Coupling between Intracellular Ca2+ Stores and the Ca2+ Permeability of the Plasma Membrane COMPARISON OF THE EFFECTS OF THAPSIGARGIN, 2,5-DI-( TERT-BUTYL)-1,4-HYDROQUINONE, AND CYCLOPIAZONIC ACID IN RAT THYMIC LYMPHOCYTES*

The regulation of Ca2+ uptake by receptors is incom- pletely understood. It has been proposed that the Ca2+ permeability of the plasma membrane increases in re- sponse to depletion of a critical intracellular Ca2+ storage compartment (Takemura, H., Hughes, A. R., Thas- trup, O., and Putney, J. W. (1989) J. Biol. Chern. 264, 12266-12271). This hypothesis is based largely on the effect of thapsigargin, an inhibitor of endomembrane Ca2+-ATPases. Due to the existence of an endogenous leak, inhibition of Ca2+ uptake by thapsigargin induces depletion of the stores. This is accompanied by in- creased plasmalemmal Ca2+ permeability, without change in the level of inositol phosphates. On the other hand, depletion of the intracellular stores by 2,5-di-(tert-butyl)- 1,4-hydroquinone (BHQ), a chemically un- related inhibitor of the Ca2+-ATPases, fails to induce Ca2+ influx (Kass, S. G. A.,

The regulation of Ca2+ uptake by receptors is incompletely understood. It has been proposed that the Ca2+ permeability of the plasma membrane increases in response to depletion of a critical intracellular Ca2+ storage compartment (Takemura, H., Hughes, A. R., Thastrup, O., and Putney, J. W. (1989) J. Biol. Chern. 264, 12266-12271). This hypothesis is based largely on the effect of thapsigargin, an inhibitor of endomembrane Ca2+-ATPases. Due to the existence of an endogenous leak, inhibition of Ca2+ uptake by thapsigargin induces depletion of the stores. This is accompanied by increased plasmalemmal Ca2+ permeability, without change in the level of inositol phosphates. On the other hand, depletion of the intracellular stores by 2,5-di-(tert-butyl)-1,4-hydroquinone (BHQ), a chemically unrelated inhibitor of the Ca2+-ATPases, fails to induce Ca2+ influx (Kass, G. E., Duddy, S. K., Moore, G. A., and Orrenius, S. (1989) J. Biol. Chern. 264, 15192-16198). In an attempt to reconcile these observations, we analyzed in lymphocytes the mode of action of thapsigargin and BHQ. In addition, we tested the effects of cyclopiazonic acid (CPA), a blocker of the skeletal muscle sarcoplasmic reticulum Ca2+-ATPase. All three compounds released Ca" from a common intracellular compartment. Thapsigargin and low concentrations of BHQ and CPA concomitantly elevated the plasmalemmal Ca" permeability. Higher concentrations of BHQ and CPA produced a secondary inhibition of the Ca2+ entry pathway, by a mechanism seemingly unrelated to their effects on the internal stores. This inhibitory side effect can account for the reported discrepancies between the effects of thapsigargin and BHQ. The data provide further support for the notion that endomembrane Ca2+ stores are functionally coupled to the plasma membrane Ca" permeability pathway.
