Home | Help | Feedback | Subscriptions | Archive | Search | Table of Contents

This Article Abstract PDF (Full Text) PPT slides of all figures Alert me when this article is cited Citation Map Services Email this article Similar articles in this journal Similar articles in PubMed Alert me to new content in the JGP Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Laver, D. R. Articles by Lamb, G. D. Search for Related Content PubMed PubMed Citation Articles by Laver, D. R. Articles by Lamb, G. D. Social Bookmarking                      What's this?

¹ School of Biomedical Sciences, University of Newcastle and Hunter Medical Research Institute, Callaghan, NSW 2308, Australia
² Department of Zoology, La Trobe University, Melbourne, Victoria 3086, Australia

Address correspondence to Derek Laver, School of Biomedical Sciences, University of Newcastle and Hunter Medical Research Institute, Callaghan, NSW 2308, Australia. Fax: 61-2-4921-7406; email: derek.laver{at}newcastle.edu.au

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
In resting muscle, cytoplasmic Mg²⁺ is a potent inhibitor of Ca²⁺ release from the sarcoplasmic reticulum (SR). It is thought to inhibit calcium release channels (RyRs) by binding both to low affinity, low specificity sites (I-sites) and to high affinity Ca²⁺ sites (A-sites) thus preventing Ca²⁺ activation. We investigate the effects of luminal and cytoplasmic Ca²⁺ on Mg²⁺ inhibition at the A-sites of skeletal RyRs (RyR1) in lipid bilayers, in the presence of ATP or modified by ryanodine or DIDS. Mg²⁺ inhibits RyRs at the A-site in the absence of Ca²⁺, indicating that Mg²⁺ is an antagonist and does not simply prevent Ca²⁺ activation. Cytoplasmic Ca²⁺ and Cs⁺ decreased Mg²⁺ affinity by a competitive mechanism. We describe a novel mechanism for luminal Ca²⁺ regulation of Ca²⁺ release whereby increasing luminal [Ca²⁺] decreases the A-site affinity for cytoplasmic Mg²⁺ by a noncompetitive, allosteric mechanism that is independent of Ca²⁺ flow. Ryanodine increases the Ca²⁺ sensitivity of the A-sites by 10-fold, which is insufficient to explain the level of activation seen in ryanodine-modified RyRs at nM Ca²⁺, indicating that ryanodine activates independently of Ca²⁺. We describe a model for ion binding at the A-sites that predicts that modulation of Mg²⁺ inhibition by luminal Ca²⁺ is a significant regulator of Ca²⁺ release from the SR. We detected coupled gating of RyRs due to luminal Ca²⁺ permeating one channel and activating neighboring channels. This indicated that the RyRs existed in stable close-packed rafts within the bilayer. We found that luminal Ca²⁺ and cytoplasmic Mg²⁺ did not compete at the A-sites of single open RyRs but did compete during multiple channel openings in rafts. Also, luminal Ca²⁺ was a stronger activator of multiple openings than single openings. Thus it appears that RyRs are effectively "immune" to Ca²⁺ emanating from their own pore but sensitive to Ca²⁺ from neighboring channels.

Key Words: ryanodine receptor • magnesium • calcium • skeletal muscle • lipid bilayer

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Contraction in skeletal and cardiac muscle occurs when Ca²⁺ is released from the sarcoplasmic reticulum (SR) through ryanodine receptor (RyR) Ca²⁺ release channels. In striated muscle, excitation–contraction coupling describes the link between depolarization of the transverse tubule and calcium release from the SR. Upon depolarization of the transverse tubule membrane, dihydropyridine receptors (DHPRs, L-type calcium channels) are activated, which trigger the RyRs. In skeletal muscle, the RyRs are stimulated by a direct coupling with the DHPRs (Tanabe et al., 1990; Melzer et al., 1995), while in cardiac muscle, it is the influx of extracellular Ca²⁺ through the DHPR that initiates Ca²⁺ release.

Calcium release is modulated by a variety of substances, including small diffusible molecules such as ATP, Ca²⁺, Mg²⁺, and protons (pH) (Meissner, 1994). RyRs are activated at µM [Ca²⁺] and inhibited at mM [Ca²⁺] in the cytoplasm (Meissner, 1994). Mg²⁺ is believed to inhibit RyRs by two mechanisms (the dual-inhibition hypothesis). Mg²⁺ can inhibit RyRs by competing with Ca²⁺ for the activation sites (A-sites; Dunnett and Nayler, 1978; Meissner et al., 1986). In addition, Mg²⁺ can close RyRs by binding to low affinity, nonselective divalent cation inhibition sites that also mediate Ca²⁺ inhibition (I-sites; Meissner et al., 1986; Soler et al., 1992; Laver et al., 1997). However, with the former mechanism, it is not clear if Mg²⁺ acts as a competitive nonagonist (i.e., prevents Ca²⁺ from activating the channel) or as an antagonist in its own right. This distinction becomes important when one considers cytoplasmic ATP. Physiological levels of ATP (8 mM) strongly activate skeletal RyRs even in the absence of cytoplasmic Ca²⁺ (Meissner et al., 1986; Laver et al., 2001). Therefore if Mg²⁺ merely prevents Ca²⁺ activation, it would not inhibit RyRs that are activated by ATP (Laver et al., 1997). Here we measure the effects of A-site Mg²⁺ inhibition on RyRs that are activated by ATP in the absence of Ca²⁺. The first major finding of this work is that Mg²⁺ is an RyR antagonist that inhibits in the absence of cytoplasmic and luminal Ca²⁺.

