Dominique Fehlmann

Distribution and characterisation of somatostatin receptor mRNA and binding sites in the brain and periphery

The distribution and nature of (somatostatin) SRIF receptors and receptor mRNAs was studied in the brain and periphery of various laboratory animals using in situ hybridisation, autoradiography and radioligand binding. The messenger RNA (mRNA) expression of SRIF receptors msst1, msst2, msst3, msst4 and msst5 was studied in the adult mouse brain by in situ hybridisation histochemistry using specific oligonucleotide probes and compared to that of adult rats. As observed in rat brain, sst3 receptor mRNA is prominently expressed across the mouse brain, although equivalent binding has not yet been identified in situ. Sst1 and sst2 receptor mRNA expression, was prominent and again comparable to that observed in rat brain, whereas sst4 and especially sst5 receptor mRNA show comparatively low levels, although the former appears to be widely distributed while the latter could only be identified in a few nuclei. Altogether, the data are compatible with current knowledge, i.e. sst1 and sst2 receptor mRNA is prominent (both receptors have been functionally identified in the brain and for sst2 in the periphery), sst3 mRNA is highly expressed but in the absence of any functional correlate remains elusive. The expression of sst4 mRNA is comparatively low (especially when compared to what is seen in the lung, where high densities of sst4 receptors are present) and it remains to be seen whether sst5 receptor mRNA, which is confined to a few nuclei, will play a role in the brain, keeping in mind that high levels are found in the pituitary. Radioligand binding studies were performed in CCL39 cells expressing the five human recombinant receptors and compared to binding in membranes of rat cerebral cortex with [125I]Tyr11-SRIF14 which in the presence of 120 mM labels primarily sst1 receptor as suggested by the better correlation hsst1 and similar rank order of potency. The profile of [125I]Tyr3-octreotide labelled sites in rat cortex correlates better with recombinant sst2 than sst3 or sst5 binding profiles. Finally, [125I]LTT-SRIF28-labelled sites in rat lung express a sst4 receptor profile in agreement with previous findings. SRIF receptor autoradiography was performed in the brain and peripheral tissue of rat and/or guinea-pig using a number of ligands known to label recombinant SRIF receptors: [125I]LTT-SRIF28, [125I]CGP 23996, [125I]Tyr10-CST, or [125I]Tyr3-octreotide. Although, [125I]Tyr10-CST has been shown to label all five recombinant SRIF receptors, it is apparent that this radioligand is not useful for autoradiographic studies. By contrast, the other three ligands show good signal to noise ratios in rat or guinea-pig brain, rat lung, rat pancreas, or guinea-pig ileum. In most tissues, [125I]Tyr3-octreotide represents a prominent part of the binding (when compared to [125I]LTT-SRIF28 and [125I]CGP 23996), suggesting that sst2 receptors are strongly expressed in most tissues; it is only in rat lung that [125I]LTT-SRIF28 and [125I]CGP 23996 show marked binding, whereas [125I]Tyr3-octreotide does apparently label no sites, in agreement with the sole presence of sst4 receptors in this tissue.

Somatostatin receptor subtypes 2 and 4 affect seizure susceptibility and hippocampal excitatory neurotransmission in mice

We have investigated the role of somatostatin receptor subtypes sst2 and sst4 in limbic seizures and glutamate‐mediated neurotransmission in mouse hippocampus. As compared to wild‐type littermates, homozygous mice lacking sst2 receptors showed a 52% reduction in EEG ictal activity induced by intrahippocampal injection of 30 ng kainic acid (P < 0.05). The number of behavioural tonic–clonic seizures was reduced by 50% (P < 0.01) and the time to onset of seizures was doubled on average (P < 0.05). Seizure‐associated neurodegeneration was found in the injected hippocampus (CA1, CA3 and hilar interneurons) and sporadically in the ipsilateral latero‐dorsal thalamus. This occurred to a similar extent in wild‐type and sst2 knock‐out mice. Intrahippocampal injection of three selective sst2 receptor agonists in wild‐type mice (Octreotide, BIM 23120 and L‐779976, 1.5–6.0 nmol) did not affect kainate seizures while the same compounds significantly reduced seizures in rats. L‐803087 (5 nmol), a selective sst4 receptor agonist, doubled seizure activity in wild‐type mice on average. Interestingly, this effect was blocked by 3 nmol octreotide. It was determined, in both radioligand binding and cAMP accumulation, that octreotide had no direct agonist or antagonist action at mouse sst4 receptors expressed in CCl39 cells, up to micromolar concentrations. In hippocampal slices from wild‐type mice, octreotide (2 µm) did not modify AMPA‐mediated synaptic responses while facilitation occurred with L‐803087 (2 µm). Similarly to what was observed in seizures, the effect of L‐803087 was reduced by octreotide. In hippocampal slices from sst2 knock‐out mice, both octreotide and L‐803087 were ineffective on synaptic responses. Our findings show that, unlike in rats, sst2 receptors in mice do not mediate anticonvulsant effects. Moreover, stimulation of sst4 receptors in the hippocampus of wild‐type mice induced excitatory effects which appeared to depend on the presence of sst2 subtypes, suggesting these receptors are functionally coupled.

