The use of dispersed pancreatic islet cells in measurements of transmembrane transport

Pancreatic beta cell line MIN6 exhibits characteristics of glucose metabolism and glucose-stimulated insulin secretion similar to those of normal islets

Glucose-stimulated insulin secretion, glucose transport, glucose phosphorylation and glucose utilization have been characterized in the insulinoma cell line MIN6, which is derived from a transgenic mouse expressing the large T-antigen of SV40 in pancreatic beta cells. Glucose-stimulated insulin secretion occurred progressively from 5 mmol/l glucose, reached the maximal level approximately seven-fold above the basal level at 25 mmol/l, and remained at this level up to 50 mmol/l. Glucose transport was very rapid with the half-maximal uptake of 3-O-methyl-D-glucose being reached within 15 s at 22 degrees C. Glucose phosphorylating activity in the cell homogenate was due mainly to glucokinase; the Vmax value of glucokinase activity was estimated to be 255 +/- 37 nmol.h-1.mg protein-1, constituting approximately 80% of total phosphorylating activity, whereas hexokinase activity constituted less than 20%. MIN6 cells exhibited mainly the high Km component of glucose utilization with a Vmax of 289 +/- 18 nmol.h-1.mg protein-1. Thus, glucose utilization quantitatively and qualitatively reflected glucose phosphorylation in MIN6 cells. In contrast, MIN7 cells, which exhibited only a small increase in insulin secretion in response to glucose, had 4.7-fold greater hexokinase activity than MIN6 cells with a comparable activity of glucokinase. These characteristics of MIN6 cells are very similar to those of isolated islets, indicating that this cell line is an appropriate model for studying the mechanism of glucose-stimulated insulin secretion in pancreatic beta cells.

A technique for the isolation of highly viable pancreatic B-cells from ob/ob mice

A method has been developed to prepare free islet cells in suspension from adult ob/ob-mice. About 200 collagenase-isolated pancreatic islets were pooled in 4 ml of calcium-free Krebs-Ringer-HEPES buffer supplemented with 1 mM EGTA and 10 micrograms/ml DNAase. The islets were gently shaken in a water-bath for 10 min at 30 degrees C. Then, the cell suspension was filtered through a nylon screen and centrifuged through ice-cold, dense albumin. The isolated cells, of which more than 99% were B-cells, appeared well preserved both in light- and electron-microscopy. Out of the isolated cells, 7.1 +/- 0.5% took up Evans Blue and were thus considered non-viable.

Accumulation of UDP-GlcNAc into intracellular vesicles and occurrence of a carrier-mediated transport. Study with plasma-membrane-permeabilized mouse thymocytes

Metabolic labelling of mouse thymocytes with radioactive mannose or glucosamine leads to the formation of labelled GDP-Man and UDP-GlcNAc. Using isotonic ammonium chloride treatment which renders the plasma membrane of thymocytes permeable to sugar nucleotides, we demonstrate that, in contrast to GDP-Man, a pool of UDP-GlcNAc remains associated with the cells after plasma membrane permeabilization. These observations are confirmed in experiments in which permeabilized thymocytes are incubated with exogenous labelled GDP-Man and UDP-GlcNAc, and we show that only UDP-GlcNAc is accumulated into sealed intracellular vesicles. This accumulation is a saturable process which can be inhibited by UDP, demonstrating the occurrence of a specific carrier. This transport mechanism can be blocked by covalent attachment of a non-permeant inhibitor UDP-dialdehyde without affecting the N-acetylglucosaminyltransferase itself. The fact that this carrier-mediated transport is not inhibited by tunicamycin indicates that this translocation process of UDP-GlcNAc does not involve lipid intermediates.

Effect of glibenclamide on the osmotic resistance of pancreatic beta-cells

The effect of glibenclamide on the osmotic resistance of beta-cells was measured using isolated beta-cells from ob/ob-mice. The beta-cells were incubated at different osmolarity and the diameters of the approximately spherical beta-cells were measured at 22 degrees C or at 37 degrees C with the aid of a screw micrometer eyepiece fitted to a light microscope. A near linear decrease of beta-cell diameter was found with increasing osmolarity (111-617 mosm/l). Control experiments showed that the membrane stabilizers, imipramine (0.1 mmol/l) or tetracaine (1 mmol/l), strongly reduced the osmotic swelling induced by low osmolarity (180 mosm/l). Glibenclamide (0.001 or 0.2 mmol/l) did not affect the beta-cell diameter at normal osmolarity (317 mosm/l) but reduced the swelling induced by hypoosmolarity (180 mosm/l) and the shrinking induced by hyperosmolarity (617 mosm/l). It is suggested that glibenclamide increases the osmotic resistance of isolated beta-cells by changing transmembrane flow of ions.

