INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

IN ENDOTHELIAL CELLS, as in other cells, cytosolic free Ca²⁺ concentration ([Ca²⁺]_i) is controlled within narrow limits. On stimulation of endothelial cells by mediators (e.g., thrombin and histamine) (4, 9, 14) or by mechanical forces (e.g., shear stress) (1, 21), endothelial cells respond with an increase in [Ca²⁺]_i due to an activation of the D-myo-inositol 3-phosphate [Ins(3)P]-sensitive Ca²⁺ release mechanism of the endoplasmic reticulum (ER). The elevated state of [Ca²⁺]_i triggers a variety of cellular functions (5, 7). The return to control of [Ca²⁺]_i is mediated by activation of ATP-dependent Ca²⁺ transport mechanisms of the ER and the plasma membrane.

An increase in [Ca²⁺]_i may also occur under pathophysiological conditions, e.g., during hypoxia or ischemia, when endothelial cells start developing an energy deficit. The relationship between energy loss and rise of [Ca²⁺]_i differs distinctly from that found in many other cell types. In cardiomyocytes, for example, [Ca²⁺]_i rises only after extensive loss of high-energy phosphates (16). In former studies (19, 20), we found that in metabolically disturbed endothelial cells, changes in [Ca²⁺]_i and permeability occur much earlier in time. Endothelial cells respond rapidly to inhibition of energy production with a biphasic increase in [Ca²⁺]_i. An initial, sudden rise occurs when less than 30% of ATP reserves are lost (20). This rise is due to the release of Ca²⁺ from an endogenous store, the ER. Thereafter, a slow progressive increase of [Ca²⁺]_i takes place because of an influx of Ca²⁺ from the extracellular space. The increase in [Ca²⁺]_i can trigger a variety of intracellular signal transduction cascades, including those regulating the endothelial barrier function. The increase in [Ca²⁺]_i and subsequent metabolic and functional changes were found to be reversible during the initial and the early part of the second phase of cytosolic Ca²⁺ overload (20). The responsiveness of endothelial Ca²⁺ homeostasis to moderate changes in the energy state may have important consequences for endothelial function in vivo, because endothelial cells can rapidly lose small amounts of ATP, e.g., in hypoxia (2, 26, 32) or under oxidative stress (13, 28).

The mechanism by which metabolic inhibition evokes the initial sudden Ca²⁺ release from ER in endothelial cells is only partly understood. In the present study, we investigated whether inhibition of cytoplasmic production of ATP through glycolysis was equally effective as an inhibitor of mitochondrial ATP production. We also studied whether the initial rise of [Ca²⁺]_i upon metabolic inhibition is due to a reduction in Ca²⁺ sequestration into or an acceleration of Ca²⁺ release from the ER. Finally, we tried to prevent the [Ca²⁺]_i rise in metabolically impaired endothelial cells.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cell culture and endothelium in situ. Endothelial cells from porcine aorta were isolated and cultured as previously described (24). Confluent cultures of primary endothelial cells were trypsinized in PBS composed of (in mM) 137 NaCl, 2.7 KCl, 1.5 KH₂PO₄, and 8.0 Na₂HPO₄, at pH 7.4, supplemented with 0.05% (wt/vol) trypsin and 0.02% (wt/vol) EDTA, and seeded at a density of 7 × 10⁴ cells/cm² on glass coverslips or 30-mm culture dishes (Falcon-type 3001) for determination of nucleotide contents, respectively. Experiments were performed with confluent monolayers, 4 days after seeding. One additional set of experiments was carried out on endothelial cells shortly after their isolation. For this purpose, the newly isolated cells were seeded directly on the coverslip and allowed to attach during a 24-h period before use. The same set of experiments was also performed using fresh endothelium in situ on the aortic vessel wall. For this purpose, disks of 8-mm diameter were cut from porcine aorta, washed, and stored in cell culture medium 199 before use.

