The Bell-shaped Ca2+ Dependence of the Inositol 1,4,5-Trisphosphate-induced Ca2+ Release Is Modulated by Ca2+/Calmodulin*

Calmodulin inhibits inositol 1,4,5-trisphosphate (IP3) binding to the IP3 receptor in both a Ca2+-dependent and a Ca2+-independent way. Because there are no functional data on the modulation of the IP3-induced Ca2+release by calmodulin at various Ca2+ concentrations, we have studied how cytosolic Ca2+ and Sr2+interfere with the effects of calmodulin on the IP3-induced Ca2+ release in permeabilized A7r5 cells. We now report that calmodulin inhibited Ca2+ release through the IP3 receptor with an IC50 of 4.6 μm if the cytosolic Ca2+ concentration was 0.3 μm or higher. This inhibition was particularly pronounced at low IP3 concentrations. In contrast, calmodulin did not affect IP3-induced Ca2+release if the cytosolic Ca2+ concentration was below 0.3 μm. Calmodulin also inhibited Ca2+ release through the IP3 receptor in the presence of at least 10 μm Sr2+. We conclude that cytosolic Ca2+ or Sr2+ are absolutely required for the calmodulin-induced inhibition of the IP3-induced Ca2+ release and that this dependence represents the formation of the Ca2+/calmodulin or Sr2+/calmodulin complex.


Calmodulin inhibits inositol 1,4,5-trisphosphate (IP 3 )
binding to the IP 3 receptor in both a Ca 2؉ -dependent and a Ca 2؉ -independent way. Because there are no functional data on the modulation of the IP 3 -induced Ca 2؉ release by calmodulin at various Ca 2؉ concentrations, we have studied how cytosolic Ca 2؉ and Sr 2؉ interfere with the effects of calmodulin on the IP 3 -induced Ca 2؉ release in permeabilized A7r5 cells. We now report that calmodulin inhibited Ca 2؉ release through the IP 3 receptor with an IC 50 of 4.6 M if the cytosolic Ca 2؉ concentration was 0.3 M or higher. This inhibition was particularly pronounced at low IP 3 concentrations. In contrast, calmodulin did not affect IP 3 -induced Ca 2؉ release if the cytosolic Ca 2؉ concentration was below 0.3 M. Calmodulin also inhibited Ca 2؉ release through the IP 3 receptor in the presence of at least 10 M Sr 2؉ . We conclude that cytosolic Ca 2؉ or Sr 2؉ are absolutely required for the calmodulin-induced inhibition of the IP 3induced Ca 2؉ release and that this dependence represents the formation of the Ca 2؉ /calmodulin or Sr 2؉ / calmodulin complex.
Inositol 1,4,5-trisphosphate (IP 3 ) 1 is a second messenger used by many cell types to release Ca 2ϩ from internal stores (1). Cytosolic Ca 2ϩ has a bell-shaped effect on the IP 3 receptor (IP 3 R), with low concentrations stimulating the release and high concentrations inhibiting it (2)(3)(4)(5). Calmodulin, a Ca 2ϩbinding protein abundantly present in many cell types (6), binds to the modulatory region of the IP 3 R in a Ca 2ϩ -dependent way (7,8). Calmodulin also interacts with the IP 3 R in a Ca 2ϩindependent way (9 -11), with one of the interaction sites located within the IP 3 -binding domain (11). The findings that calmodulin inhibited IP 3 -induced Ca 2ϩ release in a medium containing 0.2 M free Ca 2ϩ and in addition inhibited [ 3 H]IP 3 binding both in the absence and presence of cytosolic Ca 2ϩ led to the proposal that the Ca 2ϩ -independent binding of calmodulin was responsible for the regulation of the IP 3 R (9).
