Endoplasmic Reticulum Ca2+ Depletion Unmasks a Caffeine-Induced Ca2+ Influx in Human Aortic Endothelial Cells
Abstract Intracellular Ca2+ pools contribute to changes in cytosolic [Ca2+] ([Ca2+]i), which play an important role in endothelial cell signaling. Recently, endothelial ryanodine-sensitive Ca2+ stores were shown to regulate agonist-sensitive intracellular Ca2+ pools. Since caffeine binds the ryanodine Ca2+ release channel on the endoplasmic reticulum in a variety of cell types, we examined the effect of caffeine on [Ca2+]i in human aortic endothelial cell monolayers loaded with the fluorescent probe indo 1. Under baseline conditions, 10 mmol/L caffeine induced a small increase in [Ca2+]i from 86±10 to 115±17 nmol/L (mean±SEM); this effect was similar to that of 5 μmol/L ryanodine and was unaffected by buffer Ca2+ removal. After depletion of an intracellular Ca2+ store by the irreversible endoplasmic reticulum Ca2+-ATPase inhibitor thapsigargin (1 μmol/L), ryanodine did not affect [Ca2+]i. In contrast, caffeine induced a large rapid increase in [Ca2+]i (176±19 to 338±35 nmol/L, P<.001) after thapsigargin exposure; this effect of caffeine was only observed when extracellular Ca2+ was present. A similar increase in [Ca2+]i was induced by caffeine after depletion of ryanodine- and histamine-sensitive Ca2+ stores or after pretreatment with the endoplasmic reticulum Ca2+-ATPase inhibitor cyclopiazonic acid (10 μmol/L). Thus, under baseline conditions the effect of caffeine on [Ca2+]i is similar to that of ryanodine and appears to be due to the release of an intracellular store. However, after depletion of an endoplasmic reticulum Ca2+ store, caffeine, but not ryanodine, stimulates Ca2+ influx, resulting in a large increase in [Ca2+]i. The data suggest that caffeine-induced Ca2+ influx is controlled by the status of Ca2+ loading of intracellular Ca2+ stores in human aortic endothelial cells.
Intracellular Ca2+ plays an important role in endothelial cell signaling. In endothelial cells, stimulation of histamine and bradykinin receptors increases [Ca2+]i through InsP3-mediated mobilization of an ER Ca2+ store1 2 and through stimulation of extracellular Ca2+ influx.3 4 The increase in endothelial [Ca2+]i in turn triggers the release of vasoactive substances such as prostacyclin5 and endothelium-derived relaxing factor/nitric oxide.6 7 Because of this and other critical roles in modulating the functional properties of the vascular endothelium, considerable attention has recently been placed on characterizing endothelial cell intracellular Ca2+ regulation.
The ER Ca2+-ATPase inhibitor thapsigargin8 9 has recently been used as a pharmacological probe of intracellular Ca2+ pools in endothelial cells. Thapsigargin largely depletes the InsP3- and agonist-sensitive Ca2+ pools in many cell types,10 including vascular endothelial cells,3 11 without activation of the InsP3 pathway.8 Thapsigargin also promotes extracellular Ca2+ entry in vascular endothelial cells,11 resulting in a sustained elevation in [Ca2+]i in the presence of buffer Ca2+.
The plant alkaloid ryanodine has also been used to study intracellular Ca2+ pools in several cell types. Ryanodine binds to the SR Ca2+ release channel of muscle cells.
At nanomolar concentrations in experiments using isolated SR vesicles, ryanodine maintains the Ca2+ release channel in an open substate conductance, whereas at micromolar or higher concentrations, ryanodine appears to inhibit channel activation.12 The Ca2+ release channel of muscle cells is also sensitive to caffeine,13 which opens SR Ca2+ channels and promotes discharge of Ca2+ stores.14 The presence of the ryanodine receptor was recently demonstrated in vascular and endocardial endothelium.15 Endothelial ryanodine-sensitive Ca2+ stores appear to be involved in the regulation of agonist-sensitive intracellular Ca2+ pools.16 Whether ryanodine and caffeine act on the same Ca2+ pool and similarly affect [Ca2+]i in endothelial cells is unknown.
A recent study using patch-clamp techniques in bovine aortic endothelial cells17 demonstrated that caffeine releases Ca2+ from an intracellular store that is pharmacologically distinct from the InsP3-releasable Ca2+ store. Caffeine-induced Ca2+ release from intracellular stores was also recently documented by whole-cell membrane current recordings in freshly dissociated aortic endothelial cells.18 The present study further characterizes the effect of caffeine on [Ca2+]i in HAECs loaded with the membrane-permeant (acetoxymethyl ester) form of indo 1. Our results demonstrate that under baseline conditions, caffeine induces a small increase in [Ca2+]i both in the presence and absence of buffer Ca2+. This increase in [Ca2+]i appears to be due to release of an intracellular store and is similar to the effect elicited by ryanodine.16 However, after ER Ca2+ depletion in the presence of buffer Ca2+, caffeine induces a rapid and marked increase in [Ca2+]i due to extracellular Ca2+ influx. In contrast, ryanodine does not stimulate Ca2+ influx under these conditions. We propose that caffeine-stimulated Ca2+ influx is controlled by the status of ER Ca2+ stores in HAECs.
