Expression of ciliary neurotrophic factor receptor-α messenger RNA in neonatal and adult rat brain: an in situ hybridization study

Pergamon

PII:

Neuroscience Vol. 77, No. 1, pp. 233–246, 1997 Copyright ? 1997 IBRO. Published by Elsevier Science Ltd Printed in Great Britain 0306–4522/97 $17.00+0.00 S0306-4522(96)00476-9

EXPRESSION OF CILIARY NEUROTROPHIC FACTOR RECEPTOR-á MESSENGER RNA IN NEONATAL AND ADULT RAT BRAIN: AN IN SITU HYBRIDIZATION STUDY M.-Y. LEE*, H.-D. HOFMANN and M. KIRSCH† Institute of Anatomy, University of Freiburg, P.O. Box 111, D-79001 Freiburg, Germany Abstract––Ciliary neurotrophic factor is a pleiotropic molecule thought to have multiple functions in the developing and adult nervous system. To investigate the role of ciliary neurotrophic factor in the developing and mature brain by defining putative target cells the expression of the ligand-binding á-subunit of the ciliary neurotrophic factor receptor was studied in neonatal and adult rat brains using a digoxygenin-labelled probe for in situ hybridization. Neuronal populations expressing ciliary neurotrophic factor receptor-á messenger RNA were found in many functionally diverse brain areas including the olfactory bulb (mitral cells and other neurons), neocortex (layer V) and other cortical areas (pyramidal cell layers in the piriform cortex and hippocampus, granule cell layer of the dentate gyrus) and distinct nuclei in the thalamus, hypothalamus and brainstem. In the latter, reticular nuclei and both cranial motor and sensory nerve nuclei showed intense hybridization signals in the neonatal brain. The nucleus ruber, substantia nigra pars reticularis, deep cerebellar nuclei and a subpopulation of cells in the internal granular layer of the cerebellum were also labelled. In many areas (e.g., in thalamic, midbrain and pontine nuclei) ciliary neurotrophic factor receptor-á expression became undetectable with maturation; however, there were other areas (e.g., olfactory bulb, cerebral cortex and hypothalamus) where expression was higher in the adult. The neuroepithelium of the neonatal rat displayed a highly selective expression of ciliary neurotrophic factor receptor-á in areas which are known to exhibit high rates of postnatal cell proliferation in the germinal zones. Generally, neurons which have been reported to respond to exogenous ciliary neurotrophic factor were labelled by the ciliary neurotrophic factor receptor-á probe. This was not the case, however, for striatal and septal neurons. The results of this study suggest that ciliary neurotrophic factor receptor-á ligands have even broader functions than previously thought, acting on different neuronal populations in the developing and mature brain, respectively. Copyright ? 1997 IBRO. Published by Elsevier Science Ltd. Key words: neuroepithelium, olfactory bulb, cortex, diencephalon, brainstem, ciliary neurotrophic factor receptor-á.

Ciliary neurotrophic factor (CNTF) was originally purified, characterized and cloned on the basis of its survival-promoting activity for embryonic chick ciliary ganglion neurons.6,30,51 Subsequent studies demonstrated that CNTF has a much broader spectrum of target cells and that the action of CNTF is not restricted to the promotion of neuronal survival (see Refs 36 and 47 for reviews). In vitro, the protein has been shown to exert trophic influences on developing sensory, sympathetic and parasympathetic neurons of the peripheral nervous system.47 Central neurons which respond in vitro to CNTF by improved survival include spinal motoneurons,4,37 spinal and corticospinal neurons,33,35 septal neurons,25 various types of hippocampal neurons23 and retinal *M.-Y. L. is on leave from the Department of Anatomy, Catholic University Medical College, Seoul, Korea †To whom correspondence should be addressed. Abbreviations: CNTF, ciliary neurotrophic factor; CNTFRá, CNTF receptor-á; P, postnatal day; PB, phosphate buffer; PCR, polymerase chain reaction; SSC, standard saline citrate; TBS, Tris-buffered saline.

ganglion cells.28,40 In vivo, CNTF has been shown to reduce the programmed cell death of motoneurons in chick embryos.42 CNTF also influences phenotypic properties in a variety of immature neurons, indicating multiple functions of this protein during development. It stimulates the expression of transmittersynthesizing enzymes in cultures from embryonic spinal cord, substantia nigra, locus coeruleus and retina 22,31,33,34 and regulates the expression of the rod photoreceptor phenotype in retinal cultures.17,26 In addition, CNTF markedly reduces the proliferation of sympathetic precursor cells16 and induces changes in the transmitter phenotype in these cells.46 Moreover, it has also been shown to act on cultured glial cells and their precursors affecting the generation, maturation and survival of oligodendrocytes and astrocytes.7,29,39 Besides these effects on developing cells, systemic or local administration of CNTF ameliorates cell loss or degeneration of injured neurons in the facial nucleus of neonatal rats48 and in the substantia nigra, septum, anterior thalamus and retina of adult

233

234

M.-Y. Lee et al.

animals12,19,20,27 and prevents progression of motoneuron diseases in mouse mutants.41,48 These effects, together with the substantially higher levels of CNTF found in peripheral nerves and different brain regions of adult rats than in immature stages have led to the suggestion that CNTF plays a role as a lesion factor in the mature nervous system. The CNTF protein is distantly related to leukemia inhibitory factor, interleukin-6 and other cytokines.10 Its biological effects are mediated by a tripartite receptor complex consisting of a CNTF-specific ligand binding á-subunit (CNTFRá) and two signal transducing components (gp130, leukemia inhibitory factor receptor â) which are identical to those of other cytokine receptors.13,18,21,25,49 Expression of CNTFRá has been studied by northern blot analysis and radioactive in situ hybridization in embryonic and adult rats.24 Receptor mRNA is detectable from embryonic day 9 onward and in adult animals it is widely distributed in the nervous system. Identified CNTFRá-expressing cells include the known CNTFresponsive peripheral and spinal neurons and lower and upper motoneurons of the brain. In situ localization of CNTFRá in areas of neuronal precursor cell proliferation and differentiation in embryonic rats was in agreement with the observed in vitro effects on developing neurons of different origin, indicating an important role for CNTF during development. This conclusion was supported by the recent demonstration of severe motoneuron deficits in newborn mutant mice lacking CNTFRá.14 The latter observation is particularly interesting in view of contrary results obtained with mice lacking endogenous CNTF which appear to develop normally,14,38 while CNTFRádeficient mice die shortly after birth. This discrepancy was interpreted as indicating the existence of a second CNTF receptor ligand which is responsible for CNTFRá-mediated effects on developing neurons.49 Thus, there is substantial evidence that CNTFRá has essential functions during development and in the adult nervous system for a variety of neuronal cell types. In the present in situ hybridization study a digoxygenin-labelled probe was used to analyse in more detail the cellular distribution of CNTFRá mRNA in the brain of neonatal and adult rats. EXPERIMENTAL PROCEDURES

