International Union of Basic and Clinical Pharmacology. CV. Somatostatin Receptors: Structure, Function, Ligands, and New Nomenclature

1Since their discovery, research on somatostatin and its receptors has remained active with more than 700 papers published annually. Somatostatin—also known as somatotropin release-inhibiting factor (SRIF)—was originally discovered in 1973 as a hypothalamic neuropeptide based on its ability to inhibit growth hormone (GH) release from the anterior pituitary (Fig. 1). SRIF occurs in two forms, SRIF-14 and SRIF-28, with broad antisecretory activity on many hormones, including GH, insulin, glucagon, gastrin, cholecystokinin (CCK), and ghrelin. In the original report, it was suggested that SRIF could have potential for treatment of acromegaly. However, due to its short circulating half-life (t1/2) (<3 minutes), the therapeutic potential of natural SRIF-14 is limited. Consequently, highly potent and metabolically stable SRIF analogs were synthesized in 1982. The first SRIF analog approved for clinical use was octreotide. About 10 years later, the first SRIF-based radiopharmaceuticals were synthesized by conjugating a chelator to octreotide, followed by radiolabeling with a γ-emitter, which paved the way for in vivo imaging of human tumors. In the early 1990s, five subtypes of somatostatin receptors (SSTs) termed SST1 to SST5 were cloned in mice, rats, and humans. Given that only two SRIF tissue binding sites could be identified using ligands available at that time, the subsequent discovery of five different SSTs was surprising and triggered in-depth research into binding properties, localization, and regulation of the ligand. This led to classification of the clinically used SRIF analogs octreotide and lanreotide as SST2-prefering ligands, which in turn stimulated the search for novel compounds that bind either more broadly or more selectively to individual SSTs. In 1996, a structurally related neuropeptide termed cortistatin (CST) with a more restricted distribution in the cerebral cortex and hippocampus was identified. In the late 1990s, knockout (KO) mice and selective nonpeptide agonists were developed for all five SSTs, which helped to define their physiologic functions. In 1998, the development of octreotide conjugates radiolabeled with a β-emitter provided proof-of-principle for peptide-receptor radiotherapy (PRRT). A few years later, the capacity of SSTs to form homodimers and heterodimers with other G protein–coupled receptors (GPCRs) was observed, which stimulated the search for bitopic chimeric compounds. In the early 2000s, the search for multireceptor ligands led to the discovery of pasireotide, which was the first pituitary-directed drug approved for therapy of Cushing’s disease. In 2008, high-affinity peptide antagonists with utility for SST imaging and treatment were synthesized, suggesting that receptor internalization is not an absolute requirement for tumor imaging. Also in 2008, the generation of highly specific rabbit monoclonal antibodies (mAbs) facilitated detection of SSTs in human tissues and enabled correlation of SST2 and SST5 receptor expression with octreotide and pasireotide responses. Shortly thereafter, development of phosphosite-specific antibodies provided molecular insights into mechanisms for SST activation by octreotide and pasireotide. In 2009, a novel truncated variant of SST5 generated by aberrant splicing was identified [SST5 transmembrane domain (SST5TMD4)] and shown to be overexpressed in several hormone-related tumors, wherein the variant increases aggressiveness. In the past few years, orally available and subtype-selective SST agonists and antagonists have been synthesized. Some of these substances may become lead compounds for potential new therapeutic indications directed toward individual SSTs.

2SRIF, a cyclic neuroendocrine peptide, was first isolated and identified as a hypothalamic factor that inhibited GH secretion from anterior pituitary cells. SRIF exists in two main bioactive isoforms: the tetradecapeptide (SRIF-14) isolated from the hypothalamus and the 28-amino-acid isoform (SRIF-28), generated from the same prepro-SRIF precursor through post-translational processing at a distinct cleavage site and which differs from the shorter isoform by an N-terminal extension (Fig. 2). Both isoforms are expressed at variable amounts in the same tissue areas. It is not clear whether the two peptides are coexpressed by the same cells or by separate cells. The family of somatostatin peptides includes CST, a highly similar peptide reviewed below, and which is structurally and functionally related to the urotensin II peptide family. The two families (somatostatin and urotensin II) as well as those of their respective GPCRs may derive from a single ancestral ligand–receptor pair. The duplication, generating the two families, likely occurred before the emergence of vertebrates. Subsequently, each family expanded during evolution, through whole-genome duplications, followed by local duplications and gene losses. Despite their evolutional divergence, the two families conserve close functional links. The vertebrate SRIF family is composed of at least six paralogous genes named SS1 to SS6. In mammals, SRIF-14 and SRIF-28 both derive from the SS1 gene, localized on chromosome 3q27.3 in humans. The SRIF-14 primary structure is highly conserved in vertebrates, and cleavage sites generating SRIF-14 and its extended isoform have been fully conserved during evolution. An additional product of the mammalian processing of prepro-SRIF is a 13-amino-acid noncyclic amidated peptide, neuronostatin, which immediately follows the signal peptide. Bioinformatic analyses of evolutionary conserved sequences suggest the occurrence of neuronostatin in other vertebrates. A novel peptide showing structural similarity to SRIF-28 and isolated from monkey ileum comprises amino acid sequences matching the N-terminal 13 amino acids of SRIF-28. This peptide is expressed in enteric neurons and may play a possible role in food intake control.

3The structure of rat and human SRIF genes, as well as the transcriptional unit and upstream regulatory elements of the rat gene, has been characterized. Expression of the SRIF precursor gene is regulated by growth factors and cytokines, including GH, insulin-like growth factor 1 (IGF-1), insulin, leptin, and inflammatory cytokines, and by glucocorticoids, testosterone, and estradiol. cAMP is a potent activator of SRIF transcription, and SRIF release from neurons and peripheral secretory cells is triggered by membrane depolarization and increased cytosolic calcium concentrations. Several neurotransmitters, neuropeptides, hormones, and nutrients, some also altering gene transcription, affect SRIF release in the central nervous system (CNS) and in peripheral tissues. Characterization of neurotransmitter, neuropeptide, and hormone modulation of hypothalamic SRIFergic neurons has raised interest in light of the key role played by SRIF as a distal mediator for neuroendocrine and metabolic control of the GH axis activity in health and disease.

4Abundant SRIF immunoreactivity is evident in the mediobasal hypothalamus and median eminence, amygdala, preoptic area, hippocampus, striatum, cerebral cortex, olfactory regions, and the brainstem. Three main categories of SRIFergic neurons can be distinguished: hypophysiotropic neurons, long-projecting GABAergic neurons, and GABAergic interneurons acting within microcircuits.

5In the rat CNS, SRIFergic neurons regulating pituitary function are located within the periventricular nucleus and the parvocellular part of the paraventricular nucleus and send axonal projections to the median eminence at the base of the hypothalamus. SRIF-producing neuronal cell bodies are also found in the arcuate (ARC) and ventromedial nuclei. Hypophysiotropic SRIFergic neuronal axons descend toward the pituitary stalk and release SRIF into the portal blood vessel system, thereby reaching anterior pituitary cells. Some axons travel through the neural pituitary stalk into the neurohypophysis. Other fibers project outside the hypothalamus to areas such as the limbic system or may interact, through interneurons, with other hypothalamic nuclei, including the ARC where GH-releasing hormone (GHRH) is expressed, the preoptic nucleus, the ventromedial nucleus, and the suprachiasmatic nucleus, which exhibits circadian pacemaker activity. SRIF is ubiquitously expressed in mammalian brain. Extrahypothalamic SRIF immunoreactivity is found in the amygdala, preoptic area, hippocampus, striatum, cerebral cortex, sensory regions, and brainstem. SRIF neurons are classified into two main categories: interneurons acting locally within microcircuits and long-range–projecting neurons whose fibers reach distant areas. SRIF frequently colocalizes with GABA. SRIFergic interneurons likely play a role in regulation of distal dendrite excitability, and long-range–projecting SRIFergic neurons may participate in coordinating activity between distant brain regions. Accordingly, nonpyramidal cells located in the hippocampus and targeting the medial septum and the medial entorhinal cortex form inhibitory synapses on GABAergic interneurons, and may cooperate in generation and synchronization of rhythmic oscillatory activity in these areas. Recently, brain-wide SRIF neuron maps have established the precise cell number, density, and somatic morphology with anatomic references for SRIF-related specific functions.

6Peripheral SRIF is produced by secretory cells in gastrointestinal mucosa and by δ-cells in pancreatic islets, as well as by a subpopulation of C cells within the thyroid gland. In addition to SRIF-producing neuroendocrine cells, inflammatory and immune response cells and tumor cells may also express SRIF. The peptide has been immunolocalized in human epidermis, with low expression levels in keratinocytes and higher expression in subsets of Merkel and dendritic cells. SRIF is a neurotransmitter and neuromodulator, an endocrine hormone and a paracrine factor acting in the same tissue where it is expressed. Circulating levels of SRIF are very low because the peptide is rapidly degraded by ubiquitous peptidases. In the retina, SRIF was detected by immunohistochemical studies in amacrine and ganglion cells, as well as in interplexiform cells, whereas electrophysiological studies support the view that it may function as a neurotransmitter, neuromodulator, or trophic factor.

7Within the hypothalamus–pituitary system, SRIF is the main regulatory element exerting inhibitory control on both basal and stimulated GH secretion and reduces prolactin and thyroid-stimulating hormone (TSH) secretion in normal subjects. It can also suppress release of adrenocorticotropic hormone (ACTH) from tumor cells. SRIF brain actions are mediated by presynaptic or postsynaptic mechanisms. SRIF modulates neuronal excitability, and in the hippocampus, cortex, and hypothalamus it also induces presynaptic inhibition of excitatory neurotransmission. In other brain areas, SRIF also decreases GABA release. Postsynaptic mechanisms of action include membrane hyperpolarization via activation of potassium ion currents (K+ currents), in particular voltage-gated K+ currents, noninactivating potassium currents (M currents), and voltage-insensitive leak currents. SRIF is coreleased with GABA from hippocampal neurons and from axonal terminals in other brain areas. SRIF inhibits dopamine release from the midbrain as well as hypothalamic release of noradrenaline, thyrotropin-releasing hormone, and corticotropin-releasing hormone (CRH). Activation of brain SRIF signaling may alleviate endocrine, autonomic, and behavioral responses to stress mediated by central CRH and CRH receptors. SRIF has a role in cognitive functions, learning and memory processes, control of locomotor activity, control of food intake, nociception, and autonomic functions. SRIF is highly expressed in brain regions associated with seizures and has been suggested as an endogenous antiepileptic.

8Peripheral SRIF actions include inhibition of hormone secretion, exocrine secretion, and cell proliferation. In the gastrointestinal (GI) tract (GIT), SRIF exerts a generalized inhibitory effect on release of gut hormones [including gastrin, CCK, gastric inhibitory polypeptide, vasoactive intestinal peptide, enteroglucagon, motilin], gastric acid, digestive enzymes, bile, and colonic fluid. SRIF also negatively affects gallbladder contraction, small intestinal segmentation, and gastric emptying. In pancreatic islets, release of SRIF from δ-cells inhibits secretion of insulin, glucagon, and other peptides from neighboring cells. SRIF reduces TSH-induced release of triiodothyronine (T3) and thyroxine as well as calcitonin release. In the adrenals, SRIF inhibits angiotensin II–stimulated aldosterone secretion and acetylcholine-stimulated medullary catecholamine secretion. SRIF reduces release of kidney-derived renin caused by hypovolemia and vasopressin-mediated water absorption. In addition to nervous system functions and regulation of endocrine and GI functions, SRIF also may affect key cellular processes in diverse tissues by regulating the release of both growth factors and cytokines as well as cellular responses to these stimuli. SRIF can contribute to control of smooth muscle cell contractility, lymphocyte and inflammatory cell proliferation and activity, tumor cell growth, and normal tissue plasticity. In human skin, SRIF has been suggested as a negative regulator of epidermal wound healing. Finally, at variance with its nearly universal inhibitory actions, low (pM) concentrations of SRIF stimulate in vitro GH release on cultured pituitary cells derived from pigs and nonhuman primates and from human somatotroph adenomas.

9CST, a cyclic neuropeptide, highly homologous to SRIF, was identified as a region-specific brain mRNA encoding a protein of 112-amino-acid residues, which was called preproCST. CST in mammals derives from the CORT gene. The gene encoding for human and mouse CST is located on 1p36.3–1p36.2 and on chromosome 4, respectively. Similarly to preproSRIF, cleavage of preproCST gives rise to multiple mature products, CST-14 and CST-29 in rats and CST-17 and CST-29 in humans (Fig. 2). CST-14 and SRIF-14 differ in three amino acid; CST aligns with the second-amino-acid residue of SRIF on the N-terminus and terminates one-amino-acid residue beyond the C-terminal of SRIF. Human CST-17 contains an arginine for lysine substitution and is extended by three amino acids at the amino-terminal end, resulting in CST-17 sharing 10 of the 14 SRIF-14 residues. Similar to their prepropeptides, mature CST and SRIF are also highly homologous, including the two cysteine residues that render the peptides cyclic as well as a FWKT motif critical for SST binding. Consequently, CST peptides bind to all SST subtypes with similar affinity than SRIF, and yet there is no evidence for a selective cortistatin receptor. Notably, the FWKT motif is also present in urotensin II and urotensin-related peptide, which are indeed agonists of SSTs.

10PreproCST mRNA is predominantly expressed in the cerebral cortex and hippocampus. In the cortex, mainly layers II–III and VI contain CST-positive cells. Interestingly, CST-positive cell bodies are not uniformly distributed in all cortical areas, with highest numbers evident in the visual and temporal cortex and lowest in the somatosensory cortex. CST-containing neurons are also detected in the piriform cortex and entorhinal area. In the hippocampus, CST expression is found in a small subset of nonpyramidal neurons of the subiculum and in the stratum oriens of hippocampus subfields Cornu Ammonis (CA)1–3. In the hilar region of the dentate gyrus, however, CST-positive neurons are only transiently present during development. In parallel, there is a temporary increase of cortical CST expression during development, which correlates with maturation of cortical interneurons. Indeed, cortical CST-expressing neurons, similarly to SRIF-positive neurons, are also GABAergic. However, CST- and SRIF-containing neurons are expressed in distinct, only partially overlapping populations. CST is also expressed in other brain areas, such as in the olfactory bulb, in the striatum, in the periventricular nucleus of the hypothalamus, and in GABAergic interneurons of the deep layers of the spinal cord dorsal horn. No CST expression was detected in the thalamus, brainstem, or cerebellum. The projections of CST-positive neurons were not analyzed in detail, but due to their high homology it is possible that anti-SRIF antibodies may also label CST-containing axons.

11CST is also expressed in the periphery, in general at lower level than SRIF but with a broader distribution: preproCST mRNA was detected in peripheral nociceptive neurons, endocrine organs (e.g., pituitary gland, adrenal gland, thyroid gland, parathyroid gland, endocrine pancreas), digestive system (e.g., stomach, jejunum, ileum, colon, rectum, liver), kidney, lung, and gonads, and also in smooth muscle cells, endothelial cells, and immune cells (e.g., lymphocytes, monocytes, macrophages, dendritic cells). Discrepancies between mRNA expression and protein synthesis were observed in several tissues (including adrenal, thyroid, lung, and gonads). Notably, predominant CST is present in parathyroid chief cells and immune cells.

12In keeping with their similar affinities to SSTs, CST and SRIF share several biologic properties, including inhibition of neuronal activity and consequent antiepileptic activity, inhibition of cell proliferation, and regulation of hormones, and particularly inhibition of GH secretion. Double KO mice devoid of both SRIF and CST show markedly increased GH levels, although they do not display overt giant phenotypes. Nevertheless, functional differences of CST and SRIF can only partly be attributed to their distinct tissue distributions. At the cellular level, CST, similarly to SRIF, increases the M current in hippocampal neurons but also augments the hyperpolarization-activated currents, thereby modulating synaptic integration and regulation of oscillatory activity. At the behavioral level, CST induces hypomotility, whereas SRIF causes hypermotility; CST enhances slow-wave sleep, whereas SRIF increases rapid eye movement sleep. CST and SRIF also regulate differently endocrine functions as well as learning and memory processes. Consistent with its widespread distribution in the immune system, CST is a potent anti-inflammatory factor, decreasing the production of several inflammatory cytokines [tumor necrosis factor-α, interleukin (IL)-1β, IL-6, IL-12, interferon-γ], chemokines, and acute-phase proteins. CST also inhibits T helper 1- and 17-driven inflammatory responses in models of inflammation [e.g., sepsis, atherosclerosis], and autoimmune diseases [e.g., inflammatory bowel disease, rheumatoid arthritis, and multiple sclerosis]. In parallel with its potent anti-inflammatory effect, CST is also an endogenous analgesic factor acting at both the peripheral and spinal level.

13CST may activate GPCRs other than SSTs, including the ghrelin receptor 1a (GHS-R1a) and human-specific MAS-related GPR family member X2. Functions of CST not shared by SRIF in the immune and endocrine systems are likely mediated by GHS-R1a, whereas MAS-related GPR family member X2 might play an important role in the analgesic effects of CST in humans. In addition, some CST-specific functions might be mediated by a yet unidentified CST-selective receptor, and by truncated SST5 variants that selectively respond to CST.

14There is yet considerable misconception and lack of clarity regarding classification and nomenclature of SSTs. SRIF binding sites were initially defined through radioligand-binding studies performed in rat brain cerebral cortex membranes. SRIF-1 (also called SS-1) recognition sites were characterized by high affinity for SRIF-14 and SRIF-28, and for cyclic peptides such as octreotide and seglitide. In contrast, SRIF-2 (also called SS-2) sites exhibit high affinity for SRIF-14 and SRIF-28, but very low affinity for octreotide and seglitide. In fact, SS-1 and SS-2 binding sites correlate with recombinant SST2 and SST1 receptors, respectively.

15Subsequently, five distinct receptor genes have been cloned and named chronologically according to their respective publication dates, but two were regrettably given the same appellation (SST4). In 1995, a consistent nomenclature for the recombinant receptors was agreed upon according to International Union of Basic and Clinical Pharmacology (IUPHAR) guidelines (sst1, sst2, sst3, sst4, and sst5). Given that radioligands could differentiate only two distinct SRIF binding sites, the subsequent cloning of five receptors was indeed surprising. IUPHAR guidelines recommended that recombinant receptors without well-defined functional characteristics should be referred to by lower case letters, i.e., sst1, sst2, etc.. When the recombinant receptor is shown to be of functional relevance in whole tissues and is fully characterized, upper case letters should be used, i.e., SST1, SST2, etc.. Moreover, the name of a receptor should not include the letter “R” or “r” as an abbreviation for receptor. Thus, according to IUPHAR guidelines, employing receptor names such as SSTR1, SSTR2, etc., is discouraged.

16Shortly after cloning, it became apparent that the five recombinant receptors comprise two classes or groups, on the basis of their phylogeny, structural homologies, and pharmacological properties. One class was referred to as SRIF1, comprising SST2, SST3, and SST5 receptor subtypes. The other class was referred to as SRIF2, comprising the other two recombinant receptor subtypes SST1 and SST4. SST subtypes share many structural characteristics and their main intracellular signaling pathways. Conversely, individual SST subtypes can now clearly be differentiated according to their cellular and subcellular localization as well as distinct modes of regulation and functional and pharmacological properties. The IUPHAR Committee on Receptor Nomenclature and Drug Classification subcommittee now recognizes the physiologic correlates of SSTs and has decided on upper case nomenclature for all five SSTs. Thus, the new recommended nomenclature for SSTs is SST1, SST2, SST3, SST4, and SST5.

17All five SSTs are prototypical class A GPCRs that belong to the rhodopsin-like family of receptors. All possess seven transmembrane domains (TMDs) that provide the characteristic architecture of GPCRs. Receptor sequences for human SSTs range in length from 364 amino acids for SST5 to 418 amino acids for SST3 (Table 1). Unfortunately, crystal structures are not yet available for any SST. However, the five SST subtypes share common structural features such as a conserved sequence (YANSCANPILY) in transmembrane region 7 (mammalian SST signature). In addition, there is a consensus motif (X-[S/T]-X-Φ) at the end of the carboxyl-terminal tail of all mammalian SSTs. The X-S/T-X-Φ motif is regarded as a potential postsynaptic density protein (PSD)-95/discs large/ZO-1 (PDZ) domain binding site crucial for interaction with scaffolding proteins. Like all prototypical GPCRs, SSTs contain a DRY motif in the second intracellular loop (ICL) and are involved in coupling to G proteins. Genes encoding human SST1–SST5 are located on chromosomes 14, 17, 22, 20, and 16, respectively (Table 1). There is considerable sequence similarity between different SST subtypes (39%–57%) (Table 2). In fact, sequence similarity is high for a given subtype when compared across species (81%–98% for mouse, human, and rat homologs).

18Comparative genomic analysis suggests that the current set of receptors present in mammalian species arose from a single ancestral gene. This precursor was duplicated before the appearance of vertebrates, leading to genes coding for ancestral SRIF1-type and SRIF2-type receptors, and one gene coding for the ligand, SRIF. Further tetraploidizations occurred during vertebrate evolution, generating genes coding for SST2, SST3, and SST5 from the SRIF1-type precursor gene, and genes coding for SST1, SST4, and SST6 from the SRIF2-type precursor gene. The gene coding for SST6 has been lost in mammals, but is identifiable in several fish species. An additional tetraploidization in teleost fish gave rise to even more receptor species. The common ancestor also gave rise to two so-called Drostar receptors in Drosophila that are not activated by known mammalian peptides (including SRIF variants and opioids). However, their endogenous ligand allatostatin C bears only superficial similarity to SRIF. In contrast, the signature motif YANSCANPILY present in mammalian receptors is only slightly modified to YSNSAVNPILY in Drostar1, and the C-terminal PDZ ligand motif found in all SSTs is also present in the fly.

19Genes encoding SSTs are intronless within their coding sequence, except for SST2. The SST2 gene can be alternatively spliced to produce two receptor proteins, SST2A and SST2B, that differ in length and sequence of their carboxyl termini. Human tissues contain the unspliced SST2A variant exclusively, whereas both spliced forms have been identified in rodents. Although the SST5 gene does not contain CD introns, variants of SST5 mRNA formed by splicing of noncanonical donor and acceptor splice sites are identified in humans, pigs, and rodents. The human SST5 variants encode truncated receptors containing five (SST5TMD5) or four (SST5TMD4) transmembrane domains and distinct carboxyl-termini.

20Despite the prominent therapeutic role of SST2- and SST5-targeting SRIF analogs in pharmacotherapy of endocrine tumors, surprisingly few disease-associated mutations have been identified in any of the seven genes comprising the SRIF system (two peptide precursors and five receptor genes). To date, it has been reported that a single acromegaly patient resistant to octreotide treatment displayed a coding polymorphism in SST5 that clearly affected receptor signaling. The R240W mutation presumably disrupts G protein and mitogen-activated protein kinase (MAPK) signaling, abolishing the antisecretory effects of SRIF on SST5-expressing cells. Besides this unique case, loss of heterozygosity at SST5 was speculated to lead to reduced mRNA expression, but molecular mechanisms for this notion have not been conclusively elucidated. Although numerous studies have reported reduced SST2 and SST5 expression in treatment-resistant tumors, correlations with any particular polymorphism in SST genes have not been established. Molecular mechanisms underlying low SST expression in octreotide- or lanreotide-resistant tumors must therefore reside in genes outside of the SRIF system and still await identification.

