Capacitative Ca²⁺ entry in vascular endothelial cells is mediated via pathways sensitive to 2 aminoethoxydiphenyl borate and xestospongin C

Abstract

1. Agonists increase endothelial cell intracellular Ca²⁺, in part, by capacitative entry, which is triggered by the filling state of intracellular Ca²⁺ stores. It has been suggested that depletion of endoplasmic reticulum (ER) Ca²⁺ stores either leads to a physical coupling between the ER and a plasma membrane channel, or results in production of an intracellular messenger which affects the gating of membrane channels. As an axis involving the IP₃ receptor has been implicated in a physical coupling mechanism the aim of this study was to examine the effects of the putative IP₃ receptor antagonists/modulators, 2 aminoethoxydiphenyl borate (2APB) and xestospongin C, on endothelial cell Ca²⁺ entry.

2. Studies were conducted in fura 2 loaded cultured bovine aortic endothelial cells and endothelial cells isolated from rat heart.

3. 2APB (30 – 300 μM) inhibited Ca²⁺ entry induced by both agonists (ATP 1 μM, bradykinin 0.1 μM) and receptor-independent mechanisms (thapsigargin 1 μM, ionomycin 0.5 and 5 μM). 2APB did not diminish endothelial cell ATP-induced production of IP₃ nor effect in vitro binding of [³H]-IP₃ to an adrenal cortex binding protein. Capacitative Ca²⁺ entry was also blocked by disruption of the actin cytoskeleton with cytochalasin (100 nM) while the initial Ca²⁺ release phase was unaffected.

4. Similarly to 2APB, xestospongin C (3 – 10 μM) inhibited ATP-induced Ca²⁺ release and capacitative Ca²⁺ entry. Further, xestospongin C inhibited capacitative Ca²⁺ entry induced by thapsigargin (1 μM) and ionomycin (0.5 μM).

5. The data are consistent with a mechanism of capacitative Ca²⁺ entry in vascular endothelial cells which requires (a) IP₃ receptor binding and/or an event distal to the activation of the ER receptor and (b) a spatial relationship, dictated by the cytoskeleton, between Ca²⁺ release and entry pathways.

Introduction

The production of vasodilator factors such as nitric oxide (via constituitive NO synthase) and prostacyclin (via cyclo-oxygenase) by endothelial cells represent Ca²⁺-dependent processes (for example see references Martin & Michaelis, 1990; Lin et al., 2000; Mizuno et al., 2000). The increase in intracellular Ca²⁺ typically occurs following the activation of membrane receptors which initiate signal transduction pathways that lead to Ca²⁺ release from the endoplasmic reticulum (ER) and entry from the extracellular compartment.

In many cell types, including endothelial cells, release of intracellular Ca²⁺ has been shown to be coupled to subsequent Ca²⁺ entry by store depletion or capacitative Ca²⁺ entry (Putney, 1990; Schilling et al., 1992; Parekh & Penner, 1997; Fasolato & Nilius, 1998). Such coupling has been demonstrated in response to receptor-mediated stimuli (Putney, 1990; Schilling et al., 1992; Parekh & Penner, 1997; Fasolato & Nilius, 1998) and receptor-independent emptying of intracellular stores by agents such as thapsigargin and ionomycin (Putney, 1990; Schilling et al., 1992; Parekh & Penner, 1997; Fasolato & Nilius, 1998). While the exact mechanism by which the filling state of the ER is linked to Ca²⁺ entry is uncertain, it has been proposed that either a factor(s) modulating Ca²⁺ current across the cell membrane is released on emptying of the stores (Randriamampita & Tsien, 1993; Thomas & Hanley, 1995), or that store-depletion leads to a physical coupling between the ER and the plasma membrane (Irvine, 1990; Berridge, 1995). With respect to the former, evidence has been provided for the involvement of a number candidate signalling molecules including a small molecular weight phosphate-containing compound (Randriamampita & Tsien, 1993), a cytochrome P450 epoxygenase metabolite of arachidonic acid (Rzigalinski et al., 1999), and events requiring tyrosine phosphorylation (Fleming et al., 1996; Babnigg et al., 1997). In recent studies of clonal embryonic kidney cells Ma et al. (2000) have provided evidence supporting a direct coupling mechanism, and further, that this coupling requires the inositol trisphosphate (IP₃) receptor or a related molecule (Ma et al., 2001). This conclusion was based on the observations that both rearrangement of the actin cytoskeleton and blockade of the IP₃ receptor with 2 aminoethoxydiphenyl borate (2APB) uncoupled store depletion from Ca²⁺ entry. Similar results have been presented for other cell types, including human platelets, where an alternate IP₃ receptor blocker, xestospongin C, inhibited both the association of IP₃ receptors with the membrane Ca²⁺ channel protein Trp1, and capacitative Ca²⁺ entry (Rosado & Sage, 2001). However, cell or species specific differences may exist as Ribeiro et al. (1997) reported that NIH-3T3 cells do not require an intact cytoskeleton to demonstrate capacitative Ca²⁺ entry as thapsigargin-induced Ca²⁺ entry was unaffected by cytochalasin D treatment. Further, this group also reported differences in the regulation of capacitative Ca²⁺ entry between 3T3 and pancreatic acinar cells (Louzao et al., 1996). In addition to cell specific differences, per se, variation may occur as a result of the heterogeneity within the family of protein subunits which constitute the Ca²⁺ influx channels (Hofmann et al., 1999; Putney, 1999; Mery et al., 2001). In this regard some seven mammalian homologues of the Trp proteins have been identified (Putney, 1999).

