Ca2+ removal mechanisms in freshly isolated rabbit aortic endothelial cells

Ca2+ removal mechanisms in freshly isolated rabbit aortic endothelial cells

Research Cell Calcium (2002) 31(6), 265–277 0143-4160/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0143-416...

430KB Sizes 0 Downloads 13 Views

Research

Cell Calcium (2002) 31(6), 265–277 0143-4160/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0143-4160(02)00075-1, available online at http://www.idealibrary.com on

Ca2+ removal mechanisms in freshly isolated rabbit aortic endothelial cells X. Wang, S. Reznick, P. Li, W. Liang, C. van Breemen Department of Pharmacology and Therapeutics, Vancouver Vascular Biology Research Center, University of British Columbia, Vancouver, Canada V6T 1Z3

Summary Calcium removal from the cytoplasm was investigated in freshly isolated aortic endothelial cells by monitoring changes in intracellular calcium ([Ca2+ ]i ) using ratiometric fura-2 fluorimetry. Blockade of the Na+ /Ca2+ exchanger (NCX) by replacement of external sodium with equi-molar N-methyl-d-glutamine (0Na PSS) decreased the removal rate by 52%. Blockade of the sarco/endoplasmic reticulum Ca2+ ATPase (SERCA) by cyclopiazonic acid (CPA) decreased the removal rate by 50%. Simultaneous application of CPA and 0Na PSS did not reduce the removal rate any further (53%). The lack of additivity of these two procedures, suggests that SERCA and the NCX function in series to lower [Ca2+ ]i . In addition, in the absence of extracellular Ca2+ , removal of external Na+ markedly reduced the rate of loss of Ca2+ from the ER further supporting the hypothesis that NCX is functionally linked to ER calcium release channels, and thus, plays an important role in ER calcium unloading. To investigate the mechanism for the coupling of NCX and SERCA, the same protocols as described above were repeated after treating the cells with cytochalasin D, which disrupts the cytoskeleton. This treatment uncoupled the NCX from SERCA, as evidenced by the resulting additive inhibitory effects of application of CPA and removal of extracellular Na+ on the rate of Ca2+ removal from the cytoplasm. These data suggest that in endothelial cells NCX and SERCA function in series to remove about half of the free Ca2+ from the cytosol, while PMCA contributes to the other half of the Ca2+ removal process. © 2002 Elsevier Science Ltd. All rights reserved.

INTRODUCTION Vascular endothelial cells release various paracrine substances upon chemical, hormonal and physical stimulation, which serve to regulate smooth muscle tone and prevent coagulation of platelets and leukocytes [1]. Nitric oxide (NO) and prostacyclin have both vasodilatory and antithrombotic effects [2,3], while endothelium derived hyperpolarizing factor (EDHF) appears to mainly regulate smooth muscle function [4]. Endothelial cells also release vasoconstrictive agents such as thromboxane, angeotensin and endothelin [5]. Although the stimulus for the secretion of all these vasoactive compounds is elevation of [Ca2+ ]i [6,7], it is not yet clear what aspect of Ca2+ signaling determines the coupling to a specific product. Received 28 November 2001 Revised 22 March 2002 Accepted 26 March 2002 Correspondence to: Xiaodong Wang, Department of Pharmacology and Therapeutics, University of British Columbia, 2176 Health Sciences Mall, Vancouver, BC, Canada V6T 1Z3. Tel.: +1-604-822-2447; e-mail: [email protected]

The intracellular free Ca2+ concentration ([Ca2+ ]i ) is tightly regulated by a variety of ion channels and pumps. The mechanisms responsible for supplying Ca2+ during endothelial cell activation, including capacitative Ca2+ influx and Ca2+ release from IP3 sensitive intracellular stores, have been the focus of many studies [8]. In contrast only a few studies have dealt with calcium extrusion from endothelial cells [9–12]. In general, at least three mechanisms of lowering [Ca2+ ]i are present in each cell: the plasma membrane Ca2+ ATPase (PMCA), the sarcoplasmic/endoplasmic reticulum Ca2+ ATPase (SERCA) and the Na+ /Ca2+ exchanger (NCX). Although all three mechanisms have been identified in vascular endothelial cells [13,14], the relative contribution of each of the components to overall calcium extrusion and sequestration remains to be elucidated. Recent observations in vascular smooth muscle and endothelial cells have suggested that the micro-structural physical localization of Ca2+ pumps and channels in the ER and plasmamembrane (PM) relative to each other may play a critical role in determining the precise patterns of Ca2+ fluxes and the creation of cytoplasmic Ca2+ gradients [12,15–17]. The physiological relevance of these 265

