Introduction

Top Abstract Introduction Materials & Methods Results Discussion Conclusion References

SIB is often observed in clinics of mental retardation, schizophrenia and character disorders (Green, 1967; Maisto et al., 1978; Winchel and Stanley, 1991). Actually the clinical characteristics of the behavior are manifold, and its source must be plural (Winchel and Stanley, 1991). In some types of SIB, evidence is accumulating that impairment of dopaminergic neurotransmission underlies SIB. Lloyd et al. (1981) found that dopamine and its metabolites were severely decreased in the striatum using the postmortem brains of patients suffered from Lesch-Nyhan syndrome, a genetic deficiency of HPRT, in which SIB has been known to be one of the neurological symptoms, in addition to mental retardation and choreoathetosis (Christie et al., 1982; Lesch and Nyhan, 1964). Clinical studies indicating dopamine impairment in Tourette's syndrome, in which SIB is one of the main symptoms, also supports the impairment of dopaminergic neurotransmission being at the basis of SIB (Sandy, 1987).

The intimate relationship of dopamine and SIB is also suggested by the experimental animals. In primates, Goldstein et al. (1986) succeeded to induce SIB by the electric lesion of midbrain dopaminergic neurons. The recent finding by Jinnah et al. (1993) of a reduction of dopamine levels in HPRT-gene-deleted mutants presents molecular evidence that the single gene mutation of HPRT deficiency could impair dopamine neurotransmission. Importance of dopaminergic neurotransmission to SIB is strongly supported by the discovery of neonatal dopamine-lesioned rats with 6-OHDA, a catecholamine neurotoxin, manifesting SIB after loading with L-DOPA (Breese et al., 1984; Breese et al., 1990). In pharmacobehavioral studies using this rat model, Breese et al. (1985) found that dopamine D₁-like receptors have a major role in the induction and cessation of SIB. However, D₁ binding and D₁-mediated adenylate cyclase activity were not altered in the neonatal dopamine depleted rats (Breese et al., 1987a; Dewar et al., 1990; Duncan et al., 1993; Simson et al., 1992), but Luthman et al. (1990) found elevation of basal levels of striatal cyclic AMP. To date, neurochemical studies could not account for the neuronal mechanism of SIB in the animal model of neonatal 6-OHDA-treated rats.

Recently we demonstrated that the susceptibility of dopamine neurons to the neonatal 6-OHDA toxicity is different with each dopamine neuron system (Yokoyama et al., 1993a). Moreover, the 6-OHDA treatment in adults, which never exhibit SIB even when challenged by L-DOPA, showed a different pattern of the destruction of dopaminergic neurons (Ikeda et al., 1992), although the content of dopamine was similarly depleted as neonatal intracisternal 6-OHDA treatment. From these findings, we hypothesized that the specific loci related to the destruction of dopaminergic neurons and accompanied up-regulation of dopamine receptors underlies the manifestation of SIB. After this working hypothesis, we compared the dopaminergic neurons and receptors of three groups in neonatal intracisternal treated rats at their grown up adult age: the rats we compared were neonatal 6-OHDA-treated rats showing SIB (SIB(+) group), neonatal 6-OHDA-treated rats not showing SIB (SIB(-) group), and neonatal saline-treated controls (control group). A part of this study has been published in abstract form (Okamura et al., 1992; Okamura et al., 1991; Okamura et al., 1993; Yokoyama et al., 1992; Yokoyama et al., 1993b).

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion Conclusion References

Animals. Progeny of pregnant Wistar rats were used. Neonatal rats were first given desipramine (20 mg/kg, i.p.; Sigma Chemical Co., St. Louis, MO) to protect the noradrenergic neurons, and then 6-OHDA (100 µg/5 µl in saline containing 0.1% ascorbic acid; Sigma) intracisternally on postnatal day 1 and day 3 under deep cold anesthesia. For controls, the same volume of vehicle was injected. At the age of 4 to 6 wk, the behavior was evaluated. After 30 min of pretreatment with a peripheral aromatic L-amino acid decarboxylase inhibitor, Ro4-4602 (50 mg/kg i.p.; donated from Roche-Japan Co., Tokyo, Japan), the animals were loaded with L-DOPA (100 mg/kg i.p.; Sankyo Co., Tokyo, Japan) and their behavior was checked in a clear plastic cage for 150 min. 6-OHDA-treated rats were divided into those with (SIB(+)) and without (SIB(-)) occurrence. The repeated holding of their skin between their teeth was referred to SIB. After L-DOPA loading, about 70% (positive/total; 71/103) neonatal 6-OHDA-treated rats exhibited SIB. Immediately after we confirmed SIB, we performed i.p. injection of SCH-23390 (1 mg/kg; RBI, Natick, MA), a dopamine D₁-like receptor antagonist, to stop SIB. Usually, SIB ceased within 1 min after the injection, and the injured skin was covered by a sticking plaster. The protocol for this research was accepted by the Committee for Animal Research in Kyoto Prefectural University of Medicine. At the age of 3 mo, rats were killed for anatomical and biochemical studies.