Riboprobes were generated by in vitro transription of CNTFRá-specific polymerase chain reaction (PCR) products. Total RNA was isolated from adult retina according to Chomczynski and Sacchi11 and 1 µg was reverse transcribed with 200 units of M-MLV reverse transcriptase (Gibco) and oligo (dT)12-18 (Pharmacia). CNTFRá was amplified with primers selected from the published sequence of rat CNTFRá24 to produce a 350 bp fragment (bp 681–1029). The product was purified from an agarose gel and reamplified with a reverse primer to introduce the T7-polymerase promoter sequence. After purification, 1 µg of the resulting cDNA was used for in vitro transcription with T7-polymerase and digoxygenin-labelled UTP according to the manufacturer’s instructions (Boehringer

Mannheim). The same strategy was applied to generate the sense-stranded riboprobe. Probes were used at dilutions of 1:2000–6000. In situ hybridization was performed on coronal or sagittal brain sections from three newborn and five adult Sprague– Dawley rats (Charles River, Germany). Animals were deeply anaesthetized with nembutal (50 mg/kg body weight) and perfused transcardially with phosphate-buffered saline (pH 7.2) followed by 4% paraformaldehyde in 0.1 M phosphate buffer (PB; pH 7.2). Brains were dissected and postfixed in the same fixative for 2 h at room temperature. The tissue was equilibrated with 30% sucrose in PB and frozen in isopentane. Cryostat sections (30 µm) were cut and processed free floating for hybridization. Samples from newborn and adult animals were treated simultaneously to ensure identical hybridization conditions. Adjacent sections were stained with Cresyl Violet to determine the location of the sections.43,44 Appropriate sections were collected and washed twice with 2# standard saline citrate (SSC; 20# SSC: 3 M NaCl, 0.3 M sodium citrate, pH 7.0). Prehybridization was performed at 57)C for 2 h in hybridization solution containing 50% formamide, 250 µg/ml denatured salmon sperm DNA, 100 µg/ml yeast tRNA, 0.05 M sodium phosphate (pH 7.0), 4# SSC, 5% dextran sulphate and 1# Denhardt’s solution. Sections were then incubated overnight at 57)C with probes diluted (1:6000) in hybridization solution and subsequently washed with 2# SSC at room temperature followed by successive washes at 67)C in prewarmed 2# SSC, 2# SSC/50% formamide, 0.1# SSC/50% formamide and 0.1# SSC for 30 min each. After several rinses in Tris-buffered saline (TBS; 0.15 M NaCl, 0.1 M Tris, pH 7.5) and incubation in blocking solution (Boehringer Mannheim) the tissue was treated with alkaline phosphatase-conjugated sheep anti-digoxygenin antibody (1:2000; Boehringer Mannheim) overnight at 4)C. Following several rinses in TBS and equilibration in TBS/50 mM MgCl2 (pH 9.5) the enzyme reaction was performed in the dark for 8 h using 4-nitroblue tetrazolium chloride (0.35 mg/ml) and 5-bromo-4-chloro-3-indolyl phosphate (0.18 mg/ml) as substrates. Sections were air-dried on slides, mounted in Kaiser’s gelatin and viewed with either a Wild microscope or a Zeiss Axiophot equipped with Nomarski optics. RESULTS

CNTFRá gene expression was visualized using a digoxygenin-labelled riboprobe corresponding to a 350 bp sequence of rat CNTFRá. The PCR products used to produce cRNA probes by in vitro transcription were sequenced to ensure their correct structure. In situ hybridization with digoxygenin-labelled antisense-strand probe rendered consistent and reproducible labelling in three brains from postnatal day 1 (P1) rats and five brains from adult rats. CNTFRá mRNA was localized exclusively in the cytoplasm of neurons. Hybridization with sensestrand probe, even when used at three-fold higher concentrations, never resulted in detectable cellular labelling. Therefore, these controls are demonstrated only for selected brain areas (see Fig. 1B,G, 3C). Expression of CNTFRá is described in a rostral-tocaudal sequence. Olfactory bulb In neonatal rats comparison of cryostat sections of the olfactory bulb hybridized to antisense-strand

CNTFRá mRNA in rat brain

Fig. 1. CNTFRá in situ hybridization in coronal sections through the main olfactory bulb and telencephalon of neonatal and adult rat. (A) P1 main olfactory bulb; labelling is present in the mitral cell layer (Mi) and the neuroepithelium (arrowheads) of olfactory ventricle (OV); Gl, glomerular layer; IGr, internal granular layer. (B) Section as in (A) hybridized to sense-stranded probe. (C) P1 telencephalon sectioned at the level of the anterior commissure (ac) showing a moderate signal in layer V of the cerebral cortex (triangles) and in the piriform cortex (arrowheads) and a very prominent signal in the dorsolateral aspect ventricular zone of the lateral ventricle. CPu, caudate putamen; MS, medial septum. (D) Higher magnification view of the boxed area in (C). Labelling is restricted to the columnar cells of the striatal ventricular zone (Vz); SVz, subventricular zone; ChP, choroid plexus. (E) Coronal section through the adult main olfactory bulb showing intensely labelled neurons in the mitral cell layer (Mi) and external plexiform layer (EPl); juxtaglomerular cells in the glomerular layer (Gl) are less strongly labelled; IGr, internal granular layer. (F) Higher magnification of the section in (E). (G) Section as in (F) but hybridized to sense probe. (H) Low-magnification view of the sectioned adult forebrain. Strongly labelled cells are visible in layer V of the cerebral cortex (0) and the pyramidal cell layer of the piriform cortex (4). No hybridization signal is observed in the caudate putamen (CPu) and medial septum (MS). Dark spots in the striatum represent cross-sectioned fibre tracts. LV, lateral ventricle. (I) Higher magnification view of the neocortex. (J) High-power view of layer V. Scale bars=300 µm (A–C,H), 200 µm (E–G), 50 µm (D,I,J).

235

236

M.-Y. Lee et al.

