Department of Pharmacology, Juntendo University School of Medicine,
Tokyo 113-8421, Japan
 |
INTRODUCTION |
The ryanodine receptor (RyR) is a
Ca2+ release channel on the sarcoplasmic
reticulum (SR) (Coronado et al., 1994
; Meissner, 1994; Ogawa, 1994;
Sutko and Airey, 1996). The RyR is a large (~2.3 MDa) homotetrameric
protein complex and forms the "foot" structure in situ, which spans
the gap between transverse tubule and terminal cisternae of SR
(Franzini-Armstrong and Protasi, 1997; Wagenknecht and Radermacher,
1997). To date, three genetically distinct isoforms of RyR (RyR1-3)
have been identified in mammalian tissues. RyR1 is a primary isoform in
mammalian skeletal muscles and plays an important role in
excitation-contraction (E-C) coupling. In many skeletal muscles of
nonmammalian vertebrates, in contrast, two isoforms of RyR, 
- and
-RyRs, which are homologs of mammalian RyR1 and RyR3, respectively,
coexist in nearly equal amounts (Ogawa, 1994; Sutko and Airey, 1996).
The RyR shows properties of Ca2+-induced
Ca2+ release (CICR), which is modulated by
several endogenous and exogenous ligands (Coronado et al., 1994;
Meissner, 1994; Ogawa, 1994). Among them, Mg2+,
adenine nucleotides, and caffeine have attracted much interest because
of their physiological and pharmacological relevance (Endo, 1977,
1981). The myoplasmic ATP concentration of the skeletal muscle is
reported to be 3-9 mM, of which the major form is MgATP (more than
90%) (Godt and Maughan, 1988), and the free Mg2+
concentration is estimated to be ~1 mM (Westerblad and Allen, 1992;
Konishi et al., 1993).
CICR activity of RyR is biphasically regulated by
Ca2+: micromolar Ca2+
activates the channel, whereas millimolar Ca2+
inhibits it. The biphasic effect of Ca2+ suggests
the existence of two classes of Ca2+ sites: a
high-affinity Ca2+ activation site (A-site) and a
low-affinity Ca2+ inactivation site (I-site)
(Meissner, 1994; Ogawa, 1994). Mg2+ decreases the
peak value of CICR activity with a reduction in the
Ca2+ sensitivity for activation (Endo, 1977,
1981). The extent of inhibition by Mg2+ depends
on the Ca2+ concentration, in marked contrast to
the case of procaine, which shows
Ca2+-independent inhibition (Kurebayashi and
Ogawa, 1998). This is accounted for by the dual effects of
Mg2+ on the two sites as follows: a competitive
antagonist against Ca2+ in the former site and an
agonist in the latter site (Endo, 1981; Laver et al., 1997; Meissner et
al., 1997). It should be noted, however, that the effect of
Mg2+ was assumed to be due primarily to
competitive antagonism on the A-site in many experiments with skinned
or cut fibers (Lamb and Stephenson, 1991; Jacquemond and Schneider,
1992; Lacampagne et al., 1998).
It is well known that ATP increases CICR activity without changing its
Ca2+ dependence (Endo, 1981). However, it remains
unknown whether ATP affects Mg2+ sensitivity.
This is due to difficulty in analyzing the effect of
Mg2+ in the presence of ATP, because it is
unclear whether MgATP is as potent as free ATP in stimulating CICR.
Caffeine is known to stimulate CICR activity by increasing the
Ca2+ sensitivity for activation (Endo, 1981).
However, its effect on Ca2+ inactivation is
unclear, and whether the affinity for Mg2+ of
either the A-site or the I-site is affected remains to be determined.
To understand the physiological and pharmacological roles of the CICR
activity of RyR in situ, knowledge of the affinities for
Mg2+ of these A- and I-sites and their
modulations by endogenous ligands or drugs is essential. In frog
muscle, where two isoforms are coexpressed, the role of each isoform in
E-C coupling cannot be discussed without knowing the CICR activity of
each one in the presence of ~1 mM Mg2+.
In this study, we examined the affinity for Mg2+
as well as for Ca2+ of the A- and I-sites of -
and -RyRs on the basis of the model for the actions of
Ca2+ and Mg2+. We used a
[3H]ryanodine binding assay with the two
isoforms purified from bullfrog skeletal muscle, which enabled us to
analyze the effect of Mg2+ on the individual
isoforms (Murayama and Ogawa, 1992, 1996). Because it has been reported
that affinities for Ca2+ and
Mg2+ are affected by many factors, including
ionic or nonionic solutes, detergents, and phospholipids (Ogawa et al.,
1999), their affinity must be assessed under conditions that are as
close as possible to the physiological environment. Taking advantage of
the finding that there was only a minor difference at most in the
[3H]ryanodine binding activity between the two
purified isoforms, we extended the same analysis procedure to the CICR
experiments to determine affinities for Ca2+ and
Mg2+ of both A- and I-sites of the
Ca2+ release channel, using frog skinned skeletal
muscle fibers; in these the molecular organization of biological
components involved in Ca2+ release from SR is
well maintained (Kurebayashi and Ogawa, 1986, 1998). Based on the
results of these experiments, we will discuss the role of
Mg2+ in the CICR activity of RyRs in situ, with
particular reference to the effects of adenine nucleotides and
caffeine. A preliminary report appeared earlier (Murayama et al.,
1998a).
| |
MATERIALS AND METHODS |
Materials
[3H]Ryanodine (60-90 Ci/mmol) was
purchased from Du Pont-New England Nuclear. Pure ryanodine was a
generous gift from Wako Pure Chemical Industries. All other reagents
were of analytical grade. Concentrations of AMP and -,
-methyleneadenosine triphosphate (AMPPCP) were determined by
spectrometry, using molar extinction coefficients at 259 nm of
15.4 × 103 and 14.2 × 103 M
1
cm1, respectively. Free
Ca2+ and Mg2+
concentrations were calculated using dissociation constants as follows:
EGTA for Ca2+ (pKapp = 5.94 for [3H]ryanodine binding, 6.02 and 6.43 for
CICR experiments at pH 6.8 and 7.0, respectively) was taken from
Harafuji and Ogawa (1980), EGTA for Mg2+
(pKapp = 0.49 at pH 6.8) was from Martell and
Smith (1974), AMPPCP for Ca2+ and
Mg2+ (pKa = 7.7;
pKCa = 4.16; pKMg = 4.68)
was by Ogawa et al. (1986), and AMP for Ca2+ and
Mg2+ (pKa = 6.26;
pKCa = 1.86; pKMg = 1.92)
was by Khan and Martell (1967).
[3H]Ryanodine binding
- and -RyRs were purified from the heavy fraction of SR
vesicles of bullfrog skeletal muscle (Murayama and Ogawa, 1992). Assays
of [3H]ryanodine binding to each of the two
isoforms were carried out as described in Murayama and Ogawa (1996),
with some modifications. Purified RyR (1-2 µg) was incubated with
8.5 nM [3H]ryanodine for 5 h at 25°C in
200 µl of a reaction medium containing 0.17 M NaCl, 20 mM
3-(N-morpholino)-2-hydroxypropanesulfonic acid (MOPSO)/NaOH
(pH 6.8), 1%
3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid
(CHAPS), 0.5% phospholipids, 2 mM dithiothreitol, and 4 mM AMP.
A free Ca2+ concentration was set by mixing 10 mM
EGTA with a specified amount of CaCl2. In
experiments with Mg2+,
MgCl2 was added to the medium and NaCl was
reduced to keep the ionic strength constant. This care is critical,
particularly in the presence of high Ca2+ and
Mg2+ concentrations, because the increased ionic
strength may weaken the inhibition (Murayama and Ogawa, 1996; Murayama
et al., 1998b). Samples were then filtered through Whatman GF/B glass
filters that had been soaked with 2% polyethylenimine. Filters were
rinsed twice with ice-cold water and dried. Radioactivity retained on the filters was counted in a liquid scintillation counter. Nonspecific binding was determined in the presence of 50 µM unlabeled ryanodine.
CICR experiments
A single skinned muscle fiber was obtained from the
iliofibularis muscle of Rana japonica and mounted in the
experimental chamber for optical determination of
Ca2+ concentrations as described (Kurebayashi and
Ogawa, 1998; Murayama et al., 1998b). All experiments were performed at
a temperature of 16°C.
Table 1 shows the composition of
solutions used for the experiments. A relaxing solution (RS) and
Ca2+ loading solution (LS) contained 4 mM MgATP
and 1 mM free Mg2+, and the other solutions,
i.e., three kinds of washing solutions (W1-W3), test solutions (TS),
and a discharging solution (DS), did not contain ATP, to avoid active
transport of Ca2+ into SR by
Ca2+ pump. W3 and DS contained 0.1 mM EGTA and 1 µM fura-2 for determination of Ca2+ released
from SR. TS contained 10 mM EGTA and 0-10 mM total calcium to achieve
a desired cytoplasmic Ca2+ concentration
([Ca2+]C) when
[Ca2+]C 
0.1 mM. If
0.1 < [Ca2+]C < 10 mM, it contained 10 mM CaCl2 and EGTA calculated
as (10 
[Ca2+]C)
mM. AMP, AMPPCP, Mg2+, and/or caffeine were added
to the TS if necessary.
The experimental protocol for determinations of the activity of CICR in
skinned fibers was similar to that described (Kurebayashi and Ogawa,
1998). The rate of Ca2+ release was determined by
the rate of decrease in the total amount of calcium remaining in the SR
(Ca in SR) after a certain stimulus, as follows. A skinned fiber was
initially treated with DS to empty SR of Ca2+,
and then SR was actively loaded to a constant level (prescriptive loading level) by incubation with LS (pCa 6.5) for 2 min. After removal
of ATP by washing successively with W1 for 60 s and W2 for 30 s, the skinned fiber was treated with a TS for a specified period
(t 
3 s). The fiber was then successively washed
with W1, W2, and W3 and challenged with DS to discharge all
releasable Ca2+ in SR. The amount of discharged
Ca2+ (Ca in SR) was determined from the
fluorescence ratio signal of fura-2 in DS. The protocol was repeated
with the same fiber. The prescriptive loading level without TS
treatment was determined in every three to five series of experiments
for the standard in calibration, and Ca in SR remaining after
incubation for a specified period (t) in TS was expressed as
its relative value (Y). Decay of Ca in SR apparently follows
a first-order kinetics in TS, and the final steady level, S,
is dependent on cytoplasmic [Ca2+]
([Ca2+]C), as described
previously (Kurebayashi and Ogawa, 1998). The time course of
Ca2+ release can be best fitted by the equation
Y = (1 S) × exp (kapp × t) + S,
where kapp stands for an apparent rate
constant. kapp is a measure of the
Ca2+ release channel activity. S was
practically zero at 0.02 mM
[Ca2+]C (pCa 4.7) or
less. At [Ca2+]C higher
than 0.02 mM, S was determined by incubating the fiber for
300 s in the TS solution.
Corrections for kapp at mM cytoplasmic Ca2+
As described above, the obtained
kapp is the value of the decrease in
total calcium in SR (Ca in SR), i.e., the sum of free Ca2+ and bound calcium in SR, under the
assumption that its time course may be approximated in a first-order
kinetics. In principle, the Ca2+ flux is caused
by the downhill movement of free Ca2+ through a
channel showing an intrinsic permeability. Therefore, if all of the
Ca2+ stored in the SR were free, the rate
constant for the Ca2+ release would be equal to
the intrinsic rate constant. The presence of the massive
Ca2+-binding sites in the lumen, such as
calsequestrin and Ca2+-ATPase, however, may lower
the concentration of free Ca2+ with the same
total amount of calcium in SR. In this case, the decrease in the
luminal free Ca2+ concentration during
Ca2+ release is buffered by the luminal bound
Ca2+. The apparent rate constant for this
Ca2+ release,
kapp, will be smaller than the
intrinsic rate constant, even if Ca2+
instantaneously dissociates from the binding sites. In the
Ca2+-loading conditions mentioned here, the
luminal free Ca2+ concentration is estimated to
be 10 mM, because Ca in SR was unchanged after the incubation with a
solution of 10 mM Ca2+ in the absence of
Ca2+ pump activity by removal of ATP, whereas it
was increased and decreased after the incubation with a higher and
lower Ca2+ concentration, respectively
(Kurebayashi and Ogawa, 1998). We also found that the luminal
Ca2+-binding sites might be homogeneous and
independent, and that the total sites would be 14 mM with
KD = 1 mM. Therefore, Ca in SR at the
initial loading level is calculated to consist of 13 mM bound calcium
and 10 mM free Ca2+. The apparent rate constant
for the Ca2+ release from the SR at the loading
level to a solution of a high Ca2+ concentration,
say 3 mM, is greater than that to a low Ca2+
concentration solution, say 10 µM, for the reason mentioned below. The final luminal free Ca2+ and bound calcium are
calculated to be 3 and 10 mM in the former case, and 0.01 and 0.14 mM
in the latter case, respectively. In the former,
Ca2+ release was largely driven by change in free
Ca2+, whereas there was a more marked
contribution of the bound calcium in the latter. This difference in the
apparent rate constant that depends on the Ca2+
concentration of the incubation medium must be corrected for. Furthermore, we have suggested that the luminal
Ca2+ exerts an inhibitory effect on the intrinsic
rate constant of the Ca2+ release channel with
Ki 
2 mM (Kurebayashi and Ogawa,
1998). Taking these findings into consideration, the correction factors at various Ca2+ concentrations were obtained as
follows: 1.00 at pCa > 4.9, 1.02 at pCa 4.6, 1.05 at pCa 4.0, 1.15 at pCa 3.5, 1.44 at pCa 3.0, and 1.75 at pCa 2.5. For the analysis
of CICR activity at a high [Ca2+]C,
kapp was divided by this factor, and
the corrected value was denoted as
k'app. This correction, however,
actually affects only the dissociation constant for
Ca2+ in the I-site. The other parameters for the
A-site or I-site were not significantly affected.
Data analysis
The results of [3H]ryanodine binding and
CICR experiments were fitted to equations obtained according to a model
for the actions of Ca2+ and
Mg2+ on the Ca2+ release
channel to yield parameters for these divalent cations of the two
Ca2+ sites (see Results). The curve fit was
performed using nonlinear regression by Sigma Plot, version 5 for
Macintosh (Jandel Scientific). Data are expressed as means ± SE,
except as otherwise stated.
| |
RESULTS |
[3H]Ryanodine binding to the purified - and
-RyRs
Effect of Mg2+
[3H]Ryanodine binding to - and -RyRs
that were purified from bullfrog skeletal muscle was carried out in an
isotonic medium containing 0.17 M NaCl. In such a medium frog RyRs show
only a small amount of [3H]ryanodine binding
(~10 pmol/mg protein) without any added ligand other than
Ca2+, even in the presence of its optimal
concentrations (Ogawa and Harafuji, 1990; Murayama and Ogawa, 1996).
Therefore, 4 mM AMP, which showed very weak affinity for divalent
cations, was added to the medium to stimulate the activity of RyR
without significant change in its Ca2+
sensitivity. The binding at the optimal Ca2+
concentration in the absence of Mg2+ amounted to
100-120 pmol/mg protein for both isoforms.