Immunocytochemistry. Animals were anesthetized deeply with pentobarbital, and perfused with 0.1 M PBS (pH 7.4) followed by the ice-cold fixative containing 4% paraformaldehyde and 0.2% picric acid in 0.1 M phosphate buffer (Yokoyama et al., 1993a). The brains were removed and further fixed with the same fixative overnight. After cryoprotection with 0.1 M phosphate buffer containing 20% sucrose, the brain was cut into 25-µm thick sections with a cryostat. The sections were incubated with rabbit anti-TH serum (dilution 1/10000; gift from Prof. I. Nagatsu, Fujita-Gakuen University, Toyoake) for 3 days, and incubated with biotinylated anti-rabbit immunoglobulin (dilution 1/1000; Vector, Burlingame, CA) for 12 hr, and then with avidin-biotin peroxidase complex (dilution 1/1000; Vector, Burlingame, CA) for 12 hr, which were finally exposed to the 0.05 M Tris-HCl buffer solution with 0.01% 3,3-diaminobenzidine tetrahydrochloride (Sigma) and 0.25% nickel ammonium sulfate solution (Nakalai Tesque, Kyoto, Japan) containing 0.03% H₂O₂. Between each step of the reaction, the sections were rinsed with 0.1 M PBS. The characteristics and the specificity of the primary antiserum were detailed in previous report (Nagatsu et al., 1977). Sections were mounted on chrome-alum gelatin-coated slides, and were dehydrated with a graded series of ethanol and xylene and coverslipped with mounting media (Entellan, Merk, FRG).

HPLC analysis of dopamine contents in micropunched brain slice. Dopamine contents of three groups were examined by HPLC with an electrochemical detector. Rats were decapitated, and their brains were immediately frozen with dry ice, and 300-µm sections were made with a cryostat. Sections were punched out by a stainless punching needle (internal diameter: 1.0 mm) (Palkovits et al., 1979). The punched out areas were rostral, central and caudal levels of the dorsolateral part of the caudate-putamen, nucleus accumbens, tuberculum olfactorium, medial prefrontal cortex, anterior cingulate cortex, piriform cortex, entorhinal cortex, medial preoptic area, anterior hypothalamic area, dorsomedial hypothalamic nucleus, anterior olfactory nucleus, bed nucleus of the stria terminalis, amygdala, habenular nucleus, substantia nigra, VTA, rostral linear nucleus of raphe, dorsal raphe nucleus, locus ceruleus and A1 area of the medulla oblongata. For caudate-putamen, we punched out three areas in the rostral, central and caudal parts of the caudate-putamen, and for other areas, we performed two punch-outs. Tissues were homogenized with a polytron in the presence of 0.1 M perchloric acid (0.5 ml). After centrifugation at 2000 × g for 20 min, their supernatants were adjusted to pH 3.0 with sodium acetate, and frozen to -70°C until the homogenates were used. The dopamine content in the homogenate was measured by HPLC (EICOM, Kyoto, Japan) with an electrochemical detector by the internal standard method. The eluant was composed of 87.7 mM citric acid, 87.7 mM sodium acetate, 0.01 mM EDTA-2Na, 275 mg/liter l-octane-sulfonic acid and 125 ml/liter methanol, at pH 3.5. The flow rate was kept at 1.0 ml/min at 25°C. Protein concentrations were measured by Lowry's method (Lowry et al., 1951), and the concentration of dopamine was expressed as ng/mg protein. Noradrenaline content was also measured in the same way as above, and there was no difference among three groups in each punched out area.

In vitro receptor autoradiography. Rats were deeply anesthetized with pentobarbital (70 mg/kg), and perfused transcardially with ice cold 0.05 M PBS (pH 7.4) at 50 ml/min for 5 min, followed by ice cold 5% sucrose adding in the same buffer for 2 min. The brain was quickly removed, mounted onto microtome chucks and frozen by dry ice. Coronal tissue sections, 16 µm thick, were cut on a cryostat and thaw-mounted onto chrome-alm/gelatin-coated microscope slides. The slides were dried at room temperature for about 2 hr then stored at -20°C until the binding experiments.

6

Binding assay using membrane fraction of nigral homogenates. Rats (n = 5 in each group) were killed by decapitation and each brain was removed quickly. The brain was chilled in Tris-HCl buffer, and cut 2 mm long on the brain matrix and the tissue of the substantia nigra was picked out under stereoscopic microscope. The tissue was homogenized in 100 volumes of the ice-cold 50 mM Tris-HCl buffer (pH 7.4) and centrifuged at 3000 × g for 30 min. This step was repeated twice and the final pellet was resuspended with 50 mM Tris-HCl buffer (pH 7.4) containing 120 mM NaCl, 5 mM KCl, 2 mM CaCl₂ and 1 mM MgCl₂. Aliquots of 100 µl of membrane suspension were incubated in triplicate with [³H]SCH-23390 at various concentrations of 0.03 to 1.39 nM in a final volume of 1 ml at 22°C for 90 min. Nonspecific binding was determined in the presence of 10 µM (+)-butaclamol. After incubation, the contents were immediately filtered under reduced pressure through Whatman GF/B filters which had been presoaked in the binding buffer. The filters were rinsed 3 times each with 5 ml of ice-cold binding buffer. Specific binding, which was calculated by subtracting the nonspecific binding from total binding, was usually 75% of the total binding. The protein concentration was measured by the Lowry method (Lowry et al., 1951).