Determination of [Ca²⁺]_i. [Ca²⁺]_i was determined using the fluorescent Ca²⁺ indicator, fura 2. Endothelial cells cultured on round glass coverslips (diameter 25 mm) were incubated in medium 199 supplemented with 5% (vol/vol) newborn calf serum (NCS) (heat-inactivated for 10 min at 60°C) and the addition of 2.5 µM fura 2-acetoxymethyl ester (AM) at 20°C in the dark. After a 45-min incubation period, the extracellular fura 2-AM was removed by a washing step and a medium change with NCS (2% vol/vol, heat-inactivated)-supplemented HEPES buffer (composition as described in Experimental protocols). The coverslips were then mounted in a temperature-controlled incubation chamber adapted to the fluorescence microscope (IX 70, Olympus; Hamburg, Germany). To allow hydrolysis of the AM, cells were incubated for 20 min at 35°C in the same medium before measurements were started. Loading protocol of endothelium on aortic disks was the same, except that aortic disks were incubated for 60 min with fura 2-AM at a final concentration of 5 µM.

Determination of Mn²⁺ influx. The influx of divalent cations via the endothelial plasma membrane was determined by Mn²⁺-induced quenching of fura 2 fluorescence. Fluorescence of fura 2-loaded cells was measured at the isosbestic wavelength of 360 nm (an excitation wavelength at which the fura 2 fluorescence is independent of cytosolic free [Ca²⁺]) in the presence of 1 mM Ca²⁺ with or without 100 µM MnCl₂ in the incubation medium. Emitted light was detected at 510 nm and background corrected. To characterize the Mn²⁺-induced quench, the ratio between the slopes of the decline of fura 2 fluorescence after and before the addition of MnCl₂ was determined (quench ratio). Mn²⁺ influx occurs when this quench ratio rises to values above 1.

Cellular ATP contents. Confluent endothelial monolayers cultured on 30-mm culture dishes were incubated under identical conditions as described for [Ca²⁺]_i measurements. After removal of media, incubations were terminated by the addition of ice-cold perchloric acid (0.6 N). The ATP contents of neutralized extracts were determined by high-performance liquid chromatography according to Jüngling and Kammermeier (15). ATP contents (nM ATP/mg of cellular protein) were determined and were expressed as percentage of nonstimulated control dishes incubated for the same time. Protein contents were measured according to Bradford (3), with use of bovine serum albumin as standard.

Experimental protocols. To allow a comparison between Ca²⁺ development and the corresponding ATP depletion of endothelial monolayers, cells were treated identically in incubation. During the experiments, endothelial cells were incubated at 35°C in a HEPES-buffered solution composed of (in mM) 25 HEPES, 125 NaCl, 1.0 CaCl₂, 2.6 KCl, 1.2 MgCl₂, 1.2 KH₂PO₄, and pH 7.4 and supplemented with 2% (vol/vol) heat-inactivated NCS. To impair glycolytic or mitochondrial energy production, 2-deoxy-D-glucose (2-DG, 2.5-10 mM) or NaCN (1-5 mM) was added to media, respectively. In part of the experiments with 2-DG, pyruvate (5 mM) was added; in another part of the experiments with NaCN, glucose (10 mM) was added. Other agents were applied as indicated. Stock solutions of ionomycin, thapsigargin (THG), and xestospongin C (XeC) were prepared with dimethyl sulfoxide. XeC was applied 20 min before metabolic inhibition. Appropriate volumes of stock solutions were added to the cells yielding final solvent concentrations <0.1% (vol/vol). The same final concentration of dimethyl sulfoxide was also included in all respective control experiments. Stock solutions of all other substances were prepared in the specified HEPES-buffered solution. Appropriate volumes of these solutions were added to the cells. Identical additions were included in all respective control experiments.

Materials. Falcon plastic tissue culture dishes were from Becton-Dickinson (Heidelberg, Germany); fura 2-AM was from Molecular Probes (Eugene, OR); pyruvate and ATP were from Boehringer (Mannheim, Germany); ionomycin, ryanodine, THG, and XeC were from Calbiochem (Bad Soden, Germany); NaCN and glucose were from Merck (Darmstadt, Germany); 2-DG was from Sigma-Aldrich (Deisenhofen, Germany); NCS, medium 199, penicillin-streptomycin, and trypsin-EDTA were from GIBCO Life Technologies (Eggenstein, Germany). All other chemicals were of the best available quality, usually analytic grade.

Statistical analysis. Values are expressed as means ± SE of cells taken from at least three experiments using independent monolayer preparations. Statistical analysis was performed by one-way ANOVA in conjunction with the Student-Newman-Keuls test for post hoc analysis. Between-group analysis was performed, and P values <0.05 were considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Effect of 2-DG on [Ca²⁺]_i. Addition of the glycolysis inhibitor 2-DG to endothelial cells elicited a biphasic rise of cytosolic Ca²⁺ (Fig. 1). [Ca²⁺]_i peaked at 1 min, declined afterward, and started to rise slowly again after 3 min. This 2-DG-induced rise of [Ca²⁺]_i was dose dependent between 2.5 and 10 mM 2-DG. Higher concentrations of 2-DG did not augment [Ca²⁺]_i any further (data not shown). All experiments following this study were performed with 2-DG at 10 mM.

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Dose-dependent effect of 2-deoxy-D-glucose (2-DG; ) on cytosolic Ca2+ concentration ([Ca2+]i) of aortic endothelial cells. Control represents unstimulated cells in buffer solution. Data are means ± SE of n = 60 cells. *P < 0.05 vs. control.

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Experiment on the effect of 2-DG on [Ca2+]i and Mn2+-induced quenching of fura 2 fluorescence. Top: original recording of [Ca2+]i from single endothelial cell exposed to 10 mM 2-DG in buffer solution. Bottom: fura 2 fluorescence at 360-nm excitation, expressed in percentage of value at time 0. Mn2+ (100 µM) is added at time 0 (arrow), simultaneously with 10 mM 2-DG.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Effect of pyruvate (Pyr) on 2-DG-evoked rise of [Ca2+]i. Endothelial cells were preincubated in the presence () or absence () of 5 mM Pyr for 20 min, and then 10 mM 2-DG was added. Data are means ± SE of n = 60 cells. *P < 0.05 vs. control in the absence of Pyr. #P < 0.05 vs. 2-DG alone.

Effect of NaCN on [Ca²⁺]_i. Addition of the mitochondrial inhibitor cyanide also caused a rapid and dose-dependent increase in [Ca²⁺]_i at concentrations between 1 and 5 mM (Fig. 4). Higher concentrations of NaCN did not augment [Ca²⁺]_i any further (data not shown). By using the pH indicator dye BCECF, we tested whether NaCN (5 mM) alters cytosolic pH of the endothelial cells during the first 10 min after the addition of NaCN. This was not the case (data not shown) and in agreement with previous reports of others (30). As observed in the case of 2-DG, NaCN caused a biphasic rise in [Ca²⁺]_i. It peaked 1 min after the addition of cyanide showed a transient decline, and started to rise slowly during the ongoing incubation period. Compared with the effect of 2-DG, however, the maximum of the initial rise of [Ca²⁺]_i was smaller [145 ± 7 (NaCN 5 mM, n = 60) vs. 189 ± 7 (2-DG 10 mM, n = 60); P < 0.05]. As in the case of 2-DG, it was analyzed whether NaCN stimulates influx of divalent cations through the plasma membrane. The simultaneous addition of MnCl₂ (100 µM) with NaCN (5 mM) did not alter the decline of the 360-nm fluorescence of fura 2 during the first 6 min, but did so between 6 and 7 min after the addition of the inhibitor. The quench rate rose from unity in the absence of NaCN to 2.2 ± 0.1 between 6 and 7 min after the addition of NaCN (n = 60, P < 0.05).

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Dose-dependent effect of NaCN () on [Ca2+]i of aortic endothelial cells. Control represents unstimulated cells in buffer solution. Data are means ± SE of n = 60 cells. *P < 0.05 vs. control.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Effect of glucose (Glu) on NaCN-evoked rise of [Ca2+]i. Endothelial cells were preincubated in the presence () or absence () of 10 mM Glu for 20 min, and then 5 mM NaCN was added. Data are means ± SE of n = 60 cells. *P < 0.05 vs. control in the absence of Glu. #P < 0.05 vs. NaCN alone.