21Data from large-scale human sequencing studies, such as the exome aggregation consortium (which includes about 60,000 healthy individuals), further show that all five genes coding for SSTs are rather tolerant to sequence variations. In a ranking of about 18,000 human genes based on the presence of missense mutations, the genes coding for SST receptors are placed between positions 1459 (SST1) and 9488 (SST5), where 1 would be the gene that has the lowest tolerance for missense mutations.

22The best-characterized action of SRIF is its strong inhibitory effect on both endocrine and exocrine cell secretion. SRIF also inhibits neuronal excitability. All SST subtypes mediate inhibitory actions by association with Gi/Go proteins, members of the heterotrimeric guanine-nucleotide–binding protein family characterized by sensitivity to pertussis toxin (PTX). Activation of Gi/Go proteins by SSTs leads to suppression of two critical second messengers: cAMP and cytosolic Ca2+. The reduction in cAMP results from inhibition of adenylyl cyclase. SSTs act to inhibit calcium channels both directly and indirectly, the latter by opening G protein–activated inward-rectifier K channels (Kir3.x) to produce hyperpolarization and, as a consequence, inhibition of Ca2+ influx through voltage-operated calcium channels (VOCC). Reduction of either cAMP or cytosolic Ca2+ leads to inhibition of secretion, and the simultaneous reduction of both second messengers by SRIF results in synergistic inhibitory effects on hormone release. Signaling events responsible for inhibition of cell proliferation are less well understood than those that inhibit hormone secretion. One such pathway involves activation of protein tyrosine phosphatases (PTP), including the Src homology region 2 domain-containing phosphatase (SHP)-1 and SHP-2. Ensuing dephosphorylation of specific substrates may counteract growth factor–stimulated tyrosine kinase activity and thus inhibit multiple mitogenic signaling pathways (Table 1).

23Before subtype-specific antibodies became available, detailed mapping of receptor mRNA and binding sites had been described and comprehensively reviewed. SSTs are widely expressed in the CNS and the endocrine system with some overlapping distributions, but different cellular and subcellular localizations. Within the CNS, SSTs are mainly expressed on neurons in the cortex, hippocampus, amygdala, and hypothalamus. In the endocrine system, SSTs are expressed on distinct cell populations in the anterior pituitary, pancreatic islets, adrenals, and neuroendocrine cells of the GIT. SSTs are also expressed on enteric ganglion and immune cells (Table 1).

24All five SSTs bind the endogenous SRIF and CST ligands with high affinity. Many peptidic SRIF analogs have been developed, leading not only to the discovery of metabolically stable multireceptor SRIF analogs but also to subtype-selective receptor agonists and antagonists. Selective nonpeptide agonists and antagonists are now available for all five SSTs, except SST4, for which selective antagonists are still lacking (Table 1).

25Human SST1 was the first SST cDNA to be cloned by. Cloning was accomplished from a pool of GPCR-like sequences amplified from human pancreatic islet RNA by reverse-transcriptase polymerase chain reaction using a generic set of primers covering highly conserved amino acid sequences in the third and sixth GPCR transmembrane segments. The mouse homolog was then obtained by screening a mouse genomic library with a human SST1 probe. There is 99% amino acid identity between human and mouse SST1 sequences. had previously reported the sequence of a rat cDNA encoding a novel putative GPCR expressed primarily in the hypothalamus and cerebral cortex, but whose ligand could not be identified. The high identity of the predicted amino acid sequence of this orphan receptor (97% and 98% identity with human and mouse SST1, respectively) indicated that it corresponded to the rat homolog.

26In humans, SST1 is a 391-amino-acid protein encoded on chromosome 14q13 by an intronless gene whose promoter region, transcription start site, and 5′-untranslated region (UTR) have been elucidated (Fig. 3). Characterization of rabbit mAbs against SST1 (UMB-7) indicated that the protein is heavily glycosylated, migrating in SDS-PAGE at Mr 45,000–60,000, but displaying the expected mol. wt. of Mr 45,000 after protein extracts had been subjected to enzymatic deglycosylation. The gene coding for mouse and rat SST1 is localized on chromosomes 12 C1 and 6q23, respectively, and both encode a 391-amino-acid protein.

27Studies on SST1 signaling in various cell types have yielded heterogeneous results, depending on whether SST1 is expressed endogenously or heterologously by transfection, indicating that the cellular environment is important in determining the signal transduction machinery. Signal transduction studied in SST1 heterologous cell systems was first thought to be G protein-independent, but later shown to involve both PTX-sensitive and -insensitive G proteins. Development of a polyclonal antibody to a 15-amino-acid peptide corresponding to a unique sequence in the SST1 carboxyl terminus has made it possible to immunoprecipitate endogenously expressed SST1 from pituitary tumor cell lysates and to demonstrate specific coupling to PTX-sensitive G proteins. Gi proteins reported to couple to SST1 in SST1-transfected cells include Giα1,2,3, although only Giα3 dominantly couples SST1 to downstream adenylate cyclase inhibition. SST1 transduces reduction of both cAMP accumulation and intracellular Ca2+ concentrations in heterologous cell systems (Fig. 4), as well as in insulinoma cells expressing SST1 endogenously. All SSTs regulate ion channels, including potassium channels (ATP-sensitive, inward, and delayed rectifying), as recently shown using an elegant fluorescence-based membrane potential assay in pituitary cells. SST1 activation results in membrane hyperpolarization and subsequent reduction of Ca2+ influx through voltage-sensitive Ca2+ channels, as demonstrated in endogenously SST1-expressing insulinoma cells. Interestingly, SST1, but not the other SSTs endogenously expressed in the mouse pancreatic β-cell line MIN6, shows exclusive coupling with N-type voltage-sensitive Ca2+ channels, resulting in reduced intracellular Ca2+ concentrations and in inhibition of insulin secretion. Such SST specificities were also observed in pituitary tumor cells, where, in contrast to SST2, SST1 fails to stimulate phosphoinositide-specific phospholipase C (PLC) activity or PLC-dependent release of Ca2+ from intracellular stores, but transduces inhibition of phospholipase A2 activity and arachidonic acid release, similar to SST2. Additionally, SST1 (like SST3 or SST4, but not SST2 or SST5) inhibits sodium/hydrogen exchanger 1 (NHE1) activity via a PTX-independent mechanism, as demonstrated in SST1-transfected cells, resulting in decreased extracellular acidification that may be involved in inhibition of cell migration by SRIF. Hence, SST1, but not SST2, attenuated rat sarcoma (Ras) homolog (Rho)–GTP levels and subsequent Rock activity induced both by GPCR or integrin activation when expressed in Chinese hamster lung fibroblast cells (CCL39), and these inhibitory effects correlated with decreased actin stress fiber assembly and cell migration. Interestingly, a reported substrate of Rock is NHE1, which can serve as a plasma membrane-anchoring scaffold for actin filaments to control assembly of cortical stress fibers and focal adhesions. Because Rho inhibition by SST1 is PTX-independent, it may involve activation of Gα12, a trimeric Gα protein reported to inhibit both NHE1 and the Rho guanine nucleotide exchange factor p115 RhoGEF. Notably, NHE1 localizes at the invadopodia (membrane structures involved in cell invasion) of human malignant breast carcinoma cells, where it generates extracellular acidification necessary for invadopodial-dependent extracellular matrix degradation and tumor invasion.

28SRIF-induced increase of PTP activity was shown to be mainly involved in SRIF inhibitory effects on growth factor–stimulated cell growth. PTP activity was found in a membrane complex containing SRIF and SSTs. In pituitary tumor cells, SRIF-induced activation of protein phosphatases via PTX-sensitive G proteins correlates with endogenous SST1 expression. The rat membrane-associated protein tyrosine phosphatase η (PTPη) (homolog to human receptor tyrosine phosphatase type J, formerly known as DEP-1) transduces SRIF antiproliferative effects, in both insulin and/or TSH-treated thyroid PC C13 cells (which express all SSTs but predominantly SST1), and in glioblastoma cells, which express all five SSTs. In SST1-expressing heterologous cell systems, PTP-dependent inhibition of cell proliferation by SRIF was reported to rely on a complex interplay of different PTPs, comprising the receptor-like PTPη, which provides a long-lasting PTP activity (>2 hours), and the cytosolic SHP-2, which is rapidly activated (1–5 minutes). SHP-2 and PTPη are sequentially activated in a complex comprising the Janus kinase 2 (JAK2) that phosphorylates and activates SHP-2, which in turn activates (by dephosphorylation) the SRC proto-oncogene, nonreceptor tyrosine kinase (Src) that tyrosine phosphorylates and activates PTPη. The latter phosphatase is directly responsible for SRIF-mediated inhibitory effect on fibroblast growth factor (FGF)–stimulated proliferation through SST1. SRIF-activated PTPs inhibit cell proliferation by dephosphorylating tyrosine kinase receptors and/or downstream effectors, such as platelet-derived growth factor receptor, as demonstrated in SST1-expressing pancreatic cancer–associated fibroblasts, or by inducing cell cycle arrest via upregulation of p21 (cip1/Waf1) expression in SST1-transfected cells. In addition to SST3, SST1 and SST2 blunt FGF-induced nitric oxide production through inhibition of endothelial nitric oxide synthase (NOS) in Chinese hamster ovary (CHO) cells in a PTX-dependent manner.

29Sequence analyses of the rat Sstr1 gene promoter demonstrated presence of putative transcription factor binding sites [GC box transcription factor, specificity protein 1, and activator protein (AP)-2] that are often found in TATA-less promoters. Presence of binding sites for tissue-specific transcriptional factors of the POU domain protein family was also noted, including sites for pituitary-specific positive transcription factor 1 and POU family transcription factor Tst-1 that regulate tissue-specific rat Sstr1 gene expression in the pituitary and in pancreatic β-cells, respectively. The porcine Sstr1 gene promoter showed positive regulation by cAMP (through a CREBBP1 binding site), consistent with the cAMP-mediated upregulation of SST1 mRNA in rat pituitary primary cultures induced by GHRH treatment, and in pituitary adenomas expressing a mutated Gαs (gsp oncogene) that constitutively activates the cAMP pathway. SST1 mouse pituitary expression may also be controlled by testosterone because pituitary SST1 mRNA levels are decreased in gonadectomized males but restored upon testosterone injection, and are increased by testosterone treatment in rat pituitary tumor cells (GH4C1 cells). Such Sstr1 gene regulation reported in mouse, rat, or pig was also confirmed for the human SSTR1 gene. Finally, the SST1 gene promoter contains two CpG islands, putatively involved in head and neck squamous cell carcinoma tumorigenesis, where hypermethylation of the SSTR1, but also of SRIF, gene has been correlated with reduced disease-free survival.

30SRIF binding to its receptors results in internalization of receptor–ligand complexes, a critical process for receptor downregulation, resensitization, and signaling. Intriguingly, SST internalization may differ across species, explaining controversial results reported for SST1. In the rat insulinoma cell line 1046-38, which endogenously expresses SST1, a recombinant rat epitope-labeled SST1 was expressed to demonstrate that SST1 endocytosis is observed upon cell treatment with SRIF. This was also confirmed in other rat SST1-expressing heterologous cells. Interestingly, ligand-induced rat SST1 trafficking was dynamic, involving endocytosis followed by recycling, and then re-endocytosis of the receptor and of the intact and biologically active ligand, which are not directed to lysosomal degradation. In contrast, other studies showed that human SST1 expressed in heterologous cell systems demonstrates very slow, if any, internalization upon ligand binding, although it undergoes acute desensitization of adenylyl cyclase coupling that correlates with its phosphorylation status. Differences in rat and human SST1 internalization may be due to species-specific trafficking. Responsible for this species effect might be an amino acid change at a putative phospho-acceptor site (Thr383-Cys384-Thr385-Ser386) in the rat versus human SST1 C-terminal tail, where human Ser386 has been replaced by alanine in rat SST1. This substitution might explain the reported differences between rat and human SST1 affinity for β-arrestin-1 and subsequent trafficking. Confocal microscopy analyses showed bright immunoreactivity of both human and rat SST1 within the cytoplasm, both receptors accumulating the ligand (SRIF-14) into superficial compartments. Intriguingly, a fraction of SST1 stays clustered immediately beneath the plasma membrane, in as yet unidentified intracellular vesicular compartments. This peculiar localization may be caused by the absence of a domain in the SST1 N terminus required for cell surface targeting, as described for SST3. The SST1 cytoplasmic pool serves as a reservoir for short-term upregulation of human SST1 expression at the membrane upon prolonged agonist treatment. Upregulation depends on phosphorylation events at the SST1 C-terminal tail. SST1 immunoreactivity is observed both at the membrane and in the cytoplasm in primary and heterologous cell models, and in paraffin-embedded sections of diverse human tumor tissues, in contrast to SST2, which is predominantly membrane-associated.

31Unlike other SSTs, SST1 is not capable of homodimerization, prevented by structures within the C-terminal domain. SST1 was nevertheless found heterodimerized with SST5 in SST1- and SST5-coexpressing heterologous cell systems. Heterodimerization is induced by SST5- but not SST1-selective ligands and changes intracellular signaling (inhibition of forskolin-stimulated cAMP production) of the SST1/SST5 heterodimer as compared with SST5 homodimers or SST1 monomers. SST1 was also found heterodimerized with SST2 in prostate cancer cells; this complex stabilized with a bispecific (SST1/SST2) SRIF agonist, which nevertheless was less efficient than a mono-specific SST1 agonist to produce inhibition of cell proliferation. SST1, like other SSTs, harbors within its C terminus a PDZ-binding motif that interacts with membrane-associated guanylate kinase homologs, including PSD-95 and PSD-93, or synapse-associated protein SAP-97, involved in SRIF signaling to regulate neuronal growth cone stability in neurons (including retraction of filopodia and lamellipodia). Whereas members of the membrane-associated guanylate kinase homolog subfamily are believed to play a role as molecular scaffolds in the organization of postsynaptic signaling machineries, SAP97 is also prominently expressed in axons and presynaptic terminals, where it may be involved in SST1 presynaptic functions.

32Binding studies using iodinated SRIFs in mice deficient for each of the SSTs suggest that SST2 is most abundant in the murine CNS, although expression of other SST subtypes, including SST1, was confirmed by in situ hybridization in rat brain. Immunohistochemistry studies demonstrated that SST1 is highly expressed in the hypothalamic paraventricular and ARC, the median eminence, as well as other brain regions, including basal ganglia, basal forebrain regions, and hippocampus. SST1 has also been localized in SRIF-containing amacrine cells of rat and rabbit retina. SST1 immunoreactivity is also intense in the spinal cord, especially in dorsal horn and dorsal medulla. Peripherally, SST1 is expressed in neurons of mouse, rat, and human dorsal root ganglia (DRG), and on intestinal mucosal nerve fibers. Outside the nervous system, high expression of human SST1 mRNA is apparent in stomach, intestine, and endocrine pancreas (Fig. 5). Immunohistochemistry studies later confirmed that the SST1 protein is expressed mainly in these locations and also in the anterior pituitary. SST1 immunoreactivity is also positive in the parathyroid and bronchial glands, testis (staining in single cells between the tubules, resembling Leydig cells), and skeletal muscles. In the GI tract, SST1 is found expressed in stomach (Fig. 5), including antrum and corpus, in single cells resembling neuroendocrine or enterochromaffin-like cells, although expression in enterochromaffin cells (chromogranin-positive) was not always confirmed. Further SST1 expression was noted in the rectum and also in enteric ganglion cells and corresponding nerve fibers and nerve terminals. SST1 immunoreactivity was also detected in endothelial cells of blood vessels. In the immune system, SST1 protein is expressed in lymphocytes and macrophages, consistent with the positive SST1 immunoreactivity observed in cells resembling macrophages (cluster of differentiation 68 (CD68) positive) in spleen, lymph nodes, tonsils, thymus, lung, and gut mucosa, as well as in the stroma of diverse neoplasms.

33In neoplastic tissues, receptor autoradiography with SST1 subtype-selective ligands showed preferential expression of SST1 in prostate carcinomas and sarcomas. Using the rabbit mAb UMB-7 (Table 3), SST1 expression was confirmed in prostate adenocarcinomas, rhabdomyosarcoma and liposarcoma, pituitary somatotroph adenomas, pancreatic adenocarcinoma, stomach cancer, urinary bladder cancer, pheochromocytoma, GI neuroendocrine tumors (NETs), breast carcinoma, cervix carcinoma, and ovarian tumors. SST1 was abundantly expressed in bronchopulmonary NETs and positively associated with patient survival.

34In the CNS, SST1 immunoreactivity is primarily found in fibers and terminals morphologically similar to varicose axons and that exhibit the highest brain SRIF immunoreactivity or are closely apposed by SRIF-immunoreactive fibers. This observation suggests that SST1 is predominantly targeted to presynaptic compartments. In this position it negatively modulates release of SRIF itself or of hypothalamic releasing and release-inhibiting hormones, including GHRH in ARC neurons, where high SST1 mRNA concentrations are found. Accordingly, SST1 was defined as an inhibitory autoreceptor located on the mediobasal hypothalamus, basal ganglia, and retina SRIF neurons. Negative regulation of GHRH release by SRIF is consistent with the reported SST1 inhibitory role on hypothalamic regulation of GH secretion. Intriguingly, gender-related differences in both number and labeling density of SST1 mRNA-expressing cells are observed in the rat ARC (i.e., two- to threefold increase in males versus females). This observation may explain lower basal GH levels in male than in female mammals, and also the sexually dimorphic GH pulsatile secretion. More recently, the negative regulation by SRIF of GHRH neuron electrical activity was decrypted using a GHRH–GFP transgenic model. It revealed a sexual dimorphism, which is primarily attributable to a sex-dependent control of GABAergic and glutamatergic inputs by SRIF, rather than intrinsic differences in the GHRH neurons themselves. Interestingly, the positive glutaminergic neurotransmission onto GHRH neurons is an obligatory target of SRIF in female, providing a mechanism for a more tonic inhibition in female than in male, where this inhibitory signal was absent in one-third of animals. Intriguingly, this is the opposite for SRIF inhibition of GABAergic (negative) inputs, being especially robust and synchronized in males. Both SST1 and SST2 are involved in GHRH neuron rhythmicity, but SST1 receptors specifically transduce SRIF inhibitory control of GABAergic inputs, likely taking place at the presynaptic level. Seven percent of neuropeptide Y–positive neurons in the ARC coexpress SST1 mRNA, suggesting a direct interaction between the somatotropic axis and neuroendocrine regulatory loops of energy homeostasis. Hypothalamic paraventricular and ARC SST1 may account for prevention of acute stress-induced gut motor functions in mice after central injection of a SST1-selective agonist, including inhibition of gastric emptying and stimulation of colonic motility, putatively through central SRIFergic regulation of corticotropin-releasing factor (CRF) release and downstream stress-induced CRF actions. Recently, neuroanatomical connections between somatostatin and kisspeptin neurones were observed in the rat ARC and ventromedial hypothalamus, where one-third of kisspeptin neurones exhibit SST1 immunoreactivity. Because kisspeptin is a gonadotropin-releasing hormone secretagogue, these observations suggest that the regulation of kisspeptin release by SST1 may at least be partly involved in the well-known inhibition of gonadotropin-releasing hormone release by SRIF. In the basal ganglia (substantia nigra, nucleus accumbens, globus pallidus, and ventral pallidum), SST1 is also present presynaptically, where it negatively regulates SRIF release, strongly suggesting that it may serve as an autoreceptor to modulate systems regulated by SRIF (including dopamine). SST1 (together with SST2) are abundantly expressed in nerve processes of basal forebrain regions, including substantia innominata and the horizontal limb of the diagonal band, where SRIF inhibits glutamate release presynaptically through SST1, thereby regulating excitability of forebrain cholinergic neurons. Expression and function of SST1 in the rodent hippocampus have been controversial until the demonstration of SST1-mediated SRIF inhibitory action on synaptic transmission, using hippocampal slices of SST1 KO mice and a SST1 selective agonist. Hippocampal activity is regulated by SST1 through presynaptic inhibition of glutamate release induced by epileptiform treatment. In the spinal cord, SST1 may be involved in nociceptive transmission because dorsal horn and medulla regions coexpress SST1 together with SRIF, which has analgesic effects in rodents and humans. In the peripheral nervous system, SST1 may be involved in inhibitory effects of SRIF on inflammation and nociception (Pinter et al., 2006), such as in mouse models of stress-related visceral nociception or immune-mediated arthritis. In the retina, activation of SST1 with a selective ligand decreases SRIF release from retinal explants. Surprisingly, loss of SST1 expression in SST1 KO mice results in upregulated SRIF and SST2 retinal expression, together with an enlargement of axonal terminals of rod bipolar cells, where SST2 is expressed, as well as enhanced SST2 function. Conversely, in SST2 KO mice, SST1 expression is upregulated and rod bipolar cell axonal terminals are smaller. This suggests reciprocal inhibitory retinal roles of SST1 on SST2 expression, and vice versa. It certainly contributes to SRIFergic regulation of glutamatergic transmission along the vertical retinal visual pathway in which the SST2/SRIF receptor/ligand pair is probably restrained by SST1, consistent with reported SST1 autoreceptor functions.

35In the periphery, SST1 is expressed, together with SST5, in a high percentage of pancreatic β-cells, consistent with its reported role in regulating insulin secretion in studies using SST-selective agonists or in KO mouse models. In the anterior pituitary, the SST1-selective agonist CH-275 decreases GH secretion in wild-type, but not primary somatotroph cultures derived from SST1-KO mice, demonstrating the critical role for SST1 in regulating pituitary GH. SST1 was expressed in endothelial cells of normal human veins and arteries, including atherosclerotic arteries. SST1-selective agonists demonstrated consistent angio-inhibitory effects in vitro and induced vascular relaxation through cytoskeletal alterations, making SST1-specific analogs interesting for treatment of vascular diseases, including intimal hyperplasia. Intriguingly, another study localized SST1 mRNA and SST1 protein to vascular smooth muscle cells, where it showed acute upregulation of expression during vascular trauma coincidently with smooth muscle cell proliferation, making this receptor an interesting target to inhibit myointimal proliferation. The presence of SST1 in intestinal macrophages and mast cells, especially during inflammation, has been described in mice. Low SST1 expression was also found in macrophages differentiated from peripheral bone marrow cell–derived monocytes, where it mediates together with SST2 anti-inflammatory effects after activation by a multireceptor SRIF analog. During liver inflammation, such as in cirrhosis or hepatocellular carcinoma (HCC), all five SST mRNAs were expressed, whereas expression was not observed in normal human liver. The specific SST1 agonist L-797,591 was the only SST agonist to inhibit both liver cancer cell and hepatic stellate cell migration, making SST1 agonists putatively interesting to treat liver cirrhosis or HCC. Accordingly, SRIF also reduced production of collagens and inflammatory cytokines by hepatic stellate cells, although the specific receptor subtype was not identified. This putatively explains antifibrotic and immunomodulatory actions of SRIF in the liver. SST1 was also found uniquely expressed in stellate cells of pancreatic adenocarcinoma, and its activation by pasireotide reduced chemoprotective and prometastatic features of these fibroblastic cells by reducing IL-6 and collagen-1 secretion.