	Results

Top Abstract Introduction Materials & Methods Results Discussion Conclusion References

Dopaminergic Neurons and Fibers

Dorsal striatum. In the caudate-putamen, dense TH-IR terminals were detected in the whole caudate putamen, with a slightly higher density in the medial area and periventricular area just lateral to the lateral ventricle (fig. 1A). In neonatal 6-OHDA-treated rats, TH-IR terminals in the caudate-putamen were severely destroyed, in both of the SIB(+) and SIB(-) groups. We analyzed the structure dividing into four subregions; CPPV, CPDL, CPM and CPVL. In general, the toxic effect of 6-OHDA showed a latero-medial gradient; the lateral being more vulnerable than medial (fig. 1C). In CPDL, TH-immunoreactivity was completely destroyed in all cases, in both the SIB(-) and SIB(+) groups (fig. 1, C and D). In the case of SIB(-) of figure 1C, the destruction extended to the CPVL and CPPV, but in the CPM, TH-IR terminals were moderately spared. In the case of SIB(+) of figure 1D, TH-IR terminals were almost completely destroyed in all parts. Densitometric analysis of the TH-immunostaining in the SIB(-) group demonstrated 91, 86 and 70% reduction in CPDL, CPVL and CPM, respectively (fig. 2). In the SIB(+) group, TH-staining intensity shows 97, 94 and 91% reduction in CPDL, CPVL and CPM, respectively. Dopamine content was analyzed in micropunched tissue sections of the rostral, central and caudal levels of the dorsolateral part of the caudate-putamen by HPLC. In the control, the concentration of dopamine was the highest in the rostral, next in the central, and least in the caudal (table 1). At all levels, dopamine content decreased more than 90% (P < .05) in the SIB(-) group, and more than 98% in the SIB(+) group (P < .05) compared to the control (table 1).

View larger version (93K): [in this window] [in a new window]   Fig. 1.   TH immunocytochemistry in dopaminergic terminal areas in the striatum in control (A), SIB(-) (C) and SIB(+) groups (D). B represents anatomical loci of the striatal regions. The caudate-putamen is divided into four parts; the periventricular part (CPPV), the dorsolateral part (CPDL), the medial part (CPM) and the ventrolateral part (CPVL). See further details in the text. ac, anterior commissure; AcbC, nucleus accumbens core; AcbS, nucleus accumbens shell; cc, corpus callosum; ICj, insula of Calleja; mfb, medial forebrain bundle; TuO, tuberculum olfactorium. Bars = 1 mm.

View larger version (27K): [in this window] [in a new window]   Fig. 2.   Regional differences of the surviving dopaminergic terminals in the striatum after neonatal 6-OHDA treatment. Each value represents mean ± S.E.M. of arbitrary unit of optical density. The optical density of TH-staining in the locus ceruleus and the corpus callosum is shown in arbitrary units from the minimum of 0 and maximum of 10. For further details, see "Materials and Methods." Data were obtained in five rats of each group. For these analyses, we used sections at the coronal level of Bregma +1.0. For subdivision of the caudate-putamen, see figure 4. CPDL, dorsolateral part of caudate-putamen; CPVL, ventrolateral part of caudate-putamen; CPM, the medial part of caudate-putamen; CPPV, the periventricular part of the caudate-putamen. AcbC, nucleus accumbens core; AcbS, nucleus accumbens shell; TuO, tuberculum olfactorium; ICj, insula of Calleja. *P < .05 when compared to control group. P < .05 when compared to SIB(-) group.

View this table: [in this window] [in a new window]   TABLE 1 Concentration of dopamine in the control, SIB() and SIB(+) groups

View this table: [in this window] [in a new window]   TABLE 2 [3H]SCH-23390 binding and [3H]YM-09151-2 binding in the control, SIB() and SIB(+) groups

Ventral Striatum. In the nucleus accumbens of control, dense TH-IR terminals were detected in the whole part of this nucleus (fig. 1A). The density of immunostaining was slightly weaker around the anterior commissure, corresponding to the accumbens core, than surrounding part, corresponding to the accumbens shell. In the SIB(-) group, TH-IR terminals were slightly destroyed in some parts, but most were intact (fig. 1C). However, in the SIB(+) group, TH-IR terminals were almost completely destroyed in the core part (fig, 1D). In the shell part, immunoreactivity was decreased in some parts, but at its most medial portion adjacent to the lateral septal nucleus, the TH-IR terminals remained intact. Morphometric analysis of the TH-immunoreactivity in the nucleus accumbens core, confirmed the above finding demonstrating only a slight decrease (24%; not significant) in the SIB(-) group, but very high decrease (82%; P < .05) in the SIB(+) group compared with the control (fig. 2). There was a significant difference in this region between the SIB(-) and SIB(+) groups (P < .05). In the accumbens shell, the mean optical density was only 8 and 40% reduction in the SIB(-) and SIB(+) groups, respectively (both are not significant). The HPLC analysis demonstrated 40% reduction of dopamine content in the SIB(-) (not significant) group, and 97% in the SIB(+) group (P < .05) compared with the control (table 1).

Other forebrain areas. TH-IR terminals were examined in the dopamine-innervated cerebral cortex including the medial prefrontal cortex, the suprarhinal cortex of the prefrontal cortex, the anterior cingulate cortex, piriform cortex and entorhinal cortex in which many fine dopaminergic fibers were detected (Björklund and Lindvall, 1984). The central nucleus of amygdala and the lateral septum contained high levels of dopaminergic terminals (Björklund and Lindvall, 1984). There was no prominent change of TH-IR terminals even in the SIB(+) group in all these areas (data not shown). Similar to these immunocytochemical data, the dopamine content by HPLC analysis did not show any significant difference among the three experimental conditions (table 1).