Metabolic inhibition of newly isolated endothelial cells and endothelial cells in situ. Part of the experiments described above for cultured endothelial monolayers were repeated on newly isolated endothelial cells to explore the possibility that the culturing may have changed the response of the cells to metabolic inhibition. Changes of cytosolic Ca²⁺, recorded in the ratio mode of fura 2 fluorescence, are depicted in Fig. 6A. The newly isolated cells were submitted to the same protocols of 2-DG ± pyruvate or NaCN ± glucose as described for cultured cells. The changes in cytosolic Ca²⁺ control are qualitatively the same: 2-DG or NaCN caused a rapid rise in [Ca²⁺]_i, indicated by the rise of the fura 2 ratio and the effect of 2-DG was not suppressed by pyruvate, but the effect of NaCN was abolished by glucose.

View larger version (23K): [in this window] [in a new window]   Fig. 6.   A: effect of metabolic inhibition on newly isolated endothelial cells. Left, effect of Pyr on 2-DG-evoked rise of [Ca2+]i. Endothelial cells were preincubated in the absence (, top) or the presence (, bottom) of 5 mM Pyr for 20 min, and then 10 mM 2-DG was added. Right, effect of Glu on NaCN-evoked rise of [Ca2+]i. Endothelial cells were preincubated in the absence (, top) or the presence (, bottom) of 10 mM Glu for 20 min, and then 5 mM NaCN was added. Data are given as arbitrary units (means ± SE) of n = 60 cells. B: effect of metabolic inhibition on endothelial cells in situ on aortic disks. Left, effect of Pyr on 2-DG-induced rise of [Ca2+]i. Disks were preincubated in the absence (, top) or the presence (, bottom) of 5 mM Pyr for 20 min, and then 10 mM 2-DG was added. Right, effect of Glu on NaCN-evoked rise of [Ca2+]i. Disks were preincubated in the absence (, top) or the presence (, bottom) of 10 mM Glu for 20 min, and then 5 mM NaCN was added. Data are given as arbitrary units (means ± SE) of n = 6 experiments.

Cellular ATP contents. To evaluate the effect of metabolic inhibitors on endothelial energy production, ATP contents of cultured endothelial cells were determined under the same conditions as specified for [Ca²⁺]_i measurements. Addition of either 2-DG (10 mM) or NaCN (5 mM) alone caused a slight reduction of the cellular ATP contents (Fig. 7). However, the decrease in ATP contents did not reach significance during the first minute of metabolic inhibition, i.e., in the period when the initial [Ca²⁺]_i rise reached its peak value. The loss of ATP in the presence of 2-DG or NaCN was abolished when endothelial cells were incubated in the additional presence of mitochondrial (pyruvate 5 mM) or glycolytic (glucose 10 mM) substrates, respectively. Note that the presence of pyruvate, however, did not attenuate the 2-DG-induced initial rise of [Ca²⁺]_i (Fig. 3). Incubation of the cells in the simultaneous presence of both metabolic inhibitors caused an early and marked fall in cellular ATP contents.

View larger version (19K): [in this window] [in a new window]   Fig. 7.   Effect of metabolic inhibition on cellular ATP content of endothelial cells. Cells were incubated with 10 mM 2-DG in the absence () or the presence () of 5 mM Pyr, or with 5 mM NaCN in the absence () or the presence of 10 mM Glu (). Another group of cells was inhibited simultaneously with 5 mM NaCN and 5 mM 2-DG () in buffer solution. 100% corresponds to an ATP content of 17.6 ± 1.1 nmol/mg protein (n = 18). Data are means ± SE of n = 6 culture dishes of three independent cell preparations. *P < 0.05 vs. control without the addition of Pyr or Glu.

Effect of THG pretreatment on 2-DG-induced rise in [Ca²⁺]_i. We showed previously (20) that the initial rise of [Ca²⁺]_i in energy-depleted endothelial monolayers is sensitive toward a pretreatment of cells with THG, an inhibitor of the ER-Ca²⁺-ATPase (25), and that this pretreatment empties the ER. We have now applied a low dose of THG (10 nM) for a 15-min pretreatment of the cells that preempties the ER so slowly that it does not cause a significant rise of the steady-state [Ca²⁺]_i. To test for the emptying of the ER, ATP (10 µM) was added after the pretreatment. The cytosolic [Ca²⁺]_i rise elicited by ATP under control conditions was suppressed by 71 ± 8% (n = 60). After pretreatment with THG, the initial 2-DG-induced [Ca²⁺]_i rise was also markedly reduced (Fig. 8). This confirms the previous observation that energy depletion causes a [Ca²⁺]_i rise by an effect on the THG-sensitive ER.

