Somatostatin Receptor Subtype 5 Regulates Insulin Secretion and Glucose Homeostasis

Abstract

Somatostatin (SRIF) regulates pancreatic insulin and glucagon secretion. In the present study we describe the generation of SRIF receptor subtype 5 knockout (sst₅ KO) mice to examine the role of SRIF receptor subtypes (sst) in regulating insulin secretion and glucose homeostasis. Mice deficient in sst₅ were viable, fertile, appeared healthy, and displayed no obvious phenotypic abnormalities. Pancreatic islets isolated from sst₅ KO mice displayed increased total insulin content as compared with islets obtained from wild-type (WT) mice. Somatostatin-28 (SRIF-28) and the sst₅/sst₁-selective agonist compound 5/1 potently inhibited glucose-stimulated insulin secretion from WT islets. SRIF-28 inhibited insulin secretion from sst₅ KO islets with 16-fold less potency while the maximal effect of compound 5/1 was markedly diminished when compared with its effects in WT islets. sst₅ KO mice exhibited decreased blood glucose and plasma insulin levels and increased leptin and glucagon concentrations compared with WT mice. Furthermore, sst₅ KO mice displayed decreased susceptibility to high fat diet-induced insulin resistance. The results of these studies suggest sst₅ mediates SRIF inhibition of pancreatic insulin secretion and contributes to the regulation of glucose homeostasis and insulin sensitivity. Our findings suggest a potential beneficial role of sst₅ antagonists for alleviating metabolic abnormalities associated with obesity and insulin resistance.

SOMATOSTATIN (SRIF) is a peptide hormone initially isolated from hypothalamic tissue as a suppressor of GH secretion (1). SRIF and its receptors play an important role in the detection and therapy of neuroendocrine disorders including GH-secreting pituitary adenomas, and gastro-entero-pancreatic tumors such as carcinoids, vasoactive intestinal peptidomas, gastrinomas, insulinomas, and glucagonomas (2–5). In rodents, SRIF receptors inhibit secretion of GH, TSH, prolactin, and ACTH, regulate gastrointestinal motility, gastric acid flow, and exocrine pancreas secretion and play a role in neurotransmission (6–13).

The effects of SRIF are mediated via five somatostatin receptor subtypes (sst₁–sst₅), which belong to the superfamily of G protein-coupled receptors. Murine somatostatin receptors are encoded by five genes located on separate chromosomes (14). Genes encoding SRIF receptor subtypes 1, 3, 4, and 5 are intronless, whereas sst₂ exists in two alternatively spliced isoforms (sst₂a and sst₂b), which diverge in their C-terminal sequences (15, 16). SRIF receptors are heterogeneously distributed in various tissues including brain, gastrointestinal tract, pancreas, and pituitary gland (17–19).

There are two major physiological SRIF ligands, SRIF-14 and SRIF-28, which consist of 14 and 28 amino acids, respectively (1, 20, 21). Radioligand binding studies on cells expressing individual SRIF receptor subtypes have demonstrated that SRIF-14 and SRIF-28 interact with sst₁–sst₄ with similar affinity while sst₅ displays up to a 10-fold higher affinity for SRIF-28 (22–24). Both forms of somatostatin exert biological effects in a tissue-specific manner. In the pancreas SRIF-14 is a more potent inhibitor of glucagon secretion from the pancreatic α-cells, whereas SRIF-28 preferentially inhibits insulin secretion from the β-cells (25). Because SRIF-14 and SRIF-28 inhibit insulin and glucagon release with different potencies from pancreatic α- and β-cells, it has been postulated that each cell type expresses different SRIF receptor subtypes (26).

Immunohistochemical studies on rodent endocrine pancreas demonstrated expression of sst₂ on α- and sst₅ on β-cells (27, 28), supporting the hypothesis that different SRIF receptor subtypes could mediate the effects of somatostatin on insulin and glucagon secretion. Quantitative RT-PCR has identified all five SRIF receptor subtypes in isolated rat islets (29). Based on immunohistochemical analysis, all five subtypes are present in human pancreatic islets (30, 31). At the cellular level sst₂ colocalizes strongly with glucagon while sst₅ and sst₁ quantitatively predominate in insulin cells as detected by peptide-derived antibodies in human endocrine pancreas (30).

Endogenous SRIF ligands bind to all known SRIF receptor subtypes and most tissues express multiple receptor subtypes; it is difficult to assign a specific biological function to an individual sst. Synthetic SRIF ligands with improved selectivity for SRIF receptor subtypes were developed (32, 33) and used in the functional characterization of SRIF receptors. In vitro experiments using isolated pancreatic islets have suggested a role for sst₅ in regulating insulin secretion (34). The limited selectivity of peptidyl SRIF agonists, as well as the lack of sst-selective antagonists, made it difficult to clearly define the physiological role of the individual receptor subtypes. The recent generation of nonpeptidyl, subtype-selective SRIF agonists (35) provided a milestone for the pharmacological and functional exploration of SRIF receptor subtypes. These agents, when used in conjunction with the recently developed transgenic mice lacking individual SRIF receptors (36, 37), permit assignation of specific functions to particular receptor subtypes. Recently, we have used the combination of sst₂ receptor knock out mice and receptor subtype-selective SRIF agonists to demonstrate that sst₂ is a critical regulator of glucagon secretion (38). In the current study, we examine the role of SRIF receptors in regulating insulin secretion and glucose homeostasis by using sst₅ knockout mice and subtype-selective SRIF agonists.

