Endocrinology and Metabolism

Distinct functional properties of native somatostatin receptor subtype 5 compared with subtype 2 in the regulation of ACTH release by corticotroph tumor cells

Joost van der Hoek, Marlijn Waaijers, Peter M. van Koetsveld, Diana Sprij-Mooij, Richard A. Feelders, Herbert A. Schmid, Philippe Schoeffter, Daniel Hoyer, Davide Cervia, John E. Taylor, Michael D. Culler, Steven W. J. Lamberts, Leo J. Hofland


In a series of human corticotroph adenomas, we recently found predominant mRNA expression of somatostatin (SS) receptor subtype 5 (sst5). After 72 h, the multiligand SS analog SOM230, which has a very high sst5 binding affinity, but not Octreotide (OCT), significantly inhibited basal ACTH release. To further explore the role of sst5 in the regulation of ACTH release, we conducted additional studies with mouse AtT-20 cells. SOM230 showed a 7-fold higher ligand binding affinity and a 19-fold higher potency in stimulating guanosine 5′-O-(3-thiotriphosphate) binding in AtT-20 cell membranes compared with OCT. SOM230 potently suppressed CRH-induced ACTH release, which was not affected by 48-h dexamethasone (DEX) pretreatment. However, DEX attenuated the inhibitory effects of OCT on ACTH release, whereas it increased the inhibitory potency of BIM-23268, an sst5-specific analog, on ACTH release. Quantitative PCR analysis showed that DEX lowered sst2A+2B mRNA expression significantly after 24 and 48 h, whereas sst5 mRNA levels were not significantly affected by DEX treatment. Moreover, Scatchard analyses showed that DEX suppressed maximum binding capacity (Bmax) by 72% when 125I-Tyr3-labeled OCT was used as radioligand, whereas Bmax declined only by 17% when AtT-20 cells were treated with [125I-Tyr11]SS-14. These data suggest that the sst5 protein, compared with sst2, is more resistant to glucocorticoids. Finally, after SS analog preincubation, compared with OCT both SOM230 and BIM-23268 showed a significantly higher inhibitory effect on CRH-induced ACTH release. In conclusion, our data support the concept that the sst5 receptor might be a target for new therapeutic agents to treat Cushing’s disease.

  • Cushing’s disease
  • adrenocorticotropic hormone
  • glucocorticoid

cushing’s disease, the pituitary-dependent form of Cushing’s syndrome, is the hypercortisolemic state secondary to excess or dysregulated ACTH secretion caused by an ACTH-secreting pituitary adenoma (36). The significant associated morbidity, such as increased tissue fragility, poor wound healing, hypertension, and diabetes mellitus, demands a proper medical intervention (1). Transsphenoidal surgery is currently the first line of treatment, and secondary options consist of irradiation therapy either alone or in combination with adrenolytic agents (10, 34, 35, 37). Unfortunately, none of the current treatment modalities ensures a full and permanent cure, as evidenced by the number of patients developing recurrent Cushing’s disease (43). The absence of an effective medical treatment has prompted physicians to explore new medical strategies, preferably based on fundamental and (patho)physiological pathways, in the hope of increasing the curation chances in this group of patients.

The physiological role of somatostatin (SS) in the regulation of anterior pituitary function (5, 27, 41, 45), its equivocal effects on ACTH release (6, 24), and the current use of SS analogs in patients with anterior pituitary tumors (29), has led to the exploration of SS analogs in patients with (recurrent) Cushing’s disease. To date, five G protein-coupled SS receptors have been cloned (sst1–sst5), and six gene products are currently known (39, 45). The receptor subtypes sst1–5 produce single gene products, whereas sst2A (long form) and sst2B (short form) originate from a common precursor mRNA, which is spliced at the carboxyl terminus (56). Although in vitro data demonstrate the presence of sst expression in corticotroph adenomas, the sst2-preferential analog octreotide (OCT) appears to inhibit ACTH release in Nelson’s syndrome and in some patients harboring ectopic ACTH-producing tumors, but rarely in patients with Cushing’s disease (13, 30). These observations are in agreement with the observation that almost all ACTH-secreting pituitary adenomas, i.e., patients with untreated Cushing’s disease, cannot be visualized by SS receptor (sst) scintigraphy using 111In-diethylenetriamine pentaacetic acid (DTPA) OCT (12, 28), whereas 111In-DTPA scintigraphy is positive in patients with Nelson’s syndrome (11, 12). Apparently, ACTH release from corticotropinomas is sensitive to OCT only in the absence of peripheral feedback regulation by glucocorticoids, suggesting that the sst2 might be downregulated when cortisol levels are high. Additional in vitro evidence for this hypothesis comes from studies using primary cultures of human corticotroph adenomas, in which glucocorticoids downregulated the response of corticotropin-releasing hormone (CRH)-induced ACTH secretion to OCT (50).

