurons in the neocortex in senile dementia of Alzheimer's type, Brain Res., 370 (1986)
34
35
36
11-20.
28 Nemeroff, C.B., Youngblood, W.W., Manberg, P.J., Prange, A.J. and Kizer, J.S., Regional brain concentrations of neuropeptides in Huntington's chorea and schizophrenia, Science, 221 (1983) 972-975. 29 O'Kusky, J.R. and McGeer, E.G., Methylmercury poisoning of the developing nervous system in the rat: decreased activity of glutamic acid decarboxylase in cerebral cortex and neostriatum, Dev. Brain Res., 21 (1985) 299-306. 30 O'Kusky, J., Synaptic degeneration in rat visual cortex following neonatal administration of methylmercury, Exp. Neurol., 89 (1985) 32-47. 31 Peters, A., Proskauer, C.C. and Ribak, C.E., Chandelier cells in rat visual cortex, J. Comp. Neurol., 206 (1982) 397-416. 32 Radke, J.M., O'Kusky, J.R. and Vincent, S.R., Methylmercury poisoning of the developing nervous system: effects on somatostatin immunoreactivity in cerebral cortex and striatum, Soc. Neurosci. Abstr., 12 (1986) 1362. 33 Rezek, M., Havlicek, V., Leybin, L., Pinsky, C., Kroeger, E.A., Hughes, K.R. and Friesen, H., Neostriatal administration of somatostatin: differential effect of small and large
37
38
39
40
41
doses on behavior and motor control, Can. J. Physiol. Pharmacol., 55 (1977) 234-242. Ribak, C.E., Aspinous and sparsely-spinous stellate neurons in the visual cortex of rats contain glutamic acid decarboxylase, J. Neurocytol., 7 (1978) 461-478. Rice, D.C. and Gilbert, S.G., Early chronic low-level methylmercury poisoning in monkeys impairs spatial vision, Science, 216 (1982) 759-761. Spyker, J.M., Sparber, S.B. and Goldberg, A.M., Subtle consequences of methylmercury exposure: behavioral deviations in offspring of treated mothers, Science, 177 (1972) 621-623. Takagi, H., Somogyi, P., Somogyi, J. and Smith, A.D., Fine structural studies on a type of somatostatin-immunoreactive neuron and its synaptic connections in the rat neostriatum: a correlated light and electron microscopic study, J. Comp. Neurol., 214 (1983) 1-16. Vincent, S.R., Johansson, O., H6kfelt, T., Skirboll, L., Elde, R.P., Terenius, L., Kimmel, J. and Goldstein, M., NADPH-diaphorase: a selective histochemical marker for striatal neurons containing both somatostatin- and avian pancreatic polypeptide (APP)-like immunoreactivities, J. Comp. Neurol., 217 (1983) 252-263. Vincent, S.R., Mclntosh, C.H.S., Buchan, A.M.J. and Brown, J.C., Central somatostatin systems revealed with monoclonal antibodies, J. Comp. Neurol., 238 (1985) 169-186. Vincent, S.R., Staines, W.A. and Fibiger, H.C., Histochemical demonstration of separate populations of somatostatin and cholinergic neurons in the rat striatum, Neurosci. Lett., 35 (1983) 111-114. Winer, B.J., Statistical Principles in Experimental Design, McGraw-Hill, New York, 1971, pp. 591-603.
11
Elsevier BRD50713
Methylmercury-induced movement and postural disorders in developing rat: loss of somatostatin-immunoreactive interneurons in the striatum John R. O'Kusky 1, James M. Radke 2 and Steven R. Vincent 2 1Department of Pathology, Division of Medical Microbiology, and 2Department of Psychiato', Kinsmen Laboratory of Neurological Research, University of British Columbia, Vancouver, B. C. (Canada) (Accepted 13 October 1987)
Key words: Methylmercury; Somatostatin; Glutamic acid decarboxylase (GAD); y-Aminobutyric acid (GABA); Cerebral cortex; Striatum; Postnatal development; Cerebral palsy
Tissue concentrations of the neuropeptide somatostatin and the specific activities of glutamic acid decarboxylase (GAD) were measured in several regions of the central nervous system in young rats, following chronic postnatal administration of methylmercuric chloride. By the beginning of the fourth postnatal week, these animals exhibited clinical signs of a mixed spastic/dyskinetic syndrome with visual deficits. At the onset of neurological impairment, a significant decrease in GAD activity was detected in the occipital cortex (48-49%) and striatum (45-50%) when compared to either normal or weight-matched controls. At one subclinical stage of toxicity, decreased GAD activity was detected only in the occipital cortex (29-30%). Tissue levels of somatostatin did not change significantly in the occipital cortex of methylmercury-treated animals at any stage of the experiment. However, somatostatin levels in the striatum were significantlyreduced at the onset of neurological impairment (55-57%) and at one subclinical stage of toxicity (49-54%). Immunohistochemistry for somatostatin- and neuropeptide Y-immunoreactive neurons confirmed a marked loss of cells in the dorsolateral region of the striatum with atrophy of the surviving neurons. In the cerebral cortex of methylmercury-treated animals the morphology and distribution of somatostatin-positive neurons appeared normal. In view of the reported co-localization of GAD and somatostatin in some non-pyramidal neurons of the cerebral cortex, these results indicate that methylmercury-induced lesions of the developing cerebral cortex involve a subpopulation of GABAergic neurons which are not co-localized with somatostatin. In the striatum, where GAD and somatostatin are not co-localized within the same neurons, methylmercury-induced lesions involve both GABAergic and somatostatin-positive neurons.
