The effect of cholecystokinin on proinsulin gene expression in isolated mouse pancreatic islets.
S. Lindskog,B. Ahrén,D. Malm,J. Florholmen
DOI: https://doi.org/10.1097/00006676-199611000-00013
1996-11-01
Pancreas
Abstract:To the Editor: Cholecystokinin (CCK) has an established insulin-releasing effect and potentiates glucose-induced insulin secretion in vivo ( I ) and in vitro (2,3). However, so far, only the acute effects of CCK on islet physiology, e.g., insulin release and underlying intracellular mechanisms, have been the subject of extensive research (for review see ref. 4). Whether the peptide also affects islet function on a long-term basis is not known. For example, insulinotruphic agents may affect not only the release, but also the intracellular production of insulin, illustrated by glucose, which increases the rate of insulin biosynthesis (5 ) . Several studies in p cells have shown a glucose-induced increase in levels of proinsulin mRNA that involves both transcription rate (6) and mRNA stability (7). Also, gastrointestinal peptides have been found to affect insulin gene expression. In mouse insulinoma TC-I cells, glucagon-like peptide (GLP-I) stimulates proinsulin gene expression, an effect that is reversed by galanin (8). Further, a receptor binding hormone, such as glucagon, increases (9), whereas somatostatin inhibits (lo), the production of proinsulin mRNA in HIT cells. CCK stimulates acinar cell growth in the pancreas and provokes a threefold increase in the incorporation of ['Hlthymidine into DNA ( I 1 ) . Furthermore, CCK is a potent stimulator of phospholipase C , which increases the intracellular content of inositol 1,4,5-trisphosphate (Ins I ,4,5-P3) and diacylglycerol (DAG) (3). Inositol lipid hydrolysis is one of the major pathways in the control of DNA synthesis (12). I t is possible that CCK affects insulin production by control of proinsulin gene expression. Therefore, we examined in normal mouse islets whether CCK increases the enhanced content of proinsulin mRNA induced by glucose. Adult female NMRI mice (Bomholtgiird Breeding and Research Centre Ltd., Ry, Denmark), weighing 22-25 g were used. The animals had been kept on a standard pellet diet and tap water ad libitum before the experiments. Islets were isolated by the collagenase digestion technique (13,14). After rinsing, the mouse islets were hand-picked under a stereomicroscope and incubated overnight in Rosewell Park Memorial Institute (RPMI) 1640 medium. Groups of 40 islets were cultured for 48 h in RPMI 1640 medium containing a 5 or 16.7 mM glucose concentration supplemented with 1, 10, or 100 nM CCK. The medium was changed after 24 h. There were 6 1 5 groups of islets at each dose level of CCK-8 (sulfated form) or control. Thereafter, from each group of islets, total RNA was extracted by a single-step guanidinium thiocyanate extraction method (15) and kept stored at 70°C until the islet RNA was bound to nylon membrane sheets using a slot-blot apparatus. Following cross-linking by 2 min of ultraviolet-light exposure, the filters were hybridized using a digoxigenin-labeled insulin probe (R&D Systems Europe Ltd., Abingdon, UK). After hybridization and blocking, the digoxigenin-labeled insulin probe was detected using a high-affinity antidigoxigenin antibody conjugated to alkaline phosphatase (DIG Luminescent Detection Kit; Boehringer Mannheim Biochemica Scandinavia AB, Bromma, Sweden). The luminescent reaction was initiated by the addition of AMPPD substrate solution. Thereafter, luminescence was documented by a short (5 to I0-min) exposure to x-ray films. The background activity was very low. The probe is specific for insulin as specified by the manufacturer (R&D Systems Europe Ltd.). The insulin probe was then removed by 3 min of boiling in Tris (10 mM)EDTA (ImM) medium and the filters were rehybridized with a digoxigenin-labeled RNA probe specific for p-actin (Boehringer Mannheim Biochemica) and detected by chemiluminescence as for insulin. Complete removal of the insulin probe between the two hybridization reactions was checked by the chemiluminescent reaction. The hybridized RNA from the autoradiographs was measured using a computer digital image analyzing system and the amount of proinsulin mRNA was normalized to the amount of mRNA for p-actin. Means * SEM of proinsulin mRNA/p-actin mRNA are shown. Statistical comparisons were performed using the Student t test. A probability level of random difference p < 0.05 was considered significant.