Address correspondence and reprint requests to Paola Bagnoli, Department of Biology, Unit of General Physiology, University of Pisa, via, San Zeno 31, Pisa 56127, Italy. E-mail: firstname.lastname@example.org
Somatostatin-14 (SRIF) co-localizes with GABA in the hippocampus and regulates neuronal excitability. A role of SRIF in the control of hippocampal activity has been proposed, although the exact contribution of each SRIF receptor (sst1–sst5) in mediating SRIF action requires some clarification. We used hippocampal slices of wild-type and sst1 knockout (KO) mice and selective pharmacological tools to provide conclusive evidence for a role of sst1 in mediating SRIF inhibition of synaptic transmission. With single- and double-label immunohistochemistry, we determined the distribution of sst1 in hippocampal slices and we quantified sst1 colocalization with SRIF. With electrophysiology, we found that sst1 activation with CH-275 inhibited both the NMDA- and the α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)-mediated responses. Results from sst1 KO slices confirmed the specificity of CH-275 effects; sst1 activation did not affect the inhibitory transmission which was in contrast increased by sst4 activation with L-803,087 in both wild-type and sst1 KO slices. The AMPA-mediated responses were increased by L-803,087. Functional interaction between sst1 and sst4 is suggested by the finding that their combined activation prevented the CH-275-induced inhibition of AMPA transmission. The involvement of pre-synaptic mechanisms in mediating inhibitory effects of sst1 on excitatory transmission was demonstrated by the finding that CH-275 (i) increased the paired-pulse facilitation ratio, (ii) did not influence the AMPA depolarization in the presence of tetrodotoxin, and (iii) inhibited glutamate release induced by epileptiform treatment. We conclude that SRIF control of excitatory transmission through an action at sst1 may represent an important contribution to the regulation of hippocampal activity.
The peptide somatostatin-14, or SRIF (somatotropin release-inhibiting factor), is expressed by hippocampal interneurons and plays an important inhibitory role on excitatory activity including epilepsy (see Tallent and Qiu 2008). Information on the precise contribution of each SRIF receptor subtype (sst1–sst5) on the SRIF-induced inhibition of hippocampal activity is still limited. In epileptic models, for instance, there is indication that sst2 and sst4 may mediate the majority of SRIF antiepileptic actions, with a main role of sst2 in rat and of sst4 in mouse (Qiu et al. 2008; see for ref Tallent and Qiu 2008). On the other hand, there are also data suggesting excitatory actions of sst4 on seizure susceptibility and hippocampal excitatory neurotransmission in rodents (Moneta et al. 2002; Cammalleri et al. 2004). In addition, a role of sst1 in mediating anticonvulsant effects of SRIF has been demonstrated in a mouse hippocampal model of interictal-like activity (Cammalleri et al. 2004) although this has been not confirmed by Qiu et al. (2008). There are also indications that SRIF receptors may mediate SRIF antiepileptic effects by modulating each other function. For instance, in the absence of sst1, as in sst1 knockout (KO) mice, sst2 blockade causes a significant increase in the frequency of spontaneous epileptiform discharge indicating functional interaction between sst1 and sst2 (Cammalleri et al. 2004). In addition, excitatory effects of sst4 in the mouse hippocampus appear to depend on sst2 suggesting that these receptors are functionally coupled (Moneta et al. 2002).
Information on sst1 expression and localization in the rodent hippocampus is rather controversial (see for ref. Viollet et al. 2008). There are studies demonstrating that hippocampal neurons express sst1 mRNA (Pérez et al. 1994; Hannon et al. 2002; Cammalleri et al. 2004, 2006), whereas other studies have excluded the presence of sst1 protein in the rodent hippocampus. For instance, the distribution of SRIF binding sites remains almost unchanged in the brain of sst1 KO mice suggesting that sst1 does not represent a significant proportion of SRIF binding in the mouse brain (Videau et al. 2003). By immunohistochemistry in the rat brain, discrepant results have been reported using antibodies directed against the N-terminal part (Hervieu and Emson 1998) or the C-terminal part of the human (Helboe et al. 1998) or rat (Schulz et al. 2000) sst1. Recently, the presence of sst1-immunoreactivity (sst1-IR) in the rat hippocampus has been excluded by Kwak et al. (2008) and Spary et al. (2008) whereas other results demonstrated that sst1-IR is indeed localized to distinct regions of the mouse hippocampus (Rajput et al. 2009). The presence of sst1-IR in the rat hippocampus has been reported after pilocarpine treatment indicating an sst1 involvement in epilepsy (Kwak et al. 2008).
Whether pre- and/or post-synaptic mechanisms mediate the SRIF-induced inhibition of hippocampal activity still remains to be clarified. In an in vitro model of hippocampal hyperactivity, we have recently demonstrated that SRIF decreases the phosphorylation of a protein kinase A-dependent site on NMDA receptor subunit NR1 (Ristori et al. 2008a). This has been interpreted to cause a decrease of post-synaptic excitatory currents with a consequent hyperpolarization of pyramidal neurons and/or to modulate pre-synaptically glutamate release by pyramidal cells. In this respect, there are results demonstrating that in the mouse hippocampus SRIF inhibition of hippocampal activity may involve a reduction of NMDA receptor-mediated excitatory post-synaptic currents (Tallent and Siggins 1997, 1999). On the other hand, there are also results demonstrating the involvement of pre-synaptic mechanisms in the SRIF-mediated inhibition of excitatory transmission in the rodent hippocampus (Boehm and Betz 1997; Tallent and Siggins 1997; Cammalleri et al. 2006).