View larger version (108K): [in this window] [in a new window]   Fig. 3.   TH-immunoreactivity in the hypothalamus. A cluster of immunoreactive neurons in A13 area and strong fiber bundles in the lateral hypothalamic area were detected in the control group (A). Hypothalamic dopaminergic neurons in A13, and fibers between A13 and the medial forebrain bundle were not affected in both SIB(-) and SIB(+) groups (B and C). Black arrows show the TH-immunoreactive fibers in the medial forebrain bundle, which were slightly decreased in the SIB(-) group, although severely decreased in the SIB(+) group. The fibers located in the mediodorsal aspect of the internal capsule (white arrows) were severely depleted in both SIB(-) and SIB(+) groups. f, fornix; ic, internal capsule; v, third ventricle. Bars = 250 µm.

Midbrain. As found in the control group (fig. 4, A and B), the mesencephalic TH-IR neurons form an extensive and fairly continuous cell system in the ventral tegmentum. Detailed anatomical evaluation of the ventral tegmentum revealed that they are located predominantly in histologically discrete nuclei including the substantia nigra, the retroruberal area (A8), the nucleus paranigralis, the nucleus parabrachialis pigmentosus, the nucleus interfascicularis and nucleus linearis (rostral and caudal linear nucleus of raphe) (Björklund and Lindvall, 1984; Dahlstöm and Fuxe, 1964; Hökfelt et al., 1984). The nucleus paranigralis and the nucleus parabrachialis pigmentosus which were originally described in the opossum by Tsai (1925) are now put together as the VTA (Oades and Halliday, 1987; Paxinos and Watson, 1986; Phillipson, 1979; Swanson, 1982). In this study, we examined the change of TH-IR neurons in each anatomical region.

View larger version (145K): [in this window] [in a new window]   Fig. 4.   TH-immunocytochemistry in the midbrain dopaminergic areas. Left side panels (A, C and E) show the substantia nigra, the ventral tegmental area, interfascicular nucleus (IF), the rostral linear nucleus (RLi), and right side panels (B, D,F) show the caudal linear nucleus (CLi) and retroruberal area (RR), in the control (A, B), SIB(-) (C, D) and SIB(+) (E, F) groups. The ventral tegmental area is divided into the medial (VTAM) and lateral (VTAL) parts by a dotted line between the lateral edge of the mamillary peduncle (mp) and the medial edge of medial lemniscus. See further details in the text. IP, interpeduncular nucleus; SNC, substantia nigra pars compacta; SNR, substantia nigra pars reticulata; SNL, substantia nigra pars lateralis. Bars = 500 µm.

View larger version (24K): [in this window] [in a new window]   Fig. 5.   Number of TH-immunoreactive cells in the midbrain following neonatal 6-hydroxydopamine treatment. Each value represents the mean ± S.E.M. of TH-immunopositive neurons observed in a section from five rats in each group. We used sections at the level of Bregma -6.0 for the raphe nucleus between the bilateral medial longitudinal fasciculus (IMLF), caudal linear nucleus of raphe (CLi) and retroruberal area (RR). For other regions, we used coronal sections at the level of Bregma -5.6. VTA was divided into the medial part (VTAM) and lateral part (VTAL) as shown in fig. 1. RLi, rostral linear nucleus; SNC, substantia nigra pars compacta; SNR, substantia nigra pars reticulata; SNL, substantia nigra pars lateralis. *P < .05 when compared to control group. P < .05 when compared to SIB(-) group.

Dopamine Receptor Binding

Dorsal striatum, ventral striatum and other forebrain areas. In the caudate-putamen, nucleus accumbens and tuberculum olfactorium of the control group, the binding densities of both [³H]SCH-23390 and [³H]YM-09151-2 were high. For the caudate-putamen, we measured their densities in the rostral, central and caudal levels, and at the former two levels, we further compartmentalized the region into CPPV, CPDL, CPM and CPVL, similar to the TH-immunocytochemical study. In both the SIB(-) and SIB(+) groups, the [³H]SCH-23390 and [³H]YM-09151-2 binding density at each compartment showed a similar level as that in the control group (table 2). For [³H]YM-09151-2 binding at the central level of the caudate-putamen, however, the binding density tended to be the highest in the SIB(+) group, followed in decreasing order by the SIB(-) and control group, but the differences were not statistically significant. Neonatal 6-OHDA treatment did not alter the [³H]SCH-23390 and [³H]YM-09151-2 binding density in the nucleus accumbens or tuberculum olfactorium in either the SIB(-) or SIB(+) group (table 2).

Midbrain. The substantia nigra in the midbrain is a unique area in which positive bindings of [³H]SCH-23390 and [³H]YM-09151-2 are separately distributed: pars reticulata expressed a high density of [³H]SCH-23390 binding without any [³H]YM-09151-2-positive binding, while pars compacta expressed a moderate density of [³H]YM-09151-2 binding with a very low level of [³H]SCH-23390 binding. In the substantia nigra pars compacta, the binding density of [³H]YM-09151-2 binding was highest in the control, followed in decreasing order by the SIB(-) and SIB(+) groups (table 2). The differences between the SIB(-) and control group or between the SIB(+) and control group were significant (P < .05), but the difference between the SIB(-) and SIB(+) groups was not significant (table 2). In VTA, [³H]YM-09151-2 binding was highest in the control group, followed in decreasing order by the SIB(-) and SIB(+) groups. However, only the difference between control and SIB(+) was significant (P < .05).

View larger version (K): [in this window] [in a new window]   Fig. 6.   Image graphics of D1 binding labeled by [3H]SCH-23390 in the substantia nigra pars reticulata. In the control group, strong bindings was detected in the substantia nigra pars reticulata, but very weak in other adjacent parts. In both the control and SMB(-) groups, the binding density in the lateral part of the pars reticulata was weaker than that in the medial part. In the SMB(+) group, the binding density increased in all regions, and the mediolateral gradient was not detected.

