Selective increase in somatostatin mRNA expression in human basal ganglia in Parkinson's disease

Molecular Brain Research 50 Ž1997. 59–70

Research report

Selective increase in somatostatin mRNA expression in human basal ganglia in Parkinson’s disease David J. Eve a , Angus P. Nisbet a , Ann E. Kingsbury a , James Temlett b, C. David Marsden c , Oliver J.F. Foster d,) a

Parkinson’s Disease Society Brain Research Centre (Brain Bank), 1 Wakefield Street, London WC1N 1PJ, UK Department of Medicine, UniÕersity of the Witwatersrand Medical School, 7 York Road, Parktown 2193, South Africa Institute of Neurology, Parkinson’s Disease Society Brain Research Centre (Brain Bank), 1 Wakefield Street, London WC1N 1PJ, UK d Parkinson’s Disease Society Brain Research Centre (Brain Bank), St. George’s Hospital Medical School, London, UK b

c

Accepted 22 April 1997

Abstract Levels of the neurotransmitter somatostatin ŽSS. have previously been shown to be reduced in the cortex and hippocampus of demented parkinsonian patients and patients with Alzheimer’s disease. In situ hybridisation histochemistry ŽISHH. was performed with an 35 S tail-labelled oligonucleotide DNA probe to human SS mRNA, to examine its expression within the striatum, medial medullary lamina ŽMML. and reticular thalamic nucleus in Parkinson’s disease ŽPD. and in matched controls. A chronic unilaterally MPTP-lesioned L-DOPA-naive primate model was also examined for comparison of SS mRNA expression with that in human L-DOPA treated PD subjects. Quantitation of SS mRNA expression on emulsion dipped sections revealed a significant increase Ž82%. in the MML of the globus pallidus in PD Ž56.5 m m2 of silver grainrcell, n s 9 cases. compared to controls Ž26.3 m m2rcell, n s 13 cases, p - 0.01, Student’s t-test., paralleling the increase previously observed by this group for NOS mRNA. SS mRNA expression was higher in the dorsolateral than ventromedial putamen in controls Ž p - 0.001; DL: 24.89 " SEM 1.35; VM: 17.96 " SEM 2.63; n s 14. but this gradient was lost in PD cases Ž p ) 0.05; DL: 22.68 " 1.94; VM: 22.17 " 2.94; n s 10.. These findings suggest specific modification of basal ganglia SS-ergic pathways in PD. q 1997 Elsevier Science B.V. Keywords: Somatostatin; Basal ganglia; Hybridization, in situ; Parkinson’s disease; MPTP Ž1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine.; Medial medullary lamina

1. Introduction Somatostatin ŽSS. is a widely-expressed tetradecapeptide that has been implicated as a regulator of endocrine, gut and brain function. It is synthesized from its precursor molecule preprosomatostatin and exists in several forms: SS-14, SS-25 and SS-28, which feature a common sequence of amino acids. In the adult mammalian brain, SS satisfies a number of criteria for definition as a neurotransmitter w7x. SS is abundant within the hypothalamus from which it is secreted into the local circulation, and within all layers of the cerebral cortex, the striatum, and the substantia nigra pars compacta w25x. A role for cortical and hippocampal SS in cognitive function has been postulated ) Corresponding author. Parkinson’s Disease Society Brain Research Centre ŽBrain Bank., 1 Wakefield Street, London WC1N 1PJ, UK. Fax: q44 Ž71. 278-4993.

since the levels of both SS protein and mRNA are reduced in these regions in individuals suffering from cognitive deterioration, such as demented parkinsonian patients, and those with Alzheimer’s disease w10,12x. Immunocytochemical and in situ hybridization studies have shown SS to be present within neurons of the caudate putamen, hippocampus and cerebral cortex w6,25x. The majority of such SS-positive neurons are interneuronal and co-express the neurotransmitters neuropeptide Y ŽNPY. and nitric oxide ŽNO. w19,35,39x. Within the hippocampus and cerebral cortex, there is also co-localisation of these neurotransmitters within GABA-ergic interneurons w34x. Within the striatum, some SS-ergic neurons are located on the borders of striosomes, and it has been suggested that SS-ergic interneurons may allow communication between the striosomes and matrix w14,31,32x. The presence of SS within the striatum raises the possibility that SS might influence the nigrostriatal system. This is supported by a

60

D.J. EÕe et al.r Molecular Brain Research 50 (1997) 59–70

number of studies demonstrating a relationship between SS and dopamine ŽDA. expression. Intracerebroventricular injection of SS has been shown to induce increased DA synthesis and turnover in several brain regions w13x, and to regulate DA-ergic release from striatal slices w4,7x. Beal et al. w5x subsequently demonstrated this to be a specific effect of SS, which is additive with respect to the co-localised neurotransmitter, NPY. A reciprocal influence of DA on SS expression has also been demonstrated: both D1 and D2 receptor blockade has been shown to significantly decrease SS content within these interneurons w3x. Nigral innervation of the striatum in normal controls is heterogeneous with a larger number of DA-ergic terminals in the dorsolateral compared with the ventromedial putamen. In Parkinson’s disease ŽPD., the striatal DA loss, resulting from depletion of nigrostriatal neurons, is nonuniform, with a ventromedial-dorsolateral gradient Žthe latter showing the greatest degree of loss. w18x. If DA regulates SS expression in man, one might therefore expect differential expression of SS mRNA in different striatal regions in PD, as has been demonstrated for preproenkephalin ŽPPE. mRNA expression w28x. In order to explore the role of SS in basal ganglia function and its relationship with DA transmission in PD, quantitative in situ hybridisation histochemistry ŽISHH. was used to measure SS mRNA expression in putaminal subregions of human post-mortem material from PD cases and neurologically normal controls. We have recently demonstrated that nitric oxide synthase ŽNOS. mRNA expression Žthe synthetic enzyme for putative neurotransmitter NO. per cell within the medial medullary lamina ŽMML, situated between the external and internal globus pallidus., is significantly increased in patients with Parkinson’s disease compared with controls w29x. NOS and SS may be co-localised within these cells and we therefore investigated whether SS mRNA expression in the MML might show parallel upregulation in PD cases using ISHH. Previous human studies by Mengod et al. w25x demonstrated SS mRNA expression within the external medullary lamina of the thalamus ŽEML.rreticular thalamic nucleus ŽRTN., and this region was also examined. The MPTP-treated primate model of PD demonstrates nigral cell loss and similar behavioural characteristics to the human disease. Three unilaterally-MPTP-treated LDOPA-naive vervets were made available for parallel studies in this animal model, as a control for the effect of therapeutically administered L-DOPA in the PD cases studied.

2. Materials and methods Post-mortem details of the subjects examined are shown in Table 1. A total of 9 PD cases and 13 controls were studied with respect to the medial medullary lamina, 10 PD and 14 controls with respect to the putamen, 9 PD and

Table 1 Characteristics of material Case no.

