Additional related information may be found at:
Neuropsychopharmacology: The Fifth Generation of Progress

Back to Psychopharmacology - The Fourth Generation of Progress

Somatostatin in the Central Nervous System

David R. Rubinow, Candace L. Davis, and Robert M. Post


Early efforts to screen hypothalamic extracts for growth hormone (GH)-releasing activity inadvertently resulted in the discovery in 1968 of an inhibitor of GH release, the biological activity of which was localized to the median eminence and anterior hypothalamus. This inhibiting factor, somatostatin (SS, SRIF), was subsequently isolated and structurally characterized in 1973 and was found to be identical to an insulin inhibitor described in 1969. The discovery of SS, for which Guillemin received the Nobel prize in 1977, established both (a) the dual regulation of anterior pituitary peptides by hypothalamic releasing and release-inhibiting factors and (b) the existence of "brain-gut" peptides, present in both the central nervous system (CNS) and gastrointestinal system.

While the anatomical distribution of SS is widespread, it is particularly concentrated in the central and peripheral nervous systems and in the gastrointestinal (GI) tract (including the pancreas), the major source of circulatory SS. In the GI tract, SS is localized in the D cells of the stomach and islets of Langerhans, the small intestine, and the vagal neurons of the myenteric and submucosal plexuses. SS secreted from these sites acts both locally (paracrine) and distally (endocrine) to modulate nutrient homeostasis through regulation of endocrine and exocrine secretions of the stomach, intestine, and pancreas as well as regulation of motor activity of the stomach and intestine. As described below, SS is widely but discretely distributed throughout the CNS, where its range of neuroregulatory effects has expanded far beyond its originally discovered role.


While the originally identified hypothalamic GH release inhibitor is a tetradecapeptide (SS-14), SS now refers to a family of peptides that are cleavage products of a larger precursor, prepro-somatostatin (116 amino acids) (35). After removal of a leader sequence, the 92-amino-acid prohormone (pro-SS) is cleaved at monobasic or dibasic sites to yield the different molecular forms of SS (see Fig. 1). SS-14 is the C-terminal fragment of the prohormone and may be directly cleaved from pro-SS or from the SS-14 N-terminal extension fragment, SS-28, the other major bioactive product (see Fig. 2). SS-14 and SS-28 are secreted in different amounts by SS-containing cells and appear in a variety of proportions in different tissues. SS-14 is the predominant form appearing in neural tissue and the exclusive form in the retina and peripheral nerves, pancreas, and stomach. The mucosal cells of the gut secrete mainly SS-28, which represents 20–30% of SS-like immunoreactivity in the brain (35).


SS-containing nerve fibers are densely but selectively distributed throughout the brain, appearing in the median eminence (the source of anterior pituitary regulation), the posterior pituitary, the limbic system, the cortex, and the hypothalamus and hypothalamic projections to the brainstem and spinal cord. SS-containing cell bodies have similarly been identified in diverse brain regions including the neocortex, all limbic structures, the hypothalamus (particularly the anterior periventricular region but also the paraventricular, arcuate, and ventromedial nuclei), the striatum, the periaqueductal gray, the nucleus accumbens, the locus coeruleus, and the septal nuclei (48). SS neurons also appear at the synapses of the major sensory systems (somatosensory, visual, auditory, and olfactory) (35). While SS is most concentrated in the hypothalamus, it is most abundant in the cortex, which accounts for approximately 49% of brain SS. Spinal cord (30%), brainstem (12%), hypothalamus (7%), olfactory lobe (1%), and cerebellum (1%) represent the other major SS-containing brain regions (35). Postmortem studies in humans have revealed a distribution of SS neurons and fibers similar (although not identical) to that seen in rodents. Several CNS SS pathways have been identified (48). The periventricular nucleus is the primary source of somatostinergic input to the median eminence, while the SS-containing nerve terminals of the anterior hypothalamic nuclei and lateral and ventromedial hypothalamus originate in the amygdala. Periventricular hypothalamic somatostatinergic fibers travel in a variety of pathways, including short-distance projections to hypothalamic nuclei (e.g., suprachiasmatic, arcuate, and preoptic nuclei), long-distance rostral and ascending projections to many limbic structures (e.g., stria terminalis, amygdala, and arcuate nucleus), and caudal projections into the brainstem and spinal cord.