The present study aimed to demonstrate capacitative Ca²⁺ entry pathways in bovine aortic and rat heart endothelial cells and determine specifically whether such Ca²⁺ entry is dependent on an axis involving the IP₃ receptor-mediated mechanisms. Dependence on the IP₃ receptor was examined using the cell permeable, small molecular weight, inhibitors 2APB as described by Maruyama et al. (1997) and xestospongin C (Gafni et al., 1997).

Methods

Cell culture

Bovine aortae, obtained from a local slaughterhouse, were placed in cold physiological salt solution and transported to the laboratory. Vessels were then trimmed of adherent tissue, washed, and filled with serum free DMEM (5.6 mM glucose) containing 0.2 mg collagenase and incubated for 20 mins at 37°C. Endothelial cells were detached from the vessel wall by gentle agitation and the resulting cell suspension centrifuged. The cell pellet was then resuspended in fresh DMEM containing 10% heat-inactivated foetal calf serum and incubated at 37°C in an atmosphere of 5% CO₂. Cells were studied between passages 3 and 5.

For rat heart endothelial cells, hearts were removed from euthanized rats (protocol approved by the Animal Experimentation and Ethics Committee, RMIT University) and a polyethylene cannula exerted into the left ventricle. The cannulated preparation was connected to a re-circulating perfusion system and perfused with Joklik's medium containing heparin (1 u ml⁻¹) and collagenase (0.7 mg ml⁻¹) for 30 min, 37°C. Hearts were then mechanically disrupted and endothelial cells obtained by differential sieving and centrifugation (modified from Ford & Rovetto, 1987). Endothelial cells were then incubated in DMEM (5.6 mM glucose) with 20% foetal bovine serum at 37°C, 10% CO₂.

Endothelial cells were characterized by positive staining for endothelial cell nitric oxide synthase and negative staining for α smooth muscle actin.

Measurement of intracellular Ca²⁺ (Ca²⁺_i).

Following removal of endothelial cells from culture flasks (Ca²⁺ free solution containing 0.02% EDTA and 0.25% trypsin) cells (10⁵ cells ml⁻¹) were plated on glass coverslips which had been precoated with 50 μl of Engelberth Holm-Swarm (EHS) mouse sarcoma matrix (2.5 mg protein ml⁻¹). Cells were allowed to adhere for 16 h (37°C, 5% CO₂) prior to preparation for experiments. The coverslips were then transferred to a HEPES-buffered Kreb's solution containing the acetomethoxy ester of fura 2 (1 μM; 60 min). After loading cells were washed in fresh buffer to remove excess Ca²⁺ indicator.

Changes in intracellular Ca²⁺ related fluorescence were monitored using a video-based imaging system (Universal Imaging, PA, U.S.A.) coupled to an inverted microscope (20×Nikon Fluor objective lens, N.A. 0.75). As an index of changes in Ca²⁺_i, the ratio of emitted fluorescence (510 nm) intensities was calculated following excitation at 340 and 380 nm. Fluorescence excitation (340 and 380 nm) was obtained by passing light from a 75 W Xenon source through a computer-controlled filter wheel. In all experiments the responses in various protocols were averaged across 8 – 10 cells/coverslip. Preliminary experiments verified that similar responses were obtained in single cells and groups of cells.

Protocols

Initial studies established the concentration – response relationships for the inhibitory effect of 2APB (10 – 300 μM) on ATP (1 μM)-induced increases in intracellular Ca²⁺. The effects of 2APB were examined both in the presence of extracellular Ca²⁺ and in separate experiments during re-addition of Ca²⁺ following exposure to ATP in the absence of extracellular Ca²⁺. Additional experiments were performed using bradykinin to mobilize intracellular Ca²⁺ to determine whether the processes of capacitative Ca²⁺ entry were common for different modes of receptor activation.

Following establishment of the inhibitory concentrations of 2APB, the effect of 100 μM 2APB on thapsigargin (1 μM) and ionomycin (0.5 and 5 μM)-induced Ca²⁺ entry was determined.

