INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

The InsP₃ receptor gene family encodes three homologous InsP₃-binding proteins with three recognized domains (i.e., InsP₃ binding, regulatory/coupling, and channel; Mignery and Südhof, 1990, 1993; Südhof et al., 1991; Blondel et al., 1993). The ligand-binding and channel domains are highly conserved. The least conserved domain is the regulatory/coupling domain, which contains several potential regulatory sites (including a Ca²⁺-binding region; Mignery et al., 1992; Sienaert et al., 1996). The regulatory/coupling domain also physically links the InsP₃-binding and channel domains. The relatively high heterogeneity of the regulatory domain suggests that interactions between the three domains may be isoform specific. This implies that the function of the homotetrameric InsP₃R channels may be heterogeneous. Because heterotetrameric InsP₃R channels may also exist (Monkawa et al., 1996; Joseph et al., 1996), it is even more likely that InsP₃R channels may indeed be functionally heterogeneous.

To test this possibility, the function of the different InsP₃R channel isoforms must be defined. A great deal is already known about type 1 InsP₃R single-channel function (reviewed in Bezprozvanny and Ehrlich, 1995). However, very little is known about the single-channel properties of the other two InsP₃R channel isoforms. One obstacle has been the difficulty in defining a reliable way to isolate homogeneous receptor populations. Recently, our laboratory established that a essentially homogeneous population of type 2 InsP₃R channels could be isolated from ventricular cardiac myocytes (Perez et al., 1997). This provided the means to perform the first head-to-head functional evaluation of two different InsP₃R channel isoforms.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Materials

Inositol 1,4,5-trisphosphate was purchased from LC Laboratories (Woburn, MA). Heparin was purchased from Fluka Chemical Corp. (Ronkankoma, NY). 3-[(3-Cholamidopropyl)dimethylammonio]-1-propane sulfonate (CHAPS) was from Boerhinger Mannheim Biochemicals (Indianapolis, IN). L--Phosphatidylcholine, L--phosphatidylethanolamine, and L--phosphatidylserine were obtained from Avanti Polar Lipids (Pelham, AL).

Membrane preparation, sucrose gradient sedimentation, and reconstitution

Microsomal membranes from bovine cerebellum and acutely isolated ferret ventricular cardiac myocytes were prepared as described previously (Mignery et al., 1990; Perez et al., 1997). Microsomes were solubilized in 1% CHAPS and fractionated on 5-20% sucrose gradients. Gradient fractions containing the highest levels of receptor protein were identified by Western blotting and then reconstituted into proteoliposomes as described previously (Mignery et al., 1992; Perez et al., 1997).

Planar lipid/protein bilayer formation

Planar lipid bilayers were formed across a 220-µm-diameter aperture in the wall of a Delrin partition as described (Perez et al., 1997). Lipid bilayer-forming solution contained a 7:3 mixture of phosphatidylethanolamine and phosphatidylcholine dissolved in decane (50 mg/ml). Proteoliposomes were added to the solution on one side of the bilayer (defined as cis). The other side was defined as trans (virtual ground). Standard solutions (unless otherwise specified) contained 220 mM CsCH₃SO₃ cis (20 mM trans), 20 mM HEPES (pH 7.4), and 1 mM EGTA ([Ca²⁺]_FREE = 250 nM). The [Ca²⁺]_FREE was verified by using a Ca⁺² electrode. A custom current/voltage conversion amplifier was used to optimize single-channel recording. Acquisition software (pClamp; Axon Instruments, Foster City, CA), an IBM-compatible 486 computer, and a 12-bit A/D-D/A converter (Axon Instruments) were used. Single-channel data were digitized at 5-10 kHz and filtered at 1 kHz. Channel sidedness was determined by InsP₃ sensitivity. The orientation of the channels studied was such that the InsP₃-sensitive side (i.e., cytoplasmic side) was in the cis compartment.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Type 1 InsP₃R receptor was isolated from bovine cerebellum. The type 2 InsP₃R was obtained from isolated ventricular cardiac myocytes. Microsomes enriched in either type 1 or type 2 InsP₃Rs were CHAPS solubilized and then fractionated on 5-20% linear sucrose gradients. Tetrameric InsP₃R-containing fractions were identified by Western blotting and reconstituted into phosphatidylcholine:phosphatidylserine (3:1) proteoliposomes (Perez et al., 1997). Immunoblotting revealed that each proteoliposome population contained essentially one type of InsP₃R (data not shown). These proteoliposomes were then incorporated into planar lipid bilayers.

