The distribution of the mRNA and protein products of the melanin-concentrating hormone (MCH) receptor gene, slc-1, in the central nervous system of the rat

Melanin‐concentrating hormone (MCH), a 19 amino acid cyclic peptide, is largely expressed in the hypothalamus. It is implicated in the control of general arousal and goal‐orientated behaviours in mammals, and appears to be a key messenger in the regulation of food intake. An understanding of the biological actions of MCH has been so far hampered by the lack of information about its receptor(s) and their location in the brain. We recently identified the orphan G‐protein‐coupled receptor SLC‐1 as a receptor for the neuropeptide MCH. We used in situ hybridization histochemistry and immunohistochemistry to determine the distribution of SLC‐1 mRNA and its protein product in the rat brain and spinal cord. SLC‐1 mRNA and protein were found to be widely and strongly expressed throughout the brain. Immunoreactivity was observed in areas that largely overlapped with regions mapping positive for mRNA. SLC‐1 signals were observed in the cerebral cortex, caudate‐putamen, hippocampal formation, amygdala, hypothalamus and thalamus, as well as in various nuclei of the mesencephalon and rhombencephalon. The distribution of the receptor mRNA and immunolabelling was in good general agreement with the previously reported distribution of MCH itself. Our data are consistent with the known biological effects of MCH in the brain, e.g. modulation of the stress response, sexual behaviour, anxiety, learning, seizure production, grooming and sensory gating, and with a role for SLC‐1 in mediating these physiological actions.

Keywords: immunohistochemistry; in situ hybridization; neuropeptide; receptor

The neuropeptide melanin‐concentrating hormone (MCH, for reviews see [10]; [2], [3]; [29]; [20]) was first recognized as having a physiological control action on teleost skin colour and was purified from salmon pituitaries in 1983 ([19]). The mammalian orthologue was purified in 1989 from rat hypothalami ([41]). The peptide shows a total conservation of primary sequence in diverse mammals − perhaps an indication of a major role of the peptide throughout evolution (see [29]). In the rat brain, MCH is largely expressed in the lateral hypothalamus (LHA) and the adjacent zona incerta. MCH neurons project very broadly throughout the brain to neocortical, striatal, amygdaloid and hippocampal areas, as well as other structures in the diencephalon, mesencephalon, pontine formation, medulla oblongata and neurointermediate lobe of the pituitary gland ([6]). The widespread network of MCH neuronal projection sites is compatible with functional roles for MCH in generalized arousal and sensorimotor integration. The association of the LHA with specific goal‐orientated behaviours, particularly feeding‐related ones, has been recently exemplified by the fact that MCH is a key regulator in energy balance and food intake. In fact, MCH is orexigenic (demonstrated by intracerebroventricular injection, [32]; [33]; [26]), and transgenic mice lacking the MCH gene have a lean and hypophagic phenotype ([36]). Other biological roles attributed to MCH concern its involvement in the stress response ([18]; [7]; [26]), sexual behaviour, anxiety ([13]), learning ([27]), seizure production ([21]), grooming, locomotor activity ([35]) and sensory gating ([28]).

For more than 15 years, advances in understanding of MCH biology have been hampered by the lack of information about the MCH receptor(s). This was mostly due to technical difficulties inherent in the cyclical nature of the MCH molecule and the absence of a simple assay to monitor MCH bioactivity. With a reverse‐pharmacology approach (see [38]), we recently identified an MCH receptor ([8]). SLC‐1, previously an orphan receptor ([23]), was shown to be activated by MCH with high specificity and affinity. Another research group also identified the same MCH receptor by using a brain purification extract approach ([34]). Several other reports describing the identification of SLC‐1 as being an MCH receptor followed ([1]; [25]; [37]).

In our earlier study, we demonstrated that the SLC‐1 mRNA and protein were present within key hypothalamic areas involved in the control of food intake. Here we describe in detail the distribution of the MCH receptor, at both mRNA and protein levels, in the rat brain and spinal cord. Part of this work was presented at The Society for Neuroscience 1999 Miami Meeting.

Adult male Wistar rats (200–250 g, Charles River, UK) were kept in a fixed 12 h light–dark cycle with food and water provided ad libitum.

All peptides were synthesized using solid‐phase methodology on a model 432A applied Biosystem Synthesizer. Peptide purity was estimated by chromatography as being greater than 95%.

All chemicals were purchased from Merck (UK) unless otherwise stated. The bovine serum albumen (BSA) used in all experiments was of a high‐quality crystallized grade (Fraction V, A‐8022, Sigma, UK).

HEK 293 cell lines were stably transfected with a human MCH receptor cDNA subcloned using the pCDN vector (Invitrogen, UK). All cell culture reagents were purchased from Life Technologies (Paisley, UK) and plastic‐ware from Costar (High Wycombe, UK). Wild‐type and SLC‐1‐transfected CHO‐K1 cells were cultured in minimum essential medium (MEM) containing Earle's salts, l‐glutamine, 10% foetal bovine serum, and 1% MEM non‐essential amino acids and 50 mg/mL geneticin solution in humidified air containing 5% CO2 at 37 °C.

For mRNA localization studies, male Sprague–Dawley rats (300–350 g) were killed by CO2 asphyxia followed by cervical dislocation. Sixteen brain regions and spinal cord and dorsal root ganglia were dissected free. Each tissue sample was pooled from between six and 16 rats depending on the size of the individual tissue samples obtained. Total RNA was extracted from the tissue according to the manufacturer's suggested protocol with the addition of an extra chloroform extraction step and phase separation, and an extra wash of the isolated RNA in 70% ethanol. The RNA was resuspended in autoclaved, ultrapure water and the concentration calculated by A260 measurement. RNA quality was assessed by electrophoresis on a 1% agarose gel. First strand cDNA synthesis was carried out by oligo(dT) priming from 1 μg of each RNA sample [0.01 m dithiothreitol (DTT), 0.5 mm each dNTP, 0.5 μg oligo(dT) primer, 40 U RNAseOUT ribonuclease inhibitor (Life Technologies) 200 U Superscript II reverse transcriptase (Life Technologies)]. Triplicate reverse transcription reactions were performed along with an additional reaction in which the reverse transcriptase was omitted to allow for assessment of genomic DNA contamination of the RNA. The resulting cDNA products were divided into 20 aliquots for parallel Taqman polymerase chain reactions (PCR) using different primer and probe sets for quantification of multiple cDNA sequences.

Quantitative reverse transcription‐polymerase chain reaction (RT‐PCR) was carried out using an ABI 7700 sequence detector (Perkin Elmer, USA) on the cDNA samples [2.5 mm MgCl2, 0.2 mm dATP, dCTP, dGTP and dUTP, 0.1 μm each primer, 0.05 m Taqman probe, 0.01 U AmpErase uracil‐N‐glycosylase (Perkin Elmer), 0.0125 U Amplitaq Gold DNA polymerase (Perkin Elmer); 50 °C for 2 min, 95 °C for 10 min followed by 40 cycles of 95 °C for 15 s, 60 °C for 1 min]. Additional reactions were performed on each 96‐well plate using known dilutions of rat genomic DNA as a PCR template to allow construction of a standard curve relating threshold cycle to template copy number.

Taqman primer and probe sets for SLC‐1 and for the housekeeping gene glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH) were designed using Primer Express software (Perkin Elmer). Parallel Taqman PCRs were run on each sample using the GAPDH primers and probe to control for RNA integrity. Primer and probe sequences used were (forward primer, reverse primer and Taqman probe):

SLC‐1: 5′‐CTCTACGCCAGGCTCATTCC,

• 5′‐ACAGAGTGAACCAGTAGAGGTCAGTGT,
• 5′‐GGCGGATGCCACAGCCCA;

GAPDH: 5′‐GAACATCATCCCTGCATCCA,

• 5′‐CCAGTGAGCTTCCCGTTCA,
• 5′‐CTTGCCCACAGCCTTGGCAGC.

Copy numbers obtained for SLC‐1 were normalized to those for GAPDH and the resulting values expressed as arbitrary units. Taqman data have been calibrated using a standard curve using known copy numbers of genomic DNA to give an initial template number of MCH receptor cDNA molecules in the reaction. The absolute values given to the expression levels are therefore arbitrary, and it is the relative expression of this transcript in different brain areas that is shown. For more details, refer to [14].

