Cholinergic agonists prime the beta-cell to glucose stimulation

The ability of the cholinergic agonists carbachol or acetylcholine to stimulate insulin release, activate phosphoinositide hydrolysis, and prime the beta-cell to the insulin stimulatory effect of 7.5 mM glucose was assessed. In the presence of 7 mM glucose, but not 2.75 mM glucose, 1 mM carbachol evoked a sustained insulin secretory response. At both glucose levels, carbachol stimulated phosphoinositide hydrolysis, an event monitored in myo-[2-3H]inositol-prelabeled islets by increases in [3H]inositol efflux and labeled inositol phosphate accumulation. Prior exposure to carbachol (0.1-1 mM) resulted in a dose-dependent increase in the subsequent insulin secretory response to 7.5 mM glucose. This sensitization developed within 2 min and lasted for at least 45 min after carbachol removal from the perifusion medium. Carbachol pretreatment also sensitized the islet to either 200 microM tolbutamide or 10 mM arginine. Prior exposure to 1 mM acetylcholine induced a similar proemial sensitization to a subsequent challenge with glucose. These results demonstrate that even though cholinergic stimulation increases phosphoinositide hydrolysis, this event is insufficient to initiate sustained insulin secretion from islets exposed to a low (2.75 mM) glucose concentration. However, this increase in phosphoinositide hydrolysis sensitizes islets to a subsequent challenge with one of several different stimuli, including glucose. Hence, this sensitization of islets to physiologically relevant glucose concentrations may represent the major contribution of vagal stimulation to the regulation of insulin secretion.     

A link between insulin resistance and hyperinsulinemia: inhibitors of phosphatidylinositol 3-kinase augment glucose-induced insulin secretion from islets of lean, but not obese, rats

Wortmannin (5-100 nM), a specific phosphatidyinositol 3-kinase inhibitor, augmented 8 mM glucose-induced insulin secretion from control Sprague Dawley rat islets in a dose-dependent manner. This effect persisted after its removal from the perifusion medium; however, this augmenting effect was reduced by the calcium channel inhibitor nitrendipine or by lowering the glucose level to 3 mM. Wortmannin amplified insulin release induced by the combination of 6-8 mM glucose plus 1 microM carbachol; however, it had no effect on phorbol ester- or alpha-ketoisocaproate-induced insulin secretion. The potentiating action of wortmannin on 8 mM glucose-induced release was duplicated by LY294002. Wortmannin had no effect on glucose usage rates or inositol phosphate accumulation in [3H]inositol-prelabeled islets. Of particular significance, although 50 nM wortmannin potentiated 8 mM glucose-induced secretion from islets of lean Zucker control rats, the fungal metabolite had little effect on 8 mM glucose-induced release from islets of insulin-resistant Zucker fatty rats. These findings support the concept that the same biochemical process, inhibition ofphosphatidyinositol 3-kinase, that causes peripheral tissue insulin resistance enhances beta-cell sensitivity to glucose and produces a compensatory increase in insulin secretion from these cells. The efficacy of wortmannin depends on the in vivo status of the donor's insulin signaling pathways. This elegant biochemical control mechanism in beta-cells ensures the maintenance of glucose homeostasis despite a reduction in insulin action on peripheral tissues.     

Detailed protocol for evaluation of dynamic perifusion of human islets to assess β-cell function

The definitive measure of β-cell quality in an islet is the measurement of β-cell function, i.e., the ability of the islets to release insulin in a controlled manner in response to minute changes in ambient glucose levels. Continuous flow or dynamic perifusion of the solution containing glucose and secretagogues through the islets is the most accurate assessment of regulated insulin release in vitro. Here, we describe in detail a low cost, mini-perifusion system that can be adapted to any laboratory to assess islet function by examining dynamic insulin release in response to elevated glucose concentrations and addition of secretagogues. Human islets with purity >80% and viability >90% were perifused with low glucose (1 mM) and subsequently challenged with high glucose (16.8 mM ± KCl, 25 mM). A prototypical biphasic response to elevated glucose concentrations was observed with an average 8-fold (above basal) increase in insulin concentration at peak values. Similarly, perifusion with carbachol or exendin-4 (Byetta) with glucose (6 mM) resulted in 1.32- and 1.35-fold increase in insulin secretion above basal. Islets could be maintained in the perifusion apparatus and continued to respond to glucose for up to 3 h. At minimal financial cost and technical expertise, this apparatus can be set-up in any biological laboratory to evaluate regulated hormone release from many cell types in less than 6 h. This will allow other laboratories to measure insulin responses to their drug or modifier of interest in vitro, in a manner that better approximates islet function in vivo.     

Detailed protocol for evaluation of dynamic perifusion of human islets to assess β-cell function

The definitive measure of β-cell quality in an islet is the measurement of β-cell function, i.e., the ability of the islets to release insulin in a controlled manner in response to minute changes in ambient glucose levels. Continuous flow or dynamic perifusion of the solution containing glucose and secretagogues through the islets is the most accurate assessment of regulated insulin release in vitro. Here, we describe in detail a low cost, mini-perifusion system that can be adapted to any laboratory to assess islet function by examining dynamic insulin release in response to elevated glucose concentrations and addition of secretagogues. Human islets with purity > 80% and viability > 90% were perifused with low glucose (1 mM) and subsequently challenged with high glucose (16.8 mM ± KCl, 25 mM). A prototypical biphasic response to elevated glucose concentrations was observed with an average 8-fold (above basal) increase in insulin concentration at peak values. Similarly, perifusion with carbachol or exendin-4 (Byetta) with glucose (6 mM) resulted in 1.32- and 1.35-fold increase in insulin secretion above basal. Islets could be maintained in the perifusion apparatus and continued to respond to glucose for up to 3 h. At minimal financial cost and technical expertise, this apparatus can be set-up in any biological laboratory to evaluate regulated hormone release from many cell types in less than 6 h. This will allow other laboratories to measure insulin responses to their drug or modifier of interest in vitro, in a manner that better approximates islet function in vivo.     

Effects of protein kinase C inhibitors on insulin secretory responses from rodent pancreatic islets

The contribution of protein kinase C (PKC) to the regulation of insulin release from perifused islets was explored using staurosporine or G? 6976 to inhibit the enzyme. Phorbol 12-myristate 13-acetate (PMA, 500 nM) addition to rat islets resulted in a slowly rising insulin secretory response. While minimally effective alone, the addition of 500 nM forskolin together with PMA resulted in a synergistic secretory response. The conventional protein-kinase-C isoform inhibitor G? 6976 (1 microM) completely abolished PMA-induced secretion. However, the combination of forskolin plus PMA significantly enhanced secretion from G? 6976-treated islets. Similar to previous findings made with staurosporine, G? 6976 (1 microM) enhanced the first phase and reduced the second phase of 20 mM glucose-induced secretion from rat islets. Additional studies were conducted comparing the secretory responses of perifused rat or mouse islets to glucose. Dramatic species differences to the hexose were observed. For example, 35-40 min after the onset of stimulation with 8, 10 or 20 mM glucose insulin release rates from mouse islets averaged 32+/-6, 84+/-27 or 131+/-17 pg/islet per minute, respectively. The responses from rat islets averaged 115+/-28, 561+/-112 or 800+/-46 pg/islet per minute at this time point. Islet insulin stores were comparable in both species. The addition of 5 microM carbachol, 500 nM forskolin or 20 mM KCl to mouse islets together with 20 mM glucose resulted in a dramatic augmentation of insulin output. The responses to carbachol or forskolin, but not KCl, were inhibited by 50 nM staurosporine. However, staurosporine (50 nM) reduced insulin secretion from rat islets stimulated with KCl plus 20 mM glucose. G? 6976 potentiated 20 mM glucose-induced secretion from mouse islets. These studies demonstrate that 1 microM G? 6976 completely abolishes PMA-induced release from rat islets and has a modest inhibitory effect on 20 mM glucose-induced secretion. G? 6976 (1 microM) had no inhibitory effect on 20 mM glucose-induced release from mouse islets. These studies also confirm that staurosporine inhibits both PKC- and PKA-mediated events in islets and this lack of specificity may account for its more pronounced inhibition of release when compared to G? 6976. Finally, significant species differences to PKC inhibitors exist between mouse and rat islets.     

Starvation diabetes in the rat: onset, recovery, and specificity of reduced responsiveness of pancreatic beta-cells.

A 24-h starvation markedly diminished the stimulant action of 8 mM glucose on insulin secretion from isolated perifused rat islets of Langerhans. The response to a supramaximal glucose stimulus (27.5 mM) remained normal, but prolonged fasting (48 or more) also reduced its efficacy. Refeeding of 24-h fasted animals resulted in complete restoration of glucose sensitivity within 24 h. The responses to glyceraldehyde (2 mM) and alpha-ketoisocaproate (8 mM) at concentrations which elicit approximately half-maximal stimulation were unaltered by a 24-h fast, while that to a half-maximally effective dose of mannose (15 mM) was decreased. Theophylline (5 mM) could not normalize the reduced secretory response to glucose seen in this state. The islets' ability to metabolize glucose, using various in vitro pretreatment protocols and different incubation times, was not affected by a 24-h fast. Mannose and glyceraldehyde metabolism were also unaltered. Prolonged fasting (48 h) reduced glucose metabolism by 25% at both 8 and 27.5 mM. The acute adaptive changes in islet sensitivity to moderate glucose and mannose concentrations during short term fasting (24 h) cannot be explained by an altered usage of the added hexoses.     