A variety of mechanisms have been postulated to account for the receptor-mediated activation of the plasmalemmal Ca2+ permeability by growth factors, hormones, or neurotransmitters. These include a direct coupling between the * This work was supported by the Medical Research Council of Canada and the National Cancer Institute. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.  receptor and the transport pathway (Benham and Tsien, 1987), or coupling via second messengers such as inositol phosphates (Kuno and Gardner, 1987;Morris et al., 1987), cyclic AMP (Kelley et aL, 1990), or cyclic GMP (Pandol and Schoeffield-Payne, 1990). Alternatively, it has been proposed that the site of action of the second messengers is an intracellular Ca2+ storage compartment, which in turn determines the Ca2+ permeability of the plasma membrane. It is not clear how information is conveyed from the internal Ca2+ store to the surface membrane, but two mechanisms have been suggested. Some authors believe that the plasmalemmal Ca2+ permeability increases in response to the elevated [Ca2+]il which results from release of the endomembrane Ca2+ stores (Von Tscharner et al., 1986;Ng et al., 1988Ng et al., , 1990. Others have proposed that the plasma membrane is responsive to the Ca2+ content of the stores, regardless of the cytosolic Ca" concentration (Putney, 1986;Takemura et al., 1989). Support for the latter model has come from experiments designed to deplete intracellular Ca2+ pools by a method independent of receptor activation and inositol phosphate production. This can be accomplished with thapsigargin, a sesquiterpene lactone that is a potent and selective inhibitor of the microsomal Ca2+-ATPase (Thastrup, 1990). Ostensibly due to a susbstantial endogenous leak, inhibition of the ATPase by the lactone produces a rapid release of Ca2+ from an endomembrane pool that includes the IP3-sensitive compartment (Thastrup et aZ., 1990). In a variety of cells, the addition of thapsigargin also results in the concomitant elevation of plasma membrane Ca2+ permeability (Thastrup, 1990). These findings are consistent with the "capacitative" coupling model, which stipulates that the plasmalemmal permeability is controlled by the degree of filling of an intracellular Ca2+ storage compartment.
Although attractive, the capacitative model has thus far failed to explain certain observations. In murine cells of the neural line NG115-401L, exposure to thapsigargin induces depletion of the endomembrane stores, yet is not accompanied by an increased Ca" permeability of the plasma membrane (Jackson et al., 1988). Moreover, discordant results have been reported using a different inhibitor of the microsomal Ca2+-ATPase. BHQ? a synthetic compound chemically unrelated to thapsigargin, effectively depletes internal Ca2+ stores in hepatocytes, without increasing the plasmalemma1 Ca2+ permeability (Kass et aL, 1989). Unlike the results in NG115-401L cells, the latter observation cannot be attributed to tissue specificity, since thapsigargin effectively increased Ca2+ influx in hepatocytes Thastrup, 1990). Thus, the source of this discrepancy remains unclear, questioning the validity of the capacitative model. In an attempt to reconcile the results obtained with thapsigargin and BHQ, we have undertaken a systematic study of the mode of action of these two inhibitors, under otherwise identical conditions. In addition, we also tested the effects of cyclopiazonic acid (CPA), an inhibitor of the Ca2+-ATPase of skeletal muscle sarcoplasmic reticulum (Siedler et al., 1989). Because of the similarities between the Ca2'-ATPases of the endoplasmic and sarcoplasmic reticuli, we anticipated that CPA would effectively deplete intracellular Ca2+ stores in non-muscle cells. The experiments were carried out in rodent lymphocytes, which have been reported to respond to thapsigargin with increased plasma membrane Caz+ permeability (Mason et aL, 1991). Our results indicate that all three compounds effectively deplete the stores and promote Ca2+ influx. At concentrations higher than those required for depletion of the stores, BHQ and, to a lesser extent, CPA inhibit the entry of Ca2+ across the plasmalemma. Therefore, the apparent inconsistency of earlier reports can be accounted for by a secondary effect of some of the inhibitors, lending further credence to the capacitative model.
The basic Na+ solution contained 140 mM NaCl, 3 mM KCl, 1 mM CaC12, 1 mM MgC12, 0.2 mM EGTA, 10 mM D-glucose, and 20 mM HEPES-free acid, plus 1 mg/ml albumin. The solution was titrated to pH 7.25 at 37 'C with NaOH. Ca2+-free solution was made by omitting Ca2+ and increasing the EGTA concentration to 0.5 mM. When required, Mn2+ was added directly to the solution as MnC12. The osmolarity of the solution was adjusted to 295-300 mosM prior to the addition of albumin using an Osmette freezing point osmometer. All solutions and stocks were stored at -20 "C.