INTRODUCTION
motor signs included both spastic and flaccid paraly-
The toxicity of methylmercury and related alkylmercury compounds in the developing nervous system has been associated with a wide range of developmental disorders 4-6'~3. Following epidemic me-
sis, ataxia, athetosis, myoclonic jerking and generalized tonic convulsions. Visual disturbances varied from constriction of the visual field to total blindness. In animal models of methylmercury neurotoxicity during prenatal and early postnatal development,
thylmercury poisoning in Japan and Iraq, clinical studies of impairment infants and children have documented signs of neurological impairment resembling cerebral palsy syndrome 1"2'6,13"14,22. Clinical signs included severe psychomotor retardation, progressive microcephaly, persistence of primitive reflexes, hyperreflexia and hypersalivation. Additional
signs of neurological impairment have varied from subtle behavioral deficits 35'36 to the full spectrum of movement and postural disorders 29'3°. Recent studies have shown that the m o v e m e n t and postural disorders characteristic of methylmercuryinduced cerebral palsy syndrome can be produced in the developing rat following postnatal administration
Correspondence: J.R. O'Kusky, Kinsmen Laboratory of Neurological Research, 2255 Wesbrook Mall, Vancouver, B.C., V6T IW5, Canada. 0165-3806/88/$03.50© 1988 Elsevier Science Publishers B.V. (Biomedical Division)
12 of methylmercuric chloride 29"3°. At the onset of neurological impairment, lesions of the cerebral cortex were found to be more localized than those reported in human autopsy material 24. In the visual cortex, degenerating neurons were concentrated mainly in layer IV and the adjacent regions of layers III and V with a more diffuse distribution in the supra- and infragranular layers. Morphological features of degenerating neurons included a radial distribution of dendrites with few spines on degenerating dendritic profiles. A relatively selective degeneration of axon terminals forming symmetric contacts was observed in the cortical neuropil of all laminae. These morphological criteria suggest that aspinous or sparsely spinous stellate neurons, the inhibitory interneurons of the neocortex which utilize 7-aminobutyric acid (GABA) as a neurotransmitter, are most susceptible to methylmercury-induced lesions of the developing cerebral cortex. At the onset of neurological impairment, a significant decrease in the specific activity of glutamic acid decarboxylase (GAD), the GABA synthetic enzyme, was detected in both the cerebral cortex and striatum of methylmercury-treated animals 29. lmmunohistochemical studies of the localization of GAD have shown that GABAergic interneurons in the cerebral cortex comprise several morphological classes which are distributed throughout most cortical laminae 3~'34. Given the unusual laminar distribution of degenerating neurons in methylmercury-induced lesions of the cerebral cortex, it is possible that only particular subpopulations of GABAergic neurons are affected. In view of the reported co-localization of the neuropeptides somatostatin and neuropeptide Y with GAD in some interneurons of the cerebral cortex 16, it is possible that such co-localization defines such a subpopulation of GABAergic neurons. The present study was conducted to determine the extent to which somatostatin- and neuropeptide Yimmunoreactive neurons are involved in methylmercury-induced lesions of the developing cerebral cortex and striatum, The specific activity of GAD and tissue concentrations of somatostatin were determined in the cerebral cortex and striatum at the onset of neurological impairment and at two subclinical stages of toxicity. Immunohistochemistry for both somatostatin and neuropeptide Y was employed to in-
vestigate the pathology of peptidergic neurons during the course of methylmercury-induced movement and postural disorders. A histochemical method to detect reduced nicotinamide adenine dinucleotide phosphate (NADPH)-diaphorase activity was also employed, since this technique selectively stains striatal neurons which contain both somatostatin and neuropeptide y38. Preliminary results of these experiments have been published in abstract form 32. MATERIALS AND METHODS