In this study, we used hippocampal slices of wild-type (WT) and sst1 KO mice and selective pharmacological tools to provide conclusive evidence for a role of sst1 in mediating SRIF inhibition of excitatory transmission. In this respect, both single- and double-label immunohistochemistry was performed to clarify the localization pattern of sst1 in hippocampal slices and to quantify sst1 expression by SRIF-containing neurons with the use of the optical disector stereological technique. At the functional level, the effects of sst1 activation on synaptic activity were established also in respect to possible interactions between sst1 and sst4. In addition, the possibility that pre-synaptic mechanisms may mediate inhibitory effects of sst1 on excitatory transmission was also investigated with electrophysiology. Finally, a well-established model of epileptiform bursting in rodent hippocampal slices (Siniscalchi et al. 1997; Sanna et al. 2000; Cammalleri et al. 2004, 2006; Kilb et al. 2007; Ristori et al. 2008a,b) was used to determine whether sst1 activation affects glutamate release.
Materials and methods
Experiments were performed on 48 WT (C57BL/6 strain) and 23 sst1 KO mice of both sexes at 6–8 weeks after birth (20–30 g body weight). The generation of sst1 KO mice has been detailed in Kreienkamp et al. (1999). To reduce genetic background variability, the sst1 null allele was backcrossed for five successive generations onto the C57BL/6J strain to produce N5 incipient congenic mice. Animals were kept in a regulated environment (23 ± 1°C, 50 ± 5% humidity) with a 12-h light/dark cycle (lights on at 8 am) with food and water ad lib. In all experiments, mice were anesthetized with halothane (4%). Experiments were performed in compliance with the Italian law on animal care No. 116/1992 and in accordance with the European Community Council Directive (EEC/609/86). All efforts were made to reduce both animal suffering and the number of animals used.
Preparation of mouse hippocampal slices
Hippocampal slices were prepared as previously described (Cammalleri et al. 2004, 2006; Ristori et al. 2008a,b). Briefly, mice were anesthetized and decapitated. Their brains were rapidly removed, placed in ice-cold artificial CSF (aCSF) and gassed with 95% O2 and 5% CO2. The composition of aCSF was (in mM): NaCl 130, KCl 3.5, NaH2PO4 1.25, MgSO4.7H2O 1.5, CaCl2.2H2O2 2, NaHCO3 24, and glucose 10, pH 7.4. Transverse hippocampal slices (400 μm) were prepared with the use of a Vibratome (Campden Instruments, Loughborough, UK) and incubated for 1 h at 23 ± 2°C in aCSF.
Hippocampal slices were fixed for 1.5 h in 4%p-formaldehyde in 0.1 M phosphate buffer (PB) at 4°C. They were rinsed in 0.1 M PB and incubated for 48 h at 4°C in either a pre-diluted rabbit polyclonal antibody (about 6 μg/mL) and directed against the C-terminal segment of the rat protein that specifically recognizes the sst1 antigen (ab27419; Abcam plc, Cambridge Science Park, Cambridge, UK) or 1 : 50 of a rat monoclonal antibody directed to SRIF (MAB354; Chemicon, Temecula, CA, USA). Triton X-100 at 0.3% was added to either ab27419 or MAB354 diluted in 0.1 M PB. After incubation, the slices were rinsed in 0.1 M PB and incubated overnight at 4°C in the respective secondary antibody (Alexa Fluor-488 or -546; Molecular Probes, Eugene, OR, USA) at a dilution of 1 : 200 in 0.1 M PB containing 0.3% Triton X-100. Finally, the slices were rinsed in 0.1 M PB, mounted on gelatin-coated glass slides, and coverslipped with a 0.1 M PB-glycerine mixture. The specificity of IR was evaluated by the use of pre-immune serum instead of the primary antibody or pre-absorption tests using control antigen peptides (Santa Cruz Biotechnology, Santa Cruz, CA, USA). An additional control for immunohistochemistry included the use of sst1 KO hippocampal slices to assess the specificity of the signals obtained with the sst1 antibody used in this study. The controls for immunohistochemistry resulted in the absence of IR.
In double-labeling experiments, the slices were washed in 0.1 M PB and incubated in both antibodies in the presence of 0.3% Triton X-100. Incubation in appropriate secondary antibodies overnight at 23 ± 2°C followed. The slices were then washed in 0.1 M PB and coverslipped. To eliminate the possibility of cross-reaction between primary and secondary antibodies in double-labeling experiments, control sections were made by omitting either of the primary antibodies. Control experiments were also performed to ensure that the primary antibodies did not cross-react when mixed together and that the secondary antibodies reacted only with the appropriate antigen–antibody complex. Immunofluorescent materials were observed with confocal microscopy (Bio-Rad Laser Scanning Microscope Radiance Plus, Hercules, CA, USA) using 10×, 20×, or 40× objective lens. Serial optical sections (1 μm apart) were scanned through the thickness of each slice at the same predetermined z-axis and collected for the analysis of sst1- and/or SRIF-IR distribution. Electronic images from the confocal microscope were processed using Adobe Photoshop 5.0 (Adobe Systems, Inc., Mountain View, CA, USA).
Quantitative analysis of sst1- and/or SRIF-positive somatic profiles (SP) was performed with the optical disector stereological technique (Gundersen et al. 1988; West and Gundersen 1990) in agreement with previous studies (Jinno and Kosaka 2004; see for ref. Jinno and Kosaka 2006). Briefly, stacks of confocal consecutive optical sections obtained at each frame with 20× objective were projected, and a montage image of the entire hippocampus was composed by the use of the software LaserSharp Radiance Plus (Bio-Rad). Antibody penetration was examined by comparing the number of immunolabeled cells at the cut surfaces and that in the remaining immunostained zones. No appreciable differences in the numbers of either sst1- or SRIF-positive SP were observed between the optical sections at the cut surfaces and those at 60 μm from them. The optical sections 2 μm from the cut surfaces were used as look-up sections and those 3–50 μm from the cut surfaces were used as reference sections. In accordance with the disector principle, the SP contained only in the reference optical sections were counted, while those cut through the look-up optical sections were not counted. They were marked in different colors on the montage to avoid double-counting. The numerical density (ND) of immunolabeled cells was calculated by using the formula ND = SP/(A × H/SV), where SP was the number of somatic profiles counted in the reference space, A was the area of the hippocampal region under examination, H was the thickness of reference space, and SV was the volumetric shrinkage factor as calculated in agreement with previous works (see for ref. Jinno and Kosaka 2006). The software package for quantitative image analysis lucia (Nikon Instruments Inc., Melville, NY, USA) was used for counting immunolabeled cell profiles and measuring the area of the hippocampal region under examination. Quantitative analysis of immunolabeled neurons in the stratum oriens of either CA1 or CA3 regions was performed in six slices from three different animals. Values of ND are expressed as mean ± SEM.
After 1 h incubation in aCSF solution, hippocampal slices from either WT or sst1 KO mice were transferred to the recording chamber containing the same solution. Intracellular recordings were performed from the stratum pyramidale of the CA1 region before, during and after the application of the sst1 agonist with or without the sst1 antagonist. In some experiments, the effect of the sst1 agonist on the α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)-mediated post-synaptic response in the presence of tetrodotoxin (TTX) was investigated. In other experiments, the sst4 agonist was also applied either alone or in combination with the sst1 agonist. The interval between substance application and the recording of an observable effect was 6–8 min. Once the substance was washed away, responses generally recovered within 20 min. Electrophysiology studies were carried out blind as to genotype. Intracellular recordings of synaptic activity were performed in current-clamp configuration using sharp glass micropipettes filled with 4 M potassium acetate (tip resistances, 80–120 MΩ). Excitatory post-synaptic potentials (EPSPs) or inhibitory post-synaptic potentials (IPSPs) were evoked by orthodromic stimulation (80 μs stimulus duration, 0.05 Hz frequency) of Schaffer collateral/commissural fibers with a bipolar tungsten electrode placed in the stratum radiatum. Superfusion of aCSF solution containing selective blockers was used to isolate given receptor-mediated-EPSPs or -IPSPs. In particular, we used bicuculline and CGP 55845A to block GABAA and GABAB receptors, dl-2-amino-5-phosphonovalerate (APV) and 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) to block NMDA and AMPA receptors. As variations in membrane potential and input resistance have critical effects on measurements of EPSPs and IPSPs, membrane potential was maintained constant at a value near to a resting potential of at least 60 mV in the absence of a DC holding current. In addition, the bridge balance was monitored carefully throughout the experiments and adjusted when necessary. Finally, to exclude that recorded signals could be contaminated by the activation of voltage-gated channels and/or polysynaptic components, we included in our measurements synaptic potentials whose time to peak was compatible with monosynaptic events. In some experiments, we studied paired-pulse facilitation (PPF) of EPSPs evoked by paired stimuli applied to the Schaffer collateral/commissural fibers at interstimulus intervals of 100 ms using half-maximal stimulus intensity. PPF was calculated by dividing the amplitude of the second EPSP by the first. Post-synaptic potentials were amplified with an Axoclamp-2B amplifier (Axon Instruments, Foster City, CA, USA) and filtered (DC 3–10 kHz). Post-synaptic potentials were averaged from five sweeps and the peak amplitude was measured. In all electrophysiological experiments, traces were stored on a PC for data analysis using the labview software package (National Instruments, Austin, TX, USA).
Following 1 h incubation at 23 ± 2°C in aCSF, hippocampal slices from either WT or sst1 KO mice were placed in Mg2+-free aCSF containing 50 μM 4-aminopyridine (0 Mg2+/4-AP) for 3 h to induce interictal-like activity according to previous studies (Cammalleri et al. 2004, 2006; Ristori et al. 2008a,b). Each sample used for measurements of glutamate release refers to the supernatant of six slices. The rate of glutamate release was measured using HPLC–electrospray Tandem–Mass Spectrometry (ESI–MS/MS) in each sample, according to previous studies (Catalani et al. 2007; Timperio et al. 2007; Cervia et al. 2008). Briefly, the samples were lyophilized and reconstituted in 100 μL of 10% acetonitrile. For chemical dramatization, they were incubated for 30 min in darkness at 60°C in a water bath with 20 μL of 10 M KOH and 500 μL of 0.1 mM dansyl chloride. HPLC was performed using a Beckman Model 126 Programming Solvent Module pump chromatograph (Beckman Instruments, Brea, CA, USA). The chromatographic separations were carried out with ODS2 Spherisorb 250 × 4.6 mm i.d. 5 μm packed column (Waters Corporation, Milford, MA, USA) as stationary phase, preceded by a guard column of Spherisorb RPC18, 5 μm, 4 × 4 mm. The mobile phase was double-distilled water 0.5% formic acid (mobile phase A), and acetonitrile 0.5% formic acid (mobile phase B). ESI–MS was performed on an ion trap mass spectrometer (Esquire 3000 plus; Bruker Daltonik, Bremen, Germany) in the positive mode. For the analysis with pneumatically assisted ESI, an electrospray voltage of 4 kV and a Nebulizer 35 psi and Dry Gas 10 L/min were employed. The temperature of dry gas was set to 300°C. Full-scan spectra were acquired over the range 50–1500 m/z (scan duration, 1 s). After statistics, the raw data (ng/mL) from the same experiment were calculated as a percentage of the glutamate release in aCSF-treated hippocampal slices. Then, data from different experiments were represented and averaged in the same graph.
All data were analyzed by Kolmogorov–Smirnov’s test upon verification of normal distribution. Statistical significance was evaluated using unpaired Student’s t-test or anova followed by Newman–Keuls multiple comparison post-test. In electrophysiological experiments, paired Student’s t-test was also used for paired values, i.e. control versus drug application. The software packages Statsoft (Tulsa, OK, USA) and graphpad Prism (Graph Software, San Diego, CA, USA) were used. All numerical data are expressed as mean ± SEM. Differences with p < 0.05 were considered significant.
In Table 1, SRIF compounds used in this study are classified according to their receptor subtype preference and direction of action (agonists or antagonists). SRIF was purchased from Bachem AG (Bubendorf, Switzerland). CH-275 was purchased from Neosystem (Strasbourg, France). SRA-880 was provided by Novartis Pharma AG (Basel, Switzerland) whereas L-803,087 was provided by Merck Research Laboratories (Rahway, NJ, USA). Where not specified, chemicals and reagents were obtained from Sigma-Aldrich (St Louis, MO, USA). The SRIF compounds were applied at a concentration of 1 μM in line with previous studies using SRIF analogs in the rat or mouse hippocampus (Boehm and Betz 1997; Tallent and Siggins 1997, 1999; Schweitzer et al. 1998; Moneta et al. 2002; Cammalleri et al. 2004, 2006; Qiu et al. 2008).
Table 1. Selectivity of somatotropin release-inhibiting factor compounds used in this study
Ligand binding affinities are expressed as IC50 or Kd (nM), representing published minimal and maximal values. 1000+ indicates > 1000.
In the hippocampus, sst1-IR was observed in the CA1 and CA3 regions (Fig. 1). As shown by higher magnifications (insets in Fig. 1), in the CA1 region, sst1-IR was found to be associated with cell profiles of the stratum pyramidale, fibers of the stratum radiatum and neuropil of the stratum lacunosum moleculare. In the CA3 region, dense immunostaining was present in the stratum pyramidale and in the dendritic fields of the stratum lucidum. In both CA1 and CA3 regions, sst1-IR was localized to some cell somata in the stratum oriens in which IR appeared to be associated mostly with cytoplasmic compartments; sst1-IR was weak in the granular cell layer of the dentate gyrus (DG).
Somatostatinergic interneurons were found in the CA1 and CA3 regions and in the hilus of the DG (Fig. 2). As shown by higher magnifications (insets in Fig. 2), in both CA1 and CA3, dendritic trees and somata of SRIF-positive neurons were present in the stratum oriens whereas immunolabeled fibers were found in the stratum radiatum and the stratum lacunosum moleculare. Some SRIF-immunoreactive cells were also found in the stratum pyramidale and the stratum radiatum of both CA1 and CA3. Numerous SRIF-positive cells were observed in the hilus of the DG with fibers arborizing in the molecular layer of the DG.
Double-labeling experiments performed to analyze sst1 expression in somatostatinergic interneurons of the mouse hippocampus demonstrated a colocalization between SRIF and sst1 in the stratum oriens of both CA1 and CA3 (Fig. 3) and in rare hilar cells. In particular, disector-counted SRIF-immunoreactive neurons in the stratum oriens were 491 in CA1 and 151 in CA3 while the sst1-containing neurons were 156 and 86 in CA1 and CA3, respectively. The ND of SRIF-positive neurons was 3.94 ± 0.12 × 103/mm3 in CA1 and 1.54 ± 0.03 × 103/mm3 in CA3 in line with previous results (Jinno and Kosaka 2006), whereas the ND of sst1-positive cells was 1.25 ± 0.03 × 103/mm3 in CA1 and 0.88 ± 0.02 × 103/mm3 in CA3. Of disector-counted neurons, those positive for both sst1 and SRIF were 118 in CA1 and 74 in CA3. Thus, the percentage of the SRIF-positive cells displaying sst1-IR was about 24% and 49% in CA1 and CA3, respectively, while the percentage of sst1-immunostained cells that also expressed SRIF-IR was about 75% and 86% in CA1 and CA3, respectively.
Effects of sst1 on synaptic transmission and its interaction with sst4
We examined the effects of CH-275 on pharmacologically isolated post-synaptic potentials recorded from CA1 pyramidal cells in aCSF-treated hippocampal slices. CH-275 was used at 1 μM, a concentration which was previously found to reduce the frequency of spontaneous epileptiform discharge by about 39% (Cammalleri et al. 2004). Recorded cells had a mean resting membrane potential of 66.4 ± 1.3 mV and a mean input resistance of 43.0 ± 15.0 MΩ. In all cells tested, bath application of drugs does not affect either the resting membrane potential or the input resistance as determined by the voltage change in response to a constant current pulse (0.2 nA–200 ms, data not shown). We isolated the NMDA- and the AMPA-EPSP components by application of 30 μM CNQX or 30 μM APV, respectively (together with 30 μM bicuculline and 1 μM CGP 55845A to block GABAA and GABAB receptors). Isolated post-synaptic potentials were blocked by superfusion with 30 μM APV, indicating that they were mediated by NMDA receptors, or with 30 μM CNQX, indicating that they were mediated by AMPA receptors. CH-275 significantly depressed the peak amplitude of both NMDA-EPSPs (eight cells; from 5.6 ± 0.5 to 3.8 ± 0.5 mV; p < 0.05; Fig. 4a) and AMPA-EPSPs (eight cells; from 6.1 ± 0.3 to 4.3 ± 0.5 mV; p < 0.05; Fig. 4b), and this effect was reversible on washout. Specificity of CH-275 effects on NMDA- and AMPA-EPSPs was confirmed by the finding that CH-275 effects were not observed in hippocampal slices of sst1 KO mice (six cells; from 5.3 ± 0.3 to 5.4 ± 0.4 mV for NMDA-EPSPs and from 5.7 ± 0.4 to 6.2 ± 0.3 mV for AMPA-EPSPs; insets in Fig. 4a and b) or when CH-275 was applied in combination with 1 μΜ SRA-880, a concentration known to antagonize CH-275 effects in hippocampal models of interictal-like activity (Cammalleri et al. 2004). Neither SRA-880 alone (eight cells; from 5.5 ± 0.6 to 5.8 ± 0.4 mV for NMDA-EPSPs and from 5.7 ± 0.4 to 5.8 ± 0.6 mV for AMPA-EPSPs; Fig. 4c and d) nor SRA-880 in combination with CH-275 (eight cells; from 5.4 ± 0.3 to 5.7 ± 0.4 mV for NMDA-EPSPs and from 5.8 ± 0.3 to 5.6 ± 0.4 mV for AMPA-EPSPs; Fig. 4e and f) were found to affect excitatory transmission in the mouse hippocampus.
No significant effects on NMDA-mediated responses were observed after the application of 1 μΜ L-803,087 to activate sst4 (six cells; from 5.3 ± 0.4 to 5.7 ± 0.5 mV; Fig. 4g) which, in contrast, significantly increased the peak amplitude of the AMPA-mediated EPSPs (six cells; from 5.8 ± 0.3 to 7.4 ± 0.3 mV; p < 0.05; Fig. 4h). The CH-275-induced inhibition of the NMDA-mediated EPSPs was not influenced by the combined application of CH-275 and L-803,087 (six cells; from 4.8 ± 0.4 to 3.4 ± 0.3 mV; p < 0.05; Fig. 4i), whereas neither the CH-275-induced inhibition nor the L-803,087-induced increase of the AMPA-mediated responses were observed after the combined application of CH-275 and L-803,087 (six cells; from 5.5 ± 05 to 5.7 ± 0.6 mV; Fig. 4j).
In six different cells, the effects of CH-275 on GABAA-IPSPs were also investigated. These potentials were isolated in the presence of 30 μM APV, 30 μM CNQX, and 1 μM CGP 55845A to block NMDA, AMPA, and GABAB receptors, respectively. The IPSPs were blocked by application of 30 μM bicuculline, confirming that they were indeed mediated by GABAA receptors. The application of CH-275 did not affect significantly the amplitude of GABAA-IPSPs (from 4.8 ± 0.3 to 5.0 ± 0.5 mV; Fig. 5). In contrast, the application of L-803,087 was found to significantly increase the amplitude of GABAA-IPSPs (six cells; from 5.3 ± 0.4 to 6.7 ± 0.6 mV; Fig. 5). In hippocampal slices of sst1 KO mice, the amplitude of GABAA-IPSPs was 5.2 ± 04 mV (six cells), a value not significantly different from that measured in hippocampal slices of WT mice. In sst1 KO slices, the L-803,087-induced increase in the amplitude of GABAA-IPSPs was not significantly different from that measured in WT slices (six cells; from 5.3 ± 0.4 to 6.7 ± 0.6 mV and from 5.2 ± 0.4 to 6.8 ± 0.4 mV in WT and sst1 KO slices, respectively; p < 0.05; inset in Fig. 5).
The sst1 effects on paired-pulse facilitation and post-synaptic responsiveness to AMPA in the presence of TTX
The finding that sst1 activation affects both NMDA- and AMPA-EPSPs suggests the possibility that pre-synaptic mechanisms may mediate inhibitory action of sst1. This was examined using the technique of PPF. This technique involves activating the excitatory afferents to the central neurons twice with a short interval between each stimulus. The response to the second stimulus is generally facilitated in relation to the initial stimulus. PPF is attributed to an increase in the amount of transmitter release to the second stimulus (Zucker 1989). On the other hand, manipulations in pre-synaptic transmitter release may result in a change in the magnitude of PPF. If the CH-275-mediated synaptic inhibition involved a pre-synaptic mechanism of action, it would be associated with an increase in the magnitude of PPF. Alternatively, if CH-275 reduced synaptic transmission via another type of mechanism (e.g. reducing the sensitivity of post-synaptic receptors), then the PPF magnitude should be relatively unaffected. We calculated the PPF ratio as the amplitude of the second EPSP divided by the amplitude of the first EPSP. As shown in Fig. 6a, activation of sst1 increased the PPF ratio from 1.9 ± 0.03 to 2.3 ± 0.04 on average by 21.3 ± 1.5% (eight cells, p < 0.05). To determine if the reduction of the EPSPs in the presence of CH-275 was also partially due to direct modulation of the sensitivity of the post-synaptic glutamatergic receptor, we investigated whether the sst1 agonist would affect the response to an exogenous application of the glutamate receptor agonist AMPA. These experiments were done in the presence of TTX (0.5 μM) to suppress action potential generation. As shown in Fig. 6b, application of AMPA (2 μM) produced a profound membrane depolarization in the hippocampal neurons. CH-275 pre-treatment for 30 min did not significantly affect the AMPA-induced membrane depolarization. These data indicate that the inhibition of glutamatergic synaptic transmission induced by CH-275 in the hippocampal neurons is not mediated by a change in post-synaptic sensitivity to glutamate.
The sst1 effects on glutamate release induced by epileptiform treatment
We examined whether the CH-275-induced decrease in synaptically evoked responses could result from a decrease in pre-synaptic transmitter release. To this aim the effect of CH-275 on glutamate release induced by 0 Mg2+/4-AP treatment was investigated. Glutamate release in hippocampal slices incubated in aCSF (control conditions) from either WT and sst1 KO hippocampal slices was 56.7 ± 1.5 and 51.6 ± 2.1 ng/mL, respectively, both values which are in the range of those previously reported in the mouse hippocampus (Mazarati et al. 2000; Zhu et al. 2006). As shown in Fig. 7, an almost threefold increase in glutamate release was observed after 3 h incubation in 0 Mg2+/4-AP, when ictal discharges are replaced by continuous interictal events. No significant differences were found between values of glutamate release measured in WT and sst1 hippocampal slices after 0 Mg2+/4-AP. CH-275 pre-treatment for 30 min at 1 μM, significantly decreased glutamate release in 0 Mg2+/4-AP-treated hippocampal slices by 43.0 ± 1.2% (p < 0.01). The CH-275-induced decrease in glutamate release was not significantly different from that induced by the application of SRIF for 30 min at 1 μM, a concentration previously reported to inhibit hippocampal activity in rodents (see for ref. Tallent and Qiu 2008). Specificity of CH-275 effects on glutamate release induced by epileptiform treatment was confirmed by the finding that the CH-275-induced inhibition of glutamate release was not observed in hippocampal slices of sst1 KO mice (inset in Fig. 7).
Previous results from acute hippocampal models of interictal-like activity suggested that sst1 may be involved in the control of bursting frequency (Cammalleri et al. 2004) although the precise contribution of sst1 in hippocampal excitatory transmission remained to be elucidated.
Here, we report findings which demonstrate that sst1 is localized to the mouse hippocampus where it plays an important role in the control of excitatory transmission.
Immunolocalization of sst1
There is general agreement about the presence of sst1 mRNA in the rodent hippocampus (Pérez et al. 1994; Hannon et al. 2002; Cammalleri et al. 2004, 2006), whereas the presence of sst1 protein is still debated (Hervieu and Emson 1998; Schulz et al. 2000; Kwak et al. 2008; Spary et al. 2008; Rajput et al. 2009). The present results demonstrate that sst1-IR is indeed expressed in distinct regions of the mouse hippocampus in agreement with recent findings in the mouse hippocampus (Rajput et al. 2009). The reason of this discrepancy may be ascribed either to different antibodies used in the previous and in this study and/or to species-dependent expression. Indeed, the use of sst1 KO hippocampal slices confirmed the specificity of the signals obtained with the sst1 antibody used in this study. As shown by our results, the great majority of sst1-IR was found to be associated to pyramidal cells in both CA1 and CA3, indicating the post-synaptic localization of sst1 and suggesting the possibility that it may cooperate with other SRIF receptors to mediate SRIF actions on hippocampal activity. Dense sst1 immunostaining was also observed in the neuropilar zone of the stratum lacunosum moleculare, where synaptic contacts between Schaeffer/commissural fibers and pyramidal cell dendrites are localized, and in the stratum lucidum, where the mossy fiber pathway terminates. These findings suggest the possibility that sst1 is expressed by pre-synaptic glutamatergic terminals, in line with previous results of the rat brain showing that sst1 is primarily confined to varicose axons and indicating a pre-synaptic role of this receptor (Schulz et al. 2000). As shown by our quantitative analysis, a discrete number of sst1 immunolabeled cells is present in the stratum oriens of both CA1 and CA3 which also contains numerous SRIF-immunoreactive cells. The presence of SRIF-immunoreactive cells in the rodent hippocampus has been previously demonstrated (Buckmaster et al. 2002; Jinno and Kosaka 2004; see for ref. Viollet et al. 2008), although our study is the first detailed demonstration of a regional localization of SRIF-containing profiles. Our finding that a high percentage of sst1 immunoreactive cells in the stratum oriens also contain SRIF indicates that sst1 is predominantly targeted to the pre-synaptic compartment and, hence, in a position to modulate the release of SRIF itself or of other neurotransmitters.
The sst1 localization to SRIF-containing cells has been demonstrated in different structures of the rodent brain in which sst1 appears to act as an inhibitory autoreceptor (see for ref. Thermos et al. 2006).
Inhibitory effects of sst1 on excitatory transmission
Inhibitory actions of SRIF on excitatory synaptic transmission have been shown in the rodent hippocampus (Tallent and Siggins 1997; Cammalleri et al. 2004). In addition, previous results in the mouse hippocampus demonstrated that activation of sst1 reduces epileptiform discharges in a model of 0 Mg2+/4-AP-induced interictal-like discharge (Cammalleri et al. 2004). In this model, the sst1 agonist CH-275 was found to significantly reduce the frequency of spontaneous bursting by about 39% and this inhibition was not significantly different from that obtained after SRIF application. Epileptiform activity was not affected by the application of either the sst2-preferring agonist octreotide or the sst3 agonist L-796,778 whereas it was increased by the sst4 agonist L-803,087. These results were in agreement with those obtained by Moneta et al. (2002) using EEG recordings after kainic acid injection in rodents. Recently, however, Qiu et al. (2008) using specific SRIF receptor KO mice and selective pharmacological tools excluded a role of sst1 in mediating the antiepileptic actions of SRIF in the mouse hippocampus. Indeed, they demonstrated a major involvement of sst4 although a contribution of sst2 and sst3 was also found. As shown by this results, in the CA1 region of the mouse hippocampus, sst1 activation reduces the pharmacologically isolated NMDA and AMPA receptor-mediated EPSPs without affecting the amplitude of GABAergic IPSPs. Our finding that sst4 activation increases the amplitude of the AMPA-mediated responses without influencing the NMDA-mediated excitatory transmission is in line with previous results demonstrating that sst4 agonist increases either epileptiform bursting or field EPSPs slopes in the mouse hippocampus (Moneta et al. 2002; Cammalleri et al. 2006). In addition, the fact that the increase of the AMPA-mediated responses induced by sst4 activation cannot be observed in the presence of sst1 agonist and the additional result that the inhibition of the AMPA-mediated excitatory transmission induced by sst1 activation can be prevented by sst4 agonist indicate functional interactions between sst1 and sst4, either directly at the receptor level or indirectly as a crosstalk between their signaling pathways. Functional interactions between distinct SRIF receptors have been demonstrated in native systems including the mouse hippocampus. For instance, selective stimulation of sst4 enhances AMPA-mediated currents in the hippocampus of WT mice while this effect is lost in sst2 KO mice indicating functional interactions between sst2 and sst4 (Moneta et al. 2002). Interactions between sst1 and sst2 have been also shown in the mouse hippocampus as epileptiform bursting can be increased by sst2 blockade in sst1 KO but not in WT mice (Cammalleri et al. 2004). Recently, functional interactions between sst2 and sst4, which are not mediated by direct receptor coupling, have been demonstrated to underlie memory formation in the mouse hippocampus (Gastambide et al. 2009).