In parallel with its widespread distribution, SS displays multiple regulatory effects in a variety of tissues. The general processes that are regulated are neurotransmission, glandular secretion, smooth muscle contractility, and cell proliferation (35). Electrophysiologic studies have demonstrated both inhibitory and excitatory effects of SS on neuronal activity. These ostensibly contradictory findings may reflect regional specificities as well as dose–response and time-course characteristics. Intracellular recording studies have revealed a biphasic dose-related neuronal response to SS, with low doses increasing and high doses inhibiting, or not affecting, action potential generation (48). More recent studies have consistently demonstrated the ability of SS to hyperpolarize CA1 pyramidal and solitary tract neurons, with both pharmacologic and voltage-clamp evidence for SS-induced m-current (a voltage-dependent outward potassium conductance) activation as the mechanism for this hyperpolarization (48). Stimulatory and inhibitory effects of SS have been observed in the cortex [increased dopamine (DA), norepinephrine (NE); decreased histamine (HS)], hippocampus [increased acetylcholine (ACh), DA, serotonin (5-HT); decreased HS], striatum (increased DA, decreased glutamate), and hypothalamus (decreased NE, HS; increased 5-HT) (16, 48). The effects of SS on the secretion of an array of hormones and neuropeptides is so uniformly inhibitory that McCann suggested that SS be called "panhibin" (32). SS is clearly involved in the physiologic regulation of GH and thyroid-stimulating hormone (TSH), inhibiting both their basal and stimulated secretion (38). While SS has no effect on luteinizing hormone (LH), follicle-stimulating hormone (FSH), prolactin, or adrenocorticotropic hormone (ACTH) in normal subjects, contradicting reports exist regarding SS-induced suppression of elevated ACTH levels in patients with Addison's disease or Nelson's syndrome (3, 35). While the reported ability of SS to suppress stimulated ACTH secretion from mouse pituitary tumor (AtT-20) cells was not confirmed in pituitary cell culture, intracerebroventricular (i.c.v.) SS-28 was noted to inhibit stress-induced corticotropin-releasing factor (CRF) secretion in rats (48). Furthermore, SS infusion in humans was observed to blunt insulin-induced hypoglycemia-stimulated elevations of b-endorphin, b-lipotropin, and cortisol (48). Similarly, while SS has no effect on basal prolactin levels in normal subjects, it decreases elevated prolactin levels in acromegaly and inhibits both basal and TRH-stimulated prolactin secretion in vitro and in an estrogen-dependent fashion in rats (16, 64). All of the diverse actions of SS appear to be mediated by SS binding to high-affinity membrane-bound receptors.


The existence of SS receptor subtypes was suggested (6) on the basis of different patterns of desensitization after exposure to SS or SS analogues (homologous desensitization). Further support for receptor heterogeneity was derived from observations of pharmacologically distinct receptors displaying different affinities for SS agonists, regional distributions, functions, and linkage to transduction systems (see ref. 6 for review). The SRIF1 receptor is a G-protein-coupled receptor that is expressed in high concentration in the dentate gyrus of the hippocampus, neostriatum, locus coeruleus, and inner layers of the cerebral cortex and is highly sensitive to homologous desensitization (6). In contrast, the SRIF2 receptor is not coupled to G proteins, is localized in the CA1 region of the hippocampus and diffusely throughout the cerebral cortex, and is resistant to homologous desensitization (6). As reviewed by Bell and Reisine (6), three SS receptors have recently been cloned and sequenced. The three receptors, SSTR1, SSTR2, and SSTR3, consist of 391, 369, and 428 amino acids, respectively, and display approximately 50% homology. Like other G-protein-coupled receptors, they appear to have seven membrane-spanning segments, with the transmembrane regions showing the most extensive sequence identity. The receptor sequences are highly conserved across species: human and mouse SSTR1, SSTR2, and SSTR3 show 99%, 94%, and 85% amino acid identity, respectively. SSTR1 messenger ribonucleic acid (mRNA) is expressed in high levels in the human GI tract, but not in the cerebral cortex, despite the presence of high levels of transcripts in the mouse and rat brain. In contrast, high levels of SSTR2 transcripts are present in human brain and kidney, and SSTR3 transcripts are present only in the brain. In situ hybridization studies confirm the expression of all three receptor transcripts in mouse brain, displaying distinct but overlapping distributions, particularly in the cortex, hippocampus, and amygdala (6).