Although the free cytosolic [Ca 2ϩ ] is a very important regulator of the IP 3 R (2-5), there are no functional data showing how calmodulin modulates the Ca 2ϩ release induced by IP 3 at various Ca 2ϩ concentrations. We have therefore studied how Ca 2ϩ interferes with the effects of calmodulin on the IP 3 -induced Ca 2ϩ release in permeabilized A7r5 cells. All experiments were performed in the absence of Mg-ATP to avoid activation of the Ca 2ϩ -and calmodulin-dependent protein kinase CaMKII that was reported to stimulate the IP 3 R (12). We now report that calmodulin inhibited the IP 3 -induced Ca 2ϩ release if the free cytosolic [Ca 2ϩ ] was 0.3 M or higher. This inhibition occurred with an IC 50 of 4.6 M and was particularly pronounced at low IP 3 concentrations. Calmodulin did not affect the IP 3 -induced Ca 2ϩ release at lower Ca 2ϩ concentrations. The effects of Ca 2ϩ could be mimicked by Sr 2ϩ . We conclude that cytosolic Ca 2ϩ or Sr 2ϩ are absolutely required for the calmodulin-induced inhibition of the IP 3 -induced Ca 2ϩ release. As a consequence, the bell-shaped Ca 2ϩ activation curve of the IP 3 R becomes narrower in the presence of calmodulin. EXPERIMENTAL PROCEDURES 45 Ca 2ϩ fluxes were performed on saponin-permeabilized A7r5 cells from embryonic rat aorta (13). The nonmitochondrial Ca 2ϩ stores were loaded for 50 min in 120 mM KCl, 30 mM imidazole-HCl (pH 6.8), 5 mM MgCl 2 , 5 mM ATP, 0.44 mM EGTA, 10 mM NaN 3 , and 150 nM free Ca 2ϩ (23 Ci ml Ϫ1 ) and then washed once in 1 ml of efflux medium containing 120 mM KCl, 30 mM imidazole-HCl (pH 6.8), 4 M thapsigargin, and, unless otherwise indicated, 1 mM BAPTA. The efflux medium was replaced every 2 min for 20 min. The additions of IP 3 , Ca 2ϩ , Sr 2ϩ , and calmodulin are indicated in the figures. The free [Ca 2ϩ ] was calculated with the CaBuf computer program using the following decimal logarithms of the association constants for ATP: H-ATP, 6.49; H-HATP, 4.11; Ca-ATP, 3.78; Ca-HATP, 1.98; Mg-ATP, 4.00; and Mg-HATP, 2.06 (14). The association constants for BAPTA were: H-BAPTA, 6.36; H-HBAPTA, 5.47; Ca-BAPTA, 6.97; and Sr-BAPTA, 5.13 (15). At the end of the experiment the 45 Ca 2ϩ remaining in the stores was released by incubation with 1 ml of a 2% SDS solution for 30 min.
Calmodulin from bovine brain (purity Ͼ99%) was obtained from Calbiochem (San Diego, CA) and dissolved as a 1 mM stock in 30 mM imidazole-HCl (pH 6.8). Control samples were treated with the same buffer.

RESULTS AND DISCUSSION
Effect of Calmodulin on IP 3 -induced Ca 2ϩ Release in the Presence of 0.3 M Free Ca 2ϩ -Permeabilized A7r5 cells loaded to equilibrium with 45 Ca 2ϩ slowly lost their accumulated 45 Ca 2ϩ during incubation in efflux medium containing 1 mM BAPTA and no added Ca 2ϩ . Thapsigargin (4 M) was added to block the endoplasmic reticulum Ca 2ϩ pumps during the additions of Ca 2ϩ . A short exposure to 1 M IP 3 and 0.3 M free Ca 2ϩ accelerated the rate of Ca 2ϩ loss (Fig. 1a, closed circles). The release was less pronounced if 10 M calmodulin was added together with the IP 3 and Ca 2ϩ (Fig. 1a, open circles). Addition of 0.3 M free Ca 2ϩ in the absence of IP 3 by itself induced a small Ca 2ϩ release (Fig. 1b), due to the exchange of labeled Ca 2ϩ for unlabeled Ca 2ϩ (3,16). This release was identical in the presence (Fig. 1b, open circles) or absence (Fig. 1b, closed circles) of 10 M calmodulin.
Ca 2ϩ release was always measured in the absence of Mg-ATP. Moreover, because there were six wash steps between the loading of the stores in the presence of Mg-ATP and the challenge with IP 3 , all residual Mg-ATP should have been effectively removed. The involvement of the Ca 2ϩ -and calmodulindependent protein kinase CaMKII in the observed inhibition by calmodulin seems therefore unlikely. In addition, we have also tested the effect of the CaMKII inhibitor KN62. Calmodulin (10 M) inhibited the Ca 2ϩ release induced by 1 M IP 3 and 0.3 M free Ca 2ϩ by 54 Ϯ 3% in the absence of KN62 and by 51 Ϯ 4% (n ϭ 3) in the presence of 10 M KN62. These findings exclude the involvement of CaMKII in the inhibition of the IP 3 R by calmodulin.
[Ca 2ϩ ] Dependence of the Effect of Calmodulin on IP 3 -induced Ca 2ϩ Release-Similar experiments to those illustrated in Fig. 1 were performed at several free Ca 2ϩ concentrations (Fig. 2a). Calmodulin (10 M) did not inhibit the IP 3 R in the absence of added Ca 2ϩ or in the presence of low free Ca 2ϩ concentrations (0.03 or 0.1 M). The same concentration of calmodulin, however, strongly inhibited the IP 3 R at higher free Ca 2ϩ concentrations (0.3 and 1 M).