As shown by the present results, sst4 activation with L-803,087 increases the amplitude of inhibitory responses in line with the finding that SRIF increases the amplitude of inhibitory transmission in the mouse hippocampus (Cammalleri et al. 2006). Thus, SRIF inhibitory action on hippocampal activity may also involve an sst4-induced increase of GABAA-mediated inhibitory transmission in addition to an sst1-mediated inhibition of excitatory responses. This would confirm a role of sst4 in mediating anticonvulsant properties of SRIF in the mouse hippocampus (Qiu et al. 2008) although excitatory actions of sst4 have been reported previously (Moneta et al. 2002; Cammalleri et al. 2004).
Pre-synaptic mechanisms mediating sst1 action on excitatory transmission
A decrease in synaptically evoked responses could result from either a decrease in pre-synaptic transmitter release or a decrease in post-synaptic responsiveness to the transmitter. Our finding that sst1 activation with CH-275 inhibits both NMDA and AMPA glutamatergic EPSPs suggests a pre-synaptic effect of the sst1 agonist. To determine whether CH-275 effect could have a pre-synaptic origin, we used the paired-pulse paradigm (Betz 1970). Modification of the PPF ratio was found specific for pre-synaptic inhibition in various preparations of the brain (see for ref. Manita et al. 2007). In this study, the significant increase of the PPF ratio in the presence of the sst1 agonist further supports the pre-synaptic modulation of the evoked-EPSPs in the mouse hippocampus. A pre-synaptic mechanism of action of CH-275 has also been suggested by our finding that CH-275 does not influence the AMPA-induced depolarization in the presence of TTX. This observation suggests that action potentials, and thus pre-synaptic depolarization, are required for sst1 to exert its effect.
An inhibitory action of SRIF at pre-synaptic sst1 should involve an sst1-mediated decrease of the probability of glutamate release from afferent pathways. As shown by the present results, basal glutamate release increases significantly in response to 0 Mg2+/4-AP in line with clinical evidence and data from animal models indicating that impairments in glutamatergic transmission may play a key role in the regulation of epileptic seizures (Li et al. 2000; Ferraz et al. 2002; Eid et al. 2008). As also shown by the present results the 0 Mg2+/4-AP-induced glutamate release is almost halved by sst1 activation. The additional finding that CH-275 effects on glutamate release are absent in hippocampal slices with genetic deletion of sst1 confirm the specificity of CH-275 action. In agreement with the present results, sst1 has been found to decrease pre-synaptic glutamate release in cholinergic neurons of the rat basal forebrain (Momiyama and Zaborszky 2006); sst1 modulation of glutamate release is likely to represent an important contribution to the regulation of the glutamatergic transmission in the hippocampus. In particular, abnormalities in the endogenous mechanisms regulating pre-synaptic glutamate release may play a major role in the pathogenesis of epilepsy. In this respect, by limiting the amount of glutamate available to glutamate receptors, sst1 is likely to exert an important neuroprotective function against glutamate neurotoxicity that characterizes seizures. The fact that the 0 Mg2+/4-AP-induced glutamate release is not affected by sst1 deletion is in apparent contrast with the finding that epileptiform discharge in sst1 KO hippocampal slices is significantly lower than in WT slices (Cammalleri et al. 2004) and seems to exclude the possibility that sst1 basal activity may render hippocampal slices less susceptible to epileptiform activity by interfering with glutamatergic transmission. This would be confirmed by the additional finding that the amplitude of inhibitory responses does not differ between WT and sst1 KO slices. On the other hand, we have previously demonstrated that antagonizing sst1 substantially increases epileptiform discharge in the hippocampus of WT mice thus in contrast with the finding of a decreased hippocampal activity following sst1 deletion.
Numerous evidence indicate that SRIF inhibits glutamatergic transmission in different structures of the rodent brain including the mouse hypothalamus (Lanneau et al. 1998), the rat hippocampus (Boehm and Betz 1997), striatum (Arnerićet al. 1986) and the mouse retina (Dal Monte et al. 2003). The present findings confirm that SRIF drastically inhibits the 0 Mg2+/4-AP-induced increase in glutamate release in the mouse hippocampus and demonstrate that the sst1 agonist reduces the evoked release of glutamate with equipotency to SRIF indicating that SRIF controls glutamate release through sst1 and this control may represent part of a mechanism by which SRIF regulates glutamate concentration in the hippocampus.
In summary, our results concur to demonstrate that sst1 is present in the mouse hippocampus in which it participates in mediating SRIF regulation of glutamatergic transmission possibly interacting with sst4. As shown by our functional studies, sst1-mediated inhibition of excitatory transmission involves pre-synaptic mechanisms although post-synaptic functions of sst1 are also likely to be involved. Thus, targeting sst1, which has limited distribution in the brain, could lead to development of novel antiepileptic drugs. However, a potential for sst1 ligands as antiepileptics needs more work to obtain information on precise contribution of sst1 in mediating anticonvulsant effects of SRIF also in respect to better understanding the mechanisms underlying functional interaction between sst1 and sst4.
This work has been supported by the University of Pisa (‘60% Grant’ to PB). We wish to thank G. Casini for critical reading of the manuscript. We are grateful to M. Fagioni (Department of Environmental Sciences, University of Tuscia) for the excellent assistance in the HPLC–ESI–MS assay. We also thank G. Bertolini for assistance with mouse colonies at the University of Pisa.