Vol. 281, Issue 4, H1532-H1544, October 2001
Phosphorylation-dependent modulation of cardiac calcium
current by intracellular free magnesium
Siegried
Pelzer,
Chicuong
La, and
Dieter J.
Pelzer
Department of Physiology and Biophysics, Faculty of Medicine,
Dalhousie University, Halifax, Nova Scotia, Canada B3H 4H7
 |
ABSTRACT |
We compared the effects of
cytosolic free magnesium (Mg
) on L-type
Ca2+ current (ICa,L) in
patch-clamped guinea pig ventricular cardiomyocytes under basal
conditions, after inhibition of protein phosphorylation, and after
stimulation of cAMP-mediated phosphorylation. Basal ICa,L density displayed a bimodal dependence on
the concentration of Mg
([Mg2+]i;
10
6-102 M), which changed
significantly as cell dialysis progressed due to a pronounced and
long-lasting rundown of ICa,L in
low-Mg2+ dialysates. Ten minutes after patch breakthrough,
ICa,L density (at +10 mV) in
Mg-depleted cells ([Mg2+]i
~1 µM) was elevated, increased to a maximum at ~20 µM
[Mg2+]i, and declined steeply at higher
[Mg2+]i. Treatment with the broad-spectrum
protein kinase inhibitor K252a (10 µM) reduced
ICa,L density and abolished these effects of
Mg except for a negative shift of
ICa,L-voltage relations with increasing
[Mg2+]i. Maximal stimulation of cAMP-mediated
phosphorylation occluded the Mg-induced stimulation
of ICa,L and prevented inhibitory effects of the
ion at [Mg2+]i <1 mM but not at higher
concentrations. These results show that the modulation of
ICa,L by Mg requires protein
kinase activity and likely originates from interactions of the ion with
proteins involved in the regulation of protein phosphorylation/dephosphorylation. Stimulatory effects of
Mg on ICa,L seem to increase
the cAMP-mediated phosphorylation of Ca2+ channels, whereas
inhibitory effects of Mg appear to curtail and/or
reverse cAMP-mediated phosphorylation.
whole cell patch-clamp recording; guinea pig ventricular
cardiomyocytes; cAMP signaling
| |
INTRODUCTION |
MG2+ is the
major divalent cation present in the cytoplasm and plays a
fundamental role in cell function. Bound to ATP, the ion is required in
reactions involving the transfer of phosphate groups, but there is also
evidence indicating a regulatory role of free intracellular
Mg2+ (Mg). In most vertebrates, the
concentration of Mg
([Mg2+]i) appears to be between 0.3 and 1 mM
in the myocardium (cf. Refs. 3, 16, and
30). Changes of [Mg2+]i within
this concentration range affect a wide array of cardiac functions,
including the activity of L-type Ca2+ channels (2,
30-33), the release of Ca2+ by the sarcoplasmic
reticulum (cf. Refs. 8 and 15), and the Ca2+
sensitivity of the contractile apparatus (5).
In regard to the effects of Mg on cardiac L-type
Ca2+ current (ICa,L), both
inhibition and stimulation have been observed depending on the
experimental conditions. Single channel studies show that
Mg2+ binds to cation-binding sites in the conducting path
of the channel pore exhibiting characteristics of both a weak blocker
and permeator (cf. Ref. 11). A prominent inhibition of
ICa,L by Mg in frog
(29, 31-33) and mammalian (1)
cardiomyocytes persisted after stimulation (1, 29) or
inhibition (31) of cAMP-mediated phosphorylation and was
attributed to a direct interaction of the ion with Ca2+
channel protein. Two reports describe the stimulation of
ICa,L by Mg at concentrations
in the low micromolar range. Yamaoka and Seyama (Ref. 31;
Fig. 5B) observed a stimulation of
ICa,L by [Mg2+]i in
frog cardiomyocytes when the concentration of intracellular calcium
([Ca2+]i) was elevated (1 µM) but not when
[Ca2+]i was 10 nM.
Mg-induced stimulation of
ICa,L in mammalian cardiomyocytes required
the presence of ATP but seemed unrelated to protein phosphorylation
(19).