Fig. 2. CNTFRá expression in the rat diencephalon and hippocampus. (A) Coronal section of a P1 rat brain demonstrating hybridization signals in thalamic nuclei (VP indicates the location of the ventral posterolateral and posteromedial nuclei), the hippocampal formation and the medial habenular nucleus (boxed area). Strong labelling is present in the dorsolateral ventricular zone of the lateral ventricle (LV; 4); the ventricular zone of the basal third ventricle is negative (0); ic, internal capsule. Inset: Higher magnification view of the labelled medial habenular nucleus (MHb) and the adjacent neuroepithelium in the roof of the third ventricle (3V). Note the sharp boundary separating a labelled dorsal and an unlabelled ventral part of the neuroepithelium in the habenular region (arrowheads); LHb, lateral habenular nucleus. (B) Coronal section of the adult brain. Areas expressing CNTFRá at this level include the mediodorsal thalamic nucleus (upper boxed area), the supraoptic nucleus (lower boxed area), the paraventricular hypothalamic nucleus (Pa) and layer V of the cerebral cortex (triangles). Note the absence of labelled cells in other thalamic nuclei (asterisks). LV, lateral ventricle; 3V, third ventricle; CPu, caudate putamen; ic, internal capsule. (C) High-power view of the upper boxed area in (B) showing the labelled neurons of the mediodorsal thalamic nucleus; sm, stria medullaris thalami. (D) High-power view of the lower boxed area in (B) with intensely labelled neurons of the supraoptic nucleus; ox, optic chiasm. Scale bars=250 µm.

(Fig. 1A) and sense-strand (Fig. 1B) probes revealed specific labelling of differentiating mitral cells which are located concentrically in several rows around the olfactory ventricle at this stage. Mitral cells are the first neurons to be generated in the embryonic olfactory bulb, becoming postmitotic between E12 and E18.8 No hybridization was detectable in the immature glomerular and internal granular layer cells, but there was a strong signal over the ventricular zone of the olfactory ventricle. In the adult olfactory bulb mitral cells still were heavily labelled (Fig. 1E–F). In addition, weak labelling was observed in juxtaglomerular cells and stronger labelling in scattered neurons of the external plexiform layer which, according to their location and size, probably represented tufted cells. This demonstrates that different populations of olfactory bulb neurons express CNTFRá during or after differentiation. Telencephalon A coronal view of the neonatal brain at the level of the anterior commissure is shown in Fig. 1C. At this developmental stage hybridization in the cerebral cortex was relatively weak, but could be clearly

localized to layer V of neocortical areas and of the limbic cortex. Cortical areas of the rhinencephalon were also positive. In the adult brain the distribution of CNTFRá expression in the cerebral cortex had not changed significantly (Fig. 1H–J; see also Fig. 3A), but the intensity of the signal was increased throughout layer V and, more prominent, in neurons of the piriform cortex. Additionally, cells in layer II of the neocortex were labelled slightly above background, but this was not invariably visible in all sections (Fig. 1H). In sections, from adult brain fibre, tracts were of fairly dark appearance, e.g., in the basal ganglia and the anterior commissure (Fig. 1H; see also Fig. 5C). This staining was not due to the presence of specific phosphatase reaction product and was identical in sections hybridized to sense probe (not shown). Labelling in layer V was restricted to a subpopulation of large cells probably representing corticospinal neurons (Fig. 1J) which have been shown to be CNTF responsive.35 Although septal and striatal neurons have been reported to respond to CNTF,19,25 no cellular labelling was observed in the basal ganglia or septal area with our technique, in either neonatal or adult animals (see Figs 1C,H, 2A,B).

CNTFRá mRNA in rat brain

237

Fig. 3. Expression of CNTFRá mRNA in the hippocampal formation of the adult rat. (A) Coronal section at the level of the medial habenular nucleus (MHb) showing a hybridization signal of high intensity in the dentate gyrus, the hippocampus proper and the subiculum. Triangles indicate the labelled layer V neurons in the cortex; 3V, third ventricle; LV, lateral ventricle; cp, cerebral peduncle; Pir, piriform cortex. (B,C) Higher magnification view of the hippocampal formation in sections hybridized to antisense and sense probes, respectively. In (B) label is visible in neurons of the dentate gyrus granular layer, in the pyramidal cell layer of CA3, CA2, CA1 and subiculum. Inset: Scattered neurons in the hilar region (boxed area in (B)) are also labelled. Scale bars=500 µm (A–C), 100 µm (inset).

The hippocampal formation showed massive CNTFRá expression in both neonatal (Fig. 2A) and adult (Fig. 3A–C) rats. Strongly labelled cells were present in the granule cell layer of the fascia dentata and throughout the pyramidal cell layer, including the subiculum. In the adult specific hybridization was also observed in scattered cells of the hilar region (Fig. 3B). In the neonatal animal a very prominent signal was observed over the ventricular zone lining the lateral ventricle. Notably, this signal was confined to distinct regions. In the rostral part of the lateral ventricle it was present only in the lateral and dorsal walls of the ventricle, as demonstrated at higher magnification in Fig. 1D, which also shows the localization of the CNTFRá mRNA in the radially oriented neuroepithelial cells and its absence in the subventricular zone. At more caudal levels CNTFRá expression disappeared in the ventrolateral ventricular wall (Fig. 2A) and in the most caudal part of the lateral ventricle it became restricted to the dorsal neuroepithelium (not shown). Diencephalon Expression of CNTFRá mRNA in the diencephalon, as visualized by in situ hybridization,

markedly changed between P1 and the adult state. After birth many of the thalamic nuclei exhibited moderate but significant and reproducible labelling (Fig. 2A). The nuclei that were found to be positive by inspection of sections from different levels are given in Table 1. They include different classes of thalamic nuclei such as the lateral geniculate body and the mediodorsal and reticular nuclei. In the adult thalamus CNTFRá expression was highly restricted, being detectable only in the dorsomedial extreme of the mediodorsal nucleus (Fig. 2B,C). In contrast, receptor expression increased in the hypothalamus between P1 and adult. Whereas only the supraoptic nucleus was moderately labelled in the neonatal (not shown), the paraventricular hypothalamic nucleus (Fig. 2B,D) and the tuberomammillary nucleus (not shown) were additionally labelled in the adult brain. CNTFRá transcripts were also produced in two other diencephalic regions at both ages studied, the medial habenular and the subthalamic nuclei (Figs 2A, 3A). As in the lateral ventricles (see above) the neuroepithelium of the third ventricle displayed a very distinct, region-specific CNTFRá expression in the neonatal brain. At this stage hybridization was restricted to the dorsal aspect of the habenular region

238

M.-Y. Lee et al. Table 1. Synopsis of the areas expressing ciliary neurotrophic factor receptor-á in neonatal and adult rat brain