36SST1 is overexpressed in prostate cancer and mediates antiproliferative effects and inhibition of prostate-specific antigen release induced by the SST1-selective agonist BIM-23926 in prostate cancer cell lines. In the thymus, SST1 mRNA is expressed on isolated thymic epithelial cells, where SRIF inhibits proliferation. SST1 mRNA is not expressed in fresh human or rat thymocytes, but rat thymocyte activation with phytohemagglutinin or IL-1 selectively induced SST1, suggesting SST1 involvement in thymocyte proliferation and differentiation.

37The lack of available SRIF analogs with selectivity for SST1 stimulated the search for such compounds. The first to be identified is CH-275, which harbors a peptidic scaffold with selected amino acid deletions (des-aa1,2,5-SRIF) that in combination with DTrp at position 8, and 4-(N-isopropyl)-aminomethylphenylalanine (IAmp) at position 9, yields des-aa1,2,5-[DTrp8,IAmp9]-SRIF (CH-275), a SRIF agonist with nanomolar affinity for SST1, and that was 30-fold more selective for SST1 versus SST2/4/5 and 10-fold versus SST3, respectively. Using integrated combinatorial chemistry with high-throughput receptor-binding approaches, a SST1-selective nonpeptide compound (L-797,591) (Fig. 6; Table 4) displaying agonistic activity with an IC50 of 3 nM was the first pharmacological tool identified for selective SST1 in vitro and in vivo studies. BIM-23926 is a synthetic SST1-selective agonistic peptide (IC50 of 4 nM), reported to decrease cell viability of human medullary thyroid carcinoma TT cells, as well as to inhibit calcitonin release and cAMP levels (Fig. 6; Table 4). SRA880 is the first reported nonpeptide SRIF SST1 competitive antagonist, with high affinity for both native and recombinantly expressed SST1 from various species (Fig. 6; Table 4) (rat, mouse, monkey, human), while displaying low affinity for a range of other neurotransmitter receptors, except the dopamine receptor D4. The compound is bioavailable and brain penetrant. Consistent with the inhibitory autoreceptor role of SST1, SRA880 administration increases SRIF brain release and signaling, countering depressive-like symptoms in mice. A series of SST1-selective ergoline derivatives has been developed, some of which show effective oral bioavailability and brain penetration. The only clinically approved SRIF analog showing high affinity for SST1 is pasireotide, a nonselective peptidic compound displaying an IC50 of 9.3 nM for SST1, and IC50 values for SST2, SST3, and SST5 of 1, 1.5, and 0.16 nM, respectively.

38cDNAs coding human and mouse SST2 were isolated, together with cDNAs coding for SST1, in early 1992 using a polymerase chain reaction–based approach with primers directed to regions conserved in all GPCRs. In the same year, a cDNA encoding for rat SST2 was identified by expression cloning from a rat brain cDNA library. The gene coding human SST2 is localized on chromosome 17q25.1 and consists of two exons. Whereas exon 1 contains only 5′UTR, the entire coding region and 3′UTR are located on exon 2. Genes encoding for mouse and rat SST2 are localized on chromosomes 11 E2 and 10q32.1, respectively. Homology between human and rodent SST2 is 94% (mouse) and 93% (rat) at the amino acid level, respectively. In all three species, SST2 is a 369-amino-acid protein (Mr = 41,305 in humans), displaying typical seven-transmembrane segments and four putative N-glycosylation sites (Asn-9, Asn-22, Asn-29, and Asn-32) (Fig. 7). In Western blot experiments, the protein is detected as a characteristic smear migrating between 70 and 80 kDa, in keeping with the assumption that these Asn residues are extensively glycosylated. In addition to this long receptor species, a mouse SST2 splice variant codes for a shortened receptor with an alternative C-terminal tail (termed SST2B). This variant arises due to removal of a part of the second exon that codes for the C-terminal tail of the long (SST2A) variant, and that also contains some 3′UTR. Some of the remaining 3′UTR of SST2A then becomes the coding region in SST2B mRNA; in contrast to SST2A, the C-terminal sequence of SST2B is very poorly conserved between species. Several functionally relevant elements of the SST2A C terminus are lost due to the alternative splicing event: 1) elimination of phosphorylation sites that contribute to agonist-dependent desensitization and internalization and 2) SST2A contains a C-terminal consensus motif for binding PDZ domains, which is not present in in SST2B. Whereas both spliced forms have been identified in rodents, human tissues exclusively contain the unspliced SST2A variant. Consequently, we use the denomination SST2 for the long unspliced SST2A variant throughout this review.

39Signaling properties of SST2 have been investigated in several heterologous expression systems, or in endogenous SST2-expressing cells using SST2 agonists. As physiologic actions of SRIF (e.g., on GH release) are sensitive to PTX, unsurprisingly, SST2 also acts mostly through PTX-sensitive G proteins of the inhibitory/olfactory family of Gα subunit (Gi/o) type, as reported in cultured mammalian cells heterologously expressing SST2, such as CHO, human embryonic kidney (HEK)293 cells, or SV4-transformed fibroblast-like derived monkey kidney (COS-7) cells. Major effects of SST2/Gαi/o signaling are inhibition of adenylyl cyclase, inhibition of voltage-gated calcium channels, and activation of Kir3.x (Fig. 8). The inability of SRIF-bound SST2 to inhibit cAMP production in some cell lines (e.g., ) was ascribed to lack of expression of an appropriate G protein. All of these effects are complementary for the inhibition of excitable cells, such as neurons or hormone-secreting cells. Thus, whereas hormone secretion by pituitary cells is driven by Ca2+ influx through voltage-gated Ca2+ channels, activation of potassium channels by SRIF-activated SST2 hyperpolarizes the membrane and prevents depolarization induced by hypothalamic releasing hormones. In pituitary cells, SST2 activates PLC [more specifically the PLC-β3 isozyme, which is partially blocked by PTX, putatively involving a Gαq/11 and/or Gβγ protein]. This results in inositol-1,4,5-trisphosphate formation and Ca2+ release into the cytoplasm from the endoplasmic reticulum, regulating the MAPK/extracellular signal-regulated kinase (ERK) pathway.

40Besides these canonical G protein–mediated signaling pathways, SST2 also activates tyrosine phosphatase activity, associated with reduced serum-stimulated cell proliferation. This effect has been replicated in several cell types either overexpressing the receptor, or expressing lower endogenous levels of SST2. Activation of tyrosine phosphatases by SRIF is PTX-sensitive and can be mimicked by addition of Gαi/o subunits purified from brain, implicating this G protein in the SST2-initiated signaling pathway. The nonreceptor tyrosine protein phosphatases SHP-1 and SHP-2, respectively, have emerged as major effectors, being copurified with SST2 (or SSTs in general) in heterologous cell systems, and sequentially activated, dependent on G protein (Gαi3 and Gβγ) and Src activity. Through phosphatase activation, SST2 inhibits tyrosine phosphorylation events such as those following activation of tyrosine kinase receptors. This leads to cell cycle arrest and subsequent inhibition of cell proliferation, through upregulation of the cyclin-dependent kinase inhibitor p27/Kip1 and the zinc finger protein (Zac1), as mainly demonstrated in heterologous cell systems. Furthermore, similar results were obtained using SST2-preferring or selective analogs in endogenously SST2-expressing cells [e.g., pituitary tumors, insulinoma, glioma, normal and tumoral pancreatic acinar cells, or thyroid cells]. This involves regulation of several signaling pathways, including Ras/Raf/ERK, phosphatidylinositol-4,5-bisphosphate 3-kinase (PI3K)/AKT serine/threonine kinase 1 (AKT)/glycogen synthase kinase 3β/mechanistic target of rapamycin kinase (mTOR), p38, neuronal NOS, and JAK2. Depending on phosphatase activity, SST2 activation also triggers apoptosis in endogenous SST2-expressing pituitary somatotroph tumor cells, and also in pancreatic cancer cells engineered to express this receptor, where apoptosis is further stimulated by treatment with death ligands. Finally, SST2-induced dephosphorylation events can also lead to inhibition of cell migration and invasion, through inhibition of the small G protein Rac and of the subsequent ruffle formation in endogenous SST2-expressing neuroblastoma cells. In addition, SST2-induced dephosphorylation events can lead to restoration of cell-to-cell (adherens and gap junctions) and cell-to-matrix (hemidesmosomes) contacts in SST2-transfected pancreatic cancer cells.

41Similar to genes encoding the other SSTs, genes coding SST2 from various species do not contain TATA and CAAT boxes. A minimal promoter fragment close to the transcription initiation site was identified as a novel initiator element sufficient to account for transcription from the SST2 promoter in neuroblastoma cells. Interestingly, a so-called enhancer box was identified in this region, which serves as a binding site for the basic helix-loop-helix transcription factor (SEF-2). Interaction of SEF-2 with the enhancer box was identified as a major driving force for activity of the promoter in several cell lines. Transcriptional activity is further enhanced by the binding of c-myc intron binding protein 1 (MIBP1), both to SEF-2 and to a thymine-cytosine–rich transcriptional enhancer element; the expression pattern of MIBP1 matches that of SST2 in the murine brain, suggesting that MIBP1 confers expression tissue specificity at least in the CNS. In addition, negative regulatory elements have been identified in more distal regions of the promoter of the mouse Sstr2 gene. Suppression of transcriptional activity by this region may be overcome by Smad3/Smad4, which plates the activity of the Sstr2 gene under control of a transforming growth factor β–dependent signaling pathway. Finally, and relevant for tumors that escape SST2-mediated antiproliferative effect of SRIF and its analogs, an alternative 5′/upstream promoter was identified that may be silenced by methylation. High methylation levels in this region correlated with reduced SST2 expression in tumor cells.

42Regulation of SST2 depends upon molecular mechanisms, implying phosphorylation events at the C-terminal tail followed by recruitment of β-arrestins and receptor endocytosis. Combined biochemical and mutagenesis approaches identified serine and threonine residues in the C-terminal tail of the SST2 that are phosphorylated upon SRIF and/or octreotide stimulation, namely, S341, S343, S348, T353, and T354 in rat SST2-transfected CHO and GH4C1 cells, and S341, S343, T353, T354, T356, and T359 in HEK293 cells stably transfected with rat or human SST2. Agonist-dependent phosphorylation of the four threonine residues was also documented in rat pituitary tumor cells (GH3) transiently transfected with rat SST2, rat pancreatic insulinoma β-cells (INS1 cells), which endogenously express SST2, and rat pancreas in vivo. Although rat SST2 internalization was partially inhibited by mutation of threonine residues, none of the mutations resulted in a complete block of receptor internalization. By contrast, multisite phosphorylation of clusters of carboxyl-terminal serine and threonine residues of the human SST2 cytoplasmic tail is a critical event for receptor endocytosis. Accordingly, in a SRIFoma, which synthesized and secreted SRIF and in which SST2 are localized intracellularly, receptors are phosphorylated, whereas in an ileal carcinoid tumor in which SST2 are membrane-bound, receptors are found nonphosphorylated. In human NET samples, SST2 phosphorylation is observed only in octreotide-treated patients and receptors are internalized, whereas in untreated tumors SST2 are not phosphorylated and are located at the cell membrane. Of note, unlike SRIF and octreotide, pasireotide (formerly known as SOM230) stimulates only phosphorylation of S341 and S343 residues of human SST2, followed by a partial receptor internalization. In cell lines, G protein–coupled receptor kinase (GRK2) or GRK3 is involved in phosphorylation of S341 and S343 residues, whereas the threonine residues (T353, T354, T356, and T359) are phosphorylated by GRK2 and GRK3. In HEK293 cells stably expressing SST2, chemical protein phosphatase inhibitors and small interfering RNA knockdown screening lead to identification of protein phosphatase 1β (PP1β) as the GPCR phosphatase that catalyzes rapid dephosphorylation of residues T353, T354, T356, and T359.

43SST2B terminates after residue 332 and therefore does not contain the phosphorylation sites identified in the C-terminal tail of SST2. Accordingly, SST2B phosphorylation is not detectable after agonist stimulation of colonic adenocarcinoma cells, whereas SST2 is phosphorylated under the same conditions. SRIF causes rapid desensitization of SST2, but not of SST2B, in this latter cell type. However, both receptor subtypes desensitized markedly in transfected CHO cell line subclone K1 (CHO-K1) cells. Interestingly, phosphorylation in the third ICL of SST2, a sequence shared by the SST2B variant, also occurs, but does not play a role in internalization and desensitization. Differences in receptor phosphorylation might be physiologically relevant in the rodent brain and the GIT, where different expression patterns of the two SST2 subtypes have been documented.

44The relationship between SST2 phosphorylation and β-arrestin binding, a major class of adaptor proteins involved in GPCR desensitization and internalization, has been investigated in different cell lines. In HEK293 and CHO cells, or primary hippocampal neurons, cotransfected with β-arrestin enhanced green fluorescent protein and the rat SST2, both β-arrestin-1 and β-arrestin-2 are recruited to the plasma membrane after agonist stimulation, form stable complexes with the receptor, and internalize together. β-arrestin-2 recruitment also occurs after agonist stimulation of human SST2 in HEK293 cells. Together, these results suggest that the SST2 belongs to the class B GPCR subgroup, because its activation results in robust recruitment of both β-arrestin-1 and -2.

45Studying SST trafficking has received increasing attention because the fate of internalized receptors, following agonist exposure, may vary from degradation to rapid recycling to the plasma membrane, thereby affecting responsiveness to endogenous ligands and drugs of therapeutic interest. A striking SST2 property is that in the vast majority of the cell types endogenously expressing this subtype, it is almost exclusively confined to the plasma membrane, such as in central and myenteric neurons, neuroendocrine cells of the gastric antrum, anterior pituitary, pancreatic islets, as well as central and peripheral tumors. However, intracytoplasmic SST2 localization was observed in CNS in regions exhibiting dense SRIF innervation such as the central nucleus of the amygdala. Intracellular localization was also described in a rat model of middle cerebral artery occlusion in cerebrocortical neurons adjacent to the infarct, which regionally correlates with transient SRIF depletion from axonal terminals. Subcellular distribution of the receptor may be dependent on surrounding SRIF concentrations, as suggested in tumors of the nervous and neuroendocrine systems (pheochromocytomas and neuroblastomas) producing autocrine SRIF. Collectively, these studies highlight that SRIF released under physiologic or pathophysiological conditions regulates localization and trafficking of SST2 consistent with results obtained in cell lines. In hippocampal neuronal cells, SST2 trafficking was analyzed in detail at different times after acute intracerebral octreotide injections or in primary neuronal culture exposed to SRIF ligands. These experiments demonstrated for the first time that GPCR cargoes recycle through the trans-Golgi network (TGN) after endocytosis. After activation and internalization, endosomes bearing SST2 in dendrites (by far the major pool of SST2) and cell bodies fuse and migrate to a perinuclear compartment expressing trans-Golgi markers such as the integral protein of the TGN, TGN38, and syntaxin-6, but not cis-golgi markers such as cis-Golgi marker 130 (GM130). These results have been comfirmed in vivo by electron microscopy approaches that showed that SST2 cargoes were not targeted to degradative departments; rather, TGN-enriched receptors recycle to the plasma membrane (dendrites and cell bodies), where they are observed in preagonist challenge equivalent amounts. The recycling process, which depends on the length and extension of dendritic arborization, is slow, 3–6 hours in vitro and 24–48 hours in vivo. Differences in kinetics between in vivo and in vitro studies might be, at least in part, due to the persistence of the intracerebrally injected agonist, which cannot be removed or chased as in in vitro settings, implying several internalization/recycling cycles before total agonist clearance or degradation. Such trafficking of activated SST2 to the TGN was also reported in myenteric neurons, in which an intact TGN is necessary for receptor recycling. In both DRGs and dorsal horn neurons, octreotide-activated SST2 in vivo also are observed to concentrate in perinuclear regions that resemble the TGN before recycling (24 hours). The physiologic significance of this peculiar recycling pathway is not fully understood. Recycled receptor might undergo biochemical modifications and/or association with scaffolding proteins for proper delivery to the cell surface. An additional intriguing hypothesis is that SST2 targeted to the TGN could produce downstream cellular responses, such as coupling to different G proteins, as demonstrated for other GPCRs, the sphingosine 1-phosphate receptor, and the TSH receptor.

46Recently, studies have focused on factors involved in regulation of intracellular SST2 trafficking. Modulating recycling of a particular receptor can indeed impact its physiologic fate and therefore offer a potential therapeutic value. Using pharmacological and cell biologic approaches, it was demonstrated that leucyl-cysteinyl aminopeptidase (LNPEP; formerly known as insulin-regulated aminopeptidase) ligands accelerate recycling of internalized SST2 in neurons in vitro or in vivo. LNPEP, which shares common regional and subcellular distribution with internalized SST2, was shown to be involved in vesicular trafficking. Importantly, because LNPEP ligands increase the density of SST2 at the plasma membrane, they also potentiate SRIF-inhibitory effects on seizure activity. LNPEP therefore represents a therapeutic target for treatment of limbic seizures and possibly for other neurologic conditions in which downregulation of GPCRs occurs. In myenteric neurons, activated SST2 traffic to endothelin-converting enzyme 1 (ECE-1)–containing vesicles and TGN. This endosomal peptidase degrades peptide ligands in intracellular organelles and promotes receptor resensitization. SST2 recycling (30 minutes) in myenteric neurons is dependent upon endosomal acidification, ECE-1 activity, and an ECE-1 cleavable ligand, which is the case for SRIF-14, but not for SRIF-28 or analogs, such as octreotide. Hence, after activation by ECE-1–resistant SRIF-28 and analogs, SST2 remain within the TGN and are poorly recycled at 120 minutes. Assuming that SST2 signals in intracellular organelles, this could explain, at least in part, the long-lasting actions of SRIF analogs such as octreotide. Of note, although ECE-1 might be present in the hippocampus, the kinetics of SST2 recycling were the same after activation by SRIF-14 or octreotide in hippocampal neurons, suggesting that the ECE-1 role on SST2 recycling is dependent upon cell types. Filamin A (FLNA), a scaffolding protein involved in intracellular trafficking of several transmembrane proteins, has been shown to interact with the SST2 in melanoma and pancreatic cell lines. FLNA appears crucial for SST2 stabilization and signaling at the plasma membrane. In addition, FLNA may protect SST2 from degradation by facilitating targeting of the receptor to a recycling pathway during long-term agonist treatment of pancreatic tumor and GH-secreting tumor cells.

47Interestingly, all SST subtypes (SST1–SST5; with the exception of the short-splice variant SST2B) contain a consensus motif for interaction with type I PDZ domains at the intracellular C termini. This motif is conserved throughout evolution, as the closest homologs of SSTs in Drosophila, the Drostar1 and Drostar2 receptors for type C allatostatins, also contain a PDZ ligand motif. In contrast, closely related opioid receptors are devoid of such a motif. In its simplest form, a type I PDZ ligand consists of the C-terminal sequence –S/T–X–φ–COOH, in which φ is a large hydrophobic residue (Phe in SST4; Ile or Leu in SST1, 2, 3, 5; Val in many other typical PDZ ligand motifs). Whereas such a motif is quite common, flanking sequences add to the specificity of PDZ-type interactions, thus ensuring that not every PDZ ligand can interact with any PDZ domain. SST2 was the first SST for which an interaction with a PDZ domain–containing protein was reported; identified members of the Shank protein family as potential interactors through yeast two-hybrid screening. Shank proteins are important scaffold proteins of the PSD. They exhibit a complex domain structure, as the central PDZ domain is accompanied by a Ras association domain, a set of seven ankyrin repeats, and a sterile alpha motif (SAM domain) and a nuclear localization signals 1 domain. In addition, a long proline-rich stretch is involved in binding actin-binding proteins. In the PSD, Shank proteins are considered as master scaffold proteins that link receptor complexes to the actin-based cytoskeleton. As SST2 does not appear to be a postsynaptically enriched receptor, it appears likely that interactions between Shank proteins and SST2 are relevant at other, nonsynaptic sites. In further studies, PDZ domain–containing 1 (PDZ-K1) protein was identified as interaction partner for all SST subtypes, including SST2. As PDZ-K1 also interacts with a PLC isoform, this work suggests that PDZ-K1 allows for coupling of SST subtypes to PLC through ternary complex formation. Finally, C-terminal PDZ ligand motifs of a larger number of membrane proteins have been shown to promote postendocytic recycling through binding to the PDZ domain containing sorting nexin family member 27, and the PDZ ligand of SST2 may also promote similar recycling.

48Additionally, SST2 harbors two immunoreceptor tyrosine-based inhibition motif (ITIM) sequences (immunoreceptor tyrosine–based inhibitory motif: I/V/L/S-x-Y-x-x-L/V), present in the third ICL and C-terminal tail. Such ITIM consensus sequences were initially found in inhibitory immunoreceptors (e.g., programmed cell death protein 1), triggering B cell receptor inhibition of SHP-2 recruitment, and subsequent dephosphorylation of B cell receptor effector molecules. Similarly, SRIF-induced phosphorylation of SST2 tyrosine 228 and 312 residues, present in each of the two SST2 ITIMs, triggers SHP-2 (but not SHP-1) direct recruitment to SST2, SHP-2 activation, and subsequent transduction of dephosphorylation events also involving the kinase Src and SHP-1, leading to cell proliferation inhibition.

49The first SST2 ICL contains two juxtaposed binding sites for the p85 regulatory subunit of PI3K and for the actin-binding and scaffolding protein FLNA. Depending on SST2 phosphorylation of tyrosine residues 66 and 71, present in FLNA and p85 binding sites, respectively, FNLA or p85 competitively binds to SST2 first ICL. In the absence of SRIF, these tyrosine residues are phosphorylated, enabling p85, but not FLNA, binding. This state is permissive for growth factor–induced activation of PI3K activity. In the presence of SRIF, SST2 tyrosine residues 66 and 71 are dephosphorylated, enabling FLNA, but not p85 binding, the dissociation of which from SST2 triggers PI3K inactivation. One hypothesis is that SHP-2 binding to SST2, induced by SRIF through ITIM phosphorylation on SST2 third ICL and C-terminal domain, triggers dephosphorylation of FLNA and p85 binding sites in SST2 first intracellular domain. Alternatively, the scaffolding FLNA protein, once recruited onto SST2 in the presence of SRIF, brings SST2 in proximity to the phosphatase SHP-1, whose activity is critical to trigger SRIF-inhibitory effects on cell proliferation, migration, or invasion. Both rat and human SST2 form constitutive homodimers. The dynamics of ligand-induced trafficking have also been studied for pig SST2. This receptor forms constitutive homodimers/multimers in the absence of ligand, which rapidly dissociate (11 seconds) upon SRIF binding. Interestingly, in contrast to human SST2, pig SST2 rapidly reassociates (110.5 seconds) during a subsequent process that temporally overlaps with receptor internalization (half-maximal 95.1 seconds). When coexpressed heterologously, SST2 and SST3 form heterodimers with reduced SST3 activity. However, to what extent SST2 forms dimers or oligomers with other GPCRs in vivo is not known.

50Two independent laboratories using two different antibodies directed toward the carboxy-terminal tail of the SST2 have demonstrated that this SST subtype is the most abundant SST in the rodent CNS in agreement with both in situ hybridization experiments and autoradiographical studies using SST2-preferring ligands. Strong SST2 labeling is detected in the deep layers of the cerebral cortex, CA1 field, and dentate gyrus of the hippocampus, lateral septum, medial septum/diagonal band of Broca, medial habenula, bed nucleus of the stria terminalis, endopiriform nucleus, claustrum, amygdaloid complex, locus coeruleus, and nucleus tractus solitarius. In the hypothalamus, the highest densitiy of SST2 immunoreactivivity is located in the ARC and the medial tuberal nucleus as well as in the lateroanterior nucleus and the ventrocaudal part of the tuber cinereum. In the rat spinal cord, SST2 neurons are localized in the superficial layers of the dorsal horn often closely apposed by SRIF-immunoreactive terminals. In the rat retina, several neuronal cell types express SST2. In the outer layers, immunoreactivity is localized to cone photoreceptors, horizontal cells, and rod and cone bipolar cells. In the inner layers, SST2 immunostaining is present in numerous medium- to large-size amacrine cells. Regional distribution of SST2 immunostaining in the human CNS is generally congruent with that reported for the rat, although, in contrast to rodents, human cerebellum displayed significant SST2 immunostaining.

51In the rat peripheral nervous system, medium-size neurons distinct from those expressing SRIF display SST2 in the DRG. In the human and rat GIT, SST2 are localized in neurons of the myenteric and submucosal plexuses, and in fibers distributed to the muscle, mucosa, and vasculature.

52Immunohistochemical experiments have examined localization of SST2B. In the rat brain, somatodendritic labeling is evident in several regions that also exhibit SST2 immunostaining, including the olfactory bulb, cerebral cortex, hippocampal formation, septal nuclei, and superior colliculi. In contrast to SST2, the Purkinje cell layer of the cerebellum appears to be SST2B immunoreactive. In the rat spinal cord, whereas the SST2 is confined to the superficial layers, SST2B is located in neuronal perikarya and proximal dendrites throughout the gray matter of the spinal cord.

53Using specific antibodies against SST2, it appears that this receptor is largely distributed in the adult rat and human pituitary (Fig. 9). Although all anterior pituitary cell types express the SST2 protein, GH-expressing cells almost completely colocalize with the SST2, whereas 50% of gonadotrophs, 60% of corticotrophs, 30% of thyrotrophs, and 10% of prolactin cells exhibit SST2 immunoreactivity. Of note, some discrepancies with distribution of the mRNA exist because the SST2 mRNA was previously found in 40% of somatotrophs, 36% of thyrotrophs, 26% of lactotrophs, 3% of corticotrophs, and 8% of gonadotrophs.

54Immunohistochemical studies also revealed the presence of the SST2 in striated ducts of the parotid gland, in neuroendocrine and enterochromaffin-like cells of the GI mucosa, in enteric ganglia, in insulin- and glucagon-secreting cells of the pancreas (Fig. 9), in the reticular zone of the adrenal cortex, in glomeruli and tubules of the kidney, in luteinized granulosa cells of the ovary, in basal parts of testicular tubules, in granulocytopoietic cells of the bone marrow, in alveolar macrophages of the lung, and in germinal centers of lymph follicles. All in all, SST2 was the most frequently detected SST subtype, and, in most cases, SST2 predominantly immune-stained at the cell plasma membrane.

55As determined using the rabbit mAb UMB-1, SST2 also represents the most prominent SST subtype in tumor tissues, with only few exceptions, such as pituitary adenomas. In somatotroph and thyrotroph pituitary adenomas, SST2 is present at high abundance in >80% of tumor specimens, along with a high presence of SST5. In contrast, in gonadotroph, corticotroph, and nonfunctioning pituitary adenomas where SST2 expression was low or even absent, a preponderance of SST3 or SST5 has been noted. In brain tumors, the prevalence of SST2 was highest. SST2 was present in most of meningiomas, medulloblastomas, neuroblastomas, and supratentorial primitive neuroectodermal tumors of childhood. SST2 was also detected in oligodendrogliomas, but much less frequently in astrocytomas. Furthermore, noticeable SST2 expression occurred in peripheral nerve sheath tumors, especially in schwannomas.

56SST2 was present in 43%–66% of medullary as well as in papillary and follicular thyroid carcinomas (;; Mussig et al., 2012;;;; ). SST2 was also detected in a high percentage (>70%) of pheochromocytomas and paragangliomas as well as in 33% of both functioning and nonfunctioning adrenocortical adenomas. Very low levels of SST2 were also observed in lymphomas. SST2 represents by far the most prominent SST subtype detected in gastroenteropancreatic (GEP)-NETs, and, overall, it was identified in >70% of cases at a high expression intensity. However, SST2 was detected more frequently in gastrinomas (100%) and in carcinoid tumors (86%) (Fig. 9) than in insulinomas (58%). SST2 expression was also found to be higher in functioning than in nonfunctioning tumors and more pronounced in gastroenteric than in pancreatic neoplasms. Negative correlation between SST2 expression and tumor grading or proliferation rate and a positive association with patient outcomes have been shown. Furthermore, a positive correlation between SST2 expression and SST-based imaging was demonstrated. SST2 was detected in 32%–56% of bronchopulmonary NETs. However, in comparison with GEP-NETs, SST2 expression was less pronounced. Lower SST2 expression in high-grade in comparison with low-grade tumors and a positive correlation with SST-based imaging were shown. SST2 expression was also observed in NETs of other origins, including thymus, breast, cervix, or prostate. Furthermore, SST2 was detected in 88%–100% of GI stromal tumors (GIST), and, also in this tumor entity, an association with favorable patient outcomes was demonstrated. Depending on tumor grade and location, SST2 was observed in 45%–100% of colorectal carcinomas and in 41%–67% of HCCs; SST2 was expressed in 20%–79% of breast cancers, in 57% of cervical carcinomas, in 39% of endometrial cancers, and in 30% of ovarian carcinomas. Furthermore, a moderate to strong SST2 expression was observed in 13% of prostate cancers in general and in 50% of prostate cancers with endocrine differentiation. SST2 was detected in 59% of Merkel cell carcinomas and in melanomas. Finally, SST2 expression in normal exocrine pancreatic tissue is progressively lost during pancreatic ductal adenocarcinoma progression, which participates in tumor aggression, as demonstrated in mouse models of pancreatic cancer combined with SST2 KO mice. Accordingly, in vitro and in vivo re-expression of SST2 in human pancreatic cancer cell lines through SST2 cDNA transfection and through in vivo SST2 gene transfer in mouse models and in a first-in-man phase I clinical trial, respectively, demonstrated promising oncosuppressive activity in advanced pancreatic cancer.

57SRIF was originally described as an inhibitor of GH release, but it also inhibits secretion of other pituitary hormones. SST2 is predominantly responsible for regulation of physiologic secretion of GH and TSH indirectly mediated by opening of K+ channels. The subsequent K+-derived membrane hyperpolarization and reduction of L- and N-type Ca2+ influx as well as intracellular Ca2+ concentration are major mechanisms by which SRIF, through the SST2 subtype, acutely inhibits exocytosis of hormone-containing vesicles. SST2 also inhibits exocytosis of hormone-containing vesicles derived from pancreatic α- and β-cells. In rodents secretion of glucagon and in humans secretion of both glucagon and insulin are regulated by SST2. In addition, the SST2 subtype activates Kir3.x, which leads to hyperpolarization and inhibits voltage-gated P/Q-type Ca2+ channels.

58Another major role of SST2 is inhibitory neuromodulation. Effects of SST2 on neuronal excitability have been studied in several CNS cell populations. Activation of SST2 in medial septal GABAergic neurons results in decreased discharge rate and consequent reduction of hippocampal θ rhythm power. SST2 also mediates hyperpolarization of dorsal horn neurons and subsequent antinociceptive effects. In the ventrolateral medulla, SST2 activation of presympathetic neurons provokes robust sympathoinhibition with bradycardia and hypotension. Rodent and sheep gonadotropin-releasing hormone neurons are inhibited by SST2, which results in decreased luteinizing hormone secretion. Activation of SST2 in olfactory bulb mitral cells modulates dendrodendritic inhibition between mitral and granule cells, which in turn results in increased γ oscillation power of mitral cells and increased odor discrimination performances. The role of SRIF in neuronal excitability has been mostly studied in the hippocampal formation. In CA1 pyramidal neurons, SRIF has hyperpolarizing effects through activation of K+ channels.

59Postsynaptic hyperpolarization of CA3 pyramidal neurons by SRIF has also been demonstrated. Presynaptic inhibition of glutamate release by SRIF reduces excitatory synaptic input on CA1 neurons; thus, SRIF decreases both post- and presynaptic hippocampal pyramidal cell excitability. By contrast, in the dentate gyrus, SRIF has no effect on granule cell postsynaptic currents or firing properties. SRIF, however, inhibits postsynaptic N-type Ca2+ channels in granule cells, resulting in inhibition of long-term potentiation, a form of synaptic plasticity, critical in learning and memory, and also plays an important role in epileptogenesis. Converging evidence suggests that, of the five SSTs, SST2 exerts a predominant role in transduction of SRIF actions in the hippocampal formation. SRIF plays a prominent role in epilepsy in agreement with its inhibitory neuromodulatory function, and SST2 mediates most of antiepileptic actions of SRIF in rats, likely in humans, but not in mice.

60Striatal SST2 receptors are involved in control of extrapyramidal motor systems, as activation of SST2 receptors in rats increases locomotor activity, whereas disruption of the SST2 receptor gene in mice impairs motor functions.

61In accordance with the widespread distribution of SST2 in the hypothalamus, the receptor also plays a role in drinking and feeding behavior. Activation of SST2 increases food intake by suppressing satiety (i.e., a mechanism delaying onset of another meal after a completed one), but not satiation (a mechanism causing meal termination). Increased meal numbers mediated by SST2 activation likely involve lateral hypothalamic orexinergic-A neurons projecting to the arcuate neuropeptide Y neurons that express orexin receptors 1. SST2 actions on orexinergic neurons, however, seem to be indirect. Activation of SST2 also increases rapid-onset water consumption. This dipsogenic function involves activation of the angiotensin II receptor type 1 signaling system. SRIF release in the hypothalamus follows a circadian rhythm, with the highest release at the beginning of the dark phase in rats. Early nocturnal drinking and feeding in rats are therefore physiologically regulated by SST2 signaling.

62Stress responses are inhibited by SRIF, and SST2 plays a major role in inhibition of acute stress induced at several levels. First, stress-related endocrine responses are inhibited by SST2 at both the CNS and pituitary levels. Second, stress-related sympathetic activation is inhibited by SST2 at the level of brainstem presympathetic neurons. Finally, behavioral stress responses, such as suppression of food intake and anxiety, are also inhibited by SST2. In addition to anxiolytic effects, SST2 also mediates antidepressant actions of SRIF.

63In the retina, SST2 inhibits adenylyl cyclase, K+/Ca2+ conductances, as well as activates guanylyl cyclase and NOS, and plays an important role in positive control of dopamine and negative control of glutamate release. SST2 signaling through these diverse intracellular pathways converges into an important retinal neuroprotection. Therefore, SRIF administration is a promising therapeutic approach in treating retinal diseases involving ischemia and excitotoxicity, and a multicentric, phase II–III, randomized, controlled clinical trial (EUROCONDOR-278040) is underway to assess the efficacy of SRIF administration in diabetic retinopathy.

64Early efforts to develop stable SRIF analogs with potent inhibitory activity on GH release have led to the synthesis of SST2-preferring peptide ligands, two of which, octreotide and lanreotide, have later been approved for clinical use. Both exhibit potent (subnanomolar) full agonistic properties at the SST2 and modest activity (low nanomolar) at SST5. In clinical practice, however, octreotide and lanreotide fully normalize GH and IGF-1 in only ∼50% of unselected acromegalic patients, which has stimulated the search for new SRIF analogs. This has led to recent development of pasireotide, which exhibits affinity for multiple SSTs. Pasireotide is particularly potent at the SST5, but binds with modest affinity to SST2, where it exhibits only partial agonistic activity. BIM-23120 is a highly selective SST2 peptide agonist often used to study SST2 activity. L-779,976 was the first selective nonpeptide agonist (Fig. 10; Table 5) with high SST2 selectivity. However, further development of L-779,976 was halted because of low oral bioavailability. Abundant SST2 expression in human tumors has stimulated a continued search for orally available SST2-selective agonists such as L-054,264, RFE-007, as well as novel β-methyltryptophan derivatives. Several peptide antagonists of SST2 have also been identified, such as BIM23627 and BIM-23454. At high concentrations these compounds exhibit some residual agonist activity and are thus weak partial agonists. Among available SST2 peptide antagonists, JR11 is currently the most potent and selective one (Fig. 10; Table 5).

65The mouse SST3 was cloned as the third member of the SST family, and cloning of the human SST3 was reported shortly thereafter. The gene encoding for human SST3 is localized on chromosome 22q13.1 and spreads over eight exons. However, the entire coding region is localized in a single exon and encodes a protein of 418 amino acids (Fig. 11). The SST3 protein sequence shares 46% homology with SST2. Analysis of the sequence showed two potential N-glycosylation sites located in the amino-terminal domain at Asn17 and Asn30. The genes encoding mouse and rat SST3 are localized on chromosome 15 E1 and 7q34, respectively; both encode a protein of 428 amino acids. In rat SST3, Asp124 is essential for binding of the endogenous peptide ligand SRIF-14. A unique feature of mouse and rat SST3 is selective targeting to primary neuronal cilia dependent on the presence of a conserved ciliary targeting motif within the third ICL (243APSCQWVQAPACQ255). This sequence is identical in mouse and rat and contains a dual (AX[A/S]XQ) ciliary targeting motif found in GPCRs efficiently localized to cilia. In contrast, the third ICL of the human SST3 contains only a single (AX[A/S]XQ) motif. Consequently, the human SST3 is not selectively localized to primary cilia but is predominantly observed at the plasma membrane in many cell types. The calculated mol. wt. of the nonglycosylated protein is approximately 46 kDa. However, in Western blots derived from human pituitary and transfected cells, SST3 is detected as a broad smear of approximately 70–80 kDa. After peptide N-glycosidase F (PNGase F) treatment, the mol. wt. is reduced to approximately 48 kDa, and the receptor protein appears as a sharp band, indicating that the native SST3 protein is indeed heavily glycosylated. SST3 is unique among SST subtypes in that it exhibits a very long carboxyl-terminal tail. In contrast to all other SSTs, the SST3 carboxyl-terminal tail lacks a potential palmitoylation site. Amino acid sequences of both the third ICL and the carboxyl-terminal tail are not conserved across species.

66SST3 is a Gi/o-coupled receptor. Agonist activation results in increased incorporation of guanosine 5′-O-[gamma-thio]triphosphate (GTPγS) into membranes of SST3-transfected cells. Its major effector systems are inhibition of adenylyl cyclase, activation of Kir3.x currents, and modulation of VOCCs (Fig. 12). In transfected cells, transient activation of ERK can be detected. These effects are greatly diminished when cells are preincubated with PTX, strongly indicating that Gi/o proteins are major effectors of SST3. Exogenously expressed human SST3 activates PLC, which is only partly blocked by PTX, suggesting involvement of Gq proteins. However, the physiologic relevance of PLC activity is unknown. When expressed in stable rat pituitary tumor cells (GC cells), human SST3 exhibits constitutive ligand-independent activity that inhibits basal cAMP/protein kinase A (PKA) signaling and suppresses GH transcription through glycogen synthase kinase 3B activation. Heterologously expressed SST3 also induces antiproliferative or proapoptic cell-specific effects. SST3-induced apoptosis in CHO-K1 cells involves induction of transformation-related protein 53 (p53) and Bax.

67Human SST3 contains a very long carboxyl-terminal tail of 102 amino acids, compared with only 66 and 56 residues in SST2 and SST5, respectively. Sequence homology within the carboxyl-terminal region of the mouse, rat, and human SST3 is rather low. Upon agonist exposure, human SST3 is phosphorylated at four carboxyl-terminal hydroxyl amino acids, namely S337, T341, T348, and S361. Direct evidence for agonist-induced phosphorylation of these residues has been provided by generation of phosphosite-specific antibodies. For rat SST3, S341, S346, S351, and T357 have been identified as major phosphoacceptor sites using whole-cell phosphorylation assays. Interestingly, phosphorylation sites for the human and rat SST3 identified to date reside within the proximal part of the SST3 carboxyl-terminal tail. Alignment of phosphorylation motifs identified in SST2 and SST5 suggests that agonist-mediated phosphorylation of these receptors occurs preferentially at a defined distance from the NPXXY motif, which marks the end of the seventh transmembrane region and the beginning of the carboxyl-terminal tail. This observation is supported by functional analysis of receptor mutants. Exchange of S/T sites to A within the proximal part of the carboxyl-terminal tail greatly diminished β-arrestin recruitment and SST3 internalization. In contrast, mutation of S/T residues within the distal part of the carboxyl-terminal tail had no effect on receptor trafficking. SRIF-induced phosphorylation is completely blocked by the SST3-selective antagonist NVP-ACQ090. Phosphorylation occurs rapidly within seconds to minutes, whereas SST3 dephosphorylation occurs more slowly. In transfected HEK293 cells, agonist-induced phosphorylation is primarily mediated by GRK2 and GRK3. The four identified phosphorylation sites do not undergo protein kinase C- or PKA-mediated phosphorylation. Dephosphorylation of SST3 specifically requires PP1α and PP1β.

68Another unique feature of the SST3 is its rapid downregulation upon prolonged agonist exposure. This effect has been clearly documented for both human and rat SST3 in transfected cells. Downregulation of about 50% of cellular SST3 protein is observed between 6 and 12 hours of continued agonist exposure. Loss of human and rat SST3 occurred similarly with SRIF, octreotide, or pasireotide as ligands. In contrast, no such downregulation was observed with human SST2 or SST5 under similar conditions. It is not known, however, whether endogenous SST3 undergoes such rapid downregulation. Such studies are difficult to perform because human cell lines expressing sufficient levels of endogenous SST3 allowing immunochemical detection of receptor protein are not available. Degradation of SST3 was blocked by the lysosomal inhibitor chloroquine and the cell-permeable proteasome inhibitor MG132. For rat SST3, agonist-induced ubiquitination occurs at intracellular lysine residues. Mutation of all intracellular lysine residues to arginine in a K-R mutant of the human SST3 prevents downregulation only during the first 6 hours of treatment. After 24 hours, downregulation of the K-R mutant occurs to the same extent as for wild-type SST3, suggesting that ubiquitination of SST3 facilitates, but is not an absolute requirement for degradation.

69SST3 is regulated like a prototypical GPCR in that it is internalized within minutes upon exposure to the endogenous SRIF-14 ligand. The time course and extent of receptor internalization are similar to that observed for SST2. Internalization is preceded by recruitment of β-arrestins to the activated SST3. In fact, β-arrestin-2 is more efficiently recruited than β-arrestin-1. Interestingly, SRIF-14 bound to internalized SST3 is rapidly degraded, whereas octreotide is recycled as an intact peptide. Phosphorylation is a precondition for arrestin translocation and internalization. SST3 and arrestin form transient complexes. Although SST3 receptors are internalized via clathrin-coated pits into early endosomes, arrestins dissociate from the receptor and redistribute into the cytosol. This is in contrast to SST2, which forms stable complexes with arrestins that cointernalize into early endosomes. After internalization, only a proportion of SST3 recycles back to the plasma membrane. Remaining intracellular receptors are transferred to larger diameter vesicles (presumably lysosomes) for degradation. This action is also in contrast to SST2, which completely recycles to the plasma membrane within 60–90 minutes after agonist removal. Differential endosomal sorting of SST3 and SST2 has been demonstrated using an immunocytochemical pulse–chase assay that estimates the degree to which internalized receptors remain associated with transferrin-containing endocytic vesicles. Transferrin is a well-established marker of early and recycling endosomes that mediate rapid recycling. After a 30-minute SRIF-14/transferrin pulse, a high degree of colocalization of both SST3 and SST2 with transferrin was observed. However, after an additional 20-minute pulse, only SST2—but not SST3—showed a high degree of colocalization with endocytosed transferrin, suggesting that SST3 are predominantly sorted to a population of endocytic vesicles distinct from those that constitute the conserved recycling pathway marked by transferrin. Ras-related in brain (Rab) proteins are markers for specific populations of endosomes. For many GPCRs, Rab proteins are major regulators of endosomal trafficking. Using real-time imaging, it was demonstrated that SST3 traffics through Rab4-, Rab11-, and Rab21-containing endosomes. Expression of inactive variants of these specific Rab proteins inhibits passage of SST3 through different endosomal compartments.

70When expressed in HEK293 cells, mouse, rat, and human SST3 demonstrate bona fide plasma membrane localization. In contrast, SST1 is predominantly localized to intracellular vesicular compartments. In fact, transplantation of the rat SST3 N-terminal domain to the rat SST1 is sufficient to facilitate plasma membrane localization of SST1, suggesting that the N-terminal domain of SST3 contains a plasma membrane targeting sequence. However, when mouse SST3 is expressed exogenously in polarized inner-medullary–collecting duct cells or cultured hippocampal neurons, the receptor protein is concentrated in primary cilia. Primary cilia are nonmotile plasma membrane appendages that serve specialized sensory functions, such as light sensation in photoreceptors or detection of odors in olfactory neurons. Cilia are enriched in signaling proteins, such as G proteins, adenylyl cyclases, ion channels, and arrestins. Their function is defined by the presence of specific signaling receptors. Importantly, disruption of ciliary function has been associated with human ciliopathies, such as Bardet–Biedl syndrome (BBS), Joubert syndrome, and Meckel syndrome, which have prominent functional and structural CNS phenotypes. In fact, SST3 was the first signaling receptor identified in primary neuronal cilia in mouse and rat brain and is therefore a prototypical ciliary GPCR. Thereafter, additional GPCRs with selective targeting to primary cilia were identified, such as 5-hydroxytryptamine receptor 6 and melanin-concentrating hormone receptor 1. In rodents, SST3 was also identified in primary cilia in pancreatic islets and adenohypophysis. Expression of SST3 in neuronal cilia in rodents appears after birth when first cilia are formed and persists throughout the aged brain. However, selective cilia targeting appears not to be conserved across many species. The dual (AX[A/S]XQ) ciliary targeting motif in the third ICL (243APSCQWVQAPACQ255) is identical in rat and mouse. This motif is not found in any other SST subtype, and human SST3 contains only a single ciliary targeting motif (242APSCQRRRRSERR254). This may explain why subcellular localization of the human SST3 is not restricted to primary cilia in many cell types.

71The rodent SST3 is a prototypical ciliary GPCR. As such, SST3 has been intensively used to identify proteins that traffic cargo to cilia. The ciliary targeting signal of rodent SST3 is directly recognized by BBS proteins that form the BBSome complex required for targeting of membrane proteins to cilia. The BBSome is an octameric complex consisting of seven highly conserved BBS proteins, BBS1, BBS2, BBS4, BBS5, BBS7, BBS8, and BBS9, plus BBIP10. BBS is an autosomal recessive disorder characterized by retinal degeneration, polydactyly, kidney cysts, and obesity, caused by mutations in any of 14 known genes and whose etiology is associated with cilium dysfunction. In the absence of BBSome function, SST3 accumulates at the plasma membrane. In the naturally occurring tubby mutant mouse, which develops retinitis pigmentosa, hearing loss, and obesity, SST3 also fails to localize to cilia. Although it is not known whether the tubby protein product binds directly to SST3, it appears to be an accessory factor in ciliary GPCR trafficking. Once embedded into the ciliary plasma membrane, SST3 behaves as a functional GPCR. Activation by SRIF facilitates translocation of β-arrestin-2 into cilia. β-Arrestin-2 recruitment depends on SST3 phosphorylation and is required for removal of activated SST3 receptors from the ciliary space. Interestingly, when the receptor fails to undergo BBSome- or arrestin-mediated retrieval from the cilia back into the cell, SST3 concentrates into membranous buds at the tips of cilia before release into extracellular vesicles via exocytosis. For the human SST3, the carboxyl-terminal domain was shown to interact with the multiple PDZ-domain protein 1 (MUPP1). MUPP1 is a tight junction scaffold protein in epithelial cells, and, as a result of the interaction with MUPP1, SST3 is targeted to tight junctions in human keratinocytes. Interaction with MUPP1 enables the receptor to regulate transepithelial permeability in a PTX-sensitive manner, suggesting that human SST3 can activate G proteins locally at tight junctions. Both rat and human SST3 form constitutive homodimers. When coexpressed, rat SST2 and SST3 can form heterodimers with reduced SST3 activity. However, to what extent SST3 can form dimers or oligomers with other GPCRs in vivo is not known.

72Cellular and subcellular localizations of human SST3 have been studied in detail using rabbit mAb UMB-5, which is directed against the distal part of its carboxyl-terminal tail (398QLLPQEASTGEKSSTMRISYL418) (Fig. 13). In normal human tissues, SST3 is present at the plasma membrane of distinct cell populations in the anterior pituitary, pancreatic islets, enteric ganglion cells of the GIT, the zona fasciculata and zona reticularis of the adrenal cortex, as well as the adrenal medulla, glomeruli and tubules of the kidney, luteinized granulosa cells of the ovary, and immune cells. Interestingly, nonfunctioning pituitary adenomas (mostly of gonadotroph lineage) express high levels of SST3, whereas expression of SST2 and SST5 is very low or absent. In most ACTH-producing adenomas, SST3 and SST5—but not SST2—are present. In contrast, rodent SST3 receptors are selectively targeted to primary cilia in pancreatic islets and anterior pituitary and to primary neuronal cilia in many brain regions, including cerebral cortex, hippocampus, hypothalamus, and amygdala. Another brain region with high SST3 mRNA expression is the cerebellum, where SST3 is expressed on Bergmann glial cells.

73SST3 is also present in different human tumor types. It is consistently observed with high intensity of expression, especially in pituitary adenomas. In GH-producing pituitary adenomas, noticeable amounts of SST5 and SST3 are also expressed, besides high levels of SST2, whereas in most of ACTH-producing adenomas only SST5 and SST3, but not SST2, were detected. In contrast, in gonadotropic and nonfunctioning pituitary tumors, high levels of SST3 were observed, whereas SST2 and SST5 expression was low or even absent. SST3 is expressed in about 30%–50% of pheochromocytomas and paragangliomas. In addition, SST3 was found most prominently expressed in thymomas. Although SST2 is clearly the most prominent receptor expressed in gastroenteropancreatic neuroendocrine neoplasms, SST3 was also detected in 52%–90% of cases. However, and in contrast to SST2, strong SST3 expression was only noted in 5%–29% of cases. Low SST3 expression levels occur also in bronchopulmonary neuroendocrine neoplasms. Likewise, in about 50% of tumors with neuroendocrine differentiation derived from breast, cervix, or prostate, weak to moderate SST3 expression was observed. SST3 was also detected in 56% of GIST tumors.

74At the cellular level, activation of SST3 inhibits hormone release. Inhibition of GH release has been observed in GC cells. SST3 agonists inhibit insulin release from INS-1 cells. At the systemic level, highly selective SST3 antagonists such as MK-1421 or MK-4256 facilitate glucose-stimulated insulin secretion from pancreatic β-cells and block glucose excursion in wild-type, but not SST3 KO mice. This suggests that SST3 antagonism represents a new potential mechanism for treating type 2 diabetes mellitus. SST3 expressed in primary neuronal cilia in rodent brain is critical for object recognition memory. SST3 KO mice are severely impaired in discriminating novel objects, whereas they retain normal memory for object location. Similarly, systemic injection of the SST3 antagonist NVP-ACQ090 disrupts recall of familiar objects in wild-type mice. In addition, the anticonvulsant effects of CST-14 in rodents can be blocked by the selective SST3 antagonist SST3-ODN-8, suggesting that this activity is mediated in part via the SST3.

75SRIF-14 is a full agonist at the SST3, mediating strong G protein signaling, full phosphorylation, and internalization of the receptor (Fig. 14). Compared with SRIF-14, octreotide and pasireotide behave as full agonists with regard to G protein signaling and as partial agonists with regard to receptor phosphorylation and internalization. L-796,778 was the first selective nonpeptide SST3 agonist with a moderate affinity; however, compared with SRIF-14, it behaves as a weak partial agonist on G protein signaling and does not induce noticeable receptor phosphorylation or internalization (Fig. 14; Table 6). The first selective SST3 antagonist discovered, SST3-ODN-8, was successfully used to label endogenous SST3 in human tissues using autoradiographic binding studies (Fig. 14; Table 6). Great progress has been made in development of two structurally distinct classes of selective nonpeptide SST3 antagonists based on tetrahydro-β-carboline and decahydroisoquinoline derivatives. The decahydroisoquinoline derivative ACQ090 is a full neutral antagonist that blocks phosphorylation and internalization of SST3 completely (Fig. 14; Table 6). Tetrahydro-β-carboline derivatives such as MK-1421 or MK-4256 are highly selective for SST3. They were shown to facilitate glucose-stimulated insulin secretion from pancreatic β-cells and block glucose excursion in rodents in vivo. MK-4256 has been evaluated as a potential candidate for treatment of type 2 diabetes mellitus (Fig. 14; Table 6). However, development was discontinued due to adverse cardiovascular effects related to human ether-a-go-go–related gene off-target side effects. Attempts to eliminate this off-target action have led to discovery of (4-phenyl-1H-imidazol-2-yl)-methanamines as potent and selective SST3 agonists.

76cDNAs coding for rat and human SST4 were cloned in 1992 and 1993, respectively, and the rat tissue distribution of the mRNA suggested a brain-specific receptor subtype. The gene encoding human SST4 is localized on chromosome 20p11.2 in a single exon. Genes encoding mouse and rat SST4 are localized on chromosomes 2 G3 and 3q41, respectively. Human SST4 and rat SST4 are proteins of 388 and 384 amino acids, respectively, and show 88% sequence identity. Both human and rodent SST4 feature one site for N-linked glycosylation in the N-terminal domain and a putative palmitoylation motif in the C-terminal tail. The calculated mol. wt. is approximately 42 kDa. However, in Western blots derived from rat brain, SST4 protein is detected as a broad smear of 60–70 kDa. After PNGase F digestion, the mol. wt. is reduced to approximately 45 kDa and the receptor protein appears as a sharp band, indicating that the native SST4 is indeed glycosylated. SST4 shares a conserved sequence (351YANSCANPILY361) (Fig. 15) in transmembrane region 7 (mammalian SST signature) and a consensus motif (X-[S/T]-X-Φ) in its C terminus with all the other mammalian SST subtypes (UniProtKB accession: P31391). The X-S/T-X-Φ motif (the hydrophobic amino acid Φ is phenylalanine in SST4) is regarded as a potential PDZ domain binding site and might be important for interaction with scaffolding proteins. Crystal structure of the agonist-bound SST4 is not yet available, and studies on ligand binding and receptor activation have been sparse. A homology model of SST4 has been generated using the active state β2-adrenergic receptor crystal structure, and a number of reported ligands have been docked to the model-built structure. This molecular prediction analysis suggested two partially overlapping binding modes.

77SST4 is coupled to heterotrimeric Gi/o protein–mediated adenylate cyclase inhibition (Fig. 16). In cortical neurons, SST4 has been linked to activation of M currents. Native SST4 modulates Kir3.x, VOCC, and transient receptor potential cation channel subfamily V member 1 in rat DRG neurons. In rat retinal ganglion cells, SST4 also modulates VOCC. In transfected cells, agonist-stimulated SST4 activates phospholipase A2, leading to production of arachidonic acid as a second messenger, the MAPK signaling cascade, and a NHE1. In CHO cells expressing human recombinant SST4, SRIF stimulates basal proliferation through a mechanism involving prolonged activation of mitogen-activated protein kinases 1/2 (ERK1/2) and phosphorylation of signal transducer and activator of transcription 3 (STAT3). This event is transduced by Gi/o proteins and is dependent on PI3K activity. However, SST4 exerts more complex functions in regulating cell proliferation. For example, SST4 causes prolonged activation of p38 MAPK, which in turn results in the induction of the cell cycle inhibitor p21 (cip1/Waf1) and growth arrest, when cells were simultaneously exposed to somatostatin and basic FGF. When expressed in malignant pleural mesothelioma cells, human SST4 activates SHP-1 and SHP-2. Downstream signaling through SHP-2 is required for cytostatic effects of SST4 observed in these cells. In transfected CHO cells, SST4 causes PI3K-dependent activation of NHE1 and increases extracellular acidification rate. In contrast, SST4 inhibits the ubiquitous NHE1 in transfected rat fibroblasts.

78The 5′ promoter region of the human gene bears little homology between the human and the rat 5′ flanking regions. A minimal −209-bp 5′ flanking region contains elements that support human promoter activity in different cell types in vitro, including rat pituitary cells. Elements located between −459 and −984 bp enhance promoter activity, although putative binding sites for tissue-specific transcription factors were not identified in these regions. Pathophysiological factors affecting SRIF and SST1-5 expression in endothelial cells include hypoxia, which induces SST4 mRNA expression in human umbilical vein endothelial cells. In brain endothelial cells, proinflammatory cytokines and lipopolysaccharide upregulate expression of SST4 as well as SST2. Substantial differences between rat and the human SST4 have been reported in desensitization after prolonged SRIF treatment. In transfected CHO cells expressing human SST4, both activation of NHE1 and stimulation of ERK phosphorylation were susceptible to a marked desensitization in response to SRIF.

79As to the fate of the receptor upon SRIF-14 binding, low levels of internalized human SST4 were detected in transfected cells. The use of radiolabeled ligand suggested rapid dissociation of the complex and rapid recycling of the ligand to the extracellular medium and of the receptor to the plasma membrane, respectively. These observations may suggest that sustained desensitization of human SST4 is not entirely dependent on receptor sequestration. In contrast, rat SST4 is not susceptible to rapid desensitization and does not undergo internalization at all, as shown in transfected cells and in brain tissue after in vivo treatment of rats with SRIF. Although a number of potential phosphate acceptor sites are present, rat SST4 is also not subject to agonist-induced phosphorylation and does not recruit β-arrestins to the plasma membrane when activated. Site-directed mutagenesis allowed for identification of a single amino acid residue (Thr331) in rat SST4, which confers resistance to agonist-induced internalization.

80Direct interaction between SST4 and the scaffolding PSD-95 has been shown in transfected HEK293 cells and in hippocampal neurons. This interaction is mediated by binding of a PDZ domain of PSD-95 to a PDZ-domain ligand motif in the C-terminal tail of SST4. PSD-95 is not involved in regulating receptor signaling in transfected cells. Moreover, as PSD-95 and SST4 partially colocalize in hippocampal neurons in the dendritic domain, the scaffold protein may be involved in targeting SST4, mainly localized in the somatodendritic postsynaptic domain in brain. In addition, an interaction between SST4 and the membrane glycoprotein dipeptidyl peptidase-4/cluster of differentiation 26 (CD26) occurs in malignant pleural mesothelioma cells. CD26 is a 110-kDa type II membrane glycoprotein with known dipeptidyl peptidase IV activity in its extracellular domain. SST4 and CD26 are highly coexpressed and interact through their cytoplasmic domains in malignant pleural mesothelioma cells. The SST4–CD26 interaction reduces cytostatic effects of SST4 agonists. Their efficacy was enhanced by suppression of CD26 as well as by treatment of cells with anti-CD26 mAbs. Upon treatment with anti-CD26 mAbs, SST4 aggregated preferentially in lipid rafts. Interestingly, SHP-2 also clustered in lipid rafts along with SST4, which in turn facilitated SST4-mediated cytostatic and antitumor effects. Moreover, in transfected HEK293 cells, human SST4 exists as constitutive homodimers and as constitutive heterodimers when coexpressed with human SST5. However, in a different cellular background (CHO-K1 cells), cotransfection of these two receptors did not result in heterodimerization under otherwise identical conditions. Nevertheless, to what extent the SST4–SST5 interaction may be physiologically relevant is not known.

81The SST4 is localized in diverse rat brain areas, and there is substantial consistency between mRNA and SST4-like immunoreactivity distribution in the rat. High levels of SST4 are present throughout the rat forebrain, whereas the signal progressively decreased toward caudal brain regions. SST4 is abundantly expressed in the olfactory bulb and in other olfactory structures. SST4 is found throughout layers I–VI of the neocortex, in the hippocampus formation, the hilar region of the dentate gyrus, the amygdala, and the hypothalamus. SST4 immunoreactivity is distributed along neuronal processes in the striatum, nucleus accumbens, and globus pallidus. In addition, approximately 50% of DRG showed SST4-like immunoreactivity. SST4 is also abundantly present in retinal ganglion cells. At the cellular and subcellular level of adult rat CNS, SST4 is preferentially distributed to somatodendritic neuronal domains. In the neocortex, hippocampus, and striatum, SST4 is almost exclusively confined to dendrites and symmetric synapses. In the hippocampus, targeting to asymmetric, excitatory synapses was observed. Colocalization studies of SRIF and SST4 provided evidence of close apposition of SRIF-containing axons and their terminals with dendrites containing SST4, suggesting that SST4 mainly functions postsynaptically. In the periphery, SST4 is expressed in the lung, heart, and placenta and is undetectable in pancreatic islets. Given that highly specific rabbit mAbs are not yet available, the cellular expression of human SST4 is less well characterized. Using polyclonal antibodies, SST4 receptors have been found in cells of the bronchial glands, the exocrine pancreas, in the GIT (stomach and duodenum), in kidney tubules, and in the parathyroid gland. Lack of SST4 binding sites and immunoreactivity in hypophyseal tissue suggests that SST4 does not play a major role in SRIF-mediated neuroendocrine control of the human anterior pituitary. SST4 immunoreactivity in human brain was restricted to the gray matter in cerebral cortex areas. In the sensory and motor cortex, staining of the large motor neurons was not detected. Immunopositive pyramidal cells were found in cortical layers III–VI, in agreement with results of SST4-like immunoreactivity in rat neocortex layers III–V, most likely representing targets of local SRIFergic neurons in the cerebral cortex. SST4 protein was found in the hippocampal formation, with immunostaining of cell bodies and processes in the polymorphic layer of the dentate gyrus and in the thalamus, where it localized particularly in fibers. SST4 immunoreactivity was also observed in the cerebellar cortex and the medulla. In summary, data on SST4 distribution in selected human brain areas correlate well with the distribution in rat brain. Comparison between SST4 immunoreactivity and SST4 mRNA distribution by in situ hybridization in human brain showed some discrepancies in thalamus, cerebral cortex, and cerebellum expression patterns. According to receptor autoradiography using selective ligands, SST4 cannot be frequently found in human tumors. Finally, two studies reported divergent results on the expression of SST4 in human insulinomas. In the first study, immunoreactivity analysis suggested that SST4 was the most frequent receptor expressed in both benign and malignant insulinomas. In the second study, the mRNA expression analysis and binding assays suggested that SST4 expression was limited to approximately 20% of tumors.

82SST4 is expressed in areas involved in learning and memory processes. In mice, pharmacological studies suggest that activation of hippocampal SST4 leads to impaired spatial learning and enhanced cued memory. This effect suggests a switch from formation of hippocampus-based memory to striatum-based memory. In addition, behavioral studies using selective ligands showed that activation of SST4 in the striatum increases rat locomotor activity via glutamatergic systems. Finally, activation of SST4 in the CNS plays a role in modulation of behavioral responses to acute stress and of behavioral and neuroendocrine changes induced by mild chronic stress in mice, suggesting involvement of SST4 in anxiety and depression-like behavior. Experimental data suggest SST4 as a therapeutic target in Alzheimer’s disease. Administration of the SST4 agonist NNC26-9100 was found to reduce soluble amyloid-β peptide oligomers in the brain by enhancing metalloproteinase-mediated amyloid degradation in two different mouse models. This effect correlated with improved learning. Interestingly, SST4 expression levels are drastically reduced in the temporal cortex of female Alzheimer’s disease patients. SRIF is also highly expressed in brain areas associated with seizures. Activation of SST4 suppressed epileptiform activity in mouse hippocampal slices and exerts anticonvulsant effects in vivo. Moreover, SST4 KO mice showed increased susceptibility to limbic seizures but other studies suggested excitatory and proconvulsant effects of SST4 activation. In a rat model for limbic seizures, intrahippocampal administration of a SST4-selective agonist has marked anticonvulsant effects, similar to administration of SST2 and SST3 agonists. SST4 is currently being evaluated as a therapeutic target for development of anti-inflammatory and/or analgesic drugs without endocrine side effects. Mice lacking SST4 exhibit increased inflammatory and nociceptive responses, suggesting impaired defense mechanisms. Peripheral administration of selective SST4 agonists reduced formalin-induced acute nociception and mechanical allodynia in arthritic and neuropathic pain models and exhibited multiple anti-inflammatory effects in rodents. DRG neurons are most likely a primary target of SST4 agonists. SST4 activation reduces membrane excitability in DRG neurons by activating Kir3.x and inhibiting VOCC channels, and both mechanisms are presumed to contribute to analgesic effects. SST4-selective agonists reduced acute and chronic airway inflammation as well as bronchial hyper-reactivity in the mouse, and inhibited carbachol-induced bronchoconstriction.

83Compared with other SST subtypes, SST4 displays somewhat lower affinity for the common endogenous ligand SRIF-14. Among multi-SST ligands, the cyclic heptapeptide veldoreotide (COR005) (previously called somatoprim, DG3173) is unique in that it binds to SST4 in addition to SST2 and SST5. Veldoreotide is a potent suppressor of GH secretion from human pituitary adenomas, which is attributed to its affinity for SST2 and SST5. Interestingly, despite nanomolar affinity for SST5, veldoreotide has minimal effects on insulin secretion from endocrine pancreas in vivo. In phase II clinical trials for the treatment of acromegaly, veldoreotide proved to be effective when administered by s.c. bolus or s.c. infusion (ClinicalTrials.gov NCT02235987 and NCT02217800). TT-232 is a stable cyclic heptapeptide with partial activity at SST4 that also binds to SST1. Compound J-2156, a nonpeptide agonist displaying high selectivity and high affinity for SST4 (Fig. 17; Table 7), was derived by solid-phase synthesis of a series of 1-naphthalenesulfonylamino-peptidomimetics. Both TT-232 and J-2156 exhibited anti-inflammatory and antinociceptive effects after i.p. administration in rodents. NNC26-9100 is the lead compound of another structurally distinct class of highly selective SST4 agonists. L-803,087 is a nonpeptidic agonist with high affinity and selectivity for SST4 (Fig. 17; Table 7). Unexpectedly for peptide GPCRs, only SST4 agonists have been identified, and antagonists are not yet available. SST4 agonists able to penetrate the blood brain barrier would be of great interest.

84Cloning of human SST5, the last subtype of the SST family to be cloned, was preceded by that of rat SST5, which was initially termed rat SSTR4, due to the temporal proximity with the cloning of the receptor currently known as rat SST4. After an initial period of confusion, the current nomenclature was agreed upon, and subsequent studies led to detailed characterization of SST5. The human SSTR5 gene is localized on chromosome 16p13.3, and its coding sequence spans a single exon, encoding a protein of 364 amino acids, whereas the rat Sstr5 receptor gene encodes a protein of 363 amino acids. Genes encoding mouse and rat SST5 are localized on chromosomes 17 A3.3 and 10q12, respectively. Mouse SST5 was also shown to encode a protein whose length appears to vary from 362 amino acids to 363 residues, and up to 385 amino acids, differences attributed to cloning procedures or mouse strain. Initial comparative analysis had revealed that the sequence of SST5 is evolutionarily well conserved, with human SST5 showing 80% homology with the amino acid sequence of rat and mouse SST5. Nevertheless, compared with other SSTs, human SST5 seems to display lower levels of identity and similarity with SST5 from other species, particularly at the carboxyl-terminal tail, as well as with other human SSTs [ranging from 42% to 52% compared with SST1, SST2, SST3, and SST4], which portrays SST5 as the least conserved subtype among SSTs. Original cloning revealed two potential N-glycosylation sites in human SST5, located at Asn-13 and Asn-26 in the amino-terminal segment, and a third Asn-187 in the second ICL (Fig. 18). Likewise, the 385-amino-acid mouse SST5 described by contains three equivalent putative N-glycosylation sites at residues 36, 46, and 208. The estimated molecular mass of the 364-residue human SST5 is 39 kDa, whereas immunodetection of SST5 from transfected baby hamster kidney cells in Western blot showed several bands ranging from 52 to 66 kDa. Deglycosylation with PNGase F rendered a single band with an estimated 40-kDa mass, thus confirming the glycoprotein nature of the receptor. Human SST5 contains a cysteine residue at position 320 as a presumed palmitoylation anchor, the mammalian SST signature motif in transmembrane region 7 (294YANSCANPVLY304), and two cysteine residues at residues 112 in the first extracellular loop (ECL) and 186 in the second ECL, predicted to enable a disulfide bridge. Studies on structural determinants of human SST5 function using mutational analysis suggested that ECL 2 is key to form the receptor ligand-binding pocket.

85In contrast to other SSTs, a number of studies have identified single-nucleotide polymorphisms in the human SST5 gene that may imply potential pathophysiological functions in pancreatic NETs and other cancers, bipolar affective disorder, acromegaly, prostate cancer, and in the regulation of circulating IGF-1 and IGFBP3 in prostate and breast cancer. However, only a single SST5 mutation associated with acromegaly has been described in a single patient.

86The human SSTR5 gene as well as rodent orthologs uniquely undergoes noncanonical splicing to truncated variants that possess less than the typical seven TMDs that characterize all GPCRs. In particular, two truncated human SST5 receptors exist, termed SST5TMD4 and SST5TMD5, which display distinct tissue distribution, subcellular localization, response to ligands, and signaling capacities as compared with canonical full-length SST5. The most studied variant, SST5TMD4, is scarcely present in normal tissues but abundantly expressed in a number of tumors, including pituitary adenomas, breast cancer, thyroid cancer, medullary thyroid carcinoma (Mole et al., 2015), NETs (Sampedro-Nunez et al., 2016), and prostate cancer. In all those tumor types, SST5TMD4 expression is associated with features of enhanced tumor aggressiveness that vary depending on the type of tumor: increased cell survival/proliferation, migration, invasion, angiogenesis, decreased apoptosis, poor response to octreotide/lanreotide, etc. Similar, albeit not identical, truncated SST5 receptor variants have also been cloned and characterized in pig and rodents. Truncated SST5 receptor variants share a number of features, as follows: 1) preferential intracellular distribution (rather than the predominant plasma membrane localization of full-length SST5); 2) functional capacity to selectively respond to ligands (e.g., SRIF for SST5TMD5; CST for SST5TMD4), by modulating distinct signaling pathways (cAMP, Ca2+, etc.); and 3) ability to physically interact with full-length SST2 and/or SST5, to retain them in intracellular compartments, and, eventually, disrupt their normal function. As a result, it has been proposed that SST5TMD4 and other truncated variants may act as functional dominant-negative partners for their respective full-length SST2 and SST5 receptor counterparts.

87SST5 signals through a wide array of mechanisms, which include prototypical Gαi-dependent inhibition of adenylyl cyclase common to all SSTs, regulation of other crucial enzymes like PTPs and MAPK, as well as modulation of free cytosolic calcium and potassium concentrations (Fig. 19). Original cloning studies showed that both human and rat SST5 inhibited forskolin-stimulated cAMP accumulation through a Gαi protein–dependent mechanism. These functions were confirmed in subsequent studies, mainly using SST5-transfected cells and/or selective ligands, which revealed that SST5 activates additional pathways in a context (i.e., ligand, cell environment)- and species-dependent manner. Thus, SST5 activates PLC activity, thereby increasing cytosolic Ca2+ levels by release from intracellular stores. However, SST5 also blocks VOCCs, thereby decreasing Ca2+ cell entry and cytosolic levels, likely by hyperpolarizing the cell through K+ influx via activation of Kir3.x. Using wild-type and mutant SST5, the main signaling pathways activated by this receptor have been delineated, as follows: 1) SST5 couples to individual Gαi1–3 and GαoA,B; 2) Gαo mediates antisecretory and antimitogenic effects of SST5 in human pituitary somatotrophs; 3) the DRY motif is crucial for SST5 coupling with downstream effectors, whereas the BBXXB motif within the third ICL is dispensable for cAMP inhibition but essential for SST5 actions to reduce intracellular calcium levels and inhibit ERK1/2 activation, as well as for β-arrestin/receptor interaction and receptor internalization; and 4) residues 328–347 within the C terminus may play an inhibitory role in receptor internalization. In fact, earlier mutational analysis indicated that Cys320 and, by and large, the C-tail of SST5 are essential for functional effector coupling (e.g., to adenylyl cyclase) and agonist-induced receptor desensitization and internalization.

88SST5 inhibits ERK1/2 by a dephosphorylation cascade, including inhibition of guanylate cyclase and inhibition of cGMP-dependent protein kinase G. It also exerts antiproliferative effects by activating neuronal NOS via p60src kinase. SST5 activates stress-activated protein kinase or c-Jun N-terminal kinase via Gα12 proteins. In addition, it activates and upregulates N-methyl-D-aspartate receptor function by a mechanism involving calmodulin-dependent kinase II, PLC, protein kinase C, and tyrosine kinases in hippocampal noradrenergic nerve endings. In rat retinal ganglion cells, SST5 mediates suppression of 2-amino-3-(5-methyl-3-oxo-2,3-dihydro-1,2-oxazol-4-yl)propanoic acid (AMPA) responses by acting through a signaling cascade involving Gi/o/cAMP-PKA/ryanodine/Ca2+/CAM/calcineurin/PP1.

89SST5 displays constitutive activity, resulting in tonic inhibition of cAMP and ERK1/2 signaling and thus reducing hormone secretion. Finally, it has been shown that very low concentrations of SRIF actually stimulate GH release from pig and nonhuman primate somatotrophs mediated via SST5 and involving adenylyl cyclase, cAMP/PKA, and intracellular Ca2+ pathways. These actions may also be related to the presence of truncated SST5 receptor variants, which activate a wide array of signaling routes and molecular effectors, in a cell-, species-, and tumor-dependent manner, including cAMP, intracellular Ca2+, mitogen-activated protein kinases (ERK/c-Jun N-terminal kinase), AKT, cyclin D3, actin-related protein 2/3, MYC/myc-associated factor X, Wingless/int-1, and retinoblastoma protein signaling components.

90The mouse Sstr5 gene contains two introns in the 5′-flanking region, which would enable the potential use of alternative gene promoters. In humans, cell-specific expression was initially assigned to the first 900 bp of the SSTR5 gene, but a subsequent study identified a 6.1-kb intron in the 5′-UTR that unveiled a new upstream promoter, which can drive tissue-specific activation of the gene in pituitary in a Pit1-independent manner, but not in other tissues (e.g., small intestine, lung, or placenta). Similar to the mouse Sstr5 gene promoter, the human SSTR5 promoter lacks consensus sites for TATA or CAAT boxes, YY1, or a comparable initiator sequence, but contains relevant regulatory elements, including an essential GC-rich region containing SP1 binding sites, located proximal upstream of the transcription start site, two thyroid hormone response elements (between −1741 and −1269 and −317 and −101), and a cAMP-responsive element (between −101 and transcription start site). Indeed, reporter assays confirmed that forskolin and thyroid hormones enhance and SRIF inhibits promoter activity, which was not altered by other treatments, including IGF-1, estrogens, glucocorticoids, and phorbol 12-myristate 13-acetate. Additional putative binding sites were identified for basic and tissue-specific transcription factors [e.g., nuclear factor 1, SP1, octamer-binding transcription factor 1, AP-1, AP-2, pituitary-specific positive transcription factor 1, Krox, pancreas-specific transcription factor 1, MyoD] and for hormone-dependent regulation [e.g., the cAMP-responsive element and tetracycline-responsive promoter element mentioned above, retinoic acid receptor, estrogen receptor, glucocorticoid receptor sites, etc.], as well as two CpG islands.

91Expression of the SSTR5 receptor gene is under multifactorial regulation that includes the following: 1) homologous control by SRIF and its analogs; 2) heterologous regulation by key stimulatory hypothalamic hormones, such as GHRH and ghrelin, which commonly inhibit SSTR5 receptor expression; and 3) endocrine–metabolic control by hormones from the major regulatory axes, such as sex steroids (17β-estradiol, testosterone), thyroid hormones, and glucocorticoids (reviewed in;; ).

92The SST5 carboxyl-terminal tail contains only two potential phosphorylation sites at residues T333 and T347 (compared with the seven putative phosphate–acceptor sites in SST2), which seem to undergo markedly divergent dynamics: whereas T347 is constitutively phosphorylated even in the absence of ligand, T333 is phosphorylated by GRK2 immediately after agonist binding, as shown with phosphosite-specific antibodies, and as supported by mutagenesis studies, which also point to T333 as an essential residue for SST5 receptor internalization. This latter study also suggested that the third ICL of SST5 is key for β-arrestin binding and receptor internalization upon ligand exposure, whereas the 36 terminal residues of the carboxyl-terminal tail may contribute to inhibit receptor internalization. Actually, regulation of SST5 by agonist-induced phosphorylation is tightly coupled to internalization and trafficking, for β-arrestin is recruited immediately after agonist-induced T333 phosphorylation and the receptor is subsequently internalized, in contrast to SST2, the SST5–β-arrestin complex is quickly disrupted, and SST5 traffics to early endosomes without β-arrestin. As a likely consequence of these distinct dynamics, the proportion of SST5 internalized after 30 minutes of SRIF exposure is considerably lower than that observed for the SST2 receptor (30%–40% versus 80%–90%, respectively). Nevertheless, it is important to emphasize that the dynamics of SST5 phosphorylation (and also its trafficking) are ligand- and context-dependent. Although SRIF-14 induces rapid dose-dependent SST5 phosphorylation, octreotide did not cause this effect. In addition, the SST5-selective agonist L-817,818 or the multireceptor ligand pasireotide induced SST5 phosphorylation to a lesser extent than the natural ligand SRIF-14, which was only paralleled by the SST5-selective agonist BIM-23268. Additional mechanisms influencing agonist-dependent SST5 phosphorylation remain to be elucidated.

93The reverse process of SST5 dephosphorylation at T333 is driven by PP1γ and depends on sequences in the carboxy-terminal tail, and is more rapid for SST5 than for SST2. In contrast, mechanisms that operate constitutive phosphorylation of T347 at SST5 and its physiologic relevance in vivo, if any, are still to be elucidated. Likewise, possible ubiquitin-mediated degradation of SST5 remains to be elucidated, although recent results showed a relationship between the abundance and functionality of ubiquitin-specific peptidase 8 and the degree of SST5 expression at both mRNA and protein levels, which could bear clinical relevance for treating corticotroph adenomas.

94Trafficking of SST5 is intimately related to both ligand-dependent regulation of the receptor and interaction with specific proteins, which may include heterodimerization with other plasma membrane receptors. Thus, specific features of SST5 that reside within its sequence, phosphorylation/dephosphorylation kinetics, and interaction with β-arrestin and PDZ proteins are determinants for particular trafficking dynamics of this receptor. Little is known regarding the pathway of SST5 from its synthesis in the endoplasmic reticulum to the plasma membrane (i.e., export pathway), whereas trafficking involved in ligand-induced internalization and recycling has been studied in more detail (Fig. 20). Earlier studies explored the process of human and rat SST5 internalization and trafficking using different cell models. In COS-7 cells, detailed microscopic and functional examination of SRIF-driven SST5 trafficking revealed that ligand exposure induced rapid (minutes) internalization of SST5, which involved trafficking to endosomal compartments, and was followed by a subsequent process of ligand–receptor dissociation and receptor recycling to the cell surface, accompanied by recruitment of receptors residing in an intracellular pool to the plasma membrane. Subsequent studies have dissected the contribution of specific protein regions and amino acid residues in SST5 trafficking, by demonstrating, for example, the importance of the third ICL (particularly phosphorylation of S242) and the carboxyl-terminal tail for agonist-induced internalization of SST5, which seems to be species and cell context dependent. In fact, SST5-trafficking dynamics are also heavily ligand-dependent, a feature of critical importance from a clinical standpoint. Trafficking does not solely depend on ligand-binding affinity for SST5, but on alternative properties that may entail, among others, distinct agonist–receptor binding sites, specific processes of phosphorylation, and subsequent activation of downstream interactions with β-arrestin, PDZ, and other interacting proteins (, 2012;;;;;;; ).

95Like the other SSTs, SST5 contains a potential C-terminal class I PDZ ligand motif. Human and rodent SST5 interact with the PDZ domain protein interacting specifically with Tc10 (PIST, a Golgi-associated protein also known as Golgi-associated PDZ and coiled-coil motif–containing protein) and with sodium/hydrogen exchanger regulatory factor (NHERF)3/PDZ-K1 (PDZ protein expressed in kidney 1), a scaffold protein. Analysis of SST5–PIST interaction in HEK293, AtT20, and MIN6 cells suggested that PIST may accompany SST5 to the Golgi/TGN, and also, that it may contribute to recycling to the plasma membrane. More recent work identified additional PDZ domain proteins interacting with mouse SST5, such as sorting nexin family member 27 and NHERF1, and further delineated the function of PIST, which seems to retain SST5 at the Golgi/TGN compartment. In contrast, NHERF1 could release the receptors from this area and thereby facilitate access to the cell surface. PDZ-K1/NHERF3 appears to regulate specific interaction and functional activation of PLCβ3 by SST5 and other SSTs in response to SRIF by forming a ternary complex with PLCβ3 and SSTs. Thus, although the PDZ motif of SST5 does not seem to be indispensable for agonist-induced internalization of the receptor or for recycling to the plasma membrane, it may limit lysosomal degradation (and hence increase receptor stability) and enable additional signaling capabilities through selective PDZ domain–driven interactions.

96SST5 displays the ability to interact with other receptors from the SST family, forming homodimers/homomers or heterodimers/heteromers. Evidence in support of the existence and functional relevance of SST5 homodimers as well as heteromers with SST1 and SST2 has been derived in cell models, and these have delineated some molecular determinants and mechanisms involved in these interactions. However, although it has been suggested that SST2–SST5 interaction could elicit relevant functional consequences in the response to SRIF analogs in acromegaly, the precise biologic and physiologic importance of these mechanisms in vivo is still a matter of debate. Interestingly, the truncated human SST5TMD4 variant disrupts normal SST2 homodimerization, whereas it does not interfere with homodimerization of its full-length SST5 counterpart. Consequently, SST5TMD4 only, and distinctly, reduced functional responses of SST2 to SRIF, but not that of canonical SST5, which conveys key functional consequences, as the variant may disrupt the inhibitory capacity of SST2. Indeed, a comparable situation occurs for truncated pig SST5 variants. Heterodimerization of SST5 with a GPCR from a different family, the dopamine receptor D2 (D2 receptor), has also been reported, and its potential pharmacological consequences in CNS are being explored. SST5 formation of heteromers seems to be promiscuous as it has also shown to interact with the ghrelin receptor GHS-R1a, in a context enabling a fine, coordinated regulation of glucose-stimulated insulin secretion by SRIF and ghrelin. In fact, SST5 may even interact physically and functionally with another receptor class, the tyrosine kinase receptors.

97By using rabbit mAb UMB-4, SST5 was detected both at the plasma membrane and in the cytoplasm of distinct cell populations of different normal human tissues such as GH- and ACTH-producing cells of the anterior pituitary, acinar cells, and striated ducts of the parotid glands, C cells of the thyroid, neuroendocrine- and enterochromaffin-like cells of the GI mucosa, insulin- and glucagon-secreting cells of the pancreas, cells in the reticular zone of the adrenal cortex and in adrenal medulla, glomerular endothelial cells and tubules of the kidney, luteinized granulosa cells of the ovary, luminal parts of testicular tubuli, lymphocytes in the germinal centers of lymph follicles, alveolar macrophages of the lung, singular cells scattered throughout the stroma of various organs, and single cells observed occasionally in the liver, most probably also representing macrophages (Fig. 21). Very limited expression of SST5 was detected in the brain of rodents, particularly in specific nuclei in the basal forebrain.

98Regarding neoplastic tissues, SST5 was observed at a high intensity of expression in all somatotroph and in most of corticotroph adenomas, whereas in gonadotroph and nonfunctioning pituitary adenomas SST5 expression was low (Fig. 21). Furthermore, the receptor was detected in 38%–57% of medullary and in most of (>75%) papillary and follicular thyroid carcinomas. To a variable extent SST5 was additionally detected in pheochromocytomas and paragangliomas and in functioning and nonfunctioning adrenocortical adenomas. Presence of SST5 was also noticed in lymphomas. Most notably and after SST2, SST5 represents the second most common SST subtype expressed in gastroenteropancreatic neuroendocrine neoplasms. SST5 was detected in 62%–93% of tumors overall, with less frequent expression in pancreatic than in GI tumors, and higher expression rates in functioning than in nonfunctioning tumors. Additionally, tumor grade correlates negatively with receptor abundance, and, hence, a positive association with patient outcomes has been demonstrated for SST5. However, SST5 overexpression is also associated with vascular and nerve invasion and thus enhanced aggressiveness. In some studies a positive correlation between SST5 expression and SST-based imaging was shown. Furthermore, SST5 was detected in 31%–45% of bronchopulmonary neoplasms and occasionally also in other tumors with neuroendocrine differentiation. SST5 was observed in 15%–47% of GIST, and SST2 and/or SST5 immunoreactivity was associated with increased recurrence-free survival. SST5 was detected in 39%–70% of colorectal cancers, and expression was higher in well to moderately differentiated tumors than in poorly differentiated ones, with a positive correlation with favorable patient outcomes. SST5 was also detected in most breast, cervical, ovarian, and prostate carcinomas, as well as in 45% of Merkel cell tumors.

99The main physiologic functions of SST5 relate to control of pituitary and pancreatic endocrine secretions. SST5 abundance in pituitary explains its relevant role in the SRIF-mediated inhibition of GH secretion from somatotrophs, and its capacity to inhibit ACTH from corticotrophs and TSH from thyrotrophs. SST5 constitutive activity may also contribute to these actions. Conversely, SST5 does not seem to participate relevantly in the physiologic control of prolactin release, although it can inhibit its secretion in prolactinomas; likewise, there is no evidence that SST5 contributes to regulate gonadotroph function.

100In the human endocrine pancreas, SST5 plays an important role in conveying the inhibitory actions of SRIF on glucose-stimulated insulin release, although there is also evidence for a role of SST2. Conversely, in rodent β-cells, SST5 is the most abundant and the predominant inhibitory receptor for glucose-induced insulin secretion, and also appears to be involved in β-cell development. Indeed, altered glucose and insulin regulation is the most prominent phenotype of SST5 KO mice, which are otherwise devoid of overt pathologic symptoms. SST5 may also contribute to inhibit glucagon secretion from α-cells, primarily controlled by SST2.

101Presence and functional roles of SST5 in the CNS are relatively limited compared with other SSTs. In the brain, SST5 activation may inhibit stress-related stimulation of hypothalamic CRF and pituitary ACTH release. SST5 may contribute to regulate sympathetic responses; likewise, SST5 may mediate gastric emptying through activation of vagal cholinergic pathways (possibly with the contribution of other receptors), as supported by its high expression in the dorsal motor nucleus of the vagus nerve. Evidence for SST5 functions outside its endocrine and CNS actions is limited. SST5 is present in the rat retina, where its activation protects from AMPA–induced neurotoxicity. SST5 is also present in cochlea, but its role and relevance are not yet known. In the reproductive tract, SST5 is present in Sertoli cells, where its expression is developmentally regulated. In the vascular system, SST5 is present in smooth muscle cells of the human and mouse aorta, where it is coexpressed with truncated SST5TMD4 and SST2 and GHS-R1a to mediate protective actions of CST. The presence of SST5, either as mRNA or protein, has been described in a wide range of disorders, especially in tumors, where its precise role and potential value remain to be established. Of particular interest is the presence of SST5 in pituitary adenomas and NETs, for they already represent a valuable pharmacological target for SRIF analog treatment.

102Although pasireotide also exhibits affinity to SST1, SST2, and SST3, it binds with superior affinity to SST5. It also exhibits potent agonistic activity at SST5 and most likely mediates most of its pharmacological actions via SST5. The peptide agonist BIM-23268 displays moderate affinity to all SSTs (Fig. 22; Table 8); however, it appears to be unique among SST5 agonists in that it exhibits full agonistic activity. Another peptide agonist is BIM-23206, which displays about 50-fold selectivity for SST5 over SST2. L-817,818 is a moderate selective nonpeptidyl agonist (Fig. 22; Table 8). However, it displays only partial agonistic activity at SST5. A series of benzoxazole piperidines was identified as high-affinity SST5 antagonists with virtual absence of binding to other SSTs. More recently, several SST5-selective antagonists such as S5A1 were evaluated as potential treatments for diabetes mellitus (Fig. 22). S5A1 displays a subnanomolar affinity for SST5 (Table 8).

103The rationale for desired characteristics of therapeutically useful SRIF analogs has evolved as knowledge of receptor subtypes and their interactions has become available. Following the discovery of SRIF as the hypothalamic factor responsible for suppression of GH secretion, it became apparent that it was involved in multiple additional physiologic functions. Due to the rapid degradation and clearance of the native SRIF peptide, efforts were focused on creating analogs with increased metabolic stability that would be useful for treating conditions of excess GH secretion, most notably acromegaly. As the structure of native SST was modified, differences were observed in the ratio of GH-suppressing activity versus other actions, in particular the suppression of insulin, which was considered a potential problem for therapeutic application. Through screening in rodents, analogs were identified with potent GH-suppressing activity with acceptably low insulin-suppressing activity, including the two SRIF analogs still most widely used clinically for treatment of acromegaly and NETs, octreotide (Sandostatin) and lanreotide (Somatuline) (Fig. 23; Table 9).

104The reason that the GH and insulin-suppressing activities could be dissociated by structural modification of SRIF was later explained with the identification of five distinct SSTs. With the realization that there were multiple receptor subtypes, the concept emerged that different subtypes controlled different, specific biologic activities, and that functional selectivity might be achieved by analogs with preference or, ideally, selectivity for one specific receptor subtype. The task then became determining which receptor subtype controlled which specific function. This was approached by creating libraries of SRIF analogs that were fully characterized for their preferences or selectivity for the different receptor subtypes. By screening panels of analogs with varying selectivities in different biologic models, it was hoped that the receptor subtype responsible for a particular biologic action could be identified, and could thus become the basis for therapeutically useful compounds targeting a specific function.

105One of the first questions to be addressed by this approach was the involvement of the SSTs in suppressing GH secretion in humans. Although the two earlier discovered analogs, octreotide and lanreotide, are the most widely used medical therapy for acromegaly, they do not normalize GH and IGF-1 levels in a significant percentage of patients with acromegaly. A recent meta-analysis of clinical studies with sustained release octreotide and lanreotide indicated normalization of GH and IGF-1 in 56% and 55%, respectively, of treated patients with acromegaly; however, most of the included studies did not use the current normalization criteria of GH <1% and did not consider composite GH and IGF-1 normalization. The PRIMARYS study, which assessed treatment-naive, unselected acromegaly patients, found that only 30% of patients were fully controlled by these clinically approved somatostatin receptor ligands.

106With characterization of their interaction with the five SSTs, both lanreotide and octreotide were found to have potent subnanomolar affinity for the SST2 subtype, with moderate affinity for SST5 (Table 9). Lanreotide and octreotide also have moderate affinity for SST5 receptors, but that does not translate functionally. Not surprisingly, studies using subtype-selective analogs confirmed that the SST2 subtype was indeed responsible for suppression of GH in the rat, the species used for biologic optimization of lanreotide and octreotide. In contrast, however, when ligand panels were tested in cultures of human fetal pituitaries, it was found that activation of both SST2, as well as the SST5 subtype, induced suppression of GH secretion, and that activation of either of the two receptors was equally efficacious. Furthermore, when the two receptors were activated together, GH-suppressing activity was significantly enhanced, well beyond that observed with the clinically used SST2-selective analogs, lanreotide and octreotide. Considering that native SRIF potently interacts with all five receptor subtypes, it is reasonable to assume that the enhanced suppression of GH induced by dual receptor activation is the normal mechanism employed by SRIF in the physiologic control of GH. These observations also suggest that the lack of potent SST5 activity in the clinically used analogs may be the reason for the lack of full control of GH and IGF-1 in a significant number of patients with acromegaly.

107The need for greater SST5 activation for a more complete suppression of GH was confirmed in studies of cultures of pituitary adenoma cells from patients with acromegaly classified as only partially responsive to SRIF analog therapy. In keeping with clinical experience, the SST2-preferring analog, octreotide, induced only a partial suppression of GH in the cultured adenoma cells. Treatment with more potent SST2- or SST5-selective analogs produced somewhat greater suppression. However, combined SST2- and SST5-selective analog treatment to activate both receptor subtypes produced greatly enhanced suppression of GH, thus supporting the concept that greater activity at SST5 is required to normalize GH in a greater percentage of acromegaly patients than is achieved with lanreotide and octreotide.

