HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 20, 17571-17579, May 17, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/20/17571    most recent M200524200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Swatton, J. E. Articles by Taylor, C. W. Search for Related Content PubMed PubMed Citation Articles by Swatton, J. E. Articles by Taylor, C. W. Social Bookmarking                      What's this?

Fast Biphasic Regulation of Type 3 Inositol Trisphosphate Receptors by Cytosolic Calcium^*

Jane E. Swatton and Colin W. Taylor

From the Department of Pharmacology, University of Cambridge, Tennis Court Road, Cambridge CB2 1PD, United Kingdom

Received for publication, January 17, 2002, and in revised form, February 21, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In cytosol-like medium (CLM) with a free [Ca²⁺] of 200 nM, a supramaximal concentration of inositol 1,4,5-trisphosphate (IP₃) (30 µM) evoked ⁴⁵Ca²⁺ release from type 3 IP₃ receptors only after a latency of 48 ± 6 ms; this latency could not be reduced by increasing the IP₃ concentration. In CLM containing a low free [Ca²⁺] (~4 nM), 300 µM IP₃ evoked ⁴⁵Ca²⁺ release after a latency of 66 ± 11 ms; this was reduced to 14 ± 3 ms when the [Ca²⁺] was 1 mM. Preincubation with CLM containing 100 µM Ca²⁺ caused a rapid (half-time = 33 ± 9 ms), complete, and fully reversible inhibition that could not be overcome by a high concentration of IP₃ (300 µM). Hepatic (type 2) IP₃ receptors were not inhibited by Ca²⁺ once they had bound IP₃, but 100 µM Ca²⁺ rapidly inhibited type 3 IP₃ receptors whether it was delivered before addition of IP₃ or at any stage during a response to IP₃. Ca²⁺ increases the affinity of IP₃ for hepatic receptors by slowing IP₃ dissociation, but Ca²⁺ had no effect on IP₃ binding to type 3 receptors. The rate of inhibition of type 3 IP₃ receptors by Ca²⁺ was faster than the rate of IP₃ dissociation, and occurred at similar rates whether receptors had bound a high (adenophostin) or low affinity (3-deoxy-3-fluoro-IP₃) agonist. Dissociation of agonist is not therefore required for Ca²⁺ to inhibit type 3 IP₃ receptors. We conclude that type 2 and 3 IP₃ receptors are each biphasically regulated by Ca²⁺, but by different mechanisms. For both, IP₃ binding causes a stimulatory Ca²⁺-binding site to be exposed allowing Ca²⁺ to bind and open the channel. IP₃ binding protects type 2 receptors from Ca²⁺ inhibition, but type 3 receptors are inhibited by Ca²⁺ whether or not they have IP₃ bound. Increases in cytosolic [Ca²⁺] will immediately inhibit type 3 receptors, but inhibit type 2 receptors only after IP₃ has dissociated.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Inositol 1,4,5-trisphosphate (IP₃)¹ receptors belong to a family of intracellular channels that mediate the release of Ca²⁺ from intracellular stores in response to a range of physiological stimuli (1). The three mammalian subtypes of the IP₃ receptor (types 1-3), which form both homotetrameric (2) and heterotetrameric (3, 4) complexes, are regulated by both IP₃ and Ca²⁺ (5). This interplay between IP₃ and Ca²⁺ is likely to determine the complex Ca²⁺ signals evoked by receptors that stimulate IP₃ formation.

The similarities between the IP₃ receptor subtypes are presently more striking than the differences. Each subtype is predicted to form a similar structure with an N-terminal IP₃-binding domain separated by some 1500 residues from a C-terminal region that includes six membrane-spanning domains, the last two of which together with the intervening loop line the pore of the channel (6, 7). Within the functional receptor, however, the IP₃-binding domain of one subunit appears to be intimately associated with the pore region of a neighboring subunit (8). The ion permeation properties of the channel are also similar for each receptor subtype (9, 10), consistent with the highly conserved sequences found within the pore-forming region (11). The most important differences between IP₃ receptor subtypes, which are certainly differentially expressed (11), are therefore likely to be in their modulation (12-15) and perhaps in their subcellular distribution (16, 17).

Biphasic regulation of IP₃ receptors by cytosolic Ca²⁺ is widespread (5), with many reports confirming that modest increases in Ca²⁺ stimulate channel opening, while more substantial increases are inhibitory for type 1 (18) and type 2 (19) IP₃ receptors, as well as for the IP₃ receptors from Xenopus (20) and insects (21, 22). The effects of Ca²⁺ on type 3 IP₃ receptors have aroused more controversy. In bilayer recordings from the type 3 IP₃ receptors of RIN-5F cells, even very high concentrations of Ca²⁺ failed to inhibit channel activity (23). IP₃-evoked Ca²⁺ release from DT40 cells lacking types 1 and 2 IP₃ receptors also appeared to be resistant to Ca²⁺ inhibition (24). But many other studies of both cells expressing predominantly type 3 IP₃ receptors (25), including RINm5F cells (26), and of recombinant type 3 IP₃ receptors (15, 27) have reported biphasic regulation of IP₃ receptor behavior by cytosolic Ca²⁺. It seems likely, therefore, that biphasic regulation by cytosolic Ca²⁺ may be an ubiquitous feature of IP₃ receptors, although it is far from clear that the underlying mechanisms are the same (28, 29).

The rapid kinetics of IP₃ receptor regulation by IP₃ and Ca²⁺ are likely to be significant in determining the contributions of different IP₃ receptor subtypes to the complex regenerative Ca²⁺ signals in intact cells (30, 31). Despite the importance of resolving the behavior of IP₃ receptors on a time scale appropriate to understanding rapid Ca²⁺ release events in intact cells (30), most studies of IP₃ receptor gating have been limited to either steady-state measurements (15, 23) or to examining rates of Ca²⁺ release with very limited temporal resolution.