Main olfactory bulb Mitral cell layer External plexiform layer Telencephalon Olfactory cortex Taenia tecta Olfactory tubercle Neocortex, layer V Limbic cortex, layer V Piriform cortex, pyramidal cell layer Hippocampus proper, pyramidal cell layer Dentate gyrus, granule cell layer Dentate gyrus, hilar neurons Subiculum, pyramidal cell layer Entorhinal cortex, pyramidal cell layer Diencephalon Thalamus Anterodorsal nucleus Laterodorsal nucleus Ventral posterolateral nucleus Ventral posteromedial nucleus Posterior nucleus Reticular nucleus Mediodorsal nucleus Lateral geniculate body Medial habenular nucleus Supraoptic nucleus Paraventricular nucleus Tuberomammillary nucleus Subthalamic nucleus Brainstem Midbrain Substantia nigra pars reticulata Parabigeminal nucleus Interpeduncular nucleus Microcellular tegmental nucleus Medial terminal nucleus accessory optic tract Oculomotor nucleus Trochlear nucleus Red nucleus Pons Trigeminal motor nucleus Mesencephalic trigeminal nucleus Locus coeruleus Nucleus trapezoid body Ventral nucleus lateral lemniscus Pontine reticular nucleus Reticulotegmental nucleus Caudal linear nucleus Median, paramedian raphe nucleus Medulla Abducens nucleus Facial motor nucleus Accessory facial nucleus Hypoglossal nucleus Vagal motor nucleus Area postrema Vestibular nucleus Dorsal cochlear nucleus External cuneate nucleus Gigantocellular reticular nucleus Lateral reticular nucleus Inferior olive Nucleus ambiguus Spinal trigeminal nucleus Cerebellum Deep nuclei Cortex

Neonatal

Adult

+ "

+ +

(+) (+) (+) (+) (+) + + (+) + +

+ + + + + + + + + +

+ + + + + + + + + + " " +

" " " " " " + " + + + + (+)

+ + + + + + + +

+ + " " " (+) (+) "

+ + (+) + + + + (+) (+)

+ " " " " + " " "

+ + " + (+) ? + ? + + + + + +

+ + + + + + + + + + + " + "

+ +

+ +

The intensity of the hybridization signal is indicated as follows: +, moderate to intense; (+), weak; ", not detectable; ?, not identified.

CNTFRá mRNA in rat brain

239

Fig. 4. CNTFRá mRNA expression in the brainstem of the neonatal rat. (A) Coronal section through caudal midbrain and rostral pons. Labelling is seen in the oculomotor nucleus (3), red nucleus (RMC), reticulotegmental nucleus (RtTg in (C)) and nucleus of the trapezoid body (Tz in (C)). (B) Higher magnification view of the upper boxed area in A demonstrating the unlabelled neuroepithelium of the aqueduct (Aq). (C) Labelled cells in the reticulotegmental nucleus (RtTg) and the nucleus of trapezoid body (Tz) shown at higher magnification (lower boxed area in (A)). (D) Coronal section through the rostral medulla. Moderate labelling is seen in spinal and medial vestibular nuclei (SpVe, MVe). Intensely labelled neurons are present in the inferior olive (IO), spinal trigeminal nucleus (Sp5), nucleus ambiguus (AMb), gigantocellular reticular nucleus (Gi) and other reticular nuclei (arrowheads); 4V, fourth ventricle. (E) Midsagittal section through the aqueduct (Aq) and the beginning of the fourth ventricle (4V). Note the intense labelling of the dorsal ventricular zone of the aqueduct and the sharp rostral and caudal boundaries of the hybridization signal (arrowheads). Aqi, inferior tectal aqueduct. (F) Coronal section through the aqueduct (Aq) at the position indicated by the broken line in (E). Note the locally restricted CNTFRá expression in the ventricular zone (arrowheads). Labelling is absent in the floor and the dorsal midline. (G) Coronal section through the caudal medulla showing labelled cells in external cuneate nucleus (ECu), hypoglossal nucleus (12), lateral reticular nucleus (LRt) and inferior olive (IO); 4V, fourth ventricle. Scale bars=200 µm.

with a sharp boundary to the unlabelled ventral ventricular zone (Fig. 2A). Brainstem In the neonate many brainstem nuclei contained relatively high levels of CNTFRá mRNA. Labelled nuclei in midbrain and pons (Fig. 4A,C) included

functionally diverse cell groups such as the oculomotor nucleus, nucleus ruber, nucleus of the trapezoid body, reticulotegmental nucleus and others which are not shown but are listed in Table 1. Similarly, in the medulla oblongata cranial nerve motor nuclei (abducens, ambiguus, facial and hypoglossal nuclei), sensory cranial nerve nuclei (spinal trigeminal nucleus, vestibular nuclei), nuclei of the reticular formation

240

M.-Y. Lee et al.

Fig. 5. CNTFRá mRNA expression in the brainstem of the adult rat. (A) Coronal section at the midbrain level showing the unlabelled nucleus ruber (R), superior colliculus (SC) and medial geniculate nucleus (MG). Owing to the low density of labelled neurons in the substantia nigra (SN), labelling is not visible at this magnification (see (B)); Aq, aqueduct. (B) Higher magnification view of the boxed area in (A). In the substanstia nigra labelling is in neurons (see inset) of the pars reticularis (SNR) and lateralis (SNL). Cells in pars compacta (SNC) are negative; cp, cerebral peduncle. (C) Cranial pons with labelled neurons in the trigeminal motor nucleus (Mo5); Tz, nucleus of trapezoid body; SO, superior olive; py, pyramidal tract; mlf, medial longitudinal fasciculus; 4V, fourth ventricle. (D) Caudal pons with labelled neurons in the lateral and medial vestibular nuclei (LVe, MVe), facial nucleus (7) and gigantocellular reticular nucleus (Gi); 4V, fourth ventricle; DC, dorsal cochlear nucleus. (E) Coronal section of the medulla with heavily labelled neurons in the vagal dorsal motor nucleus (10) and hypoglossal nucleus (12); cc, central canal; Gr, gracile nucleus. Scale bars=300 µm (A–E), 50 µm (inset in B).

and neurons of the inferior olive displayed specific hybridization signals of varying intensity (Fig. 4D,G, Table 1). With maturation CNTFRá expression decreased in many nuclei of the midbrain and pons, becoming weak or undetectable by the in situ technique used (see Table 1), whereas it remained high in others (Fig. 5A–E). In the adult, intense hybridization was seen solely in the parabigeminal nucleus and the substantia nigra (Fig. 5A,B, Table 1). In the latter, high levels of CNTFRá mRNA were present in the pars reticularis and pars lateralis, but no significant labelling was observed in the pars compacta (Fig. 5B). In the pontine area CNTFRá expression was lost in most of the nuclei found to be positive at P1, but retained in the pontine reticular nucleus and trigeminal motor nucleus (Fig. 5C, Table 1). In contrast to the decrease

in mRNA expression in midbrain and pons, it remained high in neurons of cranial nerve motor nuclei, vestibular nuclei and reticular nuclei in the medulla oblongata (Fig. 5D,E). Only in the inferior olive had it been lost as compared to P1, while the accessory facial nucleus and the area postrema had become positive (Tabel 1). Labelled neurons were also present in the dorsal cochlear nucleus of the adult. Cerebellum An intense hybridization signal was observed in deep cerebellar nuclei and in the cerebellar cortex at P1 (Fig. 6A–C). Labelling in the cortex, however, showed a medial-to-lateral gradient, with highest intensities in the vermis (Fig. 6C), and a rostral-tocaudal gradient as demonstrated in a sagittal section