108Based on these observations, the first multisubtype-selective analog, BIM-23244, was produced with 2× greater activity at SST2, and 20× greater activity at SST5, as compared with octreotide and lanreotide. To test the feasibility of biselective receptor activity in a single compound, GH-secreting tumors were collected from 10 patients with acromegaly that were clinically classified as either fully or only partially responsive to current SRIF analog therapy. Cultured adenoma cells from the fully responsive patients responded equally to treatment with either octreotide or the SST2+5 biselective analog, BIM23244; however, octreotide produced only a partial response in cells from partially responsive patients, whereas BIM-23244 produced a greatly enhanced response, similar to that observed in the adenoma cells from fully responsive patients. These results substantiate the benefit from dual activation of both receptor subtypes and validate the concept that both activities can be presented in a single compound and retain the enhanced biologic action.

109These observations opened the possibility of receptor subtype interactions in other tissues and biologic functions. Studies of gene receptor expression had already demonstrated that various tissues contained multiple SSTs, and that expression of these subtypes could change depending on the physiologic conditions or as a result of pathology. An example of the latter is the shift in SRIF suppression of prolactin secretion from SST2-mediated in normal human pituitary cells to SST5-mediated in cells from human prolactinomas. In addition, the expression of SSTs can change temporally during the course of a specific physiologic or pathologic process. demonstrated that, following damage to the endothelial lining of the aorta, the pattern of SSTs expressed changed over the subsequent days as the various stages of repair occurred. Observations such as these raised the possibility that multireceptor-interacting ligands could not only produce enhanced actions, but by targeting selected combinations of receptor subtypes, might also provide selectivity for specific physiologic or pathologic states.

110The mechanism by which SRIF action is enhanced by activating a combination of receptors remains uncertain. The simplest possibilities would be activation of two separate transduction pathways that have a common biologic endpoint, or greater activation of a single transduction pathway shared by two or more receptors. However, another possibility with a growing body of evidence is that various receptors can physically interact to form homo- or heterodimers, with resulting changes in activity. This phenomenon was first reported to occur for members of the opioid receptor family, which are structurally related to SSTs. reported that heterodimers formed by the κ and δ opioid receptors resulted in unique ligand-binding properties and, when activated, a potentiation of signal transduction. Formation of both homo- and heterodimers has subsequently been reported to occur among the SSTs as well, including SST2 and SST5. Dimerization of receptors with resulting alterations in ligand interactions suggests a unique opportunity to develop analogs that recognize only the specific homodimer of one receptor or the heterodimer of multiple receptor types, to achieve the highest level of functional specificity and efficacy.

111The approach of using subtype-selective analogs to determine which SSTs are involved in a particular biologic action has continued using different models. In some cases, single-receptor subtypes do appear to be the dominant mechanism. Examples include suppression of insulin mediated by SST5 on human β-cells, glucagon by SST2 on α-cells, and vessel out-sprouting by SST1 from cultured human placental vein explants, a model of angiogenesis. As direct suppression of insulin is mediated by SST5, the early observation that certain analogs could suppress GH while having minimal effect on insulin explains the modest amount of SST5 activity of the analogs selected for clinical use, lanreotide and octreotide.

112As studies continued examining combinations of receptor subtype activation for potential enhancement, it was observed that receptor subtype interactions can also be antagonistic. Testing the effect of SST2 and 5 activation on proliferation of thyroid medullary carcinoma cells, it was observed that SST2-selective analogs induce dose-related inhibition of proliferation, whereas SST5-selective analogs cause an increase. When combined, increasing concentrations of the SST5-selctive analog prevents the suppression of proliferation by the SST2-selective analog, in a dose-related manner, such that at equimolar concentrations the effect of both is neutralized. In this instance, coactivation of SST2 and 5 results in an antagonistic interaction, as opposed to the enhanced biologic effect observed by coactivation of these same receptors on GH secretion. These results indicate that the biologic consequence of receptor subtype interaction is not only a function of the receptors involved, but also the specific cell type or tissue in which they are expressed.

113A partial explanation of the antagonistic effects of certain SSTs may be the inactivation of one or both receptors as a result of conformational changes following heterodimerization. In studies examining expression and function of SST2 and SST3 receptors expressed individually, demonstrated homodimerization of both subtypes and induction of specific transduction mechanisms when activated. When coexpressed, however, heterodimerization between the two subtypes was observed with the consequence of retained activation and signal transduction with SST2-selective ligands, but a complete loss of activation and signaling with SST3-selective ligands. These results clearly illustrate the exponential increase in complexity in moving from the initial targeting of individual receptor subtypes to affect a specific function to the targeting of various subtype combinations.

114Although complex, the concept of targeting the interaction between multiple receptor subtypes remains attractive for enhanced efficacy; however, due to the widespread distribution of SSTs in different tissues, the original concern still remains that a metabolically stable compound able to activate multiple receptor subtypes could induce unwanted side effects. As an example, although the previously described biselective analog, BIM-23244, with selective, potent interaction with both SST2 and 5, yields superior GH suppression, and potentially greater therapeutic benefit for a wider range of patients suffering from acromegaly, the direct suppression of insulin by SST5 raises the potential for unwanted pancreatic side effects. To examine the consequence of activating SST2 versus SST5, a study was conducted in healthy volunteers in which a SRIF analog with potent, selective SST2 activity was compared with a potent SST5 analog. Infusion of the SST2 analog resulted in a dose-related decrease in glucagon and insulin, but without effect on glucose levels. In addition, administration of an amino acid challenge during the SST2 analog infusion resulted in an appropriate insulin response, again maintaining normal glycemic control. Infusion of the SST5 analog, however, resulted in a dramatic suppression of insulin secretion, with resulting hyperglycemia, and failure of the β-cells to respond to the amino acid challenge. As a result of this potentially severe side effect mediated by SST5, BIM-23244 was not further developed, and design of subsequent analogs aimed for modest interaction with SST5. These results illustrate that the original concern of potential side effects when activating multiple receptors is a legitimate consideration.

115A different concept from teasing out the involvement of specific SSTs in specific functions with the idea of creating subtype-specific, and therefore function-specific analogs, was to create clinically useful receptor ligands that are metabolically stable, but mimic the ability of native SRIF to interact with all five receptor subtypes. A theoretical advantage of this approach is that, assuming the activities are correctly proportioned, it should be possible to take advantage of receptor subtype interactions to produce different or enhanced responses from those achieved by activation of a single receptor subtype. The risk is that receptors mediating unwanted effects could also be activated. This approach was exemplified by development of the pan-receptor–specific ligand, pasireotide (also known as Signifor), a cyclohexapeptide with reasonably high affinity for SST1, 2, and 3, no interaction with SST4, but exceptionally high, subnanomolar affinity for SST5 (Table 9). In studies in rats, pasireotide produced comparable suppression of GH secretion as octreotide, but with much greater duration owing to a significantly greater circulating t1/2. In long-term infusion studies, pasireotide was considerably more effective in lowering IGF-1 than octreotide. Because IGF-1 production is regulated by both GH-dependent and independent mechanisms, the enhanced action of pasireotide may be the result of interaction with receptor subtypes other than SST2, which is the primary mediator of octreotide action.

116In a clinical trial directly comparing sustained release formulations of pasireotide and octreotide in a large cohort of randomly allocated, medically naive patients with acromegaly, pasireotide was found to control GH and IGF-1 in a greater percentage (31.3% pasireotide treatment versus 19.2% octreotide treatment) of subjects. Furthermore, in patients resistant to octreotide or lanreotide treatment, sustained-release pasireotide was found to achieve control in 15%–20% of subjects, depending on dosage. These results may be explained by the study of, in which pasireotide was compared with octreotide for their ability to suppress GH secretion from in vitro cultures of tumor cells from acromegalic patients. Overall, the two were equivalent; however, pasireotide was found to be more effective than octreotide in a subgroup of the cultures from tumors with a comparatively lower expression of SST2 and SST2/SST5 ratio. In keeping with the high activity of pasireotide at SST5, however, a significantly higher percentage of patients experienced hyperglycemia-related adverse events in the two clinical studies with pasireotide treatment as compared with octreotide or lanreotide treatment. The potential of pasireotide has also been investigated in Cushing’s disease.

117Extending further the concept of multireceptor SRIF ligands is the creation of compounds that interact with SSTs as well as receptors outside the SST family. This concept originally derived from clinical studies indicating that combined treatment of acromegalic patients with both SRIF and dopamine analogs resulted in greater control of GH and IGF-1 than the use of either agent alone. This generated the idea to create chimeric compounds that contain structural elements of both SRIF and dopamine, and that retain the ability to bind to receptors of both. In initial studies in primary cultures of human GH-secreting adenoma cells, it was observed that whereas both pure SRIF and pure dopamine analogs were able to induce dose-related suppression of GH secretion, the combination of the two individual agents produced no greater suppression of GH than the SRIF analog alone. However, when both activities were combined in a single, chimeric compound, able to interact with both the SST2 and D2 receptor, significantly enhanced potency as well as efficacy is observed. The mechanism for this enhanced activity remains unknown; however, one possible explanation is the reported demonstration of heterodimer formation between both SST2 and 5, and D2 receptor.

118Further refinement of the ratio of activities in the chimeric compound resulted in the production of BIM-23A760, which binds to SST2 (0.03 nM), SST5 (42 nM), and D2 receptor (16 nM). The modest affinity at SST5 is intentional to avoid potential pancreatic effects, as previously discussed. The lack of glycemic side effects was confirmed in normal cynomolgous monkeys in which administration of BIM-23A760 produced potent, dose-related suppression of GH and IGF-1, but had no effect on either insulin secretion or circulating glucose. In addition to suppression of secretion, BIM-23A760 has been demonstrated to have potent antiproliferative effects, producing dose-related suppression of cultured primary human nonfunctioning pituitary adenoma cells and somatotropinoma cells. Furthermore, complete arrest of spontaneously developing, aggressive, nonfunctioning pituitary adenomas was observed in vivo in proopiomelanocortin KO mice, an effect not observed with pure SRIF or dopamine analogs, either alone or in combination. These results suggest that chimeric compounds, such as BIM-23A760, may be effective in controlling pituitary diseases of hypersecretion, as well as impacting the growth of the underlying causative tumor.

119Clinical development of BIM-23A760 was initiated and produced a clean safety profile in phase I and a significant demonstration of efficacy in a phase IIa, single-dose study in acromegalic subjects. Unfortunately, with repeated administration in humans, a long-acting, highly potent dopaminergic metabolite was produced that gradually accumulated and diminished the action of the parent compound, BIM-23A760. Subsequently, after further structure–activity studies, a second generation chimera, BIM-23B065, was produced with significantly greater potency and efficacy than BIM-23A760, as demonstrated by suppression of GH secretion from primary cultures of human GH-secreting adenoma cells from patients classified as both fully and only partially responsive to the currently used SRIF analogs, octreotide and lanreotide, and without formation of an interfering metabolite. BIM-23B065 is currently in early development.

120From the initial discovery of SRIF and its receptor subtypes, the rationale for therapeutically useful analogs has evolved from targeting a single-receptor subtype to control a specific function and to limit potential side effects, to targeting specific combinations of receptors to induce enhanced effects for a specific function. Although progress has been made in terms of elucidating specific, disease-related combinations of receptors that act together, the initial concern of inducing side effects through activation of receptors in nontargeted tissues has been demonstrated to be a legitimate consideration. Future analogs that can specifically interact with targets, such as the unique binding pockets of homo- and heterodimers formed from the SST subtypes, as well as other receptor families, may yet achieve the full potency, selectivity, and safety potential envisioned for SRIF analogs.

121Hypothalamic SRIF traverses the hypothalamic–pituitary portal vessels to impinge on anterior pituitary cells that express multiple SSTs. SRIF analogs show target selectivity for receptor subtype and functional selectivity in regulating GH, ACTH, and TSH secretion. Somatotroph cells predominantly express SST2 > SST5. SST2 signals to suppress GH secretion and may also regulate somatotroph tumor growth, whereas SST5 predominantly suppresses corticotroph ACTH release (Table 1). Studies with human GH-secreting tumor cell cultures showed a similar receptor profile and functional response to SRIF analogs. TSH-secreting pituitary adenomas (TSHomas) express SST2 and SST5. Subsequent studies revealed that anterior pituitary SSTs may also signal in a ligand-independent action. Thus, constitutive SST signaling may regulate ambient pituitary hormone secretion to maintain tonic hormone control in the absence of SRIF. These observations have supported the development of therapeutic molecules targeting different SSTs. SRIF analogs with higher affinity for SST2 are more efficacious for control of GH hypersecretion in acromegaly or TSH hypersecretion from thyrotropinomas. In contrast, pasireotide, which exhibits an affinity-binding profile more similar to natural SRIF-14, is particularly suitable for suppressing ACTH in patients with pituitary-dependent Cushing disease.

122The SST2 subtype is preferentially expressed on somatotroph cell surfaces and regulates GH secretion by suppressing intracellular cAMP levels. SRIF analog formulations with high SST2 affinity employed for treating acromegaly, namely octreotide and lanreotide, have proven safe and effective for long-term acromegaly management (Table 10). Octreotide, an octapeptide, inhibits GH secretion with a potency 45 times greater than endogenous SRIF, with minimal suppression of insulin release. As the molecule is relatively resistant to enzymatic degradation, the in vivo t1/2 is prolonged (up to 2 hours) after s.c. injection. Lanreotide is a structurally related eight-amino-acid cyclic peptide. Responsiveness to both compounds correlates with GH-secreting adenoma SST2 expression. Rebound GH hypersecretion that occurs after SRIF infusion is not apparent after administration of either peptide, offering unique advantages for safe, long-term acromegaly therapy. Pasireotide exhibits a preferential high affinity to SST5 (39-fold higher than octreotide), and also binds to SST1, SST2, and SST3. Octreotide and pasireotide similarly inhibited free cytosolic calcium and GH release in vitro, in human somatotropinoma cell cultures, where they also comparably reduced GH mRNA levels and cell viability. Indeed, using cultures derived from 33 in vitro human pituitary tumors in a head-to-head study, octreotide and pasireotide exhibited equivalent antisecretory efficacy in suppressing GH.

123Both lanreotide and octreotide exhibit similar clinical efficacy and side-effect profiles. When defining disease outcomes, it is apparent that up to 40% of patients receiving SRIF analogs exhibit discordant GH and IGF-1 levels. Measuring IGF-1 levels, rather than GH levels, during an oral glucose tolerance test appears to more rigorously reflect disease control. Injectable depot SRIF analog formulations are safe and long-acting and enable maximal biochemical control. Drug levels peak 28 days after injection of sustained release i.m. microsphere preparation of octreotide long-acting release (LAR) (20–30 mg), with concomitant GH levels suppressed for up to 49 days. In an open-label study, 70% of 151 patients responsive to octreotide showed GH levels suppressed to <2.5 ng/ml. Similarly, GH <2 ng/ml and normal IGF-1 levels were achieved in 70% of 36 patients followed for up to 18 years. The water-soluble lanreotide autogel (60, 90, or 120 mg) administered by deep s.c. injection every 28–42 days suppressed GH to <2.5 ng/ml in 130 patients at 1 year. In a randomized 12-month study, 358 patients received pasireotide LAR (40 mg) or octreotide LAR (20 mg), and biochemical control was achieved in 31% and 19% of subjects, respectively. Among those resistant to maximal doses of octreotide or lanreotide, 15% and 20% of resistant patients subsequently achieved control when placed on 40 or 60 mg pasireotide, respectively. Control of GH and IGF-1 levels is more favorable in patients with GH-secreting microadenomas. Octreotide, lanreotide, and pasireotide fall short of maximal treatment goals (i.e., normalized GH and IGF-1) in a large subset of patients. Despite medication adherence rates approaching 90%, a global meta-analysis showed ∼55% control rates for both GH and IGF-1 with SRIF analogs.

124SRIF analogs used as first-line therapy administered prior to surgery in selected patients may ameliorate preoperative morbidity, including heart failure or respiratory or metabolic disorders, thus enabling safer anesthesia. Furthermore, preoperative treatment may enhance the success of postoperative outcomes by shrinking large tumor masses prior to debulking procedures. A subset of patients with minimally or noninvasive macroadenomas is most likely to benefit from preoperative therapy. In a meta-analysis of 64 reports, SRIF analogs were shown to significantly reduce GH-secreting pituitary tumor size. Moreover, a meta-analysis showed a modestly beneficial effect of preoperative SRIF analogs on postoperative biochemical control, but subsequent longer-term follow-up has not borne out these results. Up to 60%–80% of patients harboring microadenomas, macroadenomas, and locally invasive tumors experience a reduction of pituitary adenoma size, with tumor shrinkage seen by 6 months of therapy initiation. The magnitude of shrinkage is variable, but some patients respond with >50% decrease in tumor mass. Given these observations, the use of preoperative SRIF analogs to improve surgical outcomes has been debated. Although studies have shown that postoperative biochemical control is in fact improved by presurgical SRIF analog treatment, the overall evidence is dampened by the short follow-up duration and insufficient prospective evidence.

125The beneficial impact of SRIF analogs on acromegaly comorbidities is variable, especially for cardiovascular dysfunction, and is determined by age, disease duration, and degree of biochemical disease control. Clinical benefits of SRIF analogs are achieved both by ameliorating deleterious effects of chronic GH and IGF-1 exposure, as well as likely reversal of fluid retention and swelling. For example, GH-induced epithelial sodium channel–dependent sodium transport actively leads to volume expansion and soft-tissue swelling, effects largely reversed by SRIF analogs. Hypertension, likely arising from chronic vascular damage, is usually not reversible by SRIF analogs. However, with biochemical control, doses of antihypertensive drugs required to normalize blood pressure may be decreased. Structural cardiac abnormalities, including myocardial hypertrophy and heart failure, are improved with biochemical control, especially in younger patients and in those with a shorter disease duration. Features of obstructive sleep apnea are usually improved with SRIF analog therapy, but the disorder may persist despite satisfactory biochemical control. Although joint pain and arthropathy are markedly improved symptomatically by SRIF analogs, structural joint damage and associated arthritis are usually irreversible, despite achievement of biochemical control. Headache is particularly responsive to short-acting SRIF analog therapy.

126As GH is a potent antagonist of insulin action, uncontrolled acromegaly is associated with insulin resistance, hyperglycemia, and eventually diabetes. SRIF analogs exert a dual effect on glucose control. As SST5 is expressed on the pancreatic β-cells, SRIF analogs with higher SST5 affinity (pasireotide > octreotide and lanreotide) suppress insulin secretion, leading to hyperglycemia and diabetes. By contrast, the potent GH suppression achieved by SST2-preferential SRIF analogs (octreotide and lanreotide) leads to enhanced insulin sensitivity and lowering (or normalizing) blood glucose levels.

127As SST2 and SST5 are ubiquitously expressed, especially in the GIT, it is not surprising that several off-target side effects are experienced. Transient abdominal pain, bloating, nausea, and diarrhea are commonly encountered. Asymptomatic gallstones, likely due to suppressed CCK and decreased gallbladder contractility, occur in about 20% of patients. Prolonged QT intervals have been associated with bradycardia, although distinguishing disease-related from drug-related heart conduction effects may be difficult. Elevated fasting glucose and glycated hemoglobin levels are rarely encountered. Pasireotide leads to reversible insulinopenia, hyperglycemia, and diabetes in 30% or more of patients.

128Pituitary-dependent Cushing disease is caused by a corticotroph cell adenoma hypersecreting ACTH thus leading to adrenal cortisol overproduction. As corticotroph cells abundantly express SST5, pasireotide may suppress ACTH and features of hypercortisolemia in a subset of patients. In vitro, pasireotide inhibits basal and induced ACTH release from ACTH-secreting pituitary adenomas. A double-blind, randomized phase III trial in 162 Cushing disease patients treated with pasireotide 600–900 µg twice daily showed that median urinary-free cortisol (UFC) levels were suppressed by 50%, whereas ∼24% exhibited normalized UFC levels for 6 months. Patients with mildly elevated UFC levels are most likely to respond. Most patients not controlled within 8 weeks did not achieve control by study end. Of 75 patients with a demonstrable pituitary mass receiving 900 µg pasireotide, 44% exhibited decreased mean pituitary tumor size. Blood pressure, weight, and quality of life improved, and triglyceride and low-density lipoprotein levels were reduced. Blood glucose and glycated hemoglobin levels increased in 118 of 162 patients, despite suppression of cortisol levels. As Cushing disease hypercortisolism is associated with insulin resistance and heart failure, monitoring of blood sugar and electrocardiograms for corrected QT interval prolongation and bradycardia is important.

129Central hyperthyroidism is caused by a TSHoma, a rare disease occurring both in children and adults. TSHomas typically have strong SST2 and often SST5 expression and show a good response to first-generation SRIF analogs, with about 10% of cases showing resistance. There is a single case in which cure was achieved. Pasireotide has also been used successfully in TSHoma.

130Carcinoid, GI, and pancreatic NETs express cell surface SST2. These tumors exhibit significant morbidity and mortality, and at diagnosis fewer than 50% are surgically resectable. These tumors secrete 5-hydroxytryptamine or peptide hormones with significant clinical sequelae, including GI, bronchial, and cardiac dysfunction. Accordingly, SRIF analog therapy is aimed at decreasing or stabilizing tumor mass, as well as ameliorating adverse symptoms due to circulating hormones. Overall, survival of NET patients has improved about threefold since the introduction of SRIF analog therapy.

131In randomized double-blind trials, octreotide LAR and lanreotide autogel were shown to significantly ameliorate diarrhea or flushing in up to 80% of patients with carcinoid syndrome. In the placebo-controlled PROMID trial, when 85 patients with metastatic midgut NET received octreotide LAR, median time to tumor progression was extended from 6 to 14.3 months. Furthermore, the disease was stabilized in two-thirds of patients receiving the SRIF analog therapy. In a 96-week trial by the Controlled Study of Lanreotide Antiproliferative Response in NET (CLARINET) of 204 patients randomized to receive placebo or lanreotide autogel (120 mg), prolonged disease-free survival was demonstrated. Interestingly, combination treatment of octreotide with everolimus, a mechanistic target of rapamycin kinase inhibitor, exhibited additive efficacy benefit, i.e., tumor volume reduction, in 75% of patients versus 45% for those receiving placebo plus octreotide. Based on these results, SRIF analogs appear to offer both symptomatic improvement as well as direct antitumor effects in patients harboring NET.

132SRIF analog therapeutic efficacy rates vary depending on individual patient and tumor characteristics. Understanding mechanisms driving SRIF analog responsiveness and resistance has enabled a personalized approach to acromegaly classification and management (Table 11). Retrospective studies have suggested predictors of acromegaly therapeutic responses, as well as markers of aggressive disease resistant to SRIF analogs that also correlate with adverse long-term outcomes. Increasing age, levels of GH and IGF-1, and tumor size are adverse determinants of SRIF analog responsiveness. As therapy is required to be open-ended, treatment duration is an important determinant of therapeutic sensitivity and control rates improve over years of treatment.