View larger version (18K): [in this window] [in a new window]   Fig. 7.   Saturation analysis of [3H]SCH-23390 binding to nigral membranes from the control, SMB(-) and SMB(+) groups. Aliquots of 100 µl of membrane suspension were incubated in triplicate with [3H]SCH-23390 at various concentrations of 0.03 to 1.39 nM in a final volume of 1 ml at 22°C for 90 min. Each dot is presented as the mean of these determinations. The KD and Bmax shown in the lower corner represent the mean ± S.E.M. of value from five experiments. Note D1 binding density was significantly higher in the SMB(+) group than in the SMB(-) (P < .05) and control group (*P < .05).

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion Conclusion References

Region-Specific Neuronal Damage after Neonatal 6-OHDA Treatment

This study demonstrates that the vulnerability of dopaminergic neurons to 6-OHDA toxicity is anatomically defined system-specific (Björklund and Lindvall, 1984), because the regional loss of dopaminergic neurons in the midbrain corresponds well with the regional damage of dopaminergic terminals in projection areas. In the caudate-putamen of SIB(-), the medially located dopaminergic terminals in the CPM and CPPV, tended to be more resistant than those laterally located in the CPDL and CPVL. These findings were in good agreement with the previous tracing studies of dopaminergic neuron systems: the lateral part of the substantia nigra project to the lateral part of the caudate-putamen, and the medial part of the substantia nigra to the medial part of the caudate-putamen (Björklund and Lindvall, 1984). In both SIB(-) and SIB(+) groups, the disappearance of dopaminergic terminals in all parts of the caudate-putamen was well in concordance with the high destruction of dopaminergic neurons in the substantia nigra. The severe destruction of dopaminergic terminals in the nucleus accumbens core and the tuberculum olfactorium in the SIB(+) group might correspond to the significant decrease of dopaminergic neuronal number in the lateral part of VTA (see projection studies of Beckstead et al., 1979; Fallon and Moore, 1978, a and b; Swanson, 1982). 6-OHDA resistant dopaminergic fibers in the nucleus accumbens shell, insula of Calleja, amygdala, medial prefrontal cortex and limbic cortical areas in the SIB(+) group might originate from the medial part of VTA, linear nuclei and interfascicular nucleus (Yagita et al., 1992). Although the source of dopaminergic fibers in the insula of Calleja in tuberculum olfactorium were not clarified at present (Björklund and Lindvall, 1984), the survival and hyperinnervation of dopaminergic terminals in the insula of Calleja of SIB(+) suggests that the host dopaminergic cells were located in the midbrain midline nuclei.

The selective TH-IR fiber loss was observed in the lateral hypothalamic area. Ascending dopaminergic and noradrenergic axons run through in the lateral hypothalamic area from the lower brain stem (Hökfelt et al., 1984). The destruction degree of these bundles in each group corresponded with that of mesostriatal dopaminergic projecting systems. The dorsal mesostriatal dopaminergic fibers project along the dorsomedial aspect of the internal capsule. TH-IR fibers found in the dorsomedial aspect of the internal capsule in the controls were not detected both in the SIB(-) and SIB(+) groups, correlating the complete destruction of the nigral dopaminergic neurons both in these groups. TH-IR fibers in the medial forebrain bundle were slightly decreased in the SIB(-) group, and severely decreased in the SIB(+) group. These findings suggest that many of TH-IR fibers passing in this region are ventral mesostriatal dopaminergic axons projecting to the ventral striatum. On the other hand, TH-IR fibers detected between zona incerta and the internal capsule were not altered both in the SIB(-) and SIB(+) groups, indicating that the dorsal catecholamine bundle being mainly composed of noradrenergic fibers was protected in our experimental condition.

The hypothalamic dopamine system [arcuate-median eminence (A12), anterior hypothalamic periventricular area (A14), zona incerta (A13), posterior hypothalamic area(A11)], most of them were known to project inside the hypothalamus except A11, which forms the meso-spinal cord system (Skagerberg et al., 1982), were resistant to 6-OHDA. This indicates that the growth deficits observed in neonatal 6-OHDA treated rats (Smith et al., 1973), may not be caused by the injury of endocrine-regulating dopamine neurons.

It is surprising that the classification of dopaminergic systems by their topographical organization well corresponds to the different class of the sensitivity to 6-OHDA toxicity. Then, why is the 6-OHDA toxicity unique to anatomically classified dopaminergic systems? Because 6-OHDA exerts its toxic effect to dopaminergic neurons by being selectively taken into neurons by a dopamine high uptake system (Jonsson, 1983), the different sensitivities to 6-OHDA among the dopaminergic neuron systems might be in part caused by the activity of dopamine transporter (Demarest and Moore, 1979; George and Van Loon, 1982). A recent in situ hybridization study revealed that dopaminergic neurons in the substantia nigra displayed a higher level of dopamine transporter mRNA than those in the VTA regions (Blanchard et al., 1994; Shimada et al., 1992) and in the hypothalamus (Cerrutti et al., 1993). Intracellular proteins having calcium chelating activity (e.g. calbindin 28k), and growth factors are proposed as another factor for determining the vulnerability of dopaminergic neurons to 6-OHDA toxicity (Gaspar et al., 1993; Lavoie and Parent, 1991). In the MPTP-induced primate model of Parkinsonism, calbindin 28k expressing neurons were known to be resistant to MPTP-toxicity (Burns et al., 1983). Trophic factors such as basic fibroblast growth factor were found more in VTA neurons may also be a possible protecting factor for 6-OHDA-toxicity (Tooyama et al., 1992).

Interestingly, the dopamine content of the caudate-putamen in the HPRT gene-deleted mouse was about half of that in the intact mouse, although that of nucleus accumbens was unaltered (Jinnah et al., 1993). This suggests that the system-specific impairment of dopamine neurons occurred in HPRT-deleted rats, although no apparent cell death of midbrain neurons was found by TH immunocytochemistry. Although the genes of affected neurons were undiscovered, homozygous weaver mutation shows similar region-specific dopaminergic cell-death (Graybiel et al., 1990; Triarhou et al., 1988), where the dorsal mesostriatal system was more vulnerable than the ventral mesostriatal system.

Self-Injurious Behavior and Brain Dopaminergic Neuron System

The relationship between the region specific damage of dopaminergic neuronal systems and the L-DOPA induced SIB is summarized in a diagram (fig. 8). In this study, SIB(+) rats show almost complete destruction of dorsal mesostriatal and ventral mesostriatal dopaminergic systems, with intact mesocortical and hypothalamic dopaminergic systems. The difference of SIB(+) and SIB(-) lies in the presence or absence of severe destruction of the VTA-ventral striatal (ventral mesostriatal) dopaminergic system. Thus this region may be one of the key regions, for the SIB.

View larger version (29K): [in this window] [in a new window]   Fig. 8.   Diagram of the dopamine-depleted brain regions and the self-injurious behavior (SIB) in neonatal 6-OHDA-treated rats. In control rats, brain dopamine neuron system consisted of dorsal mesostriatal (coarse dotted area), ventral mesostriatal (gray area), mesocortical (black area) and hypothalamic systems (black area). In the neonatal 6-OHDA-treated SIB(-) group, the dorsal mesostriatal system (coarse dotted area) was severely impaired, but other dopamine systems were intact or only partially impaired. In neonatal 6-OHDA-treated SIB(+) group, not only dorsal mesostriatal system (coarse dotted area), but also ventral mesostriatal system (gray area) were almost completely lost, and blackened mesocortical and hypothalamic dopaminergic systems survived. A11, dopaminergic neuronal group located in the posterior hypothalamus; A12, dopaminergic neuronal group located in the hypothalamic arcuate nucleus projecting to the median eminence; A13, dopaminergic neuronal group in the zona incerta; A14, dopaminergic neuronal group in the periventricular anterior hypothalamus; Acb, nucleus accumbens; Amy, amygdala; CLi, caudal linear nucleus of raphe; CP, caudate-putamen; EntC, entorhinal cortex; ICj, insula of Calleja; IF, interfascicular nucleus; IMLF, raphe nucleus between the bilateral medial longitudinal fasciculus; PFC, prefrontal cortex; PirC; piriform cortex; RLi, rostral linear nucleus of raphe; RR, retroruberal area (A8); SL, septum lateralis; SN, substantia nigra (A9); TuO, tuberculum olfactorium; VTA, ventral tegmental area.

However, it should be noted that we do not deny the possibility that the region showing slight differences in the amount of dopamine is important for the SIB. Because the degree of disturbance of dopamine neurotransmission in weakly dopamine-depleted area remains unknown, it is not possible to define the specific site from greater dopamine reduction, as the definitive explanation for the ability L-DOPA to induce SIB in 6-OHDA-treated rats. However, our study demonstrating the region specific alternation of dopamine systems in SIB(+) rats, offers first a frame of abnormal dopamine neuronal circuit of SIB. At least, it is safe to say that dopamine neurotransmission might be impaired in severely dopamine-depleted area, and the impaired dopamine neurotransmission in that area play some role for the SIB manifestation.

The dorsal and ventral striatum may involve different kinds of behavior; e.g., the nucleus accumbens is important for the locomotion, and caudate-putamen has major roles in oral stereotypy such as biting and licking (Arnt, 1985; Kelly et al., 1975). These differences may be closely related to the anatomical difference of the inputs from the cortical structures to these striatal areas. The function of the ventral and dorsal striatum was the process of cortical information under the influence of dopaminergic stimulation, and relay to the pallidal structures. Although the dorsal striatum is associated with extensive and highly organized projections from the neocortex, the ventral striatal areas are targets of projections from limbic, cortical and subcortical regions. For processing the sensorimotor information, the dorsal striatum output flows to the globus pallidus-subthalamic nucleus-substantia nigra pars reticulata-thalamus- (ventrolateral and ventromedial) sensorimotor cortices (Alexander and Crutcher, 1990; Hattori et al., 1975). There is a direct pathway from the dorsal striatum to the substantia nigra pars reticulata. For processing the limbic information, output from ventral striatum flows to the ventral pallidum-thalamus-(mediodorsal) prefrontal cortex (Mogenson et al., 1983; Newman and Winans, 1980a; Newman and Winans, 1980b). The impairment of dorsal striatal dopamine system in SIB(-) suggests that the sensorimotor information processing alone is not important for the SIB. The loss of dopaminergic neurons in both dorsal and ventral striatal systems in the SIB(+) rats may influence both sensorimotor and limbic cortical information processing, and these impairments may induce the characteristic behavior of SIB.

The problem whether the sole impairment of ventral mesostriatal system or the both impairment of ventral mesostriatal and dorsal mesostriatal systems important for the induction of SIB, remains unsolved. Goldstein et al. (1986) demonstrated that VTA-lesioned monkey manifest SIB. This primate study suggests the importance of ventral mesostriatal dopaminergic system for the induction of SIB. However, whether primate SIB occurs by the same mechanism as in rats, is unknown.