Because the calmodulin used was lyophilized from a dialysis buffer containing 30 M Ca 2ϩ , four Ca 2ϩ ions were bound to each molecule of calmodulin. We calculated that the addition of calmodulin inhibited the Ca 2ϩ release induced by 1 M IP 3 and 0.3 M free Ca 2ϩ by only 4.3 Ϯ 2.2% (n ϭ 3). We have also tested the effect of calmodulin in an efflux medium containing 6 mM BAPTA instead of 1 mM BAPTA. The addition of 10 M calmodulin to efflux medium containing 6 mM BAPTA and 3.897 mM total Ca 2ϩ increased the calculated free [Ca 2ϩ ] from 0.30 to only 0.31 M. Fig. 2b shows that under these conditions of strong Ca 2ϩ buffering, 10 M calmodulin still inhibited the IP 3 R in the presence of 0.3 and 0.6 M free Ca 2ϩ . The same concentration of calmodulin had again no effect on the IP 3induced Ca 2ϩ release at lower free Ca 2ϩ concentrations. Fig. 2 also illustrates that cytosolic Ca 2ϩ exerted its biphasic effect on the IP 3 R both in the presence and in the absence of calmodulin. In the presence of calmodulin, the inhibition occurred at lower Ca 2ϩ concentrations. As a consequence, the bell-shaped Ca 2ϩ activation curve of the IP 3 R became narrower in the presence of calmodulin.
The [Ca 2ϩ ]-dependence of the IP 3 -induced Ca 2ϩ release was markedly different when Ca 2ϩ was buffered with 1 mM (Fig. 2a) or 6 mM BAPTA (Fig. 2b). Both the stimulatory and inhibitory effects of Ca 2ϩ were more pronounced at the higher concentration of BAPTA. It is possible that this difference is caused by the postulated local [Ca 2ϩ ] rise in the vicinity of the IP 3 Rs as a result of the passive Ca 2ϩ leak from the stores (17,18). This [Ca 2ϩ ] rise will be less pronounced in the presence of 6 mM BAPTA, thereby reducing the IP 3 -induced Ca 2ϩ release at the lowest Ca 2ϩ concentration from 46.9 Ϯ 2.1 to 30.3 Ϯ 2.3%/2 min (n ϭ 5). As a consequence the stimulatory effect of elevating the [Ca 2ϩ ] was more pronounced in the presence of 6 mM BAPTA. An alternative possibility could be that the IP 3 R at low free Ca 2ϩ concentrations is inhibited by the Ca 2ϩ -free form of BAPTA, which is the predominant form of the chelator under these conditions (19,20). Such inhibition would be more pronounced at 6 mM BAPTA, which could again explain why the release in the absence of Ca 2ϩ was reduced at the higher concentration of BAPTA. However, we have previously shown that this inhibitory effect was relatively small in A7r5 cells (21).
The Inhibition of the IP 3 R by Calmodulin Is Dose-dependent- Fig. 3a illustrates the Ca 2ϩ release induced by 1 M IP 3 and 0.3 M free Ca 2ϩ in the presence of various concentrations of calmodulin. Calmodulin inhibited the IP 3 R with an IC 50 of 4.6 M and a Hill-coefficient of 1.0, which is consistent with a single interaction with no evidence for cooperativity between the subunits of the IP 3 R tetramer.
Inhibitory Effect of Calmodulin on the Ca 2ϩ Release Induced by Various IP 3 Concentrations- Fig. 3b shows the Ca 2ϩ release as a function of the [IP 3 ] in the absence (closed circles) and presence (open circles) of 10 M calmodulin in a medium containing 0.3 M free Ca 2ϩ . IP 3 stimulated the IP 3 R with an EC 50 of 0.25 M IP 3 in the absence of calmodulin and with an EC 50 of 2.9 M IP 3 in the presence of calmodulin. Calmodulin not only increased the EC 50 for IP 3 -induced Ca 2ϩ release but also decreased the maximal Ca 2ϩ release induced by high IP 3 concentrations. Interestingly, the inhibition was relatively more pronounced at lower IP 3 concentrations, e.g. 10 M calmodulin caused an 82% inhibition of the Ca 2ϩ release induced by 0.3 M IP 3 , whereas that in the presence of 300 M IP 3 was only inhibited by 20%.