However, Mg also regulates the activity of several
enzymes that promote or impede cAMP-mediated phosphorylation, one of
the main regulatory events experienced by L-type Ca2+
channels (e.g., Ref. 13). For example, Mg2+ is
essential for the synthesis of cAMP at the catalytic site of adenylyl
cyclase (AC), stimulates the basal activity of cardiac AC (e.g., Refs.
21-23, and 28), and increases the
sensitivity of AC to hormone stimulation (23). On the
other hand, the ion is involved in the degradation of cAMP by
phosphodiesterases (PDEs), which require Mg2+ for activity
(e.g., Ref. 9) and curtails protein phosphorylation via
Mg2+-activated protein phosphatase-2C (PP2C; e.g., Ref.
14).
The objective of the present study was to assess whether the effects of
Mg on phosphorylation/dephosphorylation processes
contribute to the Mg dependence of
ICa,L in mammalian cardiomyocytes. The whole
cell patch-clamp technique was used to measure
ICa,L density and to dialyze cells with
solutions containing various concentrations of free Mg2+
while keeping the concentration of free Ca2+ constant (180 nM) and providing a sufficient supply of ATP-bound Mg2+.
Here, we report that changes in [Mg2+]i
affect ICa,L in a complex manner. Basal
ICa,L density increased when
[Mg2+]i increased from 1 to 17 µM and
decreased at higher [Mg2+]i. The
broad-spectrum protein kinase inhibitor K252a diminished ICa,L density and abolished all regulatory
effects of Mg except for a
Mg-induced negative shift of
ICa,L-voltage relations. Maximal cAMP-mediated
phosphorylation of the Ca2+ channel occluded stimulatory
effects of the ion and diminished the sensitivity of
ICa,L to inhibition by Mg. These results show that the modulation of ICa,L
by Mg2+ requires protein kinase activity. Stimulatory
effects of Mg appear to result from an increase in
the cAMP-mediated phosphorylation of Ca2+ channels;
inhibitory actions of Mg appear to curtail and/or
reverse cAMP-dependent phosphorylation. We found no evidence for
significant effects of Mg that could be
attributed to direct interactions of Mg with
Ca2+ channel protein.
| |
MATERIALS AND METHODS |
In accordance with national and local regulations on animal
experimentation, guinea pigs (300-600 g) of either sex were killed by cervical dislocation. The heart was quickly removed, and the ascending aorta was cannulated. Single ventricular myocytes were isolated by an enzymatic dissociation procedure previously described in
detail (10). Isolated myocytes were stored in
Kraftbrühe ("KB medium") (see Ref. 34 for
composition) at room temperature before experiments. For
experimental recordings, isolated cells were transferred into a
superfusion chamber positioned on top of an inverted microscope stage
(Olympus IMT-2). Once the myocytes had adhered to the glass bottom of
the chamber, they were superfused with control Tyrode solution
containing (in mM) 140 NaCl, 5.4 KCl, 1.8 CaCl2, 1.0 MgCl2, 10 HEPES, and 10 glucose (pH 7.4 with NaOH). After 5 min, the superfusate was changed to K+-free saline (KCl
replaced by CsCl). After gigaseal formation and patch breakthrough,
cells were dialyzed via the patch pipette with a K+-free
solution containing (in mM) 50 CsCl, 110 cesium aspartate, 10 HEPES, 10 1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid
(BAPTA) or 20 BAPTA-20 EGTA, and 4 Na2ATP (pH 7.2 with
CsOH). Free Ca2+ and free Mg2+ were adjusted to
the desired concentrations by adding the appropriate amounts of
MgCl2 and CaCl2 [free Ca2+ and
free Mg2+ were calculated as described by Schoenmakers and
colleagues (27)]. MgATP was provided with the use of ATP
as a high-affinity Mg2+ buffer. To obtain 1 µM free
Mg2+, a total of >30 µM of Mg2+ was
required, most of which is bound to ATP. The concentration of free
Ca2+ was 180 nM or, for experiments with cAMP-loaded
myocytes, 90 nM. No attempts were made to exchange dialysates during
individual experiments.