266

X Wang, S Reznick, P Li, W Liang, C van Breemen

microscopic Ca2+ gradients may be to regulate diverse cellular functions through varying spatial and temporal patterns of the same second messenger, Ca2+ [15,16]. The existence of a restricted subplasmalemmal space between the PM and SR, where high local concentrations of calcium can reside without affecting global [Ca2+ ]i , has recently been identified in cultured endothelial cells [17]. Furthermore, recent studies in cultured endothelial cells have identified functional coupling between the NCX and the ER Ca2+ release channels, of the same type as found in smooth muscle, which allow efficient calcium unloading from the ER to the extracellular space [12]. In the current study, we determine the contributions of SERCA, PMCA and NCX to calcium removal from native endothelial cells freshly isolated from the rabbit aorta.

MATERIALS AND METHODS

Treatment of endothelial cells with cytochalasin D Fifty micromolar cytochalasin D (Sigma) was used to pre-treat freshly isolated cells, in order to disrupt the cytoskeleton. Different time periods were tested from 30 min to 1.5 h after the cells are seeded on cover-slips. The treatment time included the 0.5 h loading time with fura-2.

Staining of cytoskeleton using fluorescent phalloidin Freshly isolated endothelial cells before and after cytochalasin D treatment were fixed using 3.7% formaldehyde and extracted with 0.1% X-100 solution. After fixation the cells were loaded for 20 min with fluorescent phalloidin (Sigma) and viewed on a laser-scanning confocal microscope (UltraView System, Perkin Elmer) using a 60× objective.

Isolation of endothelial cells Endothelial cells were obtained from the thoracic aorta of female New Zealand White rabbits (2–2.5 kg) as reported previously [8]. The rabbits were killed with CO2 asphyxiation and exsanguinated. The thoracic aorta was rapidly removed and isolated from surrounding fat and connective tissue under a dissecting microscope. The aorta was placed in a test tube containing Ca2+ -free PSS with 0.1 mg/ml collagenase, 0.1% elastase, and 1 mg/ml bovine serum albumin for 35 min at 37 ◦ C. Endothelial cells were dispersed by trituration using a Pasteur pipette. The endothelial cells were then incubated for 2 h in Antibiotic PSS solution. The endothelial cells were seeded on a glass coverslip pre-coated with poly-d-lysine at room temperature and used within 5 h.

[Ca2+ ]i measurement using fura-2 fluorescence [Ca2+ ]i was measured using a microscope-based fluorimeter from Photon Technology International (London, Canada). The endothelial cells on the coverslip were loaded with 1 ␮M fura-2/AM (acetoxymethylester) and 1 ␮M pluronic acid (F-127) in nPSS for 30 min at room temperature. The coverslip was mounted on a Nikon inverted microscope (Nikon, Diaphot). The cells were excited alternately at 340 and 380 nm and the light collected at 510 nm. The ratio of the two intensities at 340 and 380 nm (F340/F380) constitutes a relative measurement of [Ca2+ ]i . No calibration are made because the existing uncertainty of the dissociation constant of fura-2 in the intracellular milieu. The experimental solution was infused through PE tubing via gravity. The fluid level was maintained by suction. Three milliliter of solution was used for each solution exchange to assure complete wash out. The time of each exchange is less than 15 s. Cell Calcium (2002) 31(6), 265–277

Data preparation and statistics Experimentation revealed variance between Ca2+ removal rates measured on different batches of cells, hence all four treatments (Control, NCX inhibition, SERCA inhibition, and combined NCX and SERCA inhibition) were performed on each set of cells without cytochalasin D treatment. In the case of cytochalasin D treatment, however, cells were viable in most cases only for three experimental protocols. Therefore, all data are normalized to the maximal response in [Ca2+ ]i elevation elicited by the 0Na-CPA protocol, which consistently displayed the highest peak in [Ca2+ ]i . Cells were allowed to recover for at least 15 min in nPSS solution in between each protocol and the order of application was randomly selected to eliminate possible artifacts. The calcium levels during the time course of each treatment are expressed as percentages of the maximal response using 0Na-CPA protocol in Figs. 1, 2a, 5 and 6. Two methods were used to determine the rates of Ca2+ removal. (1) In Figs. 2a and 5b, an exponential decay function of the form: y = aebt was fitted to the normalized data in the declining phase. The rate constants were then normalized to that of ACh controls. (2) In Figs. 3 and 5d, the rate of decrease of the response were calculated by means of ([Ca2+ ]/t ). The rates were plotted as percent of the maximal Ca2+ removal rate under control conditions. To be able to compare between different experimental protocols, all the data are binned at certain Ca2+ levels (5, 10, 15, 20, 40, 60 and 80% of the maximal responses) and then averaged. Statistics were performed on Ca2+ removal rates expressed as percentages of control. Data from multiple experiments are given as mean ± S.E.M. t -Test and ANOVA are performed to determine the statistical significance. © 2002 Elsevier Science Ltd. All rights reserved.

