Differential Effects of Glucose and Glyburide on Energetics and Na⁺ Levels of βHC9 Cells: Nuclear Magnetic Resonance Spectroscopy and Respirometry Studies

The β-HC9 cell line was obtained from Dr. D. Hanahan. Cells were maintained in Dulbecco’s modified Eagle’s medium with 15% FCS and 24 mmol/l glucose. β-HC9 cells of passages 22–28 were used in this study. Cells were supplemented with 25 mmol/l creatine 48 h before the experiment. The cellular protein content was measured by the method of Lowry et al. (23).

Cells were placed in agarose beads (bead size 800–1,000 μm; 2000–2500 beads/cm³) to maintain them in a stable environment and prevent them from escaping from the superfusion system (24). Briefly, cells were suspended in 1 ml Krebs-Ringer bicarbonate buffer containing: 114 mmol/l NaCl, 5 mmol/l KCl, 24 mmol/l NaHCO₃, 1 mmol/l MgCl₂ × 6H₂O, 25 mmol/l creatine, 2.2 mmol/l CaCl₂, 10 mmol/l HEPES (pH 7.4), and 3 mmol/l glucose. Cells suspended in this buffer were then added to 1 ml 6% agarose (Sigma type VII, low gelling temperature) and mixed at 37°C. This suspension was then decanted into 50 ml paraffin oil (37°C) and stirred continuously. The interaction (difference in surface tension) of the oil and agarose caused bead formation. Beads were cooled to 10°C by adding ice to the water bath under continuous stirring for 5 min to allow them to become firm and to maintain their shape. The beads were irregularly shaped with an average size of 1-mm diameter, as was determined by phase contrast microscopy (Nikon TMS). Beads containing cells were suspended in buffer and washed several times.

Mitochondria from β-HC9 cells were isolated by differential centrifugation according to the generally accepted scheme with modifications providing protection of their native state (25). Briefly, 1–2 × 10⁹ cells were washed twice with a buffer containing 250 mmol/l sucrose, 20 mmol/l HEPES (pH 7.4), 0.5 mmol/l EDTA, and 0.25% BSA and homogenized in the same buffer using a glass Potter-Elvehjem homogenator with a motor-driven Teflon pestle. After initial centrifugation (7 min at 480g), the pellet containing nuclei and other cellular debris was discarded. The supernatant was centrifuged at 10,000g for 15 min to obtain the mitochondrial fraction.

Beads containing cells were placed between two ∼100-μm filters in a flow-through column (Bio Rad). Oxygen partial pressure was recorded polarographically with Clark-type oxygen electrodes placed in inflow and outflow. The superfusion medium was maintained at 37°C using a water bath. The oxygen consumption rate was calculated from the flow rate and the difference in the oxygen partial pressure between the influent and the effluent using the Bunsen absorption coefficient (26). The effluent from the cells was collected for insulin and cAMP measurements. The composition of the perfusion solution was: 114 mmol/l NaCl, 5 mmol/l KCl, 24 mmol/l NaHCO₃, 1 mmol/l MgCl₂ × 6H₂O, 25 mmol/l creatine, 2.2 mmol/l CaCl₂, 0.1 mmol/l IBMX, and 10 mmol/l HEPES (pH 7.4).

Beads containing cells were placed in a 10-mm diameter glass NMR tube and maintained in place by a filter (100 μm pore size) (Fig. 1). Each NMR sample contained ∼1.5 ml of beads suspended in 1 ml buffer (20 mg cell protein). The cells and the capillary tube containing methylene diphosphonate (MDP) and NaCl with disprosium standards were placed within the sensitive volume of the NMR coil. The superfusion medium was equilibrated with 95% O₂ and 5% CO₂ and was maintained at 37°C using a water bath. The temperature in the probe was also maintained by the internal variable temperature controller. The transfer lines were insulated. The buffer was introduced through an in-flow line (at 2.7 ml/min). A suction line, placed above the beads, removed the superfusate. The perfusate was not recirculated. The composition of the perfusion solution was the same as for oxygen consumption measurements.