Materials and Methods
HAECs (Clonetics) were obtained as cryopreserved cells and grown to confluence to passages 4 to 7 in endothelial cell growth medium (Clonetics) with 2% fetal bovine serum, 50 μg/mL gentamicin, 50 ng/mL amphotericin B, and ≈12 μg/mL bovine brain extract at 37°C in a humidified atmosphere of 95% air/5% CO2. After treatment with 0.25% trypsin in a Ca2+- and Mg2+-free solution, cells were plated on glass coverslips or in 1-mm2 glass capillary tubes (Vitro Dynamics) precoated with 1% gelatin (Sigma Chemical Co) 1 to 2 days before experimental use.
Measurement of Endothelial [Ca2+]i
HAEC monolayers were loaded with 8 μmol/L of the ester derivative (acetoxymethyl ester form) of the fluorescent Ca2+ probe indo 1 (Molecular Probes) in a 95% air/5% CO2 incubator at room temperature for 30 minutes. Monolayers were then gently washed in indicator-free HEPES-buffered saline solution and maintained an additional 30 to 60 minutes to allow for deesterification of the indicator before the experiment. HAEC monolayers were studied on the stage of a modified inverted fluorescence microscope (Zeiss IM-35) as previously described.19 Briefly, indo 1 fluorescence collected from a field of ≈15 cells was excited at 350±5 nm by using a xenon strobe lamp (model FX 193, EG & G Electro-Optics). Bandpass interference filters (Andover) selected wavelength bands of emitted fluorescence at 391 to 434 nm (410 channel) and 457 to 507 nm (490 channel), corresponding to the Ca2+-bound and Ca2+-free forms of the indicator, respectively. The 410- to 490-nm ratio was used as an index of [Ca2+]i. Autofluorescence of a comparable field of unloaded cells was <5% of indo 1–loaded cells.
Previous work has shown that caffeine quenches indo 1 fluorescence by the same factor at both emission wavelengths without affecting either the indo 1 ratio, the Kd of indo 1 for Ca2+, or cell autofluorescence.20 In a prior study using the fluorescence system in our laboratory, caffeine decreased the fluorescence at each wavelength by ≈15% in isolated cardiac myocytes.21 In the present study, indo 1 fluorescence decreased by ≈20% in HAECs during caffeine superfusion. The comparability of the indo 1 fluorescence ratios and the intracellular Ca2+ transients in the presence and absence of caffeine in cardiac myocytes was affected only to the extent to which cell autofluorescence contributed to total fluorescence.21 Since endothelial cell autofluorescence is negligible compared with that of cardiac myocytes under our experimental conditions, the effect of caffeine on indo 1 fluorescence would not be expected to have a significant effect on determinations of [Ca2+]i.
To approximate [Ca2+] values from indo 1 fluorescence ratios, intracellular Rmin and Rmax values were determined as previously described.22 Briefly, to determine an intracellular Rmin, indo 1-loaded HAEC monolayers were superfused with a solution containing (mmol/L) NaCl 137.0, KCl 5.0, MgSO4 1.2, NaH2PO4 1.2, d-glucose 16, HEPES 10, and EGTA 2 (pH 7.40±0.01). The monolayers were then exposed to a solution of similar composition with 10 mmol/L EGTA and 0.05% Triton X-100, resulting in a rapid decrease in the indo 1 ratio before appreciable loss of indicator occurred, determined by simultaneously recording absolute fluorescence at each wavelength as well as by the 410- to 490-nm ratio. To determine the intracellular Rmax, different monolayers were exposed to a solution containing 132 mmol/L KCl, 10 mmol/L K-HEPES, 1 mmol/L MgSO4, 2 μmol/L rotenone, 2 μmol/L carbonyl cyanide p-(trifluoromethoxy)phenylhydrazone (Sigma), and 10 ng/mL valinomycin (Calbiochem). HAEC monolayers were then exposed to a similar solution containing 2 μmol/L ionomycin, 69.2 mmol/L CaCl2, and 100 mmol/L N-hydroxyethylethyl-enediaminetriacetic acid (free [Ca2+] of 5900 nmol/L). In this system, the intracellular Rmin was 0.40 and the Rmax was 2.64; these were used to provide approximations of [Ca2+]i according to the following formula: [Ca2+]=Kd(R−Rmin/Rmax−R)(Sf2/Sb2),23 where Kd is the effective dissociation constant of indo 1, and Sf2 and Sb2 are the fluorescence intensities at 490 nm of Ca2+-free and Ca2+-saturated dyes, respectively. Kd was extrapolated from the published determination by using the Chelex assay by Lattanzio.24
HAEC monolayers were exposed to thapsigargin (Calbiochem), CPA (Sigma), caffeine (Sigma), histamine (Sigma), or ryanodine (Penick Corp) in a buffer containing (mmol/L) NaCl 137.0, KCl 4.9, MgSO4 1.2, NaH2PO4 1.2, HEPES 20, d-glucose 15, and CaCl2 1.5 (pH 7.40). In some experiments, the same solution without added Ca2+ and with 1 mmol/L EGTA was used to achieve Ca2+-free conditions. All experiments were performed at 23°C, since more rapid and significant intracellular loss of indo 1 occurs at 37°C than at 23°C.19
Data Analysis and Statistics
Data are reported as mean±SEM. Statistical comparisons of endothelial [Ca2+]i values among treatments of a single endothelial monolayer were performed by paired Student’s t test. Statistical comparisons of endothelial [Ca2+]i values among different monolayers were analyzed by unpaired Student’s t test. A difference was considered significant at P<.05.