Ca2+ removal mechanisms

Solutions and chemicals Both fura-2/AM and pluronic acid (F-127) were purchased from Molecular Probes (Eugene, OR, USA). All other materials were purchased from Sigma Chemicals (St. Louis, MO, USA). Nominally free Ca2+ PSS solution used for enzyme treatment contained (mM) NaCl 126, KCl 5, MgCl2 1.2, HEPES 10, d-glucose 10. Normal PSS (nPSS) solution contained

267

(mM) NaCl 126, KCl 5, MgCl2 1.2, HEPES 10, d-glucose 10, and CaCl2 1.0 mM. Zero Ca2+ PSS (0Ca PSS) solution contained (mM) NaCl 126, KCl 5, MgCl2 1.2, HEPES 10, d-glucose 10, and EGTA 0.1. Zero Na+ PSS (0Na PSS) contained (mM) N-methyl-d-glucamine (NMDG) 126, KCl 5, MgCl2 1.2, HEPES 10, d-glucose 10, and CaCl2 1. Zero Na+ /zero Ca2+ solution (0Na/0Ca PSS) contained (mM) NMDG 126, KCl 5, MgCl2 1.2, HEPES 10, d-glucose 10, and EGTA 0.1. All solutions were adjusted to pH 7.4.

Fig. 1 Effects of inhibition of NCX and SERCA on lowering of [Ca2+ ]i . Changes in [Ca2+ ]i (F340/F380) were monitored during stimulation with ACh (10 ␮M) (top left), followed by washout of the chamber with 0Ca PSS. The same protocols were repeated in the absence of extracellular Na+ (top, right) to remove the contribution of NCX to calcium removal. In the bottom graphs, 30 ␮M CPA was used to abolish the contribution of SERCA to Ca2+ removal. The control ACh response is included as dotted lines with the traces for CPA and 0Na PSS. In the bottom right figure, the CPA response is shown as a dashed line. Either CPA or 0Na PSS slowed down the rate of Ca2+ removal, however, both treatments applied concurrently did not decrease the removal rate further than 0Na or CPA alone. © 2002 Elsevier Science Ltd. All rights reserved.

Cell Calcium (2002) 31(6), 265–277

268

X Wang, S Reznick, P Li, W Liang, C van Breemen

Fig. 2 Effects of 0Na, CPA and a combination of both on the Ca2+ removal rates. (a) The decay of [Ca2+ ]i was plotted for all protocols including the control. A one-component exponential decay function was fitted to the curve. The rate constant is expressed as fraction/second. (b) The rate constant of the [Ca2+ ]i decay was normalized to that of the ACh response. Removal of the NCX decreased the removal rate by 52 ± 7%. SERCA inhibition decreased the decay rate by 50 ± 5%, while simultaneous removal of SERCA and NCX decreased the Ca2+ removal rate by 53 ± 6%. The data is representative of six replicates. () P > 0.05 vs. CPA, () P < 0.025 vs. control and (∗) P < 0.0003 vs. CPA, 0Na PSS, and 0Na-CPA treated tissue by analysis of variance (ANOVA) and Newman–Keuls test.

Cell Calcium (2002) 31(6), 265–277

© 2002 Elsevier Science Ltd. All rights reserved.

Ca2+ removal mechanisms

All solutions were superfused through the experimental chamber and a vacuum suction pump was used in order to keep a constant fluid level. Experiments were carried out at room temperature (23 ◦ C).

RESULTS The endothelial agonist acetylcholine (ACh) was applied to nPSS to stimulate Ca2+ influx and release. After [Ca2+ ]i reached peak values, both Ca2+ and the agonist were removed from the bathing solution, which caused a rapid decrease of [Ca2+ ]i to baseline levels. The rates of decline in [Ca2+ ]i are compared between different experimental groups. The results are divided into two parts: Ca2+ removal process before and after disruption of the cytoskeleton with cytochalasin D. Part I: Ca2+ removal process under normal condition Fig. 1 shows a typical experiment where the decline of [Ca2+ ]i was monitored. Removal of extracellular calcium after [Ca2+ ]i has reached its peak value in response to administration of ACh (10 ␮M) caused a decline in [Ca2+ ]i to baseline. To determine the role of the ER in Ca2+ removal, the same experiment was repeated in the presence of the