Sex

Controls 1 F 2 F 3 M 4 F 5 F 6 M 7 M 8 M 9 F 10 M 11 M 12 M 13 M 14 M PD cases 1 M 2 M 3 M 4 M 5 F 6 F 7 F 8 M 9 F 10 F

Age Žyr.

PM delay Žh.

Cause of death a

Brain pH

69 85 80 89 90 76 63 67 77 82 40 78 91 76

34 48 50 30 30 5.5 26 22 18 33 13 19 48 16

Ruptured aortic aneurysm Ischaemic heart disease Ischaemic heart disease Pulmonary embolism Myocardial infarction Left ventricular failure Myocardial infarction Chronic obstructive airways Ruptured aortic aneurysm Pulmonary embolism Carcinoma of the lung Left ventricular failure Ruptured aortic aneurysm Ruptured thoracic aneurysm

6.12 6.12 6.21 6.23 6.24 6.51 6.53 6.55 6.7 6.79 6.59 6.7 6.36 6.71

70 81 75 67 82 81 75 88 78 72

24 30 9.25 20 23.5 15 13 8.5 24.5 16

Bronchopneumonia ‘Sudden death’ Bronchopneumonia Bronchopneumonia ‘Sudden death’ Bronchopneumonia ‘Advanced PD’ ‘Sudden death’ ‘Advanced PD’ Bronchopneumonia

6.2 6.3 6.3 6.35 6.4 6.53 6.56 6.56 6.57 6.6

PD cases: Mean age 76.9 yr, range 67–88; mean PM delay 18.4 h, range 8.5–30; mean cerebellar pH 6.44 SEM 0.05, range 6.20–6.60. Controls: Mean age 75.9 yr, range 40–91; mean PM delay 28.0 h, range 5.5–50; mean cerebellar pH 6.45 SEM 0.06, range 6.12–6.79. a Cause of death as recorded on Death Certificate.

12 controls with respect to the caudate and 8 PD and 9 controls with respect to the RTN. Clinical data including drug histories, as shown in Table 2, were obtained for all cases and no patients with a recent history of neuroleptic use were included in the study. All parkinsonian patients were treated with L-DOPA up to the time of death Žmean dosage 1122 mgr24 h.. 2.1. Tissue preparation All cadavers had been refrigerated below 48C within 4 h of death. Brains were removed and flash-frozen with a post-mortem delay of less than 48 h, as follows: whole brains were divided sagittally, one half was cut into 1 cm thick coronal slices which were placed on brass blocks pre-cooled to y808C to ensure rapid freezing while preserving anatomical orientation; the other half was fixed in 10% formalin for 8 weeks prior to cutting, processing and paraffin wax-embedding of representative tissue blocks for neuropathological evaluation. 12 m m coronal sections were cryostat cut at y158C from frozen striatal blocks at a level just posterior to the mamillary body, providing a cross-section of the medial and lateral segments of the globus

D.J. EÕe et al.r Molecular Brain Research 50 (1997) 59–70

61

Table 2 Clinical details of Parkinson’s disease cases PD case no.

Disease duration Žyr.

L-DOPA

dose Žmgr24 h.

Dopamine agonist Žmgr24 h.

Selegiline Žmgr24 h.

Anti-cholinergic Žmgr24 h.

Other CNS drugs Žmgr24 h.

1 2 3 4 5 6 7 8 9 10

17 5 11 22 13 16 13 20 19 25

1000 400 700 3500 1200 400 1500 400 1000 Unknown

– – – – – Lisuride; 0.4 – – – Bromocriptine; 2.5

– 5 10 – – 10 5 – 10 –

– – – – Benzhexol; 6 – Benzhexol; 6 Benzhexol; 16 – Benztropine; 1

– – – – – – – – – Amantadine; 200

Mean disease duration 16.1 yr, range 5–25; mean L-DOPA dose at death 1122 mgr24 h, range 400–3500 mg ŽL-DOPA was always given with a peripheral decarboxylase inhibitor..

pallidus and a mid-caudal cross-section of the striatum where severe dopamine depletion would be expected in PD material. Sections were stored at y808C prior to hybridization. 2.2. Pathological confirmation of PD PD was confirmed by neuropathological examination of haematoxylin and eosin ŽH & E., luxol fast blue, Nissl and silver stained sections from striatum, brain stem, cerebellum and cerebral cortex. All cases had marked macroscopic depigmentation of the substantia nigra and locus coeruleus with severe cell loss and frequent Lewy bodies visible on H & E staining. There was no evidence of striatal involvement. Additional neuropathology was excluded and control brains were screened to rule out previously unsuspected neurological disease. 2.3. Tissue matching PD and control cases were matched for sex, age and post-mortem brain pH which our group has previously shown to be strongly correlated with preservation of a variety of mRNA species w17x. Post-mortem delay of less than 48 h, which does not appear to be an important factor in mRNA preservation w17x was less well matched Žbut not significantly different. between the 2 groups Žcontrols 28.0 h " SEM 3.7; PD 18.4 h " SEM 2.3.. Estimation of cerebellar pH was performed as described by Kingsbury et al. w17x. All cerebellar pH values were between 6.12 and 6.79. There was no significant difference between the pH of control Žmean 6.45 " SEM 0.06. and PD cases Žmean 6.44 " SEM 0.05..

male vervets were killed by lethal barbiturate anaesthesia, at least a year after application of a left intracarotid injection of 0.6–0.8 mgrkg MPTP. During this period the vervets exhibited unilateral parkinsonian characteristics. Following death, the brains were cut into 1 cm thick coronal slices, flash-frozen on dry ice and stored at y808C until required. Post-mortem brainstem examination revealed extensive DA-ergic cell loss within the substantia nigra of the lesioned side. This was investigated by tyrosine hydroxylase immunocytochemistry and found to correspond with 80–90% loss of nigral neurons. The caudate and putamen were examined in these vervets with the right, unlesioned side acting as an internal control. 2.5. Probe and probe labelling A 30-mer oligonucleotide DNA probe, complementary to bases 2413–2442 of human somatostatin mRNA w36x was 3X tail-labelled using 35 S-dATP ŽNEN DuPont. by incubating the probe Žconcentration 0.1 pmolrm l. for 1 h at 378C with terminal deoxynucleotide transferase ŽPromega Corporation. at a concentration of 500–1000 urml in a cobalt-containing cacodylate buffer ŽPromega TDT tailing buffer.. The probe sequence was checked using the NIH Basic Local Alignment Search Tool ŽBLAST. program and was found to exhibit an 8 base mismatch compared with the sequence for the structurally similar neuropeptide, cortistatin. The sequence chosen is complementary for both the SS-14 and SS-28 forms of somatostatin. The concentration of 35 S-dATP in the reaction mixture was 10 = that of the probe Ž1 pmolrm l.. Separation of labelled probe from unincorporated bases was achieved using an ion exchange resin column ŽNEN-sorb, NEN DuPont. followed by precipitation in cold ethanol w23x.