On insulin secretion

Aspects of insulin secretory mechanisms and models of diabetogenic B cell damage are discussed. Measurements of fluxes of 3H-labelled triphenylmethylphosphonium ion, 86Rb+, 42K+, 22Na+, and 45Ca2+ in isolated islets indicate that the triggering of insulin release depends on alterations in the interaction of ions with the B cells. One difficulty in the detailed analysis of these alterations are uncertainties which arise when macroscopic concepts for homogenous phases are applied to microscopic and heterogenous compartments, as exemplified by the meaning of pH in insulin secretory granules and of membrane electric potential. Nonetheless, the importance of an apparent decreased K+ permeability in mediating the insulin-releasing action of glucose, and of an apparent increased Na+ permeability in mediating the potentiating action of acetylcholine is emphasized. Fluorescent probing of Ca2+ by chlorotetracycline revealed effects of glucose alone as well as glucose-dependent and atropine-sensitive effects of acetylcholine. Although acetylcholine, sulfonylureas, and certain thiol-blocking agents may stimulate insulin release by direct effects on the B cell plasma membrane, a high capacity for D-glucose transmembrane transport has probably evolved in order that the interior of the B cells can always sense the circulating glucose concentration. A signal to secretion is thought to be transmitted from glucose metabolism to altered ion fluxes by intervention of reduced pyridine nucleotides and hypothetical redox protein for which thioredoxin may be a model. The insulin secretory defect in hereditary diabetic C57BL/KsJ-db/db-mice is apparently linked to a decreased basal permeability for K+ and a failure of the B cells to decrease further this permeability in response to glucose. Functioning B cells are acutely damaged when exposed to heterologous serum or alloxan in vitro; cytotoxic activation of complement by the alternative pathway could perhaps occur during islet inflammation. Protection experiments with free-radical scavengers in vitro and in vivo support the theory that hydroxyl radicals are instrumental in the production of alloxan diabetes. Rapid reduction of alloxan by thioredoxin in the presence of molecular oxygen and NADPH leads to strong chemiluminescence from luminol indicative of an intense radical protection. The sensitivity of B cells to alloxan may be due to physiological specializations of their plasma membranes, involving the highly effective glucose carrier or the hypothetical oxidation/reduction systems or both.

Further studies on 5-hydroxytryptamine transport in pancreatic islets and isolated beta-cells

1 The transport mechanism for (3)H-labelled 5-hydroxytryptamine (5-HT) in isolated pancreatic islets of non-inbred ob/ob mice was further characterized.2 Isolated beta-cells accumulated 5-HT to the same degree and with the same Na(+)-dependence as whole islets.3 Imipramine inhibited the uptake in a concentration-dependent way.4 Reserpine did not affect the uptake or efflux rates.5 Glucose stimulation of insulin secretion did not affect the uptake rate.6 It is concluded that the observed islet uptake of [(3)H]-5-HT represents an intracellular accumulation by the beta-cells. Mechanisms at the level of the plasma membrane may be rate-limiting for this process.

5'-AMP hydrolysis by suspensions and homogenates of pancreatic islet cells from normal and cortisone-treated rats

Suspensions of endocrine pancreas cells were prepared by shaking collagenase-isolated rat islets of Langerhans in calcium-free buffer. When incubated with 1.0 mM substrate at pH 7.4, the cells split Pi from 5'-AMP at a rate of 87 nmol/h per microgram DNA, and from beta-glycerophosphate at a rate of 25 nmol/h per microgram DNA. Km for 5'-AMP was about 54 microM. Adenosine or theophylline inhibited the 5'-AMP hydrolysis. Homogenization of the cells increased the activity toward 5'-AMP by 23% and that toward beta-glycerophosphate by 115%. Injecting rats with cortisone had no effect on the 5'-AMP hydrolysis by whole cells but significantly increased the activity in cell homogenates; the intracellular activity toward 5'-AMP was more than doubled by the cortisone treatment. Staining whole islet cells for 5'-AMP-splitting activity resulted in a demarcation of the cell periphery in control rats. Cells from cortisone-treated rats showed heavier deposits of reaction product, and their cell periphery did not stand out as clearly. It is suggested that 5'-nucleotidase is largely an ectoenzyme in normal rat islet cells. The cells also contain an as yet unidentified intracellular phosphatase that seems to be solely responsible for the increased hydrolysis of 5'-AMP in cortisone-treated rats.