Seven litters of Sprague-Dawley rats were culled to 9 pups per litter on postnatal day 2. Littermates of the same sex and similar body weight were organized into matched triplets. Individual members of these triplets were randomly assigned to one of 3 groups (methylmercury-treated, MeHg; weight-matched control, WMC; normal control, NC) and cross-fostered to separate dams. MeHg rats received subcutaneous injections of 0.01 M methylmercuric chloride in physiological saline (5 mg Hg/kg/day) beginning on postnatal day 5 and continuing until one of 3 stages of toxicity, defined as follows: stage I (day 18) when MeHg rats continued to gain weight although less rapidly than normal controls, stage II (days 20-22) when a given MeHg animal exhibited a persistent loss of body weight over 24 h, and stage III (days 22-25) at the onset of neurological impairment. Control animals were injected daily with an equivalent volume of saline. WMC animals were periodically isolated from their dam and placed in an incubator at 37 °C, maintaining their body weight within + 8% of MeHg animals. Nineteen matched triplets were employed to assay GAD activity and somatostatin immunoreactivity at the 3 stages of toxicity. The matched triplets were sacrificed by cervical fracture and the brains were immediately removed and dissected on ice. Tissue was sampled from the cerebral cortex (occipital region, including the primary visual cortex) and the striatum (caudate-putamen) at all 3 stages. From the 6 triplets at stage III, additional tissue was sampled from the thalamus, hypothalamus, hippocampus, substantia nigra, superior colliculus, cerebellum (vermal cortex, Iobules II-X) and spinal cord (cervical enlargement). For assays of the specific activity of GAD, each tis-
13 sue sample was homogenized in 10 vols. of cold 0.25 M sucrose. The specific activity of G A D was determined by measuring the UC02formed on incubation of 0.04 ml of tissue homogenate with L-[1-1aC]glutamic acid in the presence of excess pyridoxal phosphate .... . Samples were assayed in duplicate with the mean value expressed as either (1) ~mol CO 2 per g tissue wet weight per h or (2)/~mol CO 2 per mg protein per h. Protein was assayed according to Lowry et al. > with bovine serum albumin as standard. The tissue for somatostatin radioimmunoassay was weighed, boiled 10 rain in 1.0 ml of 1 N acetic acid and then sonicated. Samples were centrifuged at 4 °C for 20 rain at 10,000 rpm. The supernatant was lyophilized and stored at -70 °C until the day of assay. Somatostatin immunoreactivity was determined using a monoclonal antibody (Ab 3) 39 and Tyr-l-somatostatin labelled with 12sI via the chloramine T method. A standard curve was constructed with the antibody, labelled somatostatin and synthetic cyclic somatostatin (Peninsula Labs., Belmont, CA) in 0.4 ml of 24 mM sodium phosphate, 3.4 mM sodium acetate, 43.6 mM sodium chloride, 0.25 mM thimerosal, 0.5% bovine serum albumin and 500 KIU/ml aprotinin (pH 7.4). Tissue samples were diluted appropriately in assay buffer to obtain results within the sensitivity range of the standard curve (2.5-100 pg/mg). After a 3-day incubation at 4 °C, the free labelled somatostatin was separated with dextran-coated charcoal and counted in a gamma counter. Samples were assayed in duplicate with the mean value expressed as pg somatostatin immunoreactivity per mg tissue wet weight. The assay detects both somatostatin-14 and somatostatin-28 in equimolar amounts. The statistical significance of differences among the 3 groups at each stage of toxicity was determined using a 3 x 3 (stages x groups) analysis of variance 41. Two matched triplets at stages II and Ill were employed for immunohistochemistry (somatostatin, neuropeptide Y) and histochemistry (NADPH-diaphorase). Animals were anesthetized by an intraperitoneal injection of sodium pentobarbital (50 mg/kg) and perfused through the ascending aorta with a fixative solution containing 4% paraformaldehyde and 1 mM MgC12 in 0.1 M phosphate buffer (pH 7.4). The brains were removed, bisected in the midline and stored overnight in 0.1 M phosphate buffer containing 1 mM MgCI 2 and 15% sucrose. From the right
hemisphere of each brain, serial frozen sections (30 1~m) were cut in the frontal plane and organized into 4 series, with each series containing every fourth section through the cerebral hemisphere. Sections for somatostatin immunohistochemistry were processed using a monoclonal antibody (Ab 8) as previously described 39. Sections for neuropeptide Y immunohistochemistry were processed using a commercially available rabbit antiserum (RAS 7172, Peninsula Laboratories) as previously described 27. Sections for NADPH-diaphorase histochemistry were processed according to the indirect method of Vincent et al. 3s. Briefly, sections were incubated in 0.1 M Tris-C1 (pH 8.0) containing 1.0 mM NADP, 0.2 mM nitro blue tetrazolium, 15 mM sodium malate and 0.4% Tween80 at 37 °C for 60 rain. Sections were rinsed in phosphate-buffered saline (PBS) and coverslipped with glycerol-PBS. A fourth series of sections was stained for Nissl substance using 0.1% thionine in acetate buffer (pH 3.7). From the left hemisphere of each brain, blocks of tissue (0.5 mm