269

SERCA blocker CPA (30 ␮M) (Fig. 1, left-bottom). The application of CPA caused a small increase in steady state [Ca2+ ]i and the subsequent ACh application induced a larger [Ca2+ ]i increase as compared to the control. This is consistent with our previous finding [16], which suggests that CPA increases Ca2+ influx and prevents partial reuptake of cytoplasmic Ca2+ by the ER. The calcium removal rate was decreased in the presence of CPA, when compared with the control (dotted line). Since the NCX has been shown to be present in vascular endothelial cells [17], we designed experiments to abolish its contribution to Ca2+ extrusion. Fig. 1 (top, right part) shows the inhibition of calcium extrusion after blockade of NCX by replacement of extracellular Na+ with NMDG. In order to determine whether the NCX and SERCA function in parallel or in series to remove intracellular calcium, both SERCA and NCX were inactivated simultaneously. A typical result of the combined effect of CPA and Na+ removal is shown in Fig. 1 (bottom, right part). Although, in this case, the peak of the ACh response was greater, the rate of [Ca2+ ]i decrease was not significantly different from that obtained by blockade of SERCA alone (also see Fig. 2a). Fig. 2a shows the declining phase of [Ca2+ ]i after removal of extracellular Ca2+ under the four different conditions as described in Fig. 1. An one-component exponential

Fig. 3 Rates of Ca2+ removal as a function of [Ca2+ ]i . The rate of Ca2+ removal, [Ca2+ ]/t, was calculated from the data of Fig. 2a, and plotted against [Ca2+ ]i . To facilitate comparison between different cells, all the [Ca2+ ]i values were normalized to the maximal response of 0Na-CPA from the same cell. After normalizing the data were binned to 0–5, 10, 15, 20, 40, 60 and 80% of the maximal response and the mean value and S.E.M. are calculated after binning. © 2002 Elsevier Science Ltd. All rights reserved.

Cell Calcium (2002) 31(6), 265–277

270

X Wang, S Reznick, P Li, W Liang, C van Breemen

decay function was fitted to the data, and the rate constants are shown as fraction/second. Fig. 2b summarizes the pooled data as percentages of the maximal Ca2+ extrusion rate seen under control conditions. Complete inhibi-

tion of NCX decreased the Ca2+ removal rate by an average of 52 ± 7%. SERCA blockade decreased the Ca2+ removal rate by 50 ± 5% and simultaneous inactivation of NCX and SERCA decreased the extrusion rate by 53 ± 6% compared

Fig. 4 Unloading of the ER is inhibited when NCX is not functioning. (a) and (b) Cells were first activated with 10 ␮M ACh in nPSS (a) or 0Na PSS (b). Seventy-five seconds after removal of extracellular calcium and ACh, cells were reactivated with a second dose of 10 ␮M ACh. The peak response provides a measure of Ca2+ content of the ER under each of the conditions. Each trace is representative of four experiments. (c) Comparison of the above Ca2+ release signals in the absence and presence of extracellular Na+ . The amplitude of the second ACh-induced [Ca2+ ]i peak is given as a percentage of the initial control response to ACh. When NCX is inactivated using 0Na/0Ca PSS, the ER calcium was 43 ± 13%, compared with only 11 ± 3% in cells with functioning NCX. Traces are representative of four experiments. (∗) P < 0.015 vs. control using non-paired t-test. Cell Calcium (2002) 31(6), 265–277

© 2002 Elsevier Science Ltd. All rights reserved.

Ca2+ removal mechanisms

271

Fig. 4 (Continued ).

to the control (the values for 0Na, CPA and 0Na-CPA were not significantly different from each other, but were significantly different from the control value). Since the activity of the various Ca2+ pumps is dependent on [Ca2+ ]i we compared the Ca2+ removal rate as a function of the relative cytoplasmic Ca2+ concentration. We have expressed all values of [Ca2+ ]i as a percentage of the maximal response elicited by the 0Na-CPA protocol. The rate of Ca2+ decline was calculated as [Ca2+ ]/t. In order to compare data from different cells, all rates were binned at 0–5, 5–10, 10–15, 15–20, 20–40, 40–60, 60–80 and 80–100% of the maximal responses. The mean values are determined using the binned data and are shown in Fig. 3. It is again clear that the values of [Ca2+ ]/t in the presence of CPA, 0Na or 0Na-CPA are different from those of the control, but are not significantly different from each other. The above data suggests that in this cell preparation, NCX and SERCA are functionally coupled in series to serve as a major (approximately 50%) pathway for Ca2+ removal. This Ca2+ removal process would be composed of first Ca2+ uptake into the ER by SERCA followed by release of Ca2+ towards the NCX and subsequent Ca2+ extrusion by NCX. According to this hypothesis interruption of the last step by Na+ removal from the extracellular space should lead to greater Ca2+ accumulation in the ER. This hypothesis was tested in the experiment illustrated in Fig. 4. The cells were stimulated first using 10 ␮M ACh in the presence of extracellular Ca2+ . Once the Ca2+ level reached its peak, extracellular Ca2+ was removed and the [Ca2+ ]i re© 2002 Elsevier Science Ltd. All rights reserved.