The voltage clamp was applied with an EPC9 amplifier (Heka;
Lambrecht/Pfalz, Germany; pipette resistance 1.2-3 M
when
immersed in control Tyrode solution) using the whole cell configuration of the patch-clamp technique (7).
ICa,L was elicited by step depolarizations from
80 to +10 mV applied at 0.03 Hz (or 0.1 Hz during the measurement of
ICa,L-voltage relations). Na+
current was minimized by 50-ms prepulses to 40 mV and the presence of
100 µM tetrodotoxin or 0.2 mM
4,4'-diisothiocyanatostilbene-2,2'-disulfonic acid (12) in
the superfusate. Cell capacitance was monitored and updated with each
depolarizing pulse. Current was recorded at a bandwidth of 2-10
kHz using Pulse (version 8.21, Instrutech; Elmont, NY) and analyzed
using PulseFit (version 8.21, Instrutech) on an IBM PC. All experiments
were performed at 22 ± 1°C. The experimental data are given as
means ± SE; n is the number of cells. Statistical
significance was assessed with an unpaired two-tailed
t-test.
All biochemicals were reagent to analytic grade from Calbiochem (San
Diego, CA), Research Biochemicals International (Natick, MA), and Sigma
(St. Louis, MO). Forskolin (10 mM stock in DMSO), isobutyl-methyl-xanthine (IBMX; 20 mM stock in DMSO), and K252a (10 mM
stock in DMSO) were added to the superfusate before the experiment.
DMSO at the concentrations used (<0.1%) had no effect on
ICa,L.
| |
RESULTS |
The experimental results are presented under three subheadings
according to the three different experimental conditions used here. We
will start with a description of the Mg dependence
of ICa,L under basal conditions and provide
evidence for the stimulatory and inhibitory effects of the ion. We will then present data showing that most regulatory effects of
Mg are occluded when protein phosphorylation is
inhibited. Finally, we will illustrate that the stimulation of
cAMP-mediated phosphorylation occludes the stimulation of
ICa,L by Mg and decreases the
sensitivity of ICa,L to inhibitory effects of Mg.
Mg
dependence of basal
ICa,L.
Figure 1 shows the recordings of
ICa,L (left) and time diaries of
ICa,L density (right) during cell
dialysis with solution containing five different concentrations of free
Mg2+ ranging from 1 µM to 10 mM.
ICa,L density shortly after patch breakthough
was usually between 5 and 6 pA/pF. With dialysate concentrations of
free Mg2+ ~1 mM, a concentration close to the
[Mg2+]i in isolated guinea pig ventricular
myocytes (3), ICa,L declined with
time. This phenomenon, known as rundown, is commonly observed during
whole cell recordings of ICa,L using the
patch-clamp technique and is likely caused by a combination of many
factors resulting from the change in the intracellular environment
during cell dialysis (cf. Ref. 13). Typically,
ICa,L rundown was most noticeable early in
dialysis. In the myocyte dialyzed with 1.1 mM Mg2+
solution, ICa,L declined by ~25% within 10 min, which accounted for 95% of total rundown during the 30-min
observation period. When higher dialysate concentrations of free
Mg2+ were used to increase
[Mg2+]i, ICa,L
declined faster and to a larger extent. In the example cell dialyzed
with 10 mM Mg2+ solution, ICa,L
decreased by ~40% within 5 min.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 1.
Comparison of L-type Ca2+ current
(ICa,L) in guinea pig ventricular cardiomyocytes
during dialysis with solution containing different concentrations of
cytosolic free Mg2+ (Mg) ranging from 1 µM to 10 mM. Left: sample currents; right: time
diaries of ICa,L density at a test potential of
+10 mV. Time 0 represents the moment of patch breakthrough
and the start of cell dialysis. Filled symbols (1-5) in
each time course correspond to sample currents.
|
|
ICa,L time diaries were quite different when low
dialysate concentrations of free Mg2+ (
100 µM) were
used to decrease [Mg2+]i. Typically, an
increase in ICa,L density for 5-8 min after patch breakthrough was followed by a long period of rundown before the
current finally stabilized at 
20 min. The increase in
ICa,L was most pronounced with solution
containing ~20 µM Mg2+. In the myocytes shown,
ICa,L density recorded after 5 min of dialysis
with 1, 17, and 82 µM Mg2+ solution exceeded that in 1.1 mM Mg2+ dialysate by 2.7, 3.2, and 1.6 times, respectively
(compare sample currents 2). With 1 and 82 µM
Mg2+ solution, the initial increase in
ICa,L was completely occluded by the following
rundown. In contrast, ICa,L with 17 µM
dialysate Mg2+ stabilized after >20 min at a significantly
elevated level (compare sample currents 4 and 5,
for a summary see Fig. 3).