thick) were sampled from the occipital cortex and striatum. These blocks were postfixed in 1% buffered osmium tetroxide, dehydrated in ascending grades of ethanol and embedded in Araldite. Semithin sections (0.5/2m) were cut and stained with 1% Toluidine blue in 0.4% sodium tetraborate. RESULTS The earliest sign of toxicity in MeHg animals was an abnormal rate of growth. The postnatal increase in body weight for MeHg animals was normal for the first 5 - 6 days of the experiment, followed by a period of decelerating growth where they continued to gain weight although less rapidly than NC animals. After attaining a maximum value between days 20 and 22, body weight gradually decreased, following within 2-4 days by the onset of neurological impairment. The average body weight for MeHg rats was significantly less than that of NC rats at all stages of toxicity (Table I). At no time during the experiment did the average body weight for MeHg rats differ significantly from that of WMC rats. At the time of sacrifice, the average brain weight for MeHg rats was also significantly less than that of NC rats at all stages (Table I). However, brain weights did not differ significantly between MeHg and WMC animals. It
14
days 22 and 25. These animals exhibited what appeared to be a mixed spastic/dyskinetic syndrome. All animals displayed a mild to marked hypertonia of flexor muscles in all 4 limbs, which was symmetrical and more pronounced in the hindlimbs. Flexion deformities were observed in a number of postures. For example, when NC and WMC animals were suspended by the tail, all 4 limbs were held in full extension. When MeHg animals were suspended by the tail, extension of the forelimbs was weak and incomplete while the hindlimbs were excessively adducted in full flexion. This adduction of the hindlimbs was similar to the phenomenon of hindlimb crossing reported in adult rats following exposure to methylmercury26. However, in these developing rats the hindlimbs were tightly flexed as opposed to the flaccid paralysis seen in adult animals. When MeHg animals were suspended by the scruff, the forelimbs were moderately flexed and abducted. When stimulated by manipulation of the hindlimbs in this position, extensor thrusts of the forelimbs were frequently observed with abnormal rotation of the limb and marked flexion of the digits. Hyperreflexia was observed in the auricular startle response and in the crossed extensor reflex. All MeHg animals exhibited some degree of visual impairment. The visual placing response was weak or absent, while the vibrissa placing and forelimb placing responses were present. Unlike control animals, they failed to avoid a visual cliff and
TABLE 1
Body weight and brain weightfor MeHg, WMC and NC animals at the three stages of toxicity Stage
Matched triplets, n =
1 6
Age of onset (days) Mean 18.0 Range 18
11
Ili
8
7
20.4 20- 22
23.8 22- 25
40.2+3.4* 30.4+1.8 31.8+2.6
50.6_+3.6* 32.3+2.8 33.7+3.2
68.2+3.9** 30.8+3.4 28.7+2.8
1.30+0.03" 1.23-+0.04 1.17+0.04
1.38_+0.04" 1.24_+0.06 1.19+0.05
1.51 _+ 0.04** 1.30_+0.06 1.25+0.05
Body weight (g) (mean +_ S.E.M. NC WMC MeHg Brain weight (g) (mean _+S~E.M. NC WMC MeHg
* P < 0.01, **P < 0.001, compared to MeHg animals.
would appear that the reduced brain weights in MeHg rats resulted from the associated malnutrition rather than from a direct neurotoxic effect of methylmercury. Clinical signs of neurological impairment in the 7 MeHg animals at stage III were observed between TABLE 1I
Tiss'ue concentrations of somatostatin and specific activities of GAD in the occipital cortex and striatum of MeHg, WMC and NC animals' at the three stages of toxicity ~Concentrations in pg/mg tissue wet weight as the mean + S.E.M. ; t'Enzyme activities in/~mol/g tissue/h as the mean _+ S.E.M.
Stage
Somatostatin"
GAD b
I
11
111
1
11
III
Matched triplets ~l-
6
7
6
6
7
6
Occipital cortex NC WM( McHg
113.9 ~ 12.3 128.6,+ 1 4 . 2 124.6± 7.9
96.2_+20.2 102.(1± 9.9 115.0,+ 15.4
182.1 ,+31.1 191.6+24.5 189.0_+26.3
12.1 + 0 . 7 9 . 2 + 1.1 9.3+0.8
15.4+0.8' 15.6+0.6" 10.9+ 1.1
16.9+0.7"* 17.3+0.9" 8 . 8 ± 1.2
146.9± 13.4 :+ 162.~_+_20.5 + 75.3~: 13.5
125.6+ 14.2 ~ 131.8±21.1 +: 56.4+_ t,~7
11.3 _+ 0.~; 10.2± 1.0 9,9-+ 1.1
13.2_+0.5 14.0_+0.7
14.3_+ 1.0'* 15.6+_0.4 :::
12.0_+ 1.1
7.8-+ 1.1
SlrJattml ~,{ ~M( ~lelle
7 7 . 9 + 14.! 6 1 . 6 + 4.~ 75.1~ ~ (,.I
i' < (> i)l, / ' < {)t)0l, compared to Mcflg animals.
15 demonstrated difficulty in negotiating obstacles in a novel environment. Myoclonic jerking of the hindlimbs was observed in 2 MeHg animals immediately prior to sacrifice.
nificant differences were detected among MeHg and control animals in the specific activity of G A D , tissue concentrations of somatostatin or concentrations of total protein.