[Sr 2ϩ ] Dependence of the Effect of Calmodulin on the IP 3induced Ca 2ϩ Release-The inhibitory effects of calmodulin were clearly dependent on the presence of Ca 2ϩ . To discriminate whether calmodulin acted by potentiating the inhibitory effects of Ca 2ϩ or whether the requirement for Ca 2ϩ to see the inhibition by calmodulin reflected the formation of Ca 2ϩ /calmodulin, we have studied the effect of calmodulin in the pres-ence of various Sr 2ϩ concentrations. Sr 2ϩ is only 3-fold less potent than Ca 2ϩ in activating the liver IP 3 R but is 600-fold less potent in inhibiting it (22). A similar effect was observed in A7r5 cells, where Sr 2ϩ up to 100 M induced a concentrationdependent decrease in the EC 50 for IP 3 -induced Ca 2ϩ release, whereas Ca 2ϩ induced a biphasic effect with low Ca 2ϩ concentrations decreasing the EC 50 and higher Ca 2ϩ concentrations increasing it (23). Fig. 4 shows the Ca 2ϩ release induced by 1 M IP 3 in the presence of increasing Sr 2ϩ concentrations. The closed bars confirm that Sr 2ϩ activated the IP 3 R and that no significant inhibition was observed at 30 M Sr 2ϩ . The hatched bars show the effect of 10 M calmodulin. Calmodulin did not inhibit the IP 3 R in the absence of added Sr 2ϩ or in the presence of low free Sr 2ϩ concentrations (1 or 3 M). The same concentration of calmodulin, however, strongly inhibited the IP 3 R at higher free Sr 2ϩ concentrations (10 and 30 M), which by themselves were not inhibitory. Because Sr 2ϩ binds to calmodulin (24) but does not inhibit the IP 3 R in the absence of calmodulin, we conclude that the dependence of the calmodulin inhibition on the presence of Sr 2ϩ or Ca 2ϩ represents the formation of Sr 2ϩ /calmodulin or Ca 2ϩ /calmodulin. The inhibition occurred at higher Sr 2ϩ concentrations than Ca 2ϩ concentrations, probably because Sr 2ϩ was 30 times less effective than Ca 2ϩ in binding to the high affinity Ca 2ϩ -binding sites of calmodulin (24).
Conclusions-We observed that calmodulin inhibited the IP 3 -induced Ca 2ϩ release if the free cytosolic [Ca 2ϩ ] was 0.3 M or higher, whereas there was no effect at lower free Ca 2ϩ concentrations. Calmodulin therefore shifted the Ca 2ϩ -dependent inhibition of the IP 3 R toward lower free Ca 2ϩ concentrations without affecting the Ca 2ϩ -dependent activation. This results in a narrower bell-shaped dependence of the IP 3 R on Ca 2ϩ , which may be important for inducing the termination of intracellular Ca 2ϩ spikes. Because all experiments were done in the absence of Mg-ATP and because the inhibition was not affected by 10 M KN62, the involvement of a Ca 2ϩ -and calmodulin-dependent protein kinase can be excluded. The IP 3 R interacts with calmodulin in a Ca 2ϩ -dependent (7-9) and a Ca 2ϩ -independent way (9 -11). It has been proposed that the Ca 2ϩ -independent interaction was responsible for the inhibition of the release (9), because calmodulin inhibited IP 3 -induced Ca 2ϩ release in the presence of 0.2 M free Ca 2ϩ and inhibited [ 3 H]IP 3 binding at all free Ca 2ϩ concentrations tested. Our data on the effects of calmodulin on the Ca 2ϩ release induced by IP 3 and a broad range of cytosolic Ca 2ϩ concentrations now indicate that the inhibition of the IP 3 -induced Ca 2ϩ release is strictly dependent on the formation of a Ca 2ϩ /calmodulin or a Sr 2ϩ /calmodulin complex.
The results in the present study showing that the inhibitory effect of calmodulin on IP 3 -induced Ca 2ϩ release was dependent upon cytosolic Ca 2ϩ or Sr 2ϩ are difficult to reconcile with previous studies demonstrating a Ca 2ϩ -independent inhibition of IP 3 binding to purified cerebellar IP 3 Rs (9), cerebellar membranes (9), type-1 IP 3 Rs expressed in Sf9 cells (10,11), and the recombinant IP 3 -binding domain of the type-1 IP 3 R (11). It was technically impossible to study the effects of calmodulin on [ 3 H]IP 3 binding to A7r5 microsomes. Indeed, despite repeated attempts at different ligand concentrations, the low density of IP 3 Rs in A7r5 cells precluded the detection of IP 3 binding at neutral pH values, although it can be easily measured at alkaline pH (25). The measurements at neutral pH were necessary, because the reported Ca 2ϩ -independent effects of calmodulin on [ 3 H]IP 3 binding disappeared at alkaline pH, probably due to a conformational change of calmodulin (9,11). The apparent discrepancy between the functional data reported in the present study and the [ 3 H]IP 3 binding data (9 -11) may be due to the existence of multiple calmodulin-binding or Ca 2ϩ -binding sites that play a role and/or to the fact that IP 3 binding, which is only one step, albeit a crucial step, for channel opening, is not equivalent to the channel activity itself.