Cell Isolation and Manipulations
Thymic lymphocytes were isolated from 140-200-g male Wistar rats (The Charles River Breeding Laboratories) as previously described (Grinstein et al., 1984). The cells were counted using a Model ZM Coulter Counter (Coulter Electronics, Hialeah, FL) and maintained at room temperature at a concentration of 30-50 X IO6 cells/ ml in bicarbonate-free RPMI 1640 culture medium buffered to pH 7.4 with 20 mM HEPES.

Fluorescence Determinations
All experiments were performed at 37 "C using a Hitachi Model F-4000 fluorescence spectrophotometer equipped with a magnetic stirrer. The cells were counted immediately after the last manipulation, prior to addition to the cuvette, to ensure that the appropriate cell number was added.
Determinution of Free Cytosolic Calcium Concentration-[Ca2+]; was determined by measuring the fluorescence of indo-1. The excitation and emission wavelengths used were 331 nm (3-nm slit width) and 410 nm (10-nm slit), respectively. Thymocyte suspensions (25 X lo6 cells/ml) were loaded with indo-1 by incubation with a 2 p~ concentration of the AM precursor for 25 min at 37 "C in basic Na+ solution devoid of albumin and EGTA. The cells were then sedimented, resuspended in basic Na+ solution devoid of albumin and EGTA, and kept at room temperature until required. To monitor fluorescence, aliquots containing the required cell number were sedimented, resuspended in Na+ solution plus or minus Ca2+, as indicated, and added to the cuvette. The fluorescence of indo-1 was calibrated using ionomycin and Mn2+ as previously described (MacDougall et al., 1988). A dissociation constant of 250 nM for the indo-l.Ca2+ complex was used to calculate [Ca2+], (Grynkiewicz et al., 1985).
Determination of Mn2+ Influx-Mn2+ uptake was monitored as the rate of quenching of indo-1 fluorescence measured at the isosbestic point for the Ca2+. indo-1 complex (excitation 345 nm, emission 455 nm). When measured at the isosbestic wavelengths, the rate of fluorescence decrease is insensitive to changes in [Ca2+] and is proportional to the rate of Mn2+ accumulation in the cytosol. Similar techniques have been employed for the measurement of Mn2+ uptake by platelets, neutrophils, and endothelial cells using fura2 and quina (Hallam and Rink, 1985;Merritt et al., 1989;Hallam et al., 1989).
To mimic the conditions used for the determination of [Ca2+];, 3 X 10" cells were loaded with 2 p~ indo-1-AM at a concentration of 25 X lo6 cells/ml for 25 min at 37 "C in basic experimental solution devoid of albumin and EGTA. To prolong the linear phase of unidirectional '%a2+ uptake, cells were simultaneously loaded with dimethyl-BAPTA by adding a 4 p~ concentration of the AM precursor to this incubation medium. 2.5 X 10" cells were then sedimented and resuspended in 0.65 ml of basic experimental solution. A 0.6-ml aliquot of the suspension was transferred to a 1.5-ml microcentrifuge tube and maintained at 37 "C, while the remaining 0.05 ml was used for cell number determinations. Measurement of Ca2+ uptake was initiated by addition of W a Z + (10 pCi/ml final concentration). Uptake was terminated after the required interval by transfering 0.1-ml aliquots of the cell suspension to 1.0 ml of ice-cold Na+ solution devoid of EGTA and containing 0.1 ml of LaCl,, followed by centrifugation through an oil phase composed of 2 parts vegetable oil (Mazola) and 10 parts dibutyl phthalate (v/v). Triplicate samples of the supernatant were counted to ascertain the specific activity. The supernatant was then removed from above the oil layer, the tip of the tube containing the pellet was cut and placed in a scintillation vial, and the pellet was lysed by the addition of 0.5 ml of distilled water. Following the addition of 10 ml of Aquasol-2, the samples were counted by liquid scintillation (LKB Rackbetta, Turku, Finland).
Unless otherwise indicated, all experiments were performed at 37 "C, and the results are presented as the mean & S.E.M. of the number of preparations indicated in parentheses. The fluorescence traces illustrated are representative of experiments in a minimum of three preparations.