Ryanodine and the disulfonic stilbene derivatives are widely used pharmacological probes for elucidating the mechanisms of muscle contraction. Diisothiocyanostilbene-2',2'-di-sulfonic acid (DIDS) has two isothiocyanate groups that can form covalent bonds with several amino acid residues. Modification of RyRs by ryanodine and DIDS is used here to help dissect the two mechanisms of Mg²⁺ inhibition. In the presence of ryanodine and DIDS, the Ca²⁺ sensitivities of the A- and I-sites are more disparate, producing wider separation of the two Mg²⁺ inhibition mechanisms and a much clearer manifestation of Mg²⁺ inhibition at the A-sites (Laver et al., 1997). DIDS has been shown to activate RyRs by reversible and nonreversible mechanisms (Kawasaki and Kasai, 1989; Zahradnikova and Zahradnik, 1993; Sitsapesan, 1999). Two independent mechanisms for nonreversible activation have been distinguished by their different kinetics (referred to as slow and fast) and their different specific effects (O'Neill et al., 2003). The fast mechanism increased the degree of Ca²⁺ activation with no significant change in sensitivity (A-sites) and reduced RyR sensitivity to Ca²⁺/Mg²⁺ inhibition (I-sites) by 10-fold. The slow mechanism activated RyRs in the absence of Ca²⁺ and ATP.

Ryanodine has a profound effect on the regulation of RyRs by intracellular constituents. Ryanodine-modified RyRs are insensitive to cytoplasmic Ca²⁺ and adenine nucleotides (Rousseau et al., 1987; Laver et al., 1995), though they are still inhibited by Mg²⁺ (Masumiya et al., 2001) and low pH (Ma and Zhao, 1994; Laver et al., 2000), albeit with reduced sensitivity. In spite of the pharmacological importance of ryanodine, its effects on RyR function are not well understood. Earlier studies report that ryanodine-modified RyRs are active (i.e., open probability 1) in the virtual absence of Ca²⁺ (Rousseau et al., 1987; Laver et al., 1995). However, two recent studies on recombinant cardiac RyRs (RyR2) report that ryanodine shifts their Ca²⁺ activation response to lower concentrations by four orders of magnitude (Du et al., 2001; Masumiya et al., 2001). We investigate the effects of cytoplasmic Ca²⁺ and Cs⁺ on A-site Mg²⁺ inhibition in ryanodine-modified RyRs to probe the competitive ion binding kinetics at the A-site. Our second major finding is that ryanodine increases the Ca²⁺ sensitivity of the A-sites by 10-fold but this is insufficient to explain the level of activation of ryanodine-modified RyRs at nM Ca²⁺.

The Ca²⁺ load of the SR is an important stimulator of Ca²⁺ release (Fabiato and Fabiato, 1977). It has been shown to regulate Ca²⁺ release in response to Ca²⁺ (Ford and Podolsky, 1972; Endo, 1985; Meissner et al., 1986), caffeine (Lamb et al., 2001), and ATP (Morii and Tonomura, 1983; Donoso et al., 1995). Raised luminal Ca²⁺ increases channel activity in both purified (Tripathy and Meissner, 1996) and native RyRs (Sitsapesan and Williams, 1995; Beard et al., 2000). Activation of RyRs by luminal Ca²⁺ has been attributed to two quite different mechanisms, and there is as yet no consensus on just how the Ca²⁺ load in the SR alters RyR activation (Sitsapesan and Williams, 1997). The "true luminal regulation" hypothesis attributes luminal Ca²⁺ activation to Ca²⁺ regulatory sites on the luminal side of the RyR (Sitsapesan and Williams, 1995). This is supported by the fact that luminal Ca²⁺ activation is susceptible to tryptic digestion from the luminal side of the membrane (Ching et al., 2000). The "feedthrough" hypothesis proposes that luminal Ca²⁺ permeates the pore and binds to the cytoplasmic activation sites (Tripathy and Meissner, 1996; Xu and Meissner, 1998). The latter is supported by the close correlation between open probability and Ca²⁺ flux (lumen to cytoplasm), which is seen under a wide range of experimental manipulations in both cardiac and skeletal RyRs. Here we investigate the effects of luminal Ca²⁺ on Mg²⁺ binding to the A-site in order to detect competition between luminal and cytoplasmic ions resulting from Ca²⁺ feedthrough. Our third and most important finding is a novel mechanism for regulation of Ca²⁺ release by luminal Ca²⁺, whereby increasing luminal [Ca²⁺] decreases the affinity of the A-sites for Mg²⁺. We show that luminal Ca²⁺ activates RyRs by both the true luminal regulation and Ca²⁺ feedthrough mechanisms in a way that depends on whether they are in close proximity to other open RyRs. Finally, we present an ion binding model of Mg²⁺ inhibition at the A-sites that predicts that modulation of Mg²⁺ inhibition by luminal Ca²⁺ is a significant regulator of Ca²⁺ release from the SR.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Lipid Bilayers, Chemicals, and Solutions
SR vesicles were prepared from the back and leg muscles of New Zealand rabbits killed by captive bolt before muscle removal. The procedure was performed by the holder of a current license granted under ACT State legislation. Native SR vesicles were isolated using techniques based on those of Chu et al. (1988), as previously described by Laver et al. (1995).