To support studies conducted with 2APB, additional experiments were performed with the IP₃ receptor antagonist xestospongin C (Gafni et al., 1997). Initial studies examined the concentration-dependent effects of xestospongin C (3 – 10 μM) on ATP (1 μM)-induced Ca²⁺ responses. The responses to xestospongin C were studied both in the presence of extracellular Ca²⁺ and during Ca²⁺ re-addition following ATP stimulation in the absence of extracellular Ca²⁺. Further studies examined the effect of xestospongin C (6 and 10 μM) on thapsigargin (1 μM) and ionomycin (0.5 μM)-stimulated Ca²⁺ entry.

An additional set of studies were performed to examine whether cytoskeletal disruption dissociated agonist-induced Ca²⁺ release from capacitative Ca²⁺ entry, and whether this manoeuvre paralleled the effects of 2APB. Cells were prepared on EHS coated glass coverslips and loaded with fura 2 as described above. The cells were then treated with 100 nM cytochalasin D (60 min, 37°C) and ATP-induced changes in Ca²⁺_i determined as above. Disruption of the actin cytoskeleton was confirmed by FITC-phalloidin staining and confocal microscopy.

To determine whether 2APB exerted effects on production of IP₃, as opposed to an effect distal to generation of the second messenger, ATP-induced changes in cellular IP₃ were determined using a radioreceptor assay (Amersham). In brief, cells were plated on EHS-coated glass coverslips, as above, and placed in six-well culture plates containing DMEM. After the 16 h adherence period cells were washed and placed in Krebs buffer and stimulated with ATP (1 μM) in the presence and absence of 2APB (30 and 100 μM). In preparation for assay, IP₃ was extracted from cells using ice-cold 20% perchloric acid followed by neutralization in 60 mM HEPES, 1.5 M KOH solution. IP₃ production was measured at baseline and 10 s after the addition of ATP. This timepoint was chosen as preliminary studies have shown that in our endothelial cell preparations, IP₃ production in response to purinergic stimulation, is maximal at 10 s (data not shown) which is comparable with published studies (e.g. Purkiss et al., 1994).

Statistical methods

Changes in Ca²⁺_i were assessed as changes in the 340 : 380 nm fluorescence ratio at baseline (i.e. in the absence of agonist/drug treatment). Baseline levels were designated as 100% and responses normalized to this value. Group data is shown as mean±s.e.mean. Statistical differences between treatments has been determined by analysis of variance (ANOVA) with appropriate post hoc tests. Simple comparison of the means of two groups was determined using the Student t-test. Statistical significance was accepted at the P<0.05 level.

Chemicals and reagents

2APB (Aldrich Chemical Co., Milwaukee, WI, U.S.A.) was prepared as a 0.5 M stock solution in dimethysulphoxide; subsequent dilutions were made in physiological salt solution. ATP (Sigma Chemical Co., St Louis, MO, U.S.A.) and bradykinin (Sigma) were dissolved in physiological salt solution. Thapsigargin (Sapphire Bioscience, New South Wales, Australia) was dissolved in DMSO and subsequent dilutions made in physiological salt solution. Ionomycin (Sigma) was dissolved in chloroform and further diluted in physiological salt solution. Xestospongin C (Calbiochem) was prepared in DMSO as a 5.6 mM stock solution with subsequent dilution in physiological salt solution.

Results

Effects of 2APB on endothelial cell capacitative Ca²⁺ entry

The effect of 2APB, in bovine aortic endothelial cells, on ATP-induced changes in Ca²⁺_i are detailed in Figure 1. Figure 1a illustrates that in the presence of extracellular Ca²⁺ (2.5 mM) ATP (1 μM) stimulation results in a biphasic change in Ca²⁺_i, with a rapid increase followed by a decline to a steady-state level which remains significantly above basal levels. 2APB alone had no effect on basal fluorescence but caused a concentration-dependent inhibition of both phases of Ca²⁺ mobilization (30 – 300 μM; Figure 1a) with 100 μM 2APB abolishing the secondary phase (Figure 1a,b). To illustrate capacitative Ca²⁺ entry, cells were exposed to ATP in the absence of extracellular Ca²⁺, after which Ca²⁺ was returned to the superfusate in the continued presence of the agonist (Figure 1b). 2APB (100 μM) added prior to ATP both inhibited the initial Ca²⁺ release and prevented the influx of Ca²⁺ following the return of the cation (Figure 1b). Similarly, when 2APB was added immediately prior to the re-addition of Ca²⁺, capacitative Ca²⁺ entry was inhibited consistent with the capacitive entry mechanism per se being inhibited in this condition rather than solely being a consequence of attenuated IP₃-mediated store release (Figure 1c).

Figure 1.