Alternation of D₁ and D₂-Like Receptor Binding after Neonatal 6-OHDA Treatment

Does the above region-specific presynaptic dopaminergic neuronal damage induce the postsynaptic up-regulation of dopaminergic receptors, and finally connected to SIB? To test the hypothesis, we compared the expression of dopaminergic receptor in SIB(-) and SIB(+) compared to controls. Caudate-putamen is the region that many investigators have studied to characterize the unique behaviors of the neonatal 6-OHDA treated rats. Although dopamine terminals were very severely depleted in most compartments of the caudate-putamen in both the SIB(-) and SIB(+) groups, the present autoradiographic study demonstrated no alternation of D₁- and D₂-like receptor binding. Since no alteration was detected in any compartments of caudate-putamen, the possibility of the compartment specific up-regulation of receptors by neonatal 6-OHDA treatment was also denied in our study. Previously, however, there are conflicting data concerning the effect of neonatal depletion of dopamine on striatal D₁ receptors. Gelbard et al. (1990) and Radja et al. (1993) reported that the treatment of 6-OHDA into neonatal rats induced the reduction of the number of striatal D₁-like receptors, which was also found in SCH-23390-treated rats during neonatal age (Kostrzewa and Saleh, 1989; Saleh and Kostrzewa, 1988). However, most other investigators have reported no change in the striatal D₁-like receptor bindings after intracisternal 6-OHDA-induced neonatal destruction of dopaminergic neurons in adults either by binding assay (Breese et al., 1987a; Caboche et al., 1991; Dewar et al., 1990; Duncan et al., 1987; Leslie et al., 1991; Luthman et al., 1990) or autoradiography (Duncan et al., 1993). The findings were also confirmed from the level of mRNA showing no alteration of D_1A receptor mRNA (Duncan et al., 1993). Our findings were in concordance with the latter majority studies, indicating that the dopaminergic depletion did not influence the normal development of D₁-like receptors in the caudate-putamen.

For D₂-like receptors in the caudate-putamen, Radja et al. (1993) found the significant increase of [³H]raclopride binding in most parts of the caudate-putamen of neonatal 6-OHDA-lesioned rats in comparison with those of the control, although the D₂-like receptor elevation was not detected at the mRNA levels. This finding may not appear to be in concordance with our finding. However, considering that our statistical analysis was performed among three behaviorally characterized groups, and that the tendency of binding density was highest in the SIB(+) group, followed in decreasing order by the SIB(-) and the control group in each compartment of the caudate-putamen, the difference may depend on the evaluation of the experimental results. Even though the elevation of D₂-like receptor binding had occurred after neonatal 6-OHDA treatment, this elevation is not related to the occurrence of SIB.

Because the depletion level of dopaminergic fibers in the ventral striatum correlates the occurrence of SIB, we expect the difference of binding level of D₁- and D₂-like receptors between SIB(-) and SIB(+). However, we found none of the compartments of the nucleus accumbens and the tuberculum olfactorium showed unique changes in D₁ and D₂ binding in the SMB(+) group compared to the SMB(-) and control groups. Although the phenomenon can be interpreted in many senses, it will indicate that the ventral striatum may not be the primary D₁ target area important for the induction of SIB after the systemic injection of L-DOPA or D1 agonist.

This is the first study to report the level of the dopamine receptors in the cerebral cortex, hippocampus, amygdala, lateral septum and hypothalamus in the neonatal 6-OHDA dopamine lesioned animals. Because the specific change of dopaminergic neurotransmission unique to SIB(+) was not detected in these areas, the mesocortical and hypothalamic dopamine neuron systems may not play a dominant role for SIB. However, the indirect dopamine-depletion effect through the activation of postsynaptic peptidergic neurons in the striatum (Barush et al., 1988; Matsumoto et al., 1992; Normand et al., 1988; Voorn et al., 1987) is possible to characterize the behavioral deficit such as the impaired acquisition of an operant response (Heffner and Seiden, 1983).

The D₂ binding activity was decreased in the substantia nigra pars compacta in the both SIB(-) and SIB(+) groups, and in VTA in the SIB(+) group. The decrease of D₂ bindings was in proportion to the loss of TH-immunoreactive neurons. Thus, the loss of D₂ binding in these regions is not astonishing, because it indicates that D₂-like receptors expressed in midbrain dopaminergic neurons as autoreceptor were severely lost in SIB(+).

Upregulation of Nigral D₁-Like Receptors and Self-Injurious Behavior

Unexpectedly, dopamine D₁-like receptors demonstrated unique alteration in the midbrain of SIB(+) rats. [³H]SCH-23390 binding density in the SNR was higher, in the SIB(+) group than in the SIB(-) or the control group. D₁-like receptor binding in the SNR has been reported to be located on the presynaptic terminals of strionigral pathway (Aiso et al., 1987; Altar and Hauser, 1987; Harrison et al., 1990; Porceddu et al., 1986; Savasta et al., 1986). Dopamine released from dendrites of dopaminergic neurons located in the substantia nigra pars compacta and VTA, may act on these D₁-like receptors in SNR (Cheramy et al., 1981). The Scatchard analysis on membrane fraction of the nigral tissue revealed the increase of B_max and no change of K_D in the [³H]SCH-23390 binding in the SIB(+) group. This finding suggests that, in the SIB(+) group, D₁-like receptor at the strionigral terminals increased in number without changing its affinity. In the SIB(+) group, the dopaminergic neurons in the substantia nigra pars compacta and the lateral part of the VTA projecting dendrites to SNR were almost completely destroyed, suggesting that dopaminergic dendritic release are severely reduced in SNR. Thus, the phenomenon of the increase in the number of D₁-like receptors could be reactive up-regulation in compensation for less supply of the dendritic dopaminergic release.