Incorporation of proteoliposomes into bilayers revealed InsP₃-sensitive, heparin-blocked Ca²⁺ channels (Fig. 1 A). These channels were insensitive to ryanodine. The InsP₃ and heparin acted only from one side of the channel (presumably the cytoplasmic side). Both type 1 and type 2 channels were characterized by frequent fast opening events, with few opening events lasting longer than a few milliseconds. The unitary Ca²⁺ (70 pS) and Cs⁺ (280 pS) conductances of the type 1 and type 2 channels were very similar (Fig. 1, B and C). This suggests that different types of InsP₃R channels have similar permeation properties.

View larger version (47K): [in this window] [in a new window]   FIGURE 1   Permeation properties of InsP3R channels. The single-channel properties of type 1 and type 2 InsP3R channels were defined in planar lipid bilayer studies. Experiments were performed in the presence of 1 µM InsP3 and 10 µM ryanodine. (A) Sample single-channel records from the type 1 and type 2 InsP3R channels. Open events (i.e., Cs+ currents) are shown as upward deflections from the zero current level. Control records were recorded before the addition of InsP3 (+InsP3) to the cis chamber (1 µM for type 1; 0.06 µM for type 2). After several minutes of single-channel recording, heparin (50 µg/ml) was added to the cis chamber (+ Heparin). Solutions contained 220/20 mM CsCH3SO3 (cis/trans), 1 mM EGTA, 250 nM free Ca2+, 20 mM HEPES (pH 7.4). (B) Current-voltage data from the type 1 () and type 2 () InsP3R channels conducting Ca2+. The two data sets were fit well by the same line (70 pS). Points represent means ± SD (n > 4). Trans solution contained 50 mM Ca(CH3SO3)2 and 20 mM HEPES (pH 7.4). Cis solution contained 1 mM EGTA, 250 nM free Ca2+, 80 mM HEPES-Tris (pH 7.4). (C) Current-voltage data from the type 1 () and type 2 () InsP3R channels conducting Cs+. Both data sets were fit well by the same line (280 pS). Points represent means ± SD (n > 4). Solutions contained 220/20 mM CsCH3SO3 (cis/trans), 1 mM EGTA, 250 nM free Ca2+, 20 mM HEPES (pH 7.4).

It is well established that cytoplasmic InsP₃ and Ca²⁺ regulate single type 1 InsP₃R channels (Bezprozvanny et al., 1991; Watras et al., 1991). The type 1 channel is activated by micromolar InsP₃ only over a narrow range of cytoplasmic Ca²⁺ concentrations. In this study the InsP₃ and Ca²⁺ sensitivities of the type 2 InsP₃R channel were defined for the first time. A monovalent cation (Cs⁺) was used as a charge carrier to eliminate Ca²⁺ flux through the pore. This ensures that local Ca²⁺ levels near the channel are precisely controlled. This is important because it has been reported that the InsP₃ affinity of the channel may be Ca²⁺ dependent (Yoneshima et al., 1997; Marshall and Taylor, 1994). The cytoplasmic InsP₃ sensitivities of both single type 1 and type 2 InsP₃R channels were defined with the free Ca²⁺ concentration buffered at 250 nM on both sides of the channel. Sample single-channel recordings at different InsP₃ concentrations are shown in Fig. 2, A and B. Average open probabilities (P_o) are plotted against InsP₃ concentration in Fig. 2 C. The type 2 InsP₃R channel had higher InsP₃ affinity (EC₅₀; 58 versus 194 nM for type 1), and the InsP₃ dependence of its activation suggests some degree of cooperativity (Hill coefficient; 1.85 versus 0.96 for type 1). The threefold difference in apparent InsP₃ affinity is similar to that observed for recombinant ligand-binding domains of the type 1 and type 2 receptors (Südhof et al., 1991; Newton et al., 1994). The dashed line (Fig. 2 C) represents previously published type 1 InsP₃R channel data (Watras et al., 1991). A notable difference between the two channels was the extent of InsP₃ activation. The open probability (P_o) of the type 2 InsP₃R channel was higher than that of the type 1 channel. The extent of this difference will depend on cytoplasmic Ca²⁺ concentration. If the type 1 InsP₃ dose dependency data were collected at the optimal cytoplasmic Ca²⁺ concentration (~750 nM), the efficacy difference between the type 1 and type 2 channels would be smaller. Nevertheless, any difference in the extent of InsP₃ activation (P_o level) implies that InsP₃ efficacy at mobilizing Ca²⁺ is InsP₃R isoform specific. Differential InsP₃ efficacy may help explain the high fidelity of InsP₃ signaling cascades in cells.

View larger version (35K): [in this window] [in a new window]   FIGURE 2   The InsP3 sensitivity of single type 1 and type 2 InsP3R channels defined in planar lipid bilayer studies. Solutions contained 220/20 mM CsCH3SO3 (cis/trans), 1 mM EGTA, 250 nM free Ca2+, 20 mM HEPES (pH 7.4). All experiments were performed in the presence of 10 µM ryanodine. The InsP3 concentration was varied in the cis chamber. Opening events are shown as upward deflections from the zero current level. Data points are plotted as means ± SD (n > 6). (A) Sample single-channel records from the type 1 InsP3R channels at three InsP3 concentrations. (B) Sample single-channel records from the type 2 InsP3R channels at three InsP3 concentrations. (C) InsP3 concentration response for single type 1 () and type 2 () InsP3R channels. The type 1 InsP3R channel data from Watras et al. (1991) are represented as a dashed line.

The cytoplasmic Ca²⁺ sensitivity of single type 1 and type 2 InsP₃R channels was defined at a fixed InsP₃ concentration (1 µM). The free Ca²⁺ concentration on the luminal side of the channels was held constant at 250 nM. Sample single-channel records at different cytoplasmic Ca²⁺ concentrations are shown in Fig. 3, A and B. The average P_o of the type 2 and type 1 channels at various cytoplasmic Ca²⁺ concentrations is plotted in Fig. 3 C. The Ca²⁺ dependence of the type 1 InsP₃R channel was sharply bell shaped. Maximum type 1 channel activity occurred near 1 µM Ca²⁺. These results are similar but not identical to those of Bezprozvanny et al. (1991) (Fig. 3 C, dashed line). The differences between the two type 1 data sets could be due to the different charge carriers used. Bezprozvanny et al. (1991) used a large luminal Ca²⁺ concentration (~50 mM) to provide the charge-carrying ion. The consequence is that the superphysiological Ca²⁺ flux through the channel may alter the occupancy of cytoplasmic Ca²⁺ and/or impact InsP₃ regulation of the channel. In our study, the use of a monovalent charge carrier eliminated this possibility. The two studies also used different methods to isolate single InsP₃R channels. Bezprozvanny et al. (1991) fused native cerebellar microsomes into the bilayer, whereas we fused InsP₃R-enriched proteoliposomes.

View larger version (35K): [in this window] [in a new window]   FIGURE 3   The Ca2+ sensitivity of single type 1 and type 2 InsP3R channels defined in planar lipid bilayer studies. Solutions contained 220/20 mM CsCH3SO3 (cis/trans), 1 mM EGTA, 250 nM free Ca2+, 1 µM InsP3, 10 µM ryanodine, and 20 mM HEPES (pH 7.4). The free Ca2+ concentration was varied in the cis chamber and directly verified by an on-line mini-Ca2+ electrode. The free Ca2+ concentration in the trans chamber was buffered at 250 nM. Opening events are shown as upward deflections from the zero current level. Data points are plotted as means ± SD (n > 6). (A) Sample single-channel records from the type 1 InsP3R channels at five Ca2+ concentrations. (B) Sample single-channel records from the type 2 InsP3R channels at five Ca2+ concentrations. (C) The Ca2+ concentration response for single type 1 () and type 2 () InsP3R channels. Type 1 InsP3R channel data from Bezprozvanny et al. (1991) are represented by the dashed line. The Ca2+ concentration response for single type 2 InsP3R channels activated by 0.1 µM InsP3 is also shown ().

The Ca²⁺ dependence of the type 2 InsP₃R channel is also illustrated in Fig. 3 C. The type 2 InsP₃R channels were active, even at relatively low Ca²⁺ concentrations (~25 nM), compared to the type 1 channels. At higher Ca²⁺ concentrations, type 2 InsP₃R channel activity (P_o  0.7) was maintained over a wide Ca²⁺ concentration range. Even at millimolar Ca²⁺ concentrations (data not shown), the P_o of type 2 channel remained high (~40%). Thus the Ca²⁺ dependence of the type 2 channel had essentially a sigmoidal shape, instead of the classical bell shape observed for the type 1 channel. This is interesting because it indicates that the type 2 channel lacks the Ca²⁺ inactivation mechanism that turns off the type 1 channel at micromolar Ca²⁺ concentrations. The type 1 and type 2 InsP₃Rs both encode a cytosolic Ca²⁺-binding site (residues 2124-2146 of the type 1 isoform) that is thought to be involved in the Ca²⁺ regulation of these receptors (Mignery et al., 1992; Sienaert et al., 1996). Therefore, the single-channel data (Fig. 3 C) show that this Ca²⁺-binding region does not contain all of the determinants of InsP₃R Ca²⁺ regulation.

Recently Kaftan et al. (1997) proposed a complex scheme explaining the interaction of Ca²⁺ and InsP₃ in the regulation of the type 1 InsP₃R channel. They argue that the inactivation phase of the Ca²⁺ dose dependency of the type 1 channel is masked at high InsP₃ concentrations. To support this claim they propose a model that employs a three-dimensional surface to describe the complex interaction between Ca²⁺ and InsP₃. In this context, it is possible that the difference between the Ca²⁺ dependencies of the type 1 and type 2 InsP₃R channels may be due to a shift of this surface along the InsP₃ scale to higher concentrations for the type 1 and lower for the type 2 channel. To test this possibility, we lowered (10-fold) the fixed InsP₃ concentration and reassessed the Ca²⁺ dependency of the type 2 InsP₃R channel (see Fig. 3, open squares). The type 2 receptor's response to Ca²⁺ did not become bell shaped, as predicted by the model of Kaftan et al. (1997). Instead, its Ca²⁺ dependency remained sigmoidal, and only its magnitude appeared to be dependent on InsP₃ concentration. This result suggests that the interaction of Ca²⁺ and InsP₃ in the regulation of type 1 and type 2 channels is different.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

We have defined the single-channel function of the type 2 InsP₃R in parallel with the type 1 isoform from cerebellum. The two receptor homologs form channels with very similar permeation properties and unitary conductances. Similar permeation properties are also consistent with the highly conserved nature of the InsP₃R channel domains. These results also suggest that the amplitude of local intracellular Ca²⁺ signals in cells is not likely due to opening of different InsP₃R channels with unique or isoform-specific conductance. It is thus more likely that the heterogeneity in intracellular Ca²⁺ signaling arises from differences in how these channels are regulated.

The regulation of the type 1 and type 2 channels by InsP₃ and that by cytosolic [Ca²⁺] were markedly different. InsP₃ activated the type 2 channels to a greater extent when the channels were exposed to 250 nM Ca²⁺ (see Fig. 2). The magnitude of this difference will vary, of course, with the cytoplasmic Ca²⁺ concentration. Nevertheless, the open probability (P_o) of the type 2 InsP₃R channel was always greater than that of the type 1 channel at saturating InsP₃ concentrations. This implies that an InsP₃ stimulus will be more effective at mobilizing Ca²⁺ if the type 2 InsP₃R channel isoform is the target. This would allow InsP₃ stimuli too small to activate type 1 InsP₃R channels to mobilize Ca²⁺ through type 2 channels. This could in effect segregate parallel InsP₃ signaling cascades that use these two different InsP₃R channels.

The type 1 and type 2 InsP₃R channels have clearly different InsP₃ binding affinities (Newton et al., 1994; Südhof et al., 1991). The type 2 InsP₃R has the highest InsP₃ binding affinity (K_D  27 nM) of the three characterized isoforms, with a relative order of type 2 > type 1  type 3 (Perez et al., 1997; Newton et al., 1994: Südhof et al., 1991). The affinity of the type 1 receptor is ~5-10-fold higher than that of the type 3 receptor. The affinity of the type 2 InsP₃R is approximately threefold higher than that of the type 1 receptor. Interestingly, our data show that the EC₅₀ of InsP₃ for the type 2 InsP₃R channels was about threefold higher than that of the type 1 channel (58 versus 194 nM). Thus the relative InsP₃ affinities measured by InsP₃ binding or InsP₃ activation of single type 1 and type 2 channels are in good agreement.

Different levels of cooperativity of InsP₃ responses in several cellular systems have been reported (for a review see Mignery and Südhof, 1993). For example, some groups suggest that more than one InsP₃ molecule must bind to the InsP₃R channel for it to open (e.g., Meyer et al., 1988; Iino and Endo, 1992). Other groups suggest that InsP₃ binding to a single site on the channel complex is sufficient to open the channel (e.g., Watras et al., 1991; Finch et al., 1991). The InsP₃ dose dependency of single InsP₃R channel activity in our study indicates that cooperativity of the InsP₃R is isoform specific (see Fig. 2). The activation of the type 2 InsP₃R appeared to involve more than one InsP₃ molecule, whereas activation of the type 1 InsP₃R did not. However, the relatively low activity level of the type 1 InsP₃R channel made it difficult to accurately predict its response at low InsP₃ levels. Thus our estimation of the InsP₃ cooperativity of the type 1 channel should be considered with care. In our view, more single-channel measurements at very low InsP₃ concentrations would be required to definitively establish the InsP₃ cooperativity of type 1 receptor.

Our results demonstrate that cytosolic Ca²⁺ differentially regulates the two receptor isoforms. Type 2 receptor channel activity was maintained at high Ca²⁺ concentrations (P_o  0.4, 1 mM Ca²⁺), whereas the type 1 channels were inactivated at Ca²⁺ concentrations above 1 µM. The shape of the Ca²⁺ dependence of the type 2 channel was sigmoidal instead of the classical bell shape observed for the type 1 channel. This is an interesting observation because it has implications for the local control of intracellular Ca²⁺ release. For example, the InsP₃R channels mediate relatively large Ca²⁺ release fluxes that must alter the free Ca²⁺ profile in the microenvironment of the channel. The bell-shaped Ca²⁺ depencence of the type 1 channel indicates that Ca²⁺ can feed back and turn the channel off, making the activity of the channel self-limiting. In contrast, the sigmodial Ca²⁺ dependency of the type 2 channel indicates that Ca²⁺ feedback will not have an impact on this channel's function. Termination of type 2 channel activity, therefore, is not mediated by a Ca²⁺-dependent inactivation mechanism. Instead, type 2 channel activity will cease upon depletion of the Ca²⁺ store or removal of the InsP₃ signal, and/or through some yet to be identified modulatory protein/factor. The consequence is that the type 1 and type 2 channels may mediate very different types of intracellular Ca²⁺ signals. For example, the type 1 channel could be ideal for mediating small transient Ca²⁺ signals, whereas the type 2 channel could be specialized to mediate large, sustained Ca²⁺ signals.

The action of cytosolic Ca²⁺ on InsP₃R activity has been explored in many different experimental systems. There are reports that the function of InsP₃R is regulated by cytosolic Ca²⁺ in a biphasic manner (e.g., Bezprozvanny et al., 1991; Iino, 1990; Finch et al., 1991). Other studies show that InsP₃R function is not governed by a Ca²⁺-dependent inhibition mechanism and thus does not respond to cytosolic Ca²⁺ in a biphasic manner (e.g., Commbettes and Champeil, 1994; Horne and Meyer, 1995). Here we show that single type 1 channel activity is a biphasic function of cytosolic Ca²⁺ concentration. We also show that type 2 channel activity is a monotonic function of cytosolic Ca²⁺ concentration. Thus the apparent disparity between the previously published results could potentially be explained by isoform specific functional InsP₃R attributes in the different experimental systems used. For example, Bezprozvanny et al. (1991) explored Ca²⁺ regulation of InsP₃R channels isolated from cerebellum, tissue rich in the type 1 InsP₃R protein. Horne and Meyer (1995) explored Ca²⁺ regulation of InsP₃R channels in basophilic leukemia (RBL) cells, a cell line rich in type 2-like InsP₃R proteins (Parys et al., 1995).

In this study, the type 2 InsP₃R channel was isolated from ventricular cardiac myocytes. The specific role of the type 2 InsP₃R channel in the myocyte is unclear. Although the myocyte is clearly specialized to optimize the ryanodine receptor (RyR)-mediated Ca²⁺ signaling events that govern cardiac contractility, it must undergo the routine cellular Ca²⁺ signaling that sustains and modulates numerous metabolic and developmental events in cells. Interestingly, the sigmoidal Ca²⁺ dependence of the type 2 InsP₃R channel makes it largely unaffected by the large, repeated, RyR-mediated Ca²⁺ signals. This would effectively limit the cross-talk between RyR-mediated and type 2 InsP₃R-mediated Ca²⁺ signals. Thus the sigmoidal Ca²⁺ dependence of the type 2 InsP₃R channel may allow InsP₃-dependent intracellular signaling cascades in the myocyte to operate independently of the cardiac contractile cycle. The identity, localization, and role of these InsP₃ dependent signaling cascades remain to be determined.

In summary, the resting Ca²⁺ concentration in most cells is ~100 nM and would rarely (if ever) exceed 1 mM, even in small, localized regions. At these free Ca²⁺ concentrations (100 nM to 1 mM), type 2 InsP₃R channels will be active in the presence of InsP₃. In contrast, the type 1 channel would not be active when cytosolic Ca²⁺ reaches the micromolar level. Furthermore, InsP₃ is more effective at mobilizing Ca²⁺ through the type 2 InsP₃R channel. These data suggest that the type 1 and type 2 channels mediate different types of intracellular Ca²⁺ signals. Type 1 InsP₃R-mediated signals would be self-limiting, as Ca²⁺ can feed back and turn off the channel. Type 2 InsP₃R signals would be larger (because of greater InsP₃ efficacy at mobilizing Ca²⁺) and would simply follow local InsP₃ levels, regardless of local Ca²⁺ concentration. Because most cells contain multiple types of InsP₃Rs (Newton et al., 1994; De Smedt et al., 1997), complex patterns of local Ca²⁺ signaling can arise. Thus it is very likely that isotype-specific functional heterogeneity contributes to the spatial and temporal complexity of intracellular Ca²⁺ signaling in mammalian cells.