2.4. VerÕet material 2.6. In situ hybridization histochemistry Primate material was made available for this study, from the University of the Witwatersrand Medical School, South Africa ŽJ. Temlett., in strict accordance with the University’s Animal Ethical Committee. 3 L-DOPA-naive

In situ hybridization was carried out as previously described w27x. Duplicate sections from all controls and PD cases were hybridized as one batch for each probe. All

62

D.J. EÕe et al.r Molecular Brain Research 50 (1997) 59–70

solutions used in pre-treatment and hybridization steps were prepared with ribonuclease-free sterile deionised distilled water. Prior to hybridization, sections were briefly warmed to room temperature, fixed for 5 min in 4% paraformaldehyde in phosphate buffered saline ŽPBS. pH 7.4 and then dehydrated through graded alcohols. Hybridization was carried out at 338C for 15 h in hybridization buffer containing 0.6 M sodium chloride and 0.06 M sodium citrate Ž4 = SSC., 50% formamide, 0.5 mgrml salmon sperm DNA, 0.02% Denhart’s solution, 0.25 mgrml tRNA, 10% dextran, 50 m M d-ATP, 10 mM dithiothreitol, 0.075 mgrml poly adenosine. Labelled probe was added to give an actual radioactive concentration of 18 million dpmrml of hybridization buffer, an estimated specific activity of 30 million dpmrpmol and therefore a probe concentration of 0.6 pmolrml of hybridization buffer. Post-hybridization washing was carried out in 4 changes of 1 = SSC at 518C for 1 h and then at room temperature for 1 h with two further changes of solution. Excess salt was removed from sections by two brief rinses in deionised water prior to air-drying at room temperature. Sections were then dipped in 1:1.5 diluted K5 nuclear emulsion ŽIlford., and exposed for 2 weeks at 48C prior to development and counter-staining with 0.05% Toluidine blue for subsequent visualisation and quantitation using light microscopy and image analysis. The vervet tissue was hybridized and dipped in an identical fashion to that for human material. 2.7. Validation of probe specificity Specificity of SS probe hybridization was confirmed by competition studies; hybridization of labelled probe being abolished by addition of a 50-fold excess of unlabelled probe but not by an un-related sequence of the same length, and by Northern analysis. For Northern analysis, RNA was extracted from human striatal blocks using the guanidinium isothiocyanate method of Chomczynski and Sacchi w8x and run on agarose gels alongside RNA standards Ž0.24–9.5 kb, Gibco BRL. according to established protocols w23x. The gels were blotted onto nylon membranes, stained with 0.04% methylene blue in sodium acetate and then examined and photographed. The SS probe was 3X tail-labelled with 32 P-dATP as described above. Following prehybridization of the membrane in buffer consisting of 15% deionised formamide, 7% SDS and 0.5 M sodium phosphate ŽpH 6.8. at 508C, labelled probe was added to the buffer Žapproximately 1.5 = 10 6 cpmrml. and hybridization carried out overnight at 508C. The membrane was washed in 1 = SSCr0.1% SDS at 508C for 30 min, followed by 30 min in 0.2 = SSCr0.1% SDS before apposition to X-ray film ŽKodak X-omat Xar5.. The specificity of the SS mRNA probe was confirmed by hybridization to a single band corresponding with a size of 1.0 kb ŽFig. 1., consistent with the known size of this mRNA species in man w10x.

Fig. 1. Validation of SS mRNA probe sequence by Northern hybridization. 32 P-dATP labelled probe sequence hybridizes to a band at approximately 1 kb as shown on the blot ŽSS.. This agrees favourably with published values w10x, and in conjunction with cold and irrelevant excess protocols confirms the selectivity of this probe sequence to SS mRNA. The values represent the position of specific kb marker bands, where O is the origin.

2.8. Quantitation Tissue sections were examined under a =40 objective; silver grain area over labelled cell bodies, as determined from the Nissl staining of Toluidine blue, was treated as a quantitative measure of mRNA expression, and measured using computerised image analysis ŽQuantimet Q570 image system.. The putamen was arbitrarily subdivided into six approximately equal subregions as shown in Fig. 2 w28x. The dorsolateral ŽDL., dorsomedial ŽDM., intermedio-lateral ŽIL., intermediomedial ŽIM., ventrolateral ŽVL., and ventromedial ŽVM. regions were studied in order to allow assessment of regional variation in SS expression between dorsal and ventral striatum. Quantitation of silver grain area was performed on all SS-positive neurons in the MML of the globus pallidus, and on an average of 30

D.J. EÕe et al.r Molecular Brain Research 50 (1997) 59–70

63

3.1. Distribution of SS mRNA in control basal ganglia

Fig. 2. Diagrammatic representation of the basal ganglia regions examined. The putamen was subdivided into six subregions for the purpose of quantitation of SS mRNA expression from emulsion autoradiograms. DL sdorsolateral; IL s intermediolateral; VL s ventrolateral; DM s dorsomedial; IM s intermediomedial; VM s ventromedial; GPE s external globus pallidus; GPIs internal globus pallidus; MML s medial medullary lamina; RTNs reticular thalamic nucleus; THAL s thalamus.

neurons in each subregion of the putamen, and in the caudate and RTN. The relative mean cross-sectional area of labelled cell bodies was also measured. Results were expressed either as mean area of silver grain overlying each labelled cell body, or grain density Ž% of cell covered by grains; derived from grain area divided by cell area = 100.. Results were analyzed using the unpaired 2-tailed Student’s t-test, with values of P - 0.05 considered significant. The subregions of the putamen were compared using a general linear model of ANOVA ŽwGLMx Minitab, Minitab Inc... A pairwise comparison of the regions was then performed using the Bonferroni transformation to determine significance Ž P - 0.0033 taken as significant.. This transformation reduces the probability of making a statistical error following pairwise comparison. The vervet data were analyzed using the non-parametric Mann–Whitney test due to the small sample size available. Grain area, cell area and grain density of all individual cells were used to construct cumulative frequency distributions for PD and control groups, and these were compared using the Kolmogorov–Smirnov test.

3. Results In the present study, post-mortem delay showed no correlation with detected levels of SS mRNA expression in any of the sites studied in either the PD or control group.