Effect of Caffeine on HAEC [Ca2+]i Under Baseline Conditions
To determine the effect of caffeine on [Ca2+]i under baseline conditions, indo 1–loaded HAEC monolayers were exposed to 10 mmol/L caffeine for 16 minutes. Caffeine produced a sustained increase in the indo 1 ratio, corresponding to an increase in [Ca2+]i from 86±10 to 115±17 nmol/L (n=8, P<.05); this effect was similar to that reported previously16 when HAEC monolayers were exposed for 15 minutes to 5 μmol/L ryanodine. The increase in [Ca2+]i upon caffeine exposure was no different when extracellular Ca2+ influx was inhibited by implementing experiments in Ca2+-free buffer (Fig 1⇓; increase in [Ca2+]i, 29±13 nmol/L in Ca2+-containing buffer [n=8] and 46±12 nmol/L in Ca2+-free buffer [n=7]; P=NS). Thus, under baseline conditions, caffeine, like ryanodine,16 induces a small increase in HAEC [Ca2+]i that appears to be due to mobilization of an intracellular store.
Effect of Caffeine on HAEC [Ca2+]i After ER Ca2+ Depletion by Thapsigargin
Since, under baseline conditions, caffeine appeared to mobilize an intracellular Ca2+ store, the effect of prior depletion of an ER Ca2+ store on the effect of caffeine was examined. For these experiments, HAECs were pretreated with the irreversible ER Ca2+-ATPase inhibitor thapsigargin, which has been shown to deplete ER Ca2+ stores in endothelial cells.3 11 Under baseline conditions in the absence of thapsigargin, histamine (100 μmol/L) induces a large rapid increase in [Ca2+]i (112±7 to 929±278 nmol/L, n=5). Whereas pretreatment with thapsigargin abolished the initial rapid increase in [Ca2+]i during exposure to 100 μmol/L histamine (Fig 2⇓), the application of caffeine following thapsigargin pretreatment resulted in a large and prolonged increase in [Ca2+]i (233±48 to 512±49 nmol/L, n=4). Upon washout of caffeine and return to control buffer, [Ca2+]i returned to levels observed before caffeine exposure over ≈10 minutes. The effect of caffeine on HAEC [Ca2+]i after a more brief exposure to thapsigargin was also determined while recording indo 1 fluorescence. Superfusion of monolayers with thapsigargin for 10 minutes increased [Ca2+]i from 83±9 to 206±16 nmol/L (n=11). After a brief thapsigargin superfusion, caffeine produced an additional increase in [Ca2+]i (176±19 to 338±35 nmol/L, n=10, P<.001, Fig 3⇓), which was approximately fourfold to fivefold greater than the caffeine-induced increase in [Ca2+]i under baseline conditions. The caffeine-induced increase in [Ca2+]i following this brief thapsigargin exposure was of shorter duration than that observed following prolonged pretreatment with thapsigargin (Fig 2⇓). This effect was not specific to thapsigargin-induced ER Ca2+-ATPase inhibition, as it was also observed when monolayers were stimulated by caffeine after ER Ca2+-ATPase inhibition by CPA, an agent without apparent structural similarity to thapsigargin.3 After a 10-minute exposure to CPA (10 μmol/L), caffeine stimulated a large increase in [Ca2+]i (167±48 to 332±64 nmol/L, P<.01, n=8), which was similar to that observed after brief thapsigargin pretreatment. When monolayers were exposed to caffeine after a 10-minute pretreatment with either 0.1 or 1 μmol/L CPA, there was a trend toward a greater increase in [Ca2+]i compared with the effect of caffeine under baseline conditions (107±15 to 163±26 nmol/L after 0.1 μmol/L CPA and 155±28 to 211±43 nmol/L after 1 μmol/L CPA, n=6 for each), but these changes were not significantly different from each other or from that initiated by caffeine under control conditions (P=NS). No additional increase in [Ca2+]i was stimulated by caffeine after pretreatment with higher concentrations of CPA (data not shown). The different durations of the caffeine-induced increase in [Ca2+]i (Figs 2⇓ and 3⇓) may relate to the extent of ER Ca2+ depletion produced by the different durations of thapsigargin pretreatment, since it has previously been shown3 that pharmacological inhibition of the ER Ca2+-ATPase by thapsigargin, CPA, or BHQ causes time-dependent depletion of the endothelial internal Ca2+ store. Alternatively, the sustained caffeine-stimulated increase in [Ca2+]i may be simply a function of time after pool depletion. The transient increase in [Ca2+]i initiated by caffeine was not observed after exposures to thapsigargin of >7.5 minutes in experiments using 20- and 30-minute thapsigargin pretreatment (data not shown).
To estimate the extent of ER Ca2+ depletion produced by thapsigargin or by different concentrations of CPA, monolayers were first pretreated with thapsigargin or CPA in the absence of extracellular Ca2+ and were then exposed to ionomycin in Ca2+-free solution to release residual Ca2+ from the intracellular pool. Under baseline conditions without CPA or thapsigargin pretreatment, ionomycin (1 μmol/L in Ca2+-free buffer with 1 mmol/L EGTA) increased [Ca2+]i by 105±39 nmol/L (n=5). This was similar to the increase in [Ca2+]i stimulated by 1 μmol/L thapsigargin in Ca2+-free buffer (109±68 nmol/L, n=3, P=NS versus ionomycin). After a 7.5-minute pretreatment with 1 μmol/L thapsigargin in Ca2+-free buffer, no further increase in [Ca2+]i was elicited by ionomycin (−9±4 nmol/L, n=5), suggesting that this concentration and duration of thapsigargin pretreatment fully depleted intracellular Ca2+ stores. Similarly, ionomycin failed to increase [Ca2+]i (−3±3 nmol/L, n=3) after a 10-minute pretreatment with 10 μmol/L CPA. In contrast, ionomycin increased [Ca2+]i after pretreatment with lower concentrations of CPA. After a 10-minute exposure to either 1 μmol/L CPA (n=6) or to 0.1 μmol/L CPA (n=4), ionomycin stimulated increases in [Ca2+]i of 16±5 and 23±12 nmol/L, respectively (P<.05 versus 10 μmol/L CPA for each).
