Inositol 1,4,5-tris-phosphate activation of inositol tris-phosphate receptor Ca²⁺ channel by ligand tuning of Ca²⁺ inhibition

Abstract

Inositol 1,4,5-tris-phosphate (IP₃) binding to its receptors (IP₃R) in the endoplasmic reticulum (ER) activates Ca²⁺ release from the ER lumen to the cytoplasm, generating complex cytoplasmic Ca²⁺ concentration signals including temporal oscillations and propagating waves. IP₃-mediated Ca²⁺ release is also controlled by cytoplasmic Ca²⁺ concentration with both positive and negative feedback. Single-channel properties of the IP₃R in its native ER membrane were investigated by patch clamp electrophysiology of isolated Xenopus oocyte nuclei to determine the dependencies of IP₃R on cytoplasmic Ca²⁺ and IP₃ concentrations under rigorously defined conditions. Instead of the expected narrow bell-shaped cytoplasmic free Ca²⁺ concentration ([Ca²⁺]_i) response centered at ≈300 nM–1 μM, the open probability remained elevated (≈0.8) in the presence of saturating levels (10 μM) of IP₃, even as [Ca²⁺]_i was raised to high concentrations, displaying two distinct types of functional Ca²⁺ binding sites: activating sites with half-maximal activating [Ca²⁺]_i (K_act) of 210 nM and Hill coefficient (H_act) ≈2; and inhibitory sites with half-maximal inhibitory [Ca²⁺]_i (K_inh) of 54 μM and Hill coefficient (H_inh) ≈4. Lowering IP₃ concentration was without effect on Ca²⁺ activation parameters or H_inh, but decreased K_inh with a functional half-maximal activating IP₃ concentration (K_IP3) of 50 nM and Hill coefficient (H_IP3) of 4 for IP₃. These results demonstrate that Ca²⁺ is a true receptor agonist, whereas the sole function of IP₃ is to relieve Ca²⁺ inhibition of IP₃R. Allosteric tuning of Ca²⁺ inhibition by IP₃ enables the individual IP₃R Ca²⁺ channel to respond in a graded fashion, which has implications for localized and global cytoplasmic Ca²⁺ concentration signaling and quantal Ca²⁺ release.

Modulation of cytoplasmic free Ca²⁺ concentration ([Ca²⁺]_i) is involved in the regulation of numerous cell physiological processes in which the second messenger inositol 1,4,5-tris-phosphate (IP₃) plays a central role in most cell types (1). Binding of extracellular ligands to G protein- or tyrosine kinase-linked receptors in the plasma membrane activates phospholipase C to generate IP₃, which binds to its receptors (IP₃R) in the endoplasmic reticulum (ER), activating them as Ca²⁺ channels to release stored Ca²⁺ from the ER lumen into the cytoplasm. The complex control of [Ca²⁺]_i is manifested temporally as repetitive spikes or oscillations (with frequencies often tuned to the level of stimulation) and spatially as propagating waves (1–3) and displays “adaptation” and “quantal release”, which are poorly understood (4, 5). IP₃-mediated Ca²⁺ release is regulated by [Ca²⁺]_i, with positive and negative feedback acting in a bell-shaped manner with peak Ca²⁺ release activity at ≈300 nM Ca²⁺ (6, 7); however, the mechanisms underlying this biphasic response—a fundamental component of [Ca²⁺]_i oscillations and wave propagation models (1, 3, 8)—are still poorly defined. The intracellular location of the IP₃R channel has necessitated the use of indirect measurements to infer its activity and has restricted studies of its single-channel properties (9–11). For this report, we systematically investigated the dependencies of single IP₃R channels on [Ca²⁺]_i and IP₃ concentration under rigorously defined conditions by using nuclear patch clamp of the IP₃R in its native ER membrane environment (12–15). Our results provide insights into the mechanism of channel activation by IP₃: whereas Ca²⁺ is a true receptor agonist, the sole function of IP₃ is to relieve Ca²⁺ inhibition of the channel. Ligand tuning of feedback inhibition by the permeant ion endows individual IP₃R Ca²⁺ channels with the ability to respond in a graded fashion to stimulus intensity.