RESULTS
The ability of thapsigargin to alter cytosolic [Ca2+Ii in rat thymic lymphocytes was examined using the fluorescent Ca2+ indicator indo-1. In the presence of external Ca2+, thapsigargin induced a sustained dose-dependent rise in (Ca2+]i, as illustrated in Fig. 1, A and B. Similar increases in [Ca2+Ii during thapsigargin exposure were previously reported in human peripheral T-lymphocytes (Scharff et al., 1988) and in the human T-cell clone P28 (Gouy et aZ., 1990). In the absence of external Ca2+, exposure of thymocytes to thapsigargin produced a transient increase in [Ca"]; (Fig. IC). In other cell types, similar transients have been interpreted to reflect unmasking of an endogenous Ca2+ leak from the endomembrane pool upon inhibition of the Ca2+-ATPase (Thastrup, 1990;Thastrup et al., 1990). The dose dependence of the rise in [Ca2+]i induced by thapsigargin in the presence or absence of extracellular Ca2+ is summarized in Fig. 1D. In the presence of external Ca2+, maximally effective concentrations of thapsigargin elevated [Ca2+]; by 1079 k 217 n M (n = 4), from a resting level of 190 k 9 nM ( n = 24). It is noteworthy that, under these conditions, the rise in [Ca2+]i was sustained at all concentrations of thapsigargin tested. The efficacy of the thapsigargin-induced release of Ca" from endomembrane pools, as measured by the transient rise in [Ca"Ii in the absence of external Ca2+, closely paralleled the ability of the lactone to generate the sustained [Ca"]i increase in the presence of external Ca2+. Maximal responses were observed in both cases at thapsigargin concentrations between 3 and 30 nM. In the absence of external Ca2+, a maximal dose of thapsigargin resulted in a [CaZ+li transient peaking at 89 f 11 nM (n = 4) above the resting level.
An identical analysis of the effect of BHQ on [Ca2+Ii was performed. The release of Ca2+ from intracellular stores induced by BHQ showed a pattern similar to that observed with thapsigargin (cf. Figs. 2C and I C ) . In the range studied, the Ca2+ released by BHQ increased monotonically with the concentration of the inhibitor, reaching a maximum of 94 f 9 nM ( n = 3 ) above the basal level between 5 and 25 PM BHQ (Fig.  2 0 ) . In contrast, the effect of BHQ in the presence of extracellular Ca2+ differed markedly from that of thapsigargin. While low doses of BHQ induced a sustained [Ca2+Ii increase (Fig. 2B), the  of the BHQ concentration, two points become apparent (Fig.  2 0 ) . First, the [Ca2+]i rise induced by 50 PM BHQ is smaller than that attained with 10 or 25 PM. This effect cannot be attributed to an inability of higher concentrations of BHQ to inhibit the endomembrane ATPase, as indicated by the results obtained in the absence of extracellular Ca2+ (see above). Second, the maximal change in [Ca2+Ii in the presence of external Ca2+ is much lower than that induced by thapsigargin (289 -+ 27 nM ( n = 4) for BHQ uers'sus 1079 nM for thapsigargin). Taken together, these data suggest that the biphasic increase in [Ca2+Ii observed in the presence of external Ca2+ (Fig. M ) and the biphasic nature of the dose-response curve generated under similar conditions (open symbols in Fig. 2 0 ) are the result of a secondary effect of high concentrations of BHQ, unrelated to its ability to inhibit the endomembrane ATPase. Using experiments like those outlined above, we also investigated the effect of a third endomembrane ATPase inhibitor, CPA, on [Ca2+]i homeostasis. Cells suspended in Ca2+-free solution and exposed to CPA demonstrated a transient increase in [Ca2+Ii. This response, which resembles the effects obtained with thapsigargin and BHQ under comparable conditions, is indicative of release of Ca2+ from an intracellular store (Fig. 3C) and suggests that CPA impaired the endomembrane Ca2+-ATPase. Ca2+ release from stores was maximal at =5 PM CPA, and comparable responses were obtained with higher concentrations (Fig. 3 0 ) . In the presence of external Ca2+, low doses of CPA induced a sustained increase in [CaZ+li (Fig. 3 B ) . Higher concentrations of CPA (50 pM) elicited a biphasic change in [Ca2+Ii: a large initial increase followed by a slow recovery toward basal levels. This secondary recovery phase was slower than that seen in the presence of high concentrations of BHQ (cf. Figs. 2A and 3A) and, unlikely the latter, was not consistently observed.