RESULTS

Generation of sst₅ KO Mice

The generation of sst₅ KO mice involved cloning of a 12-kb fragment of mouse genomic DNA containing the entire coding region of the sst₅ gene. This genomic clone was then used for construction of a targeting vector. Generation of the targeting vector involved replacement of the single coding exon for sst₅ with a neomycin resistance gene and a lacZ gene in the same orientation as the sst₅ exon (Fig. 1, a–c). Insertion of the neomycin resistance gene permitted selection of embryonic stem cell clones which contained the targeting vector. Introduction of a novel EcoRI restriction site within the lacZ gene enabled the identification of targeted embryonic stem cell clones (Fig. 1c). Introduction of the bacterial enzyme β-galactosidase encoding lacZ gene allowed analysis of the sst₅ gene expression pattern.

Approximately 600 G418-resistant embryonic stem cell clones were analyzed by Southern blots, using probes flanking both sides of the homologous sequences. Five independent clones that contained the recombined sst₅ KO construct were identified (data not shown). Embryonic stem cells obtained from two independently generated clones were injected into blastocysts that were implanted into C57BL/6J foster mice. Chimeric mice from both embryonic stem cell clones underwent successful germline transmission. Approximately 50% of agouti pups carried the sst₅ KO allele.

Heterozygous sst₅ KO mice were indistinguishable from their wild-type (WT) littermates. Homozygous F₂ sst₅ KO mice were generated by cross-mating between heterozygous F₁ sst₅ KO mice. Southern blot and PCR analysis of the F₂ generation distinguished WT and sst₅ KO mice (Fig. 1, d–f). RT-PCR analysis of RNA isolated from mouse pancreas, endocrine islets, brain, and adipose tissue identified a 481-bp sst₅ DNA fragment in WT samples, which was absent in RNA samples obtained from sst₅ KO samples (Fig. 1g; data not shown). Sequencing of the RT-PCR product confirmed the identity of the 481-bp fragment as mouse sst₅. The expression of β-galactosidase was detected by staining of brain sections isolated from mice carrying a targeted sst₅ allele; no signal was observed in brains from WT mice (Fig. 1, h and i).

sst₅ KO Mice Are Viable and Display Normal Gross Morphology

Cross-matings between heterozygous sst₅ KO mice resulted in the generation of homozygous sst₅ KO mice. The resulting sst₅ KO mice showed a normal 1:2:1 Mendelian inheritance ratio of the targeted allele, indicating that deletion of sst₅ does not produce a deleterious effect on embryogenesis (data not shown). The homozygous mutant mice are phenotypically indistinguishable from their WT counterparts throughout their life. sst₅ KO mice are viable, fertile, appear healthy, give birth to a comparable number of progeny, have a life span of at least 14 months, and consume similar amounts of food and water as their WT counterparts (data not shown). Adult male sst₅ KO mice have total adipose tissue content, wet weight of liver, heart, kidney, spleen, and perigonadal white adipose tissue comparable to that measured in wild type control mice (data not shown) and there is no difference in their growth rate (Fig. 2). Careful histological analysis of pancreatic sections from 12-wk-old male mice revealed no difference in the number, size, or distribution of islets within the pancreas of KO vs. WT mice (data not shown).

Deletion of sst₅ Alters Insulin Secretion

The effects of the sst₅ deletion on insulin and glucagon secretion from islets isolated from WT and sst₅ KO mice were investigated. Pancreatic islets isolated from sst₅ KO and from WT mice showed comparable morphology as judged by stereomicroscopic examination (data not shown). The insulin content of pancreatic islets isolated from sst₅ KO mice was increased as compared with that detected in WT islets (Fig. 3A). Incubation of isolated islets with 20 mmd-glucose for 2 h resulted in a 9- to 10-fold stimulation of insulin secretion from both sst₅ KO and WT islets (Fig. 3B). Basal and glucose-stimulated insulin secretion rates from sst₅ KO islets were significantly increased as compared with WT islets; however, both types of islets responded to glucose with a similar fold increase in insulin secretion. Pancreatic islets isolated from WT and from sst₅ KO mice showed comparable basal glucagon secretion rates and glucagon content (data not shown). Thus, deletion of sst₅ results in changes in insulin secretion rates and in insulin content without affecting parameters of α-cell function.

Deletion of sst₅ Results in Alteration of SRIF-28 and Compound 5/1-Induced Inhibition of Insulin Secretion

The recently described nonpeptidyl SRIF receptor-selective agonists (35) were used to assess the role of sst₅ in regulating murine insulin secretion in vitro. The SRIF agonists were originally characterized using the five cloned human SRIF receptor subtypes, and the pharmacological properties of these ligands with respect to murine SRIF receptors have not been reported. The SRIF agonist compound 5/1 inhibits cAMP accumulation in Chinese hamster ovary (CHO)-K1 cells transfected with human sst₅ with an IC₅₀ value of 1.3 nm (35). In CHO-K1 cells transfected with mouse sst₅, SRIF-28 inhibited forskolin-stimulated cAMP accumulation with an IC₅₀ value of 0.36 nm, and compound 5/1 inhibited forskolin-stimulated cAMP accumulation with an IC₅₀ value of 0.93 nm (Fig. 4). Compound 1 and compound 2 were inactive in this assay up to a concentration of 100 nm (Fig. 4). Thus, msst₅ specifically mediates effects of compound 5/1 in a similar fashion to its human counterpart.