Fig. 1 shows the effects of
Mg2+ on the Ca2+-dependent
[3H]ryanodine binding to -RyR. The
Ca2+-dependent
[3H]ryanodine binding was biphasic in the
absence of Mg2+: Ca2+ was
stimulatory at a concentration lower than 0.1 mM and inhibitory at a
higher concentration (Fig. 1 A, open circles).
Mg2+ (5.9 mM) depressed the peak value to about
one-third of the control value and shifted the stimulatory
Ca2+ to a higher concentration range: the
EC50 value of Ca2+ was
increased from 10 µM in control to 32 µM in the presence of 5.9 mM
Mg2+ (Fig. 1 A, filled
circles).
View larger version (21K):
[in this window]
[in a new window]
|
FIGURE 1
Scheme illustrating the procedure for determinations of
the affinities for Ca2+ and Mg2+ of the A- and
I-sites of frog RyRs. [3H]Ryanodine binding to the
purified -RyR was carried out as described in Materials and Methods
in a reaction medium containing 8.5 nM [3H]ryanodine,
0.17 M NaCl, 20 mM MOPSO/NaOH (pH 6.8), 1% CHAPS, 0.5%
phospholipids, 2 mM dithiothreitol, 4 mM AMP, and various free
Ca2+ and Mg2+. (A)
Ca2+ dependence of the ryanodine binding in the presence
( ) and absence ( ) of 5.9 mM Mg2+. Data are means ± SE (n = 4). (B) Dose-dependent
inhibition by Mg2+ of [3H]ryanodine binding
at three different Ca2+ concentrations (arrows
a-c are as indicated in A): 10.7 µM
(a), 110 µM (b), and 1.0 mM
(c). Data are means ± SE (n = 3). Parameters for Ca2+ and Mg2+ of the A- and
I-sites were determined in three consecutive steps. First, the data in
the absence of Mg2+ ( in A) were fitted
to Eq. 2 to yield the following parameters:
Bmax = 117.4 pmol/mg protein;
KA,Ca = 10.1 µM;
nA,Ca = 2.0;
KI,Ca = 2.73 mM;
nI,Ca = 1.0. Second, the dose-dependent
inhibition by Mg2+ ( in B) in the
presence of 1 mM Ca2+ (arrow c in
A), where competition of Mg2+ for the A-site
is negligible, was fitted to Eq. 3 to obtain
KI,Mg (3.16 mM) and
nI,Mg (0.9). Third, the data ( in
B) at 10.7 µM Ca2+ (arrow a
in A), where competitive inhibition by Mg2+
is prominent, were fitted to Eq. 1 to estimate
KA,Mg (396 µM) and
nA,Mg (1.1). The validity of the obtained
parameters was confirmed by comparing the curve computed according to
Eq. 1 with the data (curve b in B) at 110 µM Ca2+ (arrow b in A).
These parameter values were also verified by the curve fit to the
Ca2+-dependent [3H]ryanodine binding in the
presence of 5.9 mM Mg2+ as shown in A.
|
|
The widely accepted explanation for the inhibitory effects of
Mg2+ is as follows. The RyR or
Ca2+ release channel has both the high-affinity
Ca2+ activation site (A-site) and the
low-affinity Ca2+ inactivation site (I-site).
Mg2+ serves as a competitive antagonist on the
A-site and as an agonist on the I-site (Laver et al., 1997; Meissner et
al., 1997). Therefore, the following equations are predicted according
to the model corresponding to the above explanation. Channels with the
A-site occupied by Ca2+ (the probability of
fA) and the I-site free of
Ca2+ or Mg2+ (the
probability of (1 fI)) are in
the activated state. fA and 1 fI are expressed as follows:
where KA,Ca,
KA,Mg,
KI,Ca, and
KI,Mg represent the dissociation
constants for Ca2+ and Mg2+
of the A- and I-sites, respectively.
nA,Ca,
nA,Mg,
nI,Ca, and nI,Mg represent the Hill coefficients
of the relevant sites. Because [3H]ryanodine is
considered to bind only to the open channel (Coronado et al., 1994;
Meissner, 1994; Ogawa, 1994), the [3H]ryanodine
binding (B) in the presence of a specified concentration of
[3H]ryanodine as determined in Fig. 1
A can be expressed by Eq. 1:
|
(1)
|
where Bmax is the maximum amount
of [3H]ryanodine binding to be expected in the
presence of a specified concentration of the ligand. It should be noted
that Bmax is different from the
conventional value for maximum binding, which refers to the value in
the presence of an infinite amount of
[3H]ryanodine.
Bmax is also affected by modulators
such as adenine nucleotides and caffeine. A preliminary attempt to
determine all nine parameters, including
Bmax in Eq. 1 by curve fit of the data of Ca2+ dependence, as shown in Fig. 1
A, was unsuccessful, because there were too many parameters
to fix for the data points currently available. To determine these
parameters accurately with a relatively small number of data points, we
designed an analysis procedure comprising three consecutive steps and
confirmed of the validity of the results. This procedure was found to
work well, as shown in Fig. 2. First, in
the absence of Mg2+ ([Mg2+ = 0]), Eq. 1 is simplified into Eq. 2:
|
(2)
|
The results of Ca2+-dependent
[3H]ryanodine binding in the absence of
Mg2+ (Fig. 1 A, open
symbols) were fitted to Eq. 2 to yield
Bmax and the parameters for
Ca2+ of the A-site
(KA,Ca,
nA,Ca) and the I-site
(KI,Ca, and
nI,Ca). Second, at a
[Ca2+] much higher than
KA,Ca (e.g., Fig. 1 A, arrow
c), where competitive inhibition by Mg2+ on
the A-site is negligible, the inhibition by Mg2+
can be explained by the action on the I-site alone. Under the circumstances, Eq. 1 is simplified into Eq. 3:
|
(3)
|
The results of dose-dependent inhibition by
Mg2+ in the presence of 1 mM
Ca2+ (Fig. 1 B, triangles, curve
c) were fitted to Eq. 3 to obtain the parameters for
Mg2+ of the I-site
(KI,Mg and
nI,Mg), using the other parameters
already fixed. Third, the results of Mg2+
dependence (Fig. 1 B, squares, curve a) at a
[Ca2+] near
KA,Ca (Fig. 1 A, arrow a),
where competitive inhibition by Mg2+ is
prominent, would give the parameters for Mg2+ of
the A-site (KA,Mg and
nA,Mg) according to Eq. 1, because the other parameters used in the equation are already known. The validity of the obtained parameters was confirmed by comparing the results at a
mid-range of free Ca2+ concentration (e.g., 0.1 mM Ca2+) (Fig. 1 B, circles,
curve b; see also Fig. 1 A, arrow b).
These parameters also can predict well the
Ca2+-dependent
[3H]ryanodine binding in the presence of 5.9 mM
Mg2+, as shown in Fig. 1 A.
View larger version (22K):
[in this window]
[in a new window]
|
FIGURE 2
Effect of Mg2+ on the
Ca2+-dependent [3H]ryanodine binding to frog
RyRs. [3H]Ryanodine binding to the purified -RyR
(A) and -RyR (B) was carried out as in
Fig. 1 at various free Ca2+ concentrations in the presence
of 0 (), 0.8 ( ), 2.5 (), and 5.9 () mM free
Mg2+. Data are means ± half-range of deviations of
duplicate determinations. Curves in A and
B were drawn according to Eq. 1, using the following
parameters: Bmax = 117.4 and 113.1 pmol/mg protein, KA,Ca = 10.1 and 18.1 µM, nA,Ca = 2.0 and 2.3, KI,Ca = 2.73 and 2.72 mM,
nI,Ca = 1.0 and 1.2, KA,Mg = 396 and 350 µM,
nA,Mg = 1.1 and 0.9, KI,Mg = 3.16 and 5.23 mM, and
nI,Mg = 0.9 and 0.9 for -RyR and
-RyR, respectively.
|
|
We would like to point out that the inhibition by
Mg2+ depended on the Ca2+
concentration where the dose-effect curve for
Mg2+ was obtained. Fig. 1 B shows that
IC50 values for Mg2+ were
0.48, 2.7, and 4.1 mM at 10.7 (open squares), 110 (open circles), and 1000 (filled triangles) µM
Ca2+, respectively. The three-step procedure of
analysis is indispensable for the coherent understanding of these
effects of Mg2+, which are of apparently variable
grade. These systematic determinations and the analysis procedure are
the fundamental principle throughout these experiments.
Fig. 2 demonstrates Ca2+-dependent
[3H]ryanodine binding to -RyR (Fig. 2
A) and -RyR (Fig. 2 B) in the presence of
0-5.9 mM free Mg2+. Computed curves using
parameters determined as described above corresponded well to the
experimental data points for both RyR isoforms at every
Mg2+ concentration. A similar set of parameters
was also obtained all at once by fitting all of the data points for
each isoform (Fig. 2) in the presence of various
Ca2+ and Mg2+
concentrations to Eq. 1, using a 3D curve fitter (Sigma Plot, version
5, for Macintosh) (data not shown). These results suggest that the
model and the parameters determined by the three-step procedure well
explain the effects of Mg2+ on the
[3H]ryanodine binding activity of both - and
-RyRs.
Table 2 shows a summary of six to eight
similar determinations. Although KA,Ca
for -RyR (18.1 µM) appears to be slightly larger than that for
-RyR (11.0 µM), the difference was not great, and we may conclude
that they were similar to each other. There was no difference between
- and -RyRs in KI,Ca (2.38 versus 2.34 mM), KA,Mg (324 versus 325 µM), or KI,Mg (2.79 versus 3.06 mM).
The A-site showed 20-30-fold higher affinity for
Ca2+ than for Mg2+. It
should be noted that the Hill coefficient for
Mg2+ (nA,Mg = 0.9-1.0) was significantly smaller than that for
Ca2+ (nA,Ca = 2.1-2.3). On the other hand, the I-site showed similar affinity
(KD of 2-3 mM) for
Ca2+ and Mg2+, with a Hill
coefficient of ~1. In addition to these parameters, the
Bmax values of the two isoforms were
also similar (107 versus 118 pmol/mg protein).
View this table:
[in this window]
[in a new window]
|
TABLE 2
Summary of parameters for Ca2+ and
Mg2+ of the A- and I-sites and
Bmax, which were determined from
[3H]ryanodine binding to the purified RyRs
|
|
Effect of caffeine
Caffeine is a well-known activator of RyRs and is thought to
enhance the Ca2+ sensitivity for activation.
However, it is still unclear how the effect of the drug is modulated in
the presence of Mg2+. We examined in detail the
effect of caffeine on [3H]ryanodine binding in
the presence of various concentrations of Ca2+
and Mg2+ and analyzed the effect of the drug on
the A- and I-sites. Fig. 3 A
demonstrates the effects of various concentrations of caffeine on
Ca2+-dependent
[3H]ryanodine binding to -RyR. Caffeine
dose-dependently enhanced the apparent Ca2+
sensitivity for activation: EC50 for
Ca2+ was reduced from 9.9 µM (control) to 3.6 and 1.1 µM with 2 and 10 mM caffeine, respectively. No further
Ca2+-sensitizing effect was observed at 15 mM
caffeine (data not shown). Furthermore, it slightly reduced the
inactivation in the presence of high Ca2+
concentrations. In addition, 2 mM or more caffeine increased the
binding at optimal Ca2+ by ~20%. Similar
results were obtained with the purified -RyR (see Table 2). These
findings were consistent with previous studies of
[3H]ryanodine binding to the purified RyRs
(Murayama and Ogawa, 1996) and isolated SR vesicles (Ogawa and
Harafuji, 1990).
View larger version (24K):
[in this window]
[in a new window]
|
FIGURE 3
(A) Effect of caffeine on the
Ca2+-dependent [3H]ryanodine binding to frog
RyRs. [3H]Ryanodine binding to -RyR was carried out as
in Fig. 1 in the presence of 0 (), 2 (), and 10 () mM
caffeine. Data are means ± half-range of deviations of duplicate
determinations. (B) Effect of Mg2+ on the
Ca2+-dependent [3H]ryanodine binding in the
presence of 10 mM caffeine. [3H]Ryanodine binding to
-RyR was determined with 0 (), 0.8 ( ), 2.5 ( ), and 5.9 () mM Mg2+. Computed curves were drawn using Eq. 1 and
the following parameters: Bmax = 140.2 pmol/mg protein; KA,Ca = 1.19 µM;
nA,Ca = 1.5;
KI,Ca = 3.85 mM;
nI,Ca = 1.1;
KA,Mg = 379 µM;
nA,Mg = 1.1;
KI,Mg = 4.92 mM;
nI,Mg = 1.1.
|
|
Fig. 3 B shows the effects of Mg2+ on
the Ca2+-dependent
[3H]ryanodine binding to -RyR in the
presence of 10 mM caffeine. Mg2+ decreased both
the binding at the optimal Ca2+ and apparent
Ca2+ sensitivity for activation in a
dose-dependent manner, as is the case without caffeine (see Fig. 2).
EC50 for Ca2+ was increased
from 1.1 µM (control) to 2.4, 4.2, and 6.6 µM by 0.8, 2.5, and 5.9 mM Mg2+, respectively. The analysis was carried
out to obtain parameters as described above in the three-step
procedure. Table 2 shows the summary of parameters thus determined,
including those for -RyR. The experimental results obtained in the
presence of varying amounts of Mg2+ in Fig. 3
B coincided well with computed curves, when we used the
parameters shown in Table 2. Caffeine markedly reduced
KA,Ca for both isoforms (from 11 to 1 µM, 10-fold, with -RyR and from 18 to 3 µM, sixfold, with
-RyR) with a slight decrease in
nA,Ca (from 2.1 to 1.5 with -RyR
and from 2.3 to 1.7 with -RyR). In contrast,
KA,Mg and
nA,Mg were not changed substantially
by the reagent. These results suggest that caffeine increases the
affinity of the A-site for Ca2+ but not for
Mg2+. In addition to the effects on the A-site,
caffeine slightly increased KI,Ca
(from 2.4 to 4.0 mM for -RyR and from 2.3 to 5.1 mM for -RyR) and
KI,Mg (from 2.8 to 5.2 mM for -RyR
and from 3.1 to 7.8 mM for -RyR). This may be consistent with the decreased inactivation in the presence of high
Ca2+ concentrations. The
Bmax was increased by 20-30%. This
increase in Bmax may partly contribute
to the enhanced peak value of the binding. The enhancement appeared to
be saturated at 2 mM caffeine in these experiments (see Fig. 3
A). With SR vesicles, however, the peak values of
[3H]ryanodine binding increased up to threefold
or more as caffeine was increased to 15 mM (Ogawa and Harafuji, 1990;
Ogawa et al., 1999). This difference may be explained by the effect of
CHAPS, which potentiates the effect of adenine nucleotides (Ogawa et al., 1999).
CICR activity in skinned fibers
Effect of Mg2+
Fig. 4 shows the
Ca2+ dependence of the rates of CICR that were
determined in a chloride salt medium (Cl medium) containing 4 mM AMP
and 0-1.6 mM Mg2+. In the absence of
Mg2+, CICR shows a bell-shaped
Ca2+ dependence similar to that of
[3H]ryanodine binding (Fig. 2). The peak value
of the rate constant was ~7 min1, and
EC50 and IC50 for
Ca2+ were 3 µM and 1 mM, respectively.
Mg2+ had two distinct effects on the pCa-CICR
activity relationship: it dose-dependently decreased the peak rate of
Ca2+ release with an IC50
of 0.3 mM and lowered Ca2+ sensitivity for
activation. The EC50 for
Ca2+ was increased to 5, 10, and 20 µM by 0.4, 0.8, and 1.6 mM Mg2+, respectively.
View larger version (18K):
[in this window]
[in a new window]
|
FIGURE 4
Effect of Mg2+ on the Ca2+
dependence of CICR in skinned muscle fibers. Rate constants of the
Ca2+ release from SR using six skinned fibers were
determined in the Cl medium containing 4 mM AMP and various
Ca2+ concentrations. Effects of various Mg2+
concentrations were examined: 0 (), 0.4 (), 0.8 (), or 1.6 mM
() Mg2+. The results, corrected for as described in
Materials and Methods, are plotted. pH was adjusted to 7.0 for these
experiments. Parameters determined according to the procedure described
in Fig. 1 were kmax = 7.0 min1, KA,Ca = 3.2 µM,
nA,Ca = 1.6, KI,Ca = 1.1 mM,
KI,Mg = 0.42 mM,
KA,Mg = 0.15 mM. The drawn lines are
curves calculated for indicated Mg2+ concentrations, using
these parameters. nI,Ca,
nA,Mg, and nI,Mg
were fixed at 1.0 in these simulations.
|
|
We also followed the procedure of analysis in the case of
[3H]ryanodine binding.
kmax and
k'app, however, replace
Bmax and B in Eqs. 1-3,
respectively. kmax is the maximum
value of k'app. The curve for
circles in the absence of Mg2+ (Fig. 4) was drawn
by using KA,Ca = 3.2 µM,
nA,Ca = 1.6, KI,Ca = 1.1 mM, and
kmax = 7.0 min1 as the best fit parameters. In this
calculation, we assumed nI,Ca to be
unity in all of the skinned fiber experiments, because determinations with 21 fibers gave an average of 0.92 ± 0.15 (mean ± SD)
for the parameter. This conclusion is consistent with the results of
[3H]ryanodine binding (Table 2). The Hill
coefficient for the A-site, nA,Ca, was
calculated as a parameter to be determined, because it varied between
1.0 and 2.0, depending on species and concentrations of adenine
nucleotides. Using the parameters for Ca2+
described above (KA,Ca,
nA,Ca,
KI,Ca), the relation between
k'app and Mg2+
concentration in the presence of 63 µM Ca2+
(pCa 4.2) was analyzed according to Eq. 3. The best fit was obtained when the dissociation constant of the I-site for
Mg2+ (KI,Mg) was
0.42 mM. Furthermore, a KA,Mg of 150 µM was obtained according to Eq. 1 on the basis of
Mg2+-dependent inhibition of
k'app at pCa 5.2. Hill coefficients, nA,Mg and
nI,Mg, were also fixed at 1.0, based
on the results of [3H]ryanodine binding (Table
2). To verify whether the parameters obtained here explain all of the
data at various concentrations of Ca2+ and
Mg2+, CICR activity was recomputed using these
parameters, and curves for 0.4, 0.8, and 1.6 mM
Mg2+ were drawn in Fig. 4. All of the CICR
activity determined in the presence of varied
Ca2+ and Mg2+ is consistent
with the values anticipated by the model. A similar set of parameters
was also obtained all at once, using the 3D curve fitter (Sigma Plot,
version 5, for Macintosh) as mentioned above (data not shown),
supporting the assumption that the appropriate values for the
parameters were obtained by the three-step procedure. We would like to
point out that the corrected kapp
(k'app) was used here as described
in Materials and Methods. This correction was significant at pCa 2.5 and 3.0. Without this correction,
KI,Ca and
nI,Ca would be 80% larger and 16%
smaller, respectively. Other parameters
(KA,Ca,
nA,Ca,
KA,Mg,
nA,Mg,
KI,Mg, and
nI,Mg) were not significantly affected
by the correction. Table 3 (the first group) summarizes similar experiments at pH 6.8 with varied AMP concentrations.
View this table:
[in this window]
[in a new window]
|
TABLE 3
Summary of parameters for Ca2+ and
Mg2+ of the A- and I-sites and
kmax, which were determined from CICR
experiments with skinned fibers
|
|
The stimulatory principle of an adenine nucleotide: free or bound
form
Adenine nucleotides such as ATP, AMPPCP, ADP, and AMP are believed
to stimulate CICR through the common responsible site(s) (Endo, 1981).
To assess the effect of Mg2+ on
Ca2+ release in situ, CICR activity should be
determined in the presence of ATP. Its use, however, prevents us from
analyzing clearly the Ca2+ release itself because
it also drives Ca2+ uptake by activating the
Ca2+ pump. Therefore, we performed
experiments in the presence of AMPPCP, a nonhydrolyzable ATP
analog, which is very similar to ATP in stimulating
Ca2+ release but does not support
Ca2+ uptake (Kakuta, 1984). For these
experiments, we used methanesulfonate salt medium (Ms medium) instead
of Cl medium because the latter has been reported to have some
stimulating effects on RyR (Meissner et al., 1997). In our experiments
with skinned fibers, however, there were no significant differences in
the Ca2+ sensitivities of activation and
inactivation sites between the two media, while
kmax in the Ms medium was slightly
smaller than that in the Cl medium (averaged
kmax values in the presence of 4 mM
AMP were 5.6 ± 0.5 min1
(n = 4) and 3.6 ± 1.0 min1 (n = 3) in Cl and Ms
media, respectively).
An obstacle in analyzing the effect of Mg2+ on
CICR activity in the presence of AMPPCP is that a substantial fraction
of AMPPCP binds Mg2+ and
Ca2+ with an apparent
KD of ~0.2 mM. Therefore, the
question arises whether the nucleotide complexed with a divalent cation
is equivalent to free nucleotide in stimulating CICR. To clarify this
point, we compared CICR activities in the presence of AMPPCP with those in the presence of AMP, which has a much lower affinity to divalent cations. The top of Fig. 5 A
shows the Ca2+ dependence of the relative CICR
activities normalized with the values at pCa 4.5 in the presence of 4 mM AMP (filled symbols) and 0.2 mM AMPPCP (open
symbols). A suitable set of parameters in Eq. 2 can predict all of
the data points, irrespective of AMP or AMPPCP, as shown by the curve
in Fig. 5 A. Fractions of free to total nucleotide at
various Ca2+ concentrations were calculated and
are plotted in the lower panel. A substantial fraction of AMPPCP
binds to Ca2+ at calcium concentrations higher
than 0.1 mM, whereas AMP is almost free up to 3 mM
Ca2+ (bottom panel). The
Ca2+ dependences of
k'app, however, were not
significantly different between the two nucleotides. The
IC50 values of Ca2+ were
0.38 ± 0.12 mM and 0.43 ± 0.03 mM in the presence of AMP and AMPPCP, respectively. Similar IC50 values
with AMP were also obtained in Cl medium. Fig. 5 B shows
the Mg2+-dependent inhibition of CICR
activities in the presence of AMP or AMPPCP at 90 µM
Ca2+, where the activation site was expected to
be fully saturated with Ca2+. The
Mg2+ dependence with AMPPCP was very similar to
that with AMP. The IC50 values of
Mg2+ were 0.52 ± 0.09 mM and 0.44 ± 0.03 mM in the presence of AMP and AMPPCP, respectively. The fraction
of free AMPPCP, however, was much less than that of AMP in the range of
[Mg2+] > 0.1 mM (bottom panel).
These results are consistent with the idea that free AMPPCP, MgAMPPCP,
and CaAMPPCP are equally potent in stimulating CICR. Therefore, in the
following experiments we calculated the affinities of the A- and
I-sites for Ca2+ and Mg2+
under the assumption that the free and complexed forms of AMPPCP are
equivalent to each other in stimulating CICR.
View larger version (21K):
[in this window]
[in a new window]
|
FIGURE 5
Comparison of effects of AMP and AMPPCP on
Ca2+ and Mg2+ dependences of CICR in skinned
fibers. (A) Ca2+ dependence in the absence
of Mg2+. The top panel shows Ca2+ dependence of
the relative k'app, which was
normalized with the value at pCa 4.5. CICR was determined in the Ms
medium containing 4 mM AMP () or 0.2 mM AMPPCP (, , , ,
). Different symbols indicate different preparations. The bottom
panel indicates the calculated fractions of the free form to total AMP
() or AMPPCP (). AMP is mostly in the free form, whereas the
complexed form of AMPPCP makes up a considerable fraction of the
total. The solid line in the top panel was drawn based on data
with 0.2 mM AMPPCP according to Eq. 2, where the parameters were
KA,Ca = 3.9 µM,
nA,Ca = 1.4, KI,Ca = 0.43 mM, and
nI,Ca = 1. (B)
Mg2+-dependent inhibition in the presence of 90 µM
Ca2+ is shown in the top panel, and the calculated ratio of
free to total nucleotide is shown in the bottom panel. The results were
normalized with those in the absence of Mg2+. The drawn
line shows the curve of Eq. 3 best fit to data with 0.2 mM AMPPCP,
where parameters were KI,Mg = 0.36 mM
and nI,Mg = 1.0, in addition to those
determined in A. Note that CICR activities in the
presence of AMP and AMPPCP showed very similar Ca2+
(A) and Mg2+ dependences (B),
although their ratio of the free form to total nucleotides was very
different in the presence of high concentrations of Ca2+
and Mg2+.
|
|
We examined the effect of AMPPCP on each of the parameters for CICR in
situ by determining the Ca2+ release in the
presence of various concentrations of AMPPCP. Initially, we determined
the dependence of k'app on AMPPCP
concentration at pCa 4.05, which was around the optimum concentration
(Fig. 6 A). In the absence of
AMPPCP, the k'app was very small
(0.1-0.3 min1). The
k'app determined in a single fiber
increased with increase in AMPPCP concentration up to 3 mM (Fig. 6
A, open circles). At a concentration higher than
3 mM, k'app was too great to be
reliable in our experimental system. Results in Fig. 6 A
(open circles) can be satisfactorily predicted by the
conventional law of the mass reaction with an apparent dissociation
constant (5 mM) and the saturated value for
k'app (53 min1). Similar results were obtained in three
different fibers. The averages for
k'app values in the presence of 0.2 and 1 mM AMPPCP were 2.1 ± 0.8 (n = 9) and
13.2 ± 2.0 (n = 6) min1,
respectively (filled circles). In the presence of 1 mM
Mg2+, a similar dose-dependent increase in CICR
was also observed at pCa 4.05 (Fig. 6 A, open
triangles).
View larger version (16K):
[in this window]
[in a new window]
|
FIGURE 6
(A) Dose-dependent potentiation of CICR
by AMPPCP. The k'app values were
determined in the Ms medium containing 90 µM Ca2+ and
various levels of AMPPCP with () or without () 1 mM
Mg2+, using different single fibers. Solid lines correspond
to simple equations of the law of mass reaction with a dissociation
constant and a saturated value for
k'app of, respectively, 5 mM and 53 min1 in the absence of Mg2+ and 10 mM and 11 min1 in its presence. , Average with SE of all
determinations (n = 9 and 6 for 0.2 mM and 1 mM
AMPPCP, respectively) in the absence of Mg2+, including the
results of open circles. These results suggest that
k'app in the presence of 4 mM AMPPCP
would be ~35-50 min1 at pCa 4.05. (B)
Normalized Ca2+-dependent
k'app values in the presence of 0.2 mM () or 1 mM AMPPCP ( ). Parameters in Eq. 2 that would give a
best fit are listed in Table 3. Note that the two fitted curves are
very similar to each other. The averaged
kmax values were 2.9 min1 and
18 min1 in the presence of 0.2 mM and 1 mM AMPPCP,
respectively.
|
|
The Ca2+-dependent
k'app values in the presence of 0.2 mM and 1 mM AMPPCP were normalized by each peak value in Fig. 6
B. They were homologous to each other in their
Ca2+ dependences. The second group in Table 3
summarizes the results of analysis according to Eqs. 1-3 of
determinations with 0.2 and 1 mM AMPPCP in the presence of various
Ca2+ and Mg2+
concentrations. As AMPPCP was increased from 0.2 to 1 mM,
kmax was enhanced as much as sixfold.
Neither KA,Ca nor
KI,Ca, however, was significantly
affected. This result was in accordance with previous reports of
unchanged Ca2+ dependence with skinned fibers
(Endo, 1981) and with [3H]ryanodine binding to
SR vesicles (Ogawa and Harafuji, 1990). The Hill coefficient of the
A-site (nA,Ca), however, appears to be
slightly smaller at a higher concentration of AMPPCP: 1.4 and 1.2 at
0.2 mM and 1 mM AMPPCP, respectively. The average of
KI,Mg at 1 mM AMPPCP was 0.18 ± 0.02 mM (ranging from 0.13 to 0.22 mM), while that at 0.2 mM was
0.36 ± 0.03 mM (ranging from 0.26 to 0.46 mM). Under the
experimental conditions where Mg2+ and AMPPCP
were used, some variations of the KD
values for Mg2+ of AMPPCP may lead to a different
conclusion. For example, if pKMg is
4.38, which is two times larger than what we used (4.68), the range of
KI,Mg would be 0.27-0.49 mM and
0.17-0.28 mM in the presence of 0.2 mM and 1.0 mM AMPPCP,
respectively. We may conclude that the affinity of the I-site for
Mg2+ was similar between 0.2 and 1 mM AMPPCP.
This is also the case with the affinity of the A-site for
Mg2+ (Table 3). It should also be noted that
k'app in the absence of
Ca2+ (10 mM EGTA, pCa ~9) at 1.0 mM AMPPCP
(0.17 ± 0.07 min1, n = 3)
was significantly higher than that at 0.2 mM AMPPCP (0.03 ± 0.02 min1, n = 4). This activity was
completely suppressed by less than 1 mM Mg2+. As
AMPPCP was increased, Ca2+ release in the virtual
absence of Ca2+ was more prominent, but it would
be only a minor fraction (no more than 1%) of the peak rate of CICR
(data not shown). In summary, the stimulating effect of adenine
nucleotides such as AMP and AMPPCP can be explained by enhanced
kmax, but affinities for
Ca2+ and Mg2+ of A- and
I-sites remain unchanged, being independent of the concentrations of
nucleotides. This conclusion was proved regardless of the medium used
(Ms or Cl) (data not shown).
Effect of caffeine
We then examined the effect of caffeine on CICR in skinned fibers
in the presence of varying amounts of Ca2+ and
Mg2+ (Fig. 7), and
the results are summarized in the third group of Table 3. Fig. 7
A shows the potentiating effect of 5 mM caffeine on the
Ca2+-dependent CICR activity in the presence of
0.2 mM AMPPCP. Caffeine had two distinct effects: it increased the
maximum rate constant of Ca2+ release by fourfold
(kmax = 2.9 versus 11.2 min1) (Fig. 7 A, inset) and
increased the sensitivity of the Ca2+ activation
(KA,Ca = 3.0 versus 0.4 µM) by
sevenfold (Fig. 7 A). The averaged affinity of the I-site
for Ca2+ appeared to be lowered to one-third by
caffeine (KI,Ca = 0.40 versus 1.2 mM)
(Fig. 7 A). However, significant differences were not
detected (p > 0.05, t-test) between
normalized values in the presence and absence of caffeine at pCa < 4.5, except for pCa 3.0, because the experimental variation was
large at higher Ca2+ concentrations (Fig. 7
A). The same conclusion was also obtained by ANOVA analysis
(p 0.09) with StatView 5.0 for Macintosh.
View larger version (21K):
[in this window]
[in a new window]
|
FIGURE 7
Effect of caffeine on the affinities of A- and I-sites
for Ca2+ and Mg2+. All determinations were
carried out in the Ms medium containing 0.2 mM AMPPCP.
(A) Normalized Ca2+-dependent
k'app values in the presence ()
and absence () of 5 mM caffeine. *Significant difference between
their normalized values (p < 0.05). Parameters
that would give the best fits are listed in Table 3. Determined values
of k'app are plotted in the inset.
Note that caffeine enhanced the Ca2+ sensitivity for
activation by sevenfold and the k'app
at the optimal Ca2+ concentrations by fourfold.
(B) Inhibition by Mg2+ in the presence and
absence of 5 mM caffeine. Determinations were carried out at the pCas
indicated by arrows in A (open arrows, no
caffeine; filled arrows, 5 mM caffeine). Rate constants
of CICR at various concentrations of Mg2+ in the absence of
caffeine ( ) were determined at pCa 4.05 () and 5.44 ()
whereas in the presence of caffeine (- - -) at pCa 4.56 () and 6.03 (). The results were normalized by the value in the
absence of Mg2+ in each series of determinations. Best-fit
parameters in Eqs. 1-3 are listed in the third section of Table 3.
|
|
The inhibition by Mg2+ of the CICR activity
determined at Ca2+ concentrations where the
occupation with Ca2+ of the A-site was hardly
affected by Mg2+ (pCa 4.56 and 4.05, in the
presence and absence of caffeine, respectively) is shown in Fig. 7
B (the two right curves with circles). The
normalized Mg2+ dependence in the presence of
caffeine (Fig. 7 B, filled circles) was very similar to that
in its absence (Fig. 7 B, open circles), although the
absolute k'app values were increased
severalfold by caffeine. KI,Mg values
determined by Eq. 3 were not significantly changed by caffeine
(KI,Mg = 0.31 versus 0.36 mM). The two
left curves (triangles) in Fig. 7 B represent the
Mg2+ dependences obtained at
Ca2+ concentrations that gave ~60% of the
maximum activity (pCa 6.03 and 5.44, with and without caffeine,
respectively). The curve in the presence of caffeine (Fig. 7 B,
filled triangles) was not significantly different from that in the
absence of caffeine (Fig. 7, open triangles). Taken
together, similar Mg2+ dependences under the two
specified conditions indicate that caffeine does not change the
affinity for Mg2+ of the A- or I-sites. Actually,
calculated KA,Mg in the presence of
caffeine (74 µM) was also indistinguishable from that without caffeine (97 µM). In contrast to the case for
Ca2+, the affinity for Mg2+
of the A-site was not affected by caffeine. In the presence of 1 mM AMP
in the Cl medium, very similar results were also observed for the
effect of caffeine: an approximately sixfold increase in
kmax and an approximately sixfold
reduction in KA,Ca without a
significant change in other parameters (Table 3). The four- to sixfold
enhancement of kapp by 5 mM caffeine
at the optimal Ca2+ was consistently obtained in
the presence of 0-10 mM AMP (data not shown). These observations
indicate that caffeine and adenine nucleotide may act independently
through the different underlying mechanisms. This is consistent with
previous reports (Endo, 1981; Sitsapesan and Williams, 1990; Coronado
et al., 1994; Meissner, 1994; Ogawa, 1994; Sutko and Airey, 1996).
| |
DISCUSSION |
It is well known that Mg2+ decreases the
biphasically Ca2+-dependent CICR activity of the
Ca2+ release channel (RyR) with a shift to a
higher Ca2+ concentration range. However, there
was no quantitative consideration on this matter until recent
investigations made by Laver et al. (1997) and Meissner et al. (1997).
During the course of this investigation, we have learned that there are
some possible reasons for the difficulty. One of these is a substantial
overlap between the Ca2+ activation curve,
fA, and the Ca2+
inactivation curve, (1 fI). As
shown in Fig. 8
A, Mg2+ affects the
fA and (1 fI) curves differently. These effects of Mg2+ in particular make determination of the
parameters difficult without numerous systematic experiments such as
those shown here. To be freed from this difficulty, Meissner et al.
(1997) performed experiments under different conditions to obtain
separately parameters for the A- or I-site: some determinations were
carried out in a solution containing 0.5 M choline chloride, where the
contribution of the I-site was claimed to be negligible. It is still
unclear, however, whether these parameters can be valid in situ,
because the sensitivities for Ca2+ and
Mg2+ were largely affected by several factors,
including ionic species and their concentrations (Murayama and Ogawa,
1996; Ogawa et al., 1999). Laver et al. (1997), on the other hand,
determined these parameters with cardiac RyR, where the inactivation by
Ca2+ and Mg2+ was claimed
to be weak, and extended these findings to the skeletal muscle RyR to
obtain a simplified relationship with a necessary approximation.
View larger version (17K):
[in this window]
[in a new window]
|
FIGURE 8
Simulations of effects of Mg2+ on the CICR
activity in situ and analysis of the occupancy by divalent cations of
the A- and I-sites. (A) The fractions of
Ca2+-occupied A-site (fA) and
vacant I-site (1-fI) and the anticipated
CICR activity (k = kmax × fA × (1 fI)) were calculated in the presence or
absence of 1 mM Mg2+. fA (- - -), 1 fI (·····), and
k () were calculated as a function of free
Ca2+, using the following parameters, which are assumed to
be those for the Ca2+ release channel of SR in situ:
kmax = 45 min1;
KA,Ca = 2.5 µM;
nA,Ca = 1;
KI,Ca = 0.4 mM;
nI,Ca = 1;
KA,Mg = 75 µM;
nA,Mg = 1;
KI,Mg = 0.3 mM;
nI,Mg = 1. We assumed
nA,Ca to be 1 in the presence of 4 mM
AMPPCP, based on the finding that nA,Ca
values decreased with an increase in AMPPCP concentration (from 1.4 at
0.2 mM to 1.2 at 1 mM; see Table 3). Mg2+ (1 mM) shifts
curve fA rightward along the abscissa to
decrease the apparent Ca2+ sensitivity and reduces (1 fI) throughout all Ca2+
concentrations, resulting in thedecreased peak value. (B and C) The
probabilities of occupancy by divalent cations of A- (B)
and I- (C) sites were calculated as a function of free
Ca2+ in the presence of 1 mM Mg2+. The same
parameter values as those in A were used for
calculation. See text for details.
|
|
We performed systematic [3H]ryanodine binding
experiments with - and -RyRs purified from frog skeletal muscle
in the presence of various concentrations of Ca2+
and Mg2+ in an isotonic medium that simulated the
characteristics of the sarcoplasm as far as possible and determined the
parameters according to the three-step procedure with the aid of
computer simulation. The [3H]ryanodine binding
activity in an isotonic salt solution is too low to be precisely
analyzed in the absence of any stimulator other than
Ca2+ (Murayama and Ogawa, 1996). Here, AMP was
used to increase the activity without a significant change in
Ca2+ dependence. The adenine nucleotide has
another advantage in that it shows very weak affinity for
Ca2+ and Mg2+. Another
important precaution is to keep the ionic strength of the medium
constant, because the activity in the presence of high concentrations
of Ca2+ and Mg2+ may
otherwise be changed. After these considerations, the parameters for
Ca2+ and Mg2+ of the two
Ca2+ sites on individual isoforms can be obtained
under the same conditions. To know the parameters for the
Ca2+ release channel in situ that maintain the
organization of RyR and related proteins, furthermore, we measured the
CICR activity, using frog skinned skeletal muscle fibers under
conditions as close as possible to those of the physiological
environment. In analyzing these results, we took advantage of the
determinations of nA,Mg = nI,Ca = nI,Mg = 1 in
[3H]ryanodine binding experiments, because more
laborious maneuvers of CICR experiments impeded such numerous
determinations, as in [3H]ryanodine binding.
All of the results obtained are summarized in Table 2 for
[3H]ryanodine binding experiments and in Table
3 for CICR from SR in frog skinned skeletal muscle fibers.
Comparison between Ca2+ release channel in SR and
purified RyR
Table 2 shows that - and -RyRs are very similar in the
values for all parameters in the presence of 4 mM AMP. This means that
there was only a minor difference at most in the
Ca2+-dependent
[3H]ryanodine binding and its modulation by
Mg2+ between the two isoforms. Frog skeletal
muscles express the two isoforms in almost equal amounts (Murayama and
Ogawa, 1992; 1994). This indicates that both - and -RyRs
contribute to the CICR activity in frog skinned skeletal muscle fibers,
although we cannot exclude the possibility that either isoform might be
silent in situ.
Comparison of Tables 2 and 3 tells us that purified RyR shows lower
affinities for Ca2+ and
Mg2+ than does the Ca2+
release channel in SR: KA,Ca (~10
µM versus ~3 µM), KA,Mg (~300 µM versus ~70 µM), KI,Ca and
KI,Mg (~2 mM versus ~0.4 mM).
However, KA,Mg/KA,Ca
(= 10-30) and
KI,Mg/KI,Ca
(= 1) remained constant. This means that the selectivity between
Ca2+ and Mg2+ in each of
the A- and I-sites was unchanged, although their affinities for
divalent cations were varied. Addition of CHAPS to SR vesicles and the
following process of purification altered the
Ca2+ sensitivity of both A- and I-sites (Ogawa et
al., 1999). Changes in the environment around RyRs (e.g.,
phospholipids, detergent, associating protein(s), and so on) may affect
properties of the RyR molecule or Ca2+ release
channels. We would like to point out that troponin C in the thin
filament shows a 10-fold higher affinity for Ca2+
than purified troponin C (Ebashi and Ogawa, 1988).
Effect of adenine nucleotides on the activity of RyR
Although ATP is thought to be an endogenous activator of RyR in
situ, it has not been known whether MgATP is as potent as free ATP in
stimulating RyR. The comparison of divalent cation dependencies in the
presence of AMP and AMPPCP (Fig. 5) suggests that free AMPPCP,
CaAMPPCP, and MgAMPPCP are equipotent in stimulating CICR. Because
AMPPCP and ATP are reported to be very similar in this stimulation
(
TOP
ABSTRACT
INTRODUCTION
![f<SUB><UP>A</UP></SUB>=[<UP>Ca<SUP>2+</SUP></UP>]<SUP><UP>n</UP><SUB><UP>A,Ca</UP></SUB></SUP>/{[<UP>Ca<SUP>2+</SUP></UP>]<SUP><UP>n</UP><SUB><UP>A,Ca</UP></SUB></SUP>+K<SUP><UP>n</UP><SUB><UP>A,Ca</UP></SUB></SUP><SUB><UP>A,Ca</UP></SUB>(1+[<UP>Mg<SUP>2+</SUP></UP>]<SUP><UP>n</UP><SUB><UP>A,Mg</UP></SUB></SUP>/K<SUP><UP>n</UP><SUB><UP>A,Mg</UP></SUB></SUP><SUB><UP>A,Mg</UP></SUB>)}](/content/vol78/issue4/fulltext/1810/img001.gif)
![1−f<SUB><UP>I</UP></SUB>=1/(1+[<UP>Ca<SUP>2+</SUP></UP>]<SUP><UP>n</UP><SUB><UP>I,Ca</UP></SUB></SUP>/K<SUP><UP>n</UP><SUB><UP>I,Ca</UP></SUB></SUP><SUB><UP>I,Ca</UP></SUB>+[<UP>Mg<SUP>2+</SUP></UP>]<SUP><UP>n</UP><SUB><UP>I,Mg</UP></SUB></SUP>/K<SUP><UP>n</UP><SUB><UP>I,Mg</UP></SUB></SUP><SUB><UP>I,Mg</UP></SUB>)](/content/vol78/issue4/fulltext/1810/img002.gif)
![=B<SUB><UP>max</UP></SUB>×<FR><NU>[<UP>Ca<SUP>2+</SUP></UP>]<SUP><UP>n</UP><SUB><UP>A,Ca</UP></SUB></SUP></NU><DE>[<UP>Ca<SUP>2+</SUP></UP>]<SUP><UP>n</UP><SUB><UP>A,Ca</UP></SUB></SUP>+K<SUP><UP>n</UP><SUB><UP>A,Ca</UP></SUB></SUP><SUB><UP>A,Ca</UP></SUB>(1+[<UP>Mg<SUP>2+</SUP></UP>]<SUP><UP>n</UP><SUB><UP>A,Mg</UP></SUB></SUP>/K<SUP><UP>n</UP><SUB><UP>A,Mg</UP></SUB></SUP><SUB><UP>A,Mg</UP></SUB>)</DE></FR>](/content/vol78/issue4/fulltext/1810/img004.gif)
![×{1/(1+[<UP>Ca<SUP>2+</SUP></UP>]<SUP><UP>n</UP><SUB><UP>I,Ca</UP></SUB></SUP>/K<SUP><UP>n</UP><SUB><UP>I,Ca</UP></SUB></SUP><SUB><UP>I,Ca</UP></SUB>+[<UP>Mg<SUP>2+</SUP></UP>]<SUP><UP>n</UP><SUB><UP>I,Mg</UP></SUB></SUP>/K<SUP><UP>n</UP><SUB><UP>I,Mg</UP></SUB></SUP><SUB><UP>I,Mg</UP></SUB>)}](/content/vol78/issue4/fulltext/1810/img005.gif)
![=B<SUB><UP>max</UP></SUB>×{[<UP>Ca<SUP>2+</SUP></UP>]<SUP><UP>n</UP><SUB><UP>A,Ca</UP></SUB></SUP>/([<UP>Ca<SUP>2+</SUP></UP>]<SUP><UP>n</UP><SUB><UP>A,Ca</UP></SUB></SUP>+K<SUP><UP>n</UP><SUB><UP>A,Ca</UP></SUB></SUP><SUB><UP>A,Ca</UP></SUB>)}](/content/vol78/issue4/fulltext/1810/img007.gif)
![×{1−[<UP>Ca<SUP>2+</SUP></UP>]<SUP><UP>n</UP><SUB><UP>I,Ca</UP></SUB></SUP>/([<UP>Ca<SUP>2+</SUP></UP>]<SUP><UP>n</UP><SUB><UP>I,Ca</UP></SUB></SUP>+K<SUP><UP>n</UP><SUB><UP>I,Ca</UP></SUB></SUP><SUB><UP>I,Ca</UP></SUB>)}](/content/vol78/issue4/fulltext/1810/img008.gif)
![=B<SUB><UP>max</UP></SUB>×{1/(1+[<UP>Ca<SUP>2+</SUP></UP>]<SUP><UP>n</UP><SUB><UP>I,Ca</UP></SUB></SUP>/K<SUP><UP>n</UP><SUB><UP>I,Ca</UP></SUB></SUP><SUB><UP>I,Ca</UP></SUB>+[<UP>Mg<SUP>2+</SUP></UP>]<SUP><UP>n</UP><SUB><UP>I,Mg</UP></SUB></SUP>/K<SUP><UP>n</UP><SUB><UP>I,Mg</UP></SUB></SUP><SUB><UP>I,Mg</UP></SUB>)}](/content/vol78/issue4/fulltext/1810/img010.gif)