The D₁-like receptor binding in SNR of the normal rat had the distinct medio-lateral gradation; higher in the medial than the lateral. In the SIB(+) group, the increase of the density was more prominent in the lateral part than the medial part of SNR, and as a consequence, the D₁ binding density was very high in this area. It has been reported that the D₁ projection to the lateral part of SNR is mainly originated from the cells located in the caudal part of the caudate-putamen (Altar and Hauser, 1987). It is interesting to note that the caudal part of the caudate-putamen is the only subdivision among 11 striatal subdivisions that showed the tendency of the increase of [³H]SCH-23390 binding were detected in SIB(+) group as compared with other groups, although the increase is not statistically significant. The reason why the value does not reach significant level may be that the population of striatal neurons projecting to the substantia nigra was too minor to detect the D₁ binding change at gross macroscopic autoradiographic level, although it is still possible that the expression of D₁-like receptor in strionigral neurons were only confined to nerve terminals in the substantia nigra, not at the cell body or dendrites in the striatum. In many cases, an in situ hybridization study examining the mRNA at the cellular level in the striatal tissue may further help to understand this event.

The mechanism of the region-specific up-regulation of D₁-like receptors in the strionigral pathway is obscure at present. Striatal neurons are speculated to receive inputs from substantia nigra (dopaminergic), the cerebral cortex (glutaminergic) and raphe dorsalis (serotonergic) (Graybiel and Ragsdale, 1983). It is possible that region-specific destruction of nigrostriatal dopaminergic terminals influence the level of D₁-like receptors of strionigral neurons. However, the degree of the destruction of dopamine terminals was uniform and almost complete in the caudate-putamen of SIB rats. The hyperinnervation of serotonergic terminals was observed in the caudate-putamen in neonatal 6-OHDA-treated rats (Descarries et al., 1992; Luthman et al., 1987; Snyder et al., 1986), although the difference among striatal regions has not been reported before. No studies on region-specific change of glutaminergic transmission affecting the strionigral pathway in the neonatal dopamine-depleted rats have been performed yet. Intrinsic peptidergic systems in the caudate-putamen are also candidates of this D₁ up-regulation. Enkephalin synthesis is up-regulated and substance P synthesis is down-regulated in 6-OHDA-treated rats (Sivam, 1989; Sivam et al., 1987; Sivam and Krause, 1990). Although region-specific difference of peptidergic neurons have also not been reported as yet, it is possible that the combination of these extrinsic and intrinsic factors causes the region-specific up-regulation of D₁-like receptors in strionigral neurons.

How was the up-regulated D₁-like receptor linked to the SIB? Up-regulation of D₁ binding activity in SNR has also been reported in cases of chronic nigral administration of D₁ antagonist SCH-23390 (Porceddu et al., 1985) and systemic subchronic administration of methamphetamine (Ono and Fukuda, 1984). The former accompanies the increase of B_max, but the latter is associated with decreased K_D. The above findings suggest that the D₁ up-regulation in SNR is not especially associated with the peculiar characteristics of the behavior, SIB.

The stimulation of D₁-like receptors expressed on the strionigral GABAergic terminals in the SNR is supposed to modulate GABA release from strionigral terminals. However, the role of nigral D₁-like receptors on the GABA release is still controversial: an electrophysiological study suggests its negative role (Martin and Waszczak, 1994; Waszczak and Walters, 1986), and slice-release and behavioral studies suggest its positive role (Reubi et al., 1977; Robertson and Robertson, 1989). A recent electrophysiological evidence by Cameron and Williams (1993), however, suggests that D1 agonists release GABA in the VTA. However, GABA released from phasically acting strionigral neurons suppresses the activity of tonic acting SNR neurons (Martin and Waszczak, 1994), which connect to the extrapyramidal output system in superior colliculus and brainstem nuclei. Hyperlocomotor activity has been reported to be associated to the increased sensitivity to GABA in SNR in neonatal 6-OHDA-treated rats (Breese et al., 1987b). More interestingly, microinjection of a GABA agonist, muscimol into SNR could induce SIB (Baumeister and Frye, 1984; Breese et al., 1987b). Stimulation of D₁ or GABA receptor may also be linked to various behaviors other than SIB, and could not determine the peculiar characteristics of the SIB. However, the increased neuronal outflow of SNR neurons to extrapyramidal system may be an important "driving force" of SIB.

	Conclusion

Top Abstract Introduction Materials & Methods Results Discussion Conclusion References

Our study has demonstrated that the neonatal 6-OHDA treatment induces changes of dopaminergic neurotransmission at both presynaptic nerve terminal and postsynaptic receptor levels. In brain dopamine neuron systems, the 6-OHDA toxicity was system specific, and the region-specific damage of central dopaminergic neuron system well correlated with the induction of SIB. At receptor level, this is the first study reporting the region-specific up-regulation of D₁-like receptors in the brain of SIB model rats with neonatal 6-OHDA treatment, and the finding indicates the up-regulated D₁-like receptor in SNR as a candidate locus for the induction of SIB after L-DOPA or dopamine D₁ agonist loading.