To explore the possible role of novel SS analogs in the medical treatment of Cushing’s disease, we have further evaluated the potential significance of sst subtypes expressed in human corticotroph adenomas and determined the effects of the novel multiligand SS analog SOM230 on ACTH release in primary cultures of human corticotroph adenomas (20). SOM230, compared with OCT, has a 30-, 5-, and 40-times higher affinity to sst1, sst3, and sst5 receptors, respectively, and 2.5 times lower affinity to sst2 (32). Its elimination half-life of 23 h makes this compound suitable for clinical application (7, 60). On the basis of the observed selective expression of the sst5 receptor in this series of human corticotroph adenomas, the very high affinity of SOM230 for sst5 receptors, and the inhibition by SOM230 of basal ACTH release by human corticotroph adenoma cells, even when the cells were pretreated with DEX, it was hypothesized that this multiligand SS analog may become a new medical treatment modality in a subgroup of patients with pituitary-dependent Cushing’s disease.

The present study was carried out to further evaluate the role of sst5 in suppressing ACTH secretion from pituitary corticotrophs, with particular emphasis on the role of glucocorticoids in regulating sst2- and/or sst5-mediated ACTH suppression. Studies were carried out using the mouse ACTH-producing AtT-20 corticotroph tumor cell line, which is known to express mainly sst2A+2B and sst5 subtypes (9, 42, 56).


Cell culture.

AtT-20/D16V mouse tumor cells (from Dr. J. Tooze, European Organization for Molecular Biology) were routinely passaged by trypsinization, as described in detail previously (21). The cells were maintained in 75-cm2 flasks in Dulbecco’s minimal essential medium (DMEM) supplemented with nonessential amino acids, sodium pyruvate (1 mmol/l), 10% fetal calf serum (FCS), penicillin (1 × 105 U/l), fungizone (0.5 mg/l), l-glutamine (2 mmol/l), and sodium bicarbonate (2.2 g/l), pH 7.6. The cells were cultured at 37°C in a CO2 incubator. Medium and supplements were obtained from GIBCO Bio-cult Europe (Invitrogen, Breda, The Netherlands). For crude cell membrane preparations, AtT-20 cells were scraped and centrifuged at 4°C for 5 min at 1,000 g. The cell pellet was either stored at −80°C or used directly. For ACTH release studies, the cells were seeded at a density of 20,000 cells/well in 1 ml of culture medium. After 72 h, the medium was changed, and a 2-h incubation without or with the SS analogs in the presence of CRH (10 nM) was performed. To evaluate the effect of glucocorticoids or SS analogs on SS analog (SS-14, OCT, SOM230, and BIM-23268; Table 1) -induced inhibition of ACTH release, the cells were pretreated for 48 h with the synthetic glucocorticoid dexamethasone (DEX; 10 nM) or for 72 h with SS-14, OCT, or SOM230 (all 10 nM), respectively. After this preincubation, the medium was changed, and the effect of SS analogs on CRH-induced ACTH secretion was evaluated as described above. At the end of the incubations, the medium was collected and stored at −20°C until hormone determination.

View this table:
Table 1.

Binding selectivity of SS analogs for the 5 human sst receptor subtypes

Radioligand-binding assay.

As previously described (9), the cells were resuspended in binding assay buffer [0.5% (wt/vol) bovine serum albumin (BSA), 10 mM N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid (HEPES), pH 7.5] by homogenization with a Polytron homogenizer (Kinematica) at 50 Hz for 30 s. of Cell homogenate (150 μl; ∼10,000–25,000 cells) was incubated with 50 μl of 125I-labeled Leu8-Trp22-Tyr25-SS-28 ([125I]LTT-SS-28; 2,175 Ci/mmol, 25–75 pM final concentration) in binding assay buffer containing MgCl2 (5 mM) and bacitracin (20 μg/ml), and either 50 μl of binding assay buffer alone (total binding), supplemented with 1 μM SS-14 (nonspecific binding) or with increasing concentrations of SOM230 or OCT. Experiments were conducted in triplicate. Incubation was terminated after 1 h at room temperature by vacuum filtration through glass fiber filters presoaked in 0.25% (wt/vol) poly(ethyleneimine). The filters were washed three times with ice-cold 10 mM Tris·HCl buffer containing 154 mM NaCl (pH 7.4) and dried. Bound radioactivity was measured in a Packard TopCount using liquid scintillation (65% counting efficiency).

[35S]guanosine 5′-O-(3-thiotriphosphate)-binding assay.

