Gagerman et al., 1980; Kalkhoffand Siegesmund, 1981; Henquin and Meissner, 1984). In the present study sodium was measured with an integrating flame ...
Bioscience Reports, Vol. 6, No. 11, 1986
Glucose-induced reduction of the Sodium Content in fl-Cell-Rich Pancreatic Islets Nils Wesslen, Peter Bergsten and Bo Hellman Received December 8, 1986 KEY WORDS: glucose; sodium; pancreatic fl-cells.
The sodium contents of fl-cell-rich pancreatic islets from ob/ob-mice were measured with an integrating flame photometer. After washing to an apparent steady state with different types of ice-cold media, islets incubated in the absence of glucose contained 79-108 mmol sodium kg- 1 dry weight. Exposure to glucose resulted in 25 % reduction of the islet content of sodium. This effect became manifest in the presence of 5 mM glucose, there being no additional reduction with a further increase of glucose to 20 mM. Depression of Na § activity may partially explain why glucose, under certain conditions, can lower cytoplasmic Ca: § and even inhibit insulin release.
INTRODUCTION The sodium ion has been proposed to act as a second messenger in glucose-stimulated insulin secretion by raising cytoplasmic Ca 2§ (Hales and Milner, 1968; Lowe et at., 1976; Donatsch et at., 1977). Although more recent studies have indicated that glucose can initiate release of insulin even after blocking the entry of Na § (Pace, 1979; Herchuelz and Malaisse, 1980; Biden et al., 1986), there is no doubt that alterations of the Na § activity influence the fl-cell handling of Ca 2§ and consequently the secretory activity. Divergent opinions have been expressed as to how glucose affects the sodium content of the pancreatic fl-cells (Sehlin and Tiiljedal, 1974; Kawazu et al., 1978; Gagerman et al., 1980; Kalkhoffand Siegesmund, 1981; Henquin and Meissner, 1984). In the present study sodium was measured with an integrating flame photometer in single islets rich in fl-cells. It is shown that the effect of glucose is not to increase but, on the contrary, to reduce the sodium content, an observation which might aid to elucidate why the sugar under certain conditions reduces the cytoplasmic Ca z§ activity and even inhibits insulin release. Department of Medical Cell Biology, University of Uppsala, Biomedicum, S-751 23 Uppsala, Sweden. 967 0144-8463/86/1100.0967505.00/0 9 1986PlenumPublishin$Corporation
968
Wessl~n, Bergstenand Hellman MATERIALS AND METHODS
Adult ob/ob-mice from a non-inbred colony were starved overnight. They were killed by decapitation and pancreatic islets isolated with collagenase. Previous studies have indicated that these islets contain more than 90% /%cells, which respond normally to glucose and other stimulators of insulin release (Hellman, 1970; Hahn et al., 1974). After 45 rain of preliminary incubation at 37~ in a Hepes-buffered medium physiologically balanced in cations with Cl- as the sole anion (Hellman, 1975) and containing 3 mM glucose and 1 mg/ml albumin, the islets were transferred to media of similar composition either devoid of glucose or containing 5 or 20 mM of the sugar. After 60 min of incubation, the islets, either washed or not with ice-cold media, were placed on aluminium foil and freed of as much contaminating fluid as possible using a micropipette. The washing was performed for various periods of time in media of different compositions as described below. After submerging the aluminium foil in isopentane, chilled to its freezing point ( - 165~ with liquid nitrogen, the islets were freeze-dried overnight and weighed on a quartz fiber balance. To restrict redistribution of sodium a standardized procedure of rapid handling was employed when transferring the islets to washing media and isopentane. Sodium was measured with an integrating flame photometer (Oberg et al., t967). Single islets (dry weight 2-6 ~g) were attached to a platinum-iridium wire previously dipped in a solution of 280mM glucose and 100ram (NH4)2HPO4. The effect of glucose was to make the islet adhere to the wire, and (NH4)2HPO4 minimized anion interference. Pure hydrogen and air were used for the burner. The sample was dried for 2 min at 150~ and then brought into the flame by a metal arm attached to a solenoid. Each observation was based on the mean of 4-8 measurements. It was checked that neither the solution of glucose and (NH4)2HPO4 nor the washing media produced any sodium signal or interfered significantly with it. Statistical significances for differences between paired observations were evaluated by Student's t-test.
RESULTS Islet contents of sodium after various periods of cold washing with 100raM MgC12 (panel A) or 300 mM sucrose (panel B) are shown in Fig. 1. It is evident that the washing results in fast removal of sodium approaching a steady state level. The washout pattern differed for the two types of washing media. Using sucrose there was a more rapid removal to a lower steady state, the loss being complete after 45 sec of washing. Exposure to 20 mM glucose resulted in a reduction of the islet content of sodium (Fig. 1). The glucose effect became evident when the islets were taken for washing in both MgC12 and sucrose. In one series of experiments (panel A), a statistically significant depression was noted also without washing of the islets. When pooling the results from the 17 experiments with non-washed islets in panel A and B, the reduction obtained in the presence of glucose was equivalent to 24 + 10 mmoles kg -1 dry weight (P < 0.05).
Sodium content of pancreatic/3-cells
969
A
200
I
T
w
150
E E v
100 z O O
O
50
WITH GLUCOSE
WITH GLUCOSE
7r162
PERIFUSION TIME (MIN)
Fig. 1. Sodium contents of islets washed for different periods of time with Cold solutions of MgC12 or sucrose. After preliminary incubation with 3 mM glucose the islets were transferred to media either devoid of glucose (9 or containing 20ram of the sugar (O). After 60min of incubation the islets were either freeze-dried directly or after different periods of washing with ice-cold media consisting of 100mM MgCt2 (A) or 300mM sucrose (B). The results are expressed as mmol sodium kg ~ dry weight and represent mean values -4-SEM for nine (A) and eight (B) experiments. Statistical significances for the effects of including 20 mM glucose in the incubation medium_are indicated by * P < 0.05; 9*P