CNTFRá mRNA in rat brain

241

Fig. 6. Expression of CNTFRá mRNA in the neonatal and adult rat cerebellum. (A) Midsagittal section of a P1 cerebellum. Hybridization is seen in the medial cerebellar nucleus (Med) and in deeper layers of the cortex. Note the differences in labelling intensity between rostral and caudal cortex of the vermis. MVe, medial vestibular nucleus; IC, inferior colliculus; 4V, fourth ventricle. (B) Higher magnification of the boxed area in (A). Density of labelled cells is highest in the internal granular layer (IGL) adjacent to the Purkinje cell layer (Pk). The external germinal layer (EGL) is unlabelled. Triangles indicate the pial surface. (C) Coronal section of a P1 cerebellum showing label in the medial, interposed and lateral nucleus (Med, Int, Lat) and a medial-to-lateral gradient of labelling intensity in the cortex. (D) Coronal section of an adult cerebellum with labelled deep nuclei (see (E)) and hybridization in distinct cortical layers. Med, medial nucleus; Int, interposed nucleus; Lat, lateral nucleus; 4V, fourth ventricle. (E) At higher magnification, labelling density can be seen to be highest in the Purkinje cell layer (Pk) and low in the granular layer (GL). ML, molecular layer; inset: high-power view of the Purkinje cell layer. Purkinje cells are negative (arrowheads indicating perikaryon), but are surrounded by intensely labelled cells. Scale bars=300 µm (A,C,E), 50 µm (B, inset).

(Fig. 6A). This probably reflected regional differences in cerebellar development, since no such gradients were observed in the adult tissue. The density of labelled cells in the immature cerebellum was highest in the developing granular layer close to the immature Purkinje cells which comprise several cell layers at that stage, but scattered positive cells were also found situated more deeply towards the white matter (Fig. 6B). The external germinal layer containing proliferating cells and immature, migrating granule cells and the forming molecular layer were devoid of

labelled cells. In the mature cerebellum the labelling pattern had not changed dramatically, but a very strong signal was observed in the Purkinje cell layer (Fig. 6D,E). This signal was due to CNTFRá expression by small cells in the close vicinity of the Purkinje cell somata which themselves were not labelled (Fig. 6E, inset). Since both Bergmann glia cells and certain cerebellar interneurons are located at this position, the identity of the receptor-expressing cells remains to be established. The same is true for the scattered cells in the granular layer.

242

M.-Y. Lee et al.

Neuroepithelium CNTFRá expression in the neuroepithelial cells of the ventricular zone of the neonatal rat showed a very distinct distribution. The labelling pattern in the olfactory bulb, the lateral ventricle (dorsolateral or dorsal wall depending on the rostrocaudal level) and third ventricle (in the dorsal habenular region) has already been described in previous sections. The neuroepithelium of the cerebral aqueduct also showed region-specific expression of CNTFRá mRNA. No hybridization was observed in the rostral aqueduct of the midbrain (Fig. 4A,B,E). In the tectal region, however, intense labelling was present in the dorsal neuroepithelium lining the recess of the inferior colliculus as demonstrated in a sagittal section (Fig. 4E). More precisely, expression was intense in the dorsolateral aspect and not detectable in the floor and the dorsal midline region of the aqueduct as seen in a coronal section (Fig. 4F). The neuroepithelium of the fourth ventricle (Fig. 4E,G) and the spinal canal showed no hybridization signal. DISCUSSION

The identification of putative target cells by demonstrating the expression of the corresponding receptor at different stages of development is an important step towards understanding the function of a neurotrophic factor, although it is not sufficient to prove its physiological importance for the receptor-expressing cell type. In the present study a digoxygenin-labelled riboprobe was used to map CNTFRá mRNA expression in the neonatal and adult rat brain on a cellular level. Our results extend findings from a previous radioactive in situ hybridization study24 by defining a number of new putative CNTF targets. In addition, distinct differences were found between the immature and mature brain, stressing the importance of CNTFRá during development as well as in the adult CNS. Very recently, an immunocytochemical study on the distribution of the CNTFRá protein in the nervous system has been published.32 In those areas investigated in both studies localization of protein and mRNA were closely correlated, with one major difference concerning the expression of CNTFRá by Purkinje cells (see below). It is trivial to mention that the list of CNTFRá-expressing structures presented here may still be incomplete. Very low levels of mRNA may have escaped detection by our technique, but nevertheless could be of functional significance.34 Diversity of ciliary neurotrophic factor receptor-áexpressing brain areas At the two developmental stages studied, more than 50 brain structures displayed specific hybridization to the CNTFRá antisense probe (Table 1). Thus, a central result of our study was the demonstration that the distribution of CNTFRá mRNA is even

more widespread than previously shown.24 The overall labelling pattern, the layer specificity and the inspection of the labelled cells at higher magnification suggested that, apart from the neuroepithelium and possibly the cerebellum (see below), CNTFRá expression is confined to postmitotic neurons in the normal postnatal brain. For the final identification of the individual CNTFRá-containing cell types in the different brain areas it will be necessary to perform double-labelling experiments, but in most areas it was apparent that CNTFRá mRNA was localized in principal projection neurons, e.g., mitral cells of the olfactory bulb, layer V neurons of the neocortex, granule cells of the dentate gyrus, pyramidal cells of the hippocampus, motoneurons of the cranial nerve nuclei and deep cerebellar nuclei. In the cerebellar cortex, however, Purkinje cells were clearly negative. Instead, the distribution of labelled cells in the neonatal cortex resembled that of developing Bergmann glia obtained by [3H]thymidine autoradiography or tenascin in situ hybridization.1,15 Likewise, expression by Bergmann glia would explain the intense signal in cells surrounding the perikarya of Purkinje cells in the adult cerebellum. Thus, the cerebellar cortex may be an exception in that CNTFRá is not restricted to neurons, but additional experiments will be necessary to exclude the possibility that cerebellar interneurons are responsible for the observed labelling pattern. The absence of a hybridization signal in Purkinje cells and its presence in the granular layer24 represents a striking difference to the highly selective labelling of Purkinje cell somata and dendrites in the rat cerebellum by CNTFRá immunocytochemistry.32 There seems to be no simple explanation for this discrepancy, in particular, because the cellular distribution of mRNA and protein were in excellent agreement in other brain areas studied. In their study on the distribution of CNTFRá in PNS and CNS neurons, Ip et al.24 have already described the prominent expression of this receptor in motoneurons, layer V neurons of the neocortex and other motor system-associated CNS areas. Our results confirm and extend these observations and those of the recent immunocytochemical study by MacLennan et al.32 by identifying additional CNTFRá-expressing motor-related structures including all cranial nerve motor nuclei in the neonatal brain. In agreement with these localization studies, newborn mice homozygous for null mutations in CNTFRá exhibit severe behavioural and morphological motoneuron deficits and die shortly after birth.14 The association of CNTFRá expression with the motor system, however, is neither a general nor a specific feature. First, there are motor-related areas which did not show detectable mRNA levels and second, prominent hybridization signals were observed in other functional systems. Noticeable CNTFRá expression was present among others in structures of the limbic system (limbic cortex, hippocampus proper, dentate gyrus, entorhinal area,