Somatostatin sst2 receptor knock-out mice: localisation of sst1–5 receptor mRNA and binding in mouse brain by semi-quantitative RT–PCR, in situ hybridisation histochemistry and receptor autoradiography

The peptide hormone/neurotransmitter somatostatin (somatotropin release inhibiting factor; SRIF) and its receptors (sst(1)-sst(5)) appear to regulate many physiological functions in the CNS. Semi-quantitative analysis of the densities of mRNA expression for sst(1-5) receptors and SRIF receptor binding sites were established in sst(2) receptor knock-out (KO) mice. Patterns of sst(1-5) receptor mRNA expression were largely conserved for sst(1,3,4) and sst(5) selective oligonucleotide probes; whereas sst(2) signals were completely absent in KO mouse brain. Autoradiographic analysis demonstrated [(125)I]LTT SRIF(28), [(125)I]CGP 23996 (two radioligands known to label all five recombinant SRIF receptors) and [(125)I]Tyr(3)-octreotide (sst(2) and sst(5) receptor selective) binding in wild type (WT) mouse brain sections; yet no specific binding of [(125)I]Tyr(3)-octreotide in KO mice. In contrast, [(125)I]LTT SRIF(28) and [(125)I]CGP 23996 binding was still present in a number of brain areas in KO mice, although to a lesser degree than in those regions where [(125)I]Tyr(3)-octreotide binding was found, in WT animals. The present data suggest first, that both sst(2) receptor protein and mRNA were completely absent in the brain of these KO animals. Second, there was little evidence of compensatory regulation, at the mRNA level, of the other SRIF receptors as a consequence of the sst(2) KO. Third, the absence of any [(125)I]Tyr(3)-octreotide binding, in KO mice, suggests that this particular ligand is selective for the sst(2) receptor subtype (under the conditions utilised); or that sst(5) receptors are only marginally expressed in brain. Fourth, there were regions where the binding of [(125)I]LTT SRIF(28) and [(125)I]CGP 23996 were moderately affected by the sst(2) KO, suggesting that additional SRIF receptors may well contribute to the binding of the aforementioned radioligands. Finally, since the relative distribution of these two ligands were not entirely superimposable, it suggests that their respective selectivity profiles towards the different SRIF receptor subtypes in situ are not identical.

Pharmacological characterisation of native somatostatin receptors in AtT-20 mouse tumour corticotrophs

The mouse corticotroph tumour cell line AtT‐20 is a useful model to investigate the physiological role of native somatostatin (SRIF, Somatotropin release inhibitory factor) receptor subtypes (sst1 – sst5). The objective of this study was to characterise the pharmacological features and the functional effects of SRIF receptors expressed by AtT‐20 cells using radioligand binding and cAMP accumulation. [125I]LTT‐SRIF‐28, [125I]CGP 23996, [125I]Tyr10‐cortistatin‐14 and [125I]Tyr3‐octreotide labelled SRIF receptor binding sites with high affinity and in a saturable manner (Bmax=315, 274, 239 and 206 fmol mg−1, respectively). [125I]LTT‐SRIF‐28 labels significantly more sites than [125I]Tyr10 ‐cortistatin‐14 and [125I]Tyr3 ‐octreotide as seen previously in cells expressing pure populations of sst2 or sst5 receptors. SRIF analogues displaced the binding of the four radioligands. sst2/5 receptor‐selective ligands showed much higher affinity than sst1/3/4 receptor‐selective ligands. The binding profile of [125I]Tyr3‐octreotide was different from that of [125I]LTT‐SRIF‐28, [125I]CGP 23996 and [125I]Tyr10‐cortistatin‐14. The sst5/1 receptor‐selective ligand L‐817,818 identified two binding sites, one with subnanomolar affinity (sst5 receptors) and one with micromolar affinity (sst2 receptors); however, the proportions were different: 70 – 80% high affinity with [125I]LTT‐SRIF‐28, [125I]CGP 23996, [125I]Tyr10‐cortistatin‐14, but only 20% with [125I]Tyr3‐octreotide. SRIF analogues inhibited the forskolin‐stimulated cAMP levels depending on concentration. sst2/5 receptor‐selective ligands were highly potent, whereas sst1/3/4 receptor‐selective ligands had no significant effects. The sst2 receptor antagonist D‐Tyr8‐CYN 154806 competitively antagonised the effects of SRIF‐14 and sst2 receptor‐preferring agonists, but not those of L‐817,818. The complex binding properties of SRIF receptor analogues indicate that sst2 and sst5 receptors are the predominant SRIF receptors expressed on AtT‐20 cell membranes with no or only negligible presence of sst1, sst3 and sst4 receptors. In the functional studies using cAMP accumulation, only sst2 and sst5 receptors appear to play a role. However, the ‘predominant’ receptor appears to be the sst2 receptor, although sst5 receptors can also mediate the effect, when the ligand is not able to activate sst2 receptors. This clearly adds flexibility to SRIF‐mediated functional effects and suggests that the physiological role of SRIF and its analogues may be mediated preferentially via one subtype over another.

Compensatory changes in the hippocampus of somatostatin knockout mice: upregulation of somatostatin receptor 2 and its function in the control of bursting activity and synaptic transmission

Somatostatin‐14 (SRIF) co‐localizes with γ‐aminobutyric acid (GABA) in the hippocampus and regulates neuronal excitability. A role of SRIF in the control of seizures has been proposed, although its exact contribution requires some clarification. In particular, SRIF knockout (KO) mice do not exhibit spontaneous seizures, indicating that compensatory changes may occur in KO. In the KO hippocampus, we examined whether specific SRIF receptors and/or the cognate peptide cortistatin‐14 (CST) compensate for the absence of SRIF. We found increased levels of both sst2 receptors (sst2) and CST, and we explored the functional consequences of sst2 compensation on bursting activity and synaptic responses in hippocampal slices. Bursting was decreased by SRIF in wild‐type (WT) mice, but it was not affected by either CST or sst2 agonist and antagonist. sst4 agonist increased bursting frequency in either WT or KO. In WT, but not in KO, its effects were blocked by agonizing or antagonizing sst2, suggesting that sst2 and sst4 are functionally coupled in the WT hippocampus. Bursting was reduced in KO as compared with WT and was increased upon application of sst2 antagonist, while SRIF, CST and sst2 agonist had no effect. At the synaptic level, we observed that in WT, SRIF decreased excitatory postsynaptic potentials which were, in contrast, increased by sst2 antagonist in KO. We conclude that sst2 compensates for SRIF absence and that its upregulation is responsible for reduced bursting and decreased excitatory transmission in KO mice. We suggest that a critical density of sst2 is needed to control hippocampal activity.

Cloning, expression and pharmacological characterisation of the mouse somatostatin sst5 receptor

Abstract The mouse somatostatin (somatotropin release inhibiting factor, SRIF) sst 5 receptor coding sequence was cloned from a mouse BALB/c genomic library. It shows 97% and 81% homology with the corresponding rat and human receptors, respectively. The msst 5 receptor messenger RNA (mRNA) is present at low levels in the adult mouse brain, with significant expression in a few nuclei only, e.g. in the septum (lateral septal nuclei) or the amygdala (medial amygdaloid nucleus); very few signals were observed in the mesencephalon, metencephalon, and myelencephalon (except the dorsal motor nucleus of the vagus nerve). The msst 5 receptor was stably expressed in the hamster fibroblast cell line CCL39-SRE-Luci, which harbours the luciferase reporter gene driven by the serum responsive element. [ 125 I]LTT-SRIF-28 ([Leu 8 , D-Trp 22 , 125 I-Tyr 25 ]-SRIF-28), [ 125 I]Tyr 10 -CST, [ 125 I]CGP 23996, and [ 125 I]Tyr 3 -octreotide labelled msst 5 receptors with high affinity (pK d values: 11.0, 10.15, 9.75 and 9.43) and in a saturable manner, but defined different Bmax values: 697, 495, 540 and 144 fmoles/mg, respectively. [ 125 I]LTT-SRIF-28-labelled sites displayed the following rank order: SRIF-28> rCST-14> somatuline > CGP-23996= SRIF-14= octreotide, whereas [ 125 I]Tyr 3 -octreotide-labelled sites displayed a different profile: octreotide > SRIF-28> rCST-14= somatuline > SRIF-14> CGP-23996. The pharmacological profiles determined with [ 125 I]LTT-SRIF-28, [ 125 I]CGP 23996 and [ 125 I]Tyr 10 -CST correlated highly significantly (r 2 =0.88–0.99), whereas [ 125 I]Tyr 3 -octreotide binding was rather divergent (r 2 =0.77). Also, human and mouse sst 5 receptor profiles are very different, e.g. r 2 =0.385 for [ 125 I]Tyr 10 -CST and r 2 =0.323 for [ 125 I]LTT-SRIF-28-labelled sites. Somatostatin induces expression of luciferase reporter gene in CCL39-SRE-Luci cells. The profile was consistent with a msst 5 receptor-mediated effect although apparent potency in the luciferase assay was much reduced compared to radioligand binding data: Octreotide = SRIF-28> rCST-14= SRIF-14= CGP-23996. Octreotide, SRIF-28, BIM23052 and D Tyr Cyanamid 154806 behaved as full or nearly full agonists in comparison to SRIF-14, whereas the other compounds had relative efficacies of 40 to 70%. The present study shows that agonists radioligands define apparently different receptor populations in terms of number of sites and pharmacological profile in cells expressing a single recombinant receptor. These variations suggest that the conformation of the ligand receptor complex may vary depending on the agonist. Further, the msst 5 receptor, although primarily coupled to Gi/Go proteins, is able to stimulate luciferase gene expression driven by the serum responsive element. Finally, it is suggested that putative sst 2 selective agonists e.g. octreotide, RC160 or BIM23027 show similar or higher potency at msst 5 receptors than SRIF-14.

Antidepressants Influence Somatostatin Levels and Receptor Pharmacology in Brain

This study investigated how the administration (acute and chronic) of the antidepressants citalopram and desmethylimipramine (DMI) influences somatostatin (somatotropin release inhibitory factor, SRIF) levels and SRIF receptor density (sst1–5) in rat brain. Animals received either of the following treatments: (1) saline for 21 days (control group), (2) saline for 20 days and citalopram or DMI for 1 day (citalopram or DMI acute groups), (3) citalopram or DMI for 21 days (citalopram or DMI chronic groups). Somatostatin levels were determined by radioimmunoassay. [125I]LTT SRIF-28 binding in the absence (labeling of sst1–5) or presence of 3 nM MK678 (labeling of sst1/4) and [125I]Tyr3 octreotide (labeling of sst2/5) binding with subsequent autoradiography was performed in brains of rats treated with both antidepressants. Somatostatin levels were increased after citalopram, but not DMI administration, in the caudate-putamen, hippocampus, nucleus accumbens, and prefrontal cortex. Autoradiography studies illustrated a significant decrease in receptor density in the superficial and deep layers of frontal cortex (sst2), as well as a significant increase in the CA1 (sst1/4) hippocampal field in brains of chronically citalopram-treated animals. DMI administration increased sst1/4 receptors levels in the CA1 hippocampal region. These results suggest that citalopram and to a lesser extent DMI influence the function of the somatostatin system in brain regions involved in the emotional, motivational, and cognitive aspects of behavior.

Agonist properties of putative small-molecule somatostatin sst2 receptor-selective antagonists

The availability of antagonist ligands for somatostatin receptors is very limited, with those that are available often displaying agonist properties or limited receptor subtype selectivity. Hay et al. [Bioorg. Med. Chem. Lett. 11 (2001) 2731] recently described the development of small-molecule somatostatin receptor subtype 2 (sst(2)) selective compounds. This study investigates the binding affinity and functional characteristics of two of those antagonists (2 and 3) and the agonist compound, from which they were derived (1). In radioligand binding studies using the agonist radioligands [125I][Tyr(11)]SRIF-14 (Ala-Gly-c[Cys-Lys-Asn-Phe-Phe-Trp-Lys-Thr-(125I-Tyr)-Thr-Ser-Cys]-OH), [125I]LTT-SRIF-28 ([Leu(8),DTrp(22),125I-Tyr(25)]SRIF-28; Ser-Ala-Asn-Ser-Asn-Pro-Ala-Leu-Ala-Pro-Arg-Glu-Arg-Lys-Ala-Gly-c[Cys-Lys-Asn-Phe-Phe-DTrp-Lys-Thr-(125I-Tyr)-Thr-Ser-Cys]-OH), [125I]CGP 23996 (c[Lys-Asu-Phe-Phe-Trp-Lys-Thr-(125I-Tyr)-Thr-Ser]), [125I][Tyr(3)]octreotide (DPhe-c[Cys-(125I-Tyr)-DTrp-Lys-Thr-Cys]-Thr-OH) and [125I][Tyr(10)]cortistatin-14 (Pro-c[Cys-Lys-Asn-Phe-Phe-Trp-Lys-Thr-(125I-Tyr)-Ser-Ser-Cys]-Lys) at human recombinant somatostatin receptors expressed in Chinese hamster lung fibroblast (CCL39) cells and native rat cortex, the compounds bound with high affinity (pK(d) 6.8-9.7) and selectivity to human sst(2) receptors. Some affinity was also observed for sst(5) labelled by [125I][Tyr(3)]octreotide and [125I]CGP 23996. In functional studies at human sst(2) receptors expressed in Chinese hamster ovary (CHO) cells, both the agonist 1 and the two putative antagonists 2 and 3 concentration dependently inhibited forskolin-stimulated adenylate cyclase and stimulated luciferase reporter gene expression, with similar efficacy to the natural ligand somatotropin release inhibiting factor (SRIF)-14. Compound 1 had similar potency to SRIF-14, which was in the nanomolar range, whereas 2 and 3 were 10-100-fold less potent. The intrinsic activity of 2 and 3 was too high to allow antagonist studies to be carried out. In conclusion, in contrast to previous findings, all three compounds are potent agonists at recombinant human sst(2) receptors.

Blunted 5-HT1A receptor-mediated responses and antidepressant-like behavior in mice lacking the GABAB1a but not GABAB1b subunit isoforms

RationaleThere is accumulating evidence for a role of GABAB receptors in depression. GABAB receptors are heterodimers of GABAB1 and GABAB2 receptor subunits. The predominant GABAB1 subunit isoforms are GABAB1a and GABAB1b. GABAB1 isoforms in mice differentially influence cognition, conditioned fear, and susceptibility to stress, yet their influence in tests of antidepressant-like activity has not been fully investigated.ObjectivesGiven the interactions between GABAB receptors and the serotonergic system and the involvement of 5-HT1A receptors (5-HT1AR) in antidepressant action, we sought to evaluate 5-HT1AR function in GABAB1a−/− and GABAB1b−/− mice.MethodsGABAB1a−/− and GABAB1b−/− mice were assessed in the forced swim test (FST), and body temperature and hypothalamic-pituitary-adrenal (HPA) responses to the 5-HT1AR agonist 8-OH-DPAT were determined. Brain 5-HT1AR expression was assessed by [3H]-MPPF and [3H]-8-OH-DPAT autoradiography and 5-HT1AR G-protein coupling by [35S]GTP-γ-S autoradiography.ResultsAs previously described, GABAB1a−/− mice showed an antidepressant-like profile in the FST. GABAB1a−/− mice also demonstrated profoundly blunted hypothermic and motoric responses to 8-OH-DPAT. Furthermore, 8-OH-DPAT-induced corticosterone and adrenocorticotropic hormone (ACTH) release were both attenuated in GABAB1a−/− mice. Interestingly, [3H]-MPPF and [3H]-8-OH-DPAT binding was largely unaffected by genotype. [35S]GTP-γ-S autoradiography suggested that altered 5-HT1AR G-protein coupling only partially contributes to the functional presynaptic 5-HT1AR desensitization, and not at all to the blunted postsynaptic 5-HT1AR-mediated responses, seen in GABAB1a−/− mice.ConclusionThese data demonstrate distinct functional links between 5-HT1ARs and the GABAB1a subunit isoform and suggest that the GABAB1a isoform may be implicated in the antidepressant-like effects of GABAB receptor antagonists and in neurobiological mechanisms underlying depression.