Search

Morphological analysis of hormone content and functional assessment of hormone secretion were conducted in beta TC-6 cells, an insulin-secreting cell line derived from transgenic mice expressing the large T-antigen of simian virus 40 (SV40) in pancreatic beta-cells. We observed by immunohistochemistry and confocal microscopy that beta TC-6 cells contain abundant insulin and small amounts of glucagon and somatostatin (SRIF). Glucagon usually co-localized with insulin, whereas cells containing SRIF did not contain insulin or glucagon. Static incubation and perifusion experiments demonstrated that beta TC-6 cells at passage 30-45 secrete insulin in response to glucose. In static incubations, maximal stimulation was achieved for glucose concentrations > 2.8 mmol/l glucose, and the half-maximal effect was observed at 0.5 mmol/l. Maximal stimulation was four times greater than HIT-T15 cells at passage 72-81, although HIT cells had a greater response over their basal levels. The magnitude of the insulin response to glucose in perifusion was 1,734 +/- 384 pmol.l-1. min and was 4.6-fold greater in the presence of 3-isobutyl-1-methylxanthine. Low amounts of glucagon were released in response to amino acids. Epinephrine (EPI), and to a lesser extent SRIF, inhibited phasic glucose-induced insulin secretion. A major portion of these inhibitory effects was mediated by pertussis toxin-sensitive substrates. Immunoblots detected the presence of the G-proteins Gi alpha 2, Gi alpha 3, and Go alpha 2. These results indicate that beta TC-6 cells are a glucose-responsive cell line in which insulin exocytosis is physiologically regulated by EPI and SRIF through Gi/Go-mediated mechanisms.

Previously, we have demonstrated that somatostatin mediates all of its inhibitory effects on glucose-induced insulin secretion from the HIT-T15 cell through pertussis toxin-sensitive G-proteins and that the membrane fraction of this clonal line of pancreatic beta-cells contains six such proteins: G(i) alpha 1, G(i) alpha 2, G(i) alpha 3, and three forms of G(o) alpha. To determine the specificity of somatostatin receptor-G-protein coupling in HIT-T15 cells, we examined the ability of antisera specific for the COOH-terminus of G alpha subtypes to inhibit somatostatin-induced augmentation of membrane GTPase activity. GTPase activity increased in membranes as a function of GTP. At all concentrations of GTP studied, 1 mumol/l somatostatin stimulated GTPase activity. Pertussis-toxin pretreatment prevented the effects of somatostatin. Antisera selective for G(o) alpha subtypes reduced the effects of somatostatin on GTPase activity (GTPase activity in absence of antisera, 125 +/- 3% of control; in the presence of antisera 976, 110 +/- 2% of control; n = 13, P < 0.001), whereas antisera directed against G(i) alpha 1, G(i) alpha 2, G(i) alpha 3, and Gs alpha were without effect. Somatostatin also significantly prevented cyclic AMP accumulation during perifusion with 11.1 mmol/l glucose through a pertussis toxin-sensitive mechanism. These data indicate that the somatostatin receptor couples to G(o) alpha in the HIT-T15 cell and suggest that G(o) alpha may link somatostatin to cyclic AMP metabolism in pancreatic beta-cells.

G-proteins are important mediators of hormonal inhibition of insulin secretion. To characterize the pertussis toxin-sensitive substrates present in HIT cell membranes, we performed immunoblots with specific antisera and found evidence for the presence of Gi alpha 1, Gi alpha 2, Gi alpha 3, and three forms of Go alpha. We observed that pertussis toxin-sensitive substrates mediate all of the effects of SRIF, and a major portion of the effects of EPI, on insulin secretion from rat islets during static incubations. These results agree with our previously reported studies examining phasic glucose-induced insulin secretion from HIT cells. To ascertain whether inhibition of adenylate cyclase, presumably involving coupling of the catalytic subunit to Gi, may be a common mechanism for both hormones, we studied the effects of 8-bromo-cyclic AMP and found that this agent partially prevented the inhibitory effects of both hormones. We also observed that the inhibitory effects of SRIF and EPI on insulin were nonadditive, that both hormones were additive to nickel chloride during inhibition of insulin release, and that they noncompetitively inhibited glipizide-induced insulin secretion through pertussis toxin-sensitive mechanisms. Together, these results suggest that both hormones exert their effects on insulin secretion at multiple G-protein-regulated sites including adenylate cyclase and sites distal to the glipizide-binding site on the KATP channel.

Activated human mononuclear cells (MCs) were coincubated for 8 h with HIT cells, a clonal cell line of pancreatic islet beta-cells. Measurements of HIT cell viability and insulin secretion were determined to 1) ascertain whether activated MCs can alter beta-cell viability in the absence of exogenously provided cytokines, 2) examine this response over a range of MC-HIT cell ratios, and 3) identify mechanisms responsible for altered insulin release consequent to MC-induced HIT cell damage. HIT cell viability was markedly decreased by activated MCs during an 8-h coincubation. HIT cell lysis could be attributed to activated natural killer cells, and lysis did not occur in the presence of activated T-lymphocyte clones. Activated MCs caused a marked early increase in insulin release from HIT cells (increase at 2 h: 7.75 +/- 0.16 nM for activated MCs, 2.66 +/- 0.09 nM for control; P less than 0.001). Insulin levels by the 8th h of the coincubation were significantly lower than the 2-h peak (4.33 +/- 0.13 vs. 7.75 +/- 0.16 nM, P less than 0.001). These changes in insulin were dependent on the ratio of activated MCs to HIT cells with the effects clearly evident at an activated MC-HIT cell ratio of greater than or equal to 10:1. Pretreatment of activated MCs and HIT cells with prostaglandin-synthesis inhibitors did not prevent the cytotoxic effects of activated MCs on HIT cells. Somatostatin did not inhibit the early exaggerated insulin release, suggesting that these increased insulin levels represented leakage of insulin from damaged HIT cells rather than functional insulin secretion.(ABSTRACT TRUNCATED AT 250 WORDS)

Guanine nucleotide-binding proteins (G proteins) are critically important mediators of many signal-transduction systems. Several important sites regulating stimulus-secretion coupling and release of insulin from pancreatic beta-cells are modulated by G proteins. Gs mediates increases in intracellular cAMP associated with hormone-induced stimulation of insulin secretin. Gi mediates decreases in intracellular cAMP caused by inhibitors of insulin secretion, e.g., epinephrine, somatostatin, prostaglandin E2, and galanin. G proteins also regulate ion channels, phospholipases, and distal sites in exocytosis. Cholera and pertussis toxins irreversibly ADP ribosylate G proteins and are important tools that can be used both to manipulate G-protein-dependent modulators of insulin secretion and detect and quantify G proteins by electrophoretic techniques. The stage is set to pursue these initial observations in greater depth and ascertain whether G-protein research will provide important new insights into normal and abnormal regulation of insulin secretion.