Thus, the intracellular pool appears to be fully depleted by either a 7.5-minute pretreatment with 1 μmol/L thapsigargin or a 10-minute pretreatment with 10 μmol/L CPA, since ionomycin failed to increase [Ca2+]i after either method of pool depletion. In separate experiments performed in buffer with 1.5 mmol/L Ca2+, the caffeine-stimulated increase in [Ca2+]i was similar after either of these pretreatment protocols (≈160 nmol/L). In contrast, after pretreatment with lower concentrations of CPA, which do not fully deplete the intracellular pool, there was a trend toward a greater caffeine-stimulated [Ca2+]i increase in buffer with 1.5 mmol/L Ca2+ compared with baseline conditions, but these changes were not significant.
Effect of Ryanodine on HAEC [Ca2+]i After ER Ca2+ Depletion by Thapsigargin
The presence of the ryanodine receptor was recently demonstrated in vascular and endocardial endothelium,15 and ryanodine-sensitive Ca2+ stores were found in several endothelial cell types, including HAECs.16 However, it is unknown if ryanodine and caffeine bind the same ER Ca2+ release channel in endothelial cells or whether they have similar effects on endothelial [Ca2+]i. Therefore, experiments were performed to examine whether ryanodine, like caffeine, increases HAEC [Ca2+]i after brief exposure to thapsigargin. In contrast to caffeine, under these experimental conditions ryanodine did not increase [Ca2+]i (data not shown). In other experiments performed after thapsigargin ER Ca2+ depletion, the same HAEC monolayer was first exposed to 5 μmol/L ryanodine for 15 minutes and then to 10 mmol/L caffeine; whereas ryanodine had no effect on [Ca2+]i, caffeine induced a rapid and reversible increase in [Ca2+]i (165±24 to 318±128 nmol/L, n=8, Fig 4⇓). Thus, whereas ryanodine and caffeine have similar effects on HAEC [Ca2+]i under baseline conditions, distinct effects on HAEC [Ca2+]i are revealed by prior exposure to thapsigargin. The more sustained increase in [Ca2+]i stimulated by caffeine after brief thapsigargin pretreatment combined with ryanodine exposure (Fig 4⇓) is in contrast to the more transient caffeine-stimulated increase in [Ca2+]i following brief thapsigargin pretreatment alone (Fig 3⇑). This sustained increase in [Ca2+]i was also observed after combined thapsigargin and histamine pretreatment (data not shown). As noted above, this difference in the duration of the caffeine response may relate either to the time after pool depletion or to the extent of prior ER Ca2+ depletion.
Effect of Caffeine on HAEC [Ca2+]i After Depletion of Ryanodine- and Agonist-Sensitive ER Ca2+ Pools
To determine whether the caffeine-induced increase in HAEC [Ca2+]i following thapsigargin administration was stimulated by depletion of an ER Ca2+ store or by an effect specific to thapsigargin, monolayers were exposed to caffeine after exposure to ryanodine and histamine. HAECs were exposed for 20 minutes to 5 μmol/L ryanodine in order to deplete a ryanodine-sensitive intracellular Ca2+ store. This concentration and duration of exposure were chosen on the basis of prior work,16 which showed that the ryanodine-induced increase in HAEC [Ca2+]i reached a plateau after ≈15 to 20 minutes. In contrast to the effect of caffeine after thapsigargin, after a 20-minute ryanodine exposure only a small increase in [Ca2+]i was stimulated by caffeine (142±40 to 164±40 nmol/L, n=5, P<.05). Whether caffeine and ryanodine bind the same ER Ca2+ release channel in endothelial cells or whether caffeine acts on a ryanodine-insensitive intracellular pool as occurs in hepatocytes25 is unknown. The small increase in [Ca2+]i stimulated by caffeine under these conditions (1.5 mmol/L buffer Ca2+) may also be due to extracellular influx.
In other experiments, after release of the ryanodine-sensitive Ca2+ pool, monolayers were exposed for 5 minutes to 100 μmol/L histamine, which mobilizes Ca2+ from an InsP3-sensitive intracellular store.26 27 28 Histamine stimulation after ryanodine exposure (Fig 5⇓) resulted in only a small increase in [Ca2+]i (166±21 to 184±23 nmol/L, n=8, P=NS), consistent with an effect of ryanodine on agonist-sensitive intracellular Ca2+ stores as previously reported.16 After release of the ryanodine- and histamine-sensitive intracellular Ca2+ pools, exposure to caffeine resulted in a rapid increase in [Ca2+]i (175±22 to 254±24 nmol/L, n=8, P<.01). Thus, the effect of caffeine following depletion of an ER Ca2+ store is not specific to thapsigargin-induced ER Ca2+-ATPase inhibition, although the increase in [Ca2+]i initiated by caffeine stimulation following ryanodine and histamine exposure was smaller than that observed upon caffeine stimulation following thapsigargin exposure (80±22 nmol/L [n=8] versus 162±27 nmol/L [n=10], P<.05).