METHODS

Patch clamp of the outer membrane of individual nuclei mechanically isolated from Xenopus laevis oocytes was performed as described (12–14). The oocyte expresses only a single IP₃R isoform (type 1) and lacks other (e.g., ryanodine receptor) Ca²⁺ release channels (16). The cytoplasmic aspect of the IP₃R channel faced into the patch pipette. All experimental solutions contained 140 mM KCl, 10 mM Hepes (pH adjusted to 7.1 with KOH), and 0 or 0.5 mM Na₂ATP as indicated. By using K⁺ as the current carrier and appropriate quantities of the high-affinity Ca²⁺ chelator 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid (BAPTA), the low-affinity Ca²⁺ chelator, 5,5′-dibromoBAPTA, or ATP alone to buffer Ca²⁺ in the experimental solutions, concentrations were tightly controlled in our experiments. Total Ca²⁺ content in the solutions was determined by induction-coupled plasma mass spectrometry (Mayo Medical Laboratory, Rochester, MN). Free Ca²⁺ concentrations were calculated by using the maxchelator software (C. Patton, Stanford University, Palo Alto, CA). Pipette solutions contained various concentrations of IP₃ as stated. All experiments were performed at room temperature with the pipette electrode at +20 mV relative to the reference-bath electrode. Each data point shown is the mean of results from at least four separate patch-clamp experiments performed under the same conditions. Error bars indicate the SEM. Single-channel currents were amplified by an Axopatch-1D amplifier (Axon Instruments, Foster City, CA), filtered at 1 kHz, digitized at 5 kHz, and recorded directly on hard disk by using Pulse+PulseFit 8.02 (HEKA Electronics, Lambrecht/Pfalz, Germany) on a PowerMac 8100 with an ITC-16 interface (Instrutech, Great Neck, NY). Channel dwell times and P_os were obtained by using tac 3.03 (Bruxton, Seattle, WA). Data were fitted and modeled by using igor pro 3.12 (WaveMetrics, Lake Oswego, OR).

RESULTS AND DISCUSSION

[Ca²⁺]_i Dependence of the Gating of the IP₃R Channel.

To examine the effects of [Ca²⁺]_i on IP₃R channel activity without possible Ca²⁺ effects on IP₃ binding, a functionally saturating IP₃ concentration of 10 μM (17) was applied to the cytoplasmic (pipette) side of the channel to fully stimulate it at various [Ca²⁺]_i. With [Ca²⁺]_i corresponding to resting levels in cells (10–100 nM), the P_o of the IP₃R was low (<0.2, Fig. 1 a and c), with short open intervals (τ_o < 3 ms) separated by long closed intervals (τ_c ≈ 100 ms; Fig. 1b). When [Ca²⁺]_i was increased from 100 nM to 1 μM, P_o increased dramatically to 0.8 as the result of a marked decrease of τ_c to ≈2 ms, whereas τ_o increased moderately to ≈10 ms. Similar [Ca²⁺]_i in the absence of IP₃ did not activate the channel (data not shown). Because saturating IP₃ concentrations were used, the Ca²⁺ requirement is not caused by Ca²⁺ enhancement of IP₃ affinity (18). Surprisingly, instead of the expected narrow bell-shaped [Ca²⁺]_i response centered at ≈300 nM–1 μM (9, 10, 15, 19), P_o remained very high (≈0.8) even as [Ca²⁺]_i was raised to quite high levels (1–20 μM); with long open bursts lasting >1 sec, during which the channel only closed very briefly. As [Ca²⁺]_i was increased beyond 20 μM, P_o decreased sharply, mainly the result of τ_c increasing to ≈100 ms, whereas τ_o decreased to ≈1 ms. In the presence of saturating concentrations of IP₃, there were no systematic effects on the P_o-vs.-[Ca²⁺]_i response of the species or concentration of Ca²⁺ chelator used (Fig. 1c), or the luminal Ca²⁺ (between 0.2 and 1.5 μM) or ATP (0 or 0.5 mM; Fig. 1d) concentrations.

Figure 1.

[Ca²⁺]_i dependence of the IP₃R in the presence of 10 μM IP₃. (a) Typical single-channel current traces of the IP₃R at various [Ca²⁺]_i. Arrows indicate closed-channel current level in all traces. (b) Dependence on [Ca²⁺]_i of mean closed-channel duration τ_c and open-channel duration τ_o of the IP₃R. (c) Dependence of IP₃R P_o on [Ca²⁺]_i. Experiments were performed with different Ca²⁺ chelators (□, BAPTA; ▵, 5,5′-dibromo BAPTA, ○, ATP) in the pipette solutions. Bath solution contained 1.5 μM Ca²⁺ and 0.5 mM ATP. Numerical labels indicate calculated concentrations (in μM) of free Ca²⁺ chelators present in the pipette solutions for the corresponding data points. The solid curve is the Hill equation fit for the biphasic [Ca²⁺]_i effect on IP₃R P_o. (d) IP₃R P_o vs. [Ca²⁺]_i in various bath solutions. Small open symbols: 0.8 μM luminal [Ca²⁺], 0.5 mM ATP; large open symbols: 0.2 μM luminal [Ca²⁺], 0 ATP; solid symbols: 0.2 μM luminal [Ca²⁺], 0.5 mM ATP. Convention for symbols and the solid curve are the same as in c.