Prostaglandin E2 (PGE2) inhibits glucose-induced insulin secretion, and inhibitors of PGE2 synthesis augment this event. However, there has been confusion regarding prostaglandin regulation of insulin secretion, partly because no mechanism has been demonstrated for the inhibitory action of PGE2 on beta-cell function. These studies were performed with a clonal cell line of glucose-responsive beta-cells (HIT cells) to determine whether PGE2 effects on insulin secretion are receptor mediated and, if so, whether the postreceptor effects are mediated by inhibitory regulatory components (Ni) of adenylate cyclase. Saturable [3H]PGE2 binding to HIT cells was demonstrated. This binding was dissociable and specific for prostaglandins of the E series. Scatchard analyses of binding data indicated a single class of sites with a Kd of approximately 1 X 10(-9) M. Guinea pig islets were also demonstrated to have a single class of binding sites with a similar Kd but only 22% as many binding sites (0.060 vs. 0.013 pmol/mg protein, HIT cells vs. guinea pig islet). HIT cells were demonstrated to synthesize PGE2, and this synthesis was inhibitable by acetylsalicylic acid. Accumulation of cAMP by HIT cells was inhibited in a concentration-dependent manner by PGE2 with an IC50 of approximately 1 X 10(-9) M. Insulin secretion by HIT cells during static incubations with 11.1 mM glucose was also inhibited by PGE2 in a concentration-dependent manner with an IC50 of 1 X 10(-9) M. PGE2 was more potent than epinephrine but less potent than somatostatin in this regard. Maximum inhibition of glucose-induced insulin secretion was 26, 37, and 29% of control values for somatostatin, PGE2, and epinephrine, respectively.(ABSTRACT TRUNCATED AT 250 WORDS)

Although prostaglandin E2 (PGE2) is known to inhibit glucose-induced insulin secretion, it is uncertain whether PGE2 actions on the beta-cell are direct, whether they are equipotent for both phases of hormone secretion, and whether the same mechanism of action prevails throughout. Study of the HIT cell, a clonal line of pancreatic beta-cells, provides answers to these questions because perifusion with glucose and 3-isobutyl-1-methylxanthine stimulates biphasic insulin secretion. Perifusion with PGE2 decreased both the first and second phases of glucose-induced insulin release to 47 +/- 4% of controls. Pretreatment with pertussis toxin partly prevented PGE2 inhibition to 80 +/- 4% of controls for first phase and 79 +/- 4% of controls for second phase. To evaluate whether the partial prevention of PGE2 inhibition seen with pertussis toxin pretreatment was caused by Gi heterotrimer association between the preincubation period and the end of perifusion, PGE2 actions were also examined during continuous treatment with pertussis toxin. Under these conditions, PGE2 inhibition of both phases was totally prevented. However, no difference was observed in membrane protein ADP ribosylation when cells were examined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis after pretreatment or continuous treatment with pertussis toxin. Cyclic AMP (cAMP) accumulation was inhibited by PGE2 (from 3263 +/- 153 to 1549 +/- 158 fmol/10(6) cells) but less so after pretreatment with pertussis toxin (correlation between insulin release and cAMP accumulation during perifusion; n = 18, r = .85, P less than .001). Thus, PGE2 equally inhibits both phases of glucose-induced insulin secretion and cAMP generation through a pertussis toxin-sensitive G protein-mediated direct effect on the pancreatic beta-cell.

The HIT cell is a variably glucose-responsive clonal line of pancreatic islet beta-cells. To ascertain whether insulin responsiveness to glucose, arginine, isoproterenol, forskolin, and K+ varied in a predictable fashion, full concentration-response curves with these agonists were examined with cells from a span of 25 passages. Basal and stimulated cAMP metabolism were also examined. The findings indicate that insulin responses to glucose diminish progressively with increasing passage number and that studies of glucose-induced insulin secretion should be limited to passages 81 and earlier. This defect in insulin secretion is a general rather than a glucose-specific phenomenon in that insulin responses to the other nonglucose secretagogues also diminished with increasing passage number. All changes in glucose-stimulated responses were limited to diminutions in maximal responses; no alterations in apparent half-maximal effective concentrations (EC50s) were observed. In contrast to the continually diminishing insulin responsiveness observed, hormone inhibition by somatostatin of insulin secretion and basal cAMP metabolism remained intact throughout the passages examined. Interestingly, a reciprocal relationship between insulin responsiveness and cAMP responsiveness was observed. Dramatic cAMP responses to isoproterenol and forskolin were observed with the later passages. We conclude that loss of insulin responsivity in HIT cells is a passage-dependent process that is serial rather than sporadic and global rather than glucose specific. Dramatic reciprocal changes in cAMP metabolism occur in the later passages.