As previously described (8), the cell pellet was resuspended in 10 mM HEPES, pH 7.5, by Polytron homogenization at 50 Hz for 30 s, and centrifuged at 4°C for 30 min at 15,000 g. The microsome pellets were re-suspended in assay buffer (10 mM HEPES, 100 mM NaCl, 5 mM MgCl2, 0.1 mM EDTA, and 10 μg/ml bacitracin, pH 7.4), and either stored at −80°C or directly used. One hundred microliters per well of the microsome preparation (∼75,000 cells) were incubated in 96-well plates with [35S]guanosine 5′-O-(3-thiotriphosphate) (GTPγS; 1,030 Ci/mmol, 100–200 pM final concentration) in assay buffer containing 1 μM GDP and triplicates of assay buffer (basal) or 10 μM GTPγS, SOM230, or OCT at increasing concentrations. After a 5-min preincubation, 1.5 mg/well of wheat germ agglutinin (WGA) scintillation proximity assay (SPA) beads (Amersham) were added [beads in 50 mM Tris·HCl, 0.1% (wt/vol) sodium azide, pH 7.4], and the plates were sealed, incubated for 30 min at room temperature, and centrifuged for 10 min at 1,000 g. During assay incubation, cell membranes bind to WGA, effectively immobilizing the receptor-bearing membranes on to the SPA bead. The binding of [35S]GTPγS to such immobilized receptors brings the isotope into close proximity with the scintillant, which is incorporated within the bead. This allows the emitted radiation to stimulate the scintillant to emit light, which was measured (cpm) in a Packard TopCount. Percent stimulation of specific basal [35S]GTPγS binding was calculated as: 100 × [(experimental − basal)/(basal − nonspecific)].

Quantitative PCR.

Quantitative PCR was performed as previously described (16). Briefly, poly(A+) mRNA was isolated during Dynabeads oligo(dT)25 (Dynal, Oslo, Norway) from AtT20/D16V cell pellets containing 0.5 × 106 cells per sample. cDNA was synthesized using the poly(A+) mRNA, which was eluted from the beads in 40 μl of H2O for 2 min at 65°C, using oligo(dT)12–18 primer (Life Technologies). One-twentieth of the cDNA library was used for quantification of the sst subtype mRNA levels. A quantitative RT-PCR was performed by TaqMan Gold nuclease assay (PerlinElmer, Foster City, CA) and the ABI PRISM 7700 Sequence Detection System (PerkinElmer) for real-time amplification, according to the manufacturer’s instructions. The assay was performed using 15 μl of TaqMan Universal PCR Master Mix (Applied Biosystems, Nieuwerkerk aan den IJssel, The Netherlands), 300 nM forward primer, 300 nM reverse primer, 200 nM probe, and 10 μl of cDNA template, in a total reaction volume of 25 μl. The detection of proopiomelanocortin (POMC) mRNA was performed as a control for the negative feedback regulation by glucocorticoids on POMC gene expression. The detection of hypoxanthine phosphoribosyltransferase (HPRT) mRNA served as a control and was used for normalization of the POMC and sst subtype mRNA levels. The specific mouse primer (Biosource, Nivelles, Belgium) and probe sequences that were used are shown in Table 2.

View this table:
Table 2.

Specific mouse primer and probe sequences

The relative amounts of POMC, sst2, and sst5 mRNA were determined by means of a standard curve generated in each experiment from known amounts of mouse genomic DNA. For the determination of the amount of HPRT mRNA, a standard curve was obtained by including dilutions of a pool of cDNAs known to contain HPRT. The amounts of POMC, sst2, and sst5 mRNA were calculated relative to the amount of HPRT and is given in arbitrary units.

sst Membrane-binding studies.

The method of membrane isolation and the reaction conditions were previously described (22). Briefly, membrane preparations (corresponding to 30–50 μg protein) of cultured AtT-20 cells, in the presence or absence of 10 nM DEX (48 h), were incubated in a total volume of 100 μl at room temperature for 60 min with increasing concentrations of [125I-Tyr11]SS-14 or [125I-Tyr3]OCT with and without excess (1 μM) of unlabeled SS-14 or OCT, respectively, in HEPES buffer (10 mM HEPES, 5 mM MgCl2, and 0.02 g/l bacitracin, pH 7.6) containing 0.2% BSA. After the incubation, 1 ml of ice-cold HEPES buffer was added to the reaction mixture, and membrane-bound radioactivity was separated from unbound by centrifugation for 2 min at 14,000 rpm in an Eppendorf microcentrifuge. The remaining pellet was washed twice in ice-cold HEPES buffer, and the final pellet was counted in a γ-counter. Specific binding was taken to be total binding minus binding in the presence of 1 μM unlabeled SS-14 or OCT. As a control for binding, rat brain cortex membranes were used.

Hormone determination.

Mouse immunoreactive (ir)-ACTH concentrations were determined by a nonisotopic, automatic chemiluminescence immunoassay system (Immulite, DPC), as described previously for the detection of rodent ir-ACTH (33). The intra- and interassay coefficients of variation for ACTH were 5.6 and 7.8%, respectively. Under the conditions employed, the assay detects 1 fmol/tube of ir-ACTH. Dilution curves of the samples were parallel with those of the standard, and in addition, biological specificity of the results was in agreement with hypothalamic-pituitary-adrenal physiology.

Test substances.

OCT (Sandostatin) was obtained from Novartis Pharma (Basel, Switzerland). SOM230 was synthesized at Novartis Pharma. SS-14 was purchased from Sigma Chemical (St. Louis, MO). BIM-23268, an sst5 subtype-specific analog, was synthesized at IPSEN Pharmaceuticals (Milford, MA). DEX was derived from the pharmacy department of the Erasmus Medical Center (Rotterdam, The Netherlands). CRH was purchased from Ferring (Hoofddorp, The Netherlands). [125I]LTT-SS-28 was custom synthesized by ANAWA (Wangen, Switzerland). [125I-Tyr11]SS-14 was purchased from Amersham (Houten, The Netherlands). The SS analog [Tyr3]OCT was iodinated with 125I by the chloramine-T method and purified by HPLC, as described previously in detail (2). Specific radioactivity of all radioligands yielded ∼2,000 Ci/mmol. [35S]GTPγS was purchased from Amersham (Freiburg, Germany).

Statistical analysis.

The statistical significance of the difference between mean values regarding the effects of the SS analogs on ACTH release was determined using one-way analysis of variance. When significant overall effects were obtained by this method, comparisons were made using the Newman-Keuls multiple comparisons test. The unpaired Student’s t-test was chosen to analyze for statistical significance in the experiments determining the effects of DEX treatment on POMC and sst mRNA expression levels. pKd, EC50, and IC50 values were determined by nonlinear regression curve analysis of the concentration-effect responses by using the computer programs ActivityBase and GraphPad Prism. The unpaired Student’s t-test was chosen to analyze differences in concentration-effect curves. Data are reported as means ± SE of the indicated n values, unless otherwise specified.


Radioligand and [35S]GTPγS-binding assays.

The binding properties of SOM230 and OCT at [125I]LTT-SS-28-labeled sites in AtT-20 cell membranes were established in competition experiments. As shown in Fig. 1A, SOM230 had a sevenfold higher ligand binding affinity than OCT for [125I]LTT-SS-28 labeled sites in AtT-20 cell membranes [IC50 0.18 nM (pKd 9.74 ± 0.08, n = 8) vs. 1.2 nM (pKd 8.92 ± 0.03, n = 3), respectively, P < 0.0001]. In microsome preparations from AtT-20 cells, both SOM230 and OCT produced a concentration-dependent increase in [35S]GTPγS binding (Fig. 1B). Interestingly, SOM230 showed a clear, 19-fold higher potency in stimulating GTPγS binding compared with OCT [EC50 8.71 nM (pKd 8.09 ± 0.11, n = 12) and 169.8 nM (pKd 6.78 ± 0.07, n = 3), respectively, P < 0.0001].

Fig. 1.

A: inhibition of radioligand binding in AtT-20 cell membranes. Crude membrane preparations were incubated with 125I-labeled Leu8-Trp22-Tyr25-SS-28 ([125I]LTT-SS-28), and increasing concentrations of multiligand somatostatin (SS) analog SOM-230 (•) and octreotide (OCT; ○). Data are expressed as %specific binding. Data points represent means ± SE (bars) of data from 3–8 different experiments run in triplicate. B: stimulation of [35S]guanosine 5′-O-(3-thiotriphosphate) (GTPγS) binding in AtT-20 cell membranes. Dose-response effects of SOM230 (•) and OCT (○). Data are expressed as %stimulation of specific basal [35S]GTPγS binding. Data points represent means ± SE (bars) of data from 3–12 different experiments run in triplicate.

Effect of SS-14 and SS analogs on CRH-induced ACTH secretion in AtT-20 corticotroph cells with or without glucocorticoid pretreatment.

Full dose-response curves were performed to ascertain which concentrations of DEX and CRH should be used in our experiments. As can be seen in Fig. 2, DEX concentration-dependently suppressed POMC mRNA expression in AtT-20 cells after 24 h (Fig. 2A) and 48 h (Fig. 2B), whereby 10 nM was most effective. CRH induced ACTH release by AtT-20 cells in a dose-dependent way as well, which reached its maximum effect at 10 nM CRH (Fig. 2C). Because of the physiological negative feedback loop within the hypothalamic-pituitary-adrenal axis, the effect of 48-h DEX (10 nM) pretreatment was assessed (Fig. 2D). ACTH release by AtT-20 cells was decreased when cultured for 48 h with DEX (from 1,213 ± 32 to 75 ± 1.8 fmol·l−1·well−1, P < 0.0001). Nevertheless, DEX-treated AtT-20 cells remained responsive to CRH (10 nM) stimulation (Fig. 2D).

Fig. 2.

Effect of glucocorticoids and CRH on proopiomelanocortin (POMC) mRNA and ACTH release in AtT-20 corticotroph tumor cells. Cells were preincubated for 24 (A) and 48 h (B) with dexamethasone (DEX; range 0.01–10 nM). Subsequently, poly(A+) mRNA was isolated and used to determineexpression levels of POMC by quantitative PCR [corrected for hypoxanthine phosphoribosyltransferase (HPRT)]. Values represent %change of POMC mRNA transcripts under DEX treatment vs. control and are means ± SE of 2 independent experiments. #P < 0.001 vs. control, 0.1 and 0.01nM DEX; *P < 0.05 vs. control, 0.1 and 0.01 nM DEX. C: concentration effect curve of 3-h incubation with CRH (0.01–1,000 nM) on ACTH release in AtT-20 cells. Values represent %change of ACTH release vs. control and are means ± SE of 2 independent experiments run in quadruplicate. D: effect of DEX (10 nM; 48 h) on CRH-induced ACTH release in AtT-20 cells. After 48 h, medium was refreshed, and cells were incubated for 2 h in the absence or presence of DEX and CRH (10 nM), after which medium was collected for ACTH determination. Values are expressed as means ± SE of 4 wells/treatment group. Left: CRH-induced ACTH release in the absence of DEX (no Dex); right: CRH-induced ACTH release after 48-h 10 nM DEX treatment. #P < 0.0001 vs. basal ACTH release in the absence of DEX; *P < 0.001 CRH vs. basal.

To evaluate the effects of DEX pretreatment on CRH-induced ACTH release in AtT-20 cells, we first investigated the inhibitory potencies of the SS analogs on CRH-induced ACTH release in AtT-20 cells under basal conditions. Each SS analog (100 nM) potently inhibited CRH-induced ATCH release varying between 58 and 75% suppression (BIM-23268, P < 0.05 vs. OCT and SS-14; SOM230 vs. OCT, P < 0.05). The dose-dependent inhibition of ACTH release by SS-14 (Fig. 3A) and SOM230 (Fig. 3B) was not affected by 48-h DEX (10 nM) and the corresponding IC50 values for SS-14 [1.3 nM without vs. 0.7 nM with DEX, P = not significant (NS)] and SOM230 (0.06 nM without vs. 0.07 nM with DEX, P = NS) remained unchanged. Interestingly, the concentration-response curve for the inhibitory effects of BIM-23268 on CRH-induced ACTH release showed a shift to the left, indicating increased potency, when pretreated with DEX, with the IC50 value shifting 34-fold from 3.4 to 0.1 nM in the presence of DEX (P < 0.05; Fig. 3C). In contrast, DEX pretreatment caused the concentration-response curve of OCT to shift to the right, indicating decreased potency. The IC50 indicates a 20-fold shift from 0.2 nM without DEX to 4.3 nM in the presence of DEX (P < 0.05; Fig. 3D). Both 100 nM BIM-23268 and 100 nM SOM230 showed significantly enhanced suppressive effects on CRH-induced ACTH release in the presence of DEX (P < 0.01 and P < 0.05 vs. without DEX, respectively). No differences in the maximal inhibitory effects were observed when the cells were incubated with 100 nM OCT without or with DEX; however, as shown in Fig. 4, the inhibitory effect of 1 nM OCT on CRH-induced ACTH release was almost completely abolished by DEX treatment (−39 ± 2% without vs. −9 ± 6% with DEX, P < 0.05). In contrast, DEX treatment enhanced the ability of 1 nM BIM-23268 (−15 ± 1% without vs. −60 ± 7% with DEX, P < 0.01; Fig. 4) to suppress CRH-induced ACTH release. DEX did not affect the inhibitory effects of 1 nM SOM230. Both SOM230 and BIM-23268 were significantly more efficacious than OCT in suppressing ACTH release under DEX treatment, not only at supraphysiological but also at physiological concentrations (P < 0.05, BIM-23268 and SOM230 vs. OCT).

Fig. 3.

Effect of glucocorticoids on concentration-dependent inhibition of CRH-induced ACTH release in mouse corticotroph tumor cells by different SS analogs. AtT-20 cells were preincubated for 48 h with 10 nM DEX. After 48 h, medium was collected and exchanged for new medium containing 10 nM CRH and different concentrations of SS-14 (A), SOM230 (B), BIM-23268 (C), or OCT (D). Incubation was performed for an additional 2 h, and medium was collected for ACTH determination. Calculation of IC50 values was performed using GraphPad Prism, and the unpaired Student’s t-test was chosen to analyze differences in IC50 values. □, Control; ▪, 10 nM DEX. #P < 0.05.

Fig. 4.

Effect of glucocorticoids on SS analog- SS analog-mediated inhibition of CRH-induced ACTH release by AtT-20 cells. Cells were preincubated with or without 10 nM DEX. After 48 h, medium was refreshed, and cells were incubated for 2 h in the absence or presence of 10 nM DEX, CRH (10 nM) and indicated compounds at a concentration of 1 nM. Subsequently, medium was collected for ACTH determination. Values are expressed as means ± SE of 4 wells/treatment group. □, Control; ▪, 10 nM DEX. #P < 0.05 SOM230 and BIM-23268 vs. OCT. ∞P < 0.05 OCT vs. BIM-23268. *P < 0.05, SOM230 vs. OCT and BIM-23268.

Effect of glucocorticoid treatment on sst2 + sst5 mRNA levels.

To determine whether glucocorticoids display a regulatory role on both sst2 isoforms and/or sst5 mRNA expression levels, AtT-20 cells were exposed to 10 nM DEX for 24 and 48 h. As shown in Fig. 5, DEX significantly suppressed POMC mRNA levels after 24 and 48 h (−55 ± 2 and −74 ± 2%, respectively, both P < 0.001 to control). sst2A and sst2B mRNA levels were potently suppressed after 24 h of DEX treatment by 30 ± 6 and 45 ± 4%, respectively (Fig. 5, left: P < 0.001), which remained significantly lower after 48 h (Fig. 5, right). In contrast, sst5 mRNA expression remained constant after 24 h of DEX and was not significantly affected (20 ± 8%; P = NS) after 48 h DEX treatment.

Fig. 5.

Effect of glucocorticoids on SS receptor subtype (sst) mRNA expression levels in AtT-20 pituitary adenoma cells. AtT-20 cells were preincubated for 24 h (left) and 48 h (right) without or with 10 nM DEX. Poly(A+) mRNA was isolated from untreated and treated AtT-20 cells and used to determine expression levels of POMC or sst mRNA by quantitative PCR (corrected for HPRT). Values represent %change of different mRNA transcripts under DEX treatment vs. control and are means ± SE of 3 independent experiments. □, Control; ▪, 10 nM DEX. *P < 0.001.

Effect of glucocorticoid treatment on sst binding sites.

To determine whether glucocorticoids display a regulatory role on sst2 and/or sst5 at the protein level as well, AtT-20 cells were cultured in the presence or absence of 10 nM DEX for 48 h and subsequently collected for cell membrane-binding assays with two different radiolabeled somatostatin analogs: [125I-Tyr3]OCT, displaying superior binding affinity for only sst2, and [125I-Tyr11]SS-14, which can bind to both sst2 and sst5 with good affinity. Maximum binding capacity (Bmax) was dramatically suppressed by 72% in the presence of 10 nM DEX when [125I-Tyr3]OCT was used as radioligand (Bmax 1,404 to 390 fmol/mg in the absence and presence of DEX, respectively; Fig. 6A). As depicted in Fig. 6B, DEX attenuated radioligand binding by only 17% when [125I-Tyr11]SS-14 was used as radioligand (Bmax 1,581 and 1,315 fmol/mg in the absence and presence of 10 nM DEX, respectively).

Fig. 6.

sst Membrane-binding studies. A: Scatchard analysis of [125I-Tyr3]OCT binding to cell membranes of AtT-20 cells, cultured in the absence (Kd = 0.1 nM, n = 1,404 fmol/mg) or presence of 10 nM DEX (48 h; Kd = 0.7 nM, n = 390 fmol/mg). B: Scatchard analysis of [125I-Tyr11]SS-14 binding to cell membranes of AtT-20 cells cultured in the absence (Kd = 0.1, n = 1,581 fmol/mg) or presence of 10 nM DEX (48 h; Kd = 0.1 nM, n = 1,315 fmol/mg). □, Control; ▪, 10 nM DEX.

Effect of SS analog pretreatment.

First, it should be noted that SS analog pretreatment lowered the inhibitory effects of all SS analogs (10 nM) on CRH-induced ACTH release (from 47–69 to 18–54% range of suppression). To evaluate the effect of continued exposure of AtT-20 cells to a maximal inhibitory concentration of the different SS analogs, AtT-20 cells were preincubated for 72 h with SS-14, OCT, or SOM230 (10 nM), followed by 2 h of incubation with CRH (10 nM) and SS-14, OCT, SOM230, and/or BIM-23268 (all 10 nM). As depicted in Fig. 7A, the inhibitory effects of SOM230 (−48%) and BIM-23268 (−54%) appeared to be twofold more efficacious compared with OCT (−22%, *P < 0.01 vs. SOM230 and OCT) when AtT-20 cells were treated for 72 h with SS-14. Comparable superior suppressive effects by SOM230 and BIM-23268, compared with OCT, were observed when cells were pretreated with OCT (Fig. 7B) or SOM230 (Fig. 7C).

Fig. 7.

Effect of prior SS analog treatment on SS analog-mediated inhibition of CRH-induced ACTH release by AtT-20 cells. Cells were preincubated with 10 nM SS-14 (A), OCT (B), or SOM230 (C). After 72 h, medium was refreshed, and cells were incubated for 2 h in the absence or presence of 10 nM SS analog CRH (10 nM) and indicated compounds (10 nM), after which medium was collected for ACTH determination. Values are expressed as means ± SE of 4 wells/treatment group. *P < 0.01 vs. OCT and SS-14; #P < 0.05 vs. OCT and SS-14.


Because purified populations of corticotrophs are difficult to prepare and require substantial amounts of pituitary tissue, we based our study on the mouse pituitary corticotroph (AtT-20/D16V) cell line after first confirming by quantitative RT-PCR the expression of sst2A, sst2B, and sst5 mRNA. Although it is appreciated that cell lines are not necessarily representative of their parent cell types, we reasoned that this clonal population, which has been studied intensively as a well-accepted cellular model for corticotrophs (54), would provide a useful model in which to explore fundamental mechanisms of the effects of SS (analogs) on ACTH release and the role of sst subtypes herein.

In the present study, predominance of sst2 and sst5 was found, as both OCT and SOM230 showed high-affinity binding. These observations are well in agreement with recent extensive pharmacological studies that indicated sst2 and sst5 are mainly expressed with no or negligible presence of sst1, sst3, and sst4 in AtT-20 cells (9). The inhibitory effects of SS-14, SOM230, OCT, and BIM-23268 on CRH-induced ACTH release in AtT-20 cells are well in agreement with the concept that both sst2- and sst5-binding SS analogs potently inhibit CRH-induced ACTH release by AtT-20 cells (52). Moreover, the maximal inhibitory effect of the two SS analogs that bind with very high affinity to sst5, i.e., SOM230 and BIM-23268, was significantly higher compared with OCT. This apparent functional superiority of sst5 over sst2 seems to confirm recent observations by Cervia et al. (8), as well as our membrane- and GTPγS-binding results showing the superior profile of SOM230 compared with OCT.

In the present study, we found that glucocorticoid treatment induced remarkable differences with respect to the role of sst2 and sst5 in regulating ACTH release. We observed a profound difference in efficacy between the sst2-specific analog OCT and the multiligand SS analog SOM230 and the sst5-preferring analog BIM-23268. Because both SOM230 and BIM-23268 still potently inhibited CRH-induced ACTH release under DEX treatment whereas the suppressive effects of OCT were almost completely blocked in the “physiological” nanomolar range, it is suggested that sst2 is downregulated by glucocorticoid treatment whereas sst5 is more resistant. The intriguing differences in the functional properties of sst2 and sst5 in mediating ACTH release under DEX pressure is further supported by our observations that the IC50 values for both OCT and BIM-23268 during DEX treatment shift toward their sst5 binding affinity, i.e., a 20-fold decrease and a 34-fold increase, respectively, in their potency to suppress ACTH release. Our data not only confirm earlier observations that glucocorticoid treatment abolishes the inhibitory effect of OCT on ACTH release in vitro (50) but also seem well in agreement with other reports that demonstrate that SS only suppressed ACTH release by rat pituitary cells from long-term-adrenalectomized rats (59) or when the cells were cultured in the absence of glucocorticoids (27, 31). One group did observe ACTH suppression by SS-28, which is known to have preferable binding affinity for sst5, in a primary cell culture from a Nelson’s tumor and from an ACTH-secreting pituitary adenoma (47, 48). However, in patients with Nelson’s syndrome and adrenal insufficiency of different origin, both SS and OCT lower ACTH secretion (3, 14, 29, 55), again suggesting that glucocorticoids influence the sensitivity of the corticotroph cells for SS and OCT. Because glucocorticoids are also well known to mediate biological effects through regulation of gene expression, it was of interest to test the effects of glucocorticoids on sst2 and sst5 gene expression in AtT-20 cells. The primary transcript of the sst2 gene is alternatively spliced in a long (sst2A) and a short (sst2B) form (40, 42, 5658). The current mRNA data again support the concept that glucocorticoid treatment differentially influences sst2A+2B expression compared with sst5. It is known that the mouse sst2 gene promoter sequence is the only sst receptor that has been shown directly to be transcriptionally regulated by glucocorticoids (25, 26), whereas the mouse sst5 gene possesses multiple glucocorticoid-responsive element half-sites (17). Therefore, the immediate and powerful suppression of both sst2A and sst2B suggests a direct effect of DEX at the transcriptional level, whereas sst5 could be regulated in a different way. Support for transcriptional downregulation of sst1+2+3 mRNA expression by glucocorticoids comes from studies in rat pituitary GH4C1 cells (61). Influencing mRNA stability could be involved as well, since the addition of an RNA synthesis inhibitor produces no disruption of the ability of DEX to suppress either sst2 or sst3 mRNA levels in cultured rat pituitary cells (38). Our observation that sst2 and sst5 receptors might be regulated differently by glucocorticoids was also recently observed by Park et al. (38): in a rat model, DEX inhibited sst2 mRNA expression both in vivo and in vitro but enhanced sst5 mRNA expression. Moreover, earlier observations had already demonstrated that glucocorticoids lower [125I-Tyr11]SS-14 binding in AtT-20 cells, but sst receptors were not characterized (46). Our data confirm that [125I-Tyr11]SS-14 binding is attenuated in AtT-20 cells after DEX treatment. However, [125I-Tyr3]OCT binding, i.e., the presence of only sst2 receptor subtypes, was reduced by almost 75% by DEX, indicating that sst2 at the protein level is dramatically decreased. Because [125I-Tyr11]SS-14 binding displays both sst2 and sst5 binding affinity and is only slightly lowered by DEX, it becomes suggestive that this universal radioligand represents predominantly sst5 receptors when AtT-20 cells are treated with glucocorticoids. These experiments clearly support our mRNA data and demonstrate at the protein level that sst5, compared with sst2, seems less sensitive to DEX treatment as well.

The downregulation of sst2A+2B mRNA and protein levels in AtT-20 cells by glucocorticoids may be an explanation for the lack of efficacy of OCT in lowering ACTH and cortisol levels in patients with untreated Cushing’s disease (30). Thus the observed ability of sst5 to suppress ACTH release in AtT-20 cells, which appears to be relatively resistant to glucocorticoids, might be a new target for therapeutic agents that could lower ACTH and cortisol levels in a subgroup of patients with Cushing’s disease. Furthermore, we propose that sst2- plus sst5-preferring SS analogs, such as SOM230, might become of therapeutic interest in Cushing’s disease as well. The suppression of ACTH levels by activation of sst5 in patients with Cushing’s disease might lower cortisol levels. Because cortisol inhibits sst2 expression, these suppressive effects might subsequently be (partially) abrogated. In this relative hypocortisolemic state, enhanced ACTH inhibition via restored sst2 expression becomes suggestive. Therefore, SOM230 may be able to lower ACTH levels in Cushing’s disease even more, because it can now function via both sst5 and sst2 receptor subtypes. Nevertheless, it should be kept in mind that this tantalizing hypothesis needs further studies to confirm its rationale. These studies should be confirmed in primary cultures of rodent corticotroph cells and in living animals as well, before a well-designed clinical trial in patients with Cushing’s disease can be performed.

Internalization of receptor-ligand complexes has been shown to play a role in desensitization (23), leading to tachyphylaxis of the inhibitory effect of sst2-preferring analogs on hormone secretion in a subgroup of neuroendocrine tumors (15, 19). Prolonged treatment of AtT-20 cells with SS-14 results in desensitization of its inhibitory effect on ACTH secretion and cAMP formation (44), and prolonged exposure of AtT-20 cells to SS-14 and SS-28 has been shown to downregulate SS-14 receptors (18). Interestingly, we recently observed (20) that prolonged SOM230, but not OCT, treatment inhibited basal ACTH release both from primary cultures of ACTH-producing pituitary and from AtT-20 cells, suggesting that sst2 and sst5 appear to respond differently after continued ligand activation. The ability of the sst5-preferring analogs to continue to suppress ACTH levels, independent of prior 72-h SS analog exposure and the decreased efficacy of OCT, supports other data that sst2 desensitizes on continued ligand activation and suggests that the sst5 receptor may be more resistant to desensitization. It has already been demonstrated that sst5 receptors are rapidly recycled and restored from intracellular storage after agonist activation, which might protect this particular receptor from long-term downregulation (51). In addition, it was recently demonstrated in live, transfected AtT-20 cells that only fluorescein protein (FP)-tagged sst2 subtype, but not FP-sst5, internalized upon ligand activation (4). The FP-sst5 subtype remained localized to the membrane during treatment with either an sst5-preferring agonist, an sst2+5-biselective agonist, SS-14, or SS-28.

In summary, this study demonstrates that sst5 receptors display intriguing functional properties in regulating ACTH release in mouse corticotroph tumor cells. Moreover, the recent observation that in sst5 knockout mice serum ACTH and cortisol levels were elevated compared with wild-type mice supports the concept that sst5 receptors are important in the regulation of ACTH release in mice (53). On the basis of the potent suppressive effects on ACTH release by sst5-preferring analogs, the relative resistance of sst5 expression and action to DEX suppression as well as to prolonged SS analog exposure, and our recent observation that sst5 is the predominant sst expressed in human corticotroph adenomas (20), we propose that sst5 may become a new therapeutic target for the control of ACTH and cortisol hypersecretion in untreated patients with pituitary-dependent Cushing’s disease.


This work was supported, in part, by European Community Contract QLG3-CT-1999-00908 and Swiss Grant BBW 00-0427.


  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


View Abstract