The IP₃R P_o-vs.-[Ca²⁺]_i response in 10 μM IP₃ could be fitted to a biphasic Hill equation (Fig. 1c) so that:

1

This suggests that a tetrameric IP₃R channel can achieve a maximum open probability P_max of 0.81, with two distinct types of functional Ca²⁺ binding sites, activating sites with half-maximal activating [Ca²⁺]_i, K_act, of 210 ± 20 nM and Hill coefficient H_act of 1.9 ± 0.3, and inhibitory sites with half-maximal inhibitory [Ca²⁺]_i, K_inh, of 54 ± 3 μM and Hill coefficient H_inh of 3.9 ± 0.7. The large Hill coefficients, H_act and H_inh, indicate that activation and inhibition of the IP₃R by Ca²⁺ are both highly cooperative processes and suggest a requirement for Ca²⁺ binding to two of four monomers to open the channel in the presence of IP₃ and for Ca²⁺ binding to all four monomers to inhibit channel opening.

IP₃ Dependence of the [Ca²⁺]_i Sensitivity.

The striking insensitivity of P_o to increases of [Ca²⁺]_i until quite high levels results in a P_o-vs.-[Ca²⁺]_i curve with a broad plateau. Previous models of complex temporal and spatial [Ca²⁺]_i signaling have assumed a much higher affinity of the inhibitory Ca²⁺ binding sites than is suggested by our data. We examined whether this discrepancy arose from our use of saturating IP₃ concentrations. However, there was no significant change in the [Ca²⁺]_i dependence when the IP₃ concentration was varied between 0.1 and 180 μM (Fig. 2). Thus, there is no evidence that the IP₃R channel possesses a low-affinity IP₃ binding site with binding coefficient >100 nM, consistent with biochemical determinations (17, 18, 20, 21). Contrary to earlier observations (22–24), no inhibitory effects of the Ca²⁺ chelator BAPTA on Ca²⁺ activation of the IP₃R were detected, even at low (20–33 nM) IP₃ concentrations (Fig. 2). At IP₃ concentrations <100 nM, however, the IP₃R became more sensitive to Ca²⁺ inhibition: at 33 nM IP₃, K_inh decreased to 9.5 μM, although H_inh and Ca²⁺ activation were not affected. Further decreases in IP₃ to 10 and 20 nM caused pronounced reductions of both the maximum P_o and the range of [Ca²⁺]_i over which the IP₃R was active. The enhanced Ca²⁺ inhibition at low IP₃ concentrations cannot be explained by inhibition of IP₃ binding, because a ≈300-fold increase in [Ca²⁺]_i (from 1 to 300 μM) was required to counter the effect of only a 5-fold increase in IP₃ from 20 to 100 nM, indicating that the IP₃ relief of [Ca²⁺]_i inhibition is a highly cooperative process.

Figure 2.

IP₃-concentration dependence of the [Ca²⁺]_i sensitivity of the IP₃R. Different symbols denote data for various IP₃ concentrations as tabulated. Data points for 10 μM IP₃ were combined from experiments using various bath solutions with the same pipette [Ca²⁺]. Data points for other concentrations of IP₃ were obtained in bath solutions containing 220 nM Ca²⁺ and no ATP. The curves are Hill equation fits using Eq. 1, with K_inh varying with IP₃ concentration with values as listed in the graph, whereas P_max, K_act, H_act, and H_inh remained independent of IP₃ concentration, with the values tabulated in the graph. Numerical labels indicate calculated concentrations (in μM) of free Ca²⁺ chelator BAPTA present in pipette solutions used for the corresponding data points.

The measured P_o at all concentrations of IP₃ were fitted well by using the biphasic Hill equation (1), with K_inh being the only IP₃-concentration-sensitive parameter (Fig. 2). Even the observed reductions of maximum P_o and the range of [Ca²⁺]_i over which the channel is active at very low IP₃ concentrations can both be accounted for in this model by a continuous decrease in K_inh with decreasing IP₃, whereas the other parameters: P_max, K_act, H_act, and H_inh all remain unchanged by IP₃ concentration. These data lead to an unexpected conclusion: the effect of IP₃ binding is not to enable activation of the IP₃R by Ca²⁺, as expected for co-agonist ligands and as generally assumed (17, 18, 20, 21), but rather it is to ameliorate inhibition of the channel by Ca²⁺. The derived values of K_inh ranged from 160 nM at 10 nM IP₃ to 59 μM at 100 nM IP₃. This IP₃ dependence of K_inh was well fitted with a simple Hill equation so that

2

This equation, represented in Fig. 3a, implies that the IP₃R has a single class of functional IP₃ binding sites with a half-maximal activating IP₃ concentration, K_IP3, of 50 ± 4 nM, a Hill coefficient H_IP3 of 4 ± 0.5, and a maximum inhibitory Ca²⁺ binding coefficient K_∞ of 52 ± 4 μM at a saturating IP₃ concentration. The K_IP3 derived from this model is similar to both the dissociation constant K_D (10–100 nM) in IP₃ binding assays and IP₃ concentration required (≈10–100 nM) for stimulation of Ca²⁺ release (17, 20, 21, 25). The large Hill coefficient H_IP3 of 4 indicates that IP₃ activation of the IP₃R is highly cooperative (25–28) even in the absence of positive feedback from cytoplasmic Ca²⁺, requiring IP₃ binding to perhaps all four monomers of the channel to relieve the Ca²⁺ inhibition and gate the channel open. Because IP₃ binding to the IP₃R is not cooperative (17, 18, 20, 21), the similar Hill coefficients H_IP3 and H_inh (both ≈4) may suggest that IP₃ binding to each IP₃R monomer influences the inhibitory Ca²⁺ site in that same monomer.

Figure 3.

(a) Dependence on IP₃ concentration of K_inh as derived from Fig. 2. The curve is a simple Hill equation fit. (b) Bell-shaped normalized P_o-vs.-[Ca²⁺]_i curves calculated by using Eqs. 1 and 2 for various low concentrations of IP₃ (dotted curve, 10 nM IP₃ with maximum P_o of 0.09; dashed curve, 15 nM IP₃ with maximum P_o of 0.49; and solid curve, 20 nM IP₃ with maximum P_o of 0.71). (c) Theoretical P_o (calculated by using the model) vs. IP₃ concentration at various [Ca²⁺]_i, as labeled. Dashed curves, [Ca²⁺]_i ≤ 1 μM; thin solid curves, [Ca²⁺]_i ≥ 10 μM.

According to these results, cytoplasmic Ca²⁺ at low concentrations and IP₃ both activate the IP₃R channel (Fig. 4), but they affect the channel in fundamentally different ways. Ca²⁺ binding to the IP₃R at low [Ca²⁺]_i directly activates the channel—like a conventional agonist. In contrast, IP₃ binding to the channel activates it indirectly, solely by decreasing the affinity of the Ca²⁺ inhibitory site, with no direct stimulatory effect itself. Under conditions of low IP₃ concentration, Ca²⁺ preferentially binds to the inhibitory site because of its higher affinity (K_inh < K_act), causing the channel to be inactive. Under conditions of stimulation associated with higher IP₃ concentration, K_inh becomes >K_act, so Ca²⁺ binds preferentially to the Ca²⁺ activation site, activating the channel. A recent study of reconstituted IP₃R also observed an IP₃ dependence of Ca²⁺ inhibition (11). However, the results and interpretations are fundamentally different from ours. First, the IP₃ required to observe relief from Ca²⁺ inhibition was ≈180 μM, three orders of magnitude higher than that measured here. Furthermore, P_max was 0.03 vs. 0.81 in our studies. Second, inhibitory effects of Ca²⁺ on the channel were not observed at 180 μM IP₃, in contrast to our results. Third, an additional low-affinity (10 μM) IP₃ binding site was invoked to account for the IP₃ relief of Ca²⁺ inhibition, whereas our data indicate that only one functional IP₃ binding site is involved, with the affinity expected (≈50 nM) from biochemical studies. Importantly, our data indicate that IP₃ mediates its effects by modulating the affinity of Ca²⁺ inhibitory sites, which is conceptually distinct from agonist activity caused by IP₃ binding to a second low affinity site.

Figure 4.

Calculated P_o vs. [Ca²⁺]_i and IP₃ concentration according to our model. (a) Projection showing P_o-vs.-IP₃ concentration curves at various activating [Ca²⁺]_i. (b) Projection showing P_o-vs.-[Ca²⁺]_i curves at various IP₃ concentrations. (c) Projection showing P_o vs. IP₃ concentration curves at various inhibitory [Ca²⁺]_i.

Implications for IP₃-Mediated [Ca²⁺]_i Signaling.

Our results provide another interpretation of the role of IP₃, thus warranting reconsideration of previous observations of IP₃-mediated [Ca²⁺]_i signals in cells. Our model predicts a bell-shaped relationship between P_o and [Ca²⁺]_i, with a sharp peak at [Ca²⁺]_i <1 μM (9, 15, 19), although only under conditions of low (<20 nM) IP₃ (Fig. 4b). The low P_o and the bell shape, as well as the range of [Ca²⁺]_i over which the channel is active observed for reconstituted IP₃R channels at high (2 μM) IP₃ (9–11), are remarkably similar to the predicted channel behaviors according to our model, albeit at much lower (IP₃ ≈10 nM) concentration, suggesting an IP₃ insensitivity of the reconstituted channels in those studies. Although neither the IP₃ binding coefficient K_IP3 nor the Hill coefficient H_IP3 is [Ca²⁺]_i-dependent in our model, the interplay between the effects of [Ca²⁺]_i and IP₃ concentration on P_o can be manifested as an apparent increase in IP₃ sensitivity as [Ca²⁺]_i decreases at low [Ca²⁺]_i (Fig. 3c) and an apparent decrease in cooperativity of IP₃ activation as [Ca²⁺]_i increases (Fig. 3c). Thus, our results may qualitatively explain previous observations that submicromolar [Ca²⁺]_i lowered the functional sensitivity of the IP₃R for IP₃ (29, 30) and that the Hill coefficient for IP₃ activation of Ca²⁺ release decreased as [Ca²⁺]_i increased from the submicromolar to tens of μM range (27, 28, 31).

A major paradox has been [Ca²⁺]_i responses that are graded in proportion to the intensity of the stimulus, because the process of Ca²⁺-induced Ca²⁺ release should lead to an all-or-none signal (32, 33). One proposed mechanism suggests that increased IP₃ concentration recruits more individual elementary release units that each contribute a quantized amount of Ca²⁺ (32–35). Our model suggests another mechanism for achieving graded responses involving the IP₃R itself. Because IP₃ binding to the IP₃R relieves Ca²⁺ inhibition of the channel, higher IP₃ concentration shifts the peak of the P_o-vs.-[Ca²⁺]_i curve to higher [Ca²⁺]_i (Fig. 3b). A [Ca²⁺]_i that can inhibit channel activity at low IP₃ will be insufficient to inhibit it when IP₃ is increased (Fig. 4b). Higher [Ca²⁺]_i at the mouth of the channel pore can therefore be achieved before reaching levels that terminate release. This process therefore generates graded Ca²⁺ release from IP₃-sensitive stores through a mechanism intrinsic to the individual IP₃R. By enabling higher [Ca²⁺]_i to be achieved at the mouth of individual channels, higher IP₃ concentration associated with more intense stimuli would promote greater diffusive spread of the local [Ca²⁺]_i signal to other sites, thereby transforming highly localized signals at low levels of stimulation to more global coordinated Ca²⁺ release signals as the intensity of the stimulus is increased, without the need to invoke channels with different IP₃ sensitivities or clusters of higher channel densities (13, 32–34), although our results in no way exclude these or other mechanisms. By similar reasoning, our model can also account for the transformation of oscillating [Ca²⁺]_i at relatively low levels of agonist stimulation to sustained [Ca²⁺]_i increases during intense stimulation. Quantal Ca²⁺ release elicited by IP₃ (4, 5) has been described as a unique signaling mechanism that enables cells to retain complete responsiveness to changes in stimulus intensity (36). Our model suggests that this property can be intrinsic to the IP₃R. Because our experiments investigated the steady-state properties of the IP₃R, future kinetic studies will be required to determine the relevance of this model for quantal release in cells, as well as for other aspects of [Ca²⁺]_i signaling, including IP₃R desensitization (37) and inactivation (12, 13, 38).

ABBREVIATIONS

IP₃

inositol 1,4,5-tris-phosphate

IP₃R

IP₃ receptor

[Ca²⁺]_i

cytoplasmic free Ca²⁺ concentration

P_o

open probability

ER

endoplasmic reticulum

BAPTA

1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid