Somatostatin Depresses Excitatory but not InhibitoryNeurotransmission in Rat CA1 HippocampusMELANIE K. TALLENT AND GEORGE R. SIGGINSDepartment of Neuropharmacology, The Scripps Research Institute, La Jolla, California 92037Tallent, Melanie K. and George R. Siggins. Somatostatin de-aminobutyric acid (GABA) (Esclapez and Houser 1995;presses excitatory but not inhibitory neurotransmission in rat CA1Feldman et al. 1982; Kohler et al. 1987; Kosaka et al. 1988;hippocampus. J. Neurophysiol. 78: 3008–3018, 1997. In rat CA1Morrison et al. 1982). SST receptors are also discretelyhippocampal pyramidal neurons (HPNs), somatostatin (SST) haslocalized in the hippocampus, with four of the five clonedinhibitory postsynaptic actions, including hyperpolarization of theSST receptors (sst1–4) expressed (sst5shows no significantmembrane at rest and augmentation of the K/M-current. However,expression anywhere in the brain). In situ hybridizationthe effects of SST on synaptic transmission in this brain regionshows that the mRNAs for sst1–4occur primarily in principalhave not been well-characterized. Therefore we used intracellularcells of the CA1, CA3, dentate gyrus, and subiculum, withvoltage-clamp recordings in rat hippocampal slices to assess thelittle expression in the stratum oriens, stratum radiatum, oreffects of SST on pharmacologically isolated synaptic currents inHPNs. SST depressed both (R,S)-a-amino-3-hydroxy-5-methyl-molecular layers (Breder et al. 1992; Kong et al. 1994; Perezisoxazole-4-propionic acid (AMPA)/kainate and N-methyl-D-et al. 1994; Thoss et al. 1996). An immunocytochemicalaspartate (NMDA) receptor-mediated excitatory postsynaptic cur-study using antibodies selective for sst2showed that therents (EPSCs) in a reversible manner, with an apparent IC50of 22receptor protein is expressed largely in the soma as well asnM and a maximal effect at 100 nM. In contrast, SST at concentra-the basal and apical dendrites of CA1 hippocampal pyrami-tions up to 5mM had no direct effects on eitherg-aminobutyricdal neurons (HPNs), with no expression in CA3 (Dournaudacid-A (GABAA) or GABABreceptor–mediated inhibitory post-et al. 1996). Interestingly, autoradiographic studies usingsynaptic currents (IPSCs). The depression of EPSCs by SST wasiodinated SST ligands show little binding in the principalespecially robust during hyperexcited states when polysynapticcell layers of CA1 and CA3; instead binding is limited toEPSCs were present, suggesting that this peptide could play athe plexiform layers, mostly the stratum oriens of CA1 andcompensatory role during seizurelike activity. SST effects weregreatly attenuated by the alkylating agent N-ethylmaleimide, thusCA3, the hilus, and the subiculum (Holloway et al. 1996;implicating a transduction mechanism involving the Gi/GofamilyLeroux et al. 1993; Perez et al. 1995). These studies suggestof G-proteins. Use of 2 M Cs/in the recording electrode blockedthat SST receptor mRNAs are mostly exported out of thethe postsynaptic modulation of K/currents by SST, but did notsoma and into dendritic layers, where they would be morealter the effects of SST on EPSCs, indicating that postsynaptic K/proximate to terminals of SSTergic interneurons.currents are not involved in this action of SST. However, 2 mMThe majority of studies examining the effects of SST inexternal Ba2/blocked the effect of SST on EPSCs, suggesting thathippocampus have been done in CA1. Much is known aboutpresynaptic K/channels or other presynaptic mechanisms may bepostsynaptic effects of SST on HPNs. SST was first reportedinvolved. These findings and previous results from our laboratoryto depolarize CA1 HPNs (Dodd and Kelly 1978) but wasshow that SST has multiple inhibitory effects in hippocampus.later found to hyperpolarize these neurons (Pittman and Sig-gins 1981). Subsequent studies demonstrated that SST aug-INTRODUCTIONments the voltage-sensitive K/M-current in these cells(Moore et al. 1988), an action mediated by an arachidonicSomatostatin (SST) is a peptide widely distributedacid metabolite (Schweitzer et al. 1990, 1993). SST maythroughout the periphery and brain. SST was originally char-also augment a voltage-insensitive K/leak current in HPNsacterized as an inhibitor of growth hormone release from(P. Schweitzer and G. R. Siggins, unpublished observa-the anterior pituitary (Brazeau et al. 1972), and SST functiontions), a mechanism that would account for much of theas a hormone in the periphery is well-established (ReisineSST hyperpolarizing effect at rest. Although earlier studies1995). By contrast, the role of SST in the extrahypothalamicon hippocampal slices showed no effect of SST on Ca2/brain is less clear. This peptide is abundant in nearly allcurrents in HPNs (Schweitzer et al. 1993), recent studiesbrain regions (Crawley 1985) and is thought to play a neuro-in acutely isolated HPNs showed that SST inhibits anmodulatory role (Epelbaum 1986). However, studies exam-N-type Ca2/current in these cells (Ishibashi and Akaikeining the specific actions of SST on neurotransmission have1995). At the cellular level, these postsynaptic actions ofbeen sparse, and determining the endogenous function ofSST on ionic currents are inhibitory, in that they decreasethe peptide has been hampered by the lack of receptor antag-the probability of the neuron firing an action potential. Theseonists.data suggest that SST acts similarly to GABA, with whichHigh expression levels of SST and its receptors are foundit is colocalized (Crawley 1990; Kosaka et al. 1988).in the hippocampus. Extensive SST immunoreactivity isThe actions of SST on CA1 synaptic transmission arefound in extrinsic fibers and interneurons of the stratumless clear. Two studies found that SST attenuates inhibitorypostsynaptic potentials (IPSPs) in rabbit (Scharfman andoriens, hilus, and subiculum, where it is colocalized withg-3008 0022-3077/97 $5.00 Copyrightq1997 The American Physiological SocietyJ362-7/ 9k22$$de38 11-10-97 07:27:03 neupa LP-NeurophysSOMATOSTATIN DEPRESSES EXCITATORY NEUROTRANSMISSION 3009The area of the PSC reflects the total charge crossing the membraneSchwartzkroin 1989) and guinea pig (Xie and Sastry 1992)and thus the synaptic strength. Current traces shown for representa-HPNs. This reduction of inhibitory input would result in ative cells were normalized to a common baseline for easier compar-net excitatory effect of SST that contrasts with its directison of amplitudes. We usually voltage clamped the