Glyburide and tolbutamide induce desensitization of insulin release in rat pancreatic islets by different mechanisms.

Insulin secretion was studied in rat pancreatic islets after 24-h exposure to various glyburide or tolbutamide concentrations. Glucose-induced insulin release was significantly (P < 0.05) reduced in islets cultured with 0.1 microM glyburide or 100 microM tolbutamide (2098 +/- 187, 832 +/- 93, and 989 +/- 88 pg/islet.h in control, glyburide-exposed, and tolbutamide-exposed islets, respectively). When glyburide-treated islets were stimulated with glyburide or tolbutamide, insulin release was also impaired compared to that in control islets (P < 0.05). In contrast, tolbutamide-exposed islets showed an impaired response to tolbutamide, but a normal response to glyburide. To investigate the mechanism of the sulfonylurea-induced impairment of insulin secretion, we measured insulin release and Rb+ efflux (a marker of the K+ channel activity) in a perifusion system and islet Ca2+ uptake under static conditions. Insulin release in response to 16.7 mM glucose increased in control islets from 9.4 +/- 1.1 to 131 +/- 19 pg/islet.min (first phase secretion peak). Simultaneously, the fractional 86Rb+ efflux declined from 0.015 +/- 0.002% to 0.006 +/- 0.001% (change in decrement, -63.5%). Glucose-induced insulin release in glyburide- and tolbutamide-treated islets was significantly reduced (first phase peak, 22.1 +/- 5 and 39.7 +/- 8 pg/islet.min, respectively; P < 0.05), and the fractional 86Rb+ efflux decrement was -21 +/- 6% for glyburide (P < 0.005 vs. control islets) and -65 +/- 4% (not different from control) for tolbutamide. When glyburide- or tolbutamide-exposed islets were stimulated with the corresponding sulfonylurea, insulin release was impaired compared to that in control islets (P < 0.05), but, again, 86Rb+ efflux was impaired (P < 0.05) only in glyburide-exposed islets. When 45Ca2+ uptake was studied, the increase in glucose concentration from 2.8 to 16.7 mM increased calcium uptake in control islets from 1.76 +/- 0.58 to 7.27 +/- 1.36 pmol/islet.2 min (n = 4). Preexposure to 0.1 microM glyburide did not change calcium uptake at a glucose concentration of 2.8 mM (1.44 +/- 0.45 pmol/islet.2 min) but significantly reduced calcium uptake stimulated by 16.7 mM glucose (3.21 +/- 0.35 pmol/islet.2 min; n = 4; P < 0.005 compared to control islets). In contrast, preexposure to 100 microM tolbutamide did not change either basal or glucose-stimulated calcium uptake (1.44 +/- 0.45 and 6.90 +/- 0.81 pmol/islet.2 min, respectively; n = 4). These data show that in vitro chronic exposure of pancreatic islets to the sulfonylureas glyburide and tolbutamide impairs their ability to respond to a subsequent glucose or sulfonylurea stimulation.(ABSTRACT TRUNCATED AT 400 WORDS)     

Comparative effects of amino acids and glucose on insulin secretion from isolated rat or mouse islets

Glucose and the combination of leucine and glutamine were used to stimulate insulin secretion from rat islets during a dynamic perifusion and the responses obtained were compared with those elicited from mouse islets under identical conditions. In rat islets, glucose (15 mM) or the amino acid combination of 10 mM glutamine plus 20 mM leucine were most efficacious and peak second-phase insulin release responses were 20- to 30-fold above prestimulatory rates. In contrast to rat islet responses, sustained second-phase insulin secretory responses to the same agonists were minimally increased 1- to 2-fold from mouse islets. Parallel studies demonstrated that phospholipase C (PLC) was markedly activated in rat, but not mouse, islets by both high glucose concentrations and the amino acid combination. Additional studies documented that glucose and amino acid responses of both rat and mouse islets were amplified by carbachol or forskolin. However, wortmannin, a phosphatidylinositol 3-kinase inhibitor, amplified only the responses to glucose leaving the responses to the amino acid mixture unaltered. These observations support the concept that mitochondrial metabolism alone is minimally effective in stimulating insulin secretion from islets. The activation of the supplementary second messenger systems (PLC and/or cAMP) appears essential for the emergence of their full secretory potential. The mechanism regulating the potency and specificity of wortmannin's impact on glucose-induced secretion remains to be identified; however a unique mechanism is supported by these findings.     

Characterization of the insulinotropic potency of polyunsaturated fatty acids.

In this study we have assessed the individual abilities of the essential fatty acids, linoleic and linolenic acids, to release insulin and compared their insulinotropic potencies with those of the more established nutrient insulin secretagogues, glucose and arginine. In each experiment, a total of six islets microdissected from three mice were preperifused at the rate of 1 ml/min with Krebs-Ringer bicarbonate buffer, pH 7.4, containing 2% bovine albumin and 5.5 mM glucose (basal) with a continuous supply of 95% O2-5% CO2 at 37 C for 1 h. After collecting basal samples, the effects of 27.7 mM glucose, 20 mM arginine, 10 mM linoleic acid (18:2, omega 6), and 5 mM linolenic acid (18:3, omega 3) were tested using a sandwich protocol that entails 20-min alternating periods of stimulation with a secretagogue and a washout with basal perifusion. These nutrient concentrations were selected from initial experiments performed to characterize their dose-response effects on insulin secretion. Effluent samples were collected throughout each experiment for measurement of insulin by RIA. In one series of experiments, islets were challenged three times with 27.7 mM glucose, 10 mM linoleic acid, and 5 mM linolenic acid. In another set of experiments, islets were perifused with 20 mM arginine, 27.7 mM glucose, and 10 mM linoleic acid. All of these nutrients stimulated insulin release in a dose-dependent manner. In comparing the insulinotropic potencies of these secretagogues, we assessed insulin secretion as the integrated areas under the curve during 20 min of perifusion with a given nutrient. Thus, the mean integrated area under the curve per 20 min above basal in the presence of 27.7 mM glucose was 6,516 +/- 1,435 pg, which was not significantly different from the value of 4,772 +/- 866 pg obtained during arginine perifusion. However, the area under the curve during 20 min above basal obtained in the presence of linoleate and linolenic acid (8,712 +/- 1,949 and 10,506 +/- 1,490 pg, respectively) were significantly different (P less than 0.05) from those calculated during arginine and glucose perifusions. There was no statistically significant difference between the effects of these two fatty acids at the concentrations tested. In conclusion, our data suggest that linoleic acid and linolenic acid may be, at least in this murine islet preparation, as effective in stimulating insulin release as glucose and arginine, hitherto used to assess the abilities of nutrients to stimulate insulin secretion. However, it remains to be seen whether the efficacy of these polyunsaturated fatty acids in insulin release by murine islets will be obtained in experiments performed on human islets.     

Forskolin-induced desensitization of pancreatic beta-cell insulin secretory responsiveness: possible involvement of impaired information flow in the inositol-lipid cycle.

Isolated rat islets of Langerhans were incubated for 2 h in a myo-[2-3H]inositol-containing solution to label their phosphoinositides. Also included during this labeling period was forskolin (0.1-5 microM), a compound established to elevate islet cAMP levels. These islets were subsequently perifused, and their insulin secretory responses to 20 mM glucose or 1 microM of the phorbol ester 12-O-tetradecanoyl phorbol-13-acetate (TPA) were assessed. Determined in parallel with secretion were [3H] inositol efflux patterns and, at the termination of the perifusion, labeled inositol phosphate accumulation. The following major observations were made. 1) Forskolin had no deleterious effect on the total amount of [3H]inositol incorporated by the islets during the labeling period. 2) However, labeling in forskolin resulted in subsequent dose-dependent decreases in 20 mM glucose-induced insulin secretion, [3H]inositol efflux and inositol phosphate accumulation. 3) Inclusion of the diacylglycerol (DAG) kinase inhibitor monooleoylglycerol (50 microM) restored to a significant degree glucose-induced release from forskolin-desensitized islets. 4) Pretreatment with 5 microM forskolin had no deleterious effect on TPA-induced insulin release. 5) Prior exposure to forskolin also impaired phosphoinositide hydrolysis in response to cholecystokinin stimulation. 6) Similar to forskolin, labeling in isobutylmethylxanthine (1 mM) reduced in a parallel fashion islet [3H]inositol efflux and insulin secretion in response to 20 mM glucose stimulation. These findings demonstrate that prior chronic elevation of islet cAMP levels suppresses the activation of phospholipase-C in response to subsequent stimulation. Defective insulin secretory responsiveness of these islets appears to be the result of impaired generation of phosphoinositide-derived second messenger molecules, particularly DAG. By substituting for DAG, however, TPA circumvents this biochemical lesion and evokes a normal insulin secretory response from forskolin-pretreated islets.     

The possible mechanisms by which phanoside stimulates insulin secretion from rat islets

We recently showed that phanoside, a gypenoside isolated from the plant Gynostemma pentaphyllum, stimulates insulin secretion from rat pancreatic islets. To study the mechanisms by which phanoside stimulates insulin secretion. Isolated pancreatic islets of normal Wistar (W) rats and spontaneously diabetic Goto-Kakizaki (GK) rats were batch incubated or perifused. At both 3 x 3 and 16 x 7 mM glucose, phanoside stimulated insulin secretion several fold in both W and diabetic GK rat islets. In perifusion of W islets, phanoside (75 and 150 microM) dose dependently increased insulin secretion that returned to basal levels when phanoside was omitted. When W rat islets were incubated at 3 x 3 mM glucose with 150 muM phanoside and 0 x 25 mM diazoxide to keep K-ATP channels open, insulin secretion was similar to that in islets incubated in 150 microM phanoside alone. At 16 x 7 mM glucose, phanoside-stimulated insulin secretion was reduced in the presence of 0 x 25 mM diazoxide (P<0 x 01). In W islets depolarized by 50 mM KCl and with diazoxide, phanoside stimulated insulin release twofold at 3 x 3 mM glucose but did not further increase the release at 16 x 7 mM glucose. When using nimodipine to block L-type Ca2+ channels in B-cells, phanoside-induced insulin secretion was unaffected at 3 x 3 mM glucose but decreased at 16 x 7 mM glucose (P<0 x 01). Pretreatment of islets with pertussis toxin to inhibit exocytotic Ge-protein did not affect insulin response to 150 microM phanoside. Phanoside stimulated insulin secretion from Wand GK rat islets. This effect seems to be exerted distal to K-ATP channels and L-type Ca2+ channels, which is on the exocytotic machinery of the B-cells.     

Influence of staurosporine, nitrendipine and monooleoylglycerol on interleukin-1-induced insulin secretion and phosphoinositide hydrolysis.

The monokine interleukin-1 alpha (IL-1) induces a glucose-dependent increase in insulin secretion, an effect tentatively attributed to its ability to increase beta cell phosphoinositide (PI) hydrolysis. In the present experiments, the effects of the protein kinase C inhibitor staurosporine (20 nM), the calcium channel antagonist nitrendipine (5 microM), and the diacylglycerol kinase inhibitor monooleoylglycerol (MOG, 25 microM) on 40 nM IL-1-induced increments in insulin release from perifused islets and inositol phosphate levels in [3H]inositol prelabeled islets were assessed. In perifused islets, insulin secretion in response to IL-1 in the presence of 7 mM glucose averaged 313 +/- 43 pg/islet/min 35-40 min after the onset of stimulation. Release from control islets perifused in the presence of 7 mM glucose alone averaged 56 +/- 6 pg/islet/min at this time point. The addition of staurosporine together with IL-1 reduced insulin secretion at this time point to 88 +/- 21 pg/islet/min. This level of IL-1 caused significant increases in inositol phosphate accumulation in the presence of 7 mM glucose but not 2.75 mM glucose. Staurosporine was without a significant effect on inositol phosphate accumulation in response to the monokine. In contrast, nitrendipine (5 microM) inhibited insulin release and inositol phosphate accumulation in a parallel fashion. Finally, MOG significantly amplified release to the monokine without significantly affecting its impact on inositol phosphate accumulation. Nitrendipine or staurosporine blocked this amplifying effect of MOG on secretion. These results emphasize the role of PI hydrolysis in IL-1-induced insulin secretion and suggest further that calcium influx is essential for IL-1 to fully activate both PI hydrolysis and insulin secretion.     

Effects of short-term culturing on islet phosphoinositide and insulin secretory responses to glucose and carbachol

The ability of glucose and carbachol, alone or in combination, to stimulate islet cell phosphoinositide (PI) hydrolysis and insulin secretory responses in freshly isolated or in 20–24 h cultured rat islets was assessed. In freshly isolated,³H-inositol-prelabeled islets, 20 mM glucose alone or 1 mM carbachol alone stimulated significant increments in³H-inositol efflux and inositol phosphate (IP) accumulation. When stimulated with both agonists, a dramatic and synergistic effect on IP accumulation was noted. Carbachol (1 mM) alone had no sustained stimulatory effect on insulin secretion. Glucose (20 mM) alone induced a biphasic insulin secretory response. When compared to prestimulatory secretory rates of 18±4 pg/islet/min, peak first and second phase responses of freshly isolated islets to 20 mM glucose averaged 126±24 and 520±82 pg/islet/min, respectively. In the presence of both glucose (20 mM) and carbachol (1 mM), peak first and second phase responses now averaged 422±61 and 1016±156 pg/islet/min, respectively. In contrast to freshly studied islets, culturing islets for 20–24 h in CMRL-1066 medium attenuated all measured responses. The increases in³H-inositol efflux rates in response to glucose, carbachol, or their combination were significantly less than those observed with fresh islets. The IP responses were also attenuated. Second phase insulin secretory responses to 20 mM glucose alone (68±9 pg/islet/min) or the combination of 20 mM glucose plus 1 mM carbachol (358±85 pg/islet/min) were also significantly decreased when compared with fresh islets. We conclude from these studies that the process of culturing islets for one day in CMRL-1066 significantly decreases islet cell PI hydrolysis and insulin secretory responsiveness. These observations may help to explain the discordant conclusions reached concerning the involvement of PI hydrolysis and protein kinase C activation in the regulation of insulin release from freshly isolated versus cultured islets.     

Induction of beta-cell rest by a Kir6.2/SUR1-selective K(ATP)-channel opener preserves beta-cell insulin stores and insulin secretion in human islets cultured at high (11 mM) glucose

In health, most insulin is secreted in pulses. Type 2 diabetes mellitus (TTDM) is characterized by impaired pulsatile insulin secretion with a defect in insulin pulse mass. It has been suggested that this defect is partly due to chronic overstimulation of beta-cells imposed by insulin resistance and hyperglycemia, which results in depletion of pancreatic insulin stores. It has been reported that in TTDM overnight inhibition of insulin secretion (induction of beta-cell rest) leads to quantitative normalization of pulsatile insulin secretion upon subsequent stimulation. Recently, decreased orderliness of insulin secretion has been recognized as another attribute of impaired insulin secretion in TTDM. In the current studies we sought to address at the level of the isolated islet whether chronic elevated glucose concentrations induce both defects involved in impaired insulin secretion in TTDM: deficiency and decreased orderliness of insulin secretion. We use the concept of beta-cell rest, induced by a novel beta-cell selective K(ATP)-channel opener (KCO), NN414 (6-chloro-3-(1-methylcyclopropyl)amino-4H-thieno[3,2-e]-1,2,4-thiadiazine 1,1-dioxide), to test whether preservation of insulin stores leads to normalization of both processes in response to glucose stimulation. Human islets were isolated from three cadaveric organ donors and studied in perifusion experiments and static incubation. Acute activation of K(ATP)-channels suppressed insulin secretion from perifused human islets by approximately 90% (P < 0.0001). This KCO also inhibited glucagon secretion in a dose-dependent manner (P = 0.01). Static incubation at 11 and 16 vs. 4 mM glucose for 96 h decreased islet insulin stores by approximately 80% and 85% (P < 0.0001, respectively). In subsequent perifusion experiments, total insulin secretion ( approximately 30%; P < 0.01) from these islets and insulin pulse mass ( approximately 40%; P < 0.05) were both decreased (11 vs. 4 mM). The inhibition of insulin secretion during static incubation with KCO reduced the loss of islet insulin stores in a dose-dependent manner (P < 0.0001) and resulted in increased total insulin secretion (2.6-fold; P < 0.01) and insulin pulse mass (2.5-fold; P < 0.05) during subsequent perifusion. The orderliness of insulin secretion was significantly reduced after chronic incubation of human islets at 11 mM glucose (P = 0.04), but induction of beta-cell rest at 11 mM failed to normalize the regularity of insulin secretion during subsequent perifusion. We conclude that physiological increased glucose concentrations (11 mM), which are frequently observed in diabetes, lead to a loss of islet insulin stores and defective pulsatile insulin secretion as well as reduced orderliness of insulin secretion. Induction of beta-cell rest by selective activation of beta-cell K(ATP)-channels preserves insulin stores and pulsatile insulin secretion without restoring the orderliness of insulin secretion. Therefore, the concept of beta-cell rest may provide a strategy to protect beta-cells from chronic overstimulation and to improve islet function. Impaired glucose-regulated insulin secretion in TTDM may, however, partially involve mechanisms that are distinct from insulin stores and insulin secretion rates.     

Effects of stimulation of adenylate cyclase and protein kinase-C on cultured fetal rat B-cells

To study the maturation of fetal pancreatic B-cells, cell suspensions of pancreas from 21.5-day-old fetuses were cultured in RPMI medium containing 10 mM glucose. Forskolin (1 microM), used to stimulate adenylate cyclase, moderately delayed the neoformation of islets, slightly accelerated the proliferation of endocrine cells, and considerably increased insulin release by the cultures. The latter increase was not completely compensated for by the stimulation of insulin biosynthesis, so that the islet insulin content was decreased. The phorbol ester 12-O-tetradecanoyl-phorbol-13-acetate (TPA; 25 nM), used to stimulate protein kinase-C, had little effect on the evolution of the cultures, but increased insulin release. This increase was almost compensated for by the stimulation of insulin biosynthesis. After 9-10 days of culture, insulin release in response to 15 mM glucose or 10 mM leucine was studied with perifused islets. In control islets, glucose produced a sustained increase in insulin release, which, however, was 6-fold smaller than that produced by leucine. Addition of forskolin or TPA to the perifusion medium markedly amplified the response to glucose without causing a biphasic pattern of release. In islets cultured with forskolin, the insulin response to glucose or leucine was decreased, largely owing to the lower insulin stores. In islets cultured with TPA, the insulin response to glucose or leucine was also decreased, but these differences cannot be explained simply by changes in insulin content. Neither treatment affected the kinetics of release. In conclusion, acute stimulation of adenylate cyclase or protein kinase-C markedly increased insulin release from fetal islets without causing an adult-like biphasic pattern of secretion. Chronic stimulation did not accelerate maturation of B-cells.     

Structure-function relationships in pancreatic islets: support for intraislet modulation of insulin secretion

Pancreatic islet B cell function was studied in vitro using three structurally different preparations of islet tissues: isolated, intact islets, dispersed islet cells attached singly to microcarrier beads, and reaggregated islet cells. Mechanisms of intercellular communication are eliminated with single cell preparations, whereas in aggregates cell to cell communications are reestablished and a defined microenvironment restored. Perifusion studies measured nonstimulated and glucose- and arginine-stimulated insulin release from the three islet tissues. Insulin secretion rates were expressed as a function of cellular DNA content, permitting direct comparison between tissues. During perifusion with low (2.8 or 5.5 mM) glucose concentrations, secretion rates of single islet cells were up to 6-fold greater (P less than 0.001) than those of intact islets. Perifusion of islet cells with 2.8 mM glucose and 100 or 500 pg glucagon/ml had no effect whereas GH-release-inhibiting factor (330 and 1000 pg/ml) decreased nonstimulated insulin secretion rates by 15% (P less than 0.05). After reaggregation, basal insulin secretion rates were restored toward those of intact islets. Glucose (5.5-30 mM) and L-arginine (5-20 mM) elicited first phase insulin responses from single islet cells that were not significantly different from those observed with intact islets; in contrast, second phase responses of single islets to glucose were approximately 50% those seen with intact islets, and their second phase responses to arginine were absent. Single islet cell first and second phase insulin responses to 5.5 mM glucose were enhanced 2.2-fold (P less than 0.01) and 2.8-fold (P less than 0.05), respectively, in the presence of exogenous glucagon, resulting in secretory profiles characteristic of intact islets. Reaggregation of single islet cells was associated with markedly increased first and second phase insulin responses to both glucose and arginine stimulation. These data show that disruption of the islet microanatomy results in alteration of insulin secretory responses and that these effects can be reversed, in part by exogenous glucagon and GH-release-inhibiting factor, and by reaggregation. Although different mechanisms appear important for nonstimulated, first and second phase insulin release, the findings support a role for both direct intercellular communication and hormonal secretion by islet A and D cells in the modulation of B cell function.     

Glucose stimulates pulsatile insulin secretion from human pancreatic islets by increasing secretory burst mass: dose-response relationships

Insulin is secreted almost exclusively in discrete bursts, and physiological regulation is accomplished by modulation of the pulse mass. How the integrity of contiguous anatomic structures in the human pancreas (islets, splanchnic innervation, exocrine tissue, local hormones) directs the coordinated insulin secretion is not known. We posed the hypothesis that glucose stimulates insulin secretion from isolated human islets by an amplification of insulin pulse mass with no change in pulse frequency and that the glucose dose-response curve for the regulation of insulin pulse mass mirrors that recognized in vivo. Islets from five nondiabetic cadaveric donors were perifused in a recently validated perifusion system at 4 mM and subsequently at 8, 12, 16, or 24 mM glucose. The effluent was collected in 1-min intervals and used for the measurement of insulin (ELISA). Pulsatile insulin secretion was analyzed by deconvolution analysis. Total insulin secretion increased progressively (P < 0.0001). This augmentation was due to amplified pulse mass (3-fold, 24 mM vs. 4 mM glucose; P < 0.0001) with no change in pulse interval (approximately 4 min). Pulsatile insulin secretion was stimulated most effectively in a physiologic concentration range of 4-8 mM. The islet insulin content was significantly correlated to the magnitude of first and second phase insulin secretion (P < 0.0001). The quantifiable orderliness of pulsatile insulin secretion rose with escalating glucose concentration (P = 0.02). In conclusion, glucose stimulates pulsatile insulin secretion from isolated human islets by amplification of insulin pulse mass without altering pulse interval. The in vitro concentration-response relationship is comparable with that observed in vivo. These data imply that transplanted human islets should be able to reproduce glucose-regulated insulin secretion as observed in the intact human pancreas.     

Biochemical mechanisms involved in monomethyl succinate-induced insulin secretion.

Esters of succinic acid stimulate insulin secretion from pancreatic beta-cells. Using collagenase-isolated rat islets, the transduction mechanisms involved were investigated. In freshly isolated perifused islets, monomethyl succinate (MMSucc), in the presence of basal (2.75 mM) glucose, stimulated insulin release in a biphasic pattern. This secretory response was dependent on extracellular calcium movement into the beta-cell, since the calcium channel blocker nitrendipine (5 microM) abolished it. The glucokinase inhibitor mannoheptulose (20 mM) had no effect on its secretory action, while the protein kinase-C inhibitor staurosporine (20 nM) reduced secretion to MMSucc. In addition, while ineffective alone, the diacylglycerol kinase inhibitor monooleoylglycerol (25 microM) potentiated MMSucc-induced insulin release. A similarly amplified response occurred in the presence of forskolin (0.25 microM), a compound that elevates islet cAMP levels. The sodium salt of succinic acid (20 mM) had no effect on insulin release in the presence or absence of forskolin. Prior treatment with MMSucc in the presence of 2.75 mM glucose sensitized islets to the usually weak insulin secretory effect of 7.5 mM glucose. Other groups of islets were incubated for 2 h with myo-[2-3H]inositol to label their phosphoinositide pools. These islets were subsequently stimulated, and the kinetics of [3H]inositol efflux and insulin secretion were measured. MMSucc induced a rapid and sustained dose-dependent increase in [3H]inositol efflux rates. In batch-incubated islets, MMSucc increased inositol phosphate levels. Finally, MMSucc (20 mM), in the presence of 8 mM glucose, did not influence the detritiation of [5-3H]glucose, but reduced the oxidation of [U-14C] glucose. These results support the following conclusions. First, MMSucc is a potent activator of islet phosphoinositide hydrolysis. Second, the activation of protein kinase-C appears to contribute to the acute insulin secretory effect of MMSucc. Third, MMSucc-induced increases in phosphoinositide hydrolysis contribute at least in part to its ability to acutely stimulate insulin release and prime the beta-cell to subsequent stimulation. Finally, mitochondrial events associated with the oxidative metabolism of MMSucc may underlie its insulinotropic action.     

Diazoxide attenuates glucose-induced defects in first-phase insulin release and pulsatile insulin secretion in human islets

Humans with type-2 diabetes mellitus (TTDM) have hyperglycemia ( approximately 11 mM) and impaired glucose-mediated insulin secretion characterized by impaired first-phase insulin release (FPIR) and pulsatile insulin release. Culture of islets from nondiabetic humans in very high glucose concentrations ( approximately 20-30 mM) for 96 h causes impaired FPIR. We sought to determine 1). whether human islets cultured at a glucose concentration of approximately 11 mM (comparable to TTDM) recapitulates impaired insulin secretion in TTDM, specifically impaired FPIR and insulin pulse mass with an increased proinsulin/insulin (PI/I) secretion ratio; and 2). whether these changes can be attenuated by addition of diazoxide to islets cultured with 11 mM glucose. Islets cultured with 11 mM glucose for 96 h had 75% depleted insulin stores (P < 0.05), decreased FPIR and insulin pulse mass (P < 0.05), and an approximately 3-fold increase in the ratio of PI/I islet content and in secretion ratio (P < 0.05). Addition of diazoxide to islets cultured with 11 mM glucose decreased insulin secretion during static incubation, leading to relative preservation of insulin stores and enhanced insulin secretion during subsequent perifusion; FPIR increased by 162% (P < 0.05) and insulin pulse mass by 150% (P < 0.05) vs. no diazoxide. The mean islet PI/I content and islet PI/I secretion ratio were also decreased by approximately 70% (P < 0.05) by prior addition of diazoxide to islets during culture with 11 mM glucose. FPIR and insulin pulse mass were related to islet insulin stores (P < 0.001 for FPIR and P < 0.001 for pulse amplitude). In conclusion, the pattern of defects of insulin secretion present in TTDM (impaired FPIR and pulsatile insulin secretion, increased PI/I ratio) can be recapitulated in human islets cultured with 11 mM glucose for 96 h. These defects can be at least partially offset by concurrent inhibition of insulin secretion by diazoxide, which also preserves insulin stores. Defective insulin secretion in TTDM may be, at least in part, due to depletion of available insulin stores secondary to chronic increased demand (insulin resistance and hyperglycemia) in the setting of a decreased beta-cell mass.     

In vivo administration of interleukin-1 inhibits glucose-stimulated insulin release

Recombinant interleukin-1 beta (IL-1 beta) was administered intraperitoneally for 3 days to normal C57BL/6ByJ (B6) mice. The islets from IL-1-treated and control animals were isolated and glucose-stimulated insulin secretion studied in the perifusion system. The total islet insulin content and the ultrastructure of the islets isolated from the animals treated with IL-1 did not differ from those seen in control animals. However, glucose-stimulated insulin release was significantly impaired after 3 days of in vivo administration of IL-1, either 3 micrograms/animal/day or 0.3 micrograms/animal/day. The administration of IL-1 inhibited an acute phase of glucose-induced insulin release, whereas neither basal insulin secretion nor insulin release from 10-30 min of perifusion with glucose was impaired. There was an only partial (27%) and non-significant restoration of the insulin secretory response to glucose stimulation 4 days after discontinuation of IL-1 treatment. We conclude that IL-1 administered in vivo is capable of adversely affecting pancreatic islet response to glucose stimulation. After 3 days of administration, these changes are confined to the process of insulin release, with the islet cell morphology and total insulin content being unaffected.