We next examined the possible effects of Mg on the
voltage dependence of ICa,L. Mg2+ is
a weak blocker of the Ca2+ channel (11) and,
being the major divalent cation present in the cytoplasm, might also
affect the potential drop across the membrane by shielding negative
charges fixed at the cytoplasmic side of the membrane. Figure
2 shows example currents recorded at
10, +10, and +30 mV (left) and corresponding
ICa,L-voltage relations (right) at
five different [Mg2+]i ranging from 1 µM
(Fig. 2A) to 10 mM (Fig. 2E). The data were collected after extensive cell dialysis (28 min), when
ICa,L had reached a steady state (see Fig. 1).
Typical bell-shaped ICa,L-voltage relations were
observed at all [Mg2+]i. However, at
concentrations >100 µM, increasing [Mg2+]i
seemingly caused a progressive negative shift of the potential eliciting maximal inward current (Vmax; arrows
in Fig. 2, right) from Vmax = 13 ± 1 mV with 82 µM [Mg2+]i
(n = 5; Fig. 2C) to
Vmax = 1 ± 1 mV with 10 mM
[Mg2+]i (n = 4; Fig.
2E). ICa,L density at
Vmax was similar with 1 µM, 82 µM, and 1.1 mM [Mg2+]i (compare Fig. 2, A,
C, and D) and somewhat (~25%) smaller with 10 mM [Mg2+]i (Fig. 2E). With 17 µM
[Mg2+]i (Fig. 2B),
ICa,L was considerably larger in size, and the
ICa,L-voltage relation peaked at a less positive
potential than with 1 µM [Mg2+]i (Fig.
2A) or 82 µM [Mg2+]i (Fig.
2C). This negative shift appears to be unrelated to the leftward shift of Vmax at higher
[Mg2+]i and seems to be associated with the
mechanism causing the elevation of ICa,L (see
below).
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 2.
Voltage dependence of ICa,L with different
concentrations of Mg
([Mg2+]i) ranging from 1 µM to 10 mM [1
µM (A), 17 µM (B), 82 µM (C),
1.1 mM (D), and 10 mM (E)]. Left:
original currents recorded at test potentials (Vm) of 10 mV
( ), +10 mV ( ), and +30 mV
( ) after >28 min of cell dialysis. Right:
corresponding complete ICa,L-voltage relations.
Each point represents the average ICa,L density
measured in 3-6 cells. Where error bars are absent, they are
smaller than the symbol size. Arrows indicate the average potential
eliciting maximal inward current (Vmax) obtained
from Spline approximations of the ICa,L-voltage
relations (Origin 4.0, Microcal Software). Vmax
was 16 ± 1 mV (A), 6 ± 1 mV (B),
13 ± 1 mV (C), 8 ± 1 mV (D), and
1 ± 1 mV (E).
|
|
The data summarized in Fig. 3 illustrate
the Mg dependence of basal
ICa,L. All ICa,L
densities shown are average values measured from 5-13 myocytes.
Figure 3A depicts changes in the relation between
ICa,L at +10 mV and
[Mg2+]i during cell dialysis. Shown are the
ICa,L densities measured after 10 min of cell
dialysis, a time when we expect dialysis with the low-molecular-weight
compounds contained in our pipette solution to be complete, and
ICa,L densities after extensive cell dialysis
(28 min), when ICa,L had reached a steady
state at all [Mg2+]i (see Fig. 1). At
millimolar [Mg2+]i,
ICa,L densities were similar at both times and
decreased with increasing [Mg2+]i; at
submillimolar [Mg2+]i, cell dialysis altered
the Mg dependence of ICa,L
significantly. Ten minutes after patch breakthrough, ICa,L density with 1 µM
[Mg2+]i was elevated (1.8 times larger than
with 1 mM [Mg2+]i), increased to a maximum at
17 µM [Mg2+]i, and decreased steeply at
higher [Mg2+]i, suggesting that
Mg exerts both stimulatory and inhibitory actions on
ICa,L. After 28 min of dialysis,
ICa,L densities in 1 µM and 1 mM
[Mg2+]i were similar, and bimodal changes of
ICa,L density occurred only in a small
concentration window around 20 µM [Mg2+]i.