CNTFRá mRNA in rat brain

certain thalamic nuclei, medial habenular nucleus), in the olfactory bulb and cortex, in the hypothalamus and in nuclei of sensory systems such as the medial geniculate nucleus and vestibular and trigeminal nuclei. A corresponding diversity exists with respect to the transmitter phenotype of CNTFRá-expressing neurons. To mention only a few, CNTFRá mRNA was found in cholinergic motoneurons, glutamatergic pyramidal cells, GABAergic neurons of the substantia nigra and peptide-containing hypothalamic neurons. Apparently, there is no correlation between CNTFRá expression and the functional or anatomical characteristics of a neuron. This is similar to the neurotrophin receptors which show partially overlapping expression with CNTFRá.5,53 Ciliary neurotrophic factor receptor-á expression in ciliary neurotrophic factor-responsive neurons A series of studies has reported CNTF effects on neurons from different CNS areas in vitro and in vivo.36,47 The prominent CNTFRá expression in motoneurons is in accordance with the survivalpromoting action of CNTF on spinal motoneurons during in vivo development and in culture4,37,42,52 and on facial neurons after nerve lesion,48 and it demonstrates that CNTF can act directly on these cells. The same is true for CNTF effects on cultured hippocampal, corticospinal and injured anterior thalamic neurons.12,23,31,35 However, no CNTFRá mRNA could be detected in the septum of neonatal or adult rats, although CNTF can stimulate the differentiation of cultured embryonic septal neurons25 and prevent the degeneration of adult septal neurons following axotomy.19 Trophic effects of CNTF have also been shown for dopaminergic neurons of the substantia nigra pars compacta both in embryonic cultures and in the adult lesioned brain.20,34 The hybridization signal in the substantia nigra, however, was restricted to pars reticularis and lateralis. There are several possible explanations for this discrepancy. First, the neurons may express very low levels of CNTFRá mRNA, but nevertheless respond to high levels of exogenous CNTF. Second, the observed effects may have been mediated by other cells, for instance by astrocytes which normally do not express CNTFRá, but are induced to produce significant levels of functional CNTFRá when taken into culture and after brain lesions.45 Third, injuring septal and substantia nigra neurons either by dissociation for culturing or by axotomy may result in the induction or up-regulation of CNTFRá in these cells. We have obtained preliminary evidence that this can happen in axotomized neurons.

Differential expression of ciliary neurotrophic factor receptor-á in neonatal and adult brain Comparison of the hybridization patterns obtained for the neonatal and adult brain (Table 1) revealed

243

substantial differences between the two stages. In some regions, especially cortical areas, CNTFRá expression was found to start or increase during postnatal development. However, there were many more structures where the hybridization signal became weaker or disappeared with maturation (most pronounced in midbrain and pontine nuclei and in the neuroepithelium), indicating that CNTFRámediated processes are important during perinatal neuronal development. In contrast, the lower levels of CNTF found in the developing nervous system as compared to the adult,50,51 the absence of substantial deficits in young mice lacking CNTF14,38 and the lack of a signal sequence in the CNTF gene have led to the conclusion that the neurotrophic protein is not essential for the developing nervous system, but acts as a ‘‘lesion factor’’ which is released after injury from CNTF-producing glial cells.47 In view of these facts and their own recent results which demonstrated severe motoneuron deficits in newborn mice with null mutations in CNTFRá, DeChiara et al.14 have proposed the attractive hypothesis that the existence of a second CNTF-like ligand is responsible for the critical developmental role of CNTFRá. Thus, although we do not present data on the embryonic brain (but see Ref. 24), the widespread but specific distribution in the neonatal rat brain described here adds further evidence for the importance of CNTFRá-mediated signals in the developing CNS. Region-specific ciliary neurotrophic factor receptor-á expression in the neuroepithelium During earlier stages of embryonic development the neuroepithelium, synonymously called the ventricular zone, is the sole germinative zone in the CNS, whereas in late embryonic stages the newly forming subventricular zone represents an additional source of proliferating precursors.9 Around the time of birth, neurogenesis is ceasing and this is accompanied by a decrease in proliferative activity in the germinative layers and the gradual transformation of the multilayered neuroepithelium into the one-layer thick ependyma. However, there are distinct regions of the ventricular zone which show high rates of persisting mitotic activity in the early postnatal rat, comprising the olfactory ventricle, external wall of the lateral ventricle, habenular and pituitary regions of the third ventricle and the caudal recess of the aqueduct.1,2,3 It was exactly in these areas where we observed intense labelling of the columnar epithelial cells of the ventricular zone with sharp demarcations to the remaining unlabelled neuroepithelium (Figs 1, 2, 4). In E11 and E15 rat embryos very prominent CNTFRá mRNA has been demonstrated in the ventricular zone of developing spinal cord and brain24 and CNTFRá protein has also been shown immunocytochemically in embryonic neuroepithelial cells.32 These expression patterns suggest an important role

244

M.-Y. Lee et al.

for CNTFRá during neurogenesis which is no longer relevant in the primitive ependyma. Receptor expression, however, is not generally associated with the proliferation of neuronal or glial progenitors, since no hybridization was observed in other germinal zones such as the subventricular zone or the external germinal layer of the cerebellum. CONCLUSIONS