GAD activity/somatostatin immunoreactivity Postnatal changes in the specific activity of G A D are presented in Table II. Treatment with methylmercury resulted in significant decreases in G A D activity in the cerebral cortex at stage II (29-30%) and stage III (48-49%). In the striatum a significant decrease was detected in MeHg animals only at stage III (45-50%). The differences between NC and WMC animals were not significant at any stage, indicating that the degree of malnutrition occurring during the experiment was not sufficient to alter the normal postnatal changes in G A D activity. An analysis of G A D kinetics with respect to glutamic acid was performed using combined tissue homogenates from the cerebral cortex and striatum at stage III of toxicity. No significant differences in the affinity for substrate were detected among MeHg, WMC and NC animals, with values of K m ranging from 0.9-1.2 x 10 -3 M. The value of Vmax for MeHg animals was found to be only 69% of the value for control animals. Concentrations of total protein in the cerebral cortex and striatum did not differ significantly among MeHg, WMC and NC animals at any stage of the experiment. Consequently, expressing the specific activity of G A D as/~mol CO2/mg protein/h did not alter the experimental findings. The results of radioimmunoassay for tissue concentrations of somatostatin in the cerebral cortex and striatum are presented in Table II. Tissue concentrations of somatostatin in the cerebral cortex did not differ significantly among MeHg, WMC and NC animals at any stage of the experiment. In the striatum a significant decrease in somatostatin levels was detected in MeHg animals at stage II (49-54%) and stage III (55-57%) when compared to either WMC or NC animals. Assays for G A D activity and somatostatin immunoreactivity were performed in tissue samples from several additional regions of the central nervous system at the onset of neurological impairment (stage III). In all regions examined, including the thalamus, hypothalamus, hippocampus, substantia nigra, superior colliculus, cerebellum and spinal cord, no sig-
Histopathology in the cerebral cortex and striatum Somatostatin immunohistochemistry in the cerebral cortex confirmed the results of radioimmunoassay. The morphology and distribution of somatostatin-immunoreactive neurons in the cerebral cortex of MeHg animals appeared to be normal (Fig. 1). Neurons were distributed throughout layers I I - V I and were occassionally observed in the subcortical white matter. In layer IV the cell bodies were somewhat smaller and less intensely stained than neurons in the supra- and infragranular layers. Immunoreactive dendritic processes exhibited no systematic pattern of branching and examples of multipolar, bipolar and bitufted neurons could be identified. In MeHg animals the length and branching pattern of dendritic processes appeared normal. Neuropeptide Y-immunoreactive neurons in the cerebral cortex were similar in size, morphology and distribution to somatostatin-immunoreactive neurons, although fewer in number. In MeHg animals these neurons appeared normal. Cortical pathology was evident on Toluidine bluestained sections of the cerebral cortex at both stage II and stage III. Degenerating neurons were observed in concentrated mainly in layer IV with a more diffuse distribution in layers II, III, V and VI. The cell bodies of degenerating neurons exhibited an increased staining of the cytoplasm with an accumulation of clear vacuoles. Their nuclei were condensed and pyknotic. This pattern of pathology was similar to that reported in an earlier study 3°. Somatostatin immunohistochemistry of the striaturn revealed extensive pathology in MeHg animals (Fig. 2). There was a marked loss of neurons in the dorsolateral half of the caudate-putamen (Fig. 2A,B). The loss of somatostatin-immunoreactive neurons along the dorsolateral margin of the nucleus was virtually complete. The cell bodies of surviving neurons in the ventromedial region were smaller with a reduction in the number and length of immunoreactive dendritic processes (Fig. 2C,D). The moderately dense network of somatostatin-immunoreactive fibers, normally seen throughout the nucleus in NC
1~
and WMC animals, was markedly reduced in MeHg animals: Neuropeptide Y immunohistochemistry revealed the same pattern of pathology (Fig. 3) with a loss of neurons in the dorsolateral region of the striatum and atrophy of surviving neurons. The dense network of neuropeptide Y-immunoreactive fibers seen in control animals was virtually absent in the MeHg animals (Fig. 3C,D). NADPH-diaphorase histochemistry confirmed the loss of neuropeptideimmunoreactive neurons in the striatum (Fig. 4). Surviving NADPH-diaphorase-positive neurons were atrophic with a reduction in the number and
length of labeled dendritic processes. Labeled fibers were distinctly beaded in appearance. The presence of degenerating neurons in the striatum of MeHg animals was determined on Toluidine blue-stained sections (Fig. 5). At subclinical stage II (Fig. 5A) the nuclei of degenerating neurons were condensed and pyknotic. Increased intensity of staining was evident in the cytoplasm, which contained numerous clear vacuoles. Degenerating neurons were scattered along the dorsolateral margin of the caudate-putamen, representing less than 2% of the total number of neuronal profiles. Many adjacent
200
.~:ig. 1. Somatostatin-immunoreactive neurons in the primary visual cortex of (A) weight-matched control and (B) MeHg-treated anili~.qs at lhc onset of neurological impairment (stage III). Calibration bar in iml.
b
8
~J
3 ~
p~
~
r~
3
i ~
~
0
~Z
E~
2{}
Fig. 5. Aratdite-embedded sections (0.5 ~m, stained with Toluidine blue) from the striatum of MeHg-treated animals at (A) stage II and (B) stage III. Degenerating neurons were pyknotic with an accumulation of clear vacuoles in the cytoplasm. At stage II, isolated degenerating neurons were observed among neurons with apparently normal morphology (n). Calibration bar in/~m. neurons and the surrounding neuropil a p p e a r e d normal. At the onset of neurological i m p a i r m e n t (Fig. 5B), numerous degenerating neurons were o b s e r v e d in the dorsolateral region of the striatum. A l t h o u g h the m o r p h o l o g y of these neurons was similar to that seen at stage II, the surrounding neuropil had a spongy a p p e a r a n c e resulting from m a r k e d astrocyte cuffing and swelling of dendritic processes.