133In general, SST2 tumor expression correlates with SRIF analog responsiveness. Several studies have correlated efficacy in GH-secreting adenomas with SST2 immunostaining. Choice of rabbit mAbs to assess patterns and distribution of membrane staining is also associated with SRIF analog responsiveness. In acromegaly patients resistant to octreotide, tumors lacking SST5 immunoreactivity were resistant to pasireotide, whereas those with SST5 staining using the rabbit mAb UMB-4 had superior biochemical response. Cell culture responses to octreotide and pasireotide correlated with SST2 and SST5 expression, and lower SST2 expression was associated with superior pasireotide effects. However, in other in vitro studies on unselected pituitary tumor cell culture, responses to octreotide and pasireotide did not show an evident correspondence with the SST1–SST5 profile. Nevertheless, tumors immunopositive for SST2 expression are more likely to respond to octreotide and lanreotide, and those with a higher SST2 to SST5 ratio show improved outcomes. SST mutation has only been described in a single patient with acromegaly, in which there was a missense (Arg240Trp) variant in the SSTR5 gene. The patient displayed resistance to octreotide. The cytoskeletal actin-binding scaffolding protein filamin A regulates SST2 trafficking and stability. Lack of filamin A is associated with reduced cell surface expression of SST2 in neuroendocrine cell lines, although this was not the case in somatotroph cells that could be associated with SRIF analog resistance.

134Disrupted receptor recycling also alters SST signaling on the cell surface. Although GH-secreting adenoma SST2 expression may be less abundant following pretreatment with SRIF analogs, this does not appear to result in drug resistance in acromegaly and TSH-secreting adenomas, whereas tachyphylaxis has been observed in patients with NETs.

135Tumors with large, dense GH granules diffusely distributed throughout the cytosol are typically more responsive to SRIF analogs than are those containing small, uniform GH granules. Sparsely GH granulated somatotrophinomas express less SST2 and more SST5 and are more resistant to SRIF analogs than those that are densely granulated. They are larger and more invasive and occur in younger patients. Low E-cadherin expression also correlates with sparsely granulated adenomas and SRIF analog resistance. On magnetic resonance imaging, T2-weighted hyperintense GH-cell adenomas are frequently sparsely granulated and associated with a poor response to SRIF analogs. Hypointense adenomas are smaller and less invasive than hyper- and isointense adenomas, but interestingly exhibit higher IGF-1 levels.

136Molecular markers, including low aryl hydrocarbon receptor-interacting protein (AIP) expression, high β-arrestin expression, and presence of somatic mutation of the Gsp oncogene, have been associated with poor response to SRIF analogs, thus contributing to more adverse outcomes.

137Octreotide stimulates Zac1 mRNA expression, whereas Zac1 knockdown renders cells unresponsive to SRIF analogs. SRIF analogs upregulate AIP expression, and Aip mRNA correlates with Zac1 expression, establishing a novel pathway. As reduced AIP expression may be associated with reduced Giα2 levels, SRIF analog resistance may be encountered in AIP mutation–positive patients. Two truncated SST5 variants, SST5TMD4 with four TMD and sst5TMD5 with five TMD, may inhibit SST2 functions. SST5TMD4 correlates inversely with GH and IGF-1 reductions in response to octreotide LAR therapy.

138Somatic guanine nucleotide–binding protein Gs subunit α gene mutations, occurring in about 20%–30% of somatotrophinomas, result in smaller, less invasive, and more densely granulated tumors, are more often seen in older patients, and respond more favorably to SRIF analogs. Guanine nucleotide–binding protein Gs subunit α gene-positive patients have an approximately 10% greater reduction in GH levels in response to octreotide. In contrast, low levels of rapidly accelerated fibrosarcoma kinase inhibitory protein correlated with reduced octreotide responsiveness.

139About 20% of patients with Cushing disease achieve biochemical normalization with pasireotide. Corticotroph adenomas express high levels of SST5, followed by SST2. As glucocorticoids may suppress SST2 expression, corticotrophinomas are usually resistant to octreotide or lanreotide. Although NETs usually express SST2, insulinomas have reduced expression of SST2 compared with other NETs. SST2 is expressed in 90% of GI NETs, except insulinomas, where 50% of tumors express the receptor. Tachyphylaxis has been described at variable time intervals after commencement of treatment, with reports ranging from 3 to 27 months.

140Assessment of clinical, imaging, biochemical, and histopathological markers therefore offers a personalized approach to predict biochemical outcomes with SRIF analogs. Accordingly, rigorous phenotypic classification of acromegaly biomarkers for disease staging has been applied to further refine treatment approaches.

141Based on the metabolically stabilized synthetic octapeptide octreotide [d-Phe-Cys-Phe-d-Nal-Lys-Thr-Cys-Thr(ol)], which displays high affinity for SST2 and moderate affinity for SST5 and SST3, synthetized the 123I-radioiodinated Tyr3 analog of octreotide ([123I]Tyr3-SMS 201-995, [123I]Tyr3-octreotide, [123I]TOC) (Fig. 24; Tables 12 and 13) and exploited this targeted radioligand for the first successful noninvasive single-photon emission computed tomography (SPECT) imaging of SST receptor–rich tumors in humans. This initial study on 10 patients is considered as pioneering work in the field of SST imaging, and also as general proof-of-concept for the usefulness of peptide receptor imaging (PRI) and as a starting point for development of radiolabeled ligands for targeted PRRT. With the aim to overcome the unsuitable biodistribution of this first tracer, which was caused by high lipophilicity, predominant hepatobiliary excretion, and thus undesirably high abdominal background activity, the same group successfully developed 2 years later in collaboration with a group at Sandoz Research Institute a new derivative, [111In]diethylenetriaminepentaaceticacid-D-Phe1-octreotide ([111In]DTPA-D-Phe1-octreotide), named [111In]pentetreotide (Fig. 24). Subsequently, the favorable properties of this agent (e.g., ease of preparation, appropriate t1/2, and absence of major accumulation in the upper abdominal region due to its renal clearance) were demonstrated in a comparative evaluation in humans. For the first time, these studies introduced radiometals into the concept of PRI and PRRT, which significantly simplified the preparation of SST-targeted radiopharmaceuticals by exploiting fast and simple complexation procedures using chelator-conjugated peptide precursors, e.g., DTPA conjugated to the N-terminal d-Phe1-amino acid of the peptide as in [111In]pentetreotide. Data on [111In]pentetreotide imaging in more than 1000 patients were published in 1993, and this is still the most frequently cited paper from the European Journal of Nuclear Medicine. Because the sensitivity and specificity of [111In]pentetreotide SPECT in patients with GEP-NETs were higher than those obtained with the classic imaging modalities computer tomography or magnetic resonance imaging, [111In]pentetreotide (OctreoScan; Mallinckrodt, Staines-Upon-Thames, U.K.) was approved by the Food and Drug Administration (FDA) in 1994 as the first peptide-based imaging radiopharmaceutical on the basis of a dataset obtained in 350 European patients.

142After having established a noninvasive imaging methodology for NETs, the next logical step was the development of a treatment option, similar to the imaging and treatment of thyroid cancer with 123I- and 131I-iodide, respectively. Despite promising initial results after treatment of patients with very high doses of [111In]pentetreotide (up to 2.7 Ci in total) by means of the Auger and conversion electrons emitted by 111In, it became apparent that more efficient β-emitters, such as 90Y-yttrium (t1/2 = 64.1 hours, Eβmax= 2.28MeV), might be better suited for SST-targeted PRRT. These developments were based on the successful evaluation of 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA) as chelator for therapeutic radiometals with improved thermodynamic and kinetic stability, suitable for all commonly used M(III) radiometals, such as 90Y and 177Lu for PRRT, 111In for PRI with SPECT, and 68Ga for PRI with positron emisson tomography (PET), to mention only a few. Shortly after the introduction of 90Y-labeled SST ligands for peptide receptor therapy, it became apparent that methods to reduce the renal tracer uptake and thus to protect the kidneys were needed. were the first who introduced the concept of coinfusion of Lys/Arg solutions to reduce the renal uptake of [111In]Octreoscan by >55%. Despite this nephron protection, some patients suffered from renal failure after 90Y-DOTA-d-Tyr3-octreotide (DOTATOC) treatment. To overcome these limitations, 90Y was substituted by 177Lu [t1/2 = 6.71 days, Eβmax= 497 keV, Eγ = 113 keV (6.4%), and 208 keV (11%)], a β− emitter with shorter penetration depth and coemission of low energy photons, thus allowing therapy monitoring by means of SPECT. Consequently, Lu-labeled SST ligands, e.g., [177Lu]DOTA-Tyr3-octreotate ([177Lu]DOTATATE), became the PRRT agent of choice (Fig. 25).

143Concomitantly, and with the aim to increase the SST affinity, to modify the SST-binding profile, and to increase the tumor uptake of radiolabeled SST-binding peptides after administration of doses typically in the range of 15–30 nmol for imaging and 100–300 nmol for PRRT, various independent structural modifications of octreotide were introduced during the last 20 years, three of which should be mentioned in this work: 1) Tyr3, introduced into octreotide by to be able to label the peptide with radioiodine, was found to significantly improve binding affinity and thus replaced d-Phe3 in a variety of subsequent octreotide analogs; 2) Thr8(ol), originally introduced to increase the in vivo stability of octreotide, was substituted by Thr8, leading to Tyr3-octreotate (TATE); and 3) substitution of Tyr3 by 1-naphthylalanin, resulting in DOTA-d-Nal3-octreotide (DOTANOC) ligands (i.e., [68Ga]DOTANOC) with somewhat increased lipophilicity but high affinity to SST2, SST3, and SST5 (see below). Until today, the above-mentioned three DOTA-conjugated peptides, DOTATOC, DOTATATE, and DOTANOC, are the most often clinically used SST ligands for imaging (e.g., 68Ga-labeled DOTATOC, DOTATATE, and DOTANOC) and therapy (e.g., 177Lu-labeled DOTATOC and DOTATATE).

144The first clinical studies with [177Lu]DOTATATE started in 2000 in Rotterdam and formed the basis of a multinational neuroendocrine tumors therapy phase III trial named NETTER-1 at 41 global sites. NETTER-1 demonstrated that [177Lu]DOTATATE significantly improved progression-free survival (compared with octreotide injection, Sandostatin LAR; 60 mg; Novartis, Basel, Switzerland) in patients with advanced midgut NETs.

145Additional eight-amino-acid–containing peptide agonists such as lanreotide and vapreotide have been developed but never achieved relevance for PRI and PRRT. In 2016, a “shake and bake” kit preparation of 68Ga-labeled (t1/2 = 68 minute) DOTATOC (Netspot; Advanced Accelerator Applications, Saint-Genis-Pouilly, France) was approved by the FDA for PET imaging of NETs, whereas the corresponding 68Ga-kit of the Thr8-analog [68Ga]DOTATATE was approved by the European Medicines Agency (SomaKIT TOC; Advanced Accelerator Applications). In addition, attempts were undertaken to optimize targeting properties of the first SST-imaging agent, [123]Tyr3-octreotide, and to transfer this tracer methodology to other radiohalogens, i.e., to the most commonly used PET radioisotope, fluorine-18 (t1/2 = 109.7 minutes). Unfortunately, until recently, methods for direct radiofluorination were not available, and complex multistep preparations, which are inherently hard to automate, were developed. To overcome the limitations accompanied by radioiodination and conjugation of peptides with typically used 18F-labeled prosthetic groups, i.e., a significant increase in lipophilicity and the resulting suboptimal imaging characteristics, ) adapted an interesting approach developed in the early 1990s by to improve oral availability of SST-binding peptides by glycosylation. Compensation of the negative influence of radiohalogenation by the use of carbohydrated peptides was highly effective, both for radioiodinated and 18F-SST ligands, resulting in radiopharmaceuticals with excellent imaging properties. In a pilot study in 25 patients, Nα-(1-deoxy-D-fructosyl)-Nε-(2-[18F]fluoropropionyl)-Lys0-Tyr3-octreotate PET (Gluc-Lys[18F]FP-TATE) allowed fast, high-contrast imaging of SST-positive tumors. The biokinetics and diagnostic performance of Gluc-Lys([18F]FP)-TATE was superior to [111In]DTPA-octreotide and comparable with [68Ga]DOTATOC.

146To improve imaging quality and the availability of a suitable SPECT imaging agent, 99mTc-labeled analogs of octreotide (t1/2 = 6 hours), such as 99mTc-tricine-HYNIC-Tyr3-octreotide, have been developed and successfully established for routine clinical use. Furthermore, [99mTc]ethylenediamine-N,N′-diacetic acid/HYNIC-Tyr3-octreotide (Tektrotyd, NCBJ RC POLATOM) has been approved in a number of European and non-European countries.

147Radiolabeled ligands that bind with similar high affinity to all five SSTs, so-called panSRIF-like ligands, are expected to expand the clinical indications of currently applied predominantly SST2-targeted ligands and to significantly improve tumor targeting, imaging sensitivity, and therapeutic efficacy by crossreactivity to coexpressed SST1, SST3, SST4, and SST5. One of the first developments in that direction was [111In]DOTANOC, which showed high affinity to SST2, SST3, and SST5, and finally resulted in [68Ga]DOTANOC, one of the most frequently used PET-imaging agents (already mentioned above), whereas the use of other compounds, such as DOTA-1-naphthylalanin,Thr8-octreotide (DOTA-NOC-ATE) or DOTA-BzThi3,Thr8-octreotide (DOTA-BOC-ATE) with even higher and broader affinity remained limited in use. developed bicyclic analogs, such as DOTA-Tyr-cyclo(DAB-Arg-cyclo(Cys-Phe-D-Trp-Lys-Thr-Cys)) (AM3; affinity for SST2, SST3, and SST5). The authors concluded that due to its rapid background clearance and high tumor to nontumor ratios, 68Ga-AM3 might be an ideal PET-imaging agent. The first such peptide with high-affinity binding for all five receptor subtypes was KE108 [Y-DOTA-Tyr-cyclo(D-Dab-Arg-Phe-Phe-D-Trp-Lys-Thr-Phe)] and its DOTA analog [111In]KE88. Unfortunately, this peptide was only efficiently internalized in SST3-expressing cells and did not offer pan-receptor–imaging properties. Another cyclic peptide, DOTA-pasireotide with affinity for four SST subtypes (SST1, SST2, SST3, and SST5), has also been evaluated, with limited success. For both pasireotide- and KE108-based radioligands, the absence of SST2 internalization may turn out to be a serious disadvantage and compromise their accumulation in target cells, because in most cases SST2 overexpression prevails.

148Recently, native SRIF-14 and its d-Trp8 analog were considered for ligand development. Not unexpectedly, [111In]DOTA-SS14 and [111In]DOTA-DTrp8-SS14 showed high affinity to all human SST subtypes, and [111In]DOTA-DTrp8-SS14 localized in experimental tumors, which selectively expressed rat SST2, human SST2, SST3, and SST5. Furthermore, evaluated the SRIF mimic [111In]DOTA-LTT-SS28 [(DOTA)Ser1,Leu8,D-Trp22,Tyr25-SS28]. DOTA-LTT-SS28 exhibited a panSRIF-like binding profile (IC50 values for all SST subtypes in the low nanomolar range); behaved as an agonist at human SST2, SST3, and SST5; and efficiently stimulated internalization of the three receptor subtypes. In addition, significant and specific uptake was observed in HEK293-SST2–, HEK293-SST3–, and HEK293-SST5–expressing tumors. The authors concluded that [111In-DOTA]LTT-SS28 might be a promising ligand for multi-SST1–SST5–targeted tumor imaging.

149Taking into account that high in vivo stability of a peptide radiopharmaceutical is of utmost importance for successful tumor imaging and PRRT, the re-evaluation of native or slightly modified SRIF-14 and SRIF-28 needs to be considered in the context of peptidase activity in vivo. Neutral endopeptidase is responsible for rapid breakdown of i.v. administered SRIF-, bombesin-, and gastrin-derived peptides, and activity of neutral endopeptidase can be overcome through the mere coinjection of a protease inhibitor, such as phosphoramidone. This approach may result in enhanced supply and accumulation of these radiopeptides at tumor sites and in increased clinical diagnostic sensitivity and therapeutic efficacy.

150Radiolabeled antagonists for imaging of cerebral receptor systems were established early in the application of noninvasive SPECT and PET imaging. Internalization of the receptor after radioligand binding has been assumed to be critical for efficient retention of peptide radiopharmaceuticals in tumor cells and a prerequisite for efficient PRI and PRRT imaging. Almost all SST-binding peptide lead structures exploited for radiopharmaceutical development originate from development of SST-targeting drugs (e.g., octreotide) and exhibited agonistic behavior. The first results indicating that high-affinity SST antagonists that poorly internalize into tumor cells perform more effectively than corresponding agonists that are highly internalized into tumor cells were considered to be at the forefront of a paradigm shift in nuclear oncology imaging.

151Motivated by a study of, who found that inversion of the chirality at positions 1 and 2 of the octreotide peptide family converts an agonist to an antagonist, and by structure activity relationship studies of,, the first radiolabeled SST antagonists were evaluated. showed that two peptides with high affinity to SST2 ([111In]DOTA-sst2-ANT) and SST3 ([111In]DOTA-sst3-ODN-8), respectively, did not trigger SST2- or SST3-mediated internalization and prevented agonist-stimulated internalization. Subsequent biodistribution studies in mice bearing SST3-expressing tumors revealed strong accumulation of [111In]DOTA-SST3-ODN-8 at 1 hour with up to 60% of injected radioactivity per gram of tissue and maintained at a high level for >72 hours, whereas [111In]DOTA-NOC, with strong SST3-binding and internalization properties, showed a much lower and shorter-lasting uptake in SST3-expressing tumors. The same tendency was seen for SST2-binding ligand [111In]DOTA-SST2-ANT when compared with the highly potent SST2-selective agonist [111In]DTPA-TATE, suggesting that this observation may be valid for more than just one particular GPCR.

152A pilot study in five patients with NETs or thyroid cancer provided the first evidence that SST imaging with [111In]DOTA-sst2-ANT ([111In]DOTA-BASS) is significantly more sensitive and effective than that employing the FDA-approved radiotracer 111In-DTPA-octreotide (OctreoScan; Mallinckrodt). In a comprehensive preclinical study, three different SST2 antagonists, LM3 (p-Cl-Phe-cyclo(D-Cys-Tyr-D-Aph(Cbm)-Lys-Thr-Cys)D-Tyr-NH2), JR10 (p-NO2-Phe-c[D-Cys-Tyr-D-Aph(Cbm)-Lys-Thr-Cys]-D-Tyr-NH2), and JR11 (Cpa-c[D-Cys-Aph(Hor)-D-Aph(Cbm)-Lys-Thr-Cys]-D-Tyr-NH2), were evaluated in combination with two chelators [DOTA and 1,4,7-triazacyclononane,1-glutaric acid-4,7-acetic acid (NODAGA)] and various (radio)metals [In(III), Y(III), Lu(III), Cu(II), and Ga(III)]. Although the antagonists were found to be very sensitive to chelator modifications and complexation with distinct radiometals, the study illustrated the potential of the antagonists, because even a low-affinity antagonist was shown to be slightly superior to a high-affinity agonist, outweighing the affinity differences. This is due to the fact that a neutral antagonist labels receptors in all states (active or inactive), whereas an agonist only labels receptors in an active conformation. The active conformation may represent a limited proportion of the whole population, because most GPCRs show low levels of constitutive activity.

153JR11 was selected for clinical development as a PET-imaging agent labeled with 68Ga using the chelator NODAGA (68Ga-NODAGA-JR11 or 68Ga-OPS202) and as a therapeutic agent labeled with 177Lu using the chelator DOTA (177Lu-DOTA-JR11 or 177Lu-OPS201). In a preclinical comparison of the antagonist [177Lu]OPS201 (DOTA-JR11; DOTA-[Cpa-c(DCys-Aph(Hor)-DAph(Cbm)-Lys-Thr-Cys)-DTyr-NH2]) and the SST2 agonist [177Lu]DOTATATE, the antagonist showed 2.5-times higher tumor dose, longer tumor residence time, and 1.3-fold higher tumor-to-kidney dose ratio. A phase I/II PET/computed tomography study for interindividual comparison of 68Ga-NODAGA-JR11 (68Ga-OPS202) and 68Ga-DOTATOC indicated increased image contrast, sensitivity, and diagnostic accuracy of 68Ga-OPS202 for staging of gastroenteropancreatic NETs. The theranostic pair 68Ga-DOTA-JR11 and 177Lu-DOTA-JR11 has also been investigated in NET patients, and 177Lu-DOTA-JR11 (177Lu-OPS201) is being evaluated in phase I/II.

154More than 20 years after approval of [111In]Octreoscan, recent regulatory approvals of [68Ga]DOTATATE and [68Ga]DOTATOC and the expected authorization of [177Lu]DOTATATE will significantly advance the field and stimulate further peptide receptor-based imaging and therapy options. With respect to tracer development, recent studies with radiolabeled antagonists have generated high expectations that require verification in detailed clinical studies. Combination of [177Lu]DOTATATE radiotherapy with chemotherapy, targeted agents, or immunotherapies has been initiated. Data of a first phase III study comparing the combination of 177Lu PRRT and capecitabine (Xeloda; Roche, Basel, Switzerland), an oral chemotherapy agent, with 177Lu PRRT alone started at Erasmus MC (Rotterdam, Netherlands) are expected in 2017. Further studies on combination therapies, named peptide receptor chemoradionuclide therapy, have recently been published.

155The SRIF system comprises seven genes encoding two peptide precursors, SRIF and CST, and five receptors. Compared with many other regulatory peptides, this is a relatively high number of receptors. It remains an intriguing question why this system needs five different receptors to transduce the SRIF signal. To date, far few disease-associated mutations have been identified. KO mice for any of the SSTs exhibit rather mild phenotypes. This suggests a high functional redundancy with potential that loss of one SST can be compensated by another SST subtype. Although SSTs often show overlapping distributions, they exhibit striking differences in their subcellular localizations and trafficking. SST2 and SST5 receptors are primary targets for pharmacological treatment of pituitary adenomas and NETs. In addition, SST2 is a prototypical GPCR for development of peptide-based radiopharmaceuticals for diagnostic and therapeutic intervention. Consequently, the localization, regulation, and function of the five SSTs have been studied extensively in vitro and in vivo. However, open questions remain:

156Are there additional receptors for SRIF and/or CST?

157Does heterodimerization among SSTs and other GPCRs occur in vivo?

158Can SSTs signal from within intracellular compartments?

159What is the exact mechanism involved in the SST and AIP–Zac1 pathway?

160Why do sparsely granulated GH-secreting adenomas not respond well to SRIF analogs?

161Why do cell lines not respond well to SRIF analogs?

162What is the molecular mechanism underlying tumor imaging using SST antagonists?

163What is the therapeutic potential for development of biased SST agonists?

164Is there a potential for SST ligands that can penetrate the blood brain barrier and enter the CNS?

165How are targeting and membrane trafficking of SST1 regulated?

166What is the precise physiologic role of truncated SST5 variants?

167The future challenges include deciphering crystal structures for the five SSTs to facilitate discovery of novel SST subtype-selective agonists and antagonists, which are both safe and effective. It is expected that novel delivery ligand systems including oral formulations and longer-acting injectables will offer enhanced patient convenience for long-term therapies. In the future, SRIF-based therapies may become available for novel indications, such as treatment of type 2 diabetes with SST5 antagonists or treatment of neuropathic pain with SST4 agonists.