In their selective affinities for SS analogues and in their expression patterns in mouse brain, SSTR1 and SSTR2 resemble the SRIF2 and SRIF1 receptors, respectively (10, 30). SSTR3 displays a different pattern of pharmacologic properties and thus is not explained by the original classification. The differential effects of SS-14 and SS-28 on target cells (16) (e.g., SS-28 more potently inhibits hypothalamic CRF, GH, and TSH, while SS-14 more effectively inhibits electrical activity in the cortex) (35!popup(ch53ref35)), as well as tissue-specific differences in binding potency (SS-14 greater in brain, SS-28 greater in pituitary) and distinct labeling patterns in brain, lent empiric support to the early notion of receptor heterogeneity (35, 36). While SSTR1, SSTR2, and SSTR3 bind SS-14 and SS-28 with similar high affinities, a fourth recently cloned receptor, SSTR4, has higher affinity for SS-28 (6). Even more recently, the gene for a fifth human receptor, SSTR5, has been localized to chromosome 20, in contrast to the localization of SSTR1, SSTR2, and SSTR3 to human chromosomes 14, 17, and 22, respectively (63). Many of the known and proposed mechanisms of action of SS can be linked to and mediated by distinct properties of the five cloned receptors.

Ligand binding to SS receptors modulates one of five different cellular effector systems: adenylyl cyclase, K+ channels, Ca2+ channels, exocytosis, and tyrosine phosphatase (6). Activation of the SS receptor decreases intracellular cyclic adenosine monophosphate (cAMP) by inhibiting adenylyl cyclase (through coupling with Gia1) and decreases intracellular Ca2+ by activating the voltage-dependent outward potassium conductance or m current (possibly through Gia3) and by inhibiting Ca2+ channel activity (possibly through Goa) (24, 55). SS also blocks hormone secretion at a step distal to either cAMP formation or Ca2+ mobilization as evidenced by the ability of SS to block stimulated secretion by cAMP, adenylate cyclase stimulation or phosphodiesterase inhibition, inositol triphosphate or diacylglycerol, or elevation of intracellular Ca2+ with calcium ionophores (35). This action of exocytosis inhibition also appears to involve coupling to a G protein (36). The mechanism by which SS stimulates tyrosine phosphatase is currently unclear, although the effect is believed to underlie an observed antiproliferative effect on pancreatic and breast cancer (6, 54, 61).

As reviewed by Bell and Reisine (6), only SSTR3 is coupled to Gia1 and inhibits adenyl cyclase activity. (It is not established whether SSTR4 is linked to G protein, although it, too, appears to mediate SS inhibition of cAMP generation.) SSTR2 is coupled to Gia3 and Goa2, through which it may alter K+ and Ca2+ channel activity, respectively. The actions of SSTR1 are not G-protein-dependent and may include regulation of Na+/H+ ion exchange and, conceivably, tyrosine phosphatase stimulation. Availability of the cloned SS receptors will no doubt considerably advance attempts to precisely define the molecular events underlying the remarkable range of effects of SS.


While two separate genes code for either SS-14 or SS-28 in fish, a single gene codes for both in mammals (35). This gene consists of two exons (238 and 367 base pairs) separated by an intron (621 base pairs). Between two 5¢ upstream promoters (TATA and CAAT boxes) lies an enhancer, the cAMP response element (CRE), which was first discovered in the SS gene (20). Among the myriad factors (nutrients, neurotransmitters, neuropeptides, hormones, second messengers, growth factors) that influence SS secretion, often in a tissue-specific fashion, many have been shown as well to enhance SS genomic transcription (see Table 1). For example, SS secretion is stimulated by dopamine (via the D2 receptor), acetylcholine (muscarinic), glutamate (NMDA), growth hormone, neurotensin, CRF, insulin, insulin-like growth factor-1 (IGF-1), interleukin 1 (IL-1), and tumor necrosis factor (TNF) (7, 35, 48). In addition to stimulating secretion, the following also increase SS mRNA: NMDA agonists (cortex), GH (hypothalamus), testosterone and estradiol (hypothalamus), glucocorticoids (hypothalamus, cortex), and IL-1 and TNF (diencephalic culture) (4, 35, 64). The mechanism by which these factors alter SS genomic expression is unknown, although in some cases it will most likely involve stimulation of cAMP, which increases both SS secretion and SS genomic transcription through the CRE. Several hormones and neuroregulators have been found to inhibit SS secretion or to regulate it in a dose-related or highly tissue-specific fashion. Gamma-aminobutyric acid (GABA) is almost uniformly inhibitory, ACTH inhibits secretion in the hypothalamus, vasoactive intestinal polypeptide (VIP) is inhibitory in the hypothalamus but stimulatory in the neocortex, and glucocorticoids are stimulatory at low doses and inhibitory at high doses (35, 48).


Behavioral Effects

An important potential role for SS in CNS activity is suggested by many studies in which alterations of central SS levels result in the modulation of a variety of vegetative and related functions (48). Early studies reported decreased total, slow-wave, and rapid-eye-movement (REM) sleep following intracerebral or i.c.v. administration of SS. A more recent report observed (a) significant increases in REM sleep in rats following i.c.v. SS and (b) suppression of REM sleep following i.c.v. administration of cysteamine, a depletor of SS. No effect on electroencephalographic measures of sleep was seen, however, following intravenous infusion of SS in a study of 11 normal men by Kupfer et al. (27).

Effects of SS on hunger and locomotor activity appear to vary with the dose employed (48). Increased or dose-related biphasic food consumption have both been reported to accompany intracerebral or i.c.v. SS. Chronic i.c.v. administration of the SS analogue octreotide increased daily food intake, whereas SS antiserum significantly decreased food intake. Antagonism of CRF-induced anorexia by SS in starved rats has also been demonstrated. As reviewed by Vecsei and Widerlov (58), SS influences locomotor activity, the nature of the response being dependent upon the SS analogue, the dose, and the behavioral measure employed.

In both animals and humans, analgesia is seen following intrathecal administration of SS (48). Several SS analogues display high affinity at the mu-opiate receptor, and SS colocalizes with methionine enkephalin in neurons of the raphe nucleus and nucleus gigantocellularis (48). However, because the analgesic effects of intrathecal SS in humans are not reversed by naloxone, a non-opiate-mediated mechanism is suggested (25). Finally, a series of studies by Vecsei and Widerlov (58) and Walsh et al. (60) show that SS may enhance learning and reverse induced learning deficits, whereas SS depletion (with cysteamine) diminishes performance.

