A versatile microperifusion system

A Parallel Perifusion Slide From Glass for the Functional and Morphological Analysis of Pancreatic Islets

An islet-on-chip system in the form of a completely transparent microscope slide optically accessible from both sides was developed. It is made from laser-structured borosilicate glass and enables the parallel perifusion of five microchannels, each containing one islet precisely immobilized in a pyramidal well. The islets can be in inserted via separate loading windows above each pyramidal well. This design enables a gentle, fast and targeted insertion of the islets and a reliable retention in the well while at the same time permitting a sufficiently fast exchange of the media. In addition to the measurement of the hormone content in the fractionated efflux, parallel live cell imaging of the islet is possible. By programmable movement of the microscopic stage imaging of five wells can be performed. The current chip design ensures sufficient time resolution to characterize typical parameters of stimulus-secretion coupling. This was demonstrated by measuring the reaction of the islets to stimulation by glucose and potassium depolarization. After the perifusion experiment islets can be removed for further analysis. The live-dead assay of the removed islets confirmed that the process of insertion and removal was not detrimental to islet structure and viability. In conclusion, the present islet-on-chip design permits the practical implementation of parallel perifusion experiments on a single and easy to load glass slide. For each immobilized islet the correlation between secretion, signal transduction and morphology is possible. The slide concept allows the scale-up to even higher degrees of parallelization.

Dynamics of Insulin Secretion from EndoC-βH1 β-Cell Pseudoislets in Response to Glucose and Other Nutrient and Nonnutrient Secretagogues

The dynamics of insulin secretion were characterized in response to a variety of physiological and pharmacological stimulators and other compounds in perifused pseudoislets generated from cells of the EndoC-βH1 β-cell line. Perifusion of EndoC-βH1 pseudoislets with the physiological stimulus glucose (16.7 mM) induced sustained insulin secretion, which was inhibited by mannoheptulose. The adenylate cyclase activators IBMX and forskolin strongly potentiated this secretion. Glibenclamide, a Kir 6.2 potassium channel blocker, and Bay K 8644, an opener of the voltage-sensitive Ca²⁺ channel, also potentiated glucose-induced insulin secretion. The dynamics of insulin secretion from EndoC-βH1 pseudoislets were characterized by an insulin secretory response to glucose starting within 1-2 min and passing over without interruption into a sustained phase of insulin release for the whole stimulation period. This lack of a transient decline between the first and the second phases of insulin release is an indication for a quick supply of insulin secretory granules from the reserve pool to the docking sites below the plasma membrane. Thereby, new secretory granules are directly made available for sustained exocytosis of insulin in EndoC-βH1 β-cells. The study shows that EndoC-βH1 β-cell pseudoislets are well suited for kinetic analyses of insulin secretion.

Determination of beta-cell function: insulin secretion of isolated islets

The kinetics of insulin secretion, not just the total amount, is of decisive relevance for the physiological regulation of glucose homeostasis. Thus to characterize the relevant features of the secretory response to an insulinotropic stimulus a method is needed which is able to resolve the temporal response pattern, in particular to distinguish the first phase from the second phase response. The perifusion of collagenase-isolated islets is a method which permits to register responses of near-physiological complexity with a preparation that can also be used for cell physiological and biochemical investigations on stimulus--secretion oupling.

Novel insights into the regulation of the bound and diffusible glucokinase in MIN6 beta-cells

A stable MIN6 beta-cell clone overexpressing glucokinase as an enhanced cyan fluorescent protein (ECFP) fusion construct was generated for analysis of glucokinase regulation in these glucose-responsive insulin-secreting cells. A higher glucokinase enzyme activity accompanied by an improved glucose-induced insulin secretion indicated the integration of ECFP-glucokinase into the functional pool of glucokinase protein in MIN6-ECFP-glucokinase cells. Fluorescence recovery after photobleaching experiments of MIN6-ECFP-glucokinase cells and photoactivation of a transiently transfected photoswitchable cyan fluorescent protein (PS-CFP)-glucokinase construct in MIN6 cells indicate a higher motility of the diffusible glucokinase fraction at high glucose concentrations. In agreement with previous studies, we observed significant binding of ECFP-glucokinase to insulin secretory granules. Using fluorescence lifetime imaging, we obtained evidence for an association between glucokinase and alpha-tubulin in MIN6-ECFP-glucokinase cells. Furthermore, immunohistochemistry and fluorescence resonance energy transfer analysis by acceptor photobleaching showed distinct association between endogenous glucokinase and alpha-tubulin as well as beta-tubulin in MIN6 cells. Interestingly, glucokinase was also colocalized with kinesin, a motor protein involved in insulin secretory granule movement. Therefore, we suggest a role of a bound glucokinase protein fraction in the regulation of insulin granule movement along tubulin filaments.