SRIF-28 and SRIF receptor subtype-selective SRIF agonists did not influence basal insulin secretion from pancreatic islets isolated from adult male WT and sst₅ KO mice (data not shown); therefore, the effects of SRIF and SRIF agonists on glucose-stimulated insulin secretion were investigated. Incubation of pancreatic islets isolated from WT mice with SRIF-28 led to a concentration-dependent inhibition of glucose-stimulated insulin secretion (IC₅₀: 0.55 nm, Fig. 5A). SRIF-28 was approximately 16-fold less potent in inhibiting glucose-stimulated insulin from pancreatic islets isolated from sst₅ KO mice (IC₅₀: 8.1 nm, Fig. 5A). Compound 5/1 potently inhibited insulin secretion from WT islets with an IC₅₀ value of 0.18 nm (Fig. 5B). However, compound 5/1 inhibited insulin secretion from sst₅ KO islets only at the highest tested concentration of 100 nm (Fig. 5B). These data support the hypothesis that sst₅ is involved in regulating insulin secretion from murine pancreatic islets.

Inhibition of insulin secretion by SRIF-28 from sst₅ KO islets suggests that other SRIF receptor subtypes regulate murine β-cell secretion. To prove this hypothesis the effects of SRIF agonists with selectivity for sst₁–sst₄ on glucose-stimulated insulin secretion were studied (Fig. 5, C and D). Compound 2 inhibited insulin secretion from both WT and from sst₅ KO islets at the maximal tested concentration of 100 nm (WT: 82% ± 3, P < 0.05 vs. untreated WT controls, sst₅ KO: 78% ± 4, P < 0.05 vs. untreated sst₅ KO controls). The sst₁-, sst₃-, and sst₄-selective agonists, compounds 1, 3, and 4, did not significantly affect insulin secretion from WT or from sst₅ KO islets (Fig. 5, C and D). These data demonstrate that sst₅ is the predominant receptor involved in the regulation of insulin-secretory activity of β-cells in the mouse pancreas by SRIF.

Sst₅ KO Mice Display Improved Insulin Sensitivity

The consequences of sst₅ deletion on glucose homeostasis were investigated (Table 1). sst₅ KO mice fed a chow diet displayed lower postprandial and fasting blood glucose levels as compared with WT mice. sst₅ KO mice also displayed lower fasting and postprandial insulin levels than WT mice, indicating improved insulin sensitivity. sst₅ null mice showed significantly increased fasting and postprandial glucagon levels as compared with their WT counterparts. In addition, sst₅ KO mice had a 2.6-fold increase in fasting plasma leptin levels when compared with WT mice; however, no differences in the postprandial leptin levels were observed.

Insulin is involved in the regulation of lipid metabolism; several parameters of lipid metabolism (triglycerides, free fatty acids, and cholesterol) were measured. sst₅ KO mice displayed comparable levels of fasting and postprandial triglycerides, free fatty acids, and cholesterol to those observed in WT animals (Table 1). Thus, the sst₅ null mice experience changes in insulin, glucagon, leptin, and glucose levels.

Deletion of sst₅ Attenuates High Fat Diet-Induced Insulin Resistance

The effect of sst₅ deletion on insulin sensitivity was further studied by challenging animals with a diet containing high fat (36% wt/wt, 58.4 kcal %). High-fat diet administration resulted in body weight gain in both sst₅ KO and WT mice (Fig. 6). A transient delay in body weight gain was observed in sst₅ KO mice as compared with WT animals.

Glucose and insulin plasma levels were examined over the course of high-fat diet administration. At the beginning of high-fat diet feeding (d 0) sst₅ KO mice displayed lower postprandial blood glucose and plasma insulin levels than WT mice (Fig. 7, A and B). Over a period of 10 wk, a gradual increase in blood glucose and plasma insulin levels in both groups was observed, indicating the development of insulin resistance. The difference in postprandial blood glucose observed at the beginning of high-fat diet administration persisted throughout the course of the study. Although postprandial insulin levels increased in both groups toward the end of the study, the increase in sst₅ KO mice was delayed (Fig. 7B). Although insulin levels increased by approximately 320% in the WT group after 6 wk, the increase in sst₅ KO mice was only 48%.

The difference in insulin levels between WT and sst₅ KO mice in response to high-fat diet feeding suggested that differences in insulin secretion and/or insulin sensitivity may exist. Glucose and insulin tolerance tests in both animal groups after 6 wk of high-fat diet feeding were performed to study changes in insulin sensitivity. After ip glucose administration, blood glucose levels increased comparably in both animal groups and reached maximal levels after 30 min (Fig. 8A). However, after 30 min blood glucose levels in sst₅ KO mice decreased more rapidly and returned to basal levels by 120 min, whereas blood glucose levels in the WT group were still elevated. Intraperitoneal administration of insulin (0.3 U kg⁻¹ body weight) led to a more rapid decrease in blood glucose concentrations in sst₅ KO mice as compared with WT controls (Fig. 8B), consistent with the increased insulin sensitivity observed in the sst₅ KO mice.