DISCUSSION The results of the present study d e m o n s t r a t e that a specific subpopulation of G A B A e r g i c interneurons is involved in m e t h y l m e r c u r y - i n d u c e d lesions of the developing cerebral cortex. Many neurons in the cerebral corlex a p p e a r to use G A B A as a neurotransmittcr ~ Neurons displaying G A D immunoreactivity have been identified in all layers of the rat visual corto,: q ~. l'hcse neurons arc non-pyramidal in their morphology and have been divided into several types ~ccnrding to the size and laminar distribution of their
cell bodies and the m o r p h o l o g y and sites of termination of their axon terminals. Some G A D - i m m u n o reactive neurons in the cerebral cortex of the cat and monkey have been shown to contain the n e u r o p e p tides somatostatin and n e u r o p e p t i d e y16. All somatostatin- and n e u r o p e p t i d e Y-immunoreactive neurons in the cerebral cortex of the cat and most of these neurons in the cerebral cortex of the m o n k e y are also G A D - i m m u n o r e a c t i v e . Although they occur in all cortical laminae, they have a preferential distribution in layers II, I l i and VI of the cerebral cortex. Like other cortical interneurons they are small aspiny non-pyramidal neurons whose axons form predominantly symmetric synapses, although occasional asymmetric p e p t i d e - i m m u n o r e a c t i v e contacts have been r e p o r t e d jr. Some G A D - i m m u n o r e a c t i v e neurons do not a p p e a r to contain these neuropeptides. These neurons occur in all laminae, but they have a preferential distribution in layers IV and V. In the present study the neurotoxicity of methylmercury resulted in a significant decrease in the specific activity t}f G A D in the cerebral cortex at the on-
21 set of neurological impairment (stage III) and at subclinical stage II. Cortical concentrations of somatostatin did not change during any stage of the experiment. In the cerebral cortex of MeHg animals, degenerating neurons were observed concentrated mainly in layer IV and the adjacent regions of layers III and V with a more diffuse distribution in the supra- and infragranular layers. This laminar distribution of degenerating neurons agrees with a previous study, where degenerating axons forming predominantly symmetric synapses were observed in all cortical laminae of MeHg animal 3°. These morphological criteria are consistent with a selective degeneration of GABAergic interneurons. The normal morphology and distribution of somatostatin- and neuropeptide Y-immunoreactive neurons in the cerebral cortex of MeHg animals indicates that these neurons were spared in methylmercury-induced lesions. Thus, the neurotoxicity of methylmercury in the developing cerebral cortex appears to result in a relatively selective degeneration of those GABAergic neurons which are not co-localized with somatostatin and neuropeptide Y. This selective pathology would explain the unusual laminar distribution of degenerating GABAergic neurons. GAD-immunoreactive neurons which are not co-localized with somatostatin and neuropeptide Y are preferentially distributed in laminae IV and V with a more diffuse distribution in the supra- and infragranular layers 16. One morphologically distinct class of GAD-immunoreactive neuron in the cerebral cortex, the basket cell, appears to be among those GABAergic intemeurons which are not co-localized with somatosatin or neuropeptide y16. Basket cells have a large cell body and are distributed mainly in layers II1-V118. Their axons form pericellular nests around the cell bodies and proximal dendrites of pyramidal neurons in the deep cortical layers 12'19'21. In a previous ultrastructural study of synaptic degeneration in methylmercury-induced lesions 3°, it was found that pyramidal neurons in layers III, V and VI had multiple degenerating terminals forming symmetric synapses on their somata. These morphological criteria, combined with the neurochemical results of the present study, suggest that cortical basket cells comprise one type of GABAergic interneuron involved in methylmercury-induced lesions of the cerebral cortex.
An unexpected finding of the present study was the significant decrease in concentrations of somatostatin in the striatum at the onset of neurological impairment (stage III) and at subclinical stage II. This decrease is specific to the striatum as tissue levels of somatostatin were normal in the other 8 regions of the central nervous system examined. Decreased somatostatin levels appear to result from a loss of somatostatin-immunoreactive neurons, particularly in the dorsolateral margin of the nucleus. Similar cell losses were noted using neuropeptide Y immunohistochemistry and NADPH-diaphorase histochemistry. Given that these 3 substances have been shown to be co-localized in the same neurons in the stria° tum 3s, there would appear to be a loss of neurons rather than an inhibited synthesis of somatostatin. The presence of degenerating neurons in the striatum at stages II and III, as seen on Toluidine blue-stained sections, confirmed a neuron loss in the appropriate region of the nucleus. In the striatum of the rat, somatostatin-immunoreactive neurons comprise a relatively sparse and homogeneous population of medium-sized aspiny neurons 37'4°. These are striatal interneurons whose axons are distributed within the caudate-putamen and form symmetric synapses mainly on dendritic shafts 9'37. Although their function is largely unknown, there is preliminary evidence to suggest that somatostatin-immunoreactive neurons in the striatum have some influence on the extrapyramidal control of movement. In experimental animals, treatment with somatostatin modulates the release of dopamine 7 and alters levels of spontaneous motor activity 3'x5'33. Decreased levels of somatostatin have been detected in the cerebrospinal fluid of patients with Parkinson's disease l° and Huntington's chorea s. Autopsy material from the neostriatum of patients with Huntington's chorea has been shown to contain both altered tissue concentrations of somatostatin 2s and altered morphology of somatostatin-immunoreactive neurons and fiber networks 23. In the present study the relatively early loss of somatostatin-immunoreactive neurons in the striatum suggests that these neurons may be involved in the pathogenesis of pediatric movement and postural disorders. The results of the present study indicate that discrete populations of interneurons in the cerebral cortex and striatum are selectively impaired in methyl-
m e r c u r y - i n d u c e d lesions of the d e v e l o p i n g central
genesis of m e t h y l m e r c u r y - i n d u c e d
n e r v o u s system. In t h e c e r e b r a l c o r t e x G A B A e r g t c i n t e r n e u r o n s , which are p r o b a b l y not c o - l o c a l i z e d
postural d i s o r d e r s d u r i n g d e v e l o p m e n t .
movement
w~th the n e u r o p e p t i d e s o m a t o s t a t i n , d e g e n e r a t e at a
ACKNOWLEDGEMENTS
and
subclinical stage o f toxicity. In the s t r i a t u m i n t e r n e u rons which are s o m a t o s t a t i n - and n e u r o p e p t i d e Y-
This r e s e a r c h was s u p p o r t e d by grants f r o m the
i m m u n o r e a c t i v e d e g e n e r a t e at this subclinical stage.