Unless otherwise stated, lipid bilayers were formed from phosphatidylethanolamine (PE) and phosphatidylcholine (PC) (8:2) (Avanti Polar Lipids) dissolved in 20 µl n-decane (50 mg/ml). The bilayers were formed across an aperture of 100–200 µm diameter in a Delrin cup. The bilayer separated two solutions: cis and trans (1 ml). Vesicles were added to the cis solution and vesicle incorporation with the bilayer occurred as described by Miller and Racker (1976). Due to the orientation of RyRs in the SR vesicles, RyRs added to the cis chamber incorporated into the bilayer with the cytoplasmic face of the channel orientated to the cis solution.

The cesium salts were obtained from Sigma-Aldrich and CaCl₂ from BDH Chemicals. Unless otherwise stated, solutions were pH buffered with 10 mM N-tris[hydroxymethyl]methyl-2-aminoethanesulfonic acid (TES, obtained from ICN Biomedicals) and solutions were titrated to pH 7.4 using CsOH (optical grade from ICN Biomedicals). During SR vesicle incorporation the cis (cytoplasmic) solution contained 230 mM CsCH₃O₃S and 20 mM CsCl (250 mM Cs⁺ solution) with 1.0 or 0.1 mM CaCl₂, while the trans (luminal) solution contained 30 mM CsCH₃O₃S and 20 mM CsCl (50 mM Cs⁺ solution) and 0–1.0 mM CaCl₂. The osmotic gradient across the membrane and the Ca²⁺ in the cis solution also aided vesicle fusion with the bilayer. In experiments using symmetric [Cs⁺], vesicle fusion was performed using cis and trans baths containing 250 mM Cs⁺ solution with 500 mM mannitol in the cis bath to produce the necessary osmotic gradient. Cs⁺ was used as the current carrying ion rather than K⁺ in order to extinguish interfering signals from the SR K⁺ channels. During measurements, the composition of the cis solution was altered either by addition of aliquots of stock solutions or by continuous local perfusion of the bilayer via a tube placed in close proximity (O'Neill et al., 2003).

The required free [Ca²⁺] was attained by buffering with 4.5 mM BAPTA (1,2-bis[o-aminophenoxy]ethane-N,N,N',N'-tetraacetic acid, obtained as a tetra potassium salt from Molecular Probes) and titrated with CaCl₂. Free [Ca²⁺] in excess of 0.1 µM was measured using a Ca²⁺ electrode (Fluka). At lower concentrations, free [Ca²⁺] was estimated using published association constants (Marks and Maxfield, 1991) and the program Bound and Determined (Brooks and Storey, 1992). Solutions, which mimicked zero Ca²⁺, were made with 4.5 mM BAPTA and no added Ca²⁺. In these solutions, the total [Ca²⁺] arising from impurities was measured to be 15 µM so that with addition of 4.5 mM BAPTA the free [Ca²⁺] was calculated to be 1 nM.

In solutions containing ATP, which buffers Mg²⁺, the required free [Mg²⁺] was determined using the fluorescent magnesium indicator, Mag-fura-2 (tetra potassium salt from Molecular Probes). The ratio of fluorescence intensities at 340 and 380 nm was calibrated in the experimental solutions (50 and 250 mM Cs⁺ solutions, see above) also containing 5 µM Mag-fura-2, 4.5 mM BAPTA, (free [Ca²⁺] 1 nM–1 µM) and MgCl₂ from aliquots of a calibrated stock. This allowed determination of the ATP purity and effective Mg²⁺ binding constants under experimental conditions, which in turn allowed calculation of free [Mg²⁺] in solutions containing >1 µM Ca²⁺, where Mag-fura-2 is not a suitable indicator of [Mg²⁺].

ATP was obtained in the form of sodium salts from Sigma-Aldrich. Unless otherwise stated, DIDS (sodium salt) was prepared as a stock solution in DMSO (dimethyl sulfoxide) at concentrations between 5 and 50 mM. Care was taken to ensure that DMSO in the bath solutions did not exceed 1% since DMSO concentrations >2% inhibited RyR activity (O'Neill et al., 2003).

Acquisition and Analysis of Single Channel Recordings
Bilayer potential was controlled and currents recorded using an Axopatch 200B amplifier (Axon Instruments). The cis chamber was electrically grounded to prevent electrical interference from the perfusion tubes, and the potential of the trans chamber was varied. However, all electrical potentials are expressed here using standard physiological convention (i.e., cytoplasmic side relative to the luminal side at virtual ground). Measurements were performed at 23 ± 2°C.

During the experiments, the channel current was recorded after low pass filtering at 5 kHz and sampling at 50 kHz. The data was stored on computer disk using a data interface (Data Translation DT301) under the control of in-house software written in Visual Basic. For measurements of P_o, the current signal was digitally filtered at 1 kHz with a Gaussian filter and sampled at 5 kHz. Unitary current and time-averaged currents were measured using Channel2 software (P.W. Gage and M. Smith, Australian National University, Canberra, Australia). To calculate P_o from single channel records, a threshold discriminator was set at 50% of channel amplitude to detect channel opening and closing events. For experiments in which bilayers contained several RyRs, the time-averaged current was divided by the unitary current and the number of channels. The number of channels in each experiment could be determined during periods of strong activation (in the absence of Mg²⁺). Both methods of calculating P_o gave similar results.