Plasma leptin levels in overnight fasted sst₅ KO mice were significantly increased in the beginning (d 0) and after 14 d of high-fat diet feeding; however, no difference was observed at the end of the study (Fig. 9). Leptin has been shown to decrease food intake and body weight (39, 40). As expected from the elevated plasma leptin concentration, sst₅ KO mice showed slightly decreased body weight, but there was no detectable difference in food intake as compared with WT mice (data not shown). A small difference in total adipose tissue content was noticed while other parameters of lipid metabolism remained normal (Table 2). The increase in plasma leptin levels observed in sst₅ KO animals suggests that sst₅ inhibits leptin secretion.

DISCUSSION

Somatostatin is a potent inhibitor of insulin and glucagon secretion, which are two major hormones that regulate glucose homeostasis. Recently, the role of sst₂ in regulating glucagon secretion has been demonstrated by utilizing sst-selective SRIF agonists combined with sst₂ KO mice (38). Although several studies implicated sst₅ in regulating insulin secretion (34, 41), no conclusive evidence has been provided. The current study demonstrates that sst₅ plays an important role in regulating insulin secretion from isolated pancreatic islets using sst-selective SRIF agonists in conjunction with sst₅ KO mice. This study also reveals that sst₅ influences insulin sensitivity in an unexpected manner that is independent of its effects in the pancreas.

Several lines of evidence support the role of sst₅ as a primary receptor subtype in regulating insulin secretion from murine pancreatic islets. Islets isolated from sst₅ KO mice showed increased basal and glucose-stimulated insulin secretion rates as compared with islets isolated from WT animals. Pancreatic islets isolated from sst₅ KO mice showed increased insulin content, suggesting that increased insulin secretion may result from increased insulin synthesis. SRIF-28 potently inhibited insulin secretion from WT islets whereas deletion of sst₅ resulted in a 16-fold lower potency of SRIF-28 in inhibition of insulin secretion. Compound 5/1, an sst₅/sst₁-selective SRIF agonist, potently inhibited insulin secretion from WT islets but showed little effect on insulin secretion from sst₅ KO islets. These data are consistent with the SRIF receptor subtype distribution in the rat endocrine islets that have demonstrated a colocalization of sst₅ with insulin-secreting β-cells (27). In addition, the results of the current study support the hypothesis derived from previous studies demonstrating more potent inhibition of insulin secretion by peptidyl SRIF agonists with relative selectivity for sst₅ (41, 42).

Deletion of sst₅ did not completely abolish SRIF-28 inhibition of insulin secretion, suggesting that other SRIF receptor subtypes might be involved in regulating murine β-cell function. Rodent pancreata express all five SRIF receptor subtypes as detected by quantitative PCR methods, with sst₂ mRNA being the most abundant subtype, followed by sst₁ and sst₄ (29). By antibody analysis, sst₂ and sst₅ are expressed on glucagon and insulin cells in rodent pancreas (27, 28). The protein expression of the remaining three subtypes (sst₁, sst₃, and sst₄) in rodents is unknown. In humans, pancreatic insulin cells express high levels of sst₁ followed by sst₅ as detected by quantitative immunohistochemistry, whereas sst₃ and sst₄ show the lowest level of expression among all five subtypes (30).

High levels of sst₁ expression in humans suggest that sst₁ might have a role in regulating insulin secretion. Compound 5/1 shows only approximately 10-fold selectivity toward sst₅ as compared with sst₁; therefore, inhibition of insulin secretion from sst₅ KO islets in the current study might result from interaction of compound 5/1 with sst₁. However, the sst₁-selective SRIF agonist did not significantly influence insulin secretion from isolated mouse pancreatic islets. Furthermore, neither function nor expression of sst₁ in rodent pancreatic islets has been described. Rodent studies using currently available sst₁ antibodies (30), pharmacological experiments on islets derived from recently published sst₁ KO mice (36), and ongoing studies on sst₂/sst₅ double KO will provide more information on the possible role of sst₁ in insulin regulation.

Studies on rodent pancreatic islets using peptide-derived antibodies showed that β-cells express sst₂ (28). In addition, perfusion of human pancreas with an sst₂-selective agonist and static incubation of isolated human pancreatic islets with an sst₂-selective nonpetidergic compound demonstrated the role of sst₂ in inhibiting insulin secretion in humans (43).

Sst₃- and sst₄-selective SRIF agonists did not effect insulin secretion. These data corroborate the previously described lack of significant effects of compounds with selectivity for sst₃ and sst₄ in rodents and in humans (29, 38, 41–43). These data also correlate with the low levels of expression of RNA and protein for both SRIF receptor subtypes. Our study suggests that sst₃ and sst₄ do not play an important role in regulating murine insulin secretion.

Our phenotypic analysis of the sst₅ KO mice reveals an unexpected role for sst₅ in modulating insulin sensitivity. In vitro data demonstrate that sst₅ inhibits insulin secretion, and islets isolated from sst₅ KO mice display higher insulin content. In vivo sst₅ KO mice showed decreased plasma insulin levels. The in vitro data are obtained on tissue, which is not under the influence of physiological factors such as nutrients or hormones. Although sst₅ may negatively regulate insulin secretion in the pancreas directly, it may also affect insulin levels indirectly by acting on other tissues or cell types, and it is the combination of these effects that is reflected in the blood levels of insulin. Analysis of individual islets allows us to investigate on a cell-by-cell basis the role of sst₅ in the pancreas, and it is this type of analysis that will be required to investigate the role of sst₅ in other tissues in which it might also influence insulin regulation.