The ³¹P NMR measurements were performed with a Avance DMX-400 spectrometer at 162 MHz. ³¹P NMR spectra were acquired consecutively in 5-min periods (500 transients) for up to 8 h without proton decoupling (because the resonances are broad and the ³¹P-¹H-couplings are relatively small [3–12 Hz]). The following conditions were used: pulse width 36°, sweep width 13 kHz, 16 K data points, and repetition time 0.6 s. Free induction decays were processed with a Lorentz-Gauss window function for resolution enhancement (Bruker parameters: line broadening [LB] −8, maximum of the gaussian function [GB] 0.003).

Intracellular Na⁺ concentration was determined using thulium-DOTP (3.5 mmol/l) as shift reagent. The following conditions were used: pulse width 90°, sweep width 6.3 kHz, 4 K data points, and repetition time 0.325 s; 3 Hz LB was applied.

Mitochondrial respiratory function was measured by the polarographic method of Chance and Williams (27) using a Clark oxygen electrode. The incubation media contained 250 mmol/l sucrose, 40 mmol/l KCl, 3 mmol/l KH₂PO₄, and 10 mmol/l HEPES buffer (pH 7.4, 27°C). Mitochondrial samples (∼1.8–2 mg protein) were added to 1.5 ml incubation media containing succinate (1 mmol/l), pyruvate (0.5 mmol/l) plus malate (0.5 mmol/l), glutamate (2 mmol/l), or glutamine (2 mmol/l) as a substrate for each measurement.

Parameters measured were as follows: state 2, oxygen consumption before the addition of ADP; state 3, oxygen consumption stimulated by ADP (320 nmol); and state 4, oxygen consumption after completion of ADP phosphorylation. The respiratory control index (ratio of state 3 to state 4) was calculated. The ADP-to-O ratio (ratio between the nanomoles of ADP phosphorylated and the nanomoles of oxygen consumed during state 3) was determined and evaluated.

Mitochondrial protein concentration in the polarographic chamber at the end of each experiment was measured by the method of Lowry et al. (23), and results were used to normalize respiratory rates to mitochondrial protein.

Insulin content in efflux samples was measured by radioimmunoassay with charcoal separation (28). Rat insulin from Linco Research served as standard, and Miles anti-insulin antibody from ICN was the primary antibody. The c-AMP assay was performed according to the radioimmunoassay method of Albano and Barnes (29) using the Biotrak cAMP[¹²⁵I] assay system from Amersham Pharmacia Biotech.

Insulin and cAMP release, sodium levels, and P-creatine (PCr)–to–inorganic phosphate (P_i) ratios are presented as the mean ± SE of four to five experiments. Oxygen consumption curves are presented as the mean of the four to five experiments without SE due to the high frequency of sampling (1-s interval). In appropriate cases, significant differences between groups were determined by ANOVA with post hoc analysis using Dunnett’s multiple comparison test. P ≤ 0.05 was considered significant.

To assess the effects of GLY on oxygen consumption and insulin release in β-HC9 cells, GLY concentration was increased progressively from 0 to 2 μmol/l using a 50-nmol/l increment per min. GLY at 250–300 nmol/l had a maximal effect on insulin release (Fig. 2A) and oxygen consumption (Fig. 2D). Verapamil, a Ca²⁺ channel blocker, completely abolished the GLY effect on insulin release (Fig. 2C), implying a critical role for voltage-dependent Ca²⁺ channels in the drug’s action. Removal of both verapamil and GLY restored insulin release to a rate approximately fourfold higher than without GLY pretreatment (Fig. 2C). The latter effect may be related to the continued inhibition of the channel by GLY due to it’s tight binding to the SURs. Verapamil alone causes a transient increase in insulin release (Fig. 2B) and did not significantly affect oxygen consumption (Fig. 2D, line 2). Surprisingly, GLY in the presence of verapamil caused a dose-dependent decrease in oxygen consumption (Fig. 2D, line 3) in contrast to the effect of GLY alone (Fig. 2D, line 1). These contrasting results may indicate that GLY has dual effects on oxidative metabolism.

In the second set of experiments, we compared the effect of high glucose (30 mmol/l) and high GLY (1 μmol/l) on oxygen consumption and insulin release from perifused β-HC9 cells (Fig. 3A and B). According to Hu et al. (7), this concentration causes maximum inhibition of K_ATP currents in rat β-cells. However, other investigators reported 100% inhibition of K_ATP channels at concentrations <100 nmol/l (30). As a functional test of fuel responsiveness, the glucose was removed for 30 min, and then reintroduced at 5 mmol/l for 30 min. Subsequently, cells were exposed to either 30 mmol/l glucose or 1 μmol/l GLY plus 5 mmol/l glucose. Whereas high glucose caused a greater increase in oxygen consumption than GLY (Fig. 3A) (30 and 15%, respectively), GLY stimulated insulin release from β-cells to a higher extent than high glucose (Fig. 3B) (25- and 15-fold, respectively).

Figure 4 presents selected ³¹P-NMR spectra of superfused β-HC9 cells with low (3 or 5 mmol/l) and high (30 mmol/l) glucose and in the presence of 1 μmol/l GLY. Each spectrum clearly displays the resonances of P_i, PCr, and ATP (γ, α, and β) peaks. Results of continuous monitoring (every 5 min) of phosphorus metabolites are presented in Fig. 5A and B. Increasing the glucose concentration in perfusate from 3 to 30 mmol/l leads to a marked increase in phosphorylation potential, as evidenced by increased PCr and decreased P_i peak areas. The ATP level as indicated by the β-peak of the nucleotide remains constant. In contrast, addition of 1 μmol/l GLY to the cells superfused with 5 mmol/l glucose leads to decreased PCr and increased P_i levels without significant changes in the ATP peaks. GLY decreased the PCr-to-P_i ratio to 58% (P < 0.01) of basal level at 5 mmol/l glucose (Fig. 5C). After removal of either secretagogue, the ATP level in glucose pretreated cells remained unchanged in contrast to a decrease in ATP in the cells pretreated with GLY. The PCr decreased and P_i increased in both conditions. The addition of 30 mmol/l glucose restored the phosphorylation potential in both cases.

To better understand the mechanism of GLY action on ATP production, we extended our study by measuring the respiration and oxidative phosphorylation of mitochondria isolated from β-HC9 cells. Figure 6 presents mitochondrial oxygen consumption data in metabolic states 2, 3, and 4 during oxidation of glutamine (2 mmol/l) (Fig. 6A), succinate (1 mmol/l) (Fig. 6B), glutamate (2 mmol/l) (Fig. 6C), and pyruvate (0.5 mmol/l) plus malate (0.5 mmol/l) (Fig. 6D). The addition of 1 μmol/l GLY to mitochondria, oxidizing glutamine or glutamate, completely abolished the stimulation of respiration by ADP (state 3). However, GLY did not affect state 3 respiration when succinate or pyruvate plus malate were used as substrates. The study suggests that GLY selectively affects the oxidation of glutamine and glutamate as evidenced by absence of ADP-stimulated respiration (state 3), meanwhile leaving oxidation of other substrates unchanged.

The depolarization of the β-cell by GLY causes Ca²⁺ to enter the cell through voltage-dependent calcium channels. An increase in cytosolic-free calcium levels triggers the release of insulin (31–33). The depolarization of the cells may also cause Na⁺ to enter the cell. Increased cytosolic sodium may contribute to an increase in cytosolic-free Ca²⁺ via a plasma membrane sodium-calcium exchanger (34). Thus, Na⁺ may be an important regulator of calcium metabolism in β-cells. We applied ²³Na-NMR spectroscopy to monitor the changes in intracellular Na⁺ in β-HC9 cells superfused with either high glucose or GLY. Figure 7 shows the continuous changes in intracellular sodium during perfusion of the cells with different concentrations of glucose and with 1 μmol/l GLY during a representative experiment. Exposure to 5 mmol/l glucose decreased Na⁺ to 75% (P < 0.01) of basal level; 30 mmol/l glucose resulted in a further decrease in cytosolic Na⁺ to 60% (P < 0.01) of basal level. In contrast, Na⁺ increased to 116% (P < 0.05) when 1 μmol/l GLY was added to the perfusate containing 5 mmol/l glucose. Removal of both GLY and glucose caused a further increase of Na⁺ up to 140% (P < 0.01).

Figure 8A presents results of the continuous measurement of cAMP release from β-HC9 cells during glucose and GLY-stimulated insulin release. The pattern of cAMP release is similar to changes in insulin release and oxygen consumption in β-HC9. Both high glucose and GLY markedly enhance cAMP release in β-HC9 cells. To evaluate whether the GLY-induced changes in cAMP are related to cellular Ca²⁺ entry, we omitted Ca²⁺ from the perfusion solution containing 5 mmol/l glucose. Thirty minutes later we added 1 μmol/l GLY (Fig. 8B). Removal of Ca²⁺ from the perfusion solution initiated the appearance of two discrete spikes in both insulin and cAMP release, which may be due to Ca²⁺ release from intracellular depots. This was then followed by a decrease in insulin secretion, and the further addition of GLY did not stimulate insulin or cAMP release in Ca²⁺-free solution. However, reintroduction of Ca²⁺ led to a burst of insulin and cAMP release. After removal of GLY from the perfusate, insulin and cAMP release remained elevated, consistent with the tight binding of GLY to the SURs. Uncoupling of respiration and oxidative phosphorylation by FCCP led to decreases in both insulin and cAMP release.

Both high glucose and SUs increase oxygen consumption of β-HC9 cells. However, in the case of high glucose, an increase in oxygen consumption is associated with a large increase in the PCr-to-P_i ratio (an indication of a increased P-potential of the β-cells), which is the result of “substrate pressure” and increased production of NADH by glycolysis and Krebs cycle. GLY lowers the PCr-to-P_i ratio, contrasting with the observation after high glucose, while ATP levels in both conditions remain unchanged. The results suggest that GLY, in contrast to glucose, causes an energy deficit. These data complement recent findings of others (35) who have shown that GLY decreased the islet ATP and ADP content as well as the ATP-to-ADP ratio at 0 mmol/l glucose. However, these effects were no longer observed at 10 mmol/l glucose.

It seems likely that inhibition of K_ATP during glucose-stimulated insulin release is due to changes in free ADP concentration (36). The effective level of ATP that inhibits the channel was found to vary in different studies, with half-maximal inhibition occurring at concentrations between 15 and 200 μmol/l (37,38). Because free ATP concentrations in the β-cell are in the millimole per liter range, two explanations have been advanced to explain why K_ATPs are not closed at the ambient concentrations of ATP found in the β-cell. ADP³⁻ may compete with ATP for binding to the channels and increases the K_i of ATP, or it may be an activator per se in the form of MgADP⁻. The latter explanation is more plausible. Thus, ADP may partially reverse the inhibition of the channel by ATP. It has been shown recently that phosphatidylinositol (PI) phospholipids, like PI(4,5)P₂, antagonize nucleotide inhibition of K_ATP channels, enhancing the coupling of metabolic events to cell electrical or transport activity (39).

GLY appears to stimulate respiration by mechanisms that are distinct from those that underlie the effects of high glucose. Applying digital Ca²⁺ imaging on single β-TC3 cells, it has been shown that repaglinide and GLY induce a concentration-dependent increase in intracellular free Ca²⁺ concentration (40). The increase in [Ca²⁺]_i results from Ca²⁺ entry through voltage-dependent l-type Ca²⁺ channels because it is inhibited by verapamil. Based on studies of the heart and other tissues, it has been suggested that increases in cytosolic Ca²⁺ concentration may lead to an increase in intramitochondrial free Ca²⁺ and activation of several matrix dehydrogenases and increased turnover of the Krebs cycle (41). However, increased turnover of the Krebs cycle would lead to increased ATP synthesis, which is in contrast to the drop in ATP level seen in the GLY-stimulated β-cells. Panten et al. (42) suggested that the driving force for tolbutamide-induced oxygen uptake is a decrease in the phosphorylation potential caused by the workload imposed by stimulation of the secretion process.

Our experiments with verapamil, a calcium channel blocker, indicate that in addition to having a stimulatory effect on oxygen consumption, GLY may inhibit oxidative metabolism as seen when the voltage-dependent l-type Ca²⁺-channels are inhibited. To directly explore the mechanisms by which GLY affects oxidative phosphorylation, we studied mitochondria isolated from β-cells. It has been shown that GLY enters the β-cell and is associated with secretory granules and accumulates in concentrations (43) far in excess of those needed to occupy the SURs (44). Mitochondria might thus be a target of SU action. GLY was added to isolated mitochondria oxidizing NAD- and FAD-dependent substrates. Although a high concentration of GLY was used in this study, we found that GLY selectively affects the oxidation of glutamine and glutamate. In contrast, ADP was able to stimulate the oxidation of succinate and pyruvate plus malate in the presence of GLY, which is indicative of a viable state 3 respiration in the mitochondria.

The results also suggest that GLY does not have a direct uncoupling effect on respiration and oxidative phosphorylation (i.e., the P-to-O ratio is not changed during oxidation of pyruvate plus malate or succinate). However, we cannot exclude the possibility that in vivo GLY may be able to induce Ca²⁺ overload in cells and may lead to uncoupling of respiration and oxidative phosphorylation.

It has been demonstrated with the patch clamp technique that pancreatic β-cells are equipped with potential-dependent Na⁺ channels (22,45–47) and that glucose promotes Na⁺ entry via such channels. This effect was counteracted by both the hyperpolarizing agent diazoxide and the Na⁺ channel blocker tetrodotoxin. It also has been shown that exposure of β-cell–rich pancreatic islets from ob/ob mice to glucose resulted in a protracted lowering of the sodium content, as measured by integrating flame photometry (48,49). The maximum effect was seen at 5 mmol/l glucose, with no additional reduction when glucose was increased to 20 mmol/l (49). There was no indication that glucose-induced lowering of sodium reflected the activation of the ouabain-sensitive Na/K pump when measuring the uptake of 86Rb⁺ (48). However, other authors (35) demonstrated that d-glucose stimulates the Na⁺/K⁺ pump and further suggested that this may be part of the membrane repolarization process, following the primary depolarization by these agents. Wesslen and Hellman (1988) (48) suggested that glucose facilitates the removal of Ca²⁺ from cytoplasm by increasing the transmembrane Na⁺ gradient (48). The radioisotope studies of Kawazu et al. (50) have shown that glucose stimulates both entry and the extrusion of Na⁺ from the β-cells. Because GLY depolarizes the β-cells, it was of interest to evaluate the effect of the drug on Na⁺ transport. NMR technology allowed us to continuously monitor the changes in intracellular free Na⁺ concentration of perifused agarose embedded β-HC9 cells every 2.3 min. Results of our study show that GLY and high glucose had opposing effects on intracellular Na⁺ levels in β-cells. Both low and high glucose lower the Na⁺ level in β-HC9 cells, whereas GLY increases the intracellular ion concentration. The effect of GLY is more pronounced in glucose-free solution.

Results of the present study also indicate that both high glucose and GLY elevate cAMP (incidentally, the demonstration of this effect requires the presence of 0.1 mmol/l IBMX). These effects may be due to elevation of intracellular Ca²⁺, which leads to activation of adenylcyclase. Our results are in agreement with data showing that glucose, in addition to other agents, stimulates insulin secretion through cAMP–dependent mechanisms (20).

These data support the hypothesis that glucose activates β-cells through a “push mechanism” due to substrate pressure, enhancing fuel flux, energy production, and extrusion of Na⁺ from the cells in contrast to SUR-1 inhibitors, which may modify intermediary and energy metabolism secondarily through a “pull mechanism” due to higher energy demand resulting from increased ion fluxes and the exocytotic work load.

This work was supported in part by Grants NIDDK 22122 and 19525, a grant from the American Diabetes Association (to N.M.D.), and a grant in aid from Novo Nordisk.