Central and peripheral actions of somatostatin on the growth hormone-IGF-I axis

Abstract

Somatostatin (SRIF) analogs provide safe and effective therapy for acromegaly. In a proportion of patients, however, SRIF analogs may lead to discordant growth hormone (GH) and IGF-I suppression, which suggests a more complex mechanism than attributable to inhibition of GH release alone. To elucidate whether SRIF acts peripherally on the GH-IGF-I axis, we showed that rat hepatocytes express somatostatin receptor subtypes-2 and -3 and that IGF-I mRNA and protein levels were suppressed in a dose-dependent manner by administration of octreotide. The inhibitory effect of SRIF was not apparent without added GH and in the presence of GH was specific for IGF-I induction and did not inhibit GH-induced c-myc or extracellular signal regulated kinase (ERK) phosphorylation. Pertussis toxin treatment of hepatocytes incubated with GH and SRIF, or with GH and octreotide, abrogated the inhibitory effect on GH-induced IGF-I, which confirms the requirement for the inhibitory G-protein. Treatment with SRIF and GH increased protein tyrosine phosphatase (PTP) activity and inhibited signal transducer and activator of transcription-5b (STAT5b) phosphorylation and nuclear localization. Octreotide also inhibited GH-stimulated IGF-I protein content of ex vivo-perfused rat livers. The results demonstrate that SRIF acts both centrally and peripherally to control the GH-IGF-I axis, providing a mechanistic explanation for SRIF analog action in treating patients with GH-secreting pituitary adenomas.

Figure 1

Cultured isolated rat hepatocytes incubated with (A) GH (1–1,000 ng/ml) for 24 hours and (B) GH (500 ng/ml) or buffer for 0–24 hours. Cells were harvested for RNA and analyzed for IGF-I mRNA and β-actin by RPA. The IGF-I riboprobe corresponded to a 166-bp fragment of the mature IGF-I protein, and β-actin corresponded to a 125-bp band. Upper and lower panels: representative blots; middle panel: mean fold IGF-I mRNA induction of 3 individual studies. *P < 0.05 compared with control.

Figure 2

SRIF suppresses GH-induced hepatocyte IGF-I. (A) Isolated rat hepatocytes were incubated with GH (100 or 500 ng/ml) for 24 hours with or without 1 hour preincubation with 100 nM SRIF or octreotide (OCT). RNA was quantitated by RPA analysis. Upper panel: representative blot. Lower panel: mean fold induction of IGF-I mRNA of three individual experiments after normalization for β-actin. *P < 0.05 compared with GH-treated cells. (B) Representative RPA quantitation for IGF-I and β-actin content of isolated rat hepatocytes following incubation with 100 nM SRIF or octreotide for 24 hours with no GH stimulation. Upper panel: IGF-I; lower panel: β-actin. (C) RPA analysis of IGF-I and β-actin from isolated rat hepatocytes treated with GH (500 ng/ml) for 24 hours with and without preincubation with 0.01–100 nM SRIF (upper two panels), or 0.1–500 nM octreotide for 1 hour (lower two panels).

Figure 3

Octreotide suppresses hepatic IGF-I in vitro and in vivo. (A) Isolated rat hepatocytes incubated with vehicle or GH (500 ng/ml) in the presence and absence of 100 nM SRIF or octreotide. Medium was collected after 24 hours and assayed for IGF-I content by RIA after acid-alcohol extraction. (B) Rat livers perfused ex vivo with Krebs buffer with or without 0.5 μg/ml GH in the presence or absence of octreotide (20 ng/g body wt/h) for 60 minutes. GH was perfused from minutes 15–25 and octreotide from minutes 0–60. *P < 0.05 compared with control; ^#P < 0.05 compared with GH-perfused livers.

Figure 4

Primary hepatocytes were incubated with vehicle (Control), SRIF, or OCT for 30 minutes before addition of ¹²⁵I-GH. Additional wells were treated in the presence of 5 μg unlabeled GH to quantify residual nonspecific binding. Cells were incubated for 30 minutes in the presence of ¹²⁵I-GH, washed 3 times, and harvested and radioactivity counted. Graph represents total binding with indicated treatment minus nonspecific binding (specific binding). *P < 0.05 compared with control wells.

Figure 5

Hepatocyte GH signaling. (A) Western blot analysis of 50 μg protein from primary rat hepatocytes treated with GH in the absence or presence of SRIF for 1–30 minutes. Membranes were blotted for phospho-ERK (pERK; upper panel, 42 kDa) and ERK (lower panel). (B) Northern blot analysis of RNA from cultured primary hepatocytes treated with vehicle, GH, or GH in the presence of SRIF or octreotide for 30 minutes. Membranes were probed for c-myc (upper panel), stripped, and reprobed for GAPDH to correct for loading.

Figure 6

SRIF suppression of hepatic IGF-I does not require new protein induction. (A) Isolated primary hepatocytes were pretreated with vehicle or pertussis toxin (PTX, 100 ng/ml) for 24 hours; vehicle, 100 nM SRIF, or octreotide were added; and the cells stimulated with GH (500 ng/ml) one hour later. Cells were incubated for an additional 20 hours and cells harvested. RNA was analyzed for IGF-I content by RPA and loading normalized by β-actin quantitation. The upper panel shows a representative blot, and the lower panel shows mean fold induction of IGF-I mRNA after normalization for β-actin. (B) Primary hepatocytes were pretreated with vehicle or cycloheximide (CHX, 10 μg/ml) for 2 hours and vehicle, 100 nM SRIF, or octreotide for 1 hour before stimulation with GH (500 ng/ml). Cells were incubated for an additional 6 hours and harvested. RNA was analyzed for IGF-I content by RPA and loading normalized by quantitation of β-actin. (C) Primary hepatocytes were treated as for B and IGF-I RNA content quantitated by real-time PCR and normalized for β-actin.

Figure 7

SRIF suppression of STAT5b. (A) PTP activity was measured in primary hepatocytes treated with vehicle (Control), GH (500 ng/ml), and coincubations of GH with 100 nM SRIF or octreotide. PTP activity was assessed by incubation of cell lysates with tyrosine phosphopeptides and the phosphate production measured using a molybdate dye. *P < 0.05. (B) STAT5b was immunoprecipitated from whole-cell lysates of primary hepatocytes treated with vehicle, GH (500 ng/ml), and cotreatments of GH and 100 nM SRIF for 20 minutes. Immunoprecipitates were subject to polyacrylamide gel electrophoresis and membranes blotted for phospho-STAT5b (upper panel) and, after stripping, for STAT5b (lower panel). Phosphorylated STAT5b (pSTAT5b)/STAT5b was visible as a 92-kDa band; a second band that migrated more slowly was visible at approximately 96 kDa and likely represents serine phosphorylated STAT5b (arrows). (C) STAT5b was immunoprecipitated from hepatocyte nuclear extracts treated with GH (500 ng/ml) or cotreated with GH and 100 nM SRIF. Following electrophoresis, membranes were probed for STAT5b. The lower arrow corresponds to the heavy chain of the IgG used in the immunoprecipitation.