We previously used rapid superfusion of permeabilized rat hepatocytes loaded with ⁴⁵Ca²⁺ to examine the kinetics of IP₃-evoked Ca²⁺ release by type 2 IP₃ receptors with a temporal resolution of 9 ms (31, 32). We concluded that IP₃ binding determined whether Ca²⁺ stimulated or inhibited channel opening (31, 32). Here, we apply similar methods to examine IP₃-evoked Ca²⁺ release from permeabilized RINm5F cells, which both we (26) and others (33) have shown to express predominantly type 3 IP₃ receptors.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- IP₃ was from American Radiolabeled Chemicals (St. Louis, MO). ⁴⁵Ca²⁺ (238.5 Ci/mol) was from ICN (Thame, Oxfordshire, UK) and [³H]IP₃ (37 Ci/mmol) was from Amersham Pharmacia Biotech (Little Chalfont, UK). Ionomycin, synthetic adenophostin A, and 3-deoxy-3-fluoro IP₃ were from Calbiochem (Nottingham, UK). Thapsigargin was from Alomone Labs (Jerusalem, Israel). All other materials were from suppliers listed earlier (26).

Isolation of Cells-- RINm5F rat insulinoma cells (a gift from Dr. Peter Brown, University of Manchester, UK) were cultured at 37 °C in 5% CO₂ in RPMI 1640 medium containing 2 mM L-glutamine (Invitrogen, Paisley, UK) and 5% fetal calf serum (Sigma, Poole, UK). Cells were passaged every 3-4 days when confluent. Hepatocytes were isolated from the livers of male Wistar rates (200-300 g) by collagenase digestion (34) and kept for up to 6 h at 4 °C in Eagle's minimal essential medium containing 26 mM NaHCO₃ and 2% bovine serum albumin.

Measurement of Unidirectional ⁴⁵Ca²⁺ Efflux from Permeabilized RINm5F Cells-- Confluent RINm5F cells (passages 90-100) were scraped from a flask, washed by centrifugation (650 × g, 2 min), and resuspended (1 × 10⁶ cells/ml) in Ca²⁺-free cytosol-like medium (CLM: 140 mM KCl, 20 mM NaCl, 2 mM MgCl₂, 1 mM EGTA, 20 mM Pipes, pH 7.0, at 37 °C). The cells were permeabilized by incubation with saponin (10 µg/ml) for 5 min at 37 °C (26), washed by centrifugation (650 × g, 2 min), and resuspended in CLM containing ~200 nM free Ca²⁺ (300 µM total Ca²⁺), carbonyl cyanide p-trifluoromethoxyphenylhydrazone (10 µM), and ⁴⁵Ca²⁺ (10 µCi/ml). After addition of ATP (1.5 mM), creatine phosphate (5 mM), and creatine phosphokinase (5 units/ml), cells were incubated for 15 min at 37 °C during which their intracellular stores reached a steady-state Ca²⁺ content of 141 ± 14 pmol/10⁶ cells (n = 3). Cells were rapidly removed from the loading medium by 8-fold dilution into CLM (200 nM free Ca²⁺, 20 °C) and a brief centrifugation (650 × g, 30 s), before resuspension (6 × 10⁶ cells/ml) in CLM (200 nM free Ca²⁺, 20 °C). The effects of IP₃ or its analogues on the ⁴⁵Ca²⁺ content of the stores were determined after 2-min incubations at 20 °C (to allow direct comparison with superfusion experiments) and in the presence of thapsigargin (1 µM) to inhibit the endoplasmic reticulum Ca²⁺-ATPase. The incubations were terminated by rapid filtration through Whatman GF/C filters using a Brandel receptor binding harvester (26), and the ⁴⁵Ca²⁺ content of the stores determined by liquid scintillation counting. Active ⁴⁵Ca²⁺ accumulation was defined as the ⁴⁵Ca²⁺ that could be released by ionomycin (10 µM).

Measurement of ⁴⁵Ca²⁺ Release by Rapid Superfusion-- Permeabilized RINm5F cells loaded with ⁴⁵Ca²⁺ (30 µCi/ml) were immobilized on a nitrocellulose and glass fiber filter array held in the chamber of a rapid superfusion apparatus; the apparatus has been described elsewhere (34). Briefly it allows rates of ⁴⁵Ca²⁺ release from the immobilized cells to be measured with a temporal resolution of up to 9 ms, as CLM flows (at 2 ml/s) from pressurized cylinders to the cells and then (with the ⁴⁵Ca²⁺ released from the cells) into vials arranged around a circular fraction collector. Because the medium bathing the immobilized cells is continuously replaced during superfusion, unidirectional ⁴⁵Ca²⁺ efflux can be measured without the addition of thapsigargin used in the conventional ⁴⁵Ca²⁺ efflux experiments (see above). Inclusion of an inert marker ([³H]inulin) in the superfusing media allowed the arrival of a stimulus to be precisely related to changes in ⁴⁵Ca²⁺ efflux. Under the conditions used for these experiments, the half-time (t_1/2) for exchange of the media was 31 ± 1 ms (n = 3). All experiments were performed at 20 °C. The size of the intracellular ⁴⁵Ca²⁺ pool was calculated for each experiment by summing all ⁴⁵Ca²⁺ released during the stimulation with that released at the end of the experiment by Triton X-100 (0.5%). In some experiments (Fig. 2) very rapid increases in free [Ca²⁺] were achieved using a pulsing protocol (31): a 50-ms pulse of CLM containing 1.65 mM Ca²⁺ followed by continuous superfusion with CLM containing 1.1 mM Ca²⁺ allowed the free [Ca²⁺] to be increased from ~200 nM to 100 µM within 50 ms.

