The Existence of an Optimal Range of Cytosolic Free Calcium for Insulin-stimulated Glucose Transport in Rat Adipocytes*

We have examined the effects of extracellular and intracellular ea2+ concentrations upon basal and insu-lin-stimulated 2-deoxyglucose uptake in isolated rat adipocytes. In the absence of extracellular Ca2+, both basal and insulin-stimulated glucose uptake were sig- nificantly reduced. Insulin-stimulated glucose transport was optimal at 1 and 2 mM Ca2+. Further increases in extracellular ca2+ concentration (3 mM) signifi- cantly diminished insulin-stimulated glucose uptake. When intracellular ea2+ concentrations were aug- mented by ionomycin (1 p ~ ) , insulin-stimulated glucose uptake was significantly reduced at extracellular Ca2+ concentrations of 2 and 3 mM. The levels of intracellular free ea2+ concentrations were then measured with ea2+ indicator fura-2. The correlation between the levels of intracellular free Ca2+ and the magnitude of insulin-stimulated glucose uptake revealed that the optimal effect of insulin is observed at ea2+ levels between 140 and 370 nM. At both extremes outside of this window, both low and high levels of intracellular ea2+ result in diminished cellular responsiveness to insulin. These data suggest that intracellular calcium concen- trations may exert a dual role in the regulation of cellular sensitivity to insulin. First, there must exist a minimal concentration of intracellular calcium to pro-mote insulin action. Second, increased levels of intra- cellular calcium may provide a critical signal for diminution of insulin action.

thermore, our findings of increased [Ca"]i in adipocytes isolated from obese and hyperinsulinemic subjects (9) indicated that intracellular Ca2+ may be a critical factor in postreceptor step modulation of cellular responsiveness to insulin. In this study, we assessed the effects of various concentrations of extracellular and intracellular Ca2+ upon basal and insulinstimulated glucose transport in isolated rat adipocytes.
Preparation of Isolated Adipocytes-Male Sprague-Dawley rats weighing 200-250 g were used in these studies. Isolated adipocytes were prepared from epididymal fat pads according to the method of Rodbell (12). Animals were allowed food and water ad libitum before they were killed.
2-Deoxyglucose Transport-Adipocytes (2 X lo5 cells) were incubated in the absence and in the presence of increasing concentrations of insulin for 60 min at 37 "C. Concentrations of extracellular Ca2+ ranged between 0 and 3 mM as indicated below. Glucose uptake was initiated by addition of [3H]2-deoxyglucose (0.2 pCi). After 3 min of incubation, the reaction was terminated by transferring 200-pl aliquots of incubation mixture to the microfuge tubes (containing 100 pl of silicone oil) and centrifuging the tubes in a Beckman Microfuge. The cell pellets were counted for radioactivity present in a Beckman liquid scintillation counter. The results were corrected for diffusion by subtracting the uptake of L-glucose in the absence of insulin.
Measurements of Intracellular Calcium-The fluorescence of control and fura-2 loaded cells was measured using a Turner Model 340 spectrofluorometer as previously described (8). During Ca2+ measurements, the cells were incubated in 2.4 ml of Krebs-Hepes' buffer containing 118.4 mM NaC1, 4.69 mM KCl, 1.2 mM MgCl,, 1.18 mM KHzPO,, 1.25 mM NaHC03, 20 mM Hepes, 5 mg/ml bovine serum albumin, and 30 mg/dl glucose at pH 7.4. Extracellular Ca2+ concentration ranged from 0 to 3 mM as indicated below. The final cell concentration was approximately 2 X lo5 cells/ml. Tissue or buffer fluorescence in the absence of fura-2 did not change in response to either insulin, verapamil, or ionomycin. Similarly, neither addition of fat emulsion nor rat fat droplets to the fura-2. CaZ+.complex changed the fluorescent signal. The measurements were also performed by the dual wavelength technique (8, 13) at the excitation wave lengths 347 and 387 nm. Although the relative changes in [Ca2+Ii were the same, the absolute values were approximately 2.5 times lower, probably as a result of tissue viscosity interference with the measurements (14, 15).
Statistics-The results of these studies are presented as mean & S.E. of three to five experiments and compared using either paired or unpaired t test as indicated.

RESULTS
In initial experiments, we have assessed the influence of extracellular concentrations of Ca'+ on basal and insulinstimulated 2-deoxyglucose uptake in isolated adipocytes (Fig.  1). In these experiments, the cells were incubated without insulin (termed here, basal transport), with two submaximally effective concentrations of insulin (0.3 and 1 ng/ml) and with a maximally effective insulin concentration (25 ng/ml). The studies were performed a t varying levels of Ca2+ in the extracellular fluid (0, 0.01, 0.1, 1, 2, and 3 mM). The basal rate of glucose transport was reduced in the absence of extracellular ca2+ ( p < 0.05 uersus 0.01 mM, and <0.01 uersus other Ca2+ The abbreviation used is: Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid.

FIG. 2. The effect of intracellular CaZ+ concentrations on basal and insulin-stimulated
2-deoxyglucose uptake. The cells were exposed to 1 pM ionomycin for 10 min before the measurements of glucose uptake were initiated. The results represent the mean values of three or four separate experiments and are expressed as in Fig. 1. Insulin (ng/ml) concentrations). Regardless of the Ca2+ concentration in the extracellular fluid (when it was present), the basal rate of glucose transport was practically identical. The response to insulin stimulation was minimal in the absence of extracellular Ca2+ and optimal at 1 and 2 mM Ca". However, an increase of the extracellular Ca2+ concentration to 3 mM resulted in a significant decrease in insulin-stimulated glucose transport ( p < 0.01 uersus 1 and 2 mM at all insulin concentrations). This is particularly evident when the results of the experiments are expressed as the percent of stimulation above the basal rate of glucose transport (Fig.  1B). As can be seen, insulin-stimulated glucose transport was diminished under two conditions: 0 and 3 mM extracellular Ca2+.
T o examine the effect of intracellular Ca2+ Concentrations upon basal and insulin-stimulated glucose transport, we have performed similar experiments in the presence of the calcium ionophore, ionomycin. Ionomycin (1 FM) was added to the incubation mixture 10 min prior to initiation of glucose uptake measurements. In the absence of extracellular Ca2+, both basal and insulin-stimulated 2-deoxyglucose uptake were lower than in the presence of extracellular Ca2+ ( p < 0.01 at all points) and not significantly different from values observed in the absence of ionomycin (Fig. 2).
In the presence of extracellular Ca2+, ionomycin did not alter the rate of basal glucose transport, which was identical to that seen without ionomycin. Similarly, the rate of basal glucose transport was not significantly affected by various concentrations of extracellular Ca2+. In these experiments, the optimal response of glucose transport to insulin was seen at a ca2+ concentration of 1 mM ( p < 0.01 uersus 0.01, 2, and 3 mM ca2+ at all insulin concentrations). In the presence of ionomycin and either 2 or 3 mM extracellular Ca2+, insulinstimulated glucose transport was significantly diminished as compared with lower Ca2+ concentrations. Again, when the results were expressed as a percent above the basal rate, the inhibitory influence of high Ca2+ concentrations in the presence of ionomycin was particularly apparent (Fig. 2B). Interestingly, in the presence of ionomycin, 2 m~ extracellular Ca2+ now exerted an inhibitory effect on glucose transport ( p The cells were exposed to verapamil for 30 min before the glucose transport studies were initiated. Extracellular calcium concentration was 1 mM. The results are expressed either in absolute values (nmol/ 3 min/250,000 cells) ( A ) or as a percent above the basal rate of uptake ( B ) . The mean f S.E. of three experiments are shown.

2-Deoxyglucose uptake in the presence of 1 mM (A) or 0.1 mM (B) extracellular ea2+ and either ionomycin or verapamil
Results represent mean f S.E. of three experiments (nmol/3 min). < 0.01 versus 2 mM ca2+ in the absence of ionomycin at insulin concentrations 1 and 25 ng/ml), whereas 3 mM Ca2+ was inhibitory both alone and in the presence of the ionophore.
To further assess the influence of intracellular Ca2+ concentration upon cellular responsiveness to insulin, we studied basal and insulin-stimulated glucose transport in the presence of the Ca2+ channel blocker, verapamil (Fig. 3). The cells were preincubated with 30 ~L M verapamil for 30 min before the glucose transport studies were initiated. The basal rate of glucose transport was not affected by the presence of verapamil. However, verapamil significantly inhibited insulinstimulated glucose transport at all insulin concentrations tested. The influence of verapamil is amplified by expressing the results as the percent stimulation above the basal rate of glucose transport (Fig. 3B). The comparison of the effects of ionomycin and verapamil upon insulin-stimulated glucose uptake a t two concentrations of extracellular Ca2+ (0.1 and 1 mM) is shown in Table I. Ionomycin appears to shift the insulin response curve to the left (significant increase, p c 0.05, was observed only with 0.1 mM Ca2+), whereas verapamil clearly shifts it to the right.
In several experiments described above, parallel batches of cells were used to determine cytosolic free calcium concentrations, [Ca2+Ii. We then compared the magnitude of maximally stimulated glucose transport observed at different levels of [Ca2+], in these cells (Fig. 4). A waveform curve was obtained. The window of optimal cellular response to insulin lies with  (11, 16). In adipocytes, however, we were able to observe a dose-dependent effect of insulin and glyburide upon [Ca'+], (8,9). The effect of insulin was rapid, seen within the first 2-4 min of exposure and reaching a maximum a t 10-14 min.
Although insulin enhances both glucose transport and [Ca2+Ii, the direct relationship between these two variables has not been established. The effect of insulin upon [Ca2+Ii seems somewhat delayed when compared with the insulin effect on glucose transport. However, glucose transport meas-urements are usually performed after 30-60 min of cell exposure to insulin. It is conceivable that optimal levels of intracellular Ca2+ are achieved during this preincubation period, which then results in adequate glucose transport when the glucose tracer is added to the cells. As is evident from this and previous (17)(18)(19) communications, insulin-stimulated glucose transport is diminished in the absence of extracellular Ca2+, strongly supporting the possibility that there exists a tight relationship between insulin-stimulated increases in Ca2+ influx and glucose transport. On the other hand, we have also demonstrated (8) that ambient glucose concentrations modulate the effect of insulin on [Ca2+Ji.
Regardless of whether or not intracellular Ca2+ plays a role in initiating and/or mediating insulin action, our recent (9) and present observations strongly suggest that high [Ca2+Ii may be a mechanism for deactivation or possibly termination of insulin action. In vivo insulin and glucose infusion (euglycemic clamp) in normal volunteers resulted in higher levels of [Ca2+Ii in isolated adipocytes and in the loss of cellular responsiveness to insulin or glyburide. However, adipocytes isolated from obese subjects demonstrated high levels of intracellular Ca2+ and were also unresponsive to subsequent stimulation with insulin or glyburide (9).
Similar induction of insulin resistance was demonstrated by Garvey et al. (20) and Mandarino et al. (21). Garvey and his colleagues (20) have produced insulin resistance by incubating rat adipocytes with high concentrations of insulin and glucose for 24 h. Neither glucose nor insulin alone were active in this regard. Mandarino et al. (21) observed a significant reduction in insulin-stimulated glucose transport in adipocytes isolated from normal subjects at the 6th hour of insulin and glucose infusion (euglycemic clamp). In the light of our observations, it is possible that in both cases, exposure of adipocytes (either in vivo or in uitro) to high concentrations of insulin and glucose might have resulted in increased [Ca2+Ii which, in turn, diminished cellular responsiveness to insulin.
The mechanism whereby elevated [Ca2+Ii induced cellular insensitivity to insulin is unknown. It is intriguing to postulate that elevated levels of intracellular Ca2+ provide a feedback signal for termination of insulin action. It is conceivable that Ca2+, either itself or via activation of protein kinase C, may inhibit various processes involved in mediating insulin action. Another possibility is that intact Ca2+ fluxes, in and out of cells, are required to maintain an optimal rate of glucose transport. In recent studies with chromaffin cells, Artalejo et al. (22) demonstrated that intracellular concentrations of Ca2+ modulate the rate of Caz+ influx. In their studies, increased concentrations of intracellular Ca2+ deactivated Ca2+ channels and reduced the rate of Ca2+ influx. Furthermore, chelation of intracellular Ca2+ with quin-2 prevented deactivation of Ca2+ entry, whereas enhancement of the [Ca2+Ii with the ionophore (A23187) rapidly increased the rate of deactivation of Ca2+ influx. If the maintenance of Ca2+ influx is indeed required to optimize glucose transport, then deactivation of this process by high and/or sustained intracellular concentrations of Ca2+ may result in diminution of insulin-stimulated hexose transport.
In conclusion, there appears to be an optimal level of [Ca2+Ii for insulin-stimulated glucose transport. One must realize, however, that our method measures total intracellular Ca2+ concentration and does not take into account the possible locclized effects of Ca2+ or changes in Ca2+ concentration within discrete cellular compartments. Nevertheless, our data suggest that high and/or sustained levels of intracellular Ca2+ may function as a postreceptor feedback sensor to diminish cellular responsiveness to insulin. In obesity or in normal subjects receiving glucose/insulin infusions, increasing intracellular Ca2+ may result in overt insulin resistance (9). Further investigations are necessary to elucidate the role of Ca2+ as a physiological terminator of insulin action.