Differences in Glucose Handling by Pancreatic A- and B-cells*

Glucose exerts opposite effects upon glucagon and insulin release from the endocrine pancreas. Glucose uptake and oxidation were therefore compared in purified A- and B-cells. In purified B-cells, the intracellular concentration of glucose or 3-O-methyl-D-glucose equilibrates within 2 min with the extracellular levels, and, like in intact islets, the rate of glucose oxidation displays a sigmoidal dose-response curve for glucose. In contrast, even after 5 min of incubation, the apparent distribution space of D-glucose or 3-O-methyl-D-glucose in A-cells remains much lower than the intracellular volume. In A-cells, both the rate of 3-O-methyl-D-glucose uptake and glucose oxidation proceed proportional to the hexose concentration up to 10 mM and reach saturation at higher concentrations. Addition of insulin failed to affect 3-O-methyl-D-glucose or D-glucose uptake and glucose oxidation by purified A-cells. Glucose releases 30-fold more insulin from islets than from single B-cells, but this marked difference is not associated with differences in glucose handling. The rate of glucose oxidation is virtually identical in single and reaggregated B-cells and is not altered after addition of glucagon or somatostatin. It is concluded that the dependency of glucose-induced insulin release upon the functional coordination between islet cells is not mediated through changes in glucose metabolism.

Glucose exerts opposite effects upon glucagon and insulin release from the endocrine pancreas. Glucose uptake and oxidation were therefore compared in purified Aand B-cells. In purified B-cells, the intracellular concentration of glucose or 3-0-methyl-D-glucose equilibrates within 2 min with the extracellular levels, and, like in intact islets, the rate of glucose oxidation displays a sigmoidal dose-response curve for glucose.

In contrast, even after 5 min of incubation, the apparent distribution space of D-glucose or 3-O-methyl-~glucose in A-cells remains much lower than the intracellular volume. In A-cells, both the rate of 3-0methyl-D-glucose uptake and glucose oxidation proceed proportional to the hexose concentration up to 10 m M and reach saturation at higher concentrations. Addition of insulin failed to affect 3-O-methyl-~-glucose or D-glucose uptake and glucose oxidation by purified
Glucose releases 30-fold more insulin from islets than from single B-cells, but this marked difference is not associated with differences in glucose handling.

A-cells.
The regulation by circulating nutrients of insulin and glucagon release from the endocrine pancreas plays an essential role in the hormonal control of fuel homeostasis. The observat,ion that glucose exerts opposite effects upon insulin and glucagon release (1-7) raises questions as to the mechanisms involved in such opposite actions on neighboring cells. Glucose recognition by the pancreatic B-cell has been extensively studied in rat and mouse islets, which contain 60-90% B-cells (8)(9)(10)(11)(12). The analysis of A-cells has been hampered by their lower occurrence in normal rodent islets (10-30% of islet cells); they have been examined predominantly in other experimental models such as the A-cell-rich islets from birds (13) or from rodents with experimental diabetes (5), which contain 60-70% A-cells (14, 15). This approach raises, however, new problems in terms of the low amount of cellular * This work was supported by Grants 80-85/9 and 80-85/16 from the Belgian Ministry of Scientific Policy and Grant 3.0021.80 from the Nationaal Fonds voor Wetenschappelijk Onderzoek. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "adoertisernent" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$ To whom correspondence should be addressed at, Department of Metabolism and Endocrinology, Vrije Universiteit Brussel, Fakulteit Geneeskunde en Farmacie, Laarbeeklaan 103, B-1090 Brussels, Belgium. material and the eventual impairment of A-cell function after exposure to alloxan or streptozotocin (16).
The recent development of procedures for the purification of different types of islet cells (17, 18) makes it technically possible to conduct a comparative analysis on normal Aand B-cells. The present study examines glucose transport and oxidation rates in purified A-and B-cells; preparations of single and structurally coupled B-cells were used to assess whether glucose handling varies with cellular aggregation.