Permeabilized hepatocytes were prepared and loaded with ⁴⁵Ca²⁺ as reported previously (34), and rates of IP₃-evoked ⁴⁵Ca²⁺ release were then measured using rapid superfusion as described for RINm5F cells. Free [Ca²⁺] were predicted using the computer program WinMAXC version 2.05 (C. Patton, Stanford University, CA 93950), and then measured using either fura 2 or a Ca²⁺-sensitive electrode (19, 26).

[³H]IP₃ Equilibrium-competition Binding-- Saponin-permeabilized RINm5F cells were washed by centrifugation (650 × g, 2 min), resuspended in TE medium (50 mM Tris, 1 mM EDTA, pH 8.3, at 2 °C), and incubated (2.75 × 10⁶ cells in 200 µl) with [³H]IP₃ (3 nM) and competing ligands for 5 min at 2 °C. Bound and free ligand were separated by filtration (35), and ³H activity was determined by liquid scintillation counting. Specific [³H]IP₃ binding (~500 disintegrations/min) was typically ~70% of total binding. Equilibrium binding data were analyzed to provide dissociation constants (K_d) for competing ligands as reported previously (35).

Analysis-- Because the ⁴⁵Ca²⁺ content of the stores declines as a superfusion experiment progresses, we express many of the responses to IP₃ as fractional release rates, where the amount of ⁴⁵Ca²⁺ released by IP₃ is expressed as a fraction of the ⁴⁵Ca²⁺ remaining within the IP₃-sensitive stores at the beginning of that interval (34). By expressing rates of ⁴⁵Ca²⁺ release relative to the amount of ⁴⁵Ca²⁺ available for release, this form of analysis effectively isolates the activity of the IP₃ receptor from changes in the ⁴⁵Ca²⁺ content of the stores. We thereby expect stable fractional release rates unless the IP₃ receptor changes its behavior (36).

IP₃-evoked Ca²⁺ release was calculated by subtracting the basal rate of ⁴⁵Ca²⁺ release from the ⁴⁵Ca²⁺ detected in each fraction during stimulation with IP₃. By measuring basal rates of ⁴⁵Ca²⁺ release over protracted times (not shown), we established the following empirical relationship between the fraction of the Ca²⁺ stores released (S), the fractional rate of unstimulated ⁴⁵Ca²⁺ release at the beginning of the superfusion (I) and the fractional rate of unstimulated ⁴⁵Ca²⁺ release from partially depleted stores (P).

(Eq. 1)

This equation was used to separately compute the basal rate of ⁴⁵Ca²⁺ release from IP₃-sensitive and IP₃-insensitive Ca²⁺ stores throughout the experiment. Concentration-effect relationships were fitted to logistic equations using nonlinear curve fitting (KaleidaGraph, Synergy Software, PA) (37).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Delayed Activation of Type 3 IP₃ Receptors-- In permeabilized RINm5F cells, IP₃ caused a rapid increase in the rate of ⁴⁵Ca²⁺ release, with the half-maximal rate occurring with an IP₃ concentration of 4.96 ± 0.82 µM (n = 3) and the maximal rate of ⁴⁵Ca²⁺ release occurring with 10 µM IP₃. During stimulation with 30 µM IP₃ in normal CLM, the cells were therefore exposed to a supramaximal concentration of IP₃ within about 15 ms, but the first detectable release of ⁴⁵Ca²⁺ occurred only after a delay of 48 ± 6 ms (n = 3) (Fig. 1A, i). This long absolute latency, which is similar to that observed previously for hepatic IP₃ receptors (32), was not shortened by further increasing the IP₃ concentration (Fig. 1A, ii): the latency was 43 ± 9 ms (n = 3) when cells were stimulated with 300 µM IP₃ (Table I). In the latter experiments, cells would have been exposed to at least 7 times the maximally effective IP₃ concentration (~70 µM) within 10 ms.

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Cytosolic Ca2+ shortens the latency for IP3 receptor activation. A, rates of 45Ca2+ efflux are shown for permeabilized cells stimulated with 30 µM (i) or 300 µM (ii) IP3 in normal CLM (free [Ca2+] = 200 nM). The open bar shows the switch to CLM containing IP3 and the dotted line ([3H]inulin) denotes the arrival of IP3 in the effluent. B, rates of 45Ca2+ efflux are shown in response to 300 µM IP3 in either nominally Ca2+-free CLM () or CLM containing 1 mM free Ca2+ (). The dotted line ([3H]inulin) denotes the simultaneous delivery of the IP3 and Ca2+. C, cells superfused with Ca2+-free CLM (100 µM EGTA) were briefly (30 ms, open bar) stimulated with 300 µM IP3 before rapidly (solid bar) increasing the free [Ca2+] to 1 mM using EGTA-free CLM. The dashed line ([3H]inulin) shows the arrival of the high-Ca2+ CLM. For A-C, results (mean ± S.E., n = 3) are expressed as percentages of the total 45Ca2+ content of the stores. D and E, the free [Ca2+] of superfusing media are shown after switching from nominally Ca2+-free CLM (1 mM (D) or 100 µM (E) EGTA) to CLM containing either 2 mM Ca2+ with 1 mM EGTA (D), or EGTA-free CLM containing 1 mM Ca2+ (E). The total Ca2+ concentrations were calculated from the measured t1/2 for mixing (31 ± 1 ms) using a monoexponential equation: T = A(1-exp0.0219t), where T = the total Ca2+ concentration at time t (in ms) from the onset of mixing (when [Ca2+] = 0 mM), and A is the final total [Ca2+] (2 mM in D, 1 mM Ca2+ in E). The [EGTA] remained constant (at 1 mM) in panel D, but in E it decreased from 100 µM to EGTA-free during the mixing. The total [EGTA] (T) at time (t) from the onset of mixing was calculated from: T = B·exp0.0219t, where B is the initial [EGTA] (100 µM). At each time (D and E), the free [Ca2+] was calculated from the total [Ca2+] and [EGTA] using WinMAXC (see "Experimental Procedures").