Effect of Caffeine on HAEC [Ca2+]i After ER Ca2+ Depletion by Thapsigargin in Ca2+-Free Buffer
To determine the contribution of extracellular Ca2+ influx to the caffeine-induced [Ca2+]i increase following thapsigargin exposure, experiments were performed in Ca2+-free/EGTA buffer. When HAEC monolayers were exposed to thapsigargin in Ca2+-free buffer for 5 minutes (Fig 6⇓, top), [Ca2+]i increased and then gradually returned to control levels over a period of 5 to 10 minutes. The sustained increase in [Ca2+]i that occurred during thapsigargin application in the presence of extracellular Ca2+ was abolished under these experimental conditions; this appears to be due to Ca2+ influx.11 In contrast to its effect when extracellular Ca2+ was present, caffeine did not increase [Ca2+]i after thapsigargin exposure in the absence of buffer Ca2+ (Fig 6⇓, bottom). Thus, the caffeine-induced increase in [Ca2+]i after thapsigargin exposure is dependent on extracellular Ca2+.
Effect of Caffeine on HAEC [Ca2+]i After ER Ca2+ Depletion in Ca2+-Free Buffer Followed by a Brief Period of Ca2+ Repletion
The dependence on extracellular Ca2+ of the caffeine-induced [Ca2+]i increase following ER Ca2+ depletion suggests that caffeine stimulates Ca2+ influx under these conditions. Alternatively, ER Ca2+ depletion may stimulate extracellular Ca2+ influx and thereby refill a caffeine-sensitive intracellular Ca2+ pool only in the presence of buffer Ca2+. In this way, the observed dependence on extracellular Ca2+ of the caffeine-induced [Ca2+]i increase may actually reflect release of an intracellular store rather than caffeine-stimulated Ca2+ influx. To explore this possibility, HAECs were grown in 1-mm2 glass capillary tubes (total volume, 50 μL) to facilitate the rapid exchange of buffer solutions (perfusion rate, 1.1 mL/min). Monolayers were first exposed to thapsigargin in Ca2+-free buffer in order to deplete ER Ca2+ under conditions that prevent extracellular Ca2+ influx. The subsequent exposure to caffeine in Ca2+-free buffer did not increase [Ca2+]i (Fig 7⇓). The perfusion solution was then rapidly changed to HEPES buffer with 1.5 mmol/L Ca2+ for 2 minutes in an attempt to refill both the cytosol and intracellular Ca2+ stores not irreversibly depleted by thapsigargin. This attempt would likely be facilitated by the Ca2+ concentration gradient between the cytosolic and intracellular compartments. Changing to a Ca2+-containing solution resulted in a rapid increase in [Ca2+]i as extracellular Ca2+ influx filled the cytosol. If the repleted cytosolic compartment then filled the caffeine-sensitive Ca2+ pool, the subsequent exposure to caffeine, even in the absence of extracellular Ca2+, would be expected to increase [Ca2+]i by releasing a caffeine-sensitive intracellular store. When monolayers were exposed to caffeine in Ca2+-free buffer at this time (Fig 7A⇓), no increase in [Ca2+]i was observed; in fact, [Ca2+]i rapidly decreased from the higher level in buffer with 1.5 mmol/L Ca2+. In contrast, when caffeine exposure occurred in buffer with 1.5 mmol/L Ca2+ (Fig 7B⇓), a further increase in [Ca2+]i occurred (326±95 nmol/L, n=5). The data suggest that after ER Ca2+ depletion, caffeine stimulates extracellular Ca2+ influx.
Comparison of the Effect of Histamine With That of Caffeine on HAEC [Ca2+]i After ER Ca2+ Depletion in Ca2+-Free Buffer Followed by a Brief Period of Ca2+ Repletion
The large rapid Ca2+ influx following ER Ca2+ depletion was observed with caffeine but not with ryanodine. To determine whether this effect is seen upon histamine stimulation, experiments were performed under conditions similar to those previously described in Fig 7⇑. HAEC monolayers grown in glass capillary tubes were superfused with Ca2+-free buffer and then exposed to thapsigargin in order to deplete the ER Ca2+ store (Fig 8⇓). The solution was then rapidly changed to one with 1.5 mmol/L Ca2+ for 2 minutes, resulting in a rapid increase in [Ca2+]i. When HAECs were then exposed to histamine in Ca2+-free buffer, [Ca2+]i rapidly returned to levels noted before exposure to Ca2+-containing buffer. An increase in [Ca2+]i would have been expected if the histamine (InsP3)–sensitive intracellular Ca2+ store had been refilled during exposure to buffer with 1.5 mmol/L Ca2+. Thus, although exposure to Ca2+-containing buffer rapidly filled the cytosol, it did not fill the histamine-sensitive intracellular Ca2+ store, which had most likely been irreversibly depleted by thapsigargin.
HAEC monolayers were then exposed again to Ca2+-containing buffer (Fig 8⇑) and then to histamine in this same buffer solution. No increase in [Ca2+]i was observed when monolayers were exposed to histamine at this time; indeed [Ca2+]i decreased to a “new” baseline in the Ca2+-containing solution. However, when monolayers were exposed to caffeine after return to the new baseline [Ca2+]i, a rapid increase in [Ca2+]i (110±6 to 306±62 nmol/L, n=4, P<.05) was observed. Thus, caffeine, but not histamine, induces extracellular Ca2+ influx under these experimental conditions.