Apparently, much of the inhibitory action of Mg had
been masked by the decline of ICa,L at low
[Mg2+]i with progressing cell dialysis (see
Fig. 1). The stimulatory effect of Mg appeared to be
little altered by cell dialysis. After 10 min of dialysis,
ICa,L density in 17 µM
[Mg2+]i exceeded that in 1 µM
[Mg2+]i by 3.0 ± 1.0 pA/pF
(n = 13 for ICa,L in 17 µM
[Mg2+]i and n = 7 for
ICa,L in 1 µM
[Mg2+]i); after 28 min of dialysis, the
elevation in ICa,L density was of similar size,
3.8 ± 0.7 pA/pF (n = 6 for
ICa,L in 17 µM [Mg2+]i and n = 3 for
ICa,L in 1 µM
[Mg2+]i).
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 3.
Concentration dependence of ICa,L
on [Mg2+]i after different times of cell
dialysis (A) and at different test potentials
(B). A: ICa,L density at a
test potential of +10 mV measured after 10 min () and
30 min of cell dialysis ( ). Points represent the mean
of n = 5-13 cells. ICa,L
densities measured with GTP (200 µM)-containing dialysates after 10 min ( ) and 30 min ( ) of cell dialysis
are also shown. Points represent the mean of n = 3 cells. B: ICa,L density after 30 min
of cell dialysis at test potentials of 10, +30, and +50 mV (filled
symbols) and in GTP-containing dialysates (open symbols). Each point
represents the mean of n = 3-6 cells. Where error
bars are absent, they are smaller than the symbol size.
|
|
Both stimulatory and inhibitory effects of Mg were
also apparent at other test potentials (see Fig. 3B). The
stimulatory action of Mg appeared to be
reduced with increasing test potentials. A comparison of the
ICa,L densities measured in 1 and 17 µM
[Mg2+]i shows that
ICa,L density at 10 mV increased by 3.7-fold
compared with 2.3-, 1.6-, 1.5-, and 1.4-fold at 0, +10, +30 and +50 mV (compare Fig. 2, A and B), respectively. Also
noticeable, particularly at millimolar
[Mg2+]i, was the tendency of
ICa,L to decrease with increasing
[Mg2+]i at positive potentials but to
increase at 10 mV, which is in keeping with the
Mg-induced leftward shift of
ICa,L-voltage relations (see Fig. 2).
It is worth noting that these experiments were conducted in GTP-free
dialysates to discourage the activation of G proteins, which are
Mg2+ dependent and affect the activity of cardiac
Ca2+ channels in a complex fashion (cf. Ref.
13). ICa,L densities in
GTP-containing dialysates were similar at 1 mM
[Mg2+]i but were significantly lower than in
GTP-free dialysates at 17 µM [Mg2+]i (open
symbols in Fig. 3, A and B).
These data show that Mg affects basal
ICa,L in a complex manner, which comprises both
stimulatory and inhibitory mechanisms. Stimulatory effects of
Mg prevail at low (<20 µM)
[Mg2+]i. They appear to be little altered by
cell dialysis and enhanced at negative potentials. Inhibitory effects
of Mg become predominant at
[Mg2+]i >20 µM. However, they are masked
by the rundown of ICa,L as cell dialysis
progresses. Furthermore, the regulatory effects of
Mg appear to depend on the cellular concentration of GTP.
Mg dependence of
ICa,L after protein kinase inhibition.