The ISH experiments were performed with oligonucleotides designed from the rat SLC‐1 sequence ([24]). Various oligonucleotides with non‐overlapping sequences were designed using the PrimerSelect software of Lasergene. The program identified three sequences.

The antisense sequences corresponding to the sense sequences (with their respective position on the cDNA sequence and length) are as follows:

5′‐T CTG CAA ACC TCG TTG CTG TCC

ACT‐3′ (sense; 6–30/25‐mer)

5′‐AGT GGA CAG CAA CGA GGT TTG

CAG A‐3′ (corresponding antisense sequence);

5′‐GGC CAC CGT CCA CCC CAT CTC CTC

CAC‐3′ (sense; 429–455/27‐mer)

• 5′‐GTG GAG GAG ATG GGG TGG ACG GTG GCC‐3′ (corresponding antisense sequence);
• 5′‐AT ACT ACA GCG CAT GAC GTC TTC

GGT GGC CCC‐3′ (sense; 691–722/32‐mer)

5′‐GGG GCC ACC GAA GAC GTC ATG CCC TGT AGT AT‐3′ (corresponding antisense sequence).

Sequences were checked for uniqueness using BLAST (Advanced search; GenEMBL complete).

Oligonucleotides were used independently and provided similar patterns of distribution. In order to enhance signal intensity, in some cases a 'cocktail' (equal molar mixture) of the oligoprobes was used. Full details of the ISH protocol have been reported elsewhere ([16]). Briefly, postfixed rat brain sections [cut coronally or sagittally with a thickness of 20 μm; postfixed for 30 min in 4% PFA (paraformaldehyde) w/v in phosphate‐buffered saline (PBS)] were hybridized overnight with [35S] oligoprobes radio‐labelled by 3′‐tailing with terminal transferase (specific radioactivity > 109 dpm per microgram, incorporation > 80%), then were washed and dehydrated. The sections were exposed to Kodak Biomax film for up to 3 weeks. Control experiments included the use of the sense sequence oligoprobes, the use of excess (100 ×) cold antisense oligoprobes, and RnaseA‐pretreated sections.

A rabbit polyclonal antiserum was raised against the extreme C‐terminal hexadecapeptide H‐SNAQTADEERTESKGT‐OH (amino acids 338–353) derived from the human SLC‐1 sequence ([23]). The rat orthologue protein sequence is given in [24]. The peptide is fully identical between the rat and human orthologues. The sequence was checked for uniqueness to SLC‐1 protein using BLAST (peptide sequences against nucleotide databases, Advanced search; GenEMBL complete, April 1998, rechecked every month until September 1999). Amongst putative post‐translational modifications are three N‐glycosylation sites in the N‐terminal presumed extracellular tail (i.e. NAS, NTS, NLT; respective position of the arginine residue: 13, 16 and 23) and several phosphorylation sites in the C‐terminal presumed intracellular tail as follows: two protein kinase C phosphorylation sites, 317–319 TFR; 325–327 SVK and one casein kinase II phosphorylation site (342–345 TADE). The synthetic peptide was covalently NH2‐coupled to the carrier keyhole limpet haemocyanin (KLH) using the glutaraldehyde method. Two New Zealand rabbits were used for each peptide in the immunization procedure. Initially rabbits were injected with 0.5 mg of the peptide–KLH conjugate in Freund's complete adjuvant, then subsequently boosted four times with the same amount of antigen suspended in incomplete adjuvant. The injections were repeated monthly. The bleeds were clotted overnight at 4 °C and the serum separated from blood cells by centrifugation at 8000 g. The sera were stored at −80 °C until affinity purified.

For the initial assessment of the antiserum specificity in immunohistochemical procedures, a micropurification procedure was performed using the immunogenic peptide linked to BSA and bound to a 0.45‐μm nitrocellulose membrane (Sartorius, Gottingen, Germany). Crude antisera were incubated overnight at 4 °C with 100 μg of the peptide spotted on the filter in 100 mm Tris–HCl, pH 8.0, 0.05% (v/v) Nonidet P‐40, 150 mm NaCl and 0.5% (w/v) Marvel fat‐free dried milk. After three washes (10 min per wash in the same buffer), bound antibodies were eluted with 100 mm glycine, pH 2.5. The eluate was then neutralized with 1 m Tris–HCl, pH 8.0 (0.1 vol. of Tris per vol. of glycine).

Once the specificity of the antibody was checked (see below), a larger scale purification of the antisera was carried out using an affinity chromatography column. An N‐terminal extended immunogenic peptide with a cysteine residue was coupled to the Sulfolink matrix following the manufacturer's recommendations (Pierce & Wariner, Chester, UK). Three millilitres of the crude antiserum was applied to 1 mg of the peptide covalently bound to 2 mL of the solid gel in a 50‐mL tube on a rolling agitator overnight at 4 °C. After the resin was washed with 100 mL of 0.1 m PBS for 60 min at 4 °C, specific antibodies were eluted with glycine in a similar procedure to that above, dialysed overnight at 4 °C against 0.1 m PBS and resuspended in storage buffer [0.1 m PBS; 2 g/L BSA; 20% (v/v) normal goat serum (NGS); 30% (v/v) glycerol]. Routinely, the affinity‐purified antibodies were eluted with 10 mL of glycine and neutralized with 1 mL of Tris–HCl, pH 8. The purified antibodies were formed into aliquots and stored at −80 °C until further use.

Cell extracts were prepared by shearing cell pellets in lysis buffer [20 mm Tris, 5 mm ethylene diamine tetraacetate (EDTA), 150 mm NaCl, 0.4% Triton X‐100 (Sigma), 0.4% Nonidet P‐40 (Sigma), pH 7.4] followed by centrifugation at 3300 g. Supernatant was removed and protein concentration was measured by Bradford assay (Biorad, UK).

For Western blot analysis, cell extracts were resolved by sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS–PAGE, 4–20%), loading 10 μg protein per lane and transferred onto nitrocellulose (Schleicher and Schuell, Germany). The membrane was blocked with 10% low‐fat milk (Marvel, UK), 0.5% Tween‐20 in PBS for 1 h and revealed using an Ig fraction of the crude C antiserum (C‐terminal intracellular tail) at 1 : 2000, followed by incubation with peroxidase‐conjugated secondary antibody (Amersham, UK). Bound antibody was detected using the Enhanced ChemiLuminescence kit ECL (Amersham). The specificity of the antibody was confirmed by preabsorption of antibody with free C peptide (2 μm) and by omitting the primary antibody.

Immunosignals were revealed with an ABC peroxidase reporter system as previously described ([15]) or by immunofluorescence. Briefly, adult male Sprague–Dawley rats were anaesthetized with pentobarbital prior to being transcardially perfused with 4% paraformaldehyde in phosphate‐buffered saline (PBS × 1). The brains were removed, postfixed in 4% paraformaldehyde in PBS × 1 for 2 h at 4 °C and transferred into a 20% sucrose/PBS × 1 solution overnight. Following freezing of the brains in isopentane, 35‐μm coronal sections were cut in a cryostat. Sections were stored in 0.5% PFA in PBS × 1 at 4 °C. Sections were incubated in PBS containing 20% methanol and 1.5% hydrogen peroxide for 30 min in order to quench the endogenous peroxidase activity. Following 2 × 5 min washes in PBS × 1, sections were placed in a blocking solution (3% NGS, 2 g/L BSA, 0.1% Triton X‐100 in PBS × 1) for 30–45 min. Sections were then incubated with the primary antiserum for 48 h at 4 °C with gentle agitation. After primary incubation, sections were given three 10‐min washes in 0.3% Triton X‐100 in PBS. Sections were then processed for peroxidase immunostaining using the Vector ABC (avidin : biotinylated‐enzyme complex) amplification system following the manufacturer's recommendations (Vector, UK). Sections were incubated for 1 h at room temperature in Vector goat antirabbit IgG (H + L) with gentle shaking, then given three 10‐min washes in 0.3% Triton X‐100 in PBS × 1, and incubated under the same conditions in the preformed complex avidin–peroxidase in PBS × 1 (prepared at least 30 min prior to use). After a further three 10‐min washes in 0.3% Triton X‐100 in PBS × 1, sections were transferred into Tris‐buffered saline (TBS × 1). Sections were incubated in 3,3′‐diaminobenzidine (DAB) substrate (Vector) for 5–10 min before the reaction was stopped in distilled water. Sections were mounted onto Superfrost polished slides (BDH, UK), allowed to air dry, then coverslipped with DPX mountant.