While the latency could not be reduced by increasing the IP₃ concentration, it was reduced and ultimately abolished by increasing the free [Ca²⁺] of the CLM. In nominally Ca²⁺-free CLM (free [Ca²⁺] ~4 nM), the latency after stimulation with 300 µM IP₃ was 66 ± 11 ms and this was reduced as the free [Ca²⁺] was increased (Table I), such that when the free [Ca²⁺] was 1 mM, the latency was only 14 ± 3 ms (Table I, Fig. 1B). It is noteworthy that although increasing the free [Ca²⁺] shortened both the latency and the time taken for the rate of ⁴⁵Ca²⁺ release to reach its peak, it had no significant effect on the peak rate of ⁴⁵Ca²⁺ release (Table I).

Although there was an immediate rise in the rate of ⁴⁵Ca²⁺ release when IP₃ (300 µM) was delivered in the presence of 1 mM free Ca²⁺, there was a lag of ~40 ms before the response attained its fastest rate (Fig. 1B). We considered whether this slow take-off might simply result from a relatively slow increase in free [Ca²⁺] as Ca²⁺-free CLM (1 mM EGTA) was replaced by (t_1/2 = 31 ± 1 ms) the high-Ca²⁺ CLM (2 mM Ca²⁺, 1 mM EGTA) causing an abrupt increase in free [Ca²⁺] as the buffering capacity of the EGTA was exceeded. The simulation shown in Fig. 1D confirms that under the conditions used for these experiments, a switch to high-Ca²⁺ CLM caused the free [Ca²⁺] surrounding the cells to abruptly increase to several micromolar only after a delay of ~25 ms. By reducing the EGTA concentration in the CLM used initially to 100 µM and then stepping to CLM containing 1 mM Ca²⁺ without EGTA, the free [Ca²⁺] surrounding the cells was predicted to increase almost linearly to 1 mM and to exceed 100 µM within 8 ms (Fig. 1E).

We used this Ca²⁺-delivery protocol to very rapidly expose cells to a high free [Ca²⁺] during the latent period of the response to a supramaximal concentration of IP₃. Cells were first exposed to IP₃ (300 µM) in nominally Ca²⁺-free CLM (100 µM EGTA) for 30 ms and then to high-Ca²⁺ in EGTA-free CLM (Fig. 1C). During the initial exposure to IP₃, there was no stimulated release of ⁴⁵Ca²⁺ because this interval lies within the latent period (~66 ms in nominally Ca²⁺-free CLM). After rapidly increasing the free [Ca²⁺], there was an immediate increase in the rate of ⁴⁵Ca²⁺ release with no detectable latency (Fig. 1C).

Rapid and Reversible Inhibition of IP₃-evoked Ca²⁺ Release by Cytosolic Ca²⁺-- Stimulation of permeabilized RINm5F cells with a maximal concentration of IP₃ (10 µM) caused a rapid increase in the rate of ⁴⁵Ca²⁺ release to a peak, followed by a decay over several seconds (Fig. 2A). Preincubation of the cells for 1.2 s with CLM containing a free [Ca²⁺] of 100 µM completely abolished the subsequent response to IP₃ (Fig. 2A). The increase in free [Ca²⁺] itself only minimally increased the rate of ⁴⁵Ca²⁺ release from 0.16 ± 0.01%/100 ms to 0.23 ± 0.01%/100 ms (n = 15) (Fig. 2D), indicating that depletion of intracellular Ca²⁺ stores could not account for the loss of response to IP₃, and (see below) that increasing the free [Ca²⁺] does not itself significantly stimulate ⁴⁵Ca²⁺ release in the absence of IP₃. The reversibility of the inhibition under conditions where the stores could not re-load with Ca²⁺ (see below) further confirms that the lack of response to IP₃ results from inhibition of IP₃ receptors rather than store depletion.

View larger version (26K): [in this window] [in a new window]   Fig. 2.   Inhibition of IP3 receptors by cytosolic Ca2+. A, rates of 45Ca2+ release evoked by 10 µM IP3 are shown for cells stimulated in normal CLM (free [Ca2+] = 200 nM, ) or after preincubation for 50 ms (), 150 ms () or 1.2 s () with CLM containing 100 µM free Ca2+. The double-pulse protocol (see "Experimental Procedures") was used to increase the free [Ca2+] to 100 µM within 50 ms. B, rates of 45Ca2+ release evoked by 10 µM IP3 are shown under control conditions () or after pretreatment with 100 µM free Ca2+ for 1.2 s before stimulation with 10µM () or 300 µM () IP3. For A and B, dashed line shows the arrival of [3H]inulin (included with IP3) and results (mean ± S.E., n = 3) are expressed as percentages of the total Ca2+ content of the stores. C, effects of varying the duration of the preincubation with 100 µM Ca2+ on the peak rates of 45Ca2+ release evoked by 10 µM IP3. Results are shown as percentages of the peak rate of 45Ca2+ release under control conditions. D, cells were washed with CLM containing 200 nM free Ca2+ before the free Ca2+ was increased to 100 µM (shown by the dashed line). Results are expressed as percentages of the total Ca2+ content of stores.