Effects of 2APB on ATP-induced changes in intracellular Ca²⁺. Studies shown in (a – d) were performed in bovine aortic endothelial cells and those in (e) in rat heart endothelial cells. (a) Shows the concentration-dependent effects of 2APB on ATP-induced changes in Ca²⁺_i in the presence of extracellular Ca²⁺ (n=10). 2APB inhibits both the initial release of Ca²⁺and the subsequent plateau phase. (b) Illustrates the effect of 100 μM 2APB on ATP-induced Ca²⁺ release in the absence of extracellular Ca²⁺ and the effect of subsequent re-addition of Ca²⁺ (1 mM) to the superfusate (n=6). (c) Demonstrates the effect of 2APB specifically on the Ca²⁺ entry phase; addition of 2APB after ATP-induced Ca²⁺ release and immediately prior to the re-addition of extracellular Ca²⁺ prevented capacitative Ca²⁺ entry (n=6). (d) Shows the concentration dependence of the effects of 2APB (10 – 100 μM) on the ATP (1 μM)-induced Ca²⁺ release and influx components (n=5). (e) Demonstrates that 2APB similarly inhibits ATP (1 μM)-induced Ca²⁺ release and influx in cultured rat heart endothelial cells (n=6). Results are presented as mean±s.e.mean; n=number of coverslips. For clarity, representative error bars are shown.

To determine whether 2APB inhibits the Ca²⁺ release and entry components in bovine endothelial cells, the effects of 10, 50 and 100 μM 2APB on ATP (1 μM)-induced Ca²⁺ mobilization were compared (Figure 1d). While both components showed a 2APB concentration – dependent inhibition it appears that a residual release component persisted in the presence of 100 μM 2APB, whereas this concentration of 2APB abolished the influx component.

Endothelial cells cultured from rat heart showed a qualitatively similar response to ATP, with an initial Ca²⁺_i release phase followed by Ca²⁺ entry (Figure 1e). As observed with the bovine cells, 2APB inhibited both phases of Ca²⁺ mobilization (Figure 1e).

Similar results were obtained when bradykinin (0.1 μM) was used to mobilize Ca²⁺_i. For example, 100 μM 2APB decreased the initial Ca²⁺ release peak from 268±69% of baseline to 107±5% (P<0.05) and the secondary Ca²⁺ entry phase from 224±47% to 108±4% (P<0.05; n=5). This suggests that the effects of 2APB are not specific to ATP-receptor mediated events.

To illustrate the effects of 2APB on receptor-independent mechanisms of stimulating capacitative Ca²⁺ entry, separate preparations of bovine endothelial cells were treated with ionomycin (0.5 and 5 μM; Figure 2) or thapsigargin (1 μM; Figure 3). Ionomycin added to cells in the absence of extracellular Ca²⁺ resulted in a release of intracellular Ca²⁺, which was followed by an influx of Ca²⁺ on re-addition of the cation to the superfusate (Figure 2a,b). Concentration-dependent effects of ionomycin on both phases of Ca²⁺ mobilization were observed (Figure 2a,b). Addition of 2APB (100 μM) prior to ionomycin treatment prevented subsequent Ca²⁺ influx (Figure 2a,b). When 2APB was added to ionomycin-treated cells during the Ca²⁺ influx component there was a rapid decrease in intracellular Ca²⁺, consistent with 2APB inhibiting the Ca²⁺ entry process (Figure 2c). Washout of 2APB restored ionomycin-induced Ca²⁺ influx (data not shown).

Figure 2.

Effect of 2APB (100 μM) on ionomycin (0.5, n=6 and 5 μM, n=18)-induced changes in intracellular Ca²⁺ (a and b). (c (n=6)) Illustrates that addition of 2APB during the phase of capacitative Ca²⁺ entry inhibits entry causing a decline in Ca²⁺_i. Results are presented as mean±s.e.mean. For clarity, representative error bars are shown.

Figure 3.

Effect of 2APB (100 μM) on thapsigargin (1 μM)-induced changes in intracellular Ca²⁺. In the absence of extracellular Ca²⁺ thapsigargin induced a transient increase in Ca²⁺_i; subsequent addition of Ca²⁺ to the superfusate resulted in a large increase in Ca²⁺_i (a; n=5). (b) Illustrates that 2APB while not preventing the initial thapsigargin-induced increase in Ca²⁺ prevented subsequent capacitative Ca²⁺ entry (n=5). Washout of 2APB restored entry of Ca²⁺. Results are presented as mean±s.e.mean. For clarity, representative error bars are shown.

Exposure of bovine aortic endothelial cells to the ER Ca²⁺ ATPase inhibitor thapsigargin resulted in an increase in Ca²⁺_i and Ca²⁺ influx on re-addition of Ca²⁺ to the superfusate (Figure 3a). As in the case of ionomycin, 2APB significantly inhibited the thapsigargin-induced Ca²⁺ influx component (Figure 3a,b). Figure 3b further illustrates that washout of 2APB, in the presence of Ca²⁺ and thapsigargin, restores Ca²⁺ influx.

