Expression and localisation of somatostatin receptor subtypes sst1–sst5 in areas of the rat medulla oblongata involved in autonomic regulation

Journal of Chemical Neuroanatomy 35 (2008) 49–66 www.elsevier.com/locate/jchemneu

Expression and localisation of somatostatin receptor subtypes sst1–sst5 in areas of the rat medulla oblongata involved in autonomic regulation Emma J. Spary *, Azhar Maqbool, Trevor F.C. Batten Academic Unit of Cardiovascular Medicine, Worsley Building, University of Leeds, Leeds LS2 9JT, UK Received 11 May 2007; received in revised form 18 June 2007; accepted 20 June 2007 Available online 27 June 2007

Abstract Somatostatin is known to modulate the activity of neurones of the medulla oblongata involved in autonomic regulation, mediated through five subtypes of G protein-coupled receptors, sst1–sst5. This study utilises reverse transcription polymerase chain reaction and immunohistochemistry to investigate the expression of sst1–sst5, including the sst2A/sst2B isoforms, in the main autonomic centres of the rat medulla oblongata: nucleus of the solitary tract (NTS), dorsal motor vagal nucleus (DVN) and ventrolateral medulla (VLM). In tissue from the cerebral cortex, hippocampus and cerebellum all subtype mRNAs were detected, but sst5 signals were weak, and the distribution of sst1–sst5 immunoreactivities was consistent with previous reports. In the medulla, all sst mRNAs gave clear amplicons and subtype-specific antibodies produced characteristic patterns of immunolabelling, frequently in areas of somatostatinergic innervation. Anti-sst1 labelled beaded fibres, sst2A, sst2B, sst4 and sst5 gave somatodendritic labelling and sst3 labelled presumptive neuronal cilia. In NTS tissue, sst1, sst2A, sst4 and sst5 mRNAs were strongly expressed, while in VLM tissue sst1, sst2A, sst2B and sst4 predominated. In both areas of the medulla, neurones with intense somatodendritic sst2A immunoreactivity were principally catecholaminergic in phenotype, being double labelled for tyrosine hydroxylase (TH) and phenylethanolamineN-methyl-transferase (PNMT). Some TH/PNMT positive neurones were also sst2B and sst4 immunoreactive. Cholinergic parasympathetic neurones in the DVN were immunoreactive for the sst2A, sst2B, sst4 and sst5 subtypes. These observations are consistent with the proposal that multiple somatostatin receptor subtypes, possibly combining as heterodimers, are involved in mediating the modulatory effects of somatostatin on autonomic function, including cardiovascular, respiratory and gastrointestinal reflex activity. # 2007 Elsevier B.V. All rights reserved. Keywords: Sst subtype; Polymerase chain reaction; Nucleus of the solitary tract; Ventrolateral medulla; Catecholamine

1. Introduction Somatostatin (SOM) is widely distributed throughout the CNS and is an important neurotransmitter and modulator of neural activity (Patel, 1999), mainly due to its inhibitory effects on neuronal excitability and by its ability to modify the release of neurotransmitters (Meyer et al., 1989; Tallent and Siggins, 1999). The neuropeptide exists in two biological active forms, SOM-14 and the amino terminally extended form SOM-28 (Brazeau et al., 1973; Pradayrol et al., 1980). Both forms have diverse effects mediated via interaction with seven transmembrane spanning G-protein-coupled receptors (GPCRs) and

* Corresponding author. Tel.: +44 113 343 4172; fax: +44 113 343 4803. E-mail address: [email protected] (E.J. Spary). 0891-0618/$ – see front matter # 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.jchemneu.2007.06.002

inhibition of Ca2+ currents or augmentation of K+ conductance (Jacquin et al., 1988; Dryer et al., 1991; Selmer et al., 2000). To date five receptor genes encoding distinct somatostatin receptor (sst) subtypes have been identified, termed sst1–sst5 (Bruno et al., 1992; Kluxen et al., 1992; Li et al., 1992; Meyerhof et al., 1992; O’Carroll et al., 1992). The genes for sst1, sst3, sst4 and sst5 are not interrupted by introns in their protein coding regions; however, the rat sst2 subtype can undergo alternate mRNA splicing at its 30 end generating two separate isoforms, a long variant 2A and a short variant 2B. These splice variants are identical except for the differences in their carboxyl terminal sequences (Vanetti et al., 1992). Somatostatin receptor subtypes have been shown to form both functional hetero- and homodimers, with dimerisation modifying their functional properties (Rocheville et al., 2000a,b). The sst subtypes can also form heterodimers with

50

E.J. Spary et al. / Journal of Chemical Neuroanatomy 35 (2008) 49–66

other GPCRs, as demonstrated by the oligomerisation of sst5 with the human dopamine receptor 2 in Chinese hamster ovary cells (Rocheville et al., 2000a). In addition, there is also evidence to suggest that certain receptor subtypes can modulate the actions of other subtypes. For example sst5 has been shown to modulate the trafficking and cell surface regulation of sst2A when expressed in the same cells (Sharif et al., 2007). The distribution of somatostatin receptor binding and somatostatin-immunoreactive (IR) fibres in the brainstem suggests that the neuropeptide may be involved in a number of central regulatory functions. These include the processing of somatosensory, proprioceptive and nociceptive information, arousal and the sleep-waking cycle and autonomic functions, including cardiovascular, respiratory and gastric reflexes (Carpentier et al., 1997). Within the medulla oblongata, transient blood pressure changes are controlled by the sympathetic reflex pathway, which is subserved by a series of three nuclei: the nucleus of the solitary tract (NTS), the caudal, and the rostral ventrolateral medulla (VLM) (Dampney, 1994; Aicher et al., 2000). Studies on the NTS, the primary region for the co-ordination and integration of sensory afferent inputs derived from the cardiovascular, respiratory and gastrointestinal systems (Loewy, 1990; Van Giersbergen et al., 1992; Spyer, 1994), have shown that a substantial proportion of the GABAergic terminals contain large dense core vesicles (Maqbool et al., 1991). This suggests that they may be capable of releasing additional peptide transmitters involved in modulation of autonomic reflexes. The co-existence of somatostatin with GABA has been reported in many areas of the brain (Hendry et al., 1984; Somogyi et al., 1984). It is frequently co-released with GABA from hippocampal neurones and axonal terminals of the basolateral or central nucleus of the amygdala (Vezzani and Hoyer, 1999; McDonald and Mascagni, 2002; Saha et al., 2002). Thus, it is possible that GABA immunoreactive terminals in the NTS involved in cardiovascular regulation may also contain and release somatostatin. Indeed, microinjections of somatostatin in the NTS of anaesthetised rats have been shown to have modulatory actions on cardiovascular reflexes, producing hypotension and bradycardia (Koda et al., 1985). This raises the question of which somatostatin receptor subtypes are present in the medullary nuclei and whether the activity of the NTS and VLM neurones involved in cardiovascular regulation may be modulated by somatostatin, released from GABAergic or other axon terminals, acting upon these receptor subtypes. Previous investigations on the distribution of the sst subtypes, using either immunohistochemistry or in situ hybridisation have not examined the medulla oblongata in any detail. In the present study, we analysed the expression of sst subtype mRNAs in the medulla oblongata of the adult rat using reverse transcription polymerase chain reaction (RT-PCR) and examined the distribution of sst receptor proteins using immunohistochemistry with well characterised, subtype-specific antibodies. Sst subtype expression in the cortex, hippocampus and cerebellum

was also studied as a positive control or baseline comparison for our analyses of the medulla oblongata.

2. Materials and methods 2.1. Tissue isolation Adult, 10–12-week-old male Wistar rats (160–220 g, n = 6) were killed by decapitation under anaesthesia (5% halothane in O2) in accordance with the regulations of the UK Animals (Scientific Procedures) Act, 1986. Brains were removed and rapidly frozen on dry ice. Coronal slices of 0.5–1 mm thickness were cut from the brainstem, and tissue punches collected from the NTS and the rostral area of the VLM (RVLM) with a 0.69 mm corer under a 5 dissecting microscope. Placement of the NTS punches was confirmed as previously described (Saha et al., 2004). Tissue samples for analysis were also dissected from the cerebral cortex, hippocampus, cerebellum and the remaining area of the medulla oblongata.

2.2. RNA extraction and reverse transcription polymerase chain reaction Total RNA was isolated using the SV total RNA extraction system (Promega, Southampton, UK). Reverse transcription was initiated by adding 1 mg of RNA to 1 ml oligo (dT) primer (500 mg/ml) and heating at 70 8C for 5 min. Then 50 mM Tris–HCl pH 8.3, 1 mM deoxynucleotide triphosphates, RNase inhibitor (2U) and M-MLV reverse transcriptase (200U) were added to the mixture and the reaction incubated at 37 8C for 1 h and terminated by heating at 95 8C for 5 min. Polymerase chain reaction (PCR) (50 mM Tris–HCl, pH 8.5, 1.5 mM MgCl2, 200 mM dNTPs, 1.5 U Taq DNA polymerase) was performed in a 25 ml reaction using 2 ml of first strand product as a template, in a Perkin-Elmer GeneAmp 9700 (Applied Biosystems), with 0.4 mM subunit-specific primers (Table 1). Amplification was initiated by a 5 min pre-incubation at 95 8C, followed by 35 cycles of 95 8C (30 s), 60 8C (30 s) and 72 8C (1 min). A final extension step was performed at 72 8C for 7 min. Negative controls included amplification of RNA and water. Aliquots were separated by electrophoresis on 2% agarose gels containing ethidium bromide and visualised under UV light. The veracity of the PCR products was confirmed by DNA sequencing on an ABI 3100 genetic analyser using BigDye terminator cycle sequencing version 3.1.