Neuropsychiatric Disease-Related Alterations

Disease-related alterations in SS levels in the cerebrospinal fluid (CSF) were first reported by Patel et al. in 1977 (37). They described increased levels of SS in several inflammatory or destructive neurologic disorders (cerebral tumor, meningitis, spinal cord disease, nerve root compression, metabolic encephalopathy). These findings were subsequently replicated by other investigators (48). While these elevated CSF SS levels presumably reflected neuronal damage, the decreased CSF SS levels seen in several neuropsychiatric disorders suggested more functional neuronal alterations. These disorders include Parkinson's disease, delirium, Alzheimer's disease, depression, and multiple sclerosis (MS) during relapse (48). Su et al. (53) recently confirmed decreased SS in progressive MS, in contrast to the findings of Rosler et al. (46). Su et al. further demonstrated a significant decrease in SS in serial CSF samples over 2 years in a small group of medication-free patients with MS who manifested neurological deterioration during this interval. Three groups have reported significantly decreased SS in multi-infarct dementia patients, with nonsignificant decreases reported by a fourth group (48)). Decreased CSF SS has also been observed in patients with ACTH-dependent Cushing's syndrome (25). Finally, both decreased and normal levels of SS have been reported in patients with Huntington's disease (25, 48). Postmortem studies have demonstrated decreased SS concentrations in the brains of patients with Alzheimer's disease (see below) and Parkinson's disease (cortex and hippocampus) (48) and increases in the basal ganglia, locus coeruleus, and frontal and temporal cortex of patients with Huntington's disease (5) (presumably due to selective neuronal sparing) (42) (see also Biological Markers in Alzheimer’s Disease and Parkinson’s Disease).


Reports of low CSF SS in senile dementia and Alzheimer's disease have been remarkably uniform (48) (see Fig. 3). Human postmortem studies also demonstrate decreased SS concentrations in a variety of cortical and subcortical sites in Alzheimer's disease (48). The amount of the reduction and the brain region most clearly affected differ across studies. These differences may reflect a number of factors including postmortem changes, molecular heterogeneity of SS-like immunoreactivity, or decreases in SS masked by the loss of non-SS tissue volume (11). Additionally, Lowe et al. (29) attributed reported discrepancies to differences in the length and severity of the disease, consistent with the observation that studies with a greater representation of severely ill patients report greater decreases in brain SS than are seen in other studies. Nonetheless, SS concentrations are widely (but not without exception) reported as reduced to 21–61% of control values in the temporal cortex, with similar reductions seen in the frontal and parietal but not occipital or cingulate cortex (29).

Decreased brain and CSF SS levels in Alzheimer's disease have been observed with remarkable consistency. The clinical relevance of these findings is suggested by the following: (a) reports of correlations between the degree of cognitive impairment and reduction of SS levels in brain and CSF; (b) reports of the colocalization of SS-like immunoreactivity with senile plaques and neurofibrillary tangles; (c) observation of a correlation between the loss of SS and disease severity as determined by plaque density; (d) the demonstration that the brain areas showing the greatest reduction in SS are those that are most markedly hypometabolic as determined by positron emission tomography (PET) scan; and (e) the observation of relative increase in CSF SS concentration in association with a reversal of dementia symptoms in patients with Alzheimer's disease exposed to intensive environmental stimulation (see ref. 48 for review). This last finding is of interest given the presumption that the changes in brain SS in Alzheimer's disease reflect neuronal degeneration. It is further noteworthy given the lack of therapeutic success associated with attempts to manipulate pharmacologically central SS levels [by administering an SS analogue to patients with Alzheimer's disease or an SS-depleting agent, (i.e., cysteamine) to patients with Huntington's disease] (48). The extent to which these treatments produced changes in central SS levels is uncertain (see also Biological Markers in Alzheimer’s Disease).


Among the psychiatric disorders studied, the clearest evidence of abnormal central SS levels or activity is in depression. Seven studies document significantly decreased CSF SS in depression compared with controls (48) (see Fig. 4). Additionally, significantly lower CSF levels were seen in patients with a major depressive disorder compared with other psychiatric patients (56) and in a group of depressed patients studied during their worst week of depression compared with a group of depressed patients studied more than 2 months after their most depressed week (1). In addition to these decreased levels observed in depression, Molchan et al. (33) found a significant negative correlation between CSF SS levels and the degree of depressed mood in 60 patients with Alzheimer's disease, as well as significantly lower CSF SS levels in the depressed Alzheimer's patients compared with the nondepressed patients. Longitudinal examination of patients in different affective states demonstrated that SS values obtained during depression were significantly lower than those obtained during either the manic or improved states (47). Decreased CSF SS levels, therefore, appear to be state-related and to normalize with recovery from depression, similar to the restoration of normal CSF SS levels in MS patients during clinical remission (51).

Studies of CSF SS in samples of patients with psychiatric disorders other than depression have yielded conflicting or as yet unconfirmed results (48). Both increased and normal SS levels have been observed in manic patients. Both decreased and normal SS levels have been reported in patients with schizophrenia and anorexia nervosa (8, 15, 23). Kaye et al. (23) described a small but significant increase in CSF SS in patients with bulimia during abstinence from bingeing compared with the same patients when actively bingeing. Normal levels of CSF SS in patients with panic disorder were reported by Vecsei and Widerlov (57), who also noted normal SS levels in the premenstrual syndrome (PMS) as well as no effect of menstrual cycle phase on SS in either PMS patients or controls. Kruesi et al. (26) observed significantly higher SS levels in 10 children with obsessive–compulsive disorder (OCD) compared with 10 age- and sex-matched children with conduct disorder. Despite the absence of a normal control group in this study, these findings are of particular interest given the recent report by Altemus et al. (2) of significantly elevated CSF SS levels in adult patients with OCD relative to controls. The elevated SS levels in OCD are particularly intriguing in light of the recent demonstration in rat brain of decreased SS following administration of serotonergic agents, the treatment of choice for OCD (21).

Relatively few postmortem studies of brain SS have been performed in patients with psychiatric illness. Several early studies demonstrated no consistent abnormality in the brains of patients with schizophrenia (48). Charlton et al. (12) observed no difference in SS concentration or SS receptor affinity and binding capacity in the temporal or occipital cortex in nine primarily medicated depressed patients compared with controls. These investigators also found no difference in SS immunoreactivity in the frontal cortex in depressed patients (17) or in the frontal, motor, parietal, or temporal cortex of 12 suicide victims compared with controls (13). Significantly decreased SS content has also been observed in the temporal pole and pars opercularis of the frontal lobe (but not in other brain regions examined) in seven depressed and previously medicated patients (9). Finally, no differences in SS content were noted in the frontal cortex, amygdala, or caudate in a small series of schizophrenic (n = 13) or affectively ill suicide victims (n = 9) compared with accident victim controls (n = 24) (Davis et al., unpublished data) (see also Neuroendocrinology of Mood Disorders and Neuropeptide Alterations in Mood Disorders).


Although the precise role of SS in depression and other neuropsychiatric disorders is currently unknown, the potential relevance of alterations in SS secretion to CNS activity and seizure development is increasingly apparent. The following supportive observations (see ref. 48) exemplify the as yet undocumented mechanisms by which somatostatinergic neurons may be involved in other neuropsychiatric disorders:

1. Long-term (but not permanent) increases in SS in selective brain regions (amygdala and sensorimotor, piriform, and entorhinal cortex; striatum) following kindling (an animal model of epilepsy in which daily subictal stimuli result in progressively more intense brain activity culminating in generalized seizures) or prekindling [rats injected with pentylenetetrazol (PTZ) but not fulfilling the criteria of kindling].

2. A 200% increase in SS precursor and 60% and 80% increases in SS-14 and SS-28, respectively, in the frontal cortex after both PTZ-kindled and kainic-acid-stimulated seizures.

3. Long-lasting increases (30 days) in prepro-somatostatin mRNA in the frontal cortex and hippocampus (but not striatum or substantia nigra) following kainic-acid-induced seizures.

4. Selective increases in SS and prepro-somatostatin mRNA (but not glutamic acid decarboxylase (GAD or GAD mRNA) in the striatum and neocortex (34, 48) in stage 5 kindled rats.

5. Increases of SS in stage 3 kindled rats (forelimb clonus) and in stage 5 kindled rats (generalized seizures, rearing and falling), with the elevation in the latter observed 1 hr after the seizure and persisting up to 2 weeks following the last seizure.