Effect of xestospongin C on endothelial cell Ca²⁺ responses to ATP and capacitative Ca²⁺ entry

The effects of xestospongin C, in bovine aortic endothelial cells, on ATP-induced changes in Ca²⁺_i are shown in Figure 4. Xestospongin C alone had no effect on basal fluorescence but appeared to cause a concentration-dependent inhibition of both phases of Ca²⁺ mobilization (3 – 6 μM; Figure 4a). To specifically examine capacitative, Ca²⁺ entry, cells were exposed to ATP in the absence of extracellular Ca²⁺, after which Ca²⁺ was returned to the superfusate in the continued presence of the agonist (Figure 4b). Xestospongin C added prior to ATP both inhibited the initial Ca²⁺ release and prevented the influx of Ca²⁺ following the return of the cation (Figure 4b). For example, in the control state re-addition of Ca²⁺ caused a peak increase in fluorescence of 354±23% (n=7), while in the presence of xestospongin C (10 μM) the change in fluorescence was reduced to 172±23% (n=3; P<0.05).

Figure 4.

Effects of xestospongin C on intracellular Ca²⁺ in bovine endothelial cells. (a) Shows the concentration-dependent effects of xestospongin C on ATP-induced changes in intracellular Ca²⁺. Studies were performed in the presence of extracellular Ca²⁺. (b) Illustrates the effect of xestospongin C on ATP-induced Ca²⁺ release in the absence of extracellular Ca²⁺ and the effect of subsequent re-addition of Ca²⁺ (1 mM) to the superfusate. Results are presented as mean±s.e.mean. For clarity, representative error bars are shown.

To further confirm an inhibitory effect of xestospongin C on capacitative Ca²⁺ entry, the response to Ca²⁺ re-addition following thapsigargin (1 μM) treatment was compared in the absence and presence of the IP₃ receptor antagonist. Xestospongin C (6 and 10 μM) significantly inhibited Ca²⁺ entry on re-addition of extracellular Ca²⁺ (Figure 5a). Similarly, Ca²⁺ entry induced by ionomycin (0.5 μM) was significantly attenuated by pretreatment with xestospongin C (10 μM) (Figure 5b). To specifically demonstrate an effect on the Ca²⁺ entry component, xestospongin C (10 μM) was added to ionomycin (0.5 μM) treated cells at the time when Ca²⁺ was returned to the superfusate (Figure 5c). As with xestospongin C pretreated cells, Ca²⁺ entry was significantly (P<0.05) attenuated. Additional experiments demonstrated that xestospongin C treatment also reversed Ca²⁺ entry induced by thapsigargin (1 μM) (data not shown). Collectively these data indicate that xestospongin C inhibits the capacitative Ca²⁺ entry phase in bovine endothelial cells.

Figure 5.

Effects of xestospongin C on thapsigargin and ionomycin-induced Ca²⁺ entry in bovine endothelial cells. (a) Illustrates the inhibitory effect of xestospongin C (6 and 10 μM; n=5 and 4 respectively) compared to control (n=7) on capacitative Ca²⁺ entry induced by exposure of endothelial cells to thapsigargin (1 μM). (b) Shows the inhibitory effect of xestospongin C (10 μM) pretreatment on ionomycin (5 μM; n=3)-induced Ca²⁺ entry compared with control (n=3). (c) Illustrates that xestospongin C (10 μM) added at the time of Ca²⁺ addition to the superfusate also attenuates capacitative Ca²⁺ entry (n=4 for each condition). Results are presented as mean±s.e.mean. For clarity, representative error bars are shown.

Effect of cytoskeletal disruption on endothelial cell capacitative Ca²⁺ entry

Cytochalasin D treatment, per se, had no apparent significant effect on baseline intracellular Ca²⁺, as indicated by fluorescence ratio values measured in 0 mM extracellular Ca²⁺ conditions, prior to the addition of ATP. Control, 0.42±0.03; 100 nM cytochalasin 0.47±0.05; 1 μM cytochalasin 0.51±0.05 (n=6, ANOVA, P<0.331, n.s.)

Exposure of cells to cytochalasin D (100 nM and 1 μM) had no apparent effect on ATP-induced Ca²⁺ release, while significantly inhibiting subsequent capacitative Ca²⁺ entry (Figure 6). Effectiveness of cytochalasin in disrupting the actin cytoskeleton was confirmed by confocal microscopy of FITC-phalloidin labelled endothelial cells (data not shown). Consistent with published studies (Rosado & Sage, 2001), cytochalasin treatment appeared to cause retraction of the cytoskeleton from the plasma membrane with a clumping of phalloidin-staining material within deeper regions of the cells. In addition, cytochalasin appeared to cause concentration-dependent changes in cell morphology evident as an increased occurrence of cytoplasmic protrusions.

