In biological fluids, the insulin-like growth factors (IGFs) are associated with binding proteins (IGFBPs), which modify IGF distribution and action. Circulating IGFs are bound predominantly to IGFBP-3, of apparent hepatic origin, but regulation of IGFBP-3 has been difficult to dissect because of the lack of systems suitable for examining hepatic production of IGFBP-3 in vitro. In the present studies, IGFBP-3 expression was identified primarily in hepatic nonparenchymal cells, particularly Kupffer and sinusoidal endothelial cells. Coculture with hepatocytes enhanced the stability of nonparenchymal cells to express IGFBP-3 in vitro. IGFBP-3 in conditioned medium had apparent mol wt of 150-300 kilodaltons, suggesting formation of a ternary complex with IGFs and the acid-labile subunit. Expression and secretion of IGFBP-3 were hormonally responsive and strongly correlated (r = 0.79; P < 0.001), with 2- to 3-fold stimulation by added insulin or IGF-I (both P < 0.05), but not by added GH alone. Our findings suggest that GH may act indirectly to promote IGFBP-3 generation in vivo via increasing both the secretion of insulin and the hepatic production of IGF-I; in patients with diabetes mellitus, reduced circulating levels of IGFBP-3 despite high levels of GH may result from both insulin deficiency and inadequate hepatic production of IGF-I. Coculture of hepatic nonparenchymal and parenchymal cells should be useful for further analysis of the mechanism of IGFBP-3 regulation.

To determine the effect of the somatostatin analog octreotide on glucose tolerance in acromegaly, we examined glucose profiles, oral glucose tolerance and the insulinogenic index in patients treated with this analog. Ninety patients participated in a long-term, prospective, open-label study. There was no significant change between mean daily blood glucose profiles at baseline or during octreotide treatment. Using glucose tolerance test criteria, 61% of 90 patients had normal baseline glucose tolerance. While on octreotide, 20% and 29% of these patients, respectively, developed impaired glucose tolerance or became frankly diabetic. Conversely, three of the patients who were diabetic at baseline (N = 11) became normal (18%) or developed impaired glucose tolerance (9%) during octreotide therapy. There was no relationship between the dose of octreotide and change in glycemic state. The insulinogenic index (insulin/glucose) response to a glucose challenge decreased uniformly in octreotide-treated patients. Female patients and those with elevated baseline insulin levels were more likely to develop diabetes mellitus during octreotide therapy. In conclusion, octreotide significantly alters glucose tolerance in patients with acromegaly, mandating glucose monitoring during this form of therapy.

Short stature and decreased growth velocity are prominent features of endogenous and pharmacological glucocorticoid excess in children. Underlying processes may involve direct cellular effects or defective generation of insulin-like growth factors (IGFs) and/or IGF-binding proteins (IGFBPs), which modulate IGF-stimulated events and regulate growth. To evaluate potential mechanisms, we investigated the impact of dexamethasone (dex) on hepatic expression of IGFBP-3, the major carrier protein for IGFs. Using cocultured hepatic parenchymal and nonparenchymal cells, dex at 10(-8) and 10(-6) M decreased IGFBP-3 secretion by 67 +/- 9% and 73 +/- 9%, respectively (both P < 0.05 vs. no dex). In a separate experiment, IGFBP-3 messenger RNA (mRNA) expression was decreased by 84 +/- 2% and 75 +/- 2% (both P < 0.05 vs. no dex). In combined studies, levels of IGFBP-3 protein in conditioned medium were strongly correlated with the abundance of IGFBP-3 mRNA (r = 0.75; P < 0.01), consistent with regulation at a pretranslational level. After inhibition of transcription, levels of IGFBP-3 mRNA decreased 85% and 86% over 24 h in cells cultured in 10(-6) M and no dex, respectively; the t1/2 was 13.6 h at 10(-6) M and 12.6 h with no dex, indicating that dex had no effect on IGFBP-3 mRNA stability. To evaluate transcriptional effects, the rate of IGFBP-3 gene transcription was measured by incorporation of [alpha-32P]UTP into preinitiated message in isolated nuclei and fell 78% after the addition of 10(-6) M dex for 48 h (compared to cells cultured in 10(-9) M dex), an inhibition of a magnitude similar to the effects on protein release and mRNA abundance. We conclude that dex may reduce the production of IGFBP-3 by inhibiting IGFBP-3 gene transcription.