For measurements of channel open and closed dwell times, the current signal was filtered at 2 kHz and sampled at 10 kHz. Open and closed durations were extracted from single channel recordings using the 50% threshold (see above). Channel gating in multichannel recordings were analyzed using the Hidden Markov Model (HMM; Chung et al., 1990). The algorithm calculated the idealized, multilevel, current time course (i.e., background noise subtracted) and the transition probability matrix from the raw signal using maximum likelihood criteria. The mean channel opening and closing rates in the presence of n_o open channels, k_o⁺ and k_o^–, respectively, were calculated from the transition probability matrix, P, by Eq. 1:

(1)

where n_T is total number of channels in the bilayer and t is the sample interval.

Dissecting Mg²⁺ Inhibition at A-sites and I-sites
RyRs are inhibited when Mg²⁺ is bound to either A- or I-sites so that the open probability of the channel is the product of the Mg²⁺ occupancies of each site (Laver et al., 1997). Consequently, the half-inhibiting [Mg²⁺], K_i(Mg²⁺) depends on the Mg²⁺ affinities of both A- and I-sites (K_A and K_I, respectively). Provided the K_A << K_I then K_A K_i(Mg²⁺), with a relative error of (K_A/K_I)² (the second power in this equation stems from the Hill coefficient of 2 for Mg²⁺ inhibition at the I-sites). Since I-sites are nonselective between divalent ions (Soler et al., 1992; Laver et al., 1997; Meissner et al., 1997), then one can estimate K_I from the half-inhibitory [Ca²⁺] for RyR inhibition, K_i(Ca²⁺)_c. The subscripts c here, and l later, refer to cytoplasmic and luminal solutions, respectively. For example, in the case of ATP-activated RyRs in this study, K_i(Ca²⁺)_c is 2.5 mM (Table I) so that K_A K_i(Mg²⁺) with <15% error when K_i(Mg²⁺) < 1 mM. Since the experimental conditions were always chosen to keep the error in K_A < 15%, we denote the Mg²⁺ affinity of the A-sites by either K_A or K_i(Mg²⁺). We found that DIDS and ryanodine are useful pharmacological tools for studying the A-sites because they markedly increased K_i(Ca²⁺)_c and so increased the range of K_A that could be determined from K_i(Mg²⁺).

(2)

where

(3)

The parameter n_j is equal to the number of ions of species, j, that can bind at that site. The ions considered to compete with Mg²⁺ in this study are Cs⁺ and Ca²⁺ in the cytoplasm and Ca²⁺ in the lumen. When these ions are explicitly included Eq. 2 becomes

(4)

The term for Cs⁺ is raised to the second power because it is likely that two monovalent ions can bind to a divalent cation site (Meissner et al., 1997). Examination of Eqs. 2–4 reveals that the relative effect of a competing ion on Mg²⁺ inhibition is significant when its concentration exceeds the value of its affinity and the ratio of its concentration to affinity exceeds that of the other ions present.

In the noncompetitive model, each ion regulates channel activity at independent sites. A specific example of this, used in this work, is where the binding of luminal Ca²⁺ prevents the binding of cytoplasmic Mg²⁺ or Ca²⁺ (denoted by X²⁺) by an allosteric effect. The apparent affinities for cytoplasmic Ca²⁺ and Mg²⁺, K_app(X²⁺)_c, are related to the affinity K_m(X²⁺)_c in the absence of luminal Ca²⁺ by Eq. 5:

(5)

The Ca²⁺_c dependence of Mg²⁺ inhibition is analyzed in this study in terms of apparent affinities for Ca²⁺_c and Mg²⁺, K_app(Ca²⁺)_c, and K_app(Mg²⁺), respectively (see Fig. 4). For this analysis Eqs. 2–4 are recast into the following form:

(6)

It can be seen from Eq. 6 that the value of K_i(Mg²⁺) at low [Ca²⁺]_c approximates the value of K_m(Mg²⁺) and the value of K_m(Ca²⁺)_c determines the [Ca²⁺] threshold where K_i(Mg²⁺) becomes Ca²⁺ dependent.

Statistics and Curve Fitting
Unless otherwise stated the data are presented as mean ± SEM obtained from N bilayers and n RyRs. Theoretical curves were fitted to the data using the criteria of least squares. The open probability (P_o) of RyRs, activated or inhibited by a compound with concentration, C, was fitted by Hill equations, Eqs. 7 and 8, respectively:

(7)

(8)

where P_i and P_max are RyR open probabilities in the absence of the compound and at maximal activation, respectively. K_a and K_i are half-activation and inhibition concentrations, and n_a and n_i are the associated Hill coefficients. Ion binding models of the A-site were fitted to the logarithm of the data by minimizing the residuals weighted by the experimental error on each datum. Statistical uncertainty in the elements of the transition probability matrix produced by HMM analysis ranged from 5 to 25% relative error. This was estimated by the SEM of transition probability matrices derived from four subsections in four representative experiments.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
RyR Ca²⁺ Dependencies
RyRs in the absence of ATP and the modifying agents DIDS and ryanodine (i.e., native RyRs) are inactive in the absence of cytoplasmic [Ca²⁺] ([Ca²⁺]_c). They are activated by µM Ca²⁺ and inhibited by mM Ca²⁺, thus producing the characteristic bell-shaped Ca²⁺ dependence in open probability (P_o) (Fig. 1 A, closed circles).