Figure 6.

Effects of cytoskeletal disruption on ATP-induced changes in intracellular Ca²⁺. Cells were exposed to cytochalasin for 1 h at 37°C in a tissue culture incubator. As treatment did not cause a significant change in baseline flourescence ratio between groups, data are presented as normalized to per cent change from baseline prior to ATP stimulation. Exposure of cells to cytochalasin D (100 nM or 1 μM) had no apparent effect on the initial ATP-induced Ca²⁺ release while significantly inhibiting the capacitative Ca²⁺ entry phase (n=6 for each condition). Results are presented as mean±s.e.mean. For clarity, representative error bars are shown.

Effects of 2APB on ATP-induced inositol trisphosphate production

ATP (1 μM) caused a significant increase in IP₃ production; 2.95±2.22 pmol/well baseline compared to 9.86±3.44 pmol/well after 10 s of ATP exposure (P<0.05 compared to baseline). In the presence of 2APB (100 μM) ATP caused a similarly significant increase in IP₃ production; baseline 3.63±3.63 pmol well⁻¹ compared to 9.92±2.97 pmol well⁻¹ after 10 s of ATP exposure (P<0.05 compared to baseline). Results are presented as mean±s.e.mean for six coverslips from two independent experiments.

To verify, under cell free conditions, that 2APB did not inhibit IP₃ binding, concentration response curves for IP₃ and 2APB in displacing [³H]-IP₃ from bovine adrenal cortex were prepared. 2APB (3 – 100 μM) did not result in a measurable alteration in the binding of [³H]-IP₃ to the binding protein (data not shown).

Discussion

The results of the present study are consistent with a mechanism of capacitative Ca²⁺ entry in vascular endothelial cells which requires IP₃ receptor binding or an event distal to the activation of the ER receptor. This conclusion is based on the observation that 2APB and xestospongin C, putative IP₃ receptor antagonists or modulators, inhibit capacitative Ca²⁺ entry induced by either agonists (ATP, bradykinin) or receptor-independent Ca²⁺ mobilization (via ionomycin or thapsigargin). Control experiments demonstrated that 2APB did not lead to a reduction in IP₃ production or [³H]-IP₃ binding. Further, the finding of similar effects of 2APB on Ca²⁺ mobilization in endothelial cells from both bovine aorta and rat heart suggest that the findings are consistent across species and possibly between vascular sites.

Consistent with a number of previous studies (for example Lynch et al., 1992; Vaca & Kunze, 1994; Wang & Van Breemen, 1997) exposure of endothelial cells to ATP or bradykinin resulted in a biphasic change in intracellular Ca²⁺; an initial rapid increase that is a function of ER release and a sustained plateau that is, in part, dependent on Ca²⁺ entry from the extracellular space. As endothelial cells lack voltage gated Ca²⁺ channels, entry of this cation is considered to primarily occur through receptor/ligand gated channels and mechanisms related to the filling state of the ER, that is capacitative Ca²⁺ entry (Barritt, 1999; Lin et al., 2000; Sedova et al., 2000). The existence of the latter in the present studies was suggested by the influx of Ca²⁺ that occurred when the cation was returned to the superfusate of cells initially exposed to the agonists in the absence of extracellular Ca²⁺. Further, when the ER Ca²⁺ store was depleted by the ionophore, ionomycin, or the Ca²⁺ ATPase inhibitor, thapsigargin, Ca²⁺ entry was stimulated. As these latter compounds act on the filling state of the ER the data is consistent with a capacitative Ca²⁺ entry mechanism.

Two principal mechanisms have been proposed for the coupling of the ER filling state to Ca²⁺ entry; [1] that store depletion causes the release of a factor which acts to alter the gating properties of channels within the cell membrane (Randriamampita & Tsien, 1993; Thomas & Hanley, 1995) and [2] that store depletion results in a conformational change in an ER element which forms a direct or physical communication with the plasma membrane to allow Ca²⁺ entry (Irvine, 1990; Berridge, 1995). Recent studies of Ma et al. (2000) have been used to support a model involving a physical association between the IP₃ receptor on the ER and a Ca²⁺ entry channel on the plasma membrane (Berridge et al., 2000). The involvement of the IP₃ receptor was suggested from studies using both 2APB and xestospongin, while the physical association between the two compartments was suggested by studies altering the actin cytoskeleton. As in the present study, disruption of the cytoskeleton with cytochalasin D, in a cell line derived from pulmonary artery endothelium, was shown to inhibit capacitative Ca²⁺ entry without an effect on the initial agonist-induced release of Ca²⁺ (Holda & Blatter, 1997). Further support for a dynamic role of the actin cytoskeleton is provided by the studies of Rosado & Sage (2001) in human platelets, showing that both stabilization of the actin cytoskeleton with jasplakinolide and disruption by cytochalasin prevent capacitative Ca²⁺ entry.