Immunochemical fluorescence was carried out as previously described ([16]) using one of several different streptavidin‐conjugated fluorophores (Cy2, Cy3, Cy5, Red‐Phycoerythrin, FITC, TRITC, AMCA; all from Jackson Laboratories, USA).

Wild‐type or transfected human embryonic kidney (HEK) 293 cells were grown on LabTek slides (Life Technologies) and fixed for 15 min in 4% w/v PFA in 0.1 m PBS. The fluorescence immunocytochemical procedure was similar to that already described for immunohistochemistry.

Control experiments included the omission of the primary antiserum, the use of rabbit preimmune serum and preabsorbing the antiserum with the immunogenic peptide. Pre‐absorption controls were performed with 10 μm of the immunogenic peptide (incubated overnight with the antiserum prior to the incubation on sections). Specificity of the antisera was investigated on SLC‐1–HEK 293 transfected cells versus wild‐type cells, with or without preabsorption, with or without primary antisera in an ICC fluorescence procedure as already reported ([8]). Specificity of the C antiserum has already been described elsewhere ([8]).

Brain regions and nuclei were identified using the rat brain atlases of [31] for sagittal sections and [39] for coronal sections. The nomenclature mainly follows that adopted in the report by [6], describing the neuroanatomical distribution of the MCH system in the rat brain.

ISH data were captured with a Pulnix TM‐765 black and white camera coupled to a TAMROM SP. Peroxidase IHC data were captured with a JVC 3CCD KY‐F55B colour video camera or a Leica DC 200 digital camera. Fluorescence IHC and ICC data were analysed with a Leica TCS confocal DM RB microscope. Confocal images were either taken as a 0.8‐μm‐thick single plane or as a stack of several planes (and thereafter reconstructed in three dimensions through a Projection function). Each single plane was averaged four times.

Image acquisition for ISH and peroxidase data was carried out with Image ProPlus software (Media Cybernetics, USA). Images were taken with the highest level of contrast and brightness. Images were imported into PaintShop Pro Version 5.0 (Jasc Software, USA). Image modification involved only the transformation of colour images into black and white images using a grey‐scale function, and adjusting the levels of brightness and contrast. Some of the images had background filtered through a median cut noise function.

The primary data generated by Taqman RT‐PCR consist of a threshold cycle value, indicating the PCR cycle at which the PCR product is detectable above an arbitrary threshold. The system was calibrated using known numbers of copies of genomic DNA. When the threshold cycle generated by these standards is plotted against the genomic DNA copy number on a logarithmic scale, the data points lie on a straight line (Fig. 1A). Using the threshold cycle values resulting from the cDNA samples from various tissues, a copy number can be read from this calibration curve.

Graph: 1 Quantitative RT‐PCR analysis and mRNA localization in rat brain. (A) Calibration curve relating threshold cycle to copy number plotted on a logarithmic scale. Known quantities of genomic DNA lie on a straight line as indicated by filled circles. Threshold cycle values for the cDNA samples are fitted to this line and copy number values are given on the x‐axis. Open circles indicate hypothalamic cDNA values as an example. (B) SLC‐1 mRNA expression in rat CNS and PNS. SLC‐1 expression has been normalized to the expression of the housekeeping gene GAPDH. Bars indicate the mean values derived from three independent RT‐PCRs; error bars indicate standard error of the mean. SLC‐1 mRNA is detectable in all areas tested, with the highest levels of expression observed in amygdala, cerebral cortex (all divisions), hippocampus, hypothalamus and substantia nigra.

In this study, copy numbers for SLC‐1 have been normalized to GAPDH copies to control for variations in RNA quality and loading. Only slight variations in GAPDH mRNA expression were observed between the tissues that were included in the study (data not shown). Our results (Fig. 1B) show that SLC‐1 is widely expressed in the rat nervous system, with mRNA detected in all tissues that were tested. However, some variation in expression is observed, with higher expression in amygdala, cerebral cortex (all divisions), hippocampus, hypothalamus and substantia nigra, and lower levels of expression in striatum, thalamus, cerebellum, rhombencephalon, spinal cord and dorsal root ganglia (DRG).

Specific immunocytochemical signals (Fig. 2A) largely confined to the plasma membrane, as opposed to the control condition using the preimmune serum (Fig. 2B) or the antiserum preabsorbed with the synthetic peptide (Fig. 2C), were generated with SLC‐1‐transfected HEK‐K1 cells incubated with the antiserum.

Graph: 2 Specificity of the antisera as tested by immunocytochemistry and Western blot using SLC‐1‐transfected HEK cells. HEK 293‐SLC‐1‐transfected cells grown on chamber microscopic slides were incubated with affinity‐purified MCH receptor antiserum (A), with no primary serum (B) and with affinity‐purified MCH receptor antiserum preabsorbed with the synthetic peptide. (C) Specific immunostaining can be seen in condition A. Crude antiserum C detected several bands in transfected cells (D, lane 1) in comparison with untransfected (D, lane 3) HEK 293 cells. Bands were identified at 60 kDa and may represent the glycosylated form of the receptor. High molecular weight forms may represent receptor aggregates. Pre‐absorbed antibody failed to detect these bands in transfected (D, lane 2) or untransfected (D, lane 4) cells. Calibration bars (A–C), 150 μm.

Western blot analysis using the crude C antiserum revealed several specific bands in SLC‐1‐overexpressing cells. Two prominent molecular forms were present at 60 and 120 kDa (Fig. 2D; lane 1). Because the predicted molecular weight of SLC‐1 is 38.9 kDa, is likely that the 60‐kDa band represents a glycosylated form of the receptor, whereas the 120‐kDa band may represent receptor aggregates. Immunosignals were absent with wild‐type HEK cells (Fig. 2D; lane 3) or when the antiserum had been preabsorbed with the synthetic peptide (2 μm) and tested either with SLC‐1‐transfected (Fig. 2D; lane 2) or wild‐type HEK cells (Fig. 2D; lane 4).

A substantial amount of specific autoradiographic signal was obtained, with discrete anatomical localization, using the radiolabelled antisense oligoprobes used in the study (Fig. 4A). In control experiments, sense oligoprobes produced no substantial tissue labelling (Fig. 4B). Furthermore, preincubation of radiolabelled antisense oligoprobes with an excess of cold antisense probes (Fig. 4C), and the use of RNaseA pretreated sections (Fig. 4D), similarly resulted in a diminution of the signal intensity compared with that obtained by using antisense oligoprobes (Fig. 4A).

Graph: 4 Localization of the SLC‐1 mRNA and protein in a series of coronal forebrain sections: isocortex, olfactory system, nucleus accumbens, caudate‐putamen, medial septum, and nucleus of the diagonal band. ISH (A, C, E, G and I) and IHC (B, D, F, H and J). In particular, note also the overlapping between the mRNA (A, C, E, G and I) and the protein (B, D, F, H and J) distribution. Calibration bars, 0.2 cm. See Abbreviation list for abbreviations.

A distinctive immunostaining pattern was obtained after incubation of brain sections in the affinity‐purified C antiserum (Fig. 3E). In control experiments using the C antiserum preabsorbed with the immunogenic peptide, no signals could be observed (Fig. 3F).

Graph: 3 Specificity of the oligoprobes (A–D) and antisera (E–G) using rat brain tissue sections. mRNA in situ hybridization. Consecutive rat brain sagittal sections were hybridized with a mix of SLC‐1‐radiolabelled antisense oligonucleotides (A), a mix of SLC‐1‐radiolabelled sense oligonucleotides (B), a mix of SLC‐1‐radiolabelled antisense oligonucleotides and a 100 × excess of unlabelled radiolabelled antisense oligonucleotides (C). (D) Rat brain sagittal sections were pretreated with the RNaseA and were hybridized with a mix of SLC‐1‐radiolabelled antisense oligonucleotides. The use of sense oligoprobe (B), competed radiolabelled antisense oligoprobes with an excess of cold antisense probes (C), and the use of RnaseA‐pretreated sections (D), resulted in an absence (B) or diminution (C and D) of the signal intensity compared with that obtained by using antisense oligoprobes (A). Specific signals corresponding to the SLC‐1 mRNA were observed in the cortex (CTX), the olfactory regions (anterior olfactory nucleus and olfactory tubercle, respectively, AON and OT), the basal ganglia (caudate‐putamen and amygdala, respectively, CP and A), the hippocampal formation (hi), the diencephalon (thalamus, habenula and hypothalamus, respectively, Th, H, Hyp) and various midbrain areas (superior and inferior colliculi, pons, reticular formation, nucleus of the solitary tract and cerebellum, SC, IC, Pn, RF, NTS and Cb, respectively). Immunohistochemistry (E and G) SLC‐1‐immunoreactive patterns on rat brain sagittal sections. The rat sagittal sections were incubated with the affinity‐purified followed by immunohistochemistry reported with peroxidase (E and G). Immunostained regions were similar to the regions containing mRNA (A). Immunosignals were observed in the cortex (CTX), the olfactory regions (AON, OB and OT), the basal ganglia (CP), the hippocampal formation (hi and SUB), the diencephalon (LHA, MM) and various midbrain areas (SC, IC, Pn, NLL, NTS, SPV, Cb, MV and the reticular formation). When the section was incubated with the antiserum preabsorbed with the synthetic peptide, no signals were observed (F). There was a good overlapping between the distribution of the SLC‐1 protein‐like immunoreactivity and the MCH peptide (H; adapted from [6]; a schematic representation of the MCH projections through the rat brain). Calibration bars, 0.33 cm. See Abbreviation list for abbreviations.

At a same anatomical level, the SLC‐1‐immunostained regions in the rat brain (Fig. 3G) were largely identical with MCH‐immunostained regions in the rat brain (Fig. 3H).

SLC‐1‐like immunoreactivity was found in cells with the morphology of projection neurons and interneurons, but not in cells with glial characteristics. Immunosignals were largely restricted to the plasma membranes of labelled cells (Fig. 9, 10C,E,F, 11H,2A,G,H). Confocal microscopy of tissue sections revealed mainly discrete punctate staining of plasma membranes (Fig. 9D,F) which was in contrast to the more continuous pattern of staining of entire cell membranes seen in SLC‐1‐transfected cells (Fig. 2A).

Graph: 9 Microscopic observations of the SLC‐1 immunoreactivity in the isocortical areas of the rat brain. Numerous SLC‐1 immunostained cells were, respectively, detected throughout the isocortex, e.g. in the ventral (A) and dorsal (B) parts of the retrosplenial cortex (respectively, RSPv and RSPd), the primary motor cortical area (MOp; Fig. 10C,D; D by confocal microscopy), the secondary motor cortical area (MOs; Fig. 10B), the primary somatosensory cortex (SSs; Fig. 10E), the gustatory area (GU; Fig. 10F by confocal microscopy), the posterior‐parietal region association area (PTLp; Fig. 10G) and the primary auditory cortex (AUDp; Fig. 10H). In all cases, cells were labelled on the membrane. Calibration bars, 130 μm (A, C and H); 270 μm (B); 65 μm (E–G).

Graph: 10 Microscopic observations of the SLC‐1 immunoreactivity in the olfactory regions, hippocampal formation, basal ganglia and other forebrain regions. A strong immunostaining was observed in the olfactory tubercle (OT) illustrated with a sagittal (A) and coronal (B) section. Sparse immunostained cells were detected in the claustrum (CLA) (C). There is prominent immunostaining of the layer II in the piriform cortex (Pir) (D). Immunostained aspiny neurons were located in the caudate‐putamen (CP) (E). MCH receptor‐immunoreactive cells were found in the substantia nigra (F). In the hippocampus, immunostaining was observed in the pyramidal layer of the CA fields (G, I and J, by confocal microscopy) and in the granular cells of the dentate gyrus (DGsg) (H and I; I by confocal microscopy). Calibration bars, 540 μm (A and G); 270 μm (B, D, H and I); 65 μm (C, E and J); 130 μm (F).

Graph: 11 Microscopic observations of the SLC‐1 immunoreactivity in the diencephalon: hypothalamus, thalamus and epithalamus. A dense population of MCH receptor SLC‐1‐immunostained cells was detected in the supraoptic nucleus (SON; A) and the arcuate nucleus (AN, B). At the level of the paraventricular nucleus (PVH), there was a more pronounced staining of the lateral zone of the posterior magnocellular region (PVHpml) as opposed to a lighter immunostaining observed in the dorsal parvicellular, medio‐parvicellular and dorsal zone of the medial parvicellular region (respectively, PVHdp, PVHmpv and PVHpmd) (C). Immunostained cells were also detected in the dorsomedial and anterior nucleus (here both illustrated in the posterior region; respectively, DMHp; D and AHNp; E). In the posterior hypothalamus, MCH receptor‐immunostained cells were seen in the tuberomammillary nucleus (TMv; F). More dorsally in the diencephalon, immunostained cells were present within the medial habenula (MH; G) and in the reticular thalamic nucleus (RT; H). Calibration bars, 270 μm (A–G); 65 μm (H).

Isocortex. Strong labelling both at mRNA (e.g. Figure 4A,C,E,G,I) and protein (e.g. Figure 4B,D,F,H and J) levels was observed in the cerebral cortex (Table 1). Many isocortical subregions were positive − amongst them the primary and secondary motor area (respectively, MOp and Mos, see Fig. 4A for mRNA; Fig. 4B for protein), primary and secondary somatosensory areas (respectively, SSp and SSs, see Fig. 4C,G for mRNA; Fig. 4D,H for protein), gustatory area (GU, see Fig. 4A for mRNA; Fig. 4B for protein), anterior cingulate area (ACA, see Fig. 4A for mRNA; Fig. 4B for protein), visceral area (VISC, see Fig. 4G for mRNA; Fig. 4H for protein), agranular insular area (Alv and Ald, see Fig. 4C for mRNA; Fig. 4D for protein), retrosplenial cortex (RSP, see Fig. 5E for mRNA; Fig. 5F for protein), posterior–parietal region association area (PTLp), dorsal auditory area (AUDd), ventral auditory area (AUDv), ventral temporal association area (TEv), the ectorhinal area (ECT, see Fig. 6A for mRNA; Fig. 6B for protein), primary auditory area (AUDp), primary visual area (VISp), anterolateral, rostrolateral and anteromedial visual areas (respectively, VISa, VISrl and VISam, see Fig. 6C for mRNA; Fig. 6E for protein), posteriomedial and posterolateral visual areas (respectively, VISpm, VISpl, see Fig. 7G for protein) and the claustrum (CLA, see Fig. 4C for mRNA and Fig. 4D for protein). Immunostaining was more particularly concentrated in layers III and V (Fig. 4E). Cells immunostained in each of these areas were labelled on their surface membranes (Figs 9 and 10C). Most stained cells resembled multipolar interneurons (Fig. 9).

1 Distribution of slc‐1 mRNA and immunoreactivity in the rat brain

1 The relative density of labelling is classified as absent (–), sparse (+), moderate (++), extensive (+++), or not determined (ND).

Graph: 5 Localization of the SLC‐1 mRNA and protein in a series of intermediate coronal forebrain sections: dorsal hippocampus, amygdala, thalamus and hypothalamus. ISH (A, C, E and G) and IHC (B, D, F and H). Note the strong signals obtained in the amygdaloid regions (Amygd.) and the staining obtained in various hypothalamic (SO, AHN, PVH, MEPO, MPO, MPN, VMH, LHA, PM, PV) and thalamic (RT, AV, AD, IAD, MH, CL, PVT, OP, STN) nuclei, and the strong staining in the hippocampus. Note also the absence of staining in the corpus callosum (cc). Calibration bars, 0.2 cm. See Abbreviation list for abbreviations.

Graph: 6 Localization of the SLC‐1 mRNA and protein in a series of coronal forebrain–midbrain sections: dorsal and ventral hippocampus, amygdala, posterior thalamus and hypothalamus, substantia nigra. ISH (A, C and G) and IHC (B, D–F and H). Note the signals obtained in various posterior hypothalamic (PH, LM, MM) and thalamic nuclei (PF, VPL), in the substantia nigra pars reticulata (SNpr), the zona incerta (ZI) and the fasciculus retroflexus (fr). In the hippocampus, both the strata pyramidale and oriens (respectively, sp and so) were immunostained. Calibration bars, 0.2 cm. See Abbreviation list for abbreviations.

Graph: 7 Localization of the SLC‐1 mRNA and protein in a series of coronal midbrain sections: ventral and dorsal hippocampi, posterior thalamus, substantia nigra/geniculate nuclei, peri‐aqueductal grey matter, superior and inferior colliculi, pre‐ and post‐cerebellar nuclei. ISH (A and C) and IHC (B and D–H). Note the signals obtained in the geniculate nuclei (lateral and medial, respectively, LG and MG) and in their functionally associated colliculus nuclei (SC and IC). Note also the signals obtained in the oculomotor nerve (III), the caudal part of the substantia nigra (SNpr), the specific staining of the caudal part of the interpeduncular nucleus (IPN) and the staining in other various midbrain nuclei (PAG, VTA, APN, RN, NB,DR, VTN, TRN, NTB, POR, PB‐KF, DTN and VCO). Calibration bars, 0.2 cm. See Abbreviation list for abbreviations.

Olfactory cortex. This region was one of the richest in terms of both SLC‐1 mRNA and protein labelling. Heavy labelling, both at mRNA and protein level, of the olfactory region was seen in the anterior olfactory nucleus (AON, see Fig. 4A for mRNA and Fig. 4B for protein) and olfactory tubercle (OT, see Fig. 4C for mRNA; Figs 4D and 10A,B for protein). Dense signals were observed in the taenia tecta (TT), both in the ventral part (TTv) and the dorsal part (TTd, see Fig. 4C for mRNA; Fig. 4D for protein), and in the major islands of Calleja (islm), both at mRNA (Fig. 4E) and protein level (Fig. 4F). Strong mRNA and protein signals were observed in the piriform cortex (PIR, e.g. Figure 4A for mRNA; Fig. 4B for protein), while much weaker immunosignals were seen in the post‐piriform transition area (TR, e.g. Figure 6E). The endopiriform nucleus (EP) was also distinctly labelled for mRNA (Fig. 4I) and protein (Fig. 4J) presence.

Hippocampal formation. Immunolabelled cells were observed in the dorsal parts (SUBd, see Fig. 6G for mRNA; Fig. 6H for protein) and ventral parts (SUBv, see Fig. 7A for mRNA; Fig. 7B for protein) of the subiculum (SUB). The postsubiculum (POST) and presubiculum (PRE) were also immunostained, unlike the parasubiculum (PAR, Fig. 7D,F). The entorhinal area (ENT) was stained in its lateral part (ENTl, see Fig. 6C for mRNA; Fig. 6E for protein). In the hippocampus (hi), all subfields CA1–4 were both positive for mRNA (Fig. 6G) and protein (Fig. 6H). Immunosignals were mainly located in the stratum pyramidale (sp) in Ammon's horn (Fig. 10I and J). The granule cell layer of the dentate gyrus was densely stained both for mRNA (Fig. 5G) and protein (Fig. 10H,I).

Amygdala. The MCH receptor gene SLC‐1 is strongly expressed in the amygdaloid regions (e.g. Figure 5A,G), part of the basal ganglia. mRNA signals were seen in the basolateral, basomedial, central, medial and cortical nuclei as well as the lateral and posterior nuclei (respectively, BLA, BMA, CEA, MEA, COA, LA and PA, see Fig. 5A,G for mRNA and Fig. 5B,D,F for protein).

Septal regions. The septum was particularly enriched in SLC‐1 signals with labelling in the medial septum nucleus (MS) and the nucleus of the diagonal band of Broca (NDB) both at mRNA (e.g. Figure 4G) and protein level (Fig. 4H). On a sagittal section, immunostaining could be seen in the horizontal limb of the diagonal band (HDB, Fig. 3E) and the vertical limb of the diagonal band (VDB, Fig. 3G). At mRNA level, the dorsal (LSd) and ventral (LSv) segments of the lateral septum (LS) were labelled, as was the bed nucleus of the stria terminalis (BST). This was paralleled by protein presence (Fig. 4J).

Corpus striatum. The caudate‐putamen was labelled at both mRNA (Fig. 3A) and protein levels (Fig. 3E). Membrane‐immunostained aspiny cells were seen within the neostriatum − they most probably represent interneurons (Fig. 10E). Robust immunolabelling was present within the nucleus accumbens (Fig. 4D). The globus pallidus (GP) was not labelled for either mRNA (Fig. 4I) or protein (Fig. 4J). The substantia innominata (SI) and magnocellular preoptic nucleus (MA) were immunostained (Fig. 4H and J), as well as labelled in in situ experiments (Fig. 4I).

Epithalamus. In the epithalamus, signals were observed in the medial habenula (MH, e.g. Figure 4E for mRNA; Fig. 4F for protein) but not in the lateral habenula. Numerous immunostained cells could be seen in the MH (Fig. 11G).

Thalamic nuclei. The thalamus was reasonably enriched with SLC‐1 signals. Progressing rostrally to caudally, on a ventral‐caudal axis, mRNA and protein SLC‐1 receptor signals were detected in the paraventricular nucleus thalamus (PVT), anterodorsal nucleus thalamus (AD), interanterodorsal nucleus thalamus (IAD), anteroventral nucleus thalamus (AV), central medial nucleus thalamus (CM), anteromedial nucleus thalamus (AM), reticular nucleus thalamus (RT), lateral dorsal nucleus thalamus (LD), parathenial thalamus nucleus (PT), nucleus reuniens (RE), ventral anterior‐lateral complex thalamus (VAL, see Fig. 5C for mRNA; Fig. 5D for protein), ventral medial nucleus thalamus (VM), ventral postero‐lateral nucleus thalamus (VPL, see Fig. 5E for mRNA; Fig. 5F for protein), central lateral nucleus (CL, Fig. 5H), posterior complex thalamus (PO, see Fig. 6A for mRNA; Fig. 6B for protein), ventral postero‐medial nucleus thalamus (VPM), ventral postero‐lateral nucleus thalamus (VPL) and the parafascicular nucleus thalamus (PF, see Fig. 6C for mRNA; Fig. 6D for protein). Microscopic observations revealed immunostained cells in the RT (Fig. 11H). The subthalamic nucleus (STN) (Figs 6A and 7B) and overlying zona incerta (ZI, Fig. 6C,D) were also positive for SLC‐1 for mRNA and protein. More rostrally, the dopaminergic cell group of the zona incerta (ZIda) was also immunostained (Fig. 5G).

Metathalamus. Both the medial geniculate (MG) and lateral geniculate (LG) contained SLC‐1 mRNA (Fig. 6C) and protein (Fig. 6F).

mRNA signals were evident in the hypothalamus. The median preoptic nucleus (MEPO) and the medial preoptic area (MPO) encompassing the medial preoptic nucleus (MPN) were the most rostral hypothalamic regions to be labelled (Fig. 4I). Strong signals were obtained at the level of the anterior nucleus (AHN) and paraventricular nucleus (PVH, Fig. 5A), supraoptic (SO) and arcuate nucleus (AN, Fig. 5C), ventromedial nucleus (VMH, Fig. 5C), the LHA, the intermediate (PVi) and posterior (PVp) periventricular nuclei (Fig. 6A), medial mammillary nucleus (MM) and posterior nucleus (PH, Fig. 6C). There was a generally good overlap between mRNA‐positive regions and ‐immunostained regions. Microscopic observation revealed groups of immunostained cells in the SON (Fig. 11A), the AN (Fig. 11B), the PVH (Fig. 11C), the DMH (Fig. 11D), the AHN (Fig. 11E) and the tuberomammillary nucleus (TMv, Fig. 11F).

Visual areas. Signals were present within the superior colliculus (SC), more particularly concentrated in the superficial (SCsg) and intermediate (SCig) grey layers (Fig. 7C for mRNA; Fig. 7D for protein). In the pretectal regions, immunostaining was present in the anterior pretectal nucleus (APN) and the olivary pretectal nucleus (OP, Fig. 6H).

Somatosensory areas. The principal sensory nucleus of the trigeminal nerve (PSV) was stained both for mRNA (Fig. 6C) and protein (Fig. 6F), as well as the spinal nucleus of the trigeminal nerve (SPV, Fig. 6G–H). Oral, intermediate and caudal segments of the spinal nucleus were immunolabelled (Fig. 3E,G).

Auditory areas. Both the dorsal (DCO) and ventral (VCO) cochlear nucleus were labelled. mRNA was present within the anterior part of the VCO (VCOa, Fig. 8A) as well as the protein (Fig. 8D). Immunostaining was present in the DCO (not shown). Strong immunostaining was found in the nucleus of the lateral lemniscus (NLL, Fig. 7F), as well as in the superior olivary complex (SOC, Fig. 8F) where mRNA could also be detected (Fig. 8C). The inferior colliculus (IC) was particularly rich in SLC‐1 signals (Fig. 7H).

Graph: 8 Localization of the SLC‐1 mRNA and protein in a series of hindbrain sections and in the lumbar segment of the spinal cord. ISH (A, C, E and H) and IHC (B, D, F–H and J). Note the signals obtained in various nuclei of the reticular formation (PRN, MARN and GRN), in the cochlear nuclei (DCO and VCO), in the vestibular nuclei (MV), in the olivary nuclei (SOCl and IO), in the trigeminal system (PSV and SPV), in the facial nerve (VII), in the locus coeruleus (LC) and the grey matter of the spinal cord. Calibration bars, 0.13 cm (A–G); 0.07 cm (H); 0.065 cm (I and J). See Abbreviation list for abbreviations.

Vestibular areas. The medial vestibular nucleus (MV) as well as the rostral zone of the rostral part of the nucleus of the solitary tract (NTSrm) were immunostained (Figs 3G and 8H).

Visceral areas. The NTS contained both mRNA (Fig. 3A) and protein in its intermediate (NTSi) and ventrolateral (NTSvl) parts (Fig. 3E). Similarly, in both its lateral (PBl) and medial‐medial (PBmm) divisions, the parabranchial nucleus (PB) contained mRNA (Fig. 8C) and immunostaining (Fig. 8D). The Kolliker–Fuse nucleus (KF, Fig. 8D) was also labelled.

The nucleus of cranial nerve II (oculomotor) was immunostained (Fig. 7G).

The facial nucleus (cranial nerve VII) was immunostained (Fig. 8H).

Robust labelling was observed in the substantia nigra (SN); mRNA was evident in the substantia nigra pars compacta (SNpc, Fig. 6G), and immunolabelling in both the pars reticulata (Fig. 6E) and to a much lesser extent in the pars compacta (SNpc, Fig. 10F). The ventral tegmental area (VTA) was immunostained (Fig. 7D).

The red nucleus (RN) contained obvious immunolabelling (Fig. 7B), as well as the pontine central grey (PG) and the tegmental reticular nucleus (TRN, Fig. 7D,F). The central grey contained a particularly high amount of mRNA signal (Fig. 7C). More caudally, the inferior olivary complex (IO) was immunopositive (Fig. 8G).

There was strong mRNA labelling in the granular cell layer of the cerebellar cortex (Cbgr, Fig. 4A) as well as immunolabelling (Figs 3G and 12C). In addition, the interpositus cerebellar nucleus was immunostained in both its anterior and posterior subdivisions (IntA; IntP, Fig. 3E).

Graph: 12 Microscopic observations of the SLC‐1 immunoreactivity in the midbrain, hindbrain and lumbar spinal cord. Immunostained MCH receptor staining was detected in the peri‐aqueductal grey matter (PAG; A), the pontine grey (PG; B), the granular layer of the cerebellum (Gr; C) and in the locus coeruleus (LC; D). In the spinal cord, SLC‐1‐like immunoreactivity was present within the grey matter of the spinal cord with a more pronounced staining of the dorsal horn as opposed to the ventral one (E and F). In both horns, immunostained cells were detected (G and H). The spinal tracts, e.g. the laterocorticospinal tract (lct), laterospinocortical tract (lsc), lateral spinothalamic tract (lst) and the anterior corticothalamic tract (act) were moderately labelled (E). Calibration bars, 65 μm (A, D and H); 270 μm (B, C and F); 540 μm (E).

Central grey of the brain. Both mRNA and protein were detected in the peri‐aqueductal grey (PAG, Fig. 7A,B). Some membrane‐labelled cells were detected there (Fig. 12A). The ventral (VTN) and dorsal (DTN) tegmental nuclei were also immunostained (respectively, Figs 7H and 8B). mRNA was identified in the DTN (Fig. 8A). The locus coeruleus (LC) was positive for both the SLC‐1 mRNA (Fig. 8E) and protein (Figs 8F and 12D). The LC staining was encompassed within the immunostained pontine central grey (PCG, Fig. 8F). Immunoreactivity was also observed in the supragenual nucleus (SG, Fig. 8C).

Raphe nucleus. Immunostaining was observed in the dorsal nucleus raphe (DR) and the medial part of the superior central nucleus raphe (CSm, Fig. 7H). mRNA was also detected in the DR (Fig. 8A). More caudally, immunostaining was present in the nucleus raphe pallidus (RPA, Fig. 8H).

Interpeduncular nucleus. There was a pronounced labelling of the interpeduncular nucleus (IPN) both at mRNA (Fig. 7A) and protein (Fig. 7B) levels. The staining was restricted to the central subnucleus (IPNc).

A slight staining was observed in the caudal pontine reticular nucleus (PRN) at both mRNA (Fig. 8A) and protein (Fig. 8B) levels. A weak immunostaining was also revealed in the gigantocellular reticular nucleus (GRN, Fig. 8H), while the signal was stronger ventrally to the GRN in the magnocellular reticular nucleus (MARN, not shown). On a sagittal section (Fig. 3G), staining was confirmed in the PRN and GRN, and was also found in the medullary reticular nucleus (MDRN).

There was strong mRNA labelling (Fig. 9I) and immunolabelling (Fig. 9J) in the lumbar part of the spinal cord. All subdivisions of the grey matter (dorsal and ventral horns) were labelled (ventromedial, dorsomedial, intermediolateral, central, ventrolateral, dorsolateral and retrodorsolateral). The immunolabelling was particularly more pronounced in the dorsal horn (Fig. 12E,F). Microscopical observation revealed immunostained cell bodies in both the ventral (Fig. 12G) and dorsal (Fig. 12H) horns.

Most fibres were devoid of mRNA or immunolabelling (e.g. white matter tracts such as the corpus callosum and cerebral peduncles, Fig. 6H). However, the fasciculus retroflexus was immunostained (Fig. 7E). There was also dense immunoreactivity in some of the white matter tracts of the spinal cord, e.g. the laterocorticospinal tract (lct), the laterospinocortical tract (lsc), the lateral spinothalamic tract (lst) and the anterior corticothalamic tract (act, Fig. 12E).

By virtue of certain sequence homologies, SLC‐1 was originally described as a somatostatin‐like receptor ([23]; see also [9]). It has recently been discovered ([8]; [34]; [1]; [25]; [37]) that SLC‐1 represents a receptor for the nonadecapeptide MCH.

We generated radiolabelled oligonucleotides and polyclonal antisera to perform in situ hybridization and immunohistochemical studies to provide detailed maps of the distribution of the mRNA and protein products of this receptor throughout the rat central nervous system. Western blotting experiments, using cell lysates from SLC‐1‐transfected cells, were performed to establish the specificity of the antisera used here. Analysis of these experiments revealed several molecular forms of SLC‐1 in transfected HEK293 cells. While the predicted molecular weight of the protein product of the gene is 38.9 kDa, we detected higher molecular weight forms in the blots that may correspond to glycosylated forms of the receptor (viz. bands at 60 kDa) or receptor aggregates (120‐kDa bands). The presumed N‐terminal extracellular part of the SLC‐1 receptor contains three consensus sequences that would be suitable for N‐glycosylation (PTGPNASNTSDGPDNLT).

There was generally good agreement between the results of the present study and other studies ([23]; [34]; [25]) that used in situ hybridization to partially examine tissue‐specific slc‐1 gene expression profiles within the rat brain. We have detected slc‐1 gene expression in some brain regions that were not mentioned in the earlier studies, e.g. the cerebellum. This result is at odds with the report by [23] showing an absence of SLC‐1 mRNA in the cerebellum. Using two independent gene expression techniques (Taqman and in situ hybridization), we identified slc‐1 gene expression in the cerebellum. Both in situ studies and immunohistochemistry revealed labelling in the granular cell layer. The reason for the different findings of the two studies is unclear, but a similar lack of clarity surrounds the issue of cerebellar somatostatin receptors (e.g. see [15]).

When present, levels of expression of both mRNA and protein were often quite high with some particular 'hot‐spots' in the isocortex, olfactory cortex, amygdala, hippocampal formation, hypothalamus, thalamus and pons. Expression was generally widespread throughout the brain encompassing major cell groups in the forebrain, midbrain, hindbrain and spinal cord.

There was also a clear and generally precise overlap between the distributions of the slc‐1 mRNA and protein. Some regional differences were, however, observed. This may be because some of the receptors were axonally transported to sites distal to the somata where they originated. Interestingly, there was a very good match between the receptor distributions described in the present study with the distribution of MCH‐immunoreactive profiles within the rat brain reported by [6], [40], [11] and [5].

The immunostaining described in the present study seemed to be restricted to the neuronal population. Furthermore, signals were mainly confined to plasma membranes of cells, as would be expected for a G‐protein‐coupled receptor. Confocal fluorescence analysis revealed a different pattern of surface membrane labelling in brain sections (with punctate labelling) versus the 'continuous' labelling seen on the membranes of SLC‐1‐transfected HEK 293 cells. This may well be explained by the different densities of receptors in these cells, with a much higher abundance of receptors in the transfected cells.

Staining seen with the antiserum C and Red‐Phycoerythrin (RPE) as a fluorescent reporter was generally weak. RPE bears more than 40 fluorophore groups per molecule, and may therefore have had difficulty passing through the cell membrane to target the intracellular epitope. The antiserum revealed some nerve fibre staining. Axons with collaterals and dendrites from the many interneurons were labelled in the cerebral cortex. These findings are consistent with the MCH immunocytochemical study of [6].

The widespread and generally dense cortical distribution of the MCH receptor is consistent with the suggestion that the peptide is involved in generalized cortical arousal and sensorimotor integration ([6]; [29]). The SI and PB, both SLC‐1 immunopositive and receiving dense MCH innervation, are cell groups whose cortical projections conform to the description of non‐specific cortical afferents. Neurons in the lateral hypothalamus and zona incerta are also associated with diffused arousal functions and sensorimotor integration (see [6]). Strong immunostaining was detected in these areas as well. Together with the presence of SLC‐1 labelling in many parts of the limbic system and the medial septum, these findings may explain the role of MCH in sensory conditioning, as the peptide diminishes the ability of rats to appropriately filter sensory clues ([28]). In keeping with this, our results appeared to indicate a particular representation of SLC‐1 labelling in anatomical regions implicated in the control of vigilance. The dense labelling seen in the piriform cortex could be related to the antiseizure activity of the MCH reported by [21].

The hippocampal formation displayed dense signals for both SLC‐1 mRNA and protein. The subiculum, Ammon's horn (stata oriens and pyramidale) and dentate gyrus (granular layer) regions were strongly immunoreactive to the SLC‐1 antisera. Together with the immunostaining observed in other forebrain regions, e.g. the amygdaloid regions (see below) and the bed nucleus of the stria terminalis, these results may functionally be related to the action of MCH on passive avoidance, a behavioural paradigm associated with cognition and learning ([27]).

Elements of the basal ganglia were all immunostained, with the exception of the globus pallidus. Immunostaining was very high in the amygdala, relatively high in the substantia nigra and subthalamic nucleus, and high in the neostriatum. In the substantia nigra, mRNA was observed particularly densely in the pars compacta, while the immunostaining was heavy in the pars reticulata. This is reminiscent of the situation with the somatostatin receptor 1 (sst1) where there appeared to be a transcription of the sst1 gene in the cell bodies of the pars compacta but an expression of proteins in the dendrites in the pars reticulata (see [15]). SLC‐1 labelling was also densely represented in the extrapyramidal motor system and throughout many areas of the mesencephalic, pontine and medullary reticular formations associated with locomotor activity. Much of this labelling may also be concerned with a possible involvement of MCH in central pattern generator circuitry as well as brainstem‐controlled motor behaviour. It should be noted, however, that targeted deletion of the MCH gene does not appear to affect locomotor behaviour in the mutant mice ([36]), and MCH itself has no action on locomotor activity but antagonizes MSH‐induced hyperlocomotor behaviour ([35]).

SLC‐1 was strongly present throughout the diencephalon. Many of the thalamic nuclei were labelled and our data are in agreement with the high density of MCH fibres observed within the thalamus ([6]). Strong immunostaining was observed in the fasciculus retroflexus, a cholinergic bundle originating in the habenula and projecting to the interpeduncular nucleus and various paramedian midbrain nuclei. All of these regions have been shown to be MCH immunopositive as well ([6]). This suggests the existence of MCH modulation of cholinergic neurotransmission, acting through the SLC‐1 receptor, on an axis originating from the epithalamus and terminating in the reticular formation. MCH neurons have been shown to be responsive to acetylcholine stimulation, and carbachol has been shown to induce a rapid increase in hypothalamic MCH mRNA expression ([4]).

Dense immunostaining was found in the zona incerta, a subthalamic nucleus involved in drinking behaviour. This could be linked with the observed effects of dehydration and salt‐loading on MCH gene expression activity, and the effect of MCH on regulating the hydro‐mineral balance in the gastrointestinal tract ([17]).

Much SLC‐1 immunoreactivity appears to be related to processing systems for visual and auditory stimuli. The visual pathway conveys regulatory information through the lateral geniculate nucleus, a thalamic relay receiving afferents from the retina, the parabrachial region, the hypothalamic tuberomammillary region and the superior colliculus, and sending projections to the visual cortex and the reticular thalamic nucleus. Information is then relayed back to the dorsal thalamus. All of these regions exhibit substantial SLC‐1 immunoreactivity. Interestingly, [34] have also reported slc‐1 gene expression in the eye.

SLC‐1 receptors are also probably involved in auditory processing as immunostaining was observed in the auditory cortex, medial geniculate nucleus, inferior colliculus, the lateral lemniscus, the ventral and dorsal cochlear nuclei, and the olivary complex.

SLC‐1 is clearly involved in many diverse motor and sensory systems. Staining was also observed in specific mid‐ and hindbrain nuclei involved in these processes, e.g. the oculomotor nucleus, red nucleus and the anterior pretectal area. The latter receives afferents from the retina and visual association cortex and controls pupillary reflexes. It may be that MCH, acting through SLC‐1 receptors, plays a role in general sensorimotor integration in these systems.

More than 90 000 hypothalamic cells express the MCH gene (see [20]), and MCH gene expression levels here are just below those of oxytocin and vasopressin ([6]). In fact, hypothalamic MCH peptide levels are amongst the highest in the brain ([29]). Attention has recently been focused on the involvement of MCH in feeding and energy balance. The presence of SLC‐1 in the arcuate, ventromedial, dorsomedial and paraventricular nuclei indicates that the receptor could mediate the reported orexigenic effects of MCH ([32]; [33]; [26]). Interestingly, SLC‐1 immunoreactivity was present within the medial hypothalamus, a region where there is preproMCH mRNA and peptide expression in lactating rats ([22]). This reinforces the possible role of SLC‐1 in feeding behaviour and energy balance. In addition, the immunostaining seen in the paraventricular, supraoptic, arcuate and medial preoptic nuclei could be associated with the effect of MCH on the stress response ([18]; [7]; [26]). It may also be associated with oxytocin release ([30]), LH release ([12]), anxiety and sexual behaviour ([13]). In conclusion, most of the major roles of the hypothalamus in orchestrating visceral and survival activities and controlling homeostasis appear to depend to some extent on MCH and SLC‐1 action.

The nucleus accumbens, which was strongly SLC‐1 immunoreactive, is a major recipient of the mesolimbic dopaminergic projection from the ventral tegmental area (also SLC‐1 immunopositive) and serves a key role in reward mechanisms. This brain region may mediate positive reinforcing effects of food and thus provide an additional control by MCH on feeding behaviour. Interestingly, melanotropins are known to be implicated in drug‐seeking behaviours (see [10]).

This study describes in detail the distribution of SLC‐1 mRNA and immunoreactivity in the mature rat central nervous system. There was good anatomical agreement between these two sets of data, and the receptor distribution we report here is also in good registration with the distribution of the ligand, MCH, as reported in the literature. SLC‐1 mRNA and protein were found in a number of functionally distinct regions of the CNS, and the distributions of the receptor largely support all of the biological actions of MCH that have been reported in mammalian systems. The phylogenetic distribution of MCH has provided an interesting problem for biologists who wish to reconcile the melanophore modulation of this peptide in lower vertebrates with hitherto unknown functions in the mammalian brain. The search for a mammalian function for MCH has been hampered until now by the lack of knowledge about the receptor(s) for this peptide. The present study provides a solid basis for further characterization of the role of MCH in mammals.

We would like to thank Martyn Evans for his help with imaging technologies, Karen Davies for help with raising the antisera, Trevor Dowe for technical assistance, Amanda Perry and David Howlett for the estimation of the antisera titres by ELISA, Susan Konchar and Subinay Ganguly for providing the SLC‐1‐transfected HEK 293 cell line, Alison Muir for cell culture, and Shelagh Wilson for providing us with Fig. 4(E), and David Michalovitch for bioinformatics analysis. We are grateful to Christopher Benham and Frank Walsh for careful reading of the manuscript and helpful comments.

• Abbreviations
• ACA anterior cingulate area
• act anterior corticothalamic tract
• AD anterodorsal nucleus thalamus
• AHN (p) anterior hypothalamic nucleus (posterior part)
• Al(d)(v) agranular insular area (dorsal)(ventral) part
• AM anteromedial nucleus thalamus, AN, arcuate nucleus
• AON anterior olfactory nucleus
• APN anterior pretectal nucleus
• AUDp primary auditory area
• AUDv ventral auditory area
• AV anteroventral nucleus thalamus
• BLA basolateral nucleus amygdala
• BMA basomedial basolateral nucleus amygdala
• BSA bovine serum albumin
• BST bed nuclei of the stria terminalis
• Cb(gr) cerebellum (granular cell layer of the cerebellar cortex)
• CEAl central nucleus amygdala
• CL central lateral nucleus
• CLA claustrum
• CM central medial nucleus thalamus
• COA cortical nucleus amygdala
• CSm superior central nucleus raphe, medial part
• DAB 3,3′‐diaminobenzidine
• DCO dorsal cochlear nucleus
• DR dorsal nucleus raphe
• DRG dorsal root ganglion
• DTN dorsal tegmental nuclei
• DTT dithiothreitol
• ECT ectorhinal area
• EDTA ethylene diamine tetraacetate
• ENT entorhinal area
• EP endopiriform nucleus
• GAPDH gylceraldehyde‐3‐phosphate dehydrogenase
• GP globus pallidus
• GRN gigantocellular reticular nucleus
• GU gustatory area
• HDB nucleus horizontal limb diagonal band
• HEK human embryonic kidney
• hi hippocampus
• IA(D)(M) interantero(dorsal) (medial) nucleus thalamus
• IC inferior colliculus
• ICC immunocytochemistry
• IHC immunohistochemistry
• Int(A)(P) interpositus cerebellar nucleus (anterior) (posterior) subdivisions
• IO inferior olivary complex
• IP IPN, interpeduncular nucleus central subnucleus
• ISH in situ hybridization
• islm major islands of Calleja
• KF Kolliker–Fuse subnucleus
• KLH keyhole limpet haemocyanin
• LA lateral nucleus amygdala
• LC locus coeruleus
• lct laterocorticospinal tract
• LD latero‐dorsal thalamus
• LGv (m) (l) lateral geniculate complex, ventral (medial) (lateral) part
• LHA lateral hypothalamic area
• lsc laterospinocortical tract
• LS(d)(v) lateral septum (dorsal) (ventral) segments
• lst lateral spinothalamic tract
• MA magnocellular preoptic nucleus
• MARN magnocellular reticular nucleus
• MCH melanin‐concentrating hormone
• MDRNv medullary reticular nucleus, ventral part
• MEA median nucleus amygdala
• MEM minimum essential medium
• MEPO median preoptic nucleus
• MG(v) medial geniculate complex (ventral element)
• MH medial habenula
• MM medial mammillary nucleus
• MO(p)(s) (primary)(secondary) motor area
• MPN medial preoptic nucleus
• MPO medial preoptic area
• MS septal nucleus
• MV(m)(v) medial vestibular nucleus (magnocellular)(parvicellular) parts
• NDB nucleus of the diagonal band
• NGS normal goat serum
• NLL nucleus of lateral lemniscus
• NTS(I)(rm)(vl) nucleus of the solitary tract (intermediate) (rostral zone of the rostral part) (ventrolateral) parts
• OP olivary pretectal nucleus
• OT olfactory tubercle
• PA posterior nucleus amygdala
• PAG periaqueductal grey matter
• PAR parasubiculum
• PB(l)(mm) parabranchial nucleus (lateral)(medio‐medial) part
• PBS phosphate‐buffered saline
• PCG pontine central grey
• PF parafascicular nucleus thalamus
• PG pontine grey
• PH posterior hypothalamus nucleus
• PIR piriform cortex
• PO posterior complex thalamus
• POST postsubiculum
• PRE presubiculum
• PRN(c)(r) pontine reticular nucleus (caudal)(rostral) part
• PSV principle sensory trigeminal nerve
• PT parathenial thalamus nucleus
• PTLp posterior‐parietal region association area
• PV(i)(p) periventricular nucleus hypothalamus (intermediate)(posterior) part
• PVH(dp)(mpv)(pml)(pmd)(pv) paraventricular nucleus hypothalamus (dorsal parvicellular)(medio‐parvicellular) (posterior magnocellular‐lateral zone)(medial parvicellular‐dorsal zone) (periventricular zone)
• PVT periventricular nucleus thalamus
• RE nucleus reuniens
• RN red nucleus
• RPA nucleus raphe pallidus
• RPE Red‐Phycoerythrin
• RSP(d) (v) retrosplenial cortex (dorsal)(ventral) part
• RT reticular nucleus thalamus
• RT‐PCR reverse transcription followed by polymerase chain reaction
• SC (zo) (op) (sg) (ig) (dg) superior colliculus (zonal) (optic) (superficial grey) (intermediate grey) (deep grey) layer
• SDS–PAGE sodium dodecyl sulphate–polyacrylamide gel electrophoresis
• SG supragenual nucleus
• SI substantia innominata
• SN p(c) (r) substantia nigra pars (compacta) (reticulata)
• SO supraoptic nucleus
• SOC (l) superior olivary complex (lateral part)
• sp SP, pyramidal layer
• SPV spinal tract of the trigeminal nerve
• SS(p) (s) (primary) (secondary) somatosensory area
• STN subthalamic nucleus
• SUB(d)(v) subiculum (dorsal) (Bd) (ventral) parts
• (T)TBS (Tween 20) Tris‐buffered saline
• Tev ventral temporal association area
• TMv tuberomammillary nucleus
• TR post‐piriform transition area
• TRN tegmental reticular nucleus
• TT(d)(v) taenia tecta (dorsal)(ventral)
• VAL ventral anterior‐lateral complex thalamus
• VCO ventral cochlear nucleus
• VDB vertical limb of the diagonal band
• VIS(al)(am)(li)(ll)(p)(pl)(pm) visual area (anterolateral) (anteromedial) (intermediolateral) (laterolateral) (primary) (posterolateral) (posteriomedial)
• VISC visceral area
• VM ventral medial nucleus thalamus
• VMH ventromedial area hypothalamus
• VP(L)(M) ventral postero (lateral) (medial) nucleus thalamus
• VTA ventral tegmental area
• VTN ventral tegmental nucleus
• ZI(da) zona incerta (dopaminergic cell group of the)