View larger version (19K): [in this window] [in a new window]   FIGURE 1. The dependence of RyR open probability, Po, on cytoplasmic [Ca2+]. (A) RyRs activated by cytoplasmic Ca2+ alone (•, N = 30), modified by 10 µM ryanodine () or 4-min exposure to 100 µM DIDS (, N = 12, n = 64). Luminal [Ca2+] = 1 mM. (B) In the presence of 2 mM ATP with luminal [Ca2+] = 1 mM (, N = 9, n = 49) or 0.01 mM (, N = 15, n = 89). The membrane potential was +40 mV. The data points represent means of 2–20 measurements. The solid curves are Hill fits to the data and the parameter values are shown in Table I. The dashed curve is the Hill fit to the circles in A shown again for comparison.

View larger version (22K): [in this window] [in a new window]   FIGURE 2. Recordings from one experiment on a single skeletal RyR (RyR1) showing the effects of luminal [Ca2+] and cytoplasmic [Mg2+] on channel gating. Increasing luminal [Ca2+] markedly increased the channel open probability and decreased the channel sensitivity to Mg2+ inhibition. The cytoplasmic solution contained (in mM) 250 CsCH3O3S, 10 TES (pH 7.4), 4.5 BAPTA (1 nM free Ca2+), 2 ATP plus the various [Mg2+]. The free Mg2+ is indicated at the left and right of each row. The luminal solution contained (in mM) 30 CsCH3O3S, 20 CsCl, 10 TES, and the indicated [Ca2+]. Under these conditions, Mg2+ is thought to inhibit primarily by binding at the high affinity Ca2+ activation site. Membrane potential was held at +40 mV. The current baselines are shown by dashed lines.

DIDS produces both reversible and nonreversible activation of RyRs (O'Neill et al., 2003). Here, the effects of permanent RyR modification were measured by applying DIDS (100–500 µM) to the cytoplasmic bath for 4 min. DIDS was then removed by local perfusion before measurements commenced. DIDS modification activated RyR in the absence of cytoplasmic Ca²⁺ (Fig. 1, open circles) and shifted Ca²⁺ activation and inhibition to higher concentrations by 10-fold and by 3-fold, respectively (Table I).

Regulation of Modified RyRs by Mg²⁺
We investigated the possibility that Mg²⁺ inhibits RyRs by occluding Ca²⁺ from its activation sites on the protein. We did this by measuring the effect of Mg²⁺ on RyRs activated by ATP, DIDS, or ryanodine in the absence of cytoplasmic Ca²⁺. In all three cases, Mg²⁺ could totally and reversibly inhibit RyRs (e.g., see Figs. 2 and 3 for the case of ATP-activated RyRs). An increase in [Ca²⁺]_c produced a corresponding increase in the [Mg²⁺] required to inhibit the channels (Figs. 3 and 4). Therefore, Ca²⁺ and Mg²⁺ compete for a common site (presumably the A-site) where Mg²⁺ is an antagonist and not merely an inhibitor of Ca²⁺ activation.

View larger version (18K): [in this window] [in a new window]   FIGURE 3. Mg2+ inhibition of RyR Po in the presence of 2 mM ATP and [Ca2+]c = 10 nM (, N = 7, n = 20), 1 µM (•, N = 6, n = 13), and 10 µM (, N = 7, n = 35). Luminal [Ca2+] = 1 mM. The data points represent means of two to seven measurements. The solid curves are Hill fits to the data. Half-inhibiting Mg2+ concentrations, Ki(Mg2+), are summarized in Fig. 4 and Table III.

View larger version (12K): [in this window] [in a new window]   FIGURE 4. The [Mg2+] required to half-inhibit RyRs, Ki(Mg2+), at the indicated cytoplasmic [Ca2+]. Mg2+ inhibition was measured in RyRs that were modified by 10 µM ryanodine () or 100 µM DIDS () or activated by 2 mM ATP (). Luminal [Ca2+] = 1 mM and the membrane potential was +40 mV. The data points represent means of 4–11 measurements. The solid curves are generated by Eq. 6. The parameters of the fits are given in Table I.

Effect of Monovalent Ions on Mg²⁺ Inhibition
In native RyRs, monovalent cations are known to compete with Ca²⁺ for the activation site to prevent channel activation (Meissner et al., 1997). Therefore we compared the Mg²⁺ inhibition of RyRs at high (250 mM) and low (50 mM) Cs⁺ concentrations in the absence of cytoplasmic Ca²⁺. Both in the presence and absence of luminal Ca²⁺, raising [Cs⁺] by fivefold increased K_i(Mg²⁺) by 3- and 10-fold for ATP and ryanodine-modified RyRs, respectively (Fig. 5 and Table II, compare entries 1 and 5, 3 and 7; Table III, compare entries 1 and 3, 2 and 4). This is consistent with competition between Mg²⁺ and Cs⁺ for common sites. The relatively strong Cs⁺ dependence of Mg²⁺ inhibition in the case of ryanodine-modified channels suggests that at least two Cs⁺ can bind at a Mg²⁺ inhibition site.