In control material, scattered subpopulations of neurons labelled with SS probe were found in all subregions of the putamen, the caudate, the RTN, which was strongly labelled, and the MML. No labelling was observed within the subthalamic nucleus. A reverse phase film autoradiograph ŽFig. 3. demonstrates SS mRNA expression within all regions examined, and examples of labelled cells from emulsion autoradiographs are shown in Fig. 4. The distribution of labelled cells within the striatum appeared relatively uniform, though clusters of labelled neurons were seen in some areas Žsee Fig. 3.. The majority of labelled striatal cell bodies were medium-sized Žapproximately 160 m m2 . and either bipolar, ovoid or fusiform, and therefore not distinguishable histologically from unlabelled cell bodies. The mean cross-sectional area of labelled cell bodies in the DL-putamen was smaller than that of cells within the VM region of the putamen ŽDL-putamen; 149.49 " SEM 7.39 m m2 compared with VM-putamen; 174.18 " SEM 6.35 m m2 ., but pairwise comparison of cell body cross-sectional area across all putaminal regions showed that this difference did not reach significance. Pairwise comparison of grain density in the six regions demonstrated the presence of a DL-VM decreasing gradient in SS mRNA expression, with the lateral putamen expressing higher mRNA levels than the medial putamen. The IM and VM regions showed significantly lower grain density than the DL region Ž P - 0.001; DL: 24.89 " SEM 1.35, IM: 18.43 " SEM 0.99, VM: 17.96 " SEM 2.63; pairwise comparison following Bonferroni transformation.. Cells of the RTN were predominantly of fusiform shape with a mean cross-sectional area of 299.63 " SEM 15.22 m m2 and the majority of these appeared to be labelled. Labelled cells within the MML of controls had a mean cross-sectional area of 236.5 " SEM 29.47 m m2 but overlapping populations of large and medium sized cells could be distinguished. The majority of SS-positive cells in the MML tended to be of medium size and were bipolar, triangular or ovoid in shape, very similar to unlabelled cells, whereas the large SS-positive cells tended to be either fusiform or globular. Within the MML of controls, cellular SS mRNA expression appeared to be heterogeneous; the majority of cells showed relatively low expression, with a small subpopulation of cells showing 10-fold greater levels of expression. 3.2. SS mRNA expression in PD The anatomical distribution of labelled cells within all four nuclei examined demonstrated no gross difference in regional expression between control and parkinsonian cases. Mean cross-sectional area of labelled cells within parkinsonian DL-putamen were significantly smaller than cells within all other putaminal subdivisions except the IM putamen Ž P - 0.0005, pairwise comparison following

D.J. EÕe et al.r Molecular Brain Research 50 (1997) 59–70

64

Fig. 3. Reverse phase autoradiographic image demonstrating SS mRNA expression in ŽA. control and ŽB. parkinsonian human basal ganglia Žsilver grains showing as white dots.. Clustering of SS-ergic neurons within the putamen Žp. can be clearly seen. SS mRNA can also be seen within the caudate Žc., reticular thalamic nucleus Žrtn.. Expression within the basal forebrain ventral to the striatum Žbf., and in the cortex Žctx. dorsal to the putamen also is evident. SS mRNA expression can only clearly be seen in the medial medullary lamina Žmml. of PD patients demonstrating increased expression compared with controls. Scale bar s 2.5 mm.

Bonferroni transformation; DL: 131.2 " SEM 4.74 m m2 ; DM: 175.8 " SEM 9.19 m m2 , IL: 181.8 " SEM 8.96 m m2 , VL: 180.7 " SEM 5.74 m m2 , VM: 178.1 " SEM 11.78

m m2 .. This significant decrease in mean cell size was not observed in control cases Žsee Table 3.. A significant decrease in the DL and increase in the IL and VM regions

Table 3 Cell size and SS mRNA expression in the striatum, reticular thalamic nucleus and medial medullary lamina Žmean, SEM.; obtained by image analysis of emulsion autoradiographs. Mean values were derived from an average of 30 cells in all regions except the MML, where all visible cells were measured. Grain area covering cells measured in m m2 silver grainrcell, SEM; Cell area in m m2 ; grain densitys % of cell area covered by grains Area

Controls DL putamen DM putamen IL putamen IM putamen VL putamen VM putamen Caudate RTN MML a b c

Grain area Ž m m2 .

Number

12 6 12 14 10 6 12 9 13

PD 9 6 10 10 6 6 9 8 9

Controls 38.58, 2.99 35.09, 2.44 34.32, 2.81 31.46, 3.03 33.11, 2.03 34.56, 6.39 37.09, 3.54 59.65, 3.52 26.29, 4.48

Cell area Ž m m2 .

PD

Controls b

30.68, 3.12 41.37, 4.67 43.07, 5.09 b 31.17, 2.95 39.04, 3.13 39.43, 5.56 b 33.95, 2.69 54.89, 5.36 b 56.5, 9.3 abc

149.49, 7.39 156.57, 13.73 159.24, 7.91 153.53, 8.78 164.02, 5.48 174.18, 6.35 160.14, 6.01 299.63, 15.22 236.50, 29.47

P - 0.01 for parkinsonian compared with control cases, Student’s unpaired t-test. P - 0.05 for parkinsonian compared with control cases, Kolmogorov–Smirnov test. P - 0.01 for parkinsonian compared with control cases, Kolmogorov–Smirnov test.

Grain density Ž%. PD bc

131.2, 4.74 175.81, 9.19 b 181.79, 8.96 b 157.39, 5.03 180.66, 5.73 178.13, 11.78 146.54, 4.98 266.11, 19.43 c 287.41, 25.97

Controls

PD

24.89, 1.35 20.94, 1.97 21.02, 1.78 18.43, 0.99 20.47, 1.40 17.95, 2.63 22.26, 2.02 20.43, 0.72 10.54, 1.11

22.68, 1.94 22.18, 2.08 23.06, 2.55 18.53, 1.51 21.13, 1.71 22.17, 2.94 b 21.99, 1.19 21.43, 2.4 18.1, 1.78, abc

D.J. EÕe et al.r Molecular Brain Research 50 (1997) 59–70

65

Fig. 4. Photomicrographs of emulsion-dipped sections from a SS hybridization, counterstained with Toluidine blue showing labelled cells from striatal regions: Ža. putamen, Žb. caudate, Žc. reticular thalamic nucleus, and Žd. medial medullary lamina. Probe binding is indicated by the presence of black silver grains in the overlying autoradiographic emulsion. Scale bars s 10 m m.

in PD with respect to controls, was observed for the cumulative distribution of SS mRNA expression measured by grain area ŽKolmogorov–Smirnov test; P - 0.05. though, no significant change in mean SS mRNA expression was seen in any of the 6 putaminal subregions examined ŽStudent’s t-test; P ) 0.1. with respect to controls, including the IL putamen, where we have previously demonstrated upregulation of PPE mRNA in PD cases w28x. A significant difference in the cumulative distribution of SS mRNA expression, measured by grain area, and the cell area of SS-ergic cells was detected in PD RTN

compared with controls ŽKolmogorov-Smirnov test; P 0.05., though the mean values were not significantly different ŽStudent’s t-test; P ) 0.1.. In the MML, the total number of SS-expressing cells present in PD cases was increased by 82.1% Ž39.8 cells per section" SEM 3.4, n s 9, compared with controls: 21.8 cells per section" SEM 4.2, n s 13, P - 0.01, Student’s t-test. The cumulative distribution of SS mRNA expression within SS-ergic cells of PD MML was significantly increased compared with controls Ž P - 0.001; Kolmogorov–Smirnov test. and a significant increase Ž115%. in mean SS mRNA expression per cell Žmeasured as grain