A buffer temperature controlled perifusion system to study temperature dependence and kinetics of insulin secretion in MIN6 pseudoislets

INTRODUCTION

The perifusion of pancreatic islets is a well-known method to investigate the kinetics of insulin secretion. Nevertheless, little interest has been attributed to a precise temperature control in perifusion systems. Insulin secretion from MIN6 cells, cultured as monolayers, differs substantially from pancreatic islets, at least partly due to missing beta-to-beta cell contacts. These cellular contacts are abundant in MIN6 pseudoislets, which show a more pronounced glucose-induced insulin release. Here, a perifusion system that directly and dynamically controls the perifusion buffer temperature inside the reaction chamber is described. Additionally, the influence of small temperature changes, glucagon-like peptide 1 (GLP-1) and tolbutamide on insulin release from MIN6 pseudoislets is examined.

METHODS

MIN6 cells were cultured in suspension culture dishes to generate MIN6 pseudoislets. The pseudoislets were perifused using a newly developed 12-channel perifusion system. The buffer temperature inside the reaction chambers was dynamically controlled by a programmable proportional plus integral plus differential (PID) controller. Insulin was determined by radioimmunoassay.

RESULTS

After adjusting the PID controller, the temperature inside the reaction chambers was constant in a very narrow range. The first phase of the glucose-induced insulin secretion was enhanced from 1.0+/-0.1 to 2.8+/-0.2 ng insulin/ml and the second phase from 5.4+/-0.9 to 17.8+/-1.3 ng insulin/ml, when the temperature was elevated by 1 degrees C, from 37 to 38 degrees C. GLP-1 concentration dependently increased insulin release at 15.0 mM and was ineffective at 0.0 mM glucose. Tolbutamide induced a concentration-dependent increase in both phases of the insulin secretion.

DISCUSSION

MIN6 pseudoislets are a useful tool to study insulin secretion from beta-cells, which are arranged in clusters like pancreatic beta-cells in the islet. The strong influence of temperature on insulin release from these pseudoislets requires a perifusion system, which precisely controls the buffer temperature.

Control of insulin secretion by sulfonylureas, meglitinide and diazoxide in relation to their binding to the sulfonylurea receptor in pancreatic islets

Sulfonylureas inhibit an ATP-dependent K+ channel in the B-cell plasma membrane and thereby initiate insulin release. Diazoxide opens this channel and inhibits insulin release. In mouse pancreatic islets, we have explored whether other targets for these drugs must be postulated to explain their hypo- or hyperglycaemic properties. At non-saturating drug concentrations the rates of increase in insulin secretion declined in the order tolbutamide = meglitinide greater than glipizide greater than glibenclamide. The same rank order was observed when comparing the rates of disappearance of insulin-releasing and K+ channel-blocking effects. The different kinetics of response depend on the lipid solubility of the drugs, which controls their penetration into the intracellular space. Allowing for the different kinetics, the same maximum secretory rates were caused by saturating concentrations of tolbutamide, meglitinide, glipizide and glibenclamide. A close correlation between insulin-releasing and K+ channel-blocking potencies of these drugs was observed. The relative potencies of tolbutamide, meglitinide, glipizide and glibenclamide corresponded well to their relative affinities for binding to islet-cell membranes, suggesting that the binding site represents the sulfonylurea receptor. The biphasic time-course of dissociation of glibenclamide binding indicates a complex receptor-drug interaction. For diazoxide there was no correlation between affinity of binding to the sulfonylurea receptor and potency of inhibition of insulin secretion. Thus, opening or closing of the ATP-dependent K+ channel by diazoxide or sulfonylureas, respectively, appears to be due to interaction with different binding sites in the B-cell plasma membrane. The free concentrations of tolbutamide, glipizide, glibenclamide and diazoxide which are effective on B-cells are in the range of therapeutic plasma concentrations of the free drugs. It is concluded that the hypo- and hyperglycaemic effects of these drugs result from changing the permeability of the ATP-dependent K+ channel in the B-cell plasma membrane.