The present study documents the existence of a caffeine-stimulated Ca2+ influx in human endothelial cells that is regulated by the status of Ca2+ loading of ER Ca2+ stores. Under baseline conditions, when ER Ca2+ stores are full, caffeine induces a small increase in HAEC [Ca2+]i due to release of an intracellular Ca2+ store. This effect is similar to that of ryanodine.16 In contrast, when ER Ca2+ stores have been emptied, caffeine promotes a rapid increase in HAEC [Ca2+]i due principally to extracellular Ca2+ influx. This effect on Ca2+ influx is not observed with ryanodine or with histamine.
Although the effects of caffeine on [Ca2+]i in excitable cells are well established, the role of caffeine in endothelial cell Ca2+ regulation has only recently been examined. In the present study, the increase in HAEC [Ca2+]i stimulated by caffeine under baseline conditions was not altered when extracellular Ca2+ was absent (Fig 1⇑). This suggests the existence of a caffeine-sensitive intracellular Ca2+ store and is consistent with the findings of other recent studies.17 18 By recording the activity of Ca2+-activated K+ channels in bovine aortic endothelial cell monolayers, Thuringer and Sauve17 demonstrated caffeine-induced Ca2+ release from an InsP3-insensitive intracellular store. Application of caffeine opened Ca2+-activated K+ channels in cell-attached patch-clamp experiments in both Ca2+-free and Ca2+-containing external solutions.17 In another study of freshly dissociated single rabbit aortic endothelial cells,18 caffeine induced a transient rise in [Ca2+]i in fura 2–loaded cells in the presence and absence of extracellular Ca2+. A dose-dependent increase in unitary current activity superimposed on a large prolonged transient outward current upon application of caffeine was documented by whole-cell membrane currents recorded under voltage-clamp conditions.18 The caffeine-induced transient outward current was not affected by heparin, an InsP3 receptor antagonist. Thus, caffeine appears to release an InsP3-insensitive intracellular Ca2+ store, which may be an important source of Ca2+ for activation of K+ channels in endothelial cells.
In muscle cells, both caffeine and ryanodine bind the SR Ca2+ release channel12 13 but appear to have different mechanisms of action. At low (nanomolar) concentrations in experiments using isolated SR vesicles, ryanodine maintains the Ca2+ release channel in an open substate conductance, whereas at high concentrations, ryanodine appears to inhibit channel activation.11 Caffeine opens SR Ca2+ release channels,12 promotes discharge of Ca2+ stores,13 and activates Ca2+-induced Ca2+ release. Caffeine appears to increase the affinity of the activation site of a Ca2+-permeable SR channel for Ca2+29 and to increase the number and duration of open-channel events.30 The present study shows that in HAECs under baseline conditions, caffeine results in an increase in [Ca2+]i similar to that induced by ryanodine in this cell type.16 However, caffeine produces an additional small increase in HAEC [Ca2+]i after ryanodine stimulation. Since this experiment was performed in the presence of extracellular Ca2+, it is uncertain whether this effect is due to Ca2+ release from a ryanodine-insensitive intracellular pool as occurs in hepatocytes25 or from extracellular influx. The endothelial ER was recently found to contain a protein having homology with the ryanodine receptor/Ca2+ release channel of muscle.15 The pattern of immunofluorescence staining of endothelial cells labeled with a specific anti–ryanodine receptor antibody was consistent with the distribution of the ER.15 However, whether caffeine and ryanodine bind the same site on the ER of endothelial cells is not known.
When thapsigargin was used to deplete ER Ca2+ in the present study, caffeine did not increase [Ca2+]i when buffer Ca2+ was not present. Thus, thapsigargin appears to deplete the caffeine-sensitive Ca2+ store in HAECs, suggesting that the caffeine-sensitive intracellular Ca2+ store is a part of the ER Ca2+ store. This may not be the case in all cell types. In Jurkat T-lymphocytes,31 caffeine (15 mmol/L) increased [Ca2+]i when added after thapsigargin (1 μmol/L) in the absence of extracellular Ca2+. Under these conditions, thapsigargin abolished the response to the monoclonal anti-CD3 antibody OKT3, which releases Ca2+ from an InsP3-sensitive pool in Jurkat T-lymphocytes,31 suggesting that the caffeine-sensitive Ca2+ pool in this cell type is not part of the ER Ca2+ store.
The endothelial caffeine-sensitive intracellular Ca2+ store appears to require access, either directly or indirectly, to extracellular Ca2+. Recently, a small transient caffeine-induced hyperpolarization was recorded with microelectrodes from an intact guinea pig coronary artery preparation.32 When buffer Ca2+ was removed, the hyperpolarization decreased progressively with each successive application of caffeine, suggesting that the caffeine-sensitive pool was dependent on extracellular Ca2+ for its repletion. Similarly, exposure to caffeine produced repetitive [Ca2+]i spikes in fura 2–loaded human umbilical vein endothelial cells33 but only in the presence of extracellular Ca2+.