2.3. Perfusion fixation Adult Wistar rats (150–200 g, n = 6) were perfused transcardially with 4% paraformaldehyde and 0.2% picric acid in 0.1 M phosphate buffer as previously described (Saha et al., 2001). Brains were post-fixed for 2 h in 4% paraforTable 1 Forward (For) and reverse (Rev) primers used for detection of sst receptor subtypes by PCR Subtype

Primer sequence 50 –30

Amplicon size (bp)

Sst1

For—GCTGTCACACACAAGTCACA Rev—TTCAACAGTGCATTCGACCA

519

Sst2A/sst2B

For—GGTGACCCGAATGGTATCCA Rev—TGCCGGGTAGCTGCTTTCCA

619 (A) 305 (B)

Sst3

For—AGCAGCAACGGCCTTGCACA Rev—GTGGCTGAGGCCACAGAGCA

669

Sst4

For—GTCTCCTGGAAACAACTGGA Rev—CCCTATGCTACCACACAGCA

569

Sst5

For—CCCTCTCTCTGGCCTCCACA Rev—CGCGCTGGCCATCTTGGCTA

469

Primers are written 50 –30 with their predicted size.

E.J. Spary et al. / Journal of Chemical Neuroanatomy 35 (2008) 49–66

51

maldehyde then stored in phosphate buffered saline (PBS). Coronal 40 mm sections of the forebrain (bregma +0.2 mm to bregma 0.4 mm; Paxinos and Watson, 1986), cerebellum-pons and medulla oblongata (bregma 12.5 mm to 14.5 mm) were cut on a vibrating microtome (VT 1000S, Leica Microsystems, Milton Keynes, UK), and collected in phosphate buffered saline pH 7.6.

3. Results

2.4. Primary antisera

The presence of sst receptor subtypes mRNA (n = 6) was investigated by PCR, using individually designed primers. All reactions were performed in triplicate. To discriminate between the sst2A/sst2B isoforms, primers were designed to span the approximate 300 bp deletion splice site in areas of the mRNA sequence common to both splice variants. This approach generated a product of 619 bp for sst2A and a smaller product of 305 bp corresponding to the alternately spliced 2B isoform. Simultaneous control reactions were performed using cDNA from the cerebral cortex, hippocampus and cerebellum as positive controls, as the presence of all the subtypes in these areas is well established (Fehlmann et al., 2000; Schulz et al., 2000). Negative control reactions were also performed using water in place of cDNA and RNA that had not undergone reverse transcription. No products were detected for any of these negative controls. No variation in the expression of individual subtypes was observed between triplicates from different runs or between the cDNA from the same brain area in different animals. RT-PCR analysis of cDNA from the cortex, hippocampus and cerebellum control tissues revealed the presence of all the sst receptor subtypes, denoted by a single distinct band for each corresponding to the predicted size (Fig. 1A–C): sst1 (519 bp), sst2A (619 bp), sst2B (305 bp), sst3 (669 bp), sst4 (569 bp) and sst5 (469 bp). The relative expression of sst5 in the cortex and cerebellum was low, and absent in the hippocampus. The identity of each PCR product was verified by DNA sequencing.

Antibodies to sst1 (#5001), sst2B (#5601), sst3 (#7986) and sst4 (#6002) were gifts of Dr. S Schulz (Magdeburg, Germany) and antibodies to sst1 (RB-1556-P; Neomarkers-LabVision, Newmarket, Suffolk, UK), sst4 (ASR-004; Alamone Labs, Jerusalem, Israel), sst2A, sst4, sst5 (SS-800, SS-835, SS-838; Gramsch Laboratories, Germany) obtained from commercial suppliers. All antibodies were raised in rabbit against synthetic peptides corresponding to the C-terminal peptide sequences, except for sst4 (ASR-004), which was raised against a sequence of the 2nd extracellular loop. Antibody specificity has been characterised by immunoblot and Western blot analysis in previous studies (Schulz et al., 1998, 2000) or in the suppliers’ data, with major bands revealed in rat brain lysates of the following approximate sizes: sst1, 60 kDa; sst2A, 80 kDa; sst2B, 75 kDa; sst3, 75 kDa; sst4, 45 kDa; sst5, 50 kDa. Working dilutions and concentrations of the sst antibodies were: sst1 #5001, 1/1500 (0.05 mg/ml); sst1 RB-1556, 1/1500 (0.6 mg/ml); sst2A SS-800, 1/2000 (0.1 mg/ml); sst2B #5601, 1/500 (0.07 mg/ml); sst3 #7986, 1/2000 (0.14 mg/ml); sst4 #6002, 1/3000 (2.0 mg/ml); sst4 SS-835, 1/500 (0.5 mg/ml); sst4 ASR-004, 1/500 (1.5 mg/ml) and sst5 SS-838, 1/1500 (1.2 mg/ml). All dilutions were made in PBS containing 0.1% Triton-X100 (PBST). Antibody specificity was further assessed by pre-incubating the primary antibody at its working dilution with 10 mg/ml of the immunising peptide (obtained from Gramsch, LabVision or Santa Cruz) for 4–6 h prior to use. In all such cases, no specific immunolabelling was observed.

2.5. Immunohistochemistry Free floating sections were preincubated in 10% normal horse serum diluted in PBS for 30 min and then incubated in primary antisera at 4 8C for 16 h. After washing twice with PBST the sections were incubated in biotinylated anti-rabbit IgG (1/500; Jackson ImmunoResearch-Stratech Scientific, Newmarket, Suffolk, UK) for 2 h. The reaction was visualised with the biotin-avidin-peroxidase method (Vectastain Elite ABC Kit and SG substrate, Vector Labs, Peterborough, UK) for bright field microscopy, or with streptavidin-Alexa488 (1/1000; Molecular Probes, Leiden, Netherlands) for fluorescence microscopy. Double and triple fluorescence co-localisation experiments were performed by simultaneously incubating sections with the sst primary antisera (all raised in rabbit) and antibodies to the following markers of neural phenotype: neuronal nitric oxide synthase (NOS) raised in sheep (1/5000; gift of P. Emson), choline acetyltransferase (ChAT) raised in goat (1/1500; Chemicon, Hampshire, UK), tyrosine hydroxylase (TH) raised in mouse (1/2000; ImmunoStar Inc, Wisconsin, USA), phenylethanolamine N-methyl transferase (PNMT) raised in sheep (1/ 3000; Chemicon), glial fibrillary acidic protein (GFAP) raised in mouse (1/600; Chemicon) and somatostatin raised in rat (clone YC7; 1/800; Chemicon). The bound sst antibodies were visualised with biotinylated anti-rabbit and streptavidin-Alexa488 as above, while double labelling for other neural markers was visualised by a single-step incubation in a species-specific Cy3-conjugated secondary antibody (1/1000; Jackson). Some sections were further labelled for NOS, ChAT or PNMT using a species-specific Cy5-conjugated anti-sheep IgG (1/1000; Jackson) that was fully cross-reactive with goat IgG. The specificity of all secondary antibodies was established by the absence of labelling on sections incubated with normal rabbit serum in place of primary antibody, and the species specificity of all fluorescent secondary antibodies used in double or triple labelling was confirmed by tests on sections incubated with primary antibodies raised in an inappropriate species. Sections were examined on a Zeiss Imager Z1 microscope fitted with appropriate filter sets and an Apotome ‘pseudo-confocal’ system and digital images were captured and processed with Axiovision imaging system (Carl Zeiss, Welwyn Garden City, UK). Brain areas were defined according to the Paxinos and Watson (1986) atlas, while classification of catecholamine cell groups was according to Ho¨kfelt et al. (1984) and delineation of the subnuclei of the NTS was based on the scheme of Van Giersbergen et al. (1992).

3.1. Sst subtype expression in the cortex, hippocampus and cerebellum

3.2. Sst subtypes are expressed in medullary cardiovascular nuclei Amplification of the sst subtypes in mRNA extracted from the whole medulla oblongata showed clear expression of all sst subtypes (Fig. 1D). PCR of the NTS cDNA revealed amplicons for sst1, sst2A, sst4 and sst5, with a relatively weaker amplicon detected for sst2B but no amplification product for sst3 (Fig. 1E). In the VLM, amplicons for subtypes sst1, sst2A, sst2B and sst4 were detected, but no signals for sst3 or sst5 were observed (Fig. 1F). Results of the PCR analysis are summarised in Table 2. Table 2 Summary of the RT-PCR results Subtype

Sst1

Sst2A

Sst2B

Sst3

Sst4

Sst5

Cortex Cerebellum Hippocampus Medulla NTS VLM

+ + + + + +

+ + + + + +

+ + + + + +

+ + + +  

+ + + + + +

+ +  + + 

The presence of a positive amplicon for each sst receptor subtype is denoted by +, any tissues not expressing sst receptor subtypes are denoted by . Reactions (n = 6) were performed in triplicate.

52

E.J. Spary et al. / Journal of Chemical Neuroanatomy 35 (2008) 49–66

Fig. 1. Expression of the sst receptor subtypes in the cortex (A), cerebellum (B), hippocampus (C), medulla (D), NTS (E) and VLM (F). Amplicons representing sst1 (519 bp), sst2A (619 bp), sst2B (303 bp), sst3 (669 bp), sst4 (569 bp) and sst5 (469 bp) were detected in the cortex, cerebellum and medulla. Note that in the hippocampus (C) there is no signal detected in the sst5 lane. The NTS showed the presence of all the subtypes with the exception of sst3, and in the VLM no amplicons for sst3 or sst5 could be detected. The weaker bands on the gels are indicated by arrowheads.