Sst₅ null mice displayed lower glucose levels as compared with WT animals. The changes in both insulin and glucose were independent of their diet since a similar constellation of both hormone levels was observed in sst₅ KO mice fed either a regular chow diet or a diet containing increased fat content. WT mice responded to high fat diet feeding with a marked increase in insulin levels paralleled by an increase in blood glucose levels; these are the hallmarks of insulin resistance. The high-fat diet-induced insulin resistance manifested in WT mice was partially alleviated in sst₅ KO mice as measured by glucose and insulin tolerance tests. The increased insulin sensitivity was more surprising as sst₅ KO mice showed increased ACTH and corticosterone levels (Table 3). Both hormones have been associated with aggravating the development of insulin resistance and increasing the central deposition of adipose tissue (44, 45). Despite increased ACTH and corticosterone levels, sst₅ KO mice fed a high-fat diet displayed a slightly delayed body weight gain and showed a trend toward decreased total adipose tissue content as compared with WT mice.

The underlying mechanisms for improved insulin sensitivity in sst₅ KO mice are unknown. There could be a direct enhancement of glucose uptake by peripheral tissue in sst₅ KO mice; however, direct effects of SRIF on peripheral glucose utilization have not been demonstrated (46). Because sst₅ has a wide tissue distribution, it is possible that sst₅ regulates insulin sensitivity and/or secretion via indirect mechanisms. One potential candidate is leptin. Leptin has been demonstrated to decrease insulin secretion from isolated endocrine islets and to enhance insulin sensitivity (47, 48). sst₅ KO mice displayed increased leptin levels under fasting conditions. Although there is no report of sst₅ expression in adipose tissue, a major source of leptin, the presence of SRIF binding sites on isolated rat adipocytes has been described (49). The function of SRIF receptors on adipocytes remains unknown. In vivo studies have demonstrated lowering of leptin levels by infusion of SRIF (50) and octreotide (51), an sst₂/sst₅-selective SRIF agonist; it is possible that SRIF regulates leptin secretion directly. We have detected sst₅ expression in perigonadal adipocyte-rich tissue isolated from WT mice and in 3T3-L1 adipocytes by RT-PCR. In addition, we have evidence that sst₅ plays a role in inhibiting leptin gene expression in vitro. In our preliminary studies on differentiated 3T3-L1 cells, which provide a cell model of adipocytes, compound 5/1 and SRIF inhibited leptin gene expression. These findings suggest that sst₅ may directly regulate leptin gene synthesis and secretion. It is possible that sst₅ negatively regulates leptin secretion and that increased leptin levels in sst₅ KO mice could contribute to the improved insulin sensitivity.

Insulin levels and insulin sensitivity correlate in part with the degree of adiposity (52), and it is possible that changes in insulin sensitivity result from altered adipose tissue content. sst₅ KO mice fed a high-fat diet showed a transient delay in a body weight gain and a trend toward reduced total adipose tissue content without any differences in food intake; however, no changes in body weight and adipose tissue content in chow-fed sst₅ KO mice were observed.

sst₅ KO mice showed increased fasting and postprandial glucagon levels. Previous studies on pancreatic islets isolated from sst₂ KO mice did not provide any evidence for a direct role of sst₅ in inhibiting glucagon secretion (38). It is possible that the increase in glucagon levels in sst₅ KO mice is mediated indirectly. sst₅ KO mice have postprandial glucose levels close to the fasting glucose levels observed in WT mice. Elevation of glucagon levels in sst₅ null mice might be a counterregulatory mechanism to prevent possible hypoglycemia.

In summary, these studies show that SRIF subtype sst₅ regulates insulin secretion from murine Langerhans islets in vitro. The in vivo data provide evidence that sst₅ plays a role in regulating insulin sensitivity in rodents. Our data demonstrate that sst₅ fulfills an important physiological function in the endocrine pancreas.

MATERIALS AND METHODS

Compounds and Solutions

SRIF-14, SRIF-28, Gey’s balanced salt solution (GBSS), BSA, d-glucose were purchased from Sigma-Aldrich Corp. (St. Louis, MO); liberase RI and FuGENE 6 transfection reagent were from Roche Molecular Biochemicals (Indianapolis, IN). Tissue culture medium Roswell Park Memorial Institute (RPMI) 1640, fetal bovine serum, α-MEM, penicillin, and streptomycin were from Life Technologies, Inc. (Gaithersburg, MD). AB 2.1-prime embryonic stem cells, the Mouse Kit, and ESQ Reagent System were from Lexicon Genetics (The Woodlands, TX). Leukemia-inhibiting factor and G418 were from Life Technologies, Inc. Regular diet (7012) was from Harlan Teklad (Madison, WI); high-fat diet (36% wt/wt, 58.4% kcal %) was purchased from Research Diets, Inc. (New Brunswick, NJ).

The SRIF receptor subtype-selective compounds compound 1, compound 2, compound 3, compound 4, and compound 5/1 were provided by Dr. S. P. Rohrer (Merck \|[amp ]\| Co., Inc., Rahway, NJ). The identification and characterization of these compounds were reported previously (35). Briefly, K_i values (nm) for each human receptor are as follows: compound 1: hsst₁ (1.4), hsst₂ (1875), hsst₃ (2240), hsst₄ (170), hsst₅ (3600); compound 2: hsst₁ (2760), hsst₂(0.05), hsst₃ (729), hsst₄ (310), hsst₅ (4260); compound 3: hsst₁ (1255), hsst₂ (>10,000), hsst₃ (24), hsst₄ (8650), hsst₅ (1200); compound 4: hsst₁ (199), hsst₂ (4720), hsst₃ (1280), hsst₄ (0.7), hsst₅ (3880); compound 5/1: hsst₁ (3.3), hsst₂ (52), hsst₃ (64), hsst₄ (82), hsst₅ (0.4).

Stable Transfection of CHO-K1 Cells with the Mouse sst₅

The mouse sst₅ coding region was isolated as a 1.1-kb EcoRI-BamHI fragment. This fragment was subcloned into the corresponding restriction endonuclease sites in the mammalian expression vector pcDNA3.1 (Invitrogen, Carlsbad, CA). CHO-K1 cells were stably transfected with this mouse sst₅ plasmid construct using FuGENE 6 transfection reagent according to the manufacturer’s protocol. Transfected CHO-K1 cells were maintained in α-MEM supplemented with 10% fetal calf serum in the continuous presence of G418 (500 μg/ml). G418-resistant single cells were subsequently cloned, and membranes prepared from isolated clones were assessed for msst₅ expression using I¹²⁵-SRIF-28. The clones with the highest expression level of msst₅ were used for functional studies.

Determination of cAMP Levels

CHO-K1 cells stably transfected with mouse sst₅ were seeded in 24-well tissue culture plates and grown for 72 h to 70% confluency. The medium was removed and cells (∼1.4 × 10⁵/well) were incubated in fresh α-MEM containing 500 μm 3 isobutyl-1-methylxanthine for 30 min. Medium was removed and cell monolayers were exposed to 10 μm forskolin with or without SRIF or SRIF agonists for 10 min at 37 C. The reactions were terminated by replacing the incubation medium with 0.1 n HCl. Cells were extracted overnight at −20 C in 0.1 n HCl. Cell extracts were assayed for cAMP content by an RIA kit (Biotrak, Amersham Pharmacia Biotech, Piscataway, NJ).

Animal Husbandry

Mice were maintained under controlled conditions of 25 C with food and water available ad libitum. All experimental animal protocols were approved by the Rahway Institutional Animal Care and Use Committee.

Generation of sst₅-Deficient Mice by Gene Targeting in Mouse Embryonic Stem Cells

Design of Targeting Construct.

To delete the single coding exon of the mouse sst₅ gene, a KpnI and a NotI restriction site were introduced upstream of the ATG and downstream of the stop codon, respectively, by PCR amplification of the cloned mouse genomic DNA. The targeting vector was constructed by assembling a 5-kb SalI-KpnI genomic fragment upstream of the ATG translation initiation codon and a 1.4-kb NotI-NotI genomic fragment downstream of the sst₅ gene. A 4.7 kb KpnI-NotI fragment containing two genetic markers, a bacterial enzyme encoding lacZ under the transcriptional direction of the endogenous sst₅ promotor and a pPGK neomycin expression cassette, under the control of the phosphoglycerol kinase promotor (PGK) was inserted between these flanking sequences. The start codon for the lacZ gene was placed in the identical location as the endogenous sst₅ start codon. An additional restriction site for EcoRI was introduced at the 5′ of the lacZ gene to identify the homologous recombination event. The assembled targeting construct was subcloned into the plasmid vector Bluescript (Stratagene, La Jolla, CA).

Transfection and Cultivation of Embryonic Stem Cells

Embryonic stem cells AB2.1 derived from 129Sv/Ev mice were cultured on mitotically inactivated embryonic fibroblasts in the presence of leukemia-inhibiting factor. Approximately 1 × 10⁷ embryonic stem cells were transfected with 25 μg of the SalI-linearized targeting construct by electroporation (230 V, 500 μF). Cells were subjected to positive selection with 250 μg/ml of G418 for 12–14 d. Genomic DNA from G418-resistant embryonic stem cell clones was isolated and digested with the restriction endonuclease EcoRI to distinguish the wild type from the targeted sst₅ allele. Southern blot analysis of the EcoRI digested WT sst₅ gene, using a 900-bp BamHI-SalI fragment outside of targeting construct (5′-probe), identified a 20-kb fragment. Southern blot analysis of the EcoRI-digested targeted sst₅ allele detected a 7.5-kb fragment, due to the introduction of an additional EcoRI site. The targeted embryonic stem cell clones were injected into C57BL/6J blastocysts using standard techniques (53). The injected blastocysts were reimplanted into pseudopregnant Tac: sw (fbr) mice and allowed to develop to term. Chimeric progeny from targeted embryonic stem cells were identified by agouti coat color. Mice with chimeric coats (95–100%) were crossed to C57BL/6J and agouti progeny were genotyped.

Genotyping of Embryonic Stem Cells and Animals

Screening by Southern Blot Hybridization.