Preparation of Purified Islet Aand B-cells-Islets of Langerhans were prepared by collagenase digestion of pancreata from fed adult male Sprague-Dawley rats and were dissociated by trypsin treatment in calcium-free KRBH (17). Single B-cells (more than 95% B-cells) were separated from single non-B-cells (more than 85% A-cells and less than 3% B-cells) by fluorescence-activated cell sorting (18). Non-B-cell fractions will also be referred to as A-cell fractions. The purified islet cells were used either immediately after their separation or following a 16-h culture period at 37 "C and in humidified 95% air, 5% coz (19).
Reaggregated B-cells were obtained by culturing single B-cells a t 30,000 cells/ml in a rotatory shaker incubator (30 rpm; Braun, Melsungen, FRG); after 16 h in basal culture medium (CMRL-1066, 5 mM glucose), clusters of 10-100 cells were collected from the culture flasks. After filtering the cultured cells through a Percoll layer of density 1.045 g/ml (17), the cells were washed and finally taken up in the media used for uptake or oxidation experiments.
In both cultured and uncultured cell preparations, more than 90% of the cells accumulated neutral red.
Preparation of Fibroblasts-Fibroblasts were prepared by trypsin treatment of finely minced 14-day-old rat embryos (20); they were grown to confluency a t 37 "C and 10% COZ in 25-cm2 Falcon tissue Islet Cell Glucose Handling 1197 culture flasks (1-2 days) containing Eagle's minimal essential medium with Earl's salts and 10% (v/v) heat-inactivated fetal calf serum.
After four cycles of trypsin treatment of the cultured cells and subsequent reculturing (20), the cells were seeded into poly-L-lysinecoated (2 wg/ml in distilled water for 1 h a t 20 "C) polyvinyl chloride microtiter plates (96 x 0.25 ml; Dynatech, Zug, Switzerland) and used for uptake experiments after confluency was reached (approximately io5 cells/well).
Glucose Uptake-After preincubating the purified islet cells for 30 min at 37 "C in glucose-free KRBH, they were distributed in 0.1-ml samples containing 75,000 cells and submitted to a further 1-30 min incubation at 37 "C in the presence of 0.6 mM [6,6'-3H]sucrose (25 &i/ml) and 0.6 mM [I4C]urea (12.5 Fci/mI), 0.6-20 mM 3-O-methyl-D-[U-"C]glucose (12.5 fici/ml), or 0.6-20 mM D-[U-"C]glucose (12.5 WCi/ml). The uptake experiments were stopped by filtering the cells during a 4-s centrifugation (8000 X g; Microfuge B, Beckman Instruments) through 0.1 ml of di-n-butyl phthalate (d 1.045 g/ml) (21). The radioactivity of the pellet was counted, and the apparent distribution space of the "C-labeled compounds was calculated from the radioactivity measurements; results were corrected for extracellular contamination as judged by the [3H]sucrose space measured in the same sample (21).
Fibroblasts were kept in microtiter plates, washed with glucosefree KRBH, and then treated as islet cell suspensions. Uptake was stopped by aspirating the medium and quickly washing the cells with ice-cold KRBH. After lysing the cells in 0.15 ml of water containing 2 mM EGTA, pH 7.0, each fraction was transferred to a scintillation vial and counted in 10 ml of Dyna-Gel. Calculations were carried out as described above for islet cells.
Glucose Oxidation-Glucose oxidation experiments were performed in siliconized glass t,ubes containing 75,000 islet cells in 0.06 ml of KRBH and 0.6-30 mM D-[U14-C]glucose (10 wCi/mI). Each tube was placed in a sealed 15-ml scintillation vial, and oxidation experiments were carried out as previously described (22). Incubation time was 1 h for B-cells and 2 h for non-B-cells.
Statistical Analysis-Results are expressed as mean k S.E. for the number of experiments stated in parentheses. The statistical signif" cance of differences between experimental groups was assessed by the Student's t test for unpaired data.

RESULTS
Glucose Uptake-Raising the concentration of D-glucose or its nonmetabolizable analogue, 3-0-MG results in a rapid equilibration across the B-cell membrane. Within 2 min, the apparent distribution space of both sugars equals the cellular water space as estimated by urea distribution (Fig. 1). This phenomenon is observed at both low (0.6 mM) and high (20 mM) glucose concentrations (Table I). After this first initial rise, a further but much slower accumulation of radioactivity is noted with ['4C]glucose, but not with labeled 3-0-MG ( Fig.  1). A modest increase of the urea space in non-B-cells is noted between the 5th and 30th minute of incubation.
D-Glucose and 3-0-MG uptake proceeds a t a much slower rate in islet non-B-cells than in B-cells. Even after 30 min, their apparent distribution space does not exceed 70-80% of the corresponding urea space (Fig. 1). At 0.6 mM, D-glucose uptake is slightly, although not significantly, higher than that of 3-0-MG (Fig. 1). The initial rate of 3-0-MG uptake remains constant over the first 5 min and proceeds proportionally to the prevailing 3-0-MG levels (Fig. 2) up to at least 10 mM. Eadie-Hofstee analysis suggests a K , of  All presented uptake results were collected on uncultured islet cells; identical results were obtained for cultured Band non-B-cells (results not shown). In contrast to non-B-cells, fibroblasts incorporated 2-fold more glucose when exposed to 4 pg/ml of insulin (Table 11).
Glucose Oxidation-The dose-response curve of glucose OX- (Table 11).  with the external glucose concentration up to 10-15 mM and levels off at higher concentrations (Fig. 3). The estimated K , is comparable to that observed for glucose transport, whereas V,,,,, approximates 25-35 fmol of CO,/h/cell. The presence of insulin (0.5 Fg/ml) has no significant effect on the amount of glucose oxidized by non-B-cells in the presence of 16.7 mM glucose (Table 111). A 16-h culture period at 5.5 mM glucose did not significantly alter glucose oxidation by non-B-cells, nor did it induce responsiveness to insulin (results not shown).
Neither glucagon (1 pg/ml) nor somatostatin (50 ng/ml) alters the rate of glucose oxidation by B-cells (Table IV). Glucose oxidation rates in cultured single B-cells were similar to those measured in reaggregated B-cells (Table IV). Furthermore, the glucose oxidation rate of freshly isolated islets compared favorably with that of uncultured purified Band non-B-cells isolated from the same batch of islets. In this experiment, islets were found to contain in average 1500 Bcells and 600 A-cells as judged from the insulin and glucagon content of the different cell preparations. Taking into account the measured rates of glucose oxidation of the purified B-and non-B-cells at 8.3 and 16.7 mM glucose together with the above mentioned calculated cell numbers/islet, the predicted glucose oxidation rate/islet averaged 125 pmol of CO,/h a t 16.7 mM glucose and 75 pmol of COJh at 8.3 mM, which agreed well with our measured values of 112 f 4 ( n = 3) and 69 f 9 (n = 3), respectively. It thus appears that the amount of glucose oxidized per islet at 8.3 or 16.7 mM glucose is comparable to the sum of the quantities oxidized by an equivalent number of single islet cells.

DISCUSSION
In basal media containing no amino acids, glucose uptake proceeds much slower in A-cells than in B-cells; Eadie-Hofstee analysis of glucose uptake by non-B-cells indicates the existence of a low capacity glucose transport system with a high K , value. This glucose uptake process probably represents a rate-limiting step in glucose metabolism, as it exhibits almost an identical dose-response curve as glucose oxidation.
An increase in extracellular glucose lowers circulating glucagon levels in normal man and animals ( 5 , 13,23); it also decreases glucagon release from perfused pancreas of normal rodents (5,24). It is so far unknown whether this suppressed A-cell function is the result of a direct interaction with glucose or the consequence of glucose-induced changes in the behavior of the adjacent Bor D-cells (3,5 ) . The investigation of glucose handling by A-cells has been hampered by the lack of purified A-cell preparations from normal pancreas. A-cellenriched models have been developed from animals with experimental diabetes but have not clarified the issue (5). Using islets from streptozotocin diabetic guinea pigs, Ostenson (25) demonstrated that glucose metabolism in A-cells increases with the extracellular glucose levels and with the presence of insulin; a reduced glucagon release was only observed after addition of both insulin and glucose. In studies on diabetic ducks and dogs, glucagon suppression by glucose required the presence of insulin both before and during the exposure to high glucose (13,26). On the other hand, Pagliara et al. (24) and Matschinsky et al. (27) reported that insulin was unable to elevate glucose transport in islets from streptozotocin diabetic rats and was, in addition, not necessary for glucose suppression of glucagon release. The major question emerging from the current, sometimes conflicting, information concerns the effect of insulin upon normal A-cells, both in terms of their glucose handling and their hormone release. The availability of A-cell suspensions, prepared from normal rats and almost devoid of insulin-containing B-cells, makes it    damaged the putative insulin receptors and therefore masked the insulin effects. Indeed, fibroblasts prepared by a similar dissociation method were characterized by an insulin-dependent glucose uptake, comparable to that observed in previous reports (28-30); furthermore, a 16-h culture period, which is assumed to restore at least partially the membrane of trypsinexposed cells (31), was not followed by an insulin responsiveness. Our results are therefore compatible with those of Matschinsky et al. (27), where the addition of insulin to islets from diabetic rats did not result in an increased glucose transport (27). It appears unlikely that chronic exposure to insulin is a prerequisite for the cells to respond acutely to the hormone, since insulin failed to affect glucose uptake or oxidation in freshly isolated A-cells.
The rate of glucose uptake by purified single B-cells indicates the existence of a high capacity glucose transport system, which adjusts almost instantaneously the intracellular glucose concentration to that in the extracellular space; these results confirm the conclusions obtained from studies on unpurified islet preparations (32). In addition to the rapid equilibration phase, a second slower and continuous phase was recognized in the glucose transport curves of B-cells; as this component was not observed with the nonmetabolized 3-0-MG, it is associated to the rate of glucose metabolism in single B-cells (33). It should be noted that glucose uptake by single insulin-containing B-cells was measured a t insulin concentrations of 0.1-0.2 pg/ml, which corresponds to the hormone discharge during the preincubation period (19).
The dose-response curve of glucose oxidation by purified single B-cells displays the same sigmoidal shape as in intact islets (34); at the different glucose concentrations tested, CO, production by single B-cells was comparable to that calculated for B-cells which are incorporated into intact islets. The sigmoidal aspect of the curve is therefore a characteristic of glucose handling by individual B-cells, rather than an expression of metabolic coupling between islet cells or a result from paracrine effects of locally released glucagon or somatostatin. This view is also illustrated by the unaltered rate of glucose oxidation after aggregation of single cells or during exposure to glucagon or somatostatin.
In contrast to the similarity in glucose handling between single B-cells and intact islets, single B-cells release 30-fold less insulin in response to glucose than intact islets (19). The poor secretory activity of single B-cells is thus not caused by some metabolic defect in glucose catabolism. These results also indicate that glucose metabolism alone is not sufficient for appropriate regulation by glucose of insulin release. The secretory response is markedly amplified when single B-cells are incubated in the presence of A-cells or glucagon or when they are allowed to reaggregate (35). No alteration in glucose handling appears involved in this amplification, suggesting that the functional coordination between islet cells (19) depends on other factors, possibly messengers such as calcium or CAMP, which are known to participate in the process of insulin release (10,11,36).