6. Significant increases in SS mRNA after stage 5 kindling in hippocampus (CA1, CA2, dentate gyrus), cingulate cortex, olfactory tubercle, and rostral cortex compared with increases seen following a single electrical stimulation or control conditions (50).

7. Increases in SS-14 in focal (epileptic) compared with nonfocal (nonepileptic) regions of the temporal cortex in 33 of 35 patients with intractable seizures.

8. Precipitation of atypical seizures following i.c.v. or intracerebral injection of SS.

9. Inhibition of kindled seizures following administration of cysteamine or SS antiserum.

10. Reversal of the inhibitory effects of cysteamine by i.c.v. infusion of SS.

11. Selective loss of SS neurons in the hilus of the dentate gyrus after kainic-acid-induced seizures (52) and in human temporal lobe epilepsy (45).

12. Inhibition of kainic acid or quinolinic-acid-induced seizures (in the hippocampus) by a peptidase-resistant somatostatin analogue (59).

13. Enhancement of hippocampal kindling after infusion of SS antiserum (34).

14. Reversible decreases in SS during acute seizures stimulated by PTZ (and presumed to be related to seizure-related increased release) (48, 52).

15. Kindling-related downregulation of SS receptors in the hippocampus.

16. Periventricular SS release following electrical stimulation of several limbic sites.

Despite the conclusions of the authors of two additional studies that changes in SS may not be relevant to the development of seizures (14, 40), further support for a potential role of SS in seizure susceptibility or activity is provided by reports (described below) of the effects on SS of several anticonvulsants. The precise role of SS in seizure modulation remains to be documented but is particularly intriguing in relation to long-lasting increases of SS in some seizure models, such as kindling, but losses in others, such as kainate-precipitated seizures (see Table 2).


Decreased CSF SS levels have been observed in humans following administration of several different anticonvulsants including diphenylhydantoin and carbamazepine. This effect of carbamazepine has been observed in patients with affective disorder as well as in patients with temporal lobe epilepsy (48). Lower CSF SS levels have also been observed in rats 2 hr after intraperitoneal administration of carbamazepine (48). This effect of carbamazepine on CSF SS most likely reflects the ability of carbamazepine to alter many of the neurotransmitters involved in SS regulation (dopamine, GABA, acetylcholine, norepinephrine) (41). Carbamazepine modulation of GABA, an inhibitor of SS secretion, appears particularly relevant (48). Carbamazepine decreases GABA turnover, perhaps involving a GABA agonist mechanism. Additionally, the action of carbamazepine at the trigeminal nucleus may be mediated via baclofen-like or GABA-B receptor mechanisms. Furthermore, the ability of midazolam to inhibit SS secretion from diencephalic and cerebral cultures is postulated to occur via stimulation of the GABA-B receptor. Finally, SS appears to act presynaptically in the CA1 region of the hippocampus to dramatically depress GABA-mediated inhibitory postsynaptic potentials. Sharfman and Schwartzkroin (49) speculated that the SS-related inhibition of GABA-induced hyperpolarization might render the hippocampus hyperexcitable, facilitating the induction of long-term potentiation (LTP). While SS has been reported to facilitate LTP in CA3 (but not CA1) (31), it also may hyperpolarize CA1 neurons, which, by itself, would dampen excitability (62). Excitability of brain regions such as the hippocampus (in which SS and GABA colocalize) (62) may thus reflect the reciprocal (albeit complicated) inhibitory regulatory effects of GABA and SS. Therefore, the anticonvulsant effects of carbamazepine may in part result from carbamazepine-induced decreases in SS activity. This hypothesis is supported by several observations: (a) Carbamazepine blunts kindling-induced increases of SS in the frontal, temporal, and occipital cortex and amygdala (but not hippocampus), and (b) acute and chronic high-dose carbamazepine administration results in decreased SS and increased GABA levels in several (primarily limbic) brain regions (48). Several other studies, however, suggest that the effects of carbamazepine on basal SS concentrations in the brains of nonkindled rats are only slight or absent (28, 48).