turned to the baseline. After the extrusion period the Ca2+ remaining in the ER was measured by applying ACh for a second time in 0Ca2+ PSS. This procedure revealed that very little Ca2+ was left in the ER (second peak in Fig. 4a). However, when this same experiment was repeated in the absence of extracellular Na+ (Fig. 4b) much more Ca2+ was retained in the ER, supporting the above hypothesis. The results of these experiments are summarized in Fig. 4c. Part II: Ca2+ removal mechanisms after disrupting the cytoskeleton with cytochalasin D The results obtained in the first part of this study suggest that the two Ca2+ transporters, NCX and SERCA, are functioning in series, so that inhibition of one of the two components eliminates the entire process. Previous studies indicated that cytoskeletal components are crucial for coupling of various molecular mechanisms. For example, in endothelial cells cytochalasin D abolishes store-operated Ca2+ entry, presumably because it disrupts the putative coupling of IP3 receptors to plasmamembrane Ca2+ entry channels [18]. To investigate the role of the cytoskeleton in the complex Ca2+ removal process described above we pretreated endothelial cells with 50 ␮M cytochalasin D at room temperature for 30 min, 1 or 2 h and then repeated the same experimental protocols as in part I. Fig. 5a shows representative traces of the rates of [Ca2+ ]i decline measured under the same experimental conditions as outlined in Fig. 1, the only difference being that Cell Calcium (2002) 31(6), 265–277

272

X Wang, S Reznick, P Li, W Liang, C van Breemen

Fig. 5 Ca2+ removal after cytochalasin D pretreatment. (a) Cells were stimulated with 10 ␮M ACh in PSS (left), or after pre-treatment with either CPA (middle) or 0Na-CPA (right). After [Ca2+ ] reached the plateau phase of the ACh response, extracellular Ca2+ and ACh were removed to record the rates of Ca2+ removal. The control ACh response was superimposed in the middle and right as dashed line. The CPA response was also shown for comparison as the dotted line on the right. (b) Effects of CPA and CPA combined with 0Na+ on the rate of Ca2+ removal as calculated from data shown in (a). A one-component exponential decay function was fitted to the curves the rate constants are shown as fraction/second (see text for discussion). (c) Mean values and S.E.M. of the extrusion rate from (b) are plotted as a bar graph for all data. (d) Rates of Ca2+ removals calculated as [Ca2+ ]/t are plotted against the relative [Ca2+ ]i . To enable comparison between different cells, all the values of [Ca2+ ]i were normalized to the maximal response of 0Na-CPA from the same cell. After normalizing the data are binned to 0–5, 5–10, 10–15, 15–20, 20–40, 40–60 and 60–80% and the mean value and S.E.M. are calculated after binning. Cell Calcium (2002) 31(6), 265–277

© 2002 Elsevier Science Ltd. All rights reserved.

Ca2+ removal mechanisms

273

Fig. 5 (Continued ).

the cells were pretreated with cytochalasin D. After cytochalasin D treatment, we were able to perform only three different protocols in each cell; therefore, in order to compare the different groups of cells all data were pooled and normalized to the response measured in 0Na plus CPA. The total number for each of the experimen© 2002 Elsevier Science Ltd. All rights reserved.

tal groups were: n = 11 for ACh control, n = 21 for 0Na PSS, n = 18 for CPA and n = 25 for 0Na-CPA. The same one-component exponential fit was performed in Fig. 5b and c. In Fig. 5d, we have plotted the rates of [Ca2+ ]i decline ([Ca2+ ]i /t ) in the cytochalasin D treated cells as a Cell Calcium (2002) 31(6), 265–277

274

X Wang, S Reznick, P Li, W Liang, C van Breemen

Fig. 6 Example of Ca2+ removal after latriculin A treatment. Cells were treated with 30 ␮M latriculin A for 1.5 h at 37 ◦ C. The [Ca2+ ] decline was compared at three experimental conditions: 0Na PSS, CPA and 0Na plus CPA. Additive effect of 0Na and CPA on the rate of Ca2+ removal was observed.