U n i t e d C e r e b r a l Palsy R e s e a r c h a n d E d u c a t i o n a l
A t the o n s e t of n e u r o l o g i c a l i m p a i r m e n t , d e g e n e r a t ing striatal n e u r o n s a p p e a r to include b o t h G A B A -
F o u n d a t i o n , Inc. and the M e d i c a l R e s e a r c h C o u n c i l of C a n a d a . J . R . O . and S . R . V . are R e s e a r c h Schol-
ergic and p e p t i d e r g i c n e u r o n s . F u r t h e r r e s e a r c h is
ars of the M R C . W e t h a n k D r . J . C . B r o w n , M R C
n e e d e d to d o c u m e n t the m e t a b o l i c c h a n g e s in these
Regulatory
neurotransmitter-specific
C o l u m b i a , for the m o n o c l o n a l a n t i b o d i e s to s o m a t o -
neurons,
which
precede
Peptide
Group,
U n i v e r s i t y of British
n e u r o n a l d e g e n e r a t i o n and c o n t r i b u t e to the p a t h o -
statin.
REFERENCES
13 Harada, H., Methyl mercury poisoning due to environmental contamination (Minamata disease). In F.W. Oehme (Ed.), Toxicity of Heavy Metals in the Environment, Marcel Dekker, New York, 1978, pp. 261-302. 14 Harada, Y., Congenital (or fetal) Minamata disease. In M. Katsuma (Ed.), Minamata Disease, Kumamoto University, Kumamoto, Japan, 1968, pp. 93-117. 15 Havlicek, V., Rezek, M. and Friesen, H., Somatostatin and thyrotropin-releasing hormone: central effect on sleep and motor system, Pharmacol. Biochem. Behav., 4 (1976) 455-459. 16 Hendry, S.H.C., Jones, E.G., DeFelipe, J., Schmechel, D., Brandon, C. and Emson, P.C., Neuropeptide-containing neurons of the cerebral cortex are also GABAergic, Proc. Natl. Acad. Sci. U.S.A., 81 (1984) 6526-6530. 17 Hendry, S.H.C., Jones, E.G. and Emson, P.C., Morphology, distribution, and synaptic relations of somatostatin- and neuropeptide Y-immunoreactive neurons in rat and monkey neocortex, J. Neurosci., 4 (1984) 2497-2517. 18 Houser, C.R., Hendry, S.H.C., Jones, E.G. and Vaughn, J.E., Morphological diversity of immunocytochemically identified GABA neurons in the monkey sensory-motor cortex, J. Neurocytol., 12 (1983) 617-638. 19 Jones, E.G., Varieties and distribution of non-pyramidal cells in the somatic sensory cortex of the squirrel monkey, J. Comp. Neurol., 160 (1975) 205-268. 20 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, E.J., Protein measurement with the Folin phenol reagent, J. Biol. Chem., 193 (1951) 265-275. 21 Marin-Padilla, M., Origin of the pericellular baskets of the pyramidal cells of the human motor cortex: a Golgi study, Brain Res., 14 (1969) 633-646. 22 Marsh, D.O., Myers, G.J., Clarkson, T.W., Amin-Zaki, L., Tikriti. S. and Majeed, M.A., Fetal methylmercury poisoning: clinical and toxicological data on 29 cases, Ann. Neurol., 7 (1980) 348-353. 23 Marshall, P.E. and Landis, D.M.D., Huntington's disease is accompanied by changes in the distribution of somatostatin-containing neuronal processes, Brain Res., 329 (1985) 71-82. 24 Matsumoto. H., Koya, G. and Takeuchi, T.. Fetal Minamata disease, J. Neuropathol. Exp. Neurol., 24 (1965) 563- 574.