Genomic DNA (10 μg) from WT AB 2.1 embryonic stem cells and from three targeted clones was digested with the restriction enzyme EcoRI, electrophoresed on a 0.7% agarose gel, transferred onto nylon Hybond-N+ membrane (Amersham Pharmacia Biotech), and hybridized with a 5′-probe and a 3′-probe outside of the targeting construct (see Fig. 1). Homologous recombination between the vector and the WT sst₅ locus resulted in the deletion of the entire sst₅ coding sequence and its replacement with the neomycin resistance gene. The probe used for detecting the targeted event was a 0.9-kb BamHI-SalI fragment upstream of the 5′-homologous sequence (probe A) and a 1.3-kb BstZ17I-HindII fragment downstream of the 3′-homologous sequence (probe B). The probe to detect the WT sst₅ allele was a 400-bp Nae1-SalI fragment within the coding region of the mouse sst₅ gene (probe C).

Genomic DNA was isolated from tails of 3- to 4-wk-old pups, digested with EcoRI, transferred to membrane, and hybridized with probe A or probe B as described above for embryonic stem cell DNA.

Screening by PCR

After generation of mouse lines carrying the targeted allele, subsequent progeny were genotyped by PCR. All oligonucleotides were obtained from Operon Technologies Inc. (Alameda, CA). The PCR reaction was performed using a HotStarTaq Master Mix Kit (QIAGEN, Valencia, CA) on a PTC-200 thermocycler (MJ Research, Inc., Waltham, MA). The primers for identification of the WT locus (primer 1 5′-GTGCCAACCCCCTGCTCTATGGC-3′ and primer 2: 5′-GGTATGCACTTGACCAGAGAAGG-3′) were used under the following PCR conditions: 1 cycle of 95 C for 15 min; and 35 cycles of 94 C for 15 sec, 62 C for 15 sec, and 72 C for 180 sec to generate a DNA fragment of 2.2 kb. The primers for identifying sst₅ KO animals carrying the neomycin resistance gene (primer 3: 5′-ATCGCCTTCTATCGCCTTCTTGACGAG-3′ and the above described primer 2) were used under the following PCR conditions: 1 cycle of 95 C for 15 min; and 35 cycles of 94 C for 45 sec, 58 C for 30 sec, and 72 C for 180 sec to generate a DNA fragment of 2.3 kb.

Analysis of sst₅ Expression by RT-PCR

Total pancreas RNA from 8- to 12-wk-old WT (+/+) and homozygous (−/−) sst₅ KO mice was isolated using TriReagent (Molecular Research Center, Inc., Cincinnati, OH). RNA was treated with deoxyribonuclease (DNA-free, Ambion, Inc., Austin, TX) to remove residual DNA contamination. The expression of sst₅ was detected by RT-PCR using 1 μg of total RNA and primers specific for the mouse sst₅ gene (sense: 5′-CCGTGG CTGACGTGTTGTTTATG-3′, antisense: 5′-ATGCGAGTCACCTTGCGTTCTG-3′) under the following RT-PCR conditions: 1 cycle of 94 C for 30 sec; 35 cycles of 94 C for 30 sec, 58 C for 30 sec, and 68 C for 30 sec; and 1 cycle of 68 C for 5 min. A band of 481 bp was visualized on a 4% agarose gel, excised, and sequenced. RT-PCR with primers for 18S (Ambion, Inc.) was performed to ensure that equal quantities of RNA were used. No DNA contamination was detected by PCR without reverse transcription.

Histochemical Analysis of β-Galactosidase Expression

Analysis of β-galactosidase expression in the brain of adult 12-wk-old male WT and sst₅ KO mice was performed. Brains were quickly removed and fresh frozen on dry ice. Serial sectioning was performed; sections were fixed and stained using a β-galactosidase kit from Specialty Media (Phillipsburg, NJ) according to the manufacturer’s instructions. Sections were stained with eosin, dehydrated, and photographed.

In Vitro Insulin/Glucagon Secretion Assays

All in vitro secretion experiments on pancreatic islets isolated from overnight fasted adult 129SvEv/C57BL/6J WT and sst₅ KO mice (20–35 g, 3–6 months of age) were performed according to the previously described protocol (38). The original protocols for isolation of pancreatic islets (38, 54) have been modified to improve the quantity and quality of the islets. Enzymatic digestion of the exocrine pancreas was performed with liberase RI solution (0.2 mg/ml GBSS), which was injected into the bile-pancreatic duct (54). GBSS supplemented with 10% fetal bovine serum was used to stop the liberase RI action. Islets were hand picked under a stereomicroscope and maintained for 24 h in tissue culture medium as described previously (38). Approximately 100–130 islets per pancreas were isolated from WT and sst₅ KO mice, and 95% survived the 24-h incubation period in RPMI 1640 medium as determined by methylene blue staining. Secretion experiments were carried out on islets that showed round, smooth surfaces and no obvious attached exocrine or connective tissue. Batch incubations were performed, as already described (38). Islets were washed three times in GBSS and incubated in GBSS at 37 C for a 120-min equilibration period. Islets were washed three times with GBSS and divided by size (small, medium, and large) into three groups. Three islets of each group were plated in individual wells of the incubation plates. Nine islets per 1 ml GBSS were divided into quadruplicate and incubated for 2 h with 5 mmd-glucose without SRIF or SRIF agonists. Basal insulin secretion was determined from supernatants of islets incubated for 2 h without glucose and SRIF or SRIF agonists. The incubation was terminated by aspirating the incubation medium followed by a brief centrifugation at 1000 × g. Supernatants were assayed immediately for insulin or glucagon content by a double antibody RIA (Linco Research Inc., St. Charles, MO).

Determination of Glucagon and Insulin Content in Endocrine Islets

Animals were fasted overnight with water available ad libitum before experiments. Isolated islets in batches of 40 were cultivated for 24 h in RPMI 1640. Islets were washed three times in GBSS and incubated for 120 min. at 37 C. After washing in GBSS the islets were incubated for an additional 120 min at 37 C. Incubation medium was carefully removed and islets were extracted overnight at −20 C in 1 ml acid ethanol (0.2 n HCl in 70% ethanol) (38). The extract was centrifuged at 12,000 × g at 4 C, and hormone content in the supernatant was measured by an RIA.

In Vivo Studies

Determination of Blood Glucose Levels and Glucose Tolerance Test.

Fasting and postprandial blood glucose levels were determined with the glucometer One Touch Basic (Lifescan Inc., Milpitas, CA). Intraperitoneal glucose tolerance test (IPGTT) (2 g × kg⁻¹ body weight) was performed on age-matched adult male WT and sst₅ KO mice after fasting for 14–16 h. Blood was withdrawn from tail veins before and after glucose challenge at indicated time points for determination of glucose concentrations.

Intraperitoneal Insulin Tolerance Test.

Male adult mice were injected with 0.3 U human insulin (Humulin, Eli Lilly \|[amp ]\| Co., Indianapolis, IN) and blood glucose was measured at indicated time points.

Determination of Basal Plasma Insulin, Glucagon, Leptin, IGF-I, Corticosterone, ACTH, and Lipids.

Plasma insulin, leptin, glucagon, triglycerides, cholesterol, and free fatty acid levels were determined in fed and in 14-h fasted mice. Plasma IGF-I, corticosterone, and ACTH levels were measured in samples collected from overnight-fasted mice. For insulin, leptin, ACTH, corticosterone, triglycerides, cholesterol, and free fatty acids, blood was collected from the retroorbital sinus. For IGF-I and glucagon measurements, blood was collected between 0900 and 1100 h by a cardiac puncture. IGF-I was extracted from plasma and measured by rat RIA according to the manufacturer’s protocol (Diagnostics Systems Laboratories, Inc., Webster, TX). For ACTH and corticosterone measurements, blood was collected between 0800 and 1000 h and within 20–30 sec to avoid stress-mediated elevation of both hormones. Plasma corticosterone was detected using a rat RIA (ICN Biomedicals, Inc., Aurora, OH) and ACTH by a rat ELISA kit (Peninsula Laboratories, Inc., San Carlos, CA). Plasma leptin and insulin levels were measured by a rat ELISA kit (Alpco, Windham, NH).

Determination of Total Adipose Tissue Content and Organ Weight.

Whole-body composition was determined by dual energy x-ray absorptiometry-scan using the QDR 4500 Small Animal Studies software version 9.0 (Hologic, Inc., Bedford, MA), as described (55). The differential attenuation of low- and high-energy x-rays by the tissues stands in proportion to tissue density. This information is used by the detector and associated software in conjunction with tissue calibration phantoms to assess body composition. Fat mass consists primarily of adipose tissue. Lean mass consists primarily of skeletal muscle, blood, bones, organs, tendons, cartilage, and body fluids. To determine the organ weight, animals were euthanized by CO₂ and liver, heart, kidney, spleen, and perigonadal white adipose tissue were removed and immediately weighed.

Determination of Food and Water Consumption

Pelleted food (600 g per cage) was given to cages containing 5–10 animals. The weight of the remaining food was recorded at weekly intervals for a period of 10 wk. The amount of consumed water was estimated by the measurement of the weight of the water bottles at intervals of 48 h.

Data Analysis

All data represent mean ± sem and were obtained from four to six independent experiments unless otherwise stated. Statistical comparisons between means within a group were made using Student’s paired t test. Comparisons between means of different groups were analyzed by one-way ANOVA with a Neuman-Keuls multiple comparison posttest. Statistical comparisons and nonlinear regression analysis were performed with Prism 2.01 (GraphPad Software, Inc., San Diego, CA). Statistical significance was defined as P < 0.05.

Acknowledgments

We thank Dr. S. P. Rohrer for providing us with sst-selective nonpeptidyl agonists and Dr. S. E. Alves for help with brain and pancreas histology.

This work was supported by a grant from The Deutsche Forschungsgemeinschaft (Grant Str558/2-1) to M.Z.S.

Abbreviations:

 
• CHO,

  Chinese hamster ovary;

 
• GBSS,

  Gey’s balanced salt solution;

 
• IPGTT,

  ip glucose tolerance test;

 
• PGK,

  phosphoglycerol kinase promotor;

 
• RPMI,

  Roswell Park Memorial Institute;

 
• SRIF,

  somatostatin;

 
• sst₁₋₅,

  cloned SRIF receptor subtypes 1–5;

 
• sst₅ KO,

  homozygous SRIF receptor subtype 5 knockout;

 
• WT,

  wild-type.