function of the relative [Ca2+ ]i . Comparison with Fig. 3 shows that the treatment with cytochalasin D caused the reductions in the rates of [Ca2+ ]i decline in responses to inhibition of NCX and SERCA to become additive. Thus, cytochalasin D appeared to abolish the functional linkage between SERCA and NCX. In a different set of four experiments another cytoskeleton disrupting agent latriculin (30 ␮M), which works via a different mechanism than cytochalasin D, was used to treat the endothelial cells. The rate of Ca2+ decline was tested using the previous protocols for the three experimental conditions: 0Na-PSS, CPA and CPA in 0Na-PSS. Fig. 6 shows a representative example of latriculin treated cells suggesting that the coupling of ER and NCX was again abolished as indicated by the additive effect of 0Na-PSS and CPA. In addition phalloidin, a toxin that binds the cytoskeleton, was used to fluorescent label the cytoskeleton before and after cytochalasin D treatment. Fig. 7 shows a

Fig. 7 Fluorescent labeling of cytoskeleton using phalloidin. (a) Cells before and after cytochalasin D treatment were fixed and loaded with fluorescent phalloidin. The middle column of pictures was taken where cells appeared sharpest using a confocal microscope. The focal plane was then moved 200 nm below (left column) or above (right column) the initial focal plane in the same slide. (b) Bright field images are shown for cells before (left) and after (right) cytochalasin D treatment. Cell Calcium (2002) 31(6), 265–277

© 2002 Elsevier Science Ltd. All rights reserved.

Ca2+ removal mechanisms

comparison between control and treated cells. While in the control cells the fluorescent phalloidin was more or less evenly distributed in the periphery of the cells, the cytochalasin D treated cells exhibited a more clustered pattern of fluorescent labeling of cytoskeleton indicative of conformational changes.

DISCUSSION We report here for the first time that approximately half of Ca2+ removal from endothelial cells is mediated by an in-series process of ER uptake by SERCA, Ca2+ release towards the PM and Na+ /Ca2+ exchange and that this type of functional coupling is maintained by the cytoskeleton. Ca2+ removal in freshly isolated endothelial cells In contrast to numerous studies of Ca2+ entry in vascular endothelial cells there are very few reports on endothelial Ca2+ extrusion [9–12], and most of them concerned cultured endothelial preparations. Previous studies from our laboratory have provided the first evidence suggesting that defective endothelial Ca2+ extrusion may play critical role in the etiology of vascular disease [24]. Therefore, it is important to understand the Ca2+ removal process in native endothelial cells. We used the rate of [Ca2+ ] decline in 0Ca2+ PSS after ACh stimulation as a measurement for Ca2+ removal processes, which consists of Ca2+ extrusion through PM Ca2+ ATPase and/or NCX, and Ca2+ re-uptake into the intracellular Ca2+ stores. SERCA contribution to Ca2+ removal depends on forward NCX As shown in Fig. 1, removal of extracellular Na+ significantly reduced the rate of Ca2+ removal by 52%. This suggests that forward NCX is involved in the Ca2+ removal process. One possible side effect of the 0Na protocol is that it may cause intracellular acidification as the result of inhibition of the Na+ /H+ antiporter as suggested previously by other studies [18,19]. However, it has been reported for intact rabbit endothelial cells that inhibition of the Na+ /H+ antiporter using HMA did not alter the changes in [Ca2+ ]i caused by extracellular Na+ removal [13]. Likewise, Motley et al. [20] showed that removal of extracellular sodium did not affect intracellular pH in smooth muscle. Therefore, it is reasonable to conclude that the difference in Ca2+ removal rate upon Na+ removal represents the contribution by NCX. Fig. 1 also shows that SERCA blockade with CPA reduced the overall removal of calcium from the cytoplasm by about 50%. Moreover, extracellular sodium removal in the presence of CPA did not reduce the Ca2+ removal rate further than SERCA blockade alone, suggesting a © 2002 Elsevier Science Ltd. All rights reserved.