Glucose both inhibits and stimulates insulin secretion from isolated pancreatic islets exposed to maximally effective concentrations of sulfonylureas

Isolated pancreatic islets from mice were perifused with media containing maximally effective concentrations of glibenclamide (0.1-10 mumol/l) or glipizide (1 mumol/l). In these islets an increase of the glucose concentration from 10 mmol/l to 40 mmol/l or addition of D-glyceraldehyde (20 mmol/l) caused a temporary decrease in insulin release which was followed by a sustained enhancement of release. alpha-Ketoisocaproate (3 or 20 mmol/l) did not inhibit insulin release; at high concentration it was an even stronger secretagogue than D-glucose or D-glyceraldehyde. It is concluded that high energy phosphates couple B-cell fuel metabolism and insulin release by acting both on the ATP-dependent K+ channel and on other targets not yet identified.

Phentolamine, a deceptive tool to investigate sympathetic nervous control of insulin release

We investigated the effects of phentolamine and another more selective alpha 2-adrenoceptor antagonist, rauwolscine, on insulin release in vivo (in female Wistar-rats) and in vitro (in perfused rat pancreas and in isolated perifused mouse islets). Phentolamine was found to significantly increase glucose-induced insulin release. On the other hand, rauwolscine failed to do so, when applied in a concentration that effectively antagonized the inhibitory effect of clonidine. These results demonstrate that phentolamine is capable of directly stimulating insulin release. This effect is thus not mediated by alpha-adrenoceptors. For this reason phentolamine is not an appropriate tool to study possible inhibitory effects of the sympathetic nervous system on insulin release. An enhanced insulin response as may be observed in animals and in man in the presence of phentolamine does not furnish evidence for a tonic inhibitory control of the islet cells by the sympathetic nervous system.

Regulation of energy metabolism in pancreatic islets by glucose and tolbutamide

The kinetics of insulin secretion and oxygen uptake in response to D-glucose and tolbutamide were compared in mouse pancreatic islets. In addition, the role of decreased ATP as a driving force for secretagogue-induced oxygen consumption was examined. D-glucose (10-30 mmol/l) triggered a biphasic insulin release which always coincided with a monophasic increase in islet oxygen uptake. In the presence of D-glucose (5-30 mmol/l), tolbutamide (3-500 mumol/l) consistently elicited an initial peak of insulin secretion which was followed by a continued decline. Tolbutamide-induced secretory profiles were accompanied by similar respiratory profiles. Oxygen consumption per ng of insulin released during the test phase was higher after elevation of the glucose concentration than after addition of tolbutamide. In conjunction with 5 or 10 mmol/l D-glucose, but not with 15 or 30 mmol/l D-glucose, tolbutamide (30-100 mumol/l) lowered islet ATP content significantly (p less than 0.02). Phosphocreatine was not found in isolated islets, although they contained substantial creatine kinase activity. It is concluded that the driving force for tolbutamide-induced oxygen uptake is a decrease in the phosphorylation potential caused by the work load imposed by stimulation of the secretion process. However, a major proportion of the respiratory response to glucose also results from enhancement of biosynthesis.

Effects of isoprenaline and glucagon on insulin secretion from pancreatic islets

The effects of isoprenaline and glucagon on insulin secretion from pancreatic islets were investigated. In the presence of high concentrations of isoprenaline (10-50 mumol/l), glucose-induced (20 mmol/l) insulin secretion from isolated perifused mouse islets was inhibited. This inhibition was apparently mediated by alpha 2-adrenoceptors, as it was antagonized by rauwolscine. At low concentrations isoprenaline (0.1 or 1 mumol/l) did not affect glucose-induced (2.5; 10 or 20 mmol/l) insulin secretion from perifused mouse or rat islets, even if alpha 2-adrenoceptors were blocked by rauwolscine. A stimulatory effect of isoprenaline on insulin secretion was also not observed in the perfused rat pancreas. However, when incubated mouse islets were exposed to glucose (10 mmol/l), insulin secretion was further enhanced by isoprenaline (0.5 mumol/l). To elucidate the underlying mechanism, the effects of glucagon on insulin secretion were investigated, because glucagon is released from the pancreatic A-cells during stimulation with isoprenaline and is accumulated in the islets and the surrounding medium during incubations of pancreatic islets. Indeed, glucagon stimulated insulin secretion from perifused mouse islets in the presence of high glucose (10 or 15 mmol/l) concentrations but not of low glucose (5 mmol/l) concentrations. Thus it is concluded that direct beta-adrenergic stimulation of pancreatic B-cells does not occur in mouse or rat pancreatic islets. Augmentation of glucose-induced insulin secretion by isoprenaline observed in incubation systems can be explained as a result of stimulation by glucagon, which is released from pancreatic A-cells by isoprenaline.

The dihydropyridine derivative, Bay K 8644, enhances insulin secretion by isolated pancreatic islets

The effects of the dihydropyridine derivative Bay K 8644 upon insulin secretion by perifused isolated mouse pancreatic islets were examined. At a non-stimulatory glucose concentration (5 mmol/l) Bay K 8644 (1 mumol/l) did not stimulate insulin release. However, the same drug concentration enhanced the insulin secretory responses to an intermediate (15 mmol/l) or high (30 mmol/l) glucose concentration by 80 or 90%, respectively. Bay K 8644 was half maximally effective at 0.1 mumol/l and maximally effective at 1 mumol/l. The results are compatible with the view that voltage-dependent calcium channels are essential for stimulus-secretion coupling in pancreatic B-cells.

Regulation of insulin secretion by energy metabolism in pancreatic B-cell mitochondria. Studies with a non-metabolizable leucine analogue

In mouse pancreatic islets the kinetics of insulin secretion and O2 uptake in response to the non-metabolizable leucine analogue (+/-)-BCH (2-endo- aminonorbornane -2-carboxylic acid) were compared. In addition, the fuel-mobilizing effect of (+/-)-BCH was studied with a mitochondrial fraction from islets. (1) Within 2 min 20 mM-(+/-)-BCH markedly enhanced insulin release or O2 consumption by islets respiring in the absence of exogenous fuels. During prolonged exposure to 20 mM-(+/-)-BCH secretion declined more rapidly than O2 uptake. (2) L-Glutamine (10 mM) prevented the decrease of both insulin release and O2 uptake of islets exposed to 20mM-(+/-)-BCH. During the second phase of insulin release in response to 20 mM-(+/-)-BCH + 10 mM-L-glutamine, kinetics of secretion and respiration correlated closely. (3) Initial peaks were consistently seen in the (+/-)-BCH-induced secretory profiles, but never in the respiratory profiles. (4) In contrast with L-glycerol 3-phosphate, L-malate or pyruvate, L-glutamine or L-glutamate maintained low rates of oxidative phosphorylation in B-cell mitochondria. The effects of L-glutamine or L-glutamate were potentiated severalfold by (+/-)-BCH. (5) The effects of other branched-chain amino acids on oxidative phosphorylation resembled their effects on insulin release, redox state of nicotinamide nucleotides and glutamate dehydrogenase activity. (6) The results support the view that (+/-)-BCH stimulates insulin secretion via a primary enhancement of hydrogen supply to the respiratory chain of B-cell mitochondria.

Connecting filaments, core filaments, and side-struts: a proposal to add three new load-bearing structures to the sliding filament model

This report concerns structural forces in resting muscle and proposes three additions to the sliding filament model to account for these mechanical properties. The proposal includes: connecting filaments (C-filaments) which connect the ends of each thick filament to the neighboring Z-lines, core filaments which support the myosin of the thick filament and which attach to the C-filaments, and side-struts which bind the thick filaments together along their length and restrict their radial movement. C-filaments would act as the parallel elastic element and transmit the passive tension to the thick filaments. Isolated myofibrils (mechanically-skinned and detergent-treated frog semitendinosus fibers) when stretched progressively showed exponentially-increasing passive tension which did not disappear when filament overlap was exceeded, but continued to rise. SL was monitored with a HeNe laser. Passive tension phasically exceeded 3 X 10(5) N/M2. Electron microscopy (thin-sectioned and freeze-fracture/deep-etch specimens) of non-overlap fibers showed orderly fibril structure with clear separation of A- and I-bands. In the gap between them could be seen filaments, 40-50 A in diameter, connected to the thick filament ends. Unlike actin, these filaments did not become decorated by myosin S-1. Equatorial X-ray measurements showed that stretching relaxed skinned muscles squeezed the thick filaments closer; this radial compression continued beyond filament overlap. Extreme stretch of fibers caused the thick filaments to strain several-fold. Treatment of non-overlap fibers with a high ionic strength pyrophosphate myosin solvent caused a large drop in passive tension and stiffness, but no change in SL was detected nor was myofibril continuity detectably affected. Non-overlap fibrils, when treated with elastase, released A-segments which retain three-dimensional coherency . Deep-etch EM's of non-overlap fibers disclosed abundant structures (about 75 A) wide attaching adjacent thick filaments.

Autofluorescence-activated cell sorting of pancreatic islet cells: purification of insulin-containing B-cells according to glucose-induced changes in cellular redox state

Autofluorescence-activated cell sorting can be employed for the subfractionation of insulin-containing islet B-cells according to their responsiveness to their physiologic stimulus, glucose. The method utilizes a flow cytometric detection of the rapid variations in endogenous NAD (P) H - and FAD - fluorescence after exposure to 20 mM glucose. Under these conditions, a two-fold increase in NAD (P) H and a 40% decrease in FAD was observed in more than 75% of B-cells isolated from fed normal rats. The technique makes it possible to study the metabolic behaviour of the B-cell population in (physio)pathological conditions of impaired glucose-induced insulin release; the availability of functionally homogenous B-cell preparations facilitates studies on stimulus-secretion coupling. In view of the universal role of the cellular metabolic redox state in cell regulation, it is suggested that similar techniques can be developed for the metabolic analysis of other cell types and for their purification according to their responsiveness to specific stimuli.

Effects of alpha-adrenoceptor antagonists on clonidine-induced inhibition of insulin secretion by isolated pancreatic islets

The effects of clonidine, yohimbine, corynanthine and prazosin on glucose-induced insulin secretion by incubated or perifused mouse pancreatic islets were investigated. Clonidine (0.1 microM) inhibited glucose-induced insulin secretion alone and in the presence of yohimbine (0.1 microM), corynanthine (10 microM) or prazosin (1 microM). In higher concentrations, yohimbine (1-10 microM) antagonized the inhibitory effect of clonidine (0.1 microM) upon glucose-induced insulin secretion by incubated islets and by perifused islets. The results support the view that adrenergic inhibition of insulin secretion is mediated by alpha 2-adrenoceptors on pancreatic beta-cells.

A method for the simultaneous measurement of insulin release and B cell membrane potential in single mouse islets of Langerhans

A method has been developed for the simultaneous measurement of insulin release and electrical activity in single micro-dissected mouse islets of Langerhans. The effects of D-glucose have been studied in individual islets. Each islet was exposed to 0, 5.6, 11.1, 16.7, 22.2, 27.8, and 33.3 mmol/l glucose in a stepwise fashion. The minimum glucose concentration required to elicit spike activity is lower than that required to stimulate insulin release above basal levels and the maximum spike frequency occurs at a lower glucose concentration than does maximum insulin release. Following a reduction in glucose from 27,8 (or 33.3) to 5.6 mmol/l, membrane potentials returned to resting values within 2 min whereas insulin returned to basal values after 20 min. Increasing glucose from 5.6 to 27.8 mmol/l induced spike activity within 10 s; the insulin response was detected within 40 s. Thus, it is possible to use the single mouse islet for simultaneous measurements of insulin release and electrical activity.

Increased somatostatin secretion from pancreatic islets of streptozotocin-diabetic rats in response to glucose

Glucose stimulates somatostatin release from perifused pancreatic islets of diabetic rats 42-47 days after the induction of diabetes, and 48 h after withdrawal of insulin replacement therapy. The glucose effect is augmented by theophylline or glucagon. Basal somatostatin release and glucose-induced secretion are significantly higher in diabetic islets than in controls. It is suggested that glucose promotes somatostatin release by directly interacting with islet D cells but not via indirect pathways. Glucose-induced stimulation appears to be modulated by a D-cell adenylate cyclase/phosphodiesterase system. Reasons responsible for increased somatostatin secretion by diabetic islets include reduction in B-cell mass, suggesting that B cells may normally suppress the secretory activity of D cells.

Comparison of alpha-ketoisocaproic acid and glucose in rats: effects on insulin and somatostatin release and on islet cAMP content

The insulinotropic effects of alpha-ketoisocaproic acid and glucose reveal many common characteristics in vivo and in vitro. They qualify as initiators of insulin release, their action is amplified by potentiators of insulin release, and they have a similar potency at equimolar concentrations. The dynamics of insulin release evoked by alpha-ketoisocaproic acid and glucose are similar. Epinephrine completely inhibits the insulinotropic effect of glucose and alpha-ketoisocaproic acid. Mannoheptulose exhibits a complete, immediate and reversible blockade of glucose-induced insulin release. In contrast, inhibition of alpha-ketoisocaproic acid-induced insulin release occurs after a lag period and is not reversed by removal of the inhibitor. alpha-ketoisocaproic acid, at equimolar concentrations, is several-fold more effective than glucose in elevating cAMP content in islet. alpha ketoisocaproic acid and glucose are about equally effective in stimulating somatostatin release from isolated rat pancreatic islets. This stimulation is inhibited by epinephrine. Mannoheptulose inhibits only somatostatin release induced by glucose but not by alpha-ketoisocaproic acid. It suggested that the insulinotropic characteristics of glucose and alpha-ketoisocaproic acid reveal many common features, while their mode of action appears to be different.