View larger version (22K): [in this window] [in a new window]   FIGURE 5. The dependence of Mg2+ inhibition of ryanodine-modified RyRs on cytoplasmic [Cs+] and luminal [Ca2+]. Recordings were made in the absence of cytoplasmic [Ca2+] (1 nM). The solid curves are Hill fits to the data. Half-inhibiting Mg2+ concentrations, Ki(Mg2+), and numbers of experiments are summarized in Table II.

View larger version (17K): [in this window] [in a new window]   FIGURE 6. The effect of luminal [Ca2+] on inhibition of RyRs by cytoplasmic Mg2+. Mg2+ inhibition of RyRs was measured in the presence of 1 nM cytoplasmic [Ca2+], 2mM ATP, and various luminal [Ca2+]. (A) The mean dose–responses to Mg2+ were measured with luminal [Ca2+] of 0.1 mM (•, N = 7, n = 21), 1 mM (, N = 11, n = 26), and 3 mM (, N = 13, n = 43). The data points represent means of 3–13 measurements. The solid curves are Hill fits to the data. (B) x, half-inhibiting Mg2+ concentrations, Ki(Mg2+), obtained from seven individual RyRs at two or more [Ca2+]l; , the mean from all experiments. N and n values listed in Table III.

View larger version (35K): [in this window] [in a new window]   FIGURE 7. Recordings from a single, ryanodine-modified RyR showing the effect of luminal [Ca2+] on channel sensitivity to Mg2+ inhibition. In zero luminal Ca2+ ([Ca2+] < 10 µM), 1 mM cytoplasmic Mg2+ reduced channel activation by 80%. Increasing luminal [Ca2+] to 1 mM markedly reduced the effect of Mg2+, such that 5 mM Mg2+ was needed to produce 50% inhibition. Membrane potential was held at +40 mV. The current baselines are shown by dashed lines.

One might expect that the voltage dependence of Ca²⁺ flow through the RyR could produce a voltage dependence in their Mg²⁺ inhibition, because at negative potentials, luminal Ca²⁺ flowing through the channel should compete with cytoplasmic Mg²⁺ for the A-sites. In marked contrast with this proposition, we found that K_i(Mg²⁺) for single RyRs in a bilayer was insensitive to voltage at all luminal [Ca²⁺] tested. In the case of 1 mM luminal Ca²⁺ (cytoplasmic 2 mM ATP and 1 nM Ca²⁺) the difference in K_i(Mg²⁺) between –40 mV (100 ± 40 µM; N = 5) and +40 mV (130 ± 40 µM; N = 12) was not significant and considerably less than changes in K_i(Mg²⁺) caused by varying luminal [Ca²⁺] (this is not the case for coupled groups of RyRs, see below). This is particularly important because it indicates that the effects of luminal Ca²⁺ on Mg²⁺ inhibition of single RyRs are not due to Ca²⁺ feedthrough.

Gating Kinetics of Single RyRs
More detailed information about the mechanisms of channel regulation by cytoplasmic and luminal ligands can be obtained from the rates of channel gating. For example, it has been argued that luminal Ca²⁺ cannot affect RyR closed dwell times by acting at cytoplasmic sites since they are inaccessible to luminal Ca²⁺ when the channel is closed (Xu and Meissner, 1998). We begin with an analysis of mean open and closed dwell times (_o and _c, respectively) of single channels and proceed to analyze the analogous closing and opening rates of RyR from multichannel recordings. The general pattern of RyR activation by Ca²⁺ (both luminal and cytosolic), Mg²⁺, and ATP seen here closely follows that previously reported for the adenine nucleotides (see Fig. 7 in Laver et al., 2001). When P_o is <0.2, channel activation and inhibition was associated primarily with changes in _c, while at higher P_o, changes in channel activity were associated primarily with changes in _o (unpublished data). The dependencies of mean open and closed dwell times on luminal [Ca²⁺] and cytoplasmic [Mg²⁺] are summarized in Table VI. We find that luminal Ca²⁺ can modulate channel activity and Mg²⁺ inhibition via changes in both _o and _c, which implies a mode of action involving luminal sites on the RyR protein.

Gating Kinetics of RyRs in Coupled Groups
On average, 20% of fusion events incorporated groups of four to eight RyRs into the bilayer. Under the right experimental conditions (e.g., –40 mV and the presence of ATP), the opening of one RyR in a group tended to promote the opening of other RyRs. Fig. 8 shows the activity of four, ATP-activated RyRs at positive and negative bilayer potentials. The current trace shows transitions between the current baseline (labeled C) and four equally spaced levels (O1–O4) corresponding to one to four open channels. At positive potentials, the weighting of each current level followed a binomial distribution expected from the gating in independent channels in the bilayer (unpublished data). At negative potentials, downward current steps frequently "bypassed" some of the current levels, indicating that several channels were opening in near synchrony. The weighting of current levels in these records markedly deviated from a binomial distribution. HMM analysis of these records (see MATERIALS AND METHODS) shows that the mean RyR opening rate associated with transitions between the current baseline and level O1, k₀⁺, was significantly slower than opening rates associated with transitions between levels O1–O4, k₁⁺–k₄⁺ (Fig. 9 A, ; P = 0.003, paired t test), while there was no significant difference in the rates associated with transitions between levels O1–O4. Channel closing rates, k^-, did not depend on the number of open RyRs (Fig. 9 B). Thus, the coupling between RyRs was primarily due to the opening rate, which for each channel depended on whether or not one other channel was open in the bilayer.