The exact role that the cytoskeleton appears to play in capacitative Ca²⁺ entry appears to vary between cell types. In contrast to the findings of the present study, and those described above (Holda & Blatter, 1997; Rosado & Sage, 2001), Riberio et al. (1997) reported that in NIH-3T3 cells cytochalasin treatment specifically affects the agonist-induced release of Ca²⁺ while having no effect on capacitative Ca²⁺ entry. The differences, particularly in regard to the Ca²⁺ release phase, may relate to the extent of cytochalasin treatment as Riberio et al. (1997) used a 10 μM concentration which markedly changed cell morphology while in the present study 100 nM and 1 μM concentrations of cytochalasin were employed. The lower concentrations were chosen to avoid gross changes in cell shape and detachment of the cells from the underlying matrix. In a study examining the role of capacitative Ca²⁺ entry in regulation of adenylyl cyclase isoforms, of C6-2B rat glioma cells, Fagan et al. (1998) have suggested that regulation of the cyclase requires an intimate association with the Ca²⁺ entry pathway; although this relationship could not be altered by agents known to disrupt the cytoskeleton (cystochalasin, nocodozole, colchicine). Similarly to the studies of Riberio et al. (1997) cytoskeletal disruption did not impair capacitative Ca²⁺ entry in response to thapsigargin treatment. In studies of embryonic kidney cells (HEK293) Ma et al. (2000) reported that condensation of cortical actin by treatment with the phosphatase inhibitor calyculin A impaired capacitative Ca²⁺ entry while not affecting IP₃-mediated Ca²⁺ release. Formation of a dense cortical actin layer was considered to impair the physical interaction between the plasma membrane and endoplasmic reticulum. This group, however, has previously reported that disassembly of the actin cytoskeleton with cytochalasin D did not impair coupling between Ca²⁺ release and Ca²⁺ entry (Patterson et al., 1999). In addition to these studies of other investigators, we did not find cytochalasin treatment to markedly inhibit capacitative Ca²⁺ entry in arteriolar smooth muscle (Potocnik & Hill, 2001). Thus, collectively the available data suggest that dependence of capacitative Ca²⁺ entry on a functional cytoskeleton may vary between cell types. Conceivably the physical relationship between the endoplasmic reticulum and Ca²⁺ entry channels is dependent on different structural elements in differing cell types. Alternatively contrasting results may reflect other factors such as differences in the Ca²⁺ entry channels in the varying cell types (Putney, 1999) or variation in the amount/distribution of cytoskeletal elements.

The results of the present studies add to a growing body of evidence supporting a modulatory effect of 2APB on IP₃ receptor-mediated processes (Maruyama et al., 1997; Ascher-Landsberg et al., 1999; Gysembergh et al., 1999; Ma et al., 2000). With respect to ATP-induced Ca²⁺ mobilization 2APB concentration-dependently attenuated both the initial release phase and subsequent Ca²⁺ entry. 2APB did not, however, appear to have any effect on ATP-induced IP₃ production or basal Ca²⁺ levels as assessed by changes in the 340 : 380 nm fluorescence ratio. In contrast to the latter observation Gysembergh et al. (1999) reported an effect of 2APB (1 – 10 μM) on rabbit cardiac myocytes, while Maruyama et al. (1997) found that concentrations greater than 90 μM 2APB were required to increase baseline Ca²⁺_i in a rat cerebral microsomal preparation and greater than 200 μM 2APB was required to increase basal Ca²⁺_i in human platelets. Whether these differences reflect tissue specific or methodological differences is currently uncertain.

Recent studies have suggested that 2APB may exert a direct inhibitory effect on capacitative entry at sites other than the IP₃ receptor. Consistent with this, 2APB has been shown to inhibit Ca²⁺ entry in DT 40 B cells in which the IP₃ receptor has been deleted (Broad et al., 2001), and in excised membrane patches from rat basophilic leukaemia cells (Braun et al., 2001). In the present study, however, both putative IP₃ receptor antagonists, 2APB and xestospongin, inhibited endothelial cell capacitative Ca²⁺ entry, suggesting that in this cell type the IP₃ receptor may indeed be a component of the capacitative Ca²⁺ entry pathway. Consistent with these results, Rosado & Sage (2001) have recently shown that xestospongin inhibits capacitative Ca²⁺ entry in human platelets and the association between IP₃ receptors and Trp1. No evidence currently exists for xestospongin exerting an effect directly at the level of Trp channels, although it has been suggested that in solution 2APB may dimerise thereby resembling the molecular shape of xestospongin C (Van Rossum et al., 2000; Liu & Ambudkar, 2001). The apparently discrepant results may also relate to the particular Ca²⁺ channel proteins (for example mammalian homologues of Trp channels (Putney, 1999)) or IP₃ receptor subtypes expressed in different cell types. With respect to the latter it has been shown in A7r5 cells that specific IP₃ receptor subtypes (namely the IP₃R1 subtype) are associated with capacitative Ca²⁺ entry, while the IP₃R3 subtype appears unrelated to this mode of Ca²⁺ entry (Wang et al., 2001).

In more recent studies Ma et al. (2001) have suggested that 2APB, while being an IP₃ receptor antagonist, may also interact with a regulatory protein that exerts an action over both IP₃ receptors and capacitative Ca²⁺ entry channels. Such an effect could be consistent with both the data from the present studies and those suggesting an inhibitory effect of 2APB proximal to the ER in this signalling pathway.

That 2APB inhibited store-depletion mediated Ca²⁺ entry in the present studies was demonstrated firstly by the addition of 2APB to ATP-stimulated endothelial cells following the initial release peak (Figure 1) and secondly by inhibition of Ca²⁺ entry following exposure of cells to ionomycin (Figure 2) and thapsigargin (Figure 3). In the latter case Ca²⁺ depletion of the ER occurs independently of plasma membrane receptor activation. Somewhat surprisingly, 2APB reduced the initial Ca²⁺ release caused by 5 μM ionomycin while not affecting the initial increase in Ca²⁺_i induced by either 0.5 μM ionomycin or thapsigargin (1 μM). Conceivably the higher concentration of the ionophore, as well as depleting intracellular Ca²⁺ stores, exerts an additional effect at the level of the ER membrane. Alternatively, at the higher concentration of ionomycin unidentified effects are evident; perhaps related to the greater release of Ca²⁺ seen under this condition (relative to 0.5 μM ionomycin). Regardless of this effect, the data presented are consistent with 2APB inhibiting store depletion mediated Ca²⁺ entry in endothelial cells.

An alternate explanation of the data relates to the functional relationship between ryanodine release channels of the ER and plasma membrane K_Ca channels. Ca²⁺ release from the ER into the subplasmalemmal space has been shown to activate K_Ca channels, resulting in membrane hyperpolarization (Frieden & Graier, 2000). This change in membrane potential subsequently increases the driving force for Ca²⁺ entry and has been implicated in agonist-induced production of a number of Ca²⁺-dependent vasoactive compounds (Luckhoff & Busse, 1990a). Support of this pathway has also been derived from the observation that K⁺ channel activators lead to hyperpolarization and Ca²⁺ entry in endothelial cells (Luckhoff & Busse, 1990b). Analogous to the data obtained in the present studies disruption of the normal relationship between the superficial ER and the plasma membrane with nocodozole prevented ryanodine-induced activation of K_Ca channels in a human umbilical vein endothelial cell line (Frieden & Graier, 2000). While it is possible that such a mechanism would have been disrupted in the current studies, by exposure of cells to cytochalasin, it appears unlikely that both the 2APB (see also Maruyama et al., 1997) and xestospongin data could be explained simply by a ryanodine receptor-Ca²⁺ entry pathway. Further, if the endothelial cells studied exhibit both IP₃ receptor-dependent mechanisms and Ca²⁺-induced Ca²⁺ release (CICR) (Mozhayeva, 1996) the effectiveness of 2APB and xestospongin C in preventing Ca²⁺ entry suggests that the IP₃ receptor dependent processes occur proximally to CICR.

In summary, the data presented demonstrate the effectiveness of 2APB and xestospongin C, putative modulators of the IP₃ receptor, in inhibiting agonist-induced Ca²⁺ entry in vascular endothelial cells. Further, the fact that cytoskeletal disruption with cytochalasin specifically inhibited the Ca²⁺ influx phase, as opposed to the release component, adds support to the requirement of a physical coupling, or spatial proximity, for the coupling of store depletion to Ca²⁺ entry in endothelial cells. Differences in the effect of agents such as cytochalasin on Ca²⁺ entry, however, suggest that variation in the exact coupling mechanism may exist between cell types. Given the ability of the endothelial cell in responding to physiological forces such as shear stress (Davies et al., 1997) and the fact that the cytoskeleton is implicated in mechanotransduction (Ingber, 1997), these results may be of direct relevance to physiological mechanisms of endothelial cell autacoid production.

Abbreviations

2APB

2 aminoethoxydiphenyl borate

ATP

adenosine triphosphate

CICR

Ca²⁺-induced Ca²⁺ release

DMEM

Dulbecco's modified essential medium

DMSO

dimethyl sulphoxide

EHS

Engelberth Holm-Swarm

ER

endoplasmic reticulum

IP₃

inositol trisphosphate

Trp

transient receptor potential