A detailed analysis of the concentration dependence of the effects of CPA is summarized in Fig. 30. Like thapsigargin, the efficacy of the CPA-induced release of Ca2+ from endomembrane pools, as measured by the transient rise in [Ca2+Ii in the absence of external Ca2+, is closely paralleled by its ability to induce the larger, more sustained change in the presence of external Ca2+. Maximal responses in both the presence and absence of external Ca2+ were observed at a concentration between 5 and 10 p~ CPA. In the presence of external Ca2+, maximally effective concentrations of CPA resulted in a peak increase in [Ca2+]i of 549 f 49 nM ( n = 6), which is intermediate between the responses obtained with thapsigargin (1079 nM) and BHQ (289 nM). In the absence of external Ca2+, a maximal dose of CPA results in a [Ca2+Ii transient which peaks at a value 93 f 11 nM ( n = 3) above the resting level. This peak increase in [Ca2+Ii is virtually identical with that induced by BHQ and thapsigargin (94 nM and 89 nM, respectively).
Although it is apparent that thapsigargin, BHQ, and CPA release comparable amounts of intracellular Ca2+, it is not clear whether the same pool is affected in every instance. It is therefore conceivable that the variable rise in [Ca2++Ii induced by the different inhibitors in Ca2+-containing medium reflects the involvement of different stores, which exert variable degrees of control over the plasmalemma1 Ca" permeability. To investigate the degree of overlap between the stores released by thapsigargin, BHQ, and CPA, we monitored the changes in indo-1 fluorescence during sequential additions of the inhibitors to cells suspended in Ca2+-free solution. While the addition of 50 p~ BHQ resulted in a transient rise in [Ca2+]i, a subsequent addition of 30 nM thapsigargin was found to have no effect (Fig. 4A). If the order of presentation was reversed, thapsigargin induced a marked transient increase in [Ca2'Ii, with the secondary addition of BHQ having no detectable effect (Fig. 4B). These findings suggest that the Ca2+ pools depleted by both inhibitors overlap extensively. A similar pattern emerged when combinations of thapsigargin and CPA were investigated (Fig. 4, C and 0). While 50 p~ CPA induced a transient rise in [Ca2+]i, a subsequent addition of 30 nM thapsigargin was without effect (Fig. 4C). Conversely, CPA addition was without effect if added after thapsigargin (Fig. 40). Control experiments ruled out the possibility that the lack of effect of the second inhibitor was due to the depletion of the intracellular Ca2+ pool, as a result of the prolonged (approximately 10-min) incubation in Ca*+-free solution required to assess the effects of the first inhibitor. While some depletion does occur during this interval, a clearcut transient increase in Ca2+ can still be induced by thapsigargin addition following a 14-min incubation in Ca2+-free solution, provided no ATPase inhibitor is added previously (results not shown). Taken together, these results are consistent with the notion that thapsigargin, BHQ, and CPA release Ca2+ from an identical intracellular pool.
In view of these findings, the differential effect of the inhibitors on [Ca2+Ii in Ca2+-containing medium cannot be attributed to variable degrees of depletion of a critical Ca2+ store. We therefore considered the possibility that BHQ and, to a lesser extent, CPA reduce the level of [Ca2+Ii by either interfering with Caz+ uptake at the plasma membrane or by accelerating its extrusion. To test this possibility, cells were initially treated with 50 p~ BHQ and subsequently exposed to thapsigargin (Fig. 5A). As described above, this concentration of BHQ produced a biphasic change in [Ca2+Ii, with a secondary sustained increase to =300 nM. When added after BHQ, thapsigargin was without effect. In parallel experiments performed with the same batch of cells (Fig. 5B), thapsigargin alone increased [Ca2+]i to values in excess of 800 nM. This implies that BHQ inhibited the effect of thapsigargin on [Ca2+Ii. This can be demonstrated more clearly by reversing the order of addition of the reagents (Fig. 5C). When added after [Ca2+Ii was maximally elevated by thapsigargin, BHQ produced a precipitous drop in [Ca2+Ii to levels similar to those found in cells treated with BHQ alone.