View larger version (37K): [in this window] [in a new window]   FIGURE 8. Recordings from an experiment with 4 RyRs in the bilayer showing coupled gating at –40 mV. At +40 mV, the channels appeared to gate independently (see text). The baths contained symmetric 250 mM Cs+ solutions with [Ca2+]c = 1 nM and [Ca2+]l = 1 mM. The current baselines are shown by dashed lines and the dashed lines (O1–O4) indicate current levels associated with increasing numbers of open channels.

View larger version (14K): [in this window] [in a new window]   FIGURE 9. The dependence of mean channel opening rate (A) and closing rate (B) on the number of channels open in the bilayer under various experimental conditions at –40 mV. Cytoplasmic and luminal baths contained 250 mM Cs+ solutions plus (cytoplasmic/luminal): (, N = 7) 2 mM ATP, 1 nM Ca2+/1 mM Ca2+; (•, N = 6) 2 mM ATP, 1 nM Ca2+/0.1 mM Ca2+ (, N = 7) 60–230 µM Mg2+, 2 mM ATP, 1 nM Ca2+/1 mM Ca2+; (, N = 4) 100 µM Ca2+/1 mM Ca2+. The channel opening rate could be increased by the opening of other channels in the bilayer when ATP was present and when luminal [Ca2+]l = 1 mM. Closing rates did not significantly depend on the presence of other open channels in the bilayer.

View larger version (15K): [in this window] [in a new window]   FIGURE 10. The dependence of opening rate on the presence of open channels in the bilayer obtained from individual experiments. For each experiment, the opening rates, k0+, in the absence of open channels (opening to level O1, see Fig. 8) are shown on the abscissa while opening rates in the presence of one open channel, k1+, (opening to level O2) is on the ordinate. The dashed line indicates independent gating of uniform channels. Datum points above this line indicate coupled channel openings while points below the line can arise independent gating of a heterogeneous group of channels. The legend shows the bilayer potential and luminal [Ca2+] in the absence of cytoplasmic Mg2+ (A) or in the presence of 60–230 µM Mg2+ (B). Coupling between channels was promoted by luminal [Ca2+] and cytoplasmic Mg2+ and was abolished by positive bilayer potentials or removing cytoplasmic ATP and using 100 µM Ca2+ as the primary channel activator.

View larger version (14K): [in this window] [in a new window]   FIGURE 11. The effect of cytoplasmic Mg2+ on the channel opening rate of single RyRs (labeled single) and coupled groups of RyRs (labeled ko+ and k1+) at bilayer potentials of –40 mV (A) and +40 mV (B). The mean rate of the first channel opening ko+ (•) is similar to that seen for single channels (, N = 5 at –40 mV and N = 14 at +40 mV). The inhibitory effect of cytoplasmic Mg2+ is seen here as a decrease in the channel opening rate. However, the opening rates of RyRs in the presence of an open channel, k1+, were less sensitive to cytoplasmic Mg2+ at negative bilayer potentials than at positive potentials. The data (•, ) were compiled from experiments paired at +40 and –40 mV at 1 mM luminal Ca2+ (N = 4).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Luminal Ca²⁺ Alleviates Mg²⁺ Inhibition by an Allosteric Mechanism
Another major finding of this study is that the sensitivity of RyRs to Mg²⁺ inhibition is modulated by the level of luminal Ca²⁺. Increasing [Ca²⁺]_l from 0.1 to 1 mM produced a fourfold decrease in the Mg²⁺ affinity of the A-sites regardless of whether RyRs were in the presence of high or low cytoplasmic [Cs⁺] and [Ca²⁺], or whether they were stimulated by ATP or modified by ryanodine (Tables II and III). A number of our findings indicate that this effect is not due to competition at the A-sites between cytoplasmic Mg²⁺ and luminal Ca²⁺ that permeates the channel, but rather due to an allosteric mechanism. First, K_i(Mg²⁺) for single RyRs was independent of the bilayer potential and hence Ca²⁺ feedthrough. Second, increased luminal Ca²⁺ alleviated Mg²⁺ inhibition via changing both the mean open and closed durations of the channel (Table VI). The effect of luminal Ca²⁺ on channel closures suggests that it has access to its site of action when the channel is impermeable to Ca²⁺. Finally, the effect of luminal Ca²⁺ on Mg²⁺ inhibition is noncompetitive (Tables II and III). For example, competitive kinetics could not account for the combined effect of luminal Ca²⁺ and cytoplasmic Cs⁺ on K_i(Mg²⁺) (Table II, entries 5 and 7) or the significant effect of luminal Ca²⁺ on K_i(Mg²⁺) in the presence 10 µM cytoplasmic Ca²⁺ (Table III, entries 8 and 9).

Comparison with Cardiac RyRs
Evidence for allosteric modulation of cardiac RyRs by luminal Ca²⁺ has been reported previously (Gyorke and Gyorke, 1998). However, the effect of luminal Ca²⁺ on cardiac RyRs (RyR2 isoform) is quite different from that found here for skeletal RyRs (RyR1). With RyR2, luminal Ca²⁺ modulates channel function by sensitizing the Ca²⁺ activation site (A-sites) and by reducing inhibition at the low affinity inhibition sites (I-sites) (Gyorke and Gyorke, 1998) and a reduction in maximal level of Ca²⁺ activation (P_max) (Xu and Meissner, 1998). In contrast, with ATP-activated RyR1, luminal Ca²⁺ has no evident effect on the Ca²⁺ sensitivity of either the A- or I-sites nor does it alter P_max (Table I). In addition, at low (<1 µM) cytoplasmic Ca²⁺, luminal Ca²⁺ causes substantially more activation of RyR1 than of RyR2.

In single RyR1 channels, we did not detect competition between luminal Ca²⁺ and cytoplasmic Mg²⁺ for the A-sites, either by measuring the voltage dependence (i.e., Ca²⁺ flux dependence) of K_i(Mg²⁺) or its dependence on Cs⁺ competition. However, evidence of trans-channel competition has been reported in RyR2 by Xu and Meissner (1998) who noted that cytoplasmic Mg²⁺ induced a voltage dependence in the effect of luminal Ca²⁺, which can be interpreted as a voltage-dependent Mg²⁺ inhibition.

Ca²⁺ Feedthrough Couples RyRs in Lipid Bilayers
The coupled opening of RyRs that we observe is very similar to the phenomenon reported by Copello et al. (2003) and Porta et al. (2004). They proposed that coupled gating of RyRs occurred in bilayers when Ca²⁺ flow (luminal to cytoplasm) through one channel raised the local [Ca²⁺]_c sufficiently to activate neighboring RyRs. Several of our findings support their conclusion. First, the coupling of channel openings was promoted under conditions which favored the flow of luminal Ca²⁺ through the channel (i.e., negative potential and high [Ca²⁺]_l). Second, coupling did not occur in the presence of high (100 µM) cytoplasmic Ca²⁺ (Fig. 10). At these concentrations, the A-sites are virtually saturated and the feedthrough of luminal Ca²⁺ would have no significant activating effect. Finally, the opening of one channel in a group of RyRs caused an approximately fivefold reduction in Mg²⁺ inhibition of the other RyRs (Fig. 11), suggesting that local [Ca²⁺]_c is indeed increased and is competing with Mg²⁺ for the A-sites. From the effects of cytoplasmic [Ca²⁺] on P_o and K_i(Mg²⁺) (Figs. 1 and 4), one can estimate [Ca²⁺]_c to be 10 µM at RyRs neighboring an open RyR. The separation of these RyRs in the bilayer may be estimated from theoretical concentration profiles of [Ca²⁺] in the vicinity of a pore (Stern, 1992) (Fig. 12). The [Ca²⁺] near an open channel falls off sharply with distance because Ca²⁺ emerging from the pore is rapidly chelated by BAPTA (4.5 mM) in the bath. Fig. 12 shows that at –40 mV, 1 mM luminal Ca²⁺ generates 1 µM [Ca²⁺] in the cytoplasmic bath at 30 nm from an open channel. Reversing the polarity of the potential or reducing [Ca²⁺]_l to 0.1 mM decreases this distance to 20 nm.

View larger version (22K): [in this window] [in a new window]   FIGURE 12. Theoretical cytoplasmic [Ca2+] profiles generated by luminal Ca2+ fluxes emanating from an RyR pore into a medium containing 4.5 mM BAPTA. Ca2+ fluxes were calculated using the rate model developed by Tinker et al. (1992) and the [Ca2+] profiles were calculated using Eq. 13 from Stern (1992) with ion diffusion constants given by Xu and Meissner (1998). Solid curve, [Ca2+]l = 1 mM, –40 mV; long dashes, [Ca2+]l = 0.1 mM, –40 mV; short dashes, [Ca2+]l = 1 mM, +40 mV. [Ca2+] > 1 µM (horizontal line) significantly activates RyRs and increases Ki(Mg2+) (see Figs. 1 and 4).

The close packing of RyRs in the membrane makes it seem possible that RyRs could interact physically. In fact, a coupling phenomenon has been identified in CHAPS-purified RyRs (RyR1 and RyR2; Marx et al., 1998, 2001) that is likely to stem from a physical interaction between RyRs. Those authors found that protein fractions enriched in RyR multimers produced synchronously gated channels in 10% of instances when reconstituted with lipid bilayers. Functional coupling required FKSO6 binding protein, FKBP 12 and FKBP 12.6, but was independent of luminal Ca²⁺. Our experimental conditions were not designed to optimize observation of this type of coupling since our membrane preparations were not enriched in RyR multimers. We found such behavior in RyRs to be quite rare, observing it in only two instances out of many hundreds of recordings when using cardiac SR vesicles (Honen, B., personal communication). Therefore, RyR coupling by luminal Ca²⁺ is the dominant mechanism operating in our experiments.

Coupled and Single RyRs are Regulated Differently by Luminal Ca²⁺
Our data shows that Mg²⁺ inhibition is highly depen