The inhibition of IP₃ receptors by high Ca²⁺ could not be overcome by increasing the concentration of IP₃ (Fig. 2B). The normal peak rate of ⁴⁵Ca²⁺ release evoked by 10 µM IP₃ was 3.03 ± 0.15%/100 ms (n = 3). After preincubation with high Ca²⁺ for 1.2 s, the peak rate of Ca²⁺ release fell to 0.027 ± 0.008%/100 ms (n = 3) after stimulation with 10 µM IP₃, and to 0.24 ± 0.07%/100 ms (n = 3) after stimulation with 300 µM IP₃. Inhibition of IP₃ receptors by Ca²⁺ has been proposed to be mediated by calmodulin (38, 39), by other Ca²⁺-binding proteins (40) or by residues within the IP₃ receptor (41). We have not established the site through which Ca²⁺ inhibits type 3 IP₃ receptors, although it is unlikely to be calmodulin because neither calmidazolium (20 µM) nor a peptide inhibitor derived from Ca²⁺-calmodulin-dependent protein kinase (10 µM) (13) prevented complete inhibition of the IP₃ receptor by CLM containing 100 µM Ca²⁺ (not shown).

By varying the duration of the preincubation with high-Ca²⁺ CLM before stimulating with 10 µM IP₃, the half-time for inhibition by cytosolic Ca²⁺ was established (39 ± 3 ms, n = 3) (Fig. 2C). We were concerned that with such rapid inhibition, the time taken for IP₃ to reach its maximally effective concentration was likely to significantly affect our measurement of the kinetics of Ca²⁺ inhibition. To resolve the issue, the time course of Ca²⁺ inhibition was investigated by preincubating with high-Ca²⁺ CLM and then assessing IP₃ receptor activity using 300 µM IP₃, which allowed the IP₃ concentration to exceed 70 µM within 10 ms. The half-time for Ca²⁺ inhibition determined using this method (33 ± 9 ms) was indistinguishable from that observed using 10 µM IP₃ as the test pulse (39 ± 3 ms).

To assess whether inhibition by cytosolic Ca²⁺ was reversible, cells were pretreated for 1.2 s with CLM containing 100 µM free [Ca²⁺] to fully inhibit IP₃ receptors, and then superfused for various recovery periods in normal CLM (free [Ca²⁺] = 200 nM) before assessing IP₃ receptor activity by addition of 10 µM IP₃. The results demonstrate that inhibition by cytosolic Ca²⁺ reversed rapidly and fully with a half-time of 102 ± 20 ms (n = 3) (Figs. 3, A and B).

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Rapid recovery from Ca2+ inhibition. A, cells pretreated with 100 µM Ca2+ for 1.2 s were allowed to recover for various intervals in CLM containing 200 nM free Ca2+ before stimulation with 10 µM IP3. The control () and fully inhibited () responses are shown together with the responses recorded after recovery periods of 100 ms (), 400 ms (), 1 s (), or 8 s (). Dashed line shows arrival of [3H]inulin included with IP3. B, peak rates of 45Ca2+ release (% of control response) are plotted as a function of the recovery period. Results (mean ± S.E., n = 3) are expressed as percentages of the total Ca2+ content of the intracellular stores.

Cytosolic Ca²⁺ Inhibits Type 3 IP₃ Receptors with IP₃ Bound-- We first confirmed our earlier observation (31) that hepatic (largely type 2) IP₃ receptors are not inhibited by cytosolic Ca²⁺ when they have IP₃ bound (Fig. 4, A and C). Although preincubation with Ca²⁺ (100 µM) rapidly and completely inhibited hepatic IP₃ receptors (31), the same high-Ca²⁺ CLM delivered after exposure to 10 µM IP₃ caused a modest stimulation of Ca²⁺ release rather than inhibition (Fig. 4, A and C). Similar experiments with the type 3 IP₃ receptors of RINm5F cells gave very different results (Fig. 4, B and D). Increasing the cytosolic free [Ca²⁺] (to 100 µM) during a response to a supramaximal concentration of IP₃ (100 µM) first caused a significant increase in the rate of ⁴⁵Ca²⁺ release, but then rapidly inhibited the response (Fig. 4, B and D). This occurred whether the high Ca²⁺ was delivered at the peak of the response to IP₃ (250 ms after IP₃ addition) or during a later phase of the response (after 1 s) (Fig. 4, D and E). The stimulatory phase of the response to increased Ca²⁺ clearly reflects an interaction with the IP₃ receptor because the same stimulus in the absence of IP₃ caused no significant increase in ⁴⁵Ca²⁺ release (Fig. 2D). From the monoexponential curve fits to the decaying phases of the Ca²⁺ inhibition, we established that the half-times for Ca²⁺ inhibition were ~170 ms regardless of the duration of the prior stimulation with IP₃, or the concentration of IP₃ used. Addition of high-Ca²⁺ after stimulation with 10 or 100 µM IP₃ for 250 ms, caused the channels to close with half-times of 174 ± 32 and 170 ± 21 ms, respectively (n = 3); and when the duration of exposure to IP₃ was increased to 1 s before addition of high-Ca²⁺, the half-times for channel closure were 172 ± 22 ms (10 µM IP₃) and 173 ± 24 ms (100 µM IP₃).

View larger version (29K): [in this window] [in a new window]   Fig. 4.   Ca2+ rapidly inhibits type 3, but not type 2, IP3 receptors with IP3 bound. Permeabilized hepatocytes (A and C) or RINm5F cells (B, D, and E) were stimulated with 10 µM IP3 (hepatocytes) or 100 µM IP3 (RINm5F cells) in CLM containing 200 nM free Ca2+ either throughout the experiment (), or for 250 ms (B and D; ), 300 ms (A and C; ), or 1 s (E; ) before increasing the free [Ca2+] to 100 µM. The dashed line shows the arrival of [3H]inulin included with the IP3 in normal CLM and then (at 5 times greater [3H]inulin concentration) in the high-Ca2+ CLM. Open bars denote the presence of IP3 and solid bars the presence of high-Ca2+. Results (mean ± S.E., n = 3-5) are expressed as either percentages of the total Ca2+ content of the stores (A and B) or as fractional release rates (C-E). Note that because the percentage of the Ca2+ stores that are sensitive to IP3 is different in hepatocytes (~20%) and RINm5F cells (~55%), the relationship between peak 45Ca2+ release rate and peak fractional 45Ca2+ release rate is different for the two cell types.

In hepatocytes, cytosolic Ca²⁺ increases the affinity of the IP₃ receptor for IP₃ (42, 43). Equilibrium-competition binding experiments using permeabilized RINm5F cells demonstrated that Ca²⁺ did not increase the affinity of type 3 IP₃ receptors for IP₃. The equilibrium-dissociation constant (K_d) for IP₃ in Ca²⁺-free medium was 13 ± 3 nM (n = 4) and it was 18 ± 4 nM (n = 5) when the free [Ca²⁺] was 100 µM. We therefore considered the possibility that if IP₃ binding protects IP₃ receptors from Ca²⁺ inhibition, it might be more effective in hepatocytes where increased Ca²⁺ slows IP₃ dissociation, than in RINm5F cells where Ca²⁺ does not increase the affinity of IP₃ for its receptor. We therefore considered two possible explanations for the rapid inhibition of type 3 IP₃ receptors by Ca²⁺ in the continued presence of IP₃. First, Ca²⁺ inhibition of types 2 and 3 IP₃ receptors may be fundamentally different, with Ca²⁺ inhibiting type 3 receptors irrespective of whether they have IP₃ bound. Alternatively, IP₃ binding may prevent Ca²⁺ inhibition for both subtypes, but the rate of IP₃ dissociation from type 3 receptors may be too rapid to prevent Ca²⁺ inhibition during simultaneous exposure to IP₃ and high Ca²⁺ (Fig. 4).

We reasoned that if the latter explanation were true, the rate of Ca²⁺ inhibition could be no faster than the rate of IP₃ dissociation, and the rate of Ca²⁺ inhibition would be slower if the receptor were occupied by an agonist with a slower dissociation rate. To distinguish between the two possibilities, we therefore used two additional agonists that differ massively in their affinities for IP₃ receptors: adenophostin A (44, 45) and 3-deoxy-3-fluoro-IP₃ (35) typically have about 10-fold higher and 10-fold lower affinity than IP₃, respectively. Conventional ⁴⁵Ca²⁺ efflux assays, performed under conditions that mimic those used for rapid superfusion, established that in RINm5F cells adenophostin A (EC₅₀ = 2.5 ± 0.2 nM) was 32-fold more potent than IP₃, and 3-deoxy-3-fluoro-IP₃ (EC₅₀ = 343 ± 21 nM) was 4-fold less potent than IP₃ (EC₅₀ = 79 ± 2 nM) (Fig. 5A). To directly measure the rates of dissociation of these agonists from the active conformation of the type 3 IP₃ receptor, rates of ⁴⁵Ca²⁺ release were measured as cells were stimulated for 300 ms with 10 µM of each agonist and then as the agonist was rapidly washed out (t_1/2 = 31 ± 1 ms) (Fig. 5B). The results (Fig. 5, B and C) demonstrate that after rapid removal of IP₃, the channels close with t_1/2 = 249 ± 37 ms (n = 3), they close more quickly after removal of 3-deoxy-3-fluoro-IP₃ (t_1/2 = 156 ± 4 ms), and more slowly after removal of adenophostin A (350 ± 36 ms). The 2.2-fold difference in the dissociation rate of adenophostin A and 3-deoxy-3-fluoro-IP₃ is much less than the 137-fold difference in their relative potencies in conventional assays of ⁴⁵Ca²⁺ release (Fig. 5A), suggesting that the high affinity of adenophostin A cannot be due entirely to its slower dissociation rate (46). Nevertheless, the significantly different dissociation rates of the three agonists allowed us to assess whether agonist dissociation limits the onset of Ca²⁺ inhibition.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Ca2+ release evoked by high or low affinity agonists of the IP3 receptor. A, the concentration-dependent effects of adenophostin A (), IP3 (), or 3-deoxy-3-fluoro-IP3 () on the 45Ca2+ content of permeabilized RINm5F cells were determined during a 2-min incubation at 20 °C in CLM without ATP and containing 200 nM free [Ca2+] and 1 µM thapsigargin. B, cells were superfused for 300 ms with 10 µM adenophostin A (), IP3 (), or 3-deoxy-3-fluoro-IP3 () in CLM containing 200 nM free Ca2+ before removal of the agonist. The dashed line shows the presence of [3H]inulin, included with each of the agonists. Results (mean ± S.E., n = 3) are expressed as percentages of the total Ca2+ content of the intracellular stores. C, rates of 45Ca2+ release (from B) are shown for the period beginning 200 ms after the onset of agonist washout, with each rate expressed as a percentage of that at 200 ms. The lines were fitted to monoexponential equations.