The present study has identified a number of putative CNTF target neurons. CNTFRá could be demonstrated to be expressed in the mitotically active neuroepithelium, in a variety of postmitotic differentiating neurons of the neonatal rat and in a more restricted set of neurons in the adult brain. In situ hybridization in late embryonic brains will be required to assess the entire developmental time-course of CNTFRá expression. However, our results indicate that expression is high in proliferating neuroepithelial progenitors, absent in precursor cells of the subventricular zone and during further migration, and then induced again in distinct populations of differentiating neurons. After maturation CNTFRá apparently is down-regulated in certain types of neurons, but remains high or even increases in others. Thus, our results indicate that CNTFRá has various functions in the CNS depending on the developmen-

tal situation. Corresponding to the receptor distribution, CNTF, the only CNTFRá ligand identified so far, is widely expressed in the adult rat brain.24,50 However, its expression is significantly lower during prenatal development. Thus, it is well possible that CNTFRá-mediated processes in the adult and developing brain involve different ligands.14 For signal transduction by the CNTF receptor, interaction is required between the á-subunit and the two membrane-spanning â-subunits (leukemia inhibitory factor receptor â, gp130). By northern blot analysis, the expression of these subunits was found to parallel that of CNTFRá in brain tissues, although at lower levels,24 and the cellular responses of several of the CNTFRá-expressing neuronal cell types47 also suggest that CNS neurons express functional CNTF receptor complexes. It will be interesting, however, to study the cellular localization of leukemia inhibitory factor receptor â and gp130 in the brain to determine whether CNTFRá expression always reflects the presence of signal-transducing receptor complexes. Acknowledgements—We thank R. Bender, C. Haas and A. Straube for their helpful advice, G. Kaiser for excellent technical assistance and R. Kovacs for photographical work. This work was supported by a grant from the Deutsche Forschungsgemeinschaft, SFB 505, and the Research Fund of Songeui, Korea (M.-Y. L.).

REFERENCES

1. 2. 3. 4.

5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

15.

16. 17.

Altman J., Anderson W. J. and Wright K. A. (1968) Differential radiosensitivity of stationary and migratory primitive cells in the brains of infant rat. Expl Neurol. 22, 52–74. Altman J. and Bayer S. A. (1981) Time of origin of neurons of the rat inferior colliculus and the relations between cytogenesis and tonotopic order in the auditory pathway. Expl Brain Res. 42, 411–423. Altman J. and Das G. D. (1965) Autoradiographic and histological studies of postnatal neurogenesis. J. comp. Neurol. 126, 337–390. Arakawa Y., Sendtner M. and Thoenen H. (1990) Survival effect of ciliary neurotrophic factor (CNTF) on chick embryonic motoneurons in culture: comparison with other neurotrophic factors and cytokines. J. Neurosci. 10, 3507–3515. Barbacid M. (1994) The trk family of neurotrophin receptors. J. Neurobiol. 25, 1386–1403. Barbin G., Manthorpe M. and Varon S. (1984) Purification of the chick eye ciliary neuronotrophic factor. J. Neurochem. 43, 1468–1478. Barres B. A. and Raff M. C. (1994) Control of oligodendrocyte number in the developing rat optic nerve. Neuron 12, 935–942. Bayer S. A. (1983) 3H-Thymidine-radiographic studies of neurogenesis in the rat olfactory bulb. Expl Brain Res. 50, 329–340. Bayer S. A. and Altman J. (1991) Neocortical Development. Raven Press, New York. Bazan J. F. (1991) Neuropoietic cytokines in the hematopoietic fold. Neuron 7, 197–208. Chomczynski P. and Sacchi N. (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate–phenol– chloroform extraction. Analyt. Biochem. 162, 156–159. Clatterbuck R. E., Price D. L. and Koliatsos V. E. (1993) Ciliary neurotrophic factor prevents retrograde neuronal death in the adult central nervous system. Proc. natn. Acad. Sci. U.S.A. 90, 2222–2226. Davis S., Aldrich T. H., Stahl N., Pan L., Taga T., Kishimoto T., Ip N. Y. and Yancopoulos G. D. (1993) LIFRb and gp130 as heterodimerizing signal transducers of the tripartite CNTF receptor. Science 260, 1805–1808. DeChiara T. M., Vejsada R., Poueymirou W. T., Acheson A., Suri C., Conover J. C., Friedman B., McClain J., Pan L., Stahl N., Ip N. Y., Kato A. and Yancopoulos G. D. (1995) Mice lacking the CNTF receptor, unlike mice lacking CNTF, exhibit profound motor neuron deficits at birth. Cell 83, 313–322. Do¨rries U., Bartsch U., Nolte C., Roth J. and Schachner M. (1993) Adaptation of a non-radioactive in situ hybridization method to electron microscopy: detection of tenascin mRNAs in mouse cerebellum with digoxigeninlabelled probes and gold-labelled antibodies. Histochemistry 99, 251–262. Ernsberger U., Sendtner M. and Rohrer H. (1989) Proliferation and differentiation of embryonic chick sympathetic neurons: effects of ciliary neurotrophic factor. Neuron 2, 1275–1284. Fuhrmann S., Kirsch M. and Hofmann H.-D. (1995) Ciliary neurotrophic factor promotes chick photoreceptor development in vitro. Development 121, 2695–2706.

CNTFRá mRNA in rat brain 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50.

51.