D.J. EÕe et al.r Molecular Brain Research 50 (1997) 59–70

66

Fig. 5. Frequency histogram of grain density of SS mRNA expression in parkinsonian and control medial medullary lamina. Grain densitys % of cell area covered by grains. Larger proportion of cells express high levels of SS mRNA in PD cases, but a population of low-expressing cells persists in PD.

area. in the MML of PD cases was also observed Ž56.5 m m2 silver grainrcell" SEM 9.3, n s 9, compared with controls: 26.3 m m2 silver grainrcell" SEM 4.5, n s 13, P - 0.01, Student’s t-test.. A larger proportion of MML cells in PD cases showed very high levels of SS mRNA, particularly in those cells with profiles whose cross-sectional area was greater than 180 m m2 , although a population of cells with low expression persisted ŽFig. 5.. The possible relationship between dementia in PD and SS expression was also studied by subdividing the PD group into demented and non-demented subgroups Ž n s 3–5 for each group. but no significant correlation with SS

mRNA expression was demonstrated in any basal ganglia area examined. 3.3. Primate studies The three MPTP-lesioned vervets demonstrated parkinsonian-like symptoms and almost complete destruction of the substantia nigra on the lesioned side at post-mortem. Only the caudate and putamen were examined due to the difficulty of identifying the RTN and MML in all animals due to freeze artefact. However, SS-positive cells were observed where these structures could be recognised, and

Table 4 Cell size and SS mRNA expression in the striatum of unilaterally MPTP-lesioned vervets Žmedian.; values obtained by image analysis of emulsion autoradiographs; comparison of lesioned with unlesioned sides. Median values obtained from each region measured and derived from an average of 30 cells. Grain area covering cells measured in m m2 silver grainrcell; Cell area in m m2 ; grain densitys % of cell area covered by grains Area

Number

DL put IL put VM put Caudate

3 3 3 3

a

Grain area Ž m m2 .

Cell size Ž m m2 .

Grain density Ž%.

Unlesioned

Lesioned

Unlesioned

Lesioned

Unlesioned

Lesioned

33.47 25.52 31.66 38.95

32.62 36.57 32.76 44.81

179.18 169.41 145.27 181.98

147.51 179.41 174.7 163.89

20.16 19.01 21.31 24.18

20.03 21.01 22.87 26.59

P - 0.05 for lesioned side compared with unlesioned side in MPTP-treated vervets, Kolmogorov–Smirnov test.

a

D.J. EÕe et al.r Molecular Brain Research 50 (1997) 59–70

were generally of the fusiform type. Cells within the vervet striatum tended to be more densely congregated, but their appearance was similar to that observed in humans; medium-sized, bipolar, ovoid or fusiform. Neuronal clustering was also evident. No DL-VM putaminal gradient was apparent in caudate and lesioned putamen. The distribution of SS mRNA expression within the VM putamen of the vervets was shown to be increased in the lesioned animals Ž P - 0.05; Kolmogorov–Smirnov test., but no alteration in other areas was observed and no difference between median SS mRNA expression in the caudate, or any of the putaminal subregions examined including the VM putamen, was observed ŽMann–Whitney test; see Table 4; P ) 0.05..

4. Discussion The distribution of SS mRNA within the basal ganglia demonstrated in this study is compatible with that previously demonstrated in humans by Mengod et al. w25x using ISHH and by Bouras et al. w6x using radioimmunoassay and immunohistochemistry. In agreement with these studies, a scattered population of heavily-labelled cells was detected within the putamen, caudate, and EMLrRTN, the globus pallidus was devoid of signal, and the MML, intercalated between the external and internal globus pallidi showed an intense hybridization signal. Where cerebral cortex was present in the tissue section, signal could also be clearly observed in the deep cortical layers. Alteration of SS mRNA expression within the basal ganglia in human neurodegenerative disease has not previously been reported. SS mRNA has previously been shown to be decreased within the hippocampus of Alzheimer’s disease patients w10x and the cell size of SS-immunoreactive neurons in the striatum has been shown to be reduced, though levels of SS peptide expression were unaffected w35x. Mufson and Brandabur w26x have demonstrated no significant morphological alteration of these neurons in Parkinson’s and Alzheimer’s diseases. Studies of postmortem tissue from patients with Huntington’s disease demonstrate sparing of interneurons expressing SS, NPY and NOS mRNA in the damaged striatum; no alteration in the mRNA expression of these neurotransmitters has been detected within the striatum of patients with Huntington’s disease w24x, but radioimmunoassay studies on extracted SS from the caudate, putamen, and internal and external globus pallidus demonstrate a significant increase in SS peptide expression in Huntington’s disease, implying that SS turnover has been altered w1x. Alterations in SS expression within the cortex has been linked to dementia, so the possible influence of dementia on SS expression in the basal ganglia was studied in PD patients, half of whom had shown signs of dementia at the time of death. No correlation between dementia and SS mRNA expression was detected in any region of the basal ganglia investigated. No

67

correlation between drug treatments, such as dose of LDOPA administered, was observed. This suggests that any changes observed are symptomatic of PD rather than some confounding factor. 4.1. Putaminal SS mRNA expression Our observations concerning clustering of SS-ergic neurons in the striatum correspond to similar findings of SS-ergic cell clustering in the matrix around the striosomes of rodent striatum w3x. Gerfen w14x and Rushlow et al. w32x have shown that in rodents, SS-ergic interneurons do occasionally project fibres from the matrix to striosome, are frequently situated at the matrix-striosome border, and propose that SS cell clusters around striosomes may act as a communication between the striosome and matrix. Combined SS ISHH and acetylcholinesterase labelling might shed further light on whether the clustering we observed reflects a similar striosome-related distribution in human striatum. In control material, a DL-VM gradient of decreasing mean SS mRNA expression was observed in control patients, but was lost in PD cases. This finding was supported by cumulative distribution analysis which demonstrated increased expression in the IL and VM and decreased expression in the DL putamen of PD patients compared with controls. No significant difference in mean SS-positive cell size across the subregions of control putamen, using pairwise comparison and Bonferroni transformation was observed, but there was a trend towards the DL cells being smaller than VM cells. This trend towards a decrease in mean cell size from VM to DL striatum reached significance in PD cases Ž p - 0.0005; pairwise comparison following Bonferroni transformation.. It has been suggested that a proportion of SS-ergic neurons may not be interneuronal, particularly in the ventral region of the striatum w40x. The contribution of SS-ergic neurons with different functions and characteristics Ži.e., interneuronal and possibly projection neurons. may account for any regional variation in neuropeptide mRNA expression observed in our studies. Variations in DA-ergic innervation of the putamen might also contribute to the regional differences in SS mRNA expression observed in this study. Zones of low TH immunoreactivity comparable with striosomal regions have been identified in human striatum w15x; patches of enhanced TH immunoreactivity in dorsolateral and a more uniform innervation of the ventral striatum have been detected in rodents w9x. This pattern of innervation could underlie the DL-VM gradient in expression of SS observed in controls. Continuous stimulation with both D1 and D2 agonists increases striatal SS immunoreactivity in rodents previously lesioned with 6-hydroxydopamine w11x, whereas antagonists suppress SS mRNA expression w3x. The implication from this evidence is that DA exerts an excitatory influence on expression of SS in a population of striatal aspiny interneurons. Kubota