Possible targets for interactions with Mg2+ are the
Ca2+ channel itself as well as several systems involved in
the regulation of protein phosphorylation. To separate
phosphorylation-related effects from direct effects of
Mg on ICa,L, we used the
nonhydrolyzable ATP analog K252a. This compound abolishes the enzymatic
activity of a wide array of kinases including protein kinase A (PKA),
Ca2+/calmodulin-dependent protein kinase II, and protein
kinase C, all of which contribute to the regulation of cardiac
ICa,L (e.g., Ref. 13). Figure
4 illustrates sample
ICa,L records (left) and time diaries
of ICa,L density (right) recorded
from K252a-treated myocytes. Regardless of the concentration of free
Mg2+ in the dialysate, which ranged from 1 µM (Fig.
4A) to 5 mM (Fig. 4D),
ICa,L density at a test potential of +10 mV
immediately after patch breakthrough was considerably lower than under
basal conditions (around 3 pA/pF) and changed little during cell
dialysis.
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 4.
Comparison of ICa,L in
K252a-treated myocytes during dialysis with solution containing 1 µM
(A), 17 µM (B), 1 mM (C), and 5 mM
(D) Mg. Illustrated are sample currents
(left) and time diaries of ICa,L
density (right) recorded from myocytes that were
preincubated for 10-20 min and subsequently superfused with 10 µM K252a solution. The test potential was +10 mV. Time 0 represents the moment of patch breakthrough and the start of cell
dialysis. Filled symbols (1-4) in each time course
correspond to the sample currents shown.
|
|
The voltage dependence of ICa,L in K252a-treated
cells (Fig. 5) was determined under
steady-state conditions after 28 min of cell dialysis. Again, all
ICa,L-voltage relations had the typical bell
shape, and increasing [Mg2+]i from 1 µM
(Fig. 5A) to 5 mM (Fig. 5D) caused a progressive negative shift of Vmax (arrows in Fig. 5,
right). With 17 µM [Mg2+]i,
Vmax was similar to that observed with 1 µM
[Mg2+]i (compare Fig. 5, A and
B) but significantly more positive (P = 0.0001) than under basal conditions (compare Figs. 2B and
5B). K252a seemingly abolished the leftward shift in the
voltage dependence of basal ICa,L at 17 µM
[Mg2+]i, suggesting that this leftward shift
requires protein kinase activity.
View larger version (18K):
[in this window]
[in a new window]
|
Fig. 5.
Voltage dependence of ICa,L in K252a-treated
myocytes with different [Mg2+]i ranging from
1 µM to 5 mM [1 µM (A), 17 µM (B), 1 mM
(C), and 5 mM (D)]. Left: original
currents recorded at test potentials (Vm) of 10 mV
(), +10 mV (), and +30 mV
() after 30 min of cell dialysis. Right:
corresponding complete ICa,L-voltage relations.
Each point represents the average ICa,L density
measured in 3-6 cells. Where error bars are absent, they are
smaller than the symbol size. Arrows indicate
Vmax, which was 18 ± 2 mV (A),
17 ± 1 mV (B), 11 ± 1 mV (C), and
6 ± 1 mV (D).
|
|
The data shown in Fig. 6 summarize the
Mg-dependence of ICa,L in
K252a-treated myocytes. ICa,L densities at +10
mV were similar after 10 min and after 28 min of cell dialysis (Fig.
6A) and were unaffected by changes in
[Mg2+]i between 1 µM and 5 mM.
ICa,L at +30 and +50 mV (Fig. 6B)
decreased with increasing [Mg2+]i; at 10
mV, ICa,L increased at higher
[Mg2+]i. These changes are explicable as a
result of the Mg-induced negative shift of the
voltage dependence of ICa,L (see Fig. 5).
View larger version (12K):
[in this window]
[in a new window]
|
Fig. 6.
Concentration dependence of ICa,L
on [Mg2+]i in K252a-treated myocytes measured
after different times of cell dialysis (A) and at different
test potentials (B). A:
ICa,L density measured at a test potential of
+10 mV after 10 min () and 30 min of cell dialysis
(). Points represent the mean of n = 4-6 cells. B: ICa,L density
after 30 min of cell dialysis at test potentials of 10, +30, and +50
mV. Each point represents the mean of n = 3-6
cells. Where error bars are absent, they are smaller than the symbol
size.
|
|
These data show that the inhibition of protein phosphorylation with
K252a reduced ICa,L density and rendered the
remaining current unresponsive to the modulatory effects of
Mg. The only significant effect of
Mg on the permeation of Ca2+ through
unphosphorylated Ca2+ channels is a negative shift of
ICa,L-voltage relations, which could be
explained by the screening of intracellular negative surface charges by
Mg.