These for the experimental conditions, while a single control experiment is shown. Average uptake rates before and after additions of inhibitors were calculated over the time interval 0.25-5.5 min and 7.5-11.5 min, respectively. by BHQ of the activated plasma membrane Ca" influx pathway. Alternatively, BHQ may stimulate Ca" extrusion from the cell or activate sequestration of Ca" into an inhibitorresistant intracellular pool. To assess the first possibility, we determined the effect of the inhibitors on the rate of unidirectional Ca2+ influx, measured isotopically. To minimize backflux, thereby prolonging the linear phase of Ca2+ uptake, the cytosolic Ca2+ buffering power was increased by loading the cells with dimethyl-BAPTA. The results are summarized in Fig. 6. Addition of 300 nM thapsigargin stimulated the rate of Ca2+ uptake approximately 10-fold, from the resting value of 20 f 3 pmol/min to 207 f 34 pmol/min (n = 3). Uptake was also stimulated by BHQ and by CPA, but to a considerably lower extent. The rates attained averaged 87 +-13 pmol/ min and 88 f 3 pmol/min (n = 3), respectively. The reduced potency of CPA and BHQ to stimulate 45Ca influx parallels their smaller effect on [Ca2+]i and suggests that these agents interfere with the Ca2+ entry pathway. This conclusion was verified by measuring the unidirectional uptake of Mn2+, a Ca2+ surrogate, into indo-1-loaded thymocytes. Mn2+, which has been used successfully as a probe of Ca2+ influx pathways in platelets, neutrophils, and endothelial cells (Hallam and Rink, 1985;Merritt et al., 1989;Hallam et al., 1989), enters unstimulated thymocytes, resulting in gradual quenching of the fluorescent dye. The rate of fluorescence decrease, measured at the Ca2+ isosbestic point, provides a relative measure of the divalent cation permeability. Subsequent chelation of extracellular Mn2+ with DTPA halts this decline, but no fluorescence recovery can be observed (not shown), indicating that Mn2+ is not extruded from the cells. This reflects the inability of the Ca" pump to transport Mn2+, rather than slow dissociation of the Mn2+. indo-1 complex, since rapid fluorescence recovery was induced under these conditions by the addition of ionomycin. These findings validate the use of indo-1 quenching by Mn2+ as a discriminating measure of unidirectional divalent cation uptake in lymphocytes.
As shown in Fig. 7A, the rate of indo-1 quenching produced by Mn2+ was greatly accelerated by the addition of 30 nM thapsigargin, revealing the activation of a divalent cation permeability pathway in the plasma membrane. Subsequent addition of ionomycin, which can transport Mn2+ effectively, rapidly abolished the remaining fluorescence. At low concentrations (510 p~) , BHQ also produced a sustained acceleration of the entry of Mn2+ (Fig. 8). In contrast, a biphasic effect was recorded at 50 ~L M BHQ: a rapid increase in the rate of quenching was followed by a return toward the basal rate (Fig. 8). The latter most likely reflects the inhibitory effect of high doses of BHQ on the divalent cation entry pathway. In support of this notion, we found that addition of 50 PM BHQ to cells previously stimulated with thapsigargin led to an abrupt decline in the rate of indo-1 quenching (91 & 5%, n = 10; Fig. 7 B )   Finally, we investigated the effect of CPA on Mn2+ uptake. At 50 p~, this Ca2+-ATPase inhibitor initially increased the rate of Mn2+ entry, followed by a gradual decline toward the basal rate (Fig. 9B). This behavior resembles that reported above for 50 p~ BHQ, although the inhibitory component of CPA was less pronounced. As in the case of BHQ, the secondary phase observed after the addition of CPA results from blockade of divalent cation entry pathways. This is demonstrated in Fig. 9A, where addition of CPA following stimulation of Mn2+ uptake by thapsigargin significantly reduced the rate of indo-1 quenching. In six experiments, Mn2+ uptake induced by 30 nM thapsigargin was inhibited 62 f 20% by the addition of 50 p~ CPA.

DISCUSSION
The purpose of the present experiments was to resolve the apparent discrepancy between the reported effects of two putative inhibitors of the endomembrane Ca2+-ATPase, thapsigargin and BHQ, on plasmalemmal Ca2+ permeability. In addition, we tested if CPA, a blocker of the ATPase of muscle sarcoplasmic reticulum, also precluded Ca2+ accumulation by the reticulum of non-muscle, lymphoid cells and whether plasmalemmal Ca2+ permeability was consequently affected. We found that, at the appropriate concentrations, all three compounds triggered the rapid release of intracellular Ca2+ stores. The same endomembrane storage pool was seemingly affected by the three agents, since their effects were not additive (Fig. 4). Although we made no attempt to confirm or investigate the exact mechanism responsible for the release of Ca2+ from intracellular stores induced by these inhibitors, preliminary evidence suggests that elevation of IP3 levels is not involved. We have been unable to detect increases in IP3 levels in rat thymic lymphocytes exposed to 300 nM thapsigargin, 50 p M BHQ, or 50 p~ CPA. In the same experiments, the mitogenic lectin concanavalin A (20 pg/ml), which induces the release of Ca2+ from intracellular stores and a concomitant influx of extracellular Ca2+, increased IP3 levels approximately 6-fold (results not shown). These findings are in accordance with published reports that thapsigargin does not induce phophoinositide hydrolysis in the cultured T-cell line P28 (Gouy et dl., 1990), parotid acinar cells (Takemura et al., 1989), or the neuronal cell line NG115-4011 (Jackson et al., 1988). Similarly, BHQ failed to increase phosphoinositide hydrolysis in isolated rat hepatocytes (Kass et al., 1989). Thus, inhibition of the Ca2+-ATPase and subsequent efflux of Ca2+ through a constitutive "leak" pathway is the likely mechanism underlying the observed depletion of the stores.
In addition to releasing Ca2+ from a common intracellular store, thapsigargin, BHQ, and CPA also promoted an increase in the divalent cation permeability of the plasma membrane as measured by: 1) a relatively sustained rise in [Ca2+Ii in the presence, but not in the absence, of extracellular Ca2+; 2) an increase in the rate of 45Ca2+ uptake; and 3) an increase in the rate of Mn2+ influx, determined fluorimetrically. These findings are in good agreement with the model proposed by Putney and co-workers (Putney, 1986;Takemura et al., 1989), whereby the Ca2+ permeability of the plasma membrane is dictated by the degree of filling of an endomembrane Ca2+ storage compartment. While this model received strong support from the observations made with thapsigargin, it was difficult to rule out that this lactone had independent, direct effects on plasmalemmal permeability. The ability of BHQ and CPA to enhance Ca2+ entry into the cells provides further support for Putney's coupling model. These coumpounds, which are structurally unrelated to thapsigargin, share with this lactone the ability to deplete endomembrane Ca2+ stores. The relationship between these events is suggested by the parallel concentration dependence of the depletion of the stores (estimated in Ca2+-free medium) and the entry of divalent cations (revealed by the difference in [Ca2+Ii attained in the presence and absence of external Ca"), particularly at low doses of the drugs. The divergence noted at higher concentrations, which is due to a secondary effect on the plasma membrane, is discussed in more detail below. Increased plasmalemmal permeability to divalent cations in response to CPA and BHQ was confirmed using Mn2+.