68

D.J. EÕe et al.r Molecular Brain Research 50 (1997) 59–70

et al. w20x demonstrated that in rat striatum, there are soma and proximal dendritic synaptic connections between 50% of NPY-immunoreactive neurons and TH-immunoreactive axon terminals. The co-localisation of NPY with SS and NOS within these interneurons supports the existence of a direct synaptic connection of the soma and proximal dendrites of SS-ergic neurons, which co-express NOS and NPY, with nigrostriatal DA-ergic terminals. The loss of the DL-VM gradient of SS mRNA expression in PD may be explained by the heterogeneous loss of DA-ergic fibres innervating the striatum. Cell loss in the nigra is greatest in areas which innervate the DL putamen w18x, and removal of DA-ergic innervation of this region is likely to reduce SS mRNA expression and hence remove the gradient. A further possibility is that this effect is due to L-DOPA treatment. The loss of the heterogeneous and presumably specific influence of DA-ergic innervation on SS mRNA expression present in control cases, and its replacement by a potentially homogenous influence of exogenous L-DOPA in PD might explain the loss of the SS-ergic mRNA expression gradient in PD. The decrease in SS-ergic cell size observed in the DL-putamen of PD cases, may result from the reduced innervation by DA-ergic nigrostriatal neurons, leading to shrinking of these cells. A reduction in SS-ergic cell size has also been observed in the putamen of Alzheimer’s disease patients w35x, possibly due to reduced cholinergic innervation of these neurons. The significant decrease in the DL putamen, and increase in the VL and VM putamen observed following comparison of the distribution of SS mRNA expression ŽKolmogorov–Smirnov test. in PD compared with control cases supports the loss of the SS mRNA expression gradient in PD observed when mean values were compared ŽGLM of ANOVA., and demonstrates that changes in expression within the same structure occur depending on the specific location and connectivity of nerve cells. Rodent studies concerning the effects of nigral lesions on striatal SS mRNA expression have demonstrated either an increase in the number of cells expressing SS mRNA w22x, or a decrease w37x in SS gene expression per cell, though others have shown no change w33x in SS immunoreactivity. All these studies were carried out in acute models where the animals have been killed 2–3 weeks after lesioning and hence may not accurately reflect changes in mRNA expression in PD, where nigral damage has been present for many years. Alternatively, the relatively subtle changes in SS mRNA expression within the striatum of the PD cases examined here, may be due to exogenously administered L-DOPA mediating partial reversal of any denervation-induced changes Žall PD cases were taking L-DOPA prior to death.. It has previously been shown w2x that L-DOPA treatment reverses any changes seen in SS concentration within the striatum following MPPq-lesion of the nigrostriatal pathway in an acutely-lesioned mouse model. In our preliminary studies of MPTP-lesioned vervet

in which no L-DOPA was administered, a significant increase in the distribution of SS mRNA expression in lesioned VM putamen was apparent, compared with the unlesioned side, but no alteration in mean values or in the DL putamen were evident. In view of the small sample size of the vervet study Ž n s 3., this result requires confirmation by repeating the study using a larger group. A number of lines of evidence support the possible colocalisation of SS with NOS and NPY within the striatum, hippocampus and cortex of humans, primates and rodents w19,35,39x. Rushlow et al. w32x have demonstrated 100% co-localisation of SS with NPY in rat striatum, and 100% co-localisation of NOS with SS, except for the putaminal region bordering the globus pallidus, but only 84% colocalisation of NPY with SS, implying some heterogeneity of striatal SS neurons. Our previous finding that mean NOS mRNA expression within the striatum of PD cases treated with L-DOPA was unchanged compared with controls w29x is consistent with our current results for SS mRNA, and compatible with co-regulation of NO and SS neurotransmitter systems in human striatal interneurons.

5. RTN expression of SS mRNA Hazrati and Parent w16x have demonstrated in the squirrel monkey, the presence of fibres projecting from the pallido-pallidal, pallido-subthalamic and pallido-nigral pathways to the RTN, where co-localisation of SS and GABA w38x may occur, since this co-localisation has been shown within the hippocampus and cerebral cortex w34x. The RTN is believed to exert an inhibitory pacemaker-like influence on thalamic activity via a GABA-ergic pathway to other thalamic neurons w38x, potentially resulting in disinhibition of the thalamus, but inhibition of thalamic projections to the cerebral cortex. The cumulative distribution of SS mRNA expression was demonstrated to be reduced in PD cases, with a concomitant decrease in cell area, but no changes in mean SS mRNA expression were observed. This suggests that this interneuronal pathway may be affected in PD, whereby shrinkage of the SS-postive cells occurs, but further investigation of SS expression within L-DOPA naive vervets is required to determine whether the alterations observed are due to L-DOPA treatment or a consequence of PD. GABA-ergic expression within this area and its possible interaction with SS should also be investigated to identify the role of RTN interneurons in PD. 5.1. MML expression in SS mRNA In this study, SS mRNA expression per cell and the number of cells expressing SS mRNA were found to be increased within the MML in PD. This might be due to a significant upregulation of SS mRNA expression in PD of cells normally expressing undetectable levels or a pheno-

D.J. EÕe et al.r Molecular Brain Research 50 (1997) 59–70

typic change whereby large MML neurons develop expression of SS mRNA following the onset of PD. Soghomonian and Chesselet w37x demonstrated a significant Ž300%. increase in the number of labelled cells and in the intensity of labelling per cell of SS in the entopeduncular nucleus of the rat following a nigrostriatal lesion by 6-hydroxydopamine. This area is believed to be equivalent to the internal globus pallidus of primates, and possibly may contain cells forming a rodent equivalent of the MML. Our recent studies of NOS mRNA expression show a similar increase in mRNA expression to that demonstrated here for SS mRNA within the MML w29x. NOS mRNA expression was observed primarily in ventrally located cells of the MML whereas SS mRNA expression was observed throughout the MML ŽNisbet A, personal communication.. This finding is compatible with co-regulation of co-localised NOS and SS mRNA within the ventral MML and is the first evidence for such co-regulation of these 2 neurotransmitters in any species. Our data also suggest that the dorsally located MML cells are a separate population of cells with respect to neurotransmitter expression. The absence of SS mRNA expression within the subthalamic nucleus, where NOS mRNA has previously been demonstrated w27x confirms that 100% colocalisation of NOS and SS mRNA is not observed in humans. Recently using tyrosine hydroxylase immunohistochemistry we have demonstrated that the human MML is particularly rich in DA-ergic fibres w29x; this has also been demonstrated in primate studies w21x. In primates, these fibre tracts form part of the nigropallidal projection, but may also contain nigral axons en route to the striatum via the MML w30x. Thus SS-ergic neurons could interact with the DA-ergic system within the MML itself. Application of SS increases DA release from striatal slices in vitro w7x and this effect is additive with respect to the co-localised neurotransmitter NPY w5x providing a potential mechanism for a compensatory interaction between DA fibres and SS-ergic cells in the striatum and MML in PD. An increase in SS production in the MML in PD might therefore reflect a compensatory change to potentiate the release of DA from the surviving nigrostriatal terminals.

6. Conclusions This study demonstrates loss of a DL-VM putaminal gradient in SS mRNA expression in PD patients, perhaps reflecting loss of topographic organisation of DA inputs to the striatum. An increase in SS cellular expression has been demonstrated in the MML in PD. This parallels results obtained in human basal ganglia for NOS mRNA, Žan indicator of the putative neurotransmitter NO. w29x, which is probably co-localised with SS in ventral MML cells, suggesting co-regulation of these 2 neurotransmitter systems in these cells. The function of the MML is not

69

known, but the increase in SS expression seen in PD provides a possible mechanism by which this structure might influence the activity of DA-ergic pathways following nigral injury.

Acknowledgements All Parkinsonian brains were obtained through the UK Parkinson’s Disease Society Brain Research Centre ŽBrain Bank. ŽUK PDSBRC. donor scheme. Control brains were obtained either from routine hospital post-mortems, from registered donors via the UK PDSBRC or through the Medical Research Council Alzheimer’s Disease Brain Bank, De Crespigny Park, Denmark Hill, London SE5 8AF, by the kind assistance of Dr. N. Cairns. We wish to thank Dr. S.E. Daniel for expert neuropathological evaluation, confirmation of the diagnosis of PD and the exclusion of any unsuspected neurological disease from the control group. We are also grateful to Mrs. Angela Lee and Mrs. Sarah Brocklehurst, Institute of Neurology, for their help and guidance in statistical analysis of the data. David Eve is a PhD student sponsored by The Wellcome Trust, Dr. Nisbet is a UK Parkinson’s Disease Society Research Fellow, Ann Kingsbury is seconded by the MRC to the PDSBRC, and this work was funded by the UK Parkinson’s Disease Society, the UK Medical Research Council and The Wellcome Trust.

References w1x N. Aronin, P.E. Cooper, L.J. Lorenz, E.D. Bird, S.M. Sagar, S.E. Leeman, J.B. Martin, Somatostatin is increased in the basal ganglia in Huntington disease, Ann. Neurol. 13 Ž1983. 519–526. w2x M. Asanuma, N. Ogawa, Y.H. Sora, K. Pongdhana, K. Haba, A. Mori, Alterations of somatostatin and its modulation by levodopa in MPTP-treated mouse brain, J. Neurol. Sci. 100 Ž1990. 155–160. w3x S.J. Augood, H. Kiyama, R.L.M. Faull, P.C. Emson, Dopaminergic D1 and D2 receptor antagonists decrease prosomatostatin mRNA expression in rat striatum, Neuroscience 44 Ž1991. 35–44. w4x M.F. Beal, J.B. Martin, The effect of somatostatin on striatal catecholamines, Neurosci. Lett. 44 Ž1984. 271–276. w5x M.F. Beal, R.C. Frank, D.W. Ellison, J.B. Martin, The effect of neuropeptide Y on striatal catecholamines, Neurosci. Lett. 71 Ž1986. 118–123. w6x C. Bouras, P.J. Magistretti, J.H. Morrison, J. Constantinidis, An immunohistochemical study of pro-somatostatin-derived peptides in the human brain, Neuroscience 22 Ž1987. 781–800. w7x M.F. Chesselet, T.D. Reisine, Somatostatin regulates dopamine release in rat striatal slices and cat caudate nuclei, J. Neurosci. 3 Ž1983. 232–236. w8x P. Chomczynski, N. Sacchi, Single step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction, Anal. Biochem. 162 Ž1987. 156–159. w9x G. Doucet, L. Descarries, S. Garcia, Quantification of the dopamine innervation in adult rat neostriatum, Neuroscience 19 Ž1986. 427– 445. w10x P. Dournaud, P. CerveraPierot, E. Hirsch, F. JavoyAgid, C.L. Kordon, Y. Agid, J. Epelbaum, Somatostatin messenger RNA-contain-

70

w11x

w12x w13x w14x w15x

w16x w17x

w18x w19x

w20x

w21x w22x w23x w24x w25x w26x

D.J. EÕe et al.r Molecular Brain Research 50 (1997) 59–70 ing neurons in Alzheimer’s disease: an in situ hybridization study in hippocampus, parahippocampal cortex and frontal cortex, Neuroscience 61 Ž1994. 755–764. T.M. Engber, R.C. Boldry, S. Kuo, T.N. Chase, Dopaminergic modulation of striatal neuropeptides: differential effects of D1 and D2 receptor stimulation on somatostatin, neuropeptide Y, neurotensin, dynorphin and enkephalin, Brain Res. 581 Ž1992. 261–268. J. Epelbaum, M. Ruberg, E. Moyse, F. JavoyAgid, B. Dubois, Y. Agid, Somatostatin and dementia in Parkinson’s disease, Brain Res. 278 Ž1983. 376–379. J.A. Garcia-Sevilla, T. Magnusson, A. Carlsson, Effect of intracerebroventricularly administered somatostatin on brain monoamines turnover, Brain Res. 155 Ž1978. 159–164. C.R. Gerfen, The neostriatal mosaic; compartmentalisation of corticostriatal input and striatonigral output systems, Nature 311 Ž1984. 461–464. A.M. Graybiel, E.C. Hirsch, Y.A. Agid, Differences in tyrosine hydroxylase-like immunoreactivity characterize the mesostriatal innervation of the striosomes and extrastriosomal matrix at maturity, Proc. Natl. Acad. Sci. USA 84 Ž1987. 303–307. L.-N. Hazrati, A. Parent, Projection from the external pallidum to the reticular thalamic nucleus in the squirrel monkey, Brain Res. 550 Ž1991. 142–146. A.E. Kingsbury, O.J.F. Foster, A.P. Nisbet, N. Cairns, L. Bray, D.J. Eve, A.J. Lees, C.D. Marsden, Tissue pH as an indicator of mRNA preservation in human post-mortem brain, Mol. Brain Res. 28 Ž1995. 311–318. S.J. Kish, K. Shannak, O. Hornykiewicz, Uneven patterns of dopamine loss in the striatum of patients with idiopathic Parkinsons’ disease, N. Engl. J. Med. 318 Ž1988. 876–880. N.W. Kowall, R.J. Ferrante, M.F. Beal, E.P. Richardson Jr., M.V. Sofroniew, A.C. Cuello, J.B. Martin, Neuropeptide Y, somatostatin, and reduced nicotinamide adenine dinucleotide phosphate diaphorase in the human striatum: a combined immunocytochemical and enzyme histochemical study, Neuroscience 20 Ž1987. 817–828. Y. Kubota, S. Inagaki, S. Kito, S. Shimada, T. Okayama, H. Hatanaka, G. Pelletier, H. Tagaki, M. Tohyama, Neuropeptide Y immunoreactive neurons receive synaptic inputs from dopaminergic axon terminals in the rat neostriatum, Brain Res. 458 Ž1988. 389– 393. B. Lavoie, Y. Smith, A. Parent, Dopaminergic innervation of the basal ganglia in the squirrel monkey as revealed by tyrosine hydroxylase immunohistochemistry, J. Comp. Neurol. 289 Ž1989. 36–52. N. Lindefors, S. Brene, M. Herrera-Marschitz, H. Persson, Regulation of neuropeptide Y gene expression in rat brain, Ann. NY Acad. Sci. 611 Ž1990. 175–185. T. Maniatis, E.F. Fritsch, J. Sambrook, J. Molecular Cloning: A Laboratory Manual, Cold Spring Harbour Laboratory, 1982. G. Mengod, J.-L. Charli, J.M. Palacios, The use of in situ hybridization histochemistry for the study of neuropeptide gene expression in the human brain, Cell. Mol. Neurobiol. 10 Ž1990. 113–126. G. Mengod, M. Rigo, M. Savasta, A. Probst, J.M. Palacios, Regional distribution of neuropeptide somatostatin gene expression in the human brain, Synapse 12 Ž1992. 62–74. E.J. Mufson, M.M. Brandabur, Sparing of NADPH-diaphorase stri-

w27x

w28x

w29x

w30x

w31x w32x w33x

w34x w35x

w36x w37x

w38x

w39x w40x

atal neurons in Parkinson’s and Alzheimer’s diseases, NeuroReport 5 Ž1994. 705–708. A.P. Nisbet, O.J.F. Foster, A. Kingsbury, A.J. Lees, C.D. Marsden, Nitric oxide synthase mRNA expression in human subthalamic nucleus, striatum, and globus pallidus: implications for basal ganglia function, Mol. Brain Res. 22 Ž1994. 329–332. A.P. Nisbet, O.J.F. Foster, A.E. Kingsbury, D.J. Eve, S.E. Daniel, C.D. Marsden, A.J. Lees, Preproenkephalin and preprotachykinin messenger RNA expression in normal human basal ganglia and in Parkinson’s disease, Neuroscience 66 Ž1995. 361–376. A.P. Nisbet, O.J.F. Foster, A.E. Kingsbury, C.D. Marsden, A.J. Lees, Nitric oxide synthase mRNA expression in parkinsonian basal ganglia: a quantitative in situ hybridisation study, Movement Dis. 9 ŽSuppl. 1. Ž1994. 426. A. Parent, B. Lavoie, Y. Smith, P. Bedard, The dopaminergic nigropallidal projection in primates: distinct cellular origin and relative sparing in MPTP-treated monkeys, in: M.B. Streifler, A.D. Korczyn, E. Melamed, M.B.H. Youdim ŽEds.., Parkinson’s Disease: Anatomy, Pathology and Therapy, Advances in Neurology, vol. 53, Raven Press, New York, 1990, pp. 111-116. W. Rushlow, B.A. Flumerfelt, C.C.G. Naus, Colocalization of somatostatin, neuropeptide Y and NADPH-diaphorase in the caudateputamen of the rat, J. Comp. Neurol. 351 Ž1995. 499–508. W. Rushlow, C.C.G. Naus, B.A. Flumerfelt, Somatostatin and the patchrmatrix compartments of the rat caudate-putamen, J. Comp. Neurol. 364 Ž1996. 184–190. P. Salin, L. Kerkerian-Le Goff, V. Heidet, J. Epelbaum, A. Nieoullan, Somatostatin-immunoreactive neurons in the rat striatum: effects of corticostriatal and nigrostriatal dopaminergic lesions, Brain Res. 521 Ž1990. 23–32. D.E. Schmechel, B.G. Vickrey, D. Fitzpatrick, R.P. Elde, GABAergic neurons of mammalian cerebral cortex: widespread subclass defined by somatostatin content, Neurosci. Lett. 47 Ž1984. 227–232. N. Selden, C. Geula, L. Hersh, M.M. Mesulam, Human striatum: chemoarchitecture of the caudate nucleus putamen and ventral striatum in health and Alzheimer’s disease, Neuroscience 60 Ž1994. 621–636. L.-P. Shen, R.L. Pictet, W.J. Rutter, Human somatostatin I: sequence of the cDNA, Proc. Natl. Acad. Sci. USA 79 Ž1982. 4575– 4579. J.J. Soghomonian, M.F. Chesselet, Lesions of the dopaminergic nigrostriatal pathway alter preprosomatostatin messenger RNA levels in the striatum, entopeduncular nucleus and the lateral hypothalamus of the rat, Neuroscience 42 Ž1991. 49–59. R. Spreafico, G. Battaglia, M. De Curtis, S. De Biasi, Morphological and functional aspects of nucleus reticularis thalami ŽRTN. of the rat, in: J.M. Besson, G. Guilbaud, M. Peschanski ŽEds.., Thalamus and Pain, Elsevier, Amsterdam, 1987, pp. 111-126. L.T. Weiss, M.F. Chesselet, Regional distribution and regulation of preprosomatostatin messenger RNA in the striatum as revealed by in situ hybridisation histochemistry, Mol. Brain Res. 5 Ž1989. 121–130. R. Widmann, N. Mensdorff-Pouilly, K. Pfaller, G. Sperk, Evidence for somatostatin-containing fibres projecting from the pallidal complex to the striatum of the rat, J. Neurochem. 48 Ž1987. 1857–1861.