In the present study, the caffeine-sensitive intracellular Ca2+ store was depleted by the ER Ca2+-ATPase inhibitor thapsigargin. In Ca2+-free solution, caffeine stimulation did not increase [Ca2+]i after exposure to thapsigargin. If monolayers were then superfused with Ca2+-containing solution, a rapid increase in [Ca2+]i occurred as extracellular Ca2+ influx filled the cytosol (Fig 7⇑). This might be expected to fill an intracellular store not irreversibly depleted by thapsigargin, because of the Ca2+ concentration gradient between the cytolic and intracellular compartments. The absence of [Ca2+]i increase when monolayers were subsequently exposed to caffeine in Ca2+-free buffer (Fig 7A⇑) suggests that the caffeine-sensitive intracellular Ca2+ store was still empty at this time. In contrast, when caffeine exposure occurred in buffer with 1.5 mmol/L Ca2+, a further increase in [Ca2+]i occurred, consistent with caffeine-stimulated extracellular Ca2+ influx. This caffeine-stimulated Ca2+ influx was not observed with either ryanodine (Fig 4⇑) or with histamine (Fig 8⇑). Although increases in [Ca2+]i assessed by changes in indo 1 fluorescence could reflect either Ca2+ influx or inhibition of Ca2+ efflux, the absence of a caffeine effect in Ca2+-free buffer following intracellular store depletion suggests that Ca2+ influx is responsible for the effect of caffeine following store depletion in Ca2+-containing solution.
There are several possible mechanisms of the caffeine-stimulated Ca2+ influx after intracellular pool depletion in HAECs. One possibility is that caffeine modulates the depletion-dependent Ca2+ influx described by other investigators.3 11 In the present study, the magnitude of Ca2+ influx stimulated by caffeine correlates, at least in part, with the extent of prior ER Ca2+ depletion, since the increase in [Ca2+]i initiated by caffeine was greater after pretreatment with concentrations of thapsigargin or CPA that appeared to fully deplete intracellular Ca2+ pools than it was after pretreatment with concentrations of CPA that did not fully deplete intracellular stores. On the basis of experiments performed with ionomycin administration following thapsigargin or CPA pretreatment in Ca2+-free buffer, the intracellular Ca2+ store appears to be emptied by either a 7.5-minute pretreatment with 1 μmol/L thapsigargin or a 10-minute pretreatment with 10 μmol/L CPA. In separate experiments performed in buffer with 1.5 mmol/L Ca2+, the caffeine-stimulated increase in [Ca2+]i was similar (≈160 nmol/L) after each method of store depletion. Lower concentrations of CPA appeared to only partially deplete the intracellular store, since ionomycin caused an additional increase in [Ca2+]i after a 10-minute pretreatment with either 0.1 or 1 μmol/L CPA. After pretreatment with either of these partially depleting concentrations of CPA, there was a trend toward a greater caffeine-stimulated increase in [Ca2+]i than under baseline conditions, but these changes were not significantly different from control.
The relation of plasmalemmal Ca2+ influx to the degree of filling of intracellular Ca2+ stores has been described by Putney’s “capacitative” pathway.34 Evidence in support of this model was recently provided by patch-clamp recordings in bovine aortic endothelial cells.35 By use of different concentrations of the ER Ca2+-ATPase inhibitor BHQ, inward Ca2+ currents were found to vary with the extent of depletion of intracellular Ca2+ stores. Although the signal responsible for communicating the level of the intracellular store to the plasmalemmal influx pathway was not identified in that or the present study, it has been suggested that the intracellular store and the influx pathway are spatially in close association.35 Whether the caffeine-sensitive intracellular compartment in endothelial cells is in close association with an influx pathway is unknown. Alternatively, caffeine may stimulate the production of a second messenger that signals the opening of plasmalemmal Ca2+ channels only when intracellular stores are depleted.
Although capacitative Ca2+ influx may be involved in the effect of caffeine described in the present study, other processes that stimulate Ca2+ entry in endothelial cells may be enhanced by caffeine. Extracellular Ca2+ influx may occur by at least four other mechanisms, which have recently been reviewed by Adams et al.18 These include a passive Ca2+ “leak,” mechanosensitive (stretch-activated) Ca2+ channels, Na+-Ca2+ exchange, and receptor-mediated Ca2+ influx through receptor-operated or second messenger (eg, inositol 1,3,4,5-tetrakis-phosphate)–activated channels. The effect of caffeine is unlikely to involve dihydropyridine-sensitive L-type Ca2+ channels, since electrophysiological and unidirectional 45Ca2+ flux measurements indicate an absence of these channels in cultured endothelial cells.18 36 37 Although an effect of caffeine on [Ca2+]i mediated by an increase in cAMP through its inhibition of phosphodiesterase cannot be excluded, this appears to be an unlikely mechanism of the rapid caffeine-stimulated Ca2+ influx described in the present study. In concentrations up to 10 mmol/L, caffeine is only a weak inhibitor of endothelial phosphodiesterases, with a potency significantly less than that of other inhibitors of this enzyme, such as 3-isobutyl-1-methylxanthine (W.F. Graier, unpublished data, 1995), which has no effect on endothelial [Ca2+]i.38
The data from the present study and other work suggest that the caffeine-sensitive intracellular Ca2+ store, like the ryanodine- and InsP3-sensitive stores, is part of the ER Ca2+ store in endothelial cells. Although the caffeine-sensitive and InsP3-sensitive stores may be distinct,17 18 further investigation is necessary to determine whether ryanodine and caffeine bind the same putative Ca2+ release channel in endothelial cells. The present report shows that caffeine stimulates Ca2+ entry in endothelial cells and that this Ca2+ influx is regulated by the status of ER Ca2+ stores such that caffeine promotes Ca2+ entry when the stores are depleted, regardless of the mechanism of pool depletion. Although caffeine, ryanodine, and histamine each affect intracellular Ca2+ pools, the rapid large influx of Ca2+ following thapsigargin exposure is initiated only by caffeine and not by the other two agonists. The functional role of caffeine-stimulated Ca2+ influx requires further study.