The effects of other psychopharmacologic agents on SS have been examined in several studies. The neuroleptic fluphenazine decreased SS in the CSF of a group of schizophrenic patients (15), and the neuroleptic haloperidol decreased SS in several brain regions (striatum and nucleus accumbens) and decreased SS receptors in the rat cerebral cortex and hippocampus (39). One additional study reported an increase in CSF SS in a small group of patients during treatment with haloperidol (18!popup(ch53ref18)). No significant effects of desmethylimipramine, piribedil, or lithium carbonate were observed on CSF SS (48), nor were there effects of imipramine, maprotilline, or mianserin on SS concentrations in rat brain (22). However, Kakigi et al. (21, 22) did observe (a) significant decreases in SS concentrations in a variety of brain regions following repeated administration of serotonergic agents (clomipramine, zimelidine, 5-hydroxytryptophan) and (b) increased concentrations following treatment with a serotonin synthesis inhibitor or neurotoxin. As noted above, these latter findings are of interest given the selective efficacy of serotonergic agents in the treatment of OCD, a disorder characterized by elevated CSF SS levels. Finally, in contrast to the ability of several psychopharmacologic agents to decrease CSF SS, nimodipine, a calcium channel antagonist employed in the treatment of several cyclic mood disorders, significantly increased CSF SS in eight patients with affective disorder (unpublished data).


Many of the questions raised by the observation in 1977 of alterations in SS levels in neuropsychiatric disorders still remain: What are the determinants, and what are the consequences of altered SS levels and function? What are the mechanisms by which SS regulates behavior, and what is the range of behavioral functions that it can modulate? What is the regulatory relevance of the differential effects of pro-somatostatin-derived peptides? It seems reasonable to conclude at present that the decrease in brain and CSF SS that are uniformly observed in Alzheimer's disease appear to reflect neuronal degeneration, while the decreased CSF SS seen in depression, which, like that in MS, is transient and reversible, perhaps reflects functional alterations in SS secretion and/or metabolism. While alterations in CNS SS in neuropsychiatric disorders may merely be epiphenomenal to more etiologically relevant upstream neuroregulatory disturbances, the central role played by SS in sensory processing and complex integrated behaviors suggests that, at the very least, disturbances of SS function will influence the phenomenology of disorders in which SS dysregulation occurs. For example, the possible contribution of SS dysregulation to cognitive impairment is suggested by the following: SS administration facilitates learning and memory; cognitive impairment characterizes the neuropsychiatric disorders associated with decreased SS (Alzheimer's disease, depression, MS, Parkinson's disease, and Cushing's disease); and the extent of the decrease in CSF or brain SS has been reported by some (but not all) to be significantly correlated with the degree of cognitive impairment in patients with Alzheimer's disease and Parkinson's disease (48). Distinct roles for the SS-related peptides in neuropsychiatric disorders cannot be assigned at present, although several authors (19, 43) have suggested that abnormal processing of SS precursors may contribute to abnormal levels observed as well as influence syndromal characteristics. To the extent that dysregulation of SS may contribute to neuropsychiatric disorder-related symptomatology, therapeutic benefits may accompany successful efforts to increase brain SS activity, whether through intrathecal administration of SS, development of analogues or drug transport mechanisms that circumvent the blood–brain barrier, modulation of SS metabolism, or selective enhancement of SS gene expression.

The availability of five cloned high-affinity SS receptors offers the promise of an explosion of information about SS in the next decade. Identification of the characteristics of these receptors—including pharmacologic profile, physiologic effects, and intracellular effector systems—will permit precise definition of the consequences of alterations in specific SS fragments or receptor subtypes. The development of specific agonists and antagonists for SS receptor subtypes will not only facilitate the physiologic and pharmacologic dissection of the SS system, but may also result in the development of a new generation of therapeutically successful SS analogues. A mere 20 years after its identification, we are on the verge of understanding not only the role of SS in CNS function, but also the consequences of dysregulation of the expression of SS and its receptors.

published 2000