In rapid superfusion experiments where permeabilized RINm5F cells were stimulated for 1 s with IP₃ (10 µM), adenophostin A (10 µM), or 3-deoxy-3-fluoro-IP₃ (50 µM) in CLM containing 200 nM free Ca²⁺ before increasing the free [Ca²⁺] to 100 µM, the half-times for Ca²⁺ inhibition were indistinguishable (Fig. 6, A-C), at 125 ± 7 ms for IP₃, 130 ± 19 ms for adenophostin A, and 131 ± 16 ms for 3-deoxy-3-fluoro-IP₃ (n = 3). These results are important for two reasons. First, they demonstrate that the different rates of dissociation of these agonists from the IP₃ receptor have no effect on the time course of the inhibition by high-Ca²⁺. Second, the Ca²⁺ inhibition (t_1/2 ~130 ms) is significantly faster than the rate of dissociation of IP₃ (t_1/2 = 249 ms). Neither result is consistent with a requirement for IP₃ to dissociate from its receptor before it becomes susceptible to Ca²⁺ inhibition.

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Ca2+ rapidly inhibits Ca2+ release evoked by high or low affinity agonists of type 3 IP3 receptors. A-C, cells were superfused with IP3 (10 µM, A), adenophostin A (10 µM, B) or 3-deoxy-3-fluoro-IP3 (50 µM, C) in CLM containing 200 nM free Ca2+ throughout () or for 1 s before increasing the free Ca2+ to 100 µM (). The dashed lines show the arrival of [3H]inulin included with each agonist in 200 nM free Ca2+ and then at 5 times higher concentration with the high-Ca2+ CLM. Open bars denote the presence of agonist and solid bars the presence of high-Ca2+. Results (mean ± S.E., n = 3) are expressed as fractional release rates.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Opening of Type 3 IP₃ Receptors after Sequential Binding of IP₃ and Ca²⁺-- Our earlier rapid superfusion analysis of hepatic (type 2) IP₃ receptors demonstrated that in normal CLM (free [Ca²⁺] ~200 nM), IP₃ stimulated ⁴⁵Ca²⁺ release only after a significant delay that could not be reduced by increasing the IP₃ concentration. We concluded that delayed opening of the channel could not, therefore, be a consequence of slow binding of IP₃ to the four receptor subunits (32). Similar absolute latencies had been reported in intact cells after flash-photolysis of caged IP₃ in Xenopus oocytes (47), endothelial cells (48), and hepatocytes (49). In permeabilized hepatocytes, the latency before channel opening after addition of supramaximal concentrations of IP₃ was, however, affected by the free [Ca²⁺]; it was longer in nominally Ca²⁺-free CLM and almost abolished when the free [Ca²⁺] was very high. Because even prolonged incubation with Ca²⁺ had no greater effect on latency than Ca²⁺ delivered with IP₃, we concluded that Ca²⁺ exerted its stimulatory effect only after IP₃ had bound to the receptor (32). Our model therefore suggested that binding of IP₃ caused a Ca²⁺-binding site to become accessible and the channel opened only after that site had bound Ca²⁺ (Fig. 7A). To account for the very long delays in channel opening at low concentrations of IP₃, the delays in channel opening evoked by maximal IP₃ at limiting Ca²⁺ concentrations, and the steep concentration dependence of the effects of IP₃ on channel opening, we suggested that several of the subunits within the tetrameric receptor were required to complete this sequential binding of IP₃ and then Ca²⁺ before the channel opened to release Ca²⁺ (Fig. 7A).

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Ca regulation of types 2 and 3 IP3 receptors. A, for both subtypes, binding of IP3 causes exposure of a stimulatory Ca2+-binding site (S). Binding of IP3 and Ca2+ to several (perhaps all four) of the subunits is required for channel opening. B, Ca2+ can inhibit type 3 receptors by binding to an inhibitory site (I) whether or not they have IP3 bound, whereas IP3 binding protects type 2 receptors from Ca2+ inhibition. For simplicity, only a single receptor subunit is shown in B and in both panels the Ca2+-binding sites are shown as part of the IP3 receptor (41), although the involvement of accessory binding sites cannot be eliminated.

Our present results with type 3 IP₃ receptors are consistent with the model proposed for type 2 receptors. Maximal concentrations of IP₃ evoked ⁴⁵Ca²⁺ release only after a significant latency that could not be further reduced by even very substantial increases in IP₃ concentration. But the latency was increased in very low-Ca²⁺ CLM and almost abolished in very high-Ca²⁺ CLM (Table I). Furthermore, a rapid increase in free [Ca²⁺] during a response to even a supramaximal concentration of IP₃ (100 µM, Fig. 4, B, D, and E) or other agonists (Fig. 6) triggered a rapid increase in the rate of ⁴⁵Ca²⁺ release before the inhibitory effect of high-Ca²⁺ curtailed the response. Maximal occupancy of IP₃ receptors by IP₃ is not, therefore, sufficient to cause maximal channel opening, but it primes the receptor such that increasing the rate of Ca²⁺ binding to the newly exposed Ca²⁺-binding site stimulates further channel opening.

Our results suggest that for both types 2 and 3 receptors, IP₃ binding causes exposure of a Ca²⁺-binding site that must then bind Ca²⁺ for the channel to open (Fig. 7A). In their steady-state single channel analyses of type 3 IP₃ receptors, Foskettand co-workers (15), who also recorded biphasic effects of Ca²⁺ on channel gating for both Xenopus and type 3 IP₃ receptors, concluded that IP₃ activates the channel by decreasing its sensitivity to Ca²⁺ inhibition without modulating its sensitivity to Ca²⁺ stimulation. That interpretation is not consistent with our model, but it will be difficult to resolve the inconsistency until it becomes possible to record from single channels under conditions that allow rapid changes in IP₃ and Ca²⁺ concentration.

Rapid Inhibition of Type 3 IP₃ Receptors by Cytosolic Ca²⁺-- Despite continuing controversy (23, 24), there is substantial evidence that type 3 IP₃ receptors, like all other IP₃ receptor subtypes, are inhibited by substantial increases in cytosolic [Ca²⁺] (15, 25-27). Our present results establish that in the absence of IP₃, Ca²⁺ rapidly inhibits type 3 IP₃ receptors (t_1/2 = 39 ± 3 ms) and the inhibition cannot be overcome by even very high concentrations of IP₃; similar results were previously obtained for type 2 IP₃ receptors (31). In contrast with type 2 IP₃ receptors, however, the inhibition of type 3 receptors is rapidly (t_1/2 = 102 ± 20 ms) and completely reversible (Fig. 3).

In hepatocytes, binding of IP₃ protects the receptor from inhibition by increases in cytosolic [Ca²⁺] that would rapidly inhibit the unliganded receptor (31) (Fig. 4, A and C). This lead us to suggest that in addition to causing the stimulatory Ca²⁺-binding site to become accessible, IP₃ binding also causes occlusion of the inhibitory Ca²⁺-binding site (Fig. 7B). The results with type 3 IP₃ receptors are very different: after an initial stimulation of Ca²⁺ release (see above) an increase in cytosolic [Ca²⁺] causes rapid inhibition of IP₃-evoked Ca²⁺ release whether delivered early in the response to IP₃ or late in the response (Fig. 4, B, D, and E). In the rapid time courses, where high-Ca²⁺ was delivered either with the IP₃ (Fig. 1B) or 30 ms later (Fig. 1C), Ca²⁺ shortened the latency before channel opening, but it also caused rapid inhibition of the response. In CLM containing 200 nM free [Ca²⁺] the peak response to 300 µM IP₃ occurred after 293 ± 3 ms and then decayed slowly (t_1/2 = 1060 ± 180 ms, n = 6), but the response in 1 mM Ca²⁺ peaked much earlier (121 ± 13 ms) and then decayed much more rapidly (t_1/2 ~140 ms, Fig. 1, B and C). This pattern of response (stimulation then inhibition) and its time course are similar to the responses observed when Ca²⁺ increases were delivered at later stages of the response to IP₃ (Fig. 4, B, D, and E). We suggest that in each case the biphasic response to Ca²⁺ reflects the balance between rapid Ca²⁺ binding to stimulatory sites and a slower binding of Ca²⁺ to inhibitory sites. The time course of the inhibition is therefore slower than that recorded when the cells were pretreated with Ca²⁺ in the absence of IP₃ (Fig. 2).

Ca²⁺ slows IP₃ dissociation from hepatic IP₃ receptors (42), but it had no effect on the affinity of type 3 IP₃ receptors for IP₃. We therefore considered the possibility that even if IP₃ binding protected both receptor subtypes from Ca²⁺ inhibition, the effect might be more significant for hepatic IP₃ receptors, where Ca²⁺ would enhance the protection by slowing dissociation of IP₃ from the receptor. If IP₃ dissociation did limit the rate of Ca²⁺ binding to its inhibitory site, we would expect a receptor with a rapidly dissociating agonist bound to be more rapidly susceptible to Ca²⁺ inhibition. Our results (Fig. 6) showing that inhibition occurs with indistinguishable kinetics for receptors occupied by adenophostin, IP₃, or 3-deoxy-3-fluoro-IP₃ and that Ca²⁺ inhibition is more rapid than dissociation of even the most rapidly dissociating agonist establish that agonist dissociation is not required to allow inhibition of type 3 IP₃ receptors by Ca²⁺. We conclude that type 3 IP₃ receptors are rapidly and reversibly inhibited by increases in cytosolic [Ca²⁺] whether or not they have IP₃ bound. From single channel analyses of Xenopus (20) or type 3 (15) IP₃ receptors in the nuclear envelope, Mak et al. (15) proposed that the effect of IP₃ was to relieve inhibition by cytosolic Ca²⁺. While that model is consistent with the ability of IP₃ to block Ca²⁺ inhibition of type 2 receptors, it is difficult to reconcile with our results for type 3 receptors, where rates of inhibition by Ca²⁺ appear not to be affected by IP₃ binding (Fig. 7B).

Conclusions-- Biphasic regulation by cytosolic Ca²⁺ may be an ubiquitous feature of IP₃ receptors, but our superfusion analyses suggest that the means whereby Ca²⁺ exerts these effects are different for types 2 and 3 IP₃ receptors. Both receptors appear to open after sequential binding of first IP₃ and then Ca²⁺ to several (perhaps all four) subunits of the receptor (Fig. 7A). Ca²⁺ binding to another site (43) inhibits the receptor. Whereas that site appears to close after IP₃ binding to type 2 receptors, it remains accessible after binding of IP₃ to type 3 receptors. Ca²⁺ can therefore inhibit type 2 receptors only after IP₃ has dissociated, whereas type 3 receptors are more immediately susceptible to Ca²⁺ inhibition. We suggest that, contrary to an earlier suggestion (23), type 3 IP₃ receptors may be more susceptible than type 2 receptors to rapid inhibition by increased cytosolic Ca²⁺.

	FOOTNOTES

^* This work was supported by Wellcome Trust Grant 039662.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed. Tel./Fax: 44-1223-334058; E-mail: cwt1000@cam.ac.uk.

Published, JBC Papers in Press, March 1, 2002, DOI 10.1074/jbc.M200524200

	ABBREVIATIONS

The abbreviations used are: IP₃, inositol 1,4,5-trisphosphate; CLM, cytosol-like medium; EC₅₀, half-maximally effective concentration; Pipes, piperazine-N,N'-bis(2-ethanesulfonic acid).

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES