Effectiveness of Nateglinide on In Vitro Insulin Secretion from Rat Pancreatic Islets Desensitized to Sulfonylureas

Chronic exposure of pancreatic islets to sulfonylureas (SUs) is known to impair the ability of islets to respond to subsequent acute stimulation by SUs or glucose. Nateglinide (NAT) is a novel insulinotropic agent with a primarily site of action at β-cell KATP channels, which is common to the structurally diverse drugs like repaglinide (REP) and the SUs. Earlier studies on the kinetics, glucosedependence and sensitivity to metabolic inhibitors of the interaction between NAT and KATP channels suggested a distinct signaling pathways with NAT compared to REP, glyburide (GLY) or glimepiride (GLI). To obtain further evidence for this concept, the present study compared the insulin secretion in vitro from rat islets stimulated acutely by NAT, GLY, GLI or REP at equipotent concentrations during 1-hr static incubation following overnight treatment with GLY or tolbutamide (TOL). The islets fully retained the responsiveness to NAT stimulation after prolonged pretreatment with both SUs, while their acute response to REP, GLY, and GLI was markedly attenuated, confirming the desensitization of islets. The insulinotropic efficacy of NAT in islets desensitized to SUs may result from a distinct receptor/effector mechanism, which contributes to the unique pharmacological profile of NAT.


INTRODUCTION
Sulfonylureas (SUs) are widely used in the treatment of non-insulin-dependent (Type 2) diabetes. They stimulate insulin secretion by closing ]-cell plasma membrane KAT p channels, [1] which are formed by the molecular interaction between a high-affinity SU receptor, SUR1, and an inwardly rectifying K + channel subunit, Kir6.2. [2] The closure of KATP channels leads to the opening of voltage-dependent Ca 2 + channels, Ca 2 + influx and stimulation of Ca 2 +dependent exocytosis. [3] While SUs exert hypoglycemic action via a direct stimulation of insulin release during short term administration in type 2 diabetic patients, their activity declines during long term applica-tion, which has been suggested to be directly attributable to a desensitization of /-cells to SUs. [4][5][6][7] In vitro chronic exposure of pancreatic islets to SUs is also known to cause impairment of secretory response to subsequent stimulation by glucose or SUs. [8][9][10][11] The mechanisms underlying *Address for correspondence: Metabolic  Nateglinide (NAT), a novel D-phenylalanine derivative, shares the mechanism of action with SUs to act on KATP channels in pancreatic ]B-cells. This drug, compared to other marketed KATP channel-blocking hypoglycemic agents, has demonstrated unique characteristics including a rapid onset, short duration of action, sensitivity to ambient glucose, and resistance to metabolic inhibition, suggesting some aspects of the signaling pathway(s) mediating NAT's action are novel and distinct from those mediating the effects of SUs and REP. This study was designed to evaluate the ability of NAT and the com-parators, glyburide (GLY), glimepiride (GLI), and repaglinide (REP) to stimulate insulin secretion from isolated rat islets undergoing prolonged pre-exposure to SUs, GLY or tolbutamide (TOL). Using static incubation method, normal rat islets were incubated overnight (---18hours) in the absence and presence of therapeutically relevant concentrations of GLY or TOL before their acute responsiveness to NAT, GLY, REP, and GLI was determined. Our results showed that lasting treatment of GLY or TOL markedly attenuated the islets' acute response to GLY, GLI and REP, while NAT retained ability to stimulate insulin secretion from SU-desensitized islets. These findings are suggestive of common signaling pathways for the action of SUs and REP. On the other hand, the effectiveness of NAT in SU-desensitized islets could imply the involvement of an action site and/or a receptor/ effector pathway different from those of SUs and REP. The unique "desensitization-resistant" properties of NAT may explain some of the pharmacological differences between those of NAT and those of SUs and REP. [12,13] MATERIALS AND METHODS

Islet Isolation
Pancreas were dissected from normal fed male Sprague Dawley rats (250-275g), which were euthanized with Na pentobarbital i.p. at 120mg/kg. Islets of Langerhans were isolated by librase digestion (0.5mg/ml, Boehringer Mannheim, Germany) followed by a Ficoll gradient centrifugation. [14] Static Incubation of Islets Desensitized to SUs Freshly isolated islets were handpicked under a stereomicroscope by gentle suction through a large firepolished pipette (--4001xm diameter) into 60 15 mm Petri dishes (Corning) containing 25ml of DMEM (Dulbecco's Modified Eagle Medium, Gibco BRL) supplemented with 10% fetal calf serum, 5mM glucose (G5-DMEM) and 1% BSA (BSA was present in all incubation media throughout the experiments). Islets were preincubated in a humidified atmosphere of 95% 0 2 and 5% CO 2 at 37C for I hour.
The islets, after 1 hour preincubation, were selected in batches of two and placed into borosilicate glass tubes (1275mm) containing 500txl G8-DMEM in the presence or absence of 100nM or 10txM GLY or 100txM TOL, and incubated over night (--18hours) in a humidified atmosphere of 95% 02/5% CO 2 at 37C. Acute experiments were performed the following morning to determine the ability of islets to respond to NAT and other insulinotropic agents. 500txl 2x treatment of the concentrations of test hypoglycemic drugs was added to the tubes (to form 1 ml final volume) containing either GLY/TOL-treated or GLY/TOL-untreated islets. The concentrations of test hypoglycemic drugs were so chosen that they were comparably effective in control, i.e., 5txM NAT, 100nM GLY, 100nM GLI and 50nM REP. All test drugs were first dissolved in DMSO to form a stock solution of 10  The incubation media were diluted by factors ranging from lx to 20x depending on the concen-trations of glucose/drugs. The diluted super-natant was assayed for insulin content with SPA. [15] The assay employed commercially available products including a guinea pig antirat insulin specific antibody (Linco Research Inc) and scin-tillation proximity Type I reagent coupled to protein A (Amersham Life Science), and was performed as a single step assay. All samples were assayed in duplicate. adjusted eventually to reflect the degree of dilution. The intra-and inter-assay coeffi'cients of variation were generally between 5% and 8%.
To allow comparison of the hypoglycemic druginduced insulin secretion from overnight SUtreated and untreated islets, data were expressed as percent of the appropriate con-trois. That is to say, insulin secretion stimulat-ed by secretagogues in SU-treated and untreated groups was normalized, respectively, with the amount of insulin without secreta-gogues in SU-treated and untreated groups. This transformation corrected for the increase of control values elicited by overnight treat-ment of SUs. All values were expressed as the mean _SEM. Statistical significance was determined with t-test (single-tailed). P < 0.05 was considered significantly different.

RESULTS
Drug-induced Insulin Secretion from Islets with Overnight Treatment of 100nM GLY It was assumed that prolonged treatment with SUs would impair islet responsiveness to physiological and pharmacological stimuli. The effectiveness of four hypoglycemic agents, NAT, GLY, GLI, and REP, to stimulate insulin release from islets cultured overnight (18 hours) in the presence of 100nM GLY was evaluated and was compared to that from overnight cultured islets without GLY to determine the magnitude of desensitization induced by GLY. In the acute experiments following overnight incubation, the concentrations of test drugs were so chosen that they were comparably effective in stimulating insulin release at glucose concentration of 8mM, i.e., NAT at 5M, GLY at 100nM, GLI at 100nM, and REP at 50nM. These hypoglycemic agents stimulated insulin secretion during 1 hour static incubation from islets cultured overnight without 100nM GLY by a respective 270 44%, 292 42%, 301 26% and 430 +_ 35%. In parallel experiments on islets cultured overnight with 100nM GLY, the stimulation factors were, respectively, 354 + 19%, 302 + 13%, 340 + 12%, and 352 + 7%. These data are illustrated in Figure 2, in which the stimulation factors of all test agents were not statistically different between islet groups without and with pretreatment of 100nM GLY. Thus, prolonged treatment with glyburide at 100nM did not appear to render the islets insensitive to subsequent application of insulin secretagogues.

Drug-induced Insulin Secretion from Islets with Overnight Treatment of 10M GLY
When the concentration of GLY for overnight treatment was raised by 100-fold to 10zM, the pattern of the acute response of islets to insulin secretagogues has changed and the results are shown in Figure 3. The ability of NAT to stimulate insulin secretion did not significantly differ between the GLY untreated control (126 20%) and the GLY pretreated groups (155_ 14%). Conversely, the acute stimulation factors by GLY, REP and GLI in the groups without or  4 Insulin release during hour static incubation in response to acute challenge by antidiabetic agents from rat islets following overnight treatment without (empty bars) and with (filled bars) 1001xM TOL at G8. Data were pooled from 16 independent samples/condition and expressed as % of appropriate controls TOL). *p < 0.05 and **p < 0.005 when compare between data with and without pretreatment.
Basal insulin secretion during static incubation without and with TOL pretreatment was, respectively, 210 22 and 314 16 xU/islet/lh. treatment of islets with or without 100txM TOL are illustrated in Figure 4. NAT was able to stimulate insulin by 240 23% in control and 160 35 after TOL pretreatment. This change was not statistically significant. On the other hand, the stimulation factors in control or TOLpretreated islet groups were, respectively, 257+27% or 134+23% with GLY, 298 +27% or 137 + 13% with REP, and 260+38% or 155 +20% with GLI. The results reinforced the argument that islets desensitized to SUs retain responsiveness to NAT but not GLY, REP and GLI. The changes of insulin secretion in both GLY and TOL pretreated groups are summarized in the Table I.

DISCUSSION
The present study demonstrated that prolonged exposure of pancreatic islets to SUs, GLY or -TOL, impaired their ability to secrete insulin in response to subsequent secretagogue stimulation a phenomenon known as islet desensitization. Similar results have been previously reported for both drugs. [8][9][10]16] As KAT p channel blockade plays key role in the mechanism of action of SUs and all test drugs in this study, one possible mechanism for SU desensitization is that chronic exposure to SUs would result in lasting binding of SU to SUR1 and closing of KATP channels, and in turn, a lasting membrane depolarization. The constant depolarization would lead the ]3-cells to a refractory state, in which they respond less effectively to a further stimulation of closing KATP channels by test secretagogues. Evidence for this mechanism of SU desensitization associated with chronic exposure to GLY has been previously reported. [111 If this were indeed the case, SU desensitization is likely to be a temporary and reversible condition of refractoriness of the ]3-cells to secretagogue stimulation, since KATP channel activity would eventually recover upon withdrawal of SUs.
The key finding in the study was that NAT was the only drug of four test agents which maintained its insulinotropic efficacy in islets desensitized to SUs. Given that islets were still able to vigorously respond to NAT after 18h pretreatment of SUs, the exhaustion of a finite insulin store and/or islet secretory machinery did not appear to be the underlying defect in All data in SU-pretreated and non-pretreated groups are expressed, respectively, as percent of insulin secretion in appropriate control groups without acute stimulation by secretagogues. *p < 0.05 and **p < 0.005 on data with SU treatment compared to those without. desensitized secretion to GLY, REP and GLI. The insulinotropic action of REP, GLY and GLI was desensitized by lasting SU exposure, suggestive of common site of action for these drugs. On the other hand, the inability of both TOL and GLY to affect NAT-induced insulin secretion suggested a distinct mode of action of NAT. Provided that the mechanism of desensitization was indeed associated to the refractoriness of SUR1/KAT P channels in ]3-cells, a possible explanation for the effectiveness of NAT vs. ineffectiveness of GLY, REP and GLI in stimulation of insulin following prolonged pretreatment of SUs would be the existence of distinct binding sites on SUR1 for NAT vs. SUs and REP. It is possible that the site on SUR1 for SUs has become hypo-responsive to GLY, GLI and REP due to its desensitization induced by chronic treatment of GLY or TOL, while the site on SUR1 for NAT remained fully available to be activated with or without pretreatment of SUs. Alternatively, the present data would be reconciled by proposing a second mechanism of the insulinotropic action of NAT that is independent of KATP channels.
Being cautious not to over-interpret the in vitro findings in the present study, we only speculate that some aspects of signaling pathways mediating NAT's effect are markedly dissimilar from those involving in the effects of SUs and REP. Although the present study did not deal with time-dependent secretory pattern but cumulative stimulation of insulin secretion from islets during certain period of time, earlier works by others and us reveale essential differences of NAT action from those of GLY, GLI and REP with respect to (1) the kinetics of the interaction of NAT with SUR1 receptor/KATp channels; [17] (2) the kinetics of in vitrofin vivo insulinotropic action; [18][19][20][21][22] (3) the glucose-dependence; [23,24] and (4) the sensitivity to metabolic inhibitors. [25] These findings collectively indicated the uniqueness of NAT action and possibly a distinct receptor/ effector system(s). By virtue of these properties, NAT is able to ameliorate postprandial hyper-glycemia by augmenting early insulin release, while SUs and REP may preferentially decrease postabsorptive hyperglycemia due to their slower onset and sustained duration of action.
In conclusion, the maintenance of insulinotropic efficacy of NAT in islets desensitized to SUs adds to the list of properties of this agent that distinguish it from other SUR ligands, despite the presumed common basic mechanism of action of said agents.