Selected Abbreviations and Acronyms
|HAEC||=||human aortic endothelial cell|
|Rmax||=||maximum fluorescence ratio|
|Rmin||=||minimum fluorescence ratio|
- Received October 20, 1994.
- Accepted July 21, 1995.
- © 1995 American Heart Association, Inc.
Colden-Stanfield M, Schilling WP, Ritchie AK, Eskin SG, Navarro LT, Kunze DL. Bradykinin induced increases in cytosolic calcium and ionic currents in cultured bovine aortic endothelial cells. Circ Res.. 1987;61:632-640.
Pollock WK, Wreggett KA, Irvine RF. Inositol phosphate production and Ca2+ mobilization in human umbilical vein endothelial cells stimulated by thrombin and histamine. Biochem J.. 1988;256:371-376.
Schilling WP, Cabello OA, Rajan L. Depletion of the inositol 1,4,5-trisphosphate sensitive intracellular Ca2+ store in vascular endothelial cells activates the agonist sensitive Ca2+ influx pathway. Biochem J.. 1992;284:521-530.
Hallam T, Jacob R, Merritt JE. Evidence that agonists stimulate bivalent cation influx into human endothelial cells. Biochem J.. 1988;255:179-184.
Hallam TJ, Pearson JD, Needham LA. Thrombin stimulated elevation of human endothelial cell cytoplasmic free calcium concentration causes prostacyclin production. Biochem J.. 1988;251:243-249.
Thastrup OP, Cullen PJ, Drobak MR, Hanley MR, Dawson AP. Thapsigargin, a tumor promoter, discharges intracellular Ca2+ stores by specific inhibition of the endoplasmic reticulum Ca2+-ATPase. Proc Natl Acad Sci U S A.. 1990;87:2466-2470.
Sagara Y, Inesi G. Inhibition of the sarcoplasmic reticulum Ca2+ transport ATPase by thapsigargin at subnanomolar concentrations. J Biol Chem.. 1991;266:13503-13506.
Bian J, Ghosh TK, Wang JC, Gill DL. Identification of intracellular Ca2+ pools: selective modification by thapsigargin. J Biol Chem.. 1991;266:8801-8806.
Dolor JR, Hurtwitz LM, Mirza Z, Strauss HC, Whorton AR. Regulation of extracellular Ca2+ entry in endothelial cells: role of intracellular Ca2+ pool. Am J Physiol.. 1992;262:C171-C181.
Lai FA, Misra M, Xu L, Smith HA, Meissner G. The ryanodine receptor-Ca2+ release channel complex of skeletal muscle sarcoplasmic reticulum. J Biol Chem.. 1989;264:16776-16785.
Endo M. Calcium release from the sarcoplasmic reticulum. Physiol Rev.. 1977;57:71-108.
Fabiato A. Calcium-induced release of calcium from the cardiac sarcoplasmic reticulum. Am J Physiol.. 1983;245:C1-C14.
Lesh RE, Marks AR, Somlyo AV, Fleischer S, Somlyo AP. Antiryanodine receptor antibody binding sites in vascular and endocardial endothelium. Circ Res.. 1993;72:481-488.
Ziegelstein RC, Spurgeon HA, Pili R, Passaniti A, Cheng L, Corda S, Lakatta EG, Capogrossi MC. A functional ryanodine-sensitive intracellular Ca2+ store is present in vascular endothelial cells. Circ Res.. 1994;74:151-156.
Adams DJ, Rusko J, van Slooten G. Calcium signaling in vascular endothelial cells: Ca2+ entry and release. In: Weir EK, Hume JR, Reeves JT, eds. Ion Flux in Pulmonary Vascular Control. New York, NY: Plenum Publishing Corp; 1993:259-275.
Spurgeon HA, Stern MD, Baartz G, Raffaeli S, Hansford RG, Talo A, Lakatta EG, Capogrossi MC. Simultaneous measurement of Ca2+, contraction, and potential in cardiac myocytes. Am J Physiol.. 1990;258:H574-H586.
Ziegelstein RC, Cheng L, Blank PS, Spurgeon HA, Lakatta EG, Hansford RG, Capogrossi MC. Modulation of calcium homeostasis in cultured rat aortic endothelial cells by intracellular acidification. Am J Physiol.. 1993;265:H1424-H1433.
Grynkiewicz G, Poenie M, Tsien RY. A new generation of Ca2+ indicators with greatly improved fluorescence properties. J Biol Chem.. 1985;260:3440-3450.
McNulty TJ, Taylor CW. Caffeine-stimulated Ca2+ release from the intracellular stores of hepatocytes is not mediated by ryanodine receptors. Biochem J.. 1993;291:799-801.
Pollock WK, Wreggett KA, Irvine RF. Inositol phosphate production and Ca2+ mobilization in human umbilical-vein endothelial cells stimulated by thrombin and histamine. Biochem J.. 1988;256:371-376.
Nagasaki K, Kasai M. Channel selectivity and gating specificity of calcium-induced calcium release channel in isolated sarcoplasmic reticulum. J Biochem.. 1984;96:1769-1775.
Guse AH, Roth E, Emmrich F. Intracellular Ca2+ pools in Jurkat T-lymphocytes. Biochem J.. 1993;291:447-451.
Vaca L, Kunze DL. Depletion and refilling of intracellular Ca2+ stores induce oscillations of Ca2+ current. Am J Physiol.. 1993;264:H1319-H1322.