275

functional link between the superficial ER and NCX. Similar findings have been reported in intact smooth muscle of the rabbit vena cava [16] and in some [12] but not all [9] cultured endothelial cells The model proposed for this functional linkage involves a cytoplasmic junctional space between the PM and ER, which allows transfer of released Ca2+ to the NCX because of restricted Ca2+ diffusion to the bulk cytoplasm. Forward Na+ /Ca2+ -exchange contributes to ER Ca2+ unloading According to the above model the NCX may not only function in Ca2+ removal as has also been shown by Paltauf-Doburzynska et al. [12] in cultured endothelial cells, but may be an important regulator of ER Ca2+ content. Evidence supporting this idea is shown in Fig. 4, where blockade of the outward mode of NCX prevents effective unloading of the ER. Under conditions of 0Na and the presence of Ca2+ , which would favor the reverse mode of operation it is also conceivable that the NCX is capable of increasing ER Ca2+ content. In any case, the close interaction between plasmalemmal and ER Ca2+ transport systems allows for a superficial Ca2+ cycle in endothelial cells, similar to that proposed for smooth muscle, which would allow for regulation of [Ca2+ ]i near the PM separate from regulation of the bulk of the cytoplasm [21]. The cytoskeleton provides the physical support for the alignment of SERCA and NCX It seems unlikely that the functional linkage between the Ca2+ release channels of the ER and the NCX in the plasma membrane could be based on a completely random distribution of membranes and transport proteins. This leads to the question as to what is responsible for the creation of the putative cytoplasmic microdomains. One interesting clue is that in endothelial cells cytochalasin D pre-treatment abolished store-operated Ca2+ influx, which has been attributed to the elimination of physical coupling of ER and the Ca2+ influx channel. Cytochalasin D is an actin filament depolymerizing agent, which has been shown to disrupt the cytoskeleton in many cell types including vascular endothelial cells [22]. Thus, it becomes attractive to speculate that the cytoskeleton may serve as the structural support for the functional coupling of NCX and SERCA. In support of this theory, we found that after treatment with cytochalasin D the inhibitory effects of CPA and 0Na+ became additive, indicating that the functional coupling had been disrupted. Conformation of the changes in the cytoskeleton were obtained by using the fluorescent indicator phalloidin. Also note that brightfield microscopy showed no changes in morphology of the cells. The possible reason for that is the relative short time Cell Calcium (2002) 31(6), 265–277

276

X Wang, S Reznick, P Li, W Liang, C van Breemen

REFERENCES

Fig. 8 Diagram of calcium extrusion model in freshly isolated rabbit aortic endothelial. The intact cytoskeleton functions as a physical support for the coupling of peripheral ER and PM, where NCX in the restricted space contributes to the unloading of ER Ca2+ and consequently the Ca2+ removal process. PMCA contributes to Ca2+ extrusion, located outside the junctional space, in a parallel fashion.

frame of treatment for the cells compare to other studies [11]. A more prolonged treatment with cytochalasin D resulted in loss of responses to the agonists tested here, which makes accurate measurement impossible. As a positive control another agent latriculin A, which disrupts the cytoskeleton via a different mechanism than cytochalasin D, also led to the abolition of the coupling of NCX to ER. An alternate explanation for the apparent lack of additivity of inhibition using 0Na and CPA protocol could be that these agents caused an elevated Ca2+ level, which increased PMCA activity via calmodulin. However, as shown in Fig. 1, the difference in peak Ca2+ level between Na, CPA and 0Na + CPA was less than 10% of the maximal responses. From our previous data, the agonists increased Ca2+ level from rest (about 100 nM) to ∼600 nM [8]. According to a paper by Cox et al. [23], the increase of the relative activity of PMCA at these Ca2+ levels (500–600 nM) should not exceed 1.2 times. Therefore, we believe that the change in peak Ca2+ should not account for changes in the extrusion rate we observed in those cells, although a more quantitative determination of calmodulin and PMCA activity would help to explore this possibility. In summary, Fig. 8 depicts a hypothetical model of calcium extrusion from freshly isolated rabbit aortic endothelial cells. Accordingly the PMCA is responsible for half of the Ca2+ extrusion while the other half is mediated by SERCA uptake into the ER followed by release into the PM–ER junctional space from where the NCX pumps it out of the cell. The inferred special relationship between the superficial SR and the plasmalemmal NCX is maintained through anchoring onto actin filaments.

ACKNOWLEDGEMENTS This work is supported by CIHR (C.v.B. and X.W.) and the H&S Foundation of BC & Yukon (X.W.). Cell Calcium (2002) 31(6), 265–277

1. Jaffe EA. Physiological functions of normal endothelial cells. Ann N Y Acad Sci 1985; 454: 279–291. 2. Furchgott RF, Vanhoute PM. Endothelium-derived relaxing and contracting factors. FASEB J 1989; 3: 2007–2018. 3. Radomski MW, Palmer RMJ, Moncada S. The anti-aggregating properties of vascular endothelium: interaction between prostacyclin and nitric oxide. Brit J Pharmacol 1987; 92: 181–187. 4. Taylor SG, Weston AH. Endothelium-derived hyperpolarizing factor: a new endogenous inhibitor from the vascular endothelium. Trends Pharmacol Sci 1988; 9: 272–274. 5. Yanagisawa M, Kurihara H, Kimura S et al. A novel potent vasoconstrictor peptide produced by vascular endothelial cells. Nature 1988; 332: 411–415. 6. Luckhoff A. Release of prostacyclin and EDRF from endothelial cells is differentially controlled by extra- and intracellular calcium. Eicosanoids 1988; 1: 5–11. 7. Kruse HJ, Grunberg B, Siess W, Weber PC. Formation of biologically active autacoids is regulated by calcium influx in endothelial cells. Arterioscler Thromb 1994; 14: 1821–1828. 8. Wang X, Lau F, Li L, Yoshikawa A, van Breemen C. Acetylcholine-sensitive intracellular Ca2+ store in fresh endothelial cells and evidence for ryanodine receptors. Circ Res 1995; 77: 37–42. 9. Goto Y, Miura M, Iijima T. Extrusion mechanisms of intracellular Ca2+ in human aortic endothelial cells. Eur J Pharmacol 1996; 314 (1/2): 185–192. 10. Domotor E, Abbott NJ, Adam-Vizi V. Na+ –Ca2+ exchange and its implications for calcium homeostasis in primary cultured rat brain microvascular endothelial cells. J Physiol 1999; 515 (pt1): 147–159. 11. Sedova M, Blatter LA. Dynamic regulation of [Ca2+ ]i by plasma membrane Ca(2+ )-ATPase and Na+ /Ca2+ exchange during capacitative Ca2+ entry in bovine vascular endothelial cells. Cell Calcium 1999; 25 (5): 333–343. 12. Paltauf-Doburzynska J, Frieden M, Graier WF. Mechanisms of Ca2+ store depletion in single endothelial cells in a Ca2+ -free environment. Cell Calcium 1999; 25 (5): 345–353. 13. Li L, van Breemen C. Na+ –Ca2+ exchange in intact endothelium of rabbit cardiac valve. Circ Res 1995; 76: 396–404. 14. Misquitta CM, Mack DP, Grover AK. Sarco/endoplasmic reticulum Ca2+ (SERCA)-pumps: link to heart beats and calcium waves. Cell Calcium 1999; 25 (4): 277–290. 15. van Breemen C, Chen Q, Laher I. Superficial buffer barrier function of smooth muscle sarcoplasmic reticulum. Trends Pharmacol Sci 1995; 16: 98–105. 16. Nazer M, van Breemen C. Functional linkage of Na+ –Ca2+ exchange and sarcoplasmic reticulum Ca2+ release mediates Ca2+ cycling in vascular smooth muscle. Cell Calcium 1998; 24 (4): 275–283. 17. Cabello OA, Schilling WP. Vectorial CA2+ flux from the extracellular space to the endoplasmic reticulum via a restricted cytoplasmic compartment regulates inositol 1,4,5-trisphosphate-stimulated CA2+ release from internal stores in vascular endothelial cells. Biochem J 1993; 295: 357–366. 18. Betz AL. Sodium transport in capillaries isolated from rat brain. J Neurochem 1983; 341: 1150–1157. 19. Kitazono T, Takeshige K, Cragoe Jr EJ, Minakami S. Intracellular pH change of cultured bovine aortic endothelial cells in response to ATP addition. Biochem Biophys Res Commun 1988; 152: 1304–1309. © 2002 Elsevier Science Ltd. All rights reserved.

Ca2+ removal mechanisms

20. Motley ED, Paul RJ, Matlib MA. Role of Na+ –Ca2+ exchange in the regulation of vascular smooth muscle tension. Am J Physiol 1993; 264: H1028–H1040. 21. Nazer M, van Breemen C. A role for the sarcoplasmic reticulum in Ca2+ extrusion from rabbit inferior vena cava smooth muscle. Am J Physiol 1998; 274: H123–H131. 22. Holda JR, Blatter LA. Capacitative calcium entry is inhibited in vascular endothelial cells by disruption of cytoskeletal microfilaments. FEBS Lett 1997; 403 (2): 191–196.

© 2002 Elsevier Science Ltd. All rights reserved.

277

23. Cox JA, Comte M, Stein EA. Activation of human erythrocyte Ca2+ -dependent Mg2+ -activated ATPase by calmodulin and calcium: quantitative analysis. PNAS 1982; 79 (14): 4265–4269. 24. Skarsgard PL, Wang X, McDonald P, Lui AH, Lam EK, McManus BM, van Breemen C, Laher I. Profound inhibition of myogenic tone in rat cardiac allografts is due to eNOSand iNOS-based nitric oxide and an intrinsic defect in vascular smooth muscle contraction. Circulation 2000; 101 (11): 1303–1310.

Cell Calcium (2002) 31(6), 265–277