1 Amin-Zaki, L.. Elhassani, S., Majeed, M.A., Clarkson,
2
3
4
5 6
7
8
9
10
li
T.W., Doherty, R.A. and Greenwood, M., Intrauterine methylmercury poisoning in Iraq, Pediatrics, 54 (1974) 587--595. Amin-Zaki, k., Majeed, M.A., Elhassani, S.B., Clarkson, T.W., Greenwood, M.R. and Doherty, R.A., Prenatal methylmercury poisoning, Am. J. Dis. Child.. 133 (1979) 172- 177. Brown, M. and Vale, W., Central nervous system effects of hypothalamic peptides. Endocrinology, 96 (1975) 1333-1336. Cassidy. D.R., Toxicity of inorganic and organic mercury compounds in animals. In F.W. Oehme (Ed.), Toxicity of Heavy Metal~ in the Environment, Marcel Dekker, New York, 1978, pp. 303-330. Chang, L.W., Neurotoxic effects of mercury - - a review, Emiron. Res., 14 (1977) 329-373. ('hang, L.W. and Annau, Z., Developmental neuropathoIogy and behavioral teratology of methylmercury. In J. Yanai (Ed.), Neurobehavioral Teratology, Elsevier, New York, 1984, pp. 405-432. Chesselet. M.F, and Reisine, T.D., Somatostatin regulates dopamine release in rat striatal slices and cat caudate nuclei, J. Neurosci.. 3 (1983) 232-236. Cramcr. H.J., Kohler, G., Oepen, G., Schomburg, G. and Schratcr. E.. Huntington's chorea. Measurements of somatostatin, substance P and cyclic nucleotides in the cerebrospinal fluid, J. Neurol., 225 (1981) 183-187. DiFiglia, M. and Aronin, N.. Quantitative electron microscopic study of immunoreactive somatostatin axons in the ra~t neostriatum, Neurosci. Lett., 50 (1984) 325-331. Dupont, E., Christensen, S.E., Hansen, A.P., Olivarius, B.F. and Orskov, H., Low cerebrospinal fluid somatostatin m Parkinson's disease: an irreversible abnormality, Neurol~g), 32 (1982) 312-314. Emsan, P.C. and Lindvall, O., Distribution of putative ~etm~transmitters in the neocortex, Neuroscience, 4 (1979)
12 Fairctt, A., 1)cFelipc, J. and Regidor, J.. Non-pyramidal llcur~ms In A. Peters and E.G. Jones (Eds.). Cerebral Cor~,'~, Plenum, New York, 1984. pp. 201-253.
23 25 McGeer, P.L., McGeer, E.G., Wada, J.A. and Jung, E., Effects of globus pallidus lesions and Parkinson's disease on brain glutamic acid decarboxylase, Brain Res., 32 (1971) 425-431. 26 Miyakawa, T. and Deshimaru, M., Electron microscopical study of experimentally induced poisoning due to organic mercury compound, Acta Neuropathol.. 14 (1969) 126-133. 27 Nakamura, S. and Vincent, S.R., Somatostatin- and neuropeptide Y-immunoreactive neurons in the neocortex in senile dementia of Alzheimer's type, Brain Res., 370 (1986)
34
35
36
11-20.
28 Nemeroff, C.B., Youngblood, W.W., Manberg, P.J., Prange, A.J. and Kizer, J.S., Regional brain concentrations of neuropeptides in Huntington's chorea and schizophrenia, Science, 221 (1983) 972-975. 29 O'Kusky, J.R. and McGeer, E.G., Methylmercury poisoning of the developing nervous system in the rat: decreased activity of glutamic acid decarboxylase in cerebral cortex and neostriatum, Dev. Brain Res., 21 (1985) 299-306. 30 O'Kusky, J., Synaptic degeneration in rat visual cortex following neonatal administration of methylmercury, Exp. Neurol., 89 (1985) 32-47. 31 Peters, A., Proskauer, C.C. and Ribak, C.E., Chandelier cells in rat visual cortex, J. Comp. Neurol., 206 (1982) 397-416. 32 Radke, J.M., O'Kusky, J.R. and Vincent, S.R., Methylmercury poisoning of the developing nervous system: effects on somatostatin immunoreactivity in cerebral cortex and striatum, Soc. Neurosci. Abstr., 12 (1986) 1362. 33 Rezek, M., Havlicek, V., Leybin, L., Pinsky, C., Kroeger, E.A., Hughes, K.R. and Friesen, H., Neostriatal administration of somatostatin: differential effect of small and large
37
38
39
40
41
doses on behavior and motor control, Can. J. Physiol. Pharmacol., 55 (1977) 234-242. Ribak, C.E., Aspinous and sparsely-spinous stellate neurons in the visual cortex of rats contain glutamic acid decarboxylase, J. Neurocytol., 7 (1978) 461-478. Rice, D.C. and Gilbert, S.G., Early chronic low-level methylmercury poisoning in monkeys impairs spatial vision, Science, 216 (1982) 759-761. Spyker, J.M., Sparber, S.B. and Goldberg, A.M., Subtle consequences of methylmercury exposure: behavioral deviations in offspring of treated mothers, Science, 177 (1972) 621-623. Takagi, H., Somogyi, P., Somogyi, J. and Smith, A.D., Fine structural studies on a type of somatostatin-immunoreactive neuron and its synaptic connections in the rat neostriatum: a correlated light and electron microscopic study, J. Comp. Neurol., 214 (1983) 1-16. Vincent, S.R., Johansson, O., H6kfelt, T., Skirboll, L., Elde, R.P., Terenius, L., Kimmel, J. and Goldstein, M., NADPH-diaphorase: a selective histochemical marker for striatal neurons containing both somatostatin- and avian pancreatic polypeptide (APP)-like immunoreactivities, J. Comp. Neurol., 217 (1983) 252-263. Vincent, S.R., Mclntosh, C.H.S., Buchan, A.M.J. and Brown, J.C., Central somatostatin systems revealed with monoclonal antibodies, J. Comp. Neurol., 238 (1985) 169-186. Vincent, S.R., Staines, W.A. and Fibiger, H.C., Histochemical demonstration of separate populations of somatostatin and cholinergic neurons in the rat striatum, Neurosci. Lett., 35 (1983) 111-114. Winer, B.J., Statistical Principles in Experimental Design, McGraw-Hill, New York, 1971, pp. 591-603.