View larger version (21K): [in this window] [in a new window]   Fig. 8.   Effect of pretreatment with thapsigargin (THG), an inhibitor of endoplasmic reticulum (ER)-Ca2+-ATPase, on the 2-DG-induced rise of [Ca2+]i. Endothelial cells were either preincubated in the presence of vehicle () or 10 nM THG () for 15 min, and then 10 mM 2-DG was added. Control represents cells incubated solely with 10 nM THG () without the addition of 2-DG. Data are means ± SE of n = 60 cells. *P < 0.05 vs. THG alone. #P < 0.05 vs. 2-DG alone.

Effect of ER-Ca²⁺-ATPase inhibitors on 2-DG-induced rise in [Ca²⁺]_i. In the next set of experiments, THG (10 nM) was not used for pretreatment but applied simultaneously with 2-DG to inhibit the ER-Ca²⁺-ATPase with the onset of metabolic inhibition. As illustrated in Fig. 9 the presence of the ATPase-inhibitor accelerated and augmented the 2-DG-induced [Ca²⁺]_i overload. Note that THG (10 nM) had no effect on [Ca²⁺]_i when applied in the absence of 2-DG (Fig. 8). The time course of the [Ca²⁺]_i overload was also altered in the copresence of THG and 2-DG (Fig. 9) compared with the time course in the presence of 2-DG alone; the typical transient drop in [Ca²⁺]_i, after its initial rise, was virtually abolished. Instead, [Ca²⁺]_i stayed elevated. Analogous experiments, as here described for THG, were also performed with cyclopiazonic acid (CPA, 1 µM), a structurally different inhibitor of the ER-Ca²⁺-ATPase (18). The same changes were observed (Fig. 10). THG and CPA increased the maximum rate of the initial rise of [Ca²⁺]_i and slowed down the subsequent transient decline of [Ca²⁺]_i.

View larger version (19K): [in this window] [in a new window]   Fig. 9.   Effect of copresence of THG, an inhibitor of ER-Ca2+-ATPase, on the 2-DG-induced rise of [Ca2+]i. Endothelial cells were either incubated with vehicle and 10 mM 2-DG () or 10 nM THG plus 10 mM 2-DG (), simultaneously. Data are means ± SE of n = 80 cells. #P < 0.05 vs. 2-DG alone.

View larger version (23K): [in this window] [in a new window]   Fig. 10.   Effect of THG or cyclopiazonic acid (CPA), inhibitors of ER-Ca2+-ATPase, on the 2-DG-induced maximal rate of [Ca2+]i incline (+ Ca2+max, left) and the maximal decline of [Ca2+]i (- Ca2+max, right) during the first 3 min after the addition of 2-DG. Endothelial cells were stimulated with 10 mM 2-DG alone, or 2-DG + 10 nM THG, or 2-DG + 1 µM CPA, simultaneously. Data are means ± SE of n = 80 cells. #P < 0.05 vs. 2-DG alone.

Effect of XeC on rise in [Ca²⁺]_i. The ER of endothelial cells can release Ca²⁺ through Ins(3)P-dependent release channels, but the presence of ryanodine-sensitive Ca²⁺ release channels has also been identified on the functional level (31). Inhibition of the latter with ryanodine had no effect on the 2-DG-induced initial rise of [Ca²⁺]_i compared with 2-DG alone [180 ± 7 nM (ryanodine 25 µM + 2-DG 10 mM, n = 60) vs. 185 ± 6 nM (2-DG 10 mM alone, n = 60), P > 0.05].