245

Gearing D. P., Comeau M. R., friend D. J., Gimpel S. D., Thut C. J., McGourty J., Brasher K. K., King J. A., Gillis S., Mosley B., Ziegler S. F. and Cosman D. (1992) The IL-6 signal transducer, gp130: an oncostatin M receptor and affinity converter for the LIF receptor. Science 255, 1434–1437. Hagg T., Quon D., Higaki J. and Varon S. (1992) Ciliary neurotrophic factor prevents neuronal degeneration and promotes low affinity NGF receptor expression in the adult rat CNS. Neuron 8, 145–158. Hagg T. and Varon S. (1993) Ciliary neurotrophic factor prevents degeneration of adult rat substantia nigra dopaminergic neurons in vivo. Proc. nant. Acad. Sci. U.S.A. 90, 6315–6319. Hibi M., Murakami M., Saito M., Hirano T., Taga T. and Kishimoto T. (1990) Molecular cloning and expression of an IL-6 signal transducer, gp130. Cell 63, 1149–1157. Hofmann H.-D. (1988) Ciliary neuronotrophic factor stimulates choline acetyltransferase activity in cultured chicken retina neurons. J. Neurochem. 51, 109–113. Ip N. Y., Li Y., van de Stadt I., Panayotatos N., Alderson R. F. and Lindsay R. M. (1991) Ciliary neurotrophic factor enhances neuronal survival in embryonic rat hippocampal cultures. J. Neurosci. 11, 3124–3134. Ip N. Y., McClain J., Barrezueta N. X., Aldrich T. H., Pan L., Li Y., Wiegand S. J., Friedman B., Davis S. and Yancopoulos G. D. (1993) The a component of the CNTF receptor is required for signaling and defines potential CNTF targets in the adult and during development. Neuron 10, 89–102. Kew J. N. C. and Sofroniew M. V. (1995) Ciliary neurotrophic factor supports p75NGFR-immunoreactive non-cholinergic, but not cholinergic, developing septal neurons in vitro. Neuroscience 66, 793–804. Kirsch M., Fuhrmann S., Wiese A. and Hofmann H.-D. (1996) CNTF exerts opposite effects on in vitro development of rat and chick photoreceptors. NeuroReport 7, 697–700. LaVail M. M., Unoki K., Yasumura D., Matthes M. T., Yancopoulos G. D. and Steinberg R. H. (1992) Multiple growth factors, cytokines, and neurotrophins rescue photoreceptors from damaging effects of constant light. Proc. natn. Acad. Sci. U.S.A. 89, 11249–11253. Lehwalder D., Jeffrey P. L. and Unsicker K. (1989) Survival of purified embryonic chick retinal ganglion cells in the presence of neurotrophic factors. J. Neurosci. Res. 24, 329–337. Lillien L. E. and Raff M. C. (1990) Differentiation signals in the CNS: type-2 astrocyte development in vitro as a model system. Neuron 5, 111–119. Lin L.-F. H., Mismer D., Lile J. D., Armes L. G., Butler E. T., Vannice J. L. and Collins F. (1989) Purification, cloning, and expression of ciliary neurotrophic factor (CNTF). Science 246, 1023–1025. Louis J.-C., Magal E., Burnham P. and Varon S. (1993) Cooperative effects of ciliary neurotrophic factor and norepinephrine on tyrosine hydroxylase expression in cultured rat locus coeruleus neurons. Devl Biol. 155, 1–13. MacLennan A. J., Vinson E. N., Marks L., McLaurin D., Pfeifer M. and Lee N. (1996) Immunohistochemical localization of ciliary neurotrophic factor receptor a expression in the rat nervous system. J. Neurosci. 16, 621–630. Magal E., Burnham P. and Varon S. (1991) Effect of ciliary neurotrophic factor on rat spinal cord neurons in vitro: survival and expression of choline acetyltransferase and low-affinity nerve growth factor receptors. Devl Brain Res. 63, 141–150. Magal E., Burnham P., Varon S. and Louis J.-C. (1993) Convergent regulation by ciliary neurotrophic factor and dopamine of tyrosine hydroxylase expression in cultures of rat substantia nigra. Neuroscience 52, 867–881. Magal E., Louis J.-C., Oudega M. and Varon S. (1993) CNTF promotes the survival of neonatal rat corticospinal neurons in vitro. NeuroReport 4, 779–782. Manthorpe M., Louis J.-L. and Hagg T. and Varon S. (1993) Ciliary neuronotrophic factor. In Neurotrophic Factors (eds Loughlin S. E. and Fallon J. H.), pp. 443–473. Academic Press, New York. Martinou J. C., Martinou I. and Kato A. C. (1992) Cholinergic differentiation factor (CDF/LIF) promotes survival of isolated rat embryonic motoneurons in vitro. Neuron 8, 737–744. Masu Y., Wolf E., Holtmann B., Sendtner M., Brem G. and Thoenen H. (1993) Disruption of the CNTF gene results in motor neuron degeneration. Nature 365, 27–32. Mayer M., Bhakoo K. and Noble M. (1994) Ciliary neurotrophic factor and leukemia inhibitory factor promote the generation, maturation and survival of oligodendrocytes in vitro. Development 120, 143–153. Meyer-Franke A., Kaplan M. R., Pfrieger F. W. and Barres B. A. (1995) Characterization of the signaling interactions that promote the survival and growth of developing retinal ganglion cells in culture. Neuron 15, 805–819. Mitsumoto H., Ikeda K., Klinkosz B., Cedarbaum J. M., Wong V. and Lindsay R. M. (1994) Arrest of motoneuron disease in wobbler mice cotreated with CNTF and BDNF. Science 265, 1107–1110. Oppenheim R. W., Prevette D., Qin-Wei Y., Collins F. and MacDonald J. (1991) Control of embryonic motoneuron survival in vivo by ciliary neurotrophic factor. Science 251, 1616–1618. Paxinos G., To¨rk I. and Tecott L. H. and Valentino K. L. (1994) Developing Rat Brain. Academic Press, San Diego. Paxinos G. and Watson C. (1986) The Rat Brain. Academic Press, San Diego. Rudge J. S., Li Y., Pasnikowski E. M., Mattsson K., Pan L., Yancopoulos G. D., Wiegand S. J., Lindsay R. M. and Ip N. Y. (1994) Neurotrophic factor receptors and their signal transduction in rat astrocytes. Eur. J. Neurosci. 6, 693–705. Saadat S., Sendtner M. and Rohrer H. (1989) Ciliary neurotrophic factor induces cholinergic differentiation of rat sympathetic neurons in culture. J. Cell Biol. 108, 1807–1816. Sendtner M., Carroll P., Holtmann B., Hughes R. A. and Thoenen H. (1994) Ciliary neurotrophic factor. J. Neurobiol. 25, 1436–1453. Sendtner M., Kreutzberg G. W. and Thoenen H. (1990) Ciliary neurotrophic factor prevents the degeneration of motor neurons after axotomy. Nature 345, 440–441. Stahl N. and Yancopoulos G. D. (1994) The tripartite CNTF receptor complex: activation and signaling involves components shared with other cytokines. J. Neurobiol. 25, 1454–1466. Sto¨ckli K. A., Lillien L. E., Na¨her-Noe´ M., Breitfeld G., Hughes R. A., Raff M. C., Thoenen H. and Sendtner M. (1991) Regional distribution, developmental changes, and cellular localization of CNTF-mRNA and protein in the rat brain. J. Cell Biol. 115, 447–459. Sto¨ckli K. A., Lottspeich F., Sendtner M., Masiakowski P., Carroll P., Go¨tz R., Lindholm D. and Thoenen H. (1989) Molecular cloning, expression and regional distribution of rat ciliary neurotrophic factor. Nature 342, 920–923.

246

M.-Y. Lee et al.

52.

Wewetzer K., MacDonald J. R., Collins F. and Unsicker K. (1990) CNTF rescues motoneurons from ontogenetic cell death in-vivo, but not in-vitro. NeuroReport 1, 203–206. 53. Yan Q. and Johnson E. M. (1989) Immunohistochemical localization and biochemical characterization of nerve growth factor receptor in adult rat brain. J. comp. Neurol. 290, 585–598. (Accepted 19 August 1996)