While the above observations are compatible with the revised capacitative model, we cannot formally rule out the possibility that thapsigargin, BHQ, and CPA all directly activate Ca2+ uptake at the plasma membrane independently of their effect on the internal stores. This would require the presence of a plasmalemmal receptor for the drugs with an affinity similar to that of the endomembrane Ca2+ ATPase. The increase in [Ca2+Ii promoted by this putative receptor would involve a mechanism other than simple inhibition of the plasma membrane Ca2+-ATPase, since the unidirectional influx of 45Ca2+ and the uptake of Mn2+ (which is not pumped out of the cells) are increased by the endomembrane pump inhibitors.
While at low concentrations all three inhibitors induced a sustained rise in [Ca2+Ji, a secondary decline in [Ca2+Ii was observed in cells treated with higher concentrations of BHQ and CPA. In addition, the rise in [Ca2+li induced by CPA and BHQ was markedly lower than that observed in the presence of thapsigargin. Since maximally effective concentrations of these compounds completely depleted the intracellular store(s), the differences in plasma membrane divalent cation permeability cannot be attributed to variable degrees of store depletion. Instead, inhibition of the plasma membrane influx pathway, or uncoupling of stores and the plasma membrane, could account for the smaller increase in [Ca2+Ii induced by Calcium Permeability of Lymphocytes tion that BHQ and to a lesser extent CPA have a direct inhibitory effect on the plasma membrane divalent cation permeability. First, pretreatment of cells with BHQ precluded the stimulation of Ca2+ entry by thapsigargin (Fig. 5A). Second, addition of BHQ after stimulation with thapsigargin reduced [Ca2+]i (Fig. 5 C ) . Third, BHQ and CPA inhibited the influx of Mn2+ activated by thapsigargin (Figs. 8 and 9). Therefore, it appears that low doses of BHQ and CPA deplete internal stores and increase plasmalemmal permeability, whereas higher concentrations tend to inhibit Ca2+ entry from the medium. This dual effect readily explains the biphasic nature of the [Ca2+]i changes recorded at high concentrations of these agents and the lower [Ca2+Ii levels attained at maximally stimulatory doses of BHQ and CPA, compared with optimal doses of thapsigargin. The ability of high concentrations of BHQ to release Ca2+ from intracellular stores while simultaneously inhibiting the uptake of Ca2+ across the plasma membrane can also explain the apparent discrepancy that exists regarding the regulation of the plasma membrane Ca2+ permeability in hepatocytes. In these cells, BHQ was reported to effectively deplete the internal stores without concomitantly increasing the plasma membrane Ca2+ permeability (Kass et al., 1989). On the basis of these results, Kass and colleagues concluded that the Ca2+ content of intracellular stores is not a determining factor in the regulation of the plasmalemmal Ca2+ permeability, as proposed by Putney (1986) (see above). However, it is conceivable that the concentration of BHQ used by Kass and collaborators (1989) (25 PM) had a secondary inhibitory effect on the uptake pathway, accounting for their failure to detect increased divalent cation entry.
It is apparent from the results presented that thapsigargin is the drug of choice for the unambiguous assessment of the effects of depletion of internal Ca2+ stores. In our experiments, concentrations of thapsigargin 10-fold larger than those required to deplete the stores were without inhibitory effect on Caz+ or Mn2+ entry from the medium. However, caution must nevertheless be exercised, since preliminary experiments in leukocytes indicate that inhibition becomes significant when micromolar concentrations of thapsigargin are used.3 Because such high concentrations are required for effective depletion of stores in some tissues (Gouy et al., 1990;Thastrup et al., 1987;Jackson et al., 1988;Takemura et aZ., 1989)) secondary effects cannot be ruled out and should be considered. It is also noteworthy that CPA and particularly BHQ are considerably less expensive than thapsigargin and may be suitable for certain studies, particularly when used at lower concentrations. Lukacs and M. J. Mason, unpublished observations. In summary, the results presented here provide a means to reconcile the data obtained in other cells with thapsigargin and BHQ. In addition, CPA is introduced as an alternative agent capable of depleting intracellular Ca2+ stores in intact non-muscle cells. Together, the actions of these three compounds support the contention that the degree of filling of the stores plays a determinant role in controlling Ca2+ entry into the cell, in accordance with Putney's revised capacitative model (Takemura et al., 1989).