It is also worth noting that there was no detectable
ICa,L rundown in K252a-treated cells (see Fig.
4). Hence, one has to assume that the rundown of
ICa,L under basal conditions (see Fig. 1)
results from a decline in the activity of phosphorylated
Ca2+ channels contributing to basal
ICa,L and/or that the processes causing rundown
require protein kinase activity.
Effects of Mg2+ on ICa,L
in cAMP-loaded myocytes.
We have previously shown (35) that bath application of the
AC activator forskolin together with the PDE inhibitor IBMX elevates cAMP sufficiently to maximally stimulate the cAMP-mediated
phosphorylation of Ca2+ channels. Figure
7 shows typical
ICa,L records (left) and time diaries
of ICa,L density (right) recorded
with different concentrations of dialysate Mg2+. Shortly
after patch breakthrough, ICa,L density at +10
mV was usually in the range of 25 to 35 pA/pF. With 1.1 mM
Mg2+ dialysate (see Fig. 7D), typically an
increase in ICa,L during the first 5-7 min
of dialysis was followed by rundown that lasted for the entire 30-min
observation period. Considering that the predialysis
[Mg2+]i in guinea pig ventricular myocytes is
~1 mM (3), cell dialysis is unlikely to cause a
significant change in [Mg2+]i. Hence, one
must assume that the initial increase in ICa,L is not related to a change in [Mg2+]i. It is
more likely that the initial increase in ICa,L
represents relief from Ca2+-induced inhibition of
ICa,L due to the reduction of
[Ca2+]i from a possibly elevated predialysis
level to 90 nM, the dialysate concentration of Ca2+ used in
all experiments with cAMP-loaded cells. Similar
ICa,L time diaries were seen when submillimolar
concentrations of free Mg2+ were used to decrease
[Mg2+]i (Figs. 7, A-C). When
[Mg2+]i was increased to concentrations in
the millimolar range, ICa,L declined throughout
the entire observation period but most noticeably during early dialysis
(Fig. 7E). Note that, although Mg2+ is an
essential cofactor for the phosphate transfer by PKA, the reduction of
free dialysate Mg2+ to concentrations as low as 1 µM did
not seem to hamper the cAMP-mediated phosphorylation of
Ca2+ channels.
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 7.
Comparison of ICa,L in cAMP-loaded
myocytes during dialysis with 1 µM (A), 17 µM
(B), 82 µM (C), 1.1 mM (D) and 12.8 mM (E) Mg2+ solution. Sample currents
(left) and time diaries of ICa,L
density (right) were recorded at a test potential of +10 mV
from myocytes preincubated for 10-20 min and superfused with
solution containing 10 µM forskolin and 50 µM
isobutyl-methyl-xanthine. Time 0 represents the moment of
patch breakthrough and the start of cell dialysis. Filled symbols
(1-5) in each time course correspond to the sample
currents.
|
|
The voltage dependence of ICa,L in cAMP-loaded
myocytes is illustrated in Fig. 8. Sample
records of ICa,L at different membrane potentials (left) and ICa,L-voltage
relations (right) were measured after 30 min of dialysis
with [Mg2+]i ranging from 1 µM (Fig.
8A) to 12.8 mM (Fig. 8E). All
ICa,L-voltage relations were bell shaped, with
Vmax (indicated by arrows) clustering between
4 and +1 mV. At all [Mg2+]i,
Vmax was significantly less positive than under
basal conditions (see Fig. 2) or in K252a-treated cells (see Fig. 5).
This confirms observations by others (cf. Ref. 13) that
cAMP-mediated phosphorylation shifts the voltage dependence of the
Ca2+ channel to more negative potentials.
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 8.
Voltage dependence of cAMP-upregulated ICa,L
at different [Mg2+]i ranging from 1 µM to
12.8 mM [1 µM (A), 17 µM (B), 82 µM
(C), 1.1 mM (D), and 12.8 mM (E)].