View larger version (22K): [in this window] [in a new window]   Fig. 11.   Effect of xestospongin C (XeC) [inhibitor of Ins(3)P-dependent Ca2+ release channels] on 2-DG-evoked rise of [Ca2+]i. Endothelial cells were preincubated in the presence of 0.3 or 3 µM XeC () or vehicle () for 20 min, and then 10 mM 2-DG was added. Control represents cells preincubated in the presence of 3 µM XeC without the addition of 2-DG. Data are means ± SE of n = 60 cells. *P < 0.05 vs. XeC alone. #P < 0.05 vs. 2-DG alone.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The findings of our study show that endothelial cells react sensitively to inhibition of glycolytic energy production. This mechanism was demonstrated here for cultured and freshly isolated endothelial cells as well as for endothelial cells in situ on the intact aortic vessel wall. Analysis of the mechanism reveal that glycolytic inhibition leads to a sudden release of endoplasmic-stored Ca²⁺ through an opening of Ins(3)P-dependent release channels. At the same time, the Ca²⁺-ATPase of the ER remains functionally active. Blockade of this endogenous Ca²⁺ release with XeC prevents the early and the delayed rise of [Ca²⁺]_i.

On energy depletion, porcine aortic endothelial cells responded with a biphasic rise of [Ca²⁺]_i, very similar to that described for coronary endothelial cells from the rat (20). The first phase lasts for a few minutes, with the peak [Ca²⁺]_i occurring around 1 min after the addition of metabolic inhibitors. This initial phase is sensitive to a pretreatment of the cells with THG, which empties the ER. Mn²⁺-quenching experiments show that the transplasmalemmal flux of these divalent cations is not enhanced during the initial phase of metabolic inhibition, indicating there is no increased Ca²⁺ influx at this time. These findings corroborate the conclusion from previous experiments on coronary endothelial cells (20) and rat aortic endothelial cells (30) that the Ca²⁺ accumulating in the cytosol on metabolic inhibition originates initially from the ER. The mechanisms by which the Ca²⁺ load of the ER becomes diminished previously remained unclear.

In experiments of the present study in which inhibitors of ER-Ca²⁺-ATPase, THG, or CPA were given simultaneously with the metabolic inhibitor, the initial, sudden rise of [Ca²⁺]_i became steeper and larger, and the transient secondary fall of [Ca²⁺]_i was nearly abolished. This shows that metabolic inhibition leads to an initial activation of the ER-Ca²⁺-ATPase, inhibited in the presence of THG as well as of CPA. This activation of the ER-Ca²⁺-ATPase blunts an otherwise higher rise of [Ca²⁺]_i and even reverses some of the initial cytosolic Ca²⁺ overload. This finding has consequences for the causal understanding of the initiation of the ER-dependent rise of cytosolic Ca²⁺. This [Ca²⁺]_i rise can either be due to a reduction in Ca²⁺ uptake into the ER or an increase in Ca²⁺ release from the ER. Because the Ca²⁺ uptake is not found inactivated, the [Ca²⁺]_i rise must be due to an augmented release from the ER.

We analyzed the nature of this release mechanism with use of specific inhibitors. Ryanodine had no effect, but XeC, a potent inhibitor of the Ins(3)P-sensitive release channel that displays high selectivity over ryanodine receptors (8), blocked the rise of [Ca²⁺]_i elicited by metabolic inhibition. This indicates that the Ins(3)P-sensitive release channel is responsible for the release.

The initial sudden rise of [Ca²⁺]_i, due to Ca²⁺ release from the ER, occurred at a time when changes in total endothelial contents of ATP were small and insignificant. This is in agreement with previous observations in coronary endothelial cells (20). At high drug concentrations, Ca²⁺ release could be provoked both by inhibitors of oxidative energy production and of glycolytic energy production. But the response was much more sensitive to inhibition of glycolytic energy production. This is because the effect of mitochondrial inhibition could be blunted by the administration of glucose as a substrate for glycolysis, in agreement with Shimizu and Paul (22), and because the effect of glycolytic inhibition could not be reduced by the administration of pyruvate as mitochondrial substrate. It is worth noting that in the presence of 2-DG and pyruvate, total cellular ATP contents remained absolutely unaltered for the entire duration of the experiments, whereas [Ca²⁺]_i still rose. This specific sensitivity of endoplasmic Ca²⁺ release to inhibition of glycolytic energy production may be due to local changes in ATP concentration at the ER, because the cascade of glycolytic enzymes is associated with ER membranes (29). For other cells, the Ins(3)P-sensitive Ca²⁺ release channel may become sensitized at low concentrations of ATP (6, 23). It seems possible that such a sensitization causes an increased opening probability of the channel also in metabolically inhibited endothelial cells. It is also possible that synthesis of Ins(3)P may be rapidly upregulated on metabolic inhibition.

After the first peak, [Ca²⁺]_i declines transiently and then rises again, progressively. Mn²⁺-quenching experiments reveal that transplasmalemmal Ca²⁺ influx starts with this second progressive rise in [Ca²⁺]_i. These results are in agreement with those obtained by Noll et al. (20) who showed that the second, but not the first, phase of [Ca²⁺]_i rise is abolished when extracellular Ca²⁺ is removed at the time of metabolic inhibition. In agreement with these previous results, we now observed that the delayed rise of cytosolic Ca²⁺ can be suppressed by the addition of Ni²⁺ to extracellular medium just after the peak of the initial rise. XeC applied at this point in time had no effect on the delayed rise. It abolished the delayed rise, however, when the ability of the ER to release Ca²⁺ was initially blocked. This demonstrates that the Ca²⁺ influx causing the second rise of [Ca²⁺]_i is dependent indirectly on Ca²⁺ discharge from the ER. It seems likely that the second rise of Ca²⁺ is due to the activation of store-operated transplasmalemmal Ca²⁺ influx as is also observed in endothelial cells in response to various receptor agonists. These results are in apparent contrast to those obtained from rat aortic endothelial cells under metabolic inhibition by Ziegelstein et al. (30). These authors observed a uniformly progressive rise of [Ca²⁺]_i, lacking the sudden initial component here described. They found that this progressive rise was sensitive to THG pretreatment but insensitive to removal of external Ca²⁺. They concluded it was directly due to Ca²⁺ leakage from the ER. It is unclear whether this discrepancy of results is due to differences in experimental procedure or cell species.

The sudden release of Ca²⁺ from the ER initiated by glycolytic inhibition represents a very sensitive reaction of endothelial cells to ischemic conditions. In ischemic-reperfused tissue, a number of factors can act on endothelial cells able to inhibit glycolysis. Among these are acidosis, increased concentrations of lactate, and oxygen radicals (13, 28). It is conceivable that the rapid response of endothelial [Ca²⁺]_i to glycolytic inhibition acts as a physiological sensor for metabolic disturbance in the surrounding tissue. Rise of cytosolic Ca²⁺ in endothelial cells leads to an increase in transendothelial permeability (5, 20), which facilities exchange of substances across this barrier, but also promotes interstitial edema. Emptying of the Ca²⁺ load of the ER may also be responsible for the reduced response to endothelium-dependent vasodilators observed by some authors after prolonged glycolytic inhibition (10, 27). This is because these mediators depend on a Ca²⁺-induced activation of NO synthase. The same authors (10, 27) failed to see an endothelium-dependent vasodilatation on the addition of 2-DG alone. It is unclear how these latter findings relate to the observation made in the present study, i.e., that cytosolic Ca²⁺ rises in response to 2-DG. The following explanations seem possible. First, as argued by Griffith et al. (11), 2-DG may have inhibitory effects on NO synthesis, and these are independent of Ca²⁺. Second, in the vessels investigated in those cited studies (10, 27), the endothelial Ca²⁺ response to 2-DG may have been smaller than the one found in the models here described.

In conclusion, this study shows that endothelial cells react sensitively to changes in glycolytic energy production with Ca²⁺ release from ER. Because many functions of endothelial cells, including their production of vasoactive mediators and their function in forming a permeability barrier, are triggered by changes in the cytosolic [Ca²⁺], this mechanism seems of pathophysiological significance. The fact that under energy depletion the initial rise of [Ca²⁺]_i, which is also responsible for the later sustained [Ca²⁺]_i overload of endothelial cells, can be inhibited by XeC, a blocker of Ins(3)P-release channels, may give rise to new therapeutic strategies.