45Ca2+ uptake by dispersed pancreatic islet cells: effect of D-glucose and the calcium probe, chlorotetracycline

Uptake of 45Ca2+ was studied in dispersed pancreatic islet cells from non-inbred ob/ob-mice. Like whole islets the dispersed cells responded to 20 mM D-glucose with a markedly increased 45Ca2+-labeling of both the lanthanum-nondisplaceable and the lanthanum-displaceable calcium pools. The pronounced effect of D-glucose could not be reproduced with 3-O-methyl-D-glucose, L-glucose, D-mannose, L-leucine, or D-leucine; however, 45Ca2+ uptake was greater in the presence of L-leucine as compared with D-leucine. 45Ca2+ uptake by dispersed cells or whole islets was stimulated severalfold by 100 microM or more chlorotetracycline. At the concentration of only 10 microM, chlorotetracycline had no effect on whole islets and partially inhibited 45Ca2+ uptake by the dispersed cells. The ability of D-glucose to stimulate 45Ca2+ uptake by islets or dispersed cells remained in the presence of 10 microM chlorotetracycline. Islet cell suspensions apparently represent a valid model for studying how Ca2+ interacts with the cells. However, when using chlorotetracycline as fluorescent Ca2+ probe, attention must be paid to it potential ionophoric activity. At only 10 microM, the drug seems to monitor a peripheral pool of Ca2+, some of which may reside in normal transport channels.

Anomeric preferences of D-glucose uptake and utilization by cerebral cortex slices of rats

On aerobic incubation of rat cerebral cortex slices with anomers of D-glucose and with 2-deoxy-D-glucose (2DG) for 5 min, the disappearance of beta-D-glucose from the incubation mixture was greater than that of alpha-D-glucose and both anomers had a greater rate of disappearance than that of 2DG. In addition, there were significantly greater consumption of oxygen and production of lactate with the beta-anomer than with the alpha-anomer. In similar experiments with 3H-labeled D-glucose anomers and [1-3H]-3-O-methyl-D-glucose (3MG), the accumulation of [1-3H]-beta-D-glucose (up to 5 min) by rat cerebral cortex slices was greater than that of [1-3H]-alpha-D-glucose. Although initially lower than that of the anomers, the accumulation of [1-3H]-3MG increased at a greater rate and, by 5 min of incubation, was greater than that of both glucose anomers. This preferential accumulation was seen to disappear when the slices were preincubated with 2DG (hexokinase inhibitor) or when the temperature of incubation was reduced to 20 degrees C. Under those conditions the data with the glucose anomers were similar to those obtained with 3MG. Our data then suggested that the greater accumulation of beta-D-glucose than of alpha-D-glucose by the slices was probably not due to differences in transport through brain cell membranes but rather to the preferential metabolism of the beta-D-glucose.

Islet-cell-surface antibodies in juvenile diabetes mellitus

Using an indirect immunofluorescence test on suspensions of viable, insulin-producing islet cells from rats, we found that 32 per cent (28/88) of insulin-treated patients with juvenile diabetes have islet-cell-surface antibodies in their circulation. These antibodies also occurred in four of nine children with glucose intolerance, in one of 24 healthy children and in nondiabetic children with thyroid disorders. In the diabetic children, the immunofluorescent reaction was inhibited by preadsorption of serum to islet cells but was little affected by preadsorption to rat hepatocytes or erythrocytes or to acetone powders of various rat tissues, including pancreas. These results show that organ-specific, nonspecies-specific antibodies reactive with the cell surface of the islet cells can be present in serum from diabetic children, and provide an approach to investigation of immunopathological aspects of diabetes mellitus.

Possible toxic effects of normal and diabetic patient serum on pancreatic B-cells

Serum from normal blood-donors and juvenile diabetic patients inhibited Rb+ accumulation and stimulated release of 51Cr and insulin in suspensions of dispersed pancreatic islet cells prepared from ob/ob mouse islets, which are rich in B-cells. The effects indicate the presence of a B-cytotoxic factor in human serum. Serum from mouse and fetal calf also inhibited the islet cell accumulation of Rb+. Toxicity was not suppressed by treating serum with protein A-Sepharose and did not correlate with islet cell binding of fluorescent antibodies to human immunoglobulin. Whereas all sera inhibited Rb+ accumulation, 3 of 6 diabetic patient sera, but no blood-donor serum, made the cells fluoresce on exposure to the fluorescent antibodies. Supporting a dependence on complement, toxicity remained after dialysis, but was destroyed by treating serum with zymosan-A or heating at 56 degrees for 30 min.

The dynamics of insulin release from mouse pancreatic islet cells in suspension

The overall dynamics of glucose-induced insulin release was strikingly similar in dispersed cells and intact islets perifused in parallel. Both preparations exhibited a latency of 1-2 min, after which period there was a brisk rise of insulin release followed by a sustained second phase. During the second phase, insulin release from dispersed cells attained a stable plateau rate, whereas the release from intact islets continued to rise. Epinephrine (1 muM) inhibited the release in both preparations, but the return to basal rate was faster in the dispersed cells than in the intact islets. The dispersed cells oxidized glucose at a constant rate for at least 60 min; the glucose oxidation was markedly sensitive to changes of the glucose concentration in the range of 3-20 mM.