3.3. Localisation of sst subtypes in cortex, hippocampus and cerebellum Somatostatin receptor subtype labelling was investigated using the specific antibodies in the cerebral cortex, hippocampus and cerebellum, with sections processed concurrently under standardised conditions. The patterns of labelling observed throughout these brain areas in most cases closely resemble those already described (Schulz et al., 2000), and so will only be summarised briefly here. The distribution and characteristics of labelling observed with the two different sst1 antibodies and three different sst4 antibodies used were identical, and so will not be described separately. Consistent patterns of labelling were observed in all areas of the neocortex examined, but with variations between the layers (Table 3). Although sst1 immunoreactivity was found in beaded fibres and puncta in many areas of the forebrain, such as the ventral telencephalon and hypothalamus, these were extremely scarce in the cortex (Fig. 2A). The staining pattern for sst2A presented as widespread labelling of neuronal cell bodies and dendrites, decorated with intense immunoreactivity along the cell membranes, most prominently in the deeper layers of the cortex (Fig. 2B). While sst2B was localised to broadly the same areas, the antibody produced lighter, more homogeneous somatodendritic labelling of neurones with some labelling of processes that could originate from glial cells (Fig. 2C). The sst3 antibody produced very characteristic punctate and rod-

like labelling, with occasional light labelling of neuronal perikarya and possibly glial cells (Fig. 2D). Somatodendritic sst4 immunoreactivity was widespread throughout the cortex and was particularly intense in pyramidal cells of layers III and V (Fig. 2E), while the sst5 antibody produced a variable labelling pattern of weakly immunoreactive cells scattered throughout the cortex, most frequently in layers IV and V (Fig. 2F). Sst1 immunoreactivity was not found in any area of the hippocampus. Labelling for sst2A and sst2B exhibited generally similar patterns in the pyramidal cells and the dentate gyrus granular cells (Table 3). Small rod-like sst3-IR profiles were very prominent in the pyramidal cell layer and stratum lucidum of CA3; however, in the dentate gyrus clearly labelled interneurones were also observed (Fig. 2G). The sst4 antibodies strongly labelled the somata and dendrites of pyramidal cells, particularly those in CA3 (Fig. 2H), while sst5-IR cells were sparsely distributed in the stratum oriens and pyramidal and granule cell layers (Table 3). Weak staining for sst1 was confined to the granular layer and the Purkinje cells in certain parts of the cerebellum (Fig. 3A). The Purkinje cells were unstained for the sst2A isoform, with patches of sst2A immunoreactivity being restricted to the neuropil and Bergmann glia of the Purkinje layer (Fig. 3B). In contrast, the Purkinje cells were labelled intensely for the sst2B isoform, with diffuse immunoreactivity also present in the neuropil of the granule layer (Fig. 3C). Rod-like profiles and

E.J. Spary et al. / Journal of Chemical Neuroanatomy 35 (2008) 49–66

53

Table 3 Distribution of sst receptor immunoreactivities in the cerebral cortex, hippocampus and cerebellum Brain area a

Receptor subtypeb Sst1

Sst2A

Sst2B

Sst3

Sst4

Sst5

Neocortex Layer I Layer II Layer III Layer IV Layer V Layer VI

– – – – – F+

SN++ SD+ SD++ SD++ SD+++ SDN+++

SD+ SD+++ SD+++ SD++ SD+++ SD++

– R+ R+ R+ R++ R++

D++ SD++ SD++ D+ SD+++ SD++

– SD+ SD+ SD++ SD++ –

Hippocampus DG granule cell layer DG molecular layer DG polymorph/hilus Pyramidal layer (CA1) Pyramidal layer (CA2) Pyramidal layer (CA3) Stratum oriens Stratum radiatum Stratum lucidum (CA3) Stratum lac-mol (CA1)

– – – – – – – – – –

SD+ N+ SD++ SN+ SN++ SDN+++ N++ NF++ SD++ NF+++

SD+ N+ SD++ SD++ SD++ SD++ N+ – N++ N+

R+++ – SR++ R+ R+ R++ – – R+ –

SD+++ N+ SD++ SD++ SD++ SD++++ ND+ NP++ DP++++ N+

SD++ N+ SD+ SD++ SD+++ SD++ SD++ SD+ SD+ N+

Cerebellum Molecular layer Purkinje layer Granule cell layer Deep cerebellar nuclei

– S+ N+ –

N++ SN++ N+ SN+

ND+ S+++ N++ SN+

R+++ R++ R+ R+

D++++ SD+++ SN+ SD+++

– S+ SN++ SD+++

c

a

Abbreviations: DG, dentate gyrus; lac-mol, lacunosum moleculare. Labelling intensity assessed on a 5 point scale of – (not detectable) to ++++ (very intense); key to structures labelled: S = somata, D = dendritic processes; P = puncta, F = fibres, R = rod-like profiles, V, varicosities, N = diffuse neuropil labelling. c Based on examination of the following cortical areas: somatomotor, somatosensory, parietal, visceral, temporal, insular, auditory. b

puncta representing sst3 immunoreactivity were very evident in the molecular layer of the cerebellum, and may be attributed to the labelling of cilia of the apical dendrites of the Purkinje cells (Fig. 3D). Moderately sst4-IR Purkinje cells were scattered throughout the cerebellum with their heavily labelled dendrites extending into the molecular layer (Fig. 3E). A small number of weakly sst5-IR Purkinje cells were present in the cerebellar lobules, as well as some labelled cells in the fastigial nucleus (Fig. 3F).

paratrigeminal region. Some long beaded axons were scattered throughout the reticular formation. These were most numerous in the dorsal, lateral and intermediate reticular areas and along the ventral border of the RVLM. Although these sst1-IR axons were present in areas containing somatostatin-IR axons, double fluorescence labelling did not reveal any colocalisation (Fig. 5A–F); neither did the presumptive boutons of sst1-IR axons appear to be associated with any neurochemically defined neuronal group (Fig. 5G–L).

3.4. Sst subtype immunoreactivity in the medulla oblongata

3.4.2. Sst2A Sst2A immunoreactivity in the medulla (Table 4) was characteristically seen as intense labelling along the plasmalemma of the somata and proximal dendrites of many neurones, together with finer dendritic labelling throughout the neuropil (Fig. 4B). Large sst2A-positive cells were seen in both the NTS and DVN, and the area postrema (AP) displayed a very strongly sst2A-IR neuropil obscuring small immunoreactive cells. Within the NTS, sst2A positive cells were mostly grouped in the midline com, dorsomedial (dm) and dorsolateral (dl) subnuclei, along the ventricular surface of the NTS more rostrally, and extending laterally from the ventrolateral NTS into the dorsal reticular nucleus (DRt, Fig. 7R and S). Multiple labelling showed that nearly all of these strongly immunoreactive cells were also TH-IR (Fig. 6A and B), but only a proportion of TH positive cells of the A2/C2 catecholamine group were sst2A-IR, and this proportion

3.4.1. Sst1 Sst1 immunoreactivity in the medulla was present only in beaded axons and varicosities (Figs. 4A and 5A, D, G; Table 4). These were numerous in the DVN and NTS, especially in the commissural subnucleus (com) and around the tractus solitarius (ts) at levels caudal to the obex (Fig. 4A), and had a mainly transverse orientation in both nuclei. Few axons or varicosities were seen in the hypoglossal (XII), gracile (Gr), cuneate (Cu) or external cuneate (ECu) nuclei, but a few prominent beaded axons ran along the subependymal layer bordering the central canal and the fourth ventricle overlying the medial vestibular nucleus (MeV) at rostral levels. The raphe´ obscurus (ROb) contained a few scattered fibres and varicosities at all levels, as did the spinal trigeminal nucleus (SpV), but particularly in its more dorsal part, along the spinal tract, and in the

54

E.J. Spary et al. / Journal of Chemical Neuroanatomy 35 (2008) 49–66

Fig. 2. Immunofluorescent localisation of sst subtypes in the cerebral cortex (A–F) and hippocampus (G–H). (A) Labelling for sst1 is very weak or absent in cells throughout the cortex, with a few positive fibres (arrows) in the deeper layers, as illustrated in layer VI of the viscerosensory cortex. (B) Immunoreactivity for sst2A is intense, with a somatodendritic distribution on cell bodies (arrows) and a tangle of processes, at the transition between layers V and VI of the somatosensory cortex. (C) Labelling for the sst2B subtype is less extensive in cell bodies (larger arrows) and processes (smaller arrows) in layer V of the parietal cortex. (D) Sst3 immunoreactivity is present throughout the cortex in many small rod-like profiles of neuronal cilia (arrowheads), with a few cells stained that may represent neurones (larger arrow) or glia (smaller arrow). (E) Labelling for sst4 is moderate-to-strong throughout all areas of the cortex, including the pyramidal cells of layer V (arrows). (F) There is little specific labelling for sst5 in layers IV and V of the cortex, except for a few weak to moderate labelled neuronal somata (arrows). (G) Strongly sst3 immunoreactive neuronal cilia in the granule cell layer (Gr) of the dentate gyrus, with a few labelled interneurones in the polymorphic (Poly) and molecular (Mol) layers. (H) Somatodendritic labelling for sst4 is especially strong in the pyramidal cells (Pyr) of CA3; or, stratum oriens; Luc, stratum lucidum; Rad, stratum radiatum.

increased towards rostral levels and into the C3 group. Some of the TH/sst2A positive cells were also NOS-IR, but very few non-TH-positive NOS-IR cells were sst2A-IR (Fig. 6A–C). Many, but not all the sst2A-IR cells occurred in areas of dense somatostatin-IR innervation, and had cell bodies and dendrites lying in close apposition to somatostatin-IR puncta (Fig. 7A– Q). A similar situation was observed in the VLM, with a high proportion of coexistence of sst2A and TH (NOS) immunoreactivities in the RVLM (Fig. 6D–F) but a lower proportion at more caudal levels of the medulla. The positions of the sst2A-IR cells in the dorsomedial and ventrolateral

medulla, and the presence of PNMT immunoreactivity in many of them (Fig. 8A–D) suggested expression of the receptor by adrenaline containing neurones of the C1 and C2 groups. Other areas of the medulla containing well labelled, relatively large non-TH, non-NOS neurones and processes in the neuropil were the intermediate reticular area (IRt) and nucleus ambiguus (NAmb), where sst2A and ChAT immunoreactivities colocalised in cells both in the external formation (Fig. 5M–O), the dorsal part of the SpV and the ventral strip of the lateral paragigantocellular nucleus in the rostral medulla. The XII, Gr, Cu, ECu and inferior olivary nuclei

E.J. Spary et al. / Journal of Chemical Neuroanatomy 35 (2008) 49–66

55

Fig. 3. Immunoperoxidase visualisation of sst subtypes in the cerebellum. (A) Sst1 labelling is weak in the neuropil of the granule cell layer (gr) and in Purkinje cells (Pk, arrows). (B) The granule cell and molecular (mol) layers are mostly devoid of staining for sst2A, but patches of immunoreactivity (arrows), probably representing Bergmann glia are present in the Purkinje cell layer. (C) The Purkinje cells are moderately stained for the sst2B subtype. (D) Fine puncta immunoreactive for sst3 distributed throughout the cerebellum, are most noticeable in the Purkinje layer and the outermost part of the molecular layer (arrows). (E) Immunoreactivity for sst4 is intense in many Purkinje cell bodies (arrows) and their apical dendrites in the molecular layer. (F) Some patches of sst5 immunoreactivity (arrows) are seen in the granule cell layer and along the membranes of neural somata in the fastigial nucleus (FN). arb, arbour vitae.

(IO) contained few, weakly sst2A-IR cells and a moderately labelled neuropil. 3.4.3. Sst2B Sst2B immunoreactivity in the medulla appeared as labelling of processes and puncta in many nuclear areas, together with many neurones with weak or moderately labelled peripheral cytoplasm (Fig. 4C; Table 4). The most intensely labelled neuropil, with scattered small cells, was seen in the Cu, ECu AP, IO and lateral reticular nucleus (LRt). A weak to moderately labelled neuropil with scattered weak cells was evident in the Gr, SpV and most of the reticular formation; however, the DVN (Fig. 8E), DRt and NAmb region contained many clearly labelled cells and possibly glial processes. Cranial motoneurones, including those of the XII, were surrounded with patches of sst2B staining around the cell periphery (Fig. 8E). In the NTS, a group of weakly, but clearly labelled cells was embedded in a neuropil containing immunoreactive processes surrounding the ts (Fig. 6G). There was little evidence for colocalisation of sst2B

in catecholamine cells of the dorsomedial medulla (Fig. 8E and F), but TH immunoreactivity was seen in subsets of sst2B-IR cells in the VLM. The sst2B-IR cells throughout the VLM were distinct from NOS-IR cells (Fig. 6I and J). 3.4.4. Sst3 The distribution and characteristics of sst3 immunoreactivity showed a very similar pattern at all levels of the medulla examined. The most obvious feature was the presence of intensely fluorescent small puncta and rod-shaped profiles in many nuclear areas (Table 4), often those areas containing a dense somatostatin-IR innervation (Fig. 9A–D). These profiles were sparsely distributed, but clearly seen in the NTS, particularly around the ts, where labelling of the neuropil was absent. Similar profiles were present but appeared less numerous and/or distinct in the dorsal column nuclei, DVN, XII and SpV, where a labelling of processes in the neuropil was sometimes seen. At rostral levels of the NTS and in the dorsal medullary reticular formation, weakly labelled cells were

56

E.J. Spary et al. / Journal of Chemical Neuroanatomy 35 (2008) 49–66

Fig. 4. Immunofluorescent localisation of sst receptor subtypes in coronal sections of the dorsomedial medulla at mid area postrema (AP) level. Note the differential distribution of sst1 (A), sst2A (B), sst2B (C), sst3 (D), sst4 (E) and sst5 (F) in the NTS subnuclei, dorsal vagal nucleus (DVN), AP and hypoglossal motor nucleus (XII). Abbreviations for other areas: cc, central canal; com, commissural subnucleus; Cu, cuneate nucleus; dm, dorsomedial subnucleus; me, medial subnucleus; ts, tractus solitarius.

observed that appeared to be neuronal perikarya rather than astrocytes (Fig. 9C–F). 3.4.5. Sst4 The area of the medulla exhibiting the strongest sst4 immunoreactivity was the DVN, with all the parasympathetic preganglionic neurones appearing to be labelled (Figs. 4E and 10A). In general, many other areas (Gr, Cu, ECu, XII, together with much of the reticular formation) contained only faint labelling of the neuropil and scattered weakly sst4-IR cells (Table 4). The caudal DRt showed clear labelling of cells and processes, as did the more dorsal and superficial parts of the SpV, together with the paratrigeminal nucleus (PaV), and the parvicellular division of the LRt. Neuropil labelling in the AP and IO was strong, with scattered small cells. The NTS contained some scattered cells, which were rather weakly, but clearly immunoreactive and mostly located near the ts (Figs. 4E and 10D). Most of the clearly labelled sst4-IR cells in the NTS were not TH-IR or NOS-IR (Fig. 10A–E), but some sst4-IR cells in the RVLM were TH and PNMT-IR (Figs. 10F, G and 8G, H).

3.4.6. Sst5 The sst5 immunoreactivity in the medulla appeared as light to moderate, often ‘feathery’ labelling of the neuropil and light cellular labelling. In contrast to the other subtypes, the dorsal column nuclei (Gr, Cu, ECu) exhibited the strongest labelling (Fig. 4F), together with the IO and MeV (Table 4). Much of the dorsal parts of the reticular formation, the DVN, the NAmb, the ROb and the SpV contained scattered weakly or moderately labelled cells in a light to moderately immunoreactive neuropil. Much of the NTS was devoid of sst5 immunoreactivity, but a few moderately labelled neurones and processes were observed within the ts, and dorsal and lateral to the ts (Fig. 4F). A few TH-IR neurones in the NTS and VLM were found to express sst5 immunoreactivity, but NOS-IR neurones were not found to be sst5-IR (Fig. 10H–J). 4. Discussion The results presented in this study provide evidence from PCR for the expression of somatostatin receptor subtype

E.J. Spary et al. / Journal of Chemical Neuroanatomy 35 (2008) 49–66

57

Fig. 5. Double fluorescence labelling for sst1 and sst2A relative to other neuronal markers. (A–C) Long beaded sst1-IR fibres (arrows in A) in the ventrolateral NTS, are surrounded by somatostatin-IR varicosities (arrows in B), but the merged image (C) provides little evidence for coexistence. (D–F) Apotome confocal images from the commissural NTS, confirming the lack of coexistence of sst1 and somatostatin in beaded fibres (arrowed) in this region. (G–I) Beaded sst1-IR fibres in the medial NTS (arrows in G) among groups of NOS-IR neurones (arrows in H). The merged image (I) suggests that sst1 fibres may form boutons (arrowheads) in the close vicinity of the NOS positive neurones. (J–L) Similarly, the scattered sst1 fibres (arrows) in the RVLM may form appositions onto the TH positive catecholamine neurones. (M–O) External formation of nucleus ambiguus; sst2A-IR cells (arrows) in M, identified as presumptive parasympathetic neurones by ChAT labelling in N.

mRNAs in the medulla oblongata of the Wistar rat, and this was further confirmed by the immunolocalisation of the receptor subtype proteins using specific antibodies. These results suggest the presence of sst receptor subtypes in several areas of the medulla concerned with autonomic regulation. For some sst subtypes (sst2A, sst2B, sst4) the results of PCR and immunohistochemistry concurred, whereas for other subtypes (sst1, sst3 and sst5) there were some notable discrepancies in certain brain areas that will be discussed below.

4.1. Sst1–sst5 expression in the cortex, hippocampus and cerebellum The cortex, hippocampus and cerebellum were chosen as positive control tissues for the investigation of sst subtype expression, since previous studies using in situ hybridisation and immunohistochemistry have documented the distribution of subtype mRNAs and proteins in these areas of the brain. To our knowledge, there has been no previous investigation of the

58

E.J. Spary et al. / Journal of Chemical Neuroanatomy 35 (2008) 49–66

Table 4 Distribution of sst receptor immunoreactivities in the medulla oblongata Brain area

Gracile nucleus Cuneate nucleus External cuneate nucleus Area postrema NTS (caudal)a NTS (intermediate) NTS (rostral) Dorsal vagal nucleus Nucleus ambiguus Hypoglossal nucleus Paratrigeminal nucleus Medial vestibular nucleus Spinal trigeminal nucleus Raphe´ obscurus Inferior olivary nuclei Lateral reticular nucleus A1 region, caudal VLM A1/C1 transition region C1 region, rostral VLM Dorsal medullary reticular Paramedian reticular Gigantocellular reticular Intermediate reticular Parapyramidal reticular Paragigantocellular reticular

Receptor subtype Sst1

Sst2A

Sst2B

Sst3

Sst4

Sst5

– FV+ FV+ FV++ FV+++ FV++ FV+ FV++ F+ F+ F+ F+ FV+ FV+ – – F++ F++ F+++ FV++ – – F+ F+ F+

SN+ SN+ SN++ SDN++++ SD++++ SD+++ SD++ SD++ SDN++ N+ N+++ SN++ SDN++ N+ SN++ SN++ SDN+++ SDN+++ SDN++++ SDN++ SN+ SN+ SDN++ SDN++ SDN++

SN++ SN+++ SDN+++ SDN++++ SDN++ SDN++ SDN+ S+++ SN+++ SN++ SDN+++ SN++ SN+ SN+ SN+++ SDN+++ SDN+++ SDN+++ SDN+++ SN+++ SN+ SN++ SN++ SN++ SN++

RN+ RN+ RN++ R+ R++ R++ SR++ R+ RN+ RN+ RN+++ RN++ RN++ – N++ R+ R+ R+ R+ SR++ – – R+ – –

SN+ SN+ SN++ SN+++ S+ SD++ S+ S+++ SD++ SN+ SN+++ SN+ SDN++ N+ SN++ SN++ SDN++ SDN++ SDN++ SN++ SN+ SN+ S+ SN+ SN+

SN+++ SN+++ SN+++ – SN+ SN+ SN+ SN++ SN++ SN+ N+ SN+++ SN+++ SN++ SN+++ SN+ SN+ SN+ SN+ SN++ SN+ SN+ SN++ SN+ SN+

See Table 3 legend for explanation of scale of labelling intensity. a Caudal NTS defined as levels between bregma 14.7 and 13.6 mm, intermediate NTS as between bregma 13.6 and 12.8 mm and rostral NTS as between bregma 12.8 and 12.3 mm (Paxinos and Watson, 1986).

relative expression of sst receptor mRNAs in the rat brain by PCR with subtype and splice variant-specific primers. Relative expression levels in the three control brain areas were similar, with strong signals for sst1–sst4; however, sst5 signal was very weak (cortex and cerebellum) or not detected (hippocampus). The absence of an amplicon was determined not to be a ‘falsenegative’ result. All the reactions were run concurrently and positive expression was seen for other receptor subtypes in the same cDNA sample and for sst5 in cDNA samples from other areas. Therefore we are confident that the cDNA concentration and primers were optimal for these reactions. In addition, increasing the amount of starting template also failed to produce a positive signal. The expression levels with RT-PCR for sst2A, sst2B, sst3, sst4 and sst5 were broadly consistent with our immunohistochemical observations, and those of previous studies on the rat and other species. Moderate to strong somatodendritic labelling throughout the cortex and hippocampus, was seen for each of the sst2 isoforms, with slightly different distributions (Dournaud et al., 1996; Schindler et al., 1996; Helboe et al., 1999; Braun et al., 2002) with the clearest difference being observed in the cerebellum where sst2B appears to be expressed by Purkinje cells and sst2A by the Bergmann glia. The staining pattern for sst3 is characterised by small rod-shaped profiles in many brain areas, including the cortex, CA2 and CA3 pyramidal layer, dentate gyrus granule layer and Purkinje and molecular layer of the cerebellum. This pattern is

consistent with this sst subtype being localised to neuronal cilia (Ha¨ndel et al., 1999; Braun et al., 2002). However, in situ hybridisation suggested widespread expression of sst3 in many neurones (Ha¨ndel et al., 1999) and using a different, N-terminal directed antibody recognising the non-glycosylated receptor, labelled soma and processes were revealed (Hervieu and Emson, 1999). The sst4 subtype is strongly expressed with a somatodendritic distribution in neurones throughout the cortex, in the hippocampal pyramidal cells and the cerebellar Purkinje cells (Selmer et al., 2000; Schreff et al., 2000; Braun et al., 2002; Kang et al., 2003). The few previous studies of sst5 reported sparse, weak immunoreactivity in cortical neurones (Stroh et al., 1999) and in the hippocampus (Kang et al., 2003). The cerebellum was not examined in these studies, but weak to moderate levels of sst5 mRNA expression were reported in the granular layer and Purkinje cells, consistent with the results observed in this present study (Fehlmann et al., 2000). The main discrepancy in comparing the results of PCR and immunohistochemistry for the ‘positive control’ brain areas concerns the sst1 subtype. PCR gave a strong signal for this subtype in all three control tissues, but immunoreactivity, although prominent in fibres in other areas of the brain such as the hypothalamus, was generally weak (present study) or absent (Helboe et al., 1998). It is very unlikely that our cortical or hippocampal tissue samples were contaminated with tissue from the hypothalamus or other brain areas, and so the most likely explanation appears to be that the sst1 subtype mRNA is

E.J. Spary et al. / Journal of Chemical Neuroanatomy 35 (2008) 49–66

59

Fig. 6. Immunoreactivity for the sst2A and sst2B isoforms in neurochemically defined cells of the NTS and RVLM. (A–C) Sst2A positive cells in the rostral dorsomedial NTS (arrows in A) are immunolabelled for TH (arrows in B), with one cell (larger arrow in C) also being NOS immunoreactive. (D–F) Many of the sst2A cells in the RVLM (arrows in D) are also labelled for TH (arrows in E), with one cell (larger arrow in F) also being NOS positive. Arrowheads in A, B, D and E indicate sst2A/TH cells that are not NOS-IR; thin arrows in C and F indicate NOS positive neurones that are not sst2A immunoreactive; small arrowheads in E indicate TH-IR cells that are neither sst2A nor NOS-IR. (G and H) A group of sst2B-IR cells in the medial NTS (med) close to the tractus solitarius (ts). Two of these cells are also THIR (arrows in G and H), but other cells (arrowheads) are singled labelled for sst2B or TH. (I and J) A group of sst2B-IR cells in the RVLM (arrowheads in I) that are distinct from NOS positive cells (arrowheads in J).

60

E.J. Spary et al. / Journal of Chemical Neuroanatomy 35 (2008) 49–66

Fig. 7. Apotome confocal imaging of sst2A immunoreactivity in relation to somatostatin-IR varicosities in the NTS (A–Q) and dorsal reticular nucleus (DRt, R and S). (A–O) Multi-channel images at a series of focal planes (3.4 mm to +3.4 mm) around the centre of a sst2A-IR cell body in the dorsomedial NTS. Somatostatin-IR axons appear to make close appositions at various points on the soma (arrows in merged images in the right hand column). (P, Q) A sst2A-IR cell in the rostral medial NTS imaged by conventional fluorescence microscopy (P) and in a multi-channel merged confocal image (Q), showing that this immunoreactive cell (green) does not appear to be contacted by somatostatin-IR varicosities (red). (R and S) Similar pair of conventional (R) and Apotome (S) images from the DRt, located ventrolateral to the NTS, showing a number of sites of apposition (arrows) between somatostatin-IR terminals and sst2A-IR cells or dendrites.

E.J. Spary et al. / Journal of Chemical Neuroanatomy 35 (2008) 49–66

61

Fig. 8. Double labelling for sst subtypes and PNMT in the dorsomedial and ventrolateral medulla. (A and B) A group of sst2A-IR cells in the RVLM (A), and PNMTIR cells in the same area (B). (C and D) Similar double labelling in the medial NTS. Cell bodies that are double labelled are marked by arrows in A–D; those labelled only for sst2A are marked by arrowheads in A and D. (E) Weakly sst2B-IR cells in the rostral DVN (arrowheads), and patches of immunoreactivity around the membrane of motoneurones (arrows) in the hypoglossal nucleus (XII). (F) Same area, showing a large PNMT-IR C3 catecholamine cell, not immunoreactive for sst2B. (G and H) Neurones labelled for sst4 (G) and PNMT (H) in the RVLM. Cells that are double labelled are marked by arrows in G and H; those labelled for sst2A alone are marked by arrowheads in G.

expressed in neuronal perikarya, but the protein is present in a form only poorly recognised by the antiserum, or is rapidly transported and targeted to axons in other areas of the brain. 4.2. Sst1–sst5 expression in the medulla oblongata PCR performed on with cDNA from the whole medulla revealed expression of all the subtypes, but in tissue isolated from the NTS there was predominant expression of the sst1 and sst5 subtypes, with relatively moderate levels of sst2A and sst4, low levels of sst2B and no expression of the sst3 subtype. A previous study using in situ hybridisation (ISH) has shown high

to moderate expression of sst1 and sst2A/sst2B mRNA in the NTS with low levels of sst3 and no sst4 or sst5 expression (Fehlmann et al., 2000). The few previous immunohistochemical studies that examined the NTS reported the presence of moderate diffuse labelling for sst2A (Dournaud et al., 1998), and sparse rod-like labelling for sst3 (Ha¨ndel et al., 1999). Whilst ISH studies failed to detect sst4 mRNA in the NTS, immunohistochemistry revealed a moderate expression in processes in the NTS (Selmer et al., 2000). The localisations of sst1 and sst5 have not been examined, while sst2B immunoreactivity was not reported in the NTS, although it was localised to other medullary areas associated with motor functions (Schindler et al., 1999).

62

E.J. Spary et al. / Journal of Chemical Neuroanatomy 35 (2008) 49–66

Fig. 9. Labelling for the sst3 subtype in the NTS. (A and B) Apotome confocal imaging of sst3 immunoreactivity (A) in relation to somatostatin (SOM)-IR varicosities (B) in the ventral NTS, close to the tractus solitarius (TS). Arrows point to sst3-IR profiles believed to represent cilia and somatostatin-IR varicosities that are in close proximity. (C and D) The most rostral part of the medial NTS, showing a group of sst3-IR neuronal cell bodies (arrows in C), which are located in an area also containing sst3-IR cilia and densely innervated by somatostatin-IR terminals (D). Asterisks in D show positions of cells marked by arrows in C. (E) sst3-IR cilia (arrows) and small cells (arrowheads) in the interstitial NTS, between the fascicles of the TS. (F) The sst3-IR cells shown in E (arrowheads) do not correspond to GFAP-IR astrocytes (arrowed in F). Nuclei of sst3-IR cells are marked by asterisks.

Many sst1 positive beaded axons were prominently distributed throughout the NTS and adjacent DVN. This subtype has been implicated as an autoreceptor, and was previously localised to somatostatinergic axons in the rat hypothalamus (Helboe et al., 1999). However, with double fluorescence labelling we were unable to demonstrate colocalisation of sst1 and somatostatin immunoreactivities in the NTS and surrounding areas of the medulla. One explanation for this may be that the two antigens are localised to different compartments of the same axons, since sst1 appeared to be present in pre-terminal beaded axons, while somatostatin immunoreactivity present in numerous fine terminal varicosities. The sst2A and sst2B subtypes, along with sst4, were localised to the somatodendritic compartments of neurones in the NTS, but with a differential appearance throughout the nucleus. Intensely sst2A positive neurones were present mostly in the commissural NTS and extending in an arc through the lateral and dorsomedial areas, whereas the more weakly labelled sst2B and sst4 positive cells formed small clusters in the medial NTS and near the tractus solitarius. The distribution of sst3 immunoreactivity can be correlated with the weak signal for sst3 mRNA, since the

stained rod-like profiles believed to represent neuronal cilia (Ha¨ndel et al., 1999), despite being clearly visible, were more sparsely scattered than in many other brain areas. Therefore, the absence of a corresponding PCR signal could be attributed to insufficient sensitivity in finding these small immunoreactive structures. The few sst3-IR cell bodies were located in the most rostral NTS, an area not sampled for PCR analysis. Even though the NTS was largely devoid of sst5 immunolabelling, more extensive labelling was present in the adjacent DVN and dorsal column nuclei, and ISH has revealed very high levels of sst5 mRNA in the DVN (Fehlmann et al., 2000). Therefore, the presence of a PCR signal for sst5 in NTS tissue samples may not necessarily indicate expression in neurones of this nucleus, but may be due to the amplification of sst5 mRNA from the adjacent areas, which may be unavoidably isolated along with the NTS tissue punches, or to mRNA expressed in the dendrites of DVN neurones that are known to extend dorsally into the NTS (Rinaman et al., 1989). Besides sst5, many cell bodies of DVN neurones showed moderate or strong labelling for sst2B and sst4, possibly accounting for some of the mRNA signal observed in NTS tissue punches.

E.J. Spary et al. / Journal of Chemical Neuroanatomy 35 (2008) 49–66

63

Fig. 10. Labelling for sst4 (A–G) and sst5 (H–J) in neurochemically identified cells in the dorsomedial and ventrolateral medulla. (A–C) Triple labelling for sst4, TH and NOS in the ventral NTS (vNTS) and dorsal vagal nucleus (DVN). A sst4 positive cell in the vNTS (arrow in A) is also TH-IR (arrow in B) and NOS-IR (arrow in C). Sst4-IR cells in the adjacent DVN (A) are not TH-IR (B), and other NOS-IR cells in the DVN (arrowheads in C) do not correspond to sst4-IR cells shown in A. (D and E) The sst4-IR cells (arrowheads in D) in the medial NTS do not correspond to TH-IR cells (arrowheads in E). (F and G) Colocalisation of sst4-IR (F) and TH-IR (G) in many cells in the caudal VLM (arrows). A proportion of cells are single labelled for sst4 (arrowhead in F) or for TH (arrowhead in G). (H–J) Triple labelling for sst5, TH and NOS in the intermediate reticular formation. A few of the sst5-IR cells (arrow in H) are also TH-IR (arrow in I), but they do not correspond to NOS-IR cells (arrowhead in J). Most of the sst5 and TH positive cells in this area are single labelled (arrowheads in H and I).

In the VLM, the PCR results indicated expression only of sst1, sst2A and sst4. The neurones in this area of the medulla are functionally and phenotypically diverse, and while our tissue punches were aimed at the caudal and rostral areas of the VLM that are implicated in autonomic regulation (Blessing, 1997; Phillips et al., 2001), they would also include the reticular areas such as the nucleus ambiguus and lateral reticular nucleus. There have been no previous immunohistochemical descriptions of sst subtype distribution in these areas, but our observations, largely consistent with the PCR, suggested that sst2A, sst2B and sst4 are the main subtypes expressed by cell bodies and dendrites, with sst1 immunoreactivity present in many beaded fibres.

4.3. Functional significance of sst receptor subtypes in NTS and VLM Co-existence of somatostatin in GABAergic neurones and axon terminals has been reported in many areas of the brain (Hendry et al., 1984; Kubota et al., 1994). Somatostatin has been shown to modulate GABA-mediated inhibitory postsynaptic potentials (IPSPs) in neurones of areas where coexistence is found, such as the hippocampus (Pittman and Siggins, 1981; Scharfman and Schwartzkroin, 1989) and in areas such as the septum where coexistence does not occur (Twery and Gallagher, 1990). A presynaptic action of somatostatin also is indicated by the inhibition of 3H GABA release from striatal

64

E.J. Spary et al. / Journal of Chemical Neuroanatomy 35 (2008) 49–66

nerve terminals (Meyer et al., 1989), and the inhibition of the frequency of miniature IPSPs and the amplitude of evoked IPSPs in neurones of the sensory thalamus, without affecting the membrane properties (Leresche et al., 2000). In addition, somatostatin has been shown to have an action on the release of other transmitters, including dopamine (Chesselet and Reisine, 1983), noradrenaline (Gothert, 1980), glutamate (Wang et al., 1993; Boehm and Betz, 1997) and acetylcholine (Gray et al., 1990). It is probable that these actions are mediated by different sst receptor subtypes, since evidence suggests that each subtype has its own distinct pharmacological properties and functional characteristics, through interaction of the receptor with different G-proteins and intracellular signalling pathways (Csaba and Dournaud, 2001). The NTS, the central site of termination of vagal and glossopharyngeal sensory afferent fibres, contains numerous somatostatin-IR terminals, particularly in the subnuclei associated with cardiovascular and respiratory functions (Kalia et al., 1984; Maley, 1996). Most NTS cells expressing sst subtype immunoreactivities occurred in these areas of dense somatostatin-IR innervation, but with some clear exceptions. It may be that such neurones are influenced by the peptide diffusing through the extracellular space, or might be innervated by terminals containing peptides structurally related to somatostatin, such as cortistatin (de Lecea et al., 1996). Somatostatin has been shown to depress the excitability of NTS neurones via an action on K+ conductance (Jacquin et al., 1988) or Ca2+ currents (Rhim et al., 1996), involving sst receptors coupled to inhibitory G-proteins. In vivo studies reported that microinjection of somatostatin peptides into the NTS resulted in hypotensive and bradycardic responses (Koda et al., 1985) and an elevation of the baroreflex response inhibited by antagonists or antiserum to somatostatin (Lin et al., 1991). The origin of the somatostatinergic innervation in the NTS is mainly intrinsic neurones within the medulla (Helke, 1984; Millhorn et al., 1987), with an additional contribution from certain forebrain areas, including the amygdala (Higgins and Schwaber, 1983; Veening et al., 1984; Saha et al., 2002). A previous study by our group showed that the majority (>80%) of somatostatin-IR terminals in the caudal NTS were GABA immunopositive (Saha et al., 2002) and went on to provide evidence that somatostatin released from GABAergic axon terminals, originating from central nucleus of the amygdala neurones, may be able to modulate cardiovascular reflexes by acting at sst2A receptors expressed by NTS neurones onto which their synapses are formed. The distribution of sst2B and sst4 immunoreactivities suggests that they might also be involved in the postsynaptic responses of NTS neurones to somatostatin. As in other brain regions, the sst1 subtype is more likely to modulate presynaptic responses, although the phenotypes of the axons expressing this subtype are not yet known and there is no firm evidence for such effects in the medulla. The characteristics of the sst3 immunoreactivity, expressed on neuronal cilia, exclude this subtype from a classical pre-/post-synaptic involvement and suggest that they may be able to function as chemical sensors of the somatostatin concentration of the extra-cellular space in

areas that are rich in somatostatinergic input (Ha¨ndel et al., 1999). Somatostatin subtypes are known to assemble as both homoand heterodimers, which can enhance their functional properties (Rocheville et al., 2000b). This raises the question of whether the presence of multiple subtypes in the NTS results in dimerisation to allow greater receptor diversity. Regrettably, due to a lack of availability of reliable antibodies raised in different species we could not address this question directly by performing co-localisation studies for multiple subtypes. However, analysis of sst subtype expression on groups of neurochemically identified neurones hints that individual neurones involved in the autonomic circuitry in the medulla may express multiple subtypes. Groups of cholinergic vagal preganglionic neurones in the DVN and nucleus ambiguus, identified by ChAT immunoreactivity, were immunoreactive for sst2A, sst2B, sst4 and sst5. These neurones supply parasympathetic output to the heart, stomach and visceral organs, and studies in vivo and on slice preparations have shown that somatostatin affects visceral reflex activity by modulation of the activity of these neurones (Oomura and Mizuno, 1986; Yoneda et al., 1991; Wang et al., 1991a,b) in the case of output to the pancreas this was shown to involve sst2 receptors (Liao et al., 2007). The TH positive catecholamine neurones of the dorsomedial and ventrolateral medulla, implicated in the control of blood pressure (Colombari et al., 2001), mostly showed strong sst2A immunoreactivity. However, many of the PNMT-IR cells of the adrenergic C1 group of the RVLM, believed to contribute to the sympathetic premotor output to the intermediolateral horn (IML) of the spinal cord (Phillips et al., 2001; Guyenet, 2006), were strongly associated with sst2A and sst2B immunoreactivities, and less often with sst4 immunoreactivity. While somatostatin administered intracerebroventricularly causes sympathoinhibition (Rettig et al., 1989), and when applied topically to the ventral surface of the medulla elicits a fall in blood pressure (Haxhiu et al., 1993), there have been no neurophysiological studies directly investigating the effects of somatostatin on neurones in the RVLM shown to project to the spinal cord or with activity correlated to blood pressure changes. Intense somatodendritic labelling for sst2A was also observed in non-TH, non-PNMT neurones of the RVLM, and these may represent glutamatergic sympathetic premotor neurones (Guyenet, 2006) or respiratory neurones, which are known to be inhibited by somatostatin, causing apnoea (Chen et al., 1991). 5. Conclusion The detection and localisation of the sst1–sst5 receptor subtypes in areas of the medulla concerned with autonomic regulation, i.e. NTS, DVN, nucleus ambiguus and RVLM, suggests that somatostatin may have modulatory effects on neurones at several points in the autonomic reflex pathway. However, further investigations of the expression of sst subtypes on functionally identified neurones in these areas, together with neuropharmacological studies with subtypespecific agonists and antagonists are required before the

E.J. Spary et al. / Journal of Chemical Neuroanatomy 35 (2008) 49–66

combinations of subtypes mediating cardiovascular and other autonomic responses can be determined. References Aicher, S.A., Milner, T.A., Pickel, V.M., Reis, D.J., 2000. Anatomical substrates for baroreflex sympathoinhibition in the rat. Brain Res. Bull. 51, 107–110. Blessing, W.W., 1997. The Lower Brainstem and Bodily Homeostasis. Oxford University Press, New York. Boehm, S., Betz, H., 1997. Somatostatin inhibits excitatory transmission at rat hippocampal synapses via presynaptic receptors. J. Neurosci. 17, 4066– 4075. Braun, H., Schulz, S., Ho¨llt, V., 2002. Expression changes of somatostatin receptor subtypes sst2A, sst2B, sst3 and sst4 after a cortical contusion trauma in rats. Brain Res. 930, 191–199. Brazeau, P., Vale, W., Burgus, R., Ling, N., Butcher, M., Rivier, J., Guillemin, R., 1973. Hypothalamic polypeptide that inhibits the secretion of immunoreactive pituitary growth hormone. Science 179, 77–79. Bruno, J.F., Xu, Y., Song, J., Berelowitz, M., 1992. Molecular cloning and functional expression of a brain-specific somatostatin receptor. Proc. Natl. Acad. Sci. U.S.A. 89, 11151–11155. Carpentier, V., Vaudry, H., Mallet, E., Tayot, J., Laquerriere, A., Leroux, P., 1997. Ontogeny of somatostatin binding sites in respiratory nuclei of the human brainstem. J. Comp. Neurol. 381, 461–472. Chen, Z.B., Engberg, G., Hedner, T., Hedner, J., 1991. Antagonistic effects of somatostatin and substance P on respiratory regulation in the rat ventrolateral medulla oblongata. Brain Res. 556, 13–21. Chesselet, M.F., Reisine, T., 1983. Somatostatin regulates dopamine release in rat striatal slices and cat caudate nucleus. Neuroscience 3, 232–236. Colombari, E., Sato, M.A., Cravo, S.L., Bergamaschi, C.T., Campos, R.R., Lopes, O.U., 2001. Role of the medulla oblongata in hypertension. Hypertension 38, 549–554. Csaba, Z., Dournaud, P., 2001. Cellular biology of somatostatin receptors. Neuropeptides 35, 1–23. Dampney, R.A.L., 1994. Nucleus of the Solitary Tract. CRC Press, Boca Raton. de Lecea, L., Criado, J.R., Prospero-Garcia, O., Gautvik, K.M., Schweitzer, P., Danielson, P.E., Dunlop, C.L., Siggins, G.R., Henriksen, S.J., Sutcliffe, J.G., 1996. A cortical neuropeptide with neuronal depressant and sleep-modulating properties. Nature 381, 242–245. Dournaud, P., Boudin, H., Schonbrunn, A., Tannenbaum, G.S., Beaudet, A., 1998. Interrelationships between somatostatin sst2A receptors and somatostatin-containing axons in rat brain: evidence for regulation of cell surface receptors by endogenous somatostatin. J. Neurosci. 18, 1056–1071. Dournaud, P., Gu, Y.Z., Schonbrunn, A., Mazella, J., Tannenbaum, G.S., Beaudet, A., 1996. Localization of the somatostatin receptor SST2A in rat brain using a specific anti-peptide antibody. J. Neurosci. 16, 4466– 4478. Dryer, S.E., Dourado, M.M., Wisgirda, M.E., 1991. Properties of Ca2+ currents in acutely dissociated neurons of the chick ciliary ganglion: inhibition by somatostatin-14 and somatostatin-28. Neuroscience 44, 663–672. Fehlmann, D., Langenegger, D., Schuebach, E., Siehler, S., Feuerbach, D., Hoyer, D., 2000. Distribution and characterisation of somatostatin receptor mRNA and binding sites in the brain and periphery. J. Physiol. Paris 94, 265–281. Gothert, M., 1980. Somatostatin selectively inhibits noradrenaline release from hypothalamic neurones. Nature 288, 86–88. Gray, D.B., Zelazny, D., Manthay, N., Pilar, G., 1990. Endogenous modulation of ACh release by somatostatin and the differential roles of Ca2+ channels. J. Neurosci. 10, 2687–2698. Guyenet, P.G., 2006. The sympathetic control of blood pressure. Nature Rev. Neurosci. 7, 335–346. Ha¨ndel, M., Schulz, S., Stanarius, A., Schreff, M., Erdtmann-Vourliotis, M., Schmidt, H., Wolf, G., Ho¨llt, V., 1999. Selective targeting of somatostatin receptor 3 to neuronal cilia. Neuroscience 89, 909–926. Haxhiu, M.A., Deal, E.C., Van, L.E., Cherniack, N.S., 1993. Central effects of somatostatin and atrial natriuretic peptide on tracheal tone. J. Appl. Physiol. 75, 2353–2359.

65

Helboe, L., Schmidt, A., Stidsen, C.E., Møller, M., 1999. Immunohistochemical localization of the somatostatin receptor subtype 2 (sst2) in the central nervous system of the golden hamster (Mesocricetus auratus). J. Comp. Neurol. 405, 247–261. Helboe, L., Stidsen, C.E., Møller, M., 1998. Immunohistochemical and cytochemical localization of the somatostatin receptor subtype sst1 in the somatostatinergic parvocellular neuronal system of the rat hypothalamus. J. Neurosci. 18, 4938–4945. Helke, C.J., 1984. Studies on the origin and release of somatostatin-immunoreactivity in the nucleus of the solitary tract. Brain Res. Bull. 12, 51–55. Hendry, S.H.C., Jones, E.G., DeFilipe, J., Schmechel, D., Brandon, C., Emson, P.C., 1984. Neuropeptide containing neurons of the cerebral cortex are also GABAergic. Proc. Natl. Acad. Sci. U.S.A. 81, 6526–6530. Hervieu, G., Emson, P.C., 1999. Visualisation of somatostatin receptor (3) in the rat central nervous system. Brain Res. Mol. Brain Res. 71, 290–303. Higgins, G.A., Schwaber, J.S., 1983. Somatostatinergic projections from the central nucleus of the amygdala to the vagal nuclei. Peptides 4, 657– 662. Ho¨kfelt, T., Johansson, O., Goldstein, M., 1984. Central catecholamine neurons as revealed by immunohistochemistry with special reference to adrenaline neurons. In: Bjo¨rklund, A., Ho¨kfelt, T. (Eds.), Handbook of Chemical Neuroanatomy, Vol.2, Classical Transmitters in the CNS, Part I. Elsevier, Amsterdam, pp. 157–276. Jacquin, T., Champagnet, J., Madamba, S., Denavit-Saubie, M., Siggins, G.R., 1988. Somatostatin depresses excitability in neurons of the solitary tract complex through hyperpolarization and augmentation of IM, a non-inactivating voltage-dependent outward current blocked by muscarinic agonists. Proc. Natl. Acad. Sci. U.S.A. 85, 948–952. Kalia, M., Fuxe, K., Ho¨kfelt, T., Johansson, O., Lang, R., Ganten, D., Cuello, C., Terenius, L., 1984. Distribution of neuropeptide immunoreactive nerve terminals within the subnuclei of the nucleus of the tractus solitarius of the rat. Neuroscience 222, 409–444. Kang, T.C., An, S.J., Park, S.K., Hwang, I.K., Seo, M.O., Kim, H.S., Kang, J.H., Kwon, O.S., Won, M.H., 2003. The somatostatin receptors in the normal and epileptic hippocampus of the gerbil: subtype-specific localization and its alteration. Brain Res. 986, 91–102. Kluxen, F.-W., Bruns, C., Lu¨bbert, H., 1992. Expression cloning of a rat brain somatostatin receptor cDNA. Proc. Natl. Acad. Sci. U.S.A. 89, 4618–4622. Koda, L.Y., Ling, N., Benoit, P., Madamba, S., Bakhit, C., 1985. Blood pressure following microinjection of somatostatin related peptides into the rat nucleus tractus solitarii. Eur. J. Neurosci. 113, 425–430. Kubota, Y., Hattori, R., Yui, Y., 1994. Three distinct subpopulations of GABAergic neurons in rat frontal agranular cortex. Brain Res. 649, 159–173. Leresche, N., Asprodini, E., Emri, Z., Cope, D.W., Crunelli, V., 2000. Somatostatin inhibits GABAergic transmission in the sensory thalamus via presynaptic receptors. Neuroscience 98, 513–522. Li, M., Forte, M., North, R.A., Ross, C.A., Synder, S.H., 1992. Cloning and expression of a rat somatostatin receptor enriched in brain. J. Biol. Chem. 267, 21307–21312. Liao, Z., Li, Z.-S., Lu, Y., Wang, W.-Z., 2007. Microinjection of exogenous somatostatin in the dorsal vagal complex inhibits pancreatic secretion via somatostatin receptor-2 in rats. Am. J. Physiol. Gastrointest. Liver Physiol. 292, 746–752. Lin, S.S., Chan, J.Y., Chan, S.H., 1991. Facilitation of baroreceptor reflex response by endogenous somatostatin in the rat. Regul. Pept. 33, 239–250. Loewy, A.D., 1990. Central Regulation of Autonomic Functions. Oxford University Press, New York. Maley, B.E., 1996. Immunohistochemical localization of neuropeptides and neurotransmitters in the nucleus solitarius. Chem. Senses 21, 367–376. Maqbool, A., Batten, T.F.C., McWilliam, P.N., 1991. Ultrastructural relationships between GABAergic terminals and cardiac vagal preganglionic motoneurones and vagal afferents in the cat: a combined HRP tracing and immunogold labelling study. Eur. J. Neurosci. 3, 501–513. McDonald, A.J., Mascagni, F., 2002. Immunohistochemical characterisation of somatostatin containing interneurons in the rat basolateral amygdala. Brain Res. 943, 237–244.

66

E.J. Spary et al. / Journal of Chemical Neuroanatomy 35 (2008) 49–66

Meyer, D., Conzelmann, U., Schultheiss, K., 1989. Effects of somatostatin on the in vitro release of (3H)-GABA from slices of rat caudate-putamen. Neuroscience 28, 61–68. Meyerhof, W., Wulfsen, I., Scho¨nrock, C., Fehr, S., Richter, D., 1992. Molecular cloning of a somatostatin-28 receptor and comparison of its expression pattern with that of a somatostatin-14 receptor in rat brain. Proc. Natl. Acad. Sci. U.S.A. 89, 10267–10271. Millhorn, D.E., Seroogy, K., Ho¨kfelt, T., Schmued, L.C., Terenius, L., Buchan, A., Brown, J.C., 1987. Neurons of the ventral medulla oblongata that contain both somatostatin and enkephalin immunoreactivities project to nucleus tractus solitarii and spinal cord. Brain Res. 424, 99– 108. O’Carroll, A.-M., Lolait, S.J., Konig, M., Mahan, L., 1992. Molecular cloning and expression of a pituitary somatostatin receptor with preferential affinity for somatostatin-28. Mol. Pharmacol. 42, 939–946. Oomura, Y., Mizuno, Y., 1986. Effect of somatostatin on the vagal motor neuron in the rat. Brain Res. Bull. 17, 397–401. Patel, Y.C., 1999. Somatostatin and its receptor family. Neuroendocrinology 20, 157–198. Paxinos, G., Watson, C., 1986. The Rat Brain in Stereotaxic Coordinates, 2nd edition. Academic Press, San Diego. Phillips, J.K., Goodchild, A.K., Dubey, R., Sesiashvili, E., Takeda, M., Chalmers, J., Pilowsky, P.M., Lipski, J., 2001. Differential expression of catecholamine synthetic enzymes in the rat ventrolateral medulla. J. Comp. Neurol. 431, 20–34. Pittman, Q.J., Siggins, G.R., 1981. Somatostatin hyperpolarizes hippocampal pyramidal cells in vitro. Brain Res. 221, 402–408. Pradayrol, L., Jornvall, H., Mutt, V., Ribet, A., 1980. N-terminally extended somatostatin: the primary structure of somatostatin-28. FEBS Lett. 109, 55– 58. Rettig, R., Geist, R., Sauer, U., Rohmeiss, P., Unger, T., 1989. Central effects of somatostatin: pressor response, AVP release, and sympathoinhibition. Am. J. Physiol. 257, 588–594. Rhim, H., Toth, P.T., Miller, R.J., 1996. Mechanism of inhibition of calcium channels in rat nucleus tractus solitarius by neurotransmitters. Br. J. Pharmacol. 118, 1341–1350. Rinaman, L., Card, J.P., Schwaber, J.S., Miselis, R.R., 1989. Ultrastructural demonstration of a gastric monosynaptic vagal circuit in the nucleus of the solitary tract in rat. J. Neurosci. 9, 1985–1996. Rocheville, M., Lange, D.C., Kumar, U., Patel, S.C., Patel, R.C., Patel, Y.C., 2000a. Receptors for dopamine and somatostatin: formation of heterooligomers with functional activity. Science 288, 154–157. Rocheville, M., Lange, D.C., Kumar, U., Sasi, R., Patel, R.C., Patel, Y.C., 2000b. Subtypes of the somatstatin receptor assemble as functional homoand heterodimers. Journal of Biological Chemistry 275, 7862–7869. Saha, S., Henderson, Z., Batten, T.F.C., 2002. Somatostatin immunoreactivity in axon terminals in rat nucleus tractus solitarii arising from central nucleus of amygdala: coexistence with GABA and postsynaptic expression of sst2A receptor. J. Chem. Neuroanat. 24, 1–13. Saha, S., Sieghart, W., Fritschy, J.-M., McWilliam, P.N., Batten, T.F.C., 2001. gAminobutyric acid receptor (GABAA) subunits in rat nucleus tractus solitarii (NTS) revealed by polymerase chain reaction (PCR) and immunohistochemistry. Molecular and Cellular Neuroscience 17, 241–257. Saha, S., Spary, E.J., Maqbool, A., Asipu, A., Corbett, E.K.A., Batten, T.F.C., 2004. Increased expression of AMPA receptor subunits in the nucleus of the solitary tract in the spontaneously hypertensive rat. Brain Research Molecular Brain Research 121, 37–49. Scharfman, H.E., Schwartzkroin, P.A., 1989. Selective depression of GABAmediated IPSPs by somatostatin in area CA1 of rabbit hippocampal slices. Brain Res. 493, 205–211.

Schindler, M., Humphrey, P.P.A., Emson, P.C., 1996. Somatostatin receptors in the central nervous system. Prog. Neurobiol. 50, 9–47. Schindler, M., Humphrey, P.P.A., Lo¨hrke, S., Friauf, E., 1999. Immunohistochemical localization of the somatostatin sst2(b) receptor splice variant in the rat central nervous system. Neuroscience 90, 859–874. Schreff, M., Schulz, S., Ha¨ndel, M., Keilhoff, G., Braun, H., Pereira, G., Klutzny, M., Schmidt, H., Wolf, G., Ho¨llt, V., 2000. Distribution, targeting, and internalization of the sst4 somatostatin receptor in rat brain. J. Neurosci. 3785–3797. Schulz, S., Schulz, S., Schmitt, J., Wiborny, D., Schmidt, H., Olbricht, S., Weise, W., Roessner, A., Gramsch, C., Ho¨llt, V., 1998. Immunocytochemical detection of somatostatin receptors sst1, sst2A, sst2B, and sst3 in paraffinembedded breast cancer tissue using subtype-specific antibodies. Clin. Cancer Res. 4, 2047–2052. Schulz, S., Ha¨ndel, M., Schreff, M., Schmidt, H., Ho¨llt, V., 2000. Localisation of five somatostatin receptors in the rat central nervous system using subtype-specific antibodies. J. Physiol. Paris 94, 259–264. Selmer, I., Schindler, M., Allen, J.P., Humphrey, P.P.A., Emson, P.C., 2000. Advances in understanding neuronal somatostatin receptors. Regul. Pept. 90, 1–18. Sharif, N., Gendron, L., Wowchuk, J., Sarret, P., Mazella, J., Beaudet, A., Stroh, T., 2007. Coexpression of somatostatin receptor 5 affects internalization and trafficking of somatostatin receptor 2. Endocrinology 148, 2095–2105. Somogyi, P., Hodgson, A.J., Smith, A.D., Nunzi, M.G., Gorio, A., Wu, J.-Y., 1984. Different populations of GABAergic neurons in the visual cortex and hippocampus of cat contain somatostatin- or cholecystokinin-immunoreactive material. J. Neurosci. 4, 2590–2603. Spyer, K.M., 1994. Central nervous mechanisms contributing to cardiovascular control. J. Physiol. (Lond.) 474, 1–19. Stroh, T., Kreienkamp, H.J., Beaudet, A., 1999. Immunohistochemical distribution of the somatostatin receptor subtype 5 in the adult brain: predominant expression in the basal forebrain. J. Comp. Neurol. 412, 69–82. Tallent, M.K., Siggins, G.R., 1999. Somatostatin acts in CA1 and CA3 to reduce hippocampal epileptiform activity. J. Neurophysiol. 81, 1626–1635. Twery, M.J., Gallagher, J.P., 1990. Somatostatin depresses GABA receptormediated inhibition in the rat dorsolateral septal nucleus. Brain Res. 519, 277–286. Van Giersbergen, P.L.M., Palkovits, M., De Jong, W., 1992. Involvement of neurotransmitters in the nucleus tractus solitarii in cardiovascular regulation. Physiol. Rev. 72, 789–824. Vanetti, M., Kouba, M., Vogt, G., Ho¨llt, V., 1992. Cloning and expression of a novel mouse somatostatin receptor (SSTR2B). FEBS Lett. 311, 290–294. Veening, J.G., Swanson, L.W., Sawchenko, P.E., 1984. The organization of projections from the central nucleus of the amygdala to brainstem sites involved in central autonomic regulation: a combined retrograde transportimmunohistochemical study. Brain Res. 303, 337–357. Vezzani, A., Hoyer, D., 1999. Brain somatostatin: a candidate inhibitory role in seizures and epileptogenesis. Eur. J. Neurosci. 11, 3767–3776. Wang, Y.T., Zhang, M., Neuman, R.S., Bieger, D., 1993. Somatostatin regulates excitatory amino acid receptor-mediated fast excitatory postsynaptic potential components in vagal motoneurons. Neuroscience 53, 7–9. Wang, Y.T., Neuman, R.S., Bieger, D., 1991a. Somatostatin inhibits nicotinic cholinoceptor mediated-excitation in rat ambigual motoneurons in vitro. Neurosci. Lett. 123, 236–239. Wang, Y.T., Neuman, R.S., Bieger, D., 1991b. Nicotinic cholinoceptormediated excitation in ambigual motoneurons of the rat. Neuroscience 40, 759–767. Yoneda, M., Raybould, H., Tache, Y., 1991. Central action of somatostatin analog, SMS 201-995, to stimulate gastric acid secretion in rats. Peptides 12, 401–406.