Left: original currents recorded at test potentials (Vm) of
10 mV (), +10 mV (), and +30 mV
() after 30 min of cell dialysis. Right:
corresponding complete ICa,L-voltage relations.
Each point represents the average ICa,L density
measured in 2-5 cells. Where error bars are absent, they are
smaller than the symbol size. Arrows indicate
Vmax, which was 3 ± 1 mV (A),
3 ± 1 mV (B), 4 ± 1 mV (C),
2 ± 2 mV (D), and 1 ± 1 mV (E).
|
|
Figure 9 summarizes the dependence of
ICa,L on [Mg2+]i in
cAMP-loaded cells. The
ICa,L-[Mg2+]i
relations shown in Fig. 9A were determined at +10 mV after 10 and 30 min of cell dialysis. At both times, cAMP-upregulated ICa,L was unresponsive to variations of
[Mg2+]i between 1 µM and 1 mM; increasing
[Mg2+]i in the millimolar range caused a
large decrease in ICa,L density. ICa,L at 10 mV displayed a very similar
dependence on [Mg2+]i (see Fig.
9B). Although at more positive test potentials a decline of
ICa,L density at submillimolar
[Mg2+]i was noticeable, the stimulation of
cAMP-mediated phosphorylation appears to confer considerable protection
against the inhibitory effects of Mg on
ICa,L seen under basal conditions (see Fig. 3).
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 9.
[Mg2+]i-dependence of
cAMP-upregulated ICa,L measured after different
times of cell dialysis (A) and at different test potentials
(B). A: ICa,L density
measured at a test potential of +10 mV after 10 min ()
and 30 min of cell dialysis (). Points represent the
mean of n = 4-6 cells. B:
ICa,L density after 30 min of cell dialysis at
test potentials of 10, +30, and +50 mV. Each point represents the
mean of n = 2-6 cells. Where error bars are
absent, they are smaller than the symbol size.
|
|
These data show that the Mg dependence of
ICa,L under basal conditions and in cAMP-loaded
cells differs considerably in several respects. In cAMP-loaded cells, the stimulatory effects of Mg were occluded, whereas the inhibitory effects of
Mg occurred only at millimolar
concentrations of the ion (compare Figs. 3 and 9). Other than under
basal conditions or in K252a-treated cells, in cAMP-loaded cells there
was no clear effect of [Mg2+]i on the voltage
dependence of ICa,L (compare Figs. 2, 5, and 7).
Furthermore, in cAMP-loaded cells, ICa,L
declined even after prolonged cell dialysis. This is in keeping with
the observation that the rundown of single channel current in excised
patches is significantly slowed down in the presence of compounds that promote the cAMP-mediated phosphorylation of the Ca2+
channel (cf. Ref. 13). Hence, one would expect that after
the stimulation of cAMP-mediated phosphorylation, the rundown of
ICa,L is prolonged.
| |
DISCUSSION |
The experiments described in this study confirm the importance of
Mg in the regulation of Ca2+ influx via
L-type Ca2+ channels in mammalian cardiomyocytes. Evidence
is provided that Mg has a dual stimulatory and
inhibitory action on ICa,L. Additional
experimental data demonstrate that the inhibition of protein
phosphorylation prevents both stimulatory and inhibitory effects of
Mg on ICa,L. Finally, it is
shown that the stimulation of cAMP-mediated phosphorylation occludes
the stimulation of ICa,L by
Mg and diminishes the sensitivity of
ICa,L to inhibition by Mg. These results strongly suggest that the modulation of
ICa,L by Mg originates
from effects of Mg on enzymes that regulate protein
phosphorylation and, in particular, the cAMP-mediated phosphorylation
of the Ca2+ channel.
Inhibitory effects of
Mg.
In the absence of neuronal or hormonal stimuli, the main regulatory
influence of Mg on cardiac L-type Ca2+
channels seems to be inhibitory (see Fig. 3). Our data possibly underestimate the full extent to which Mg curtails
basal ICa,L in intact cardiomyocytes, because
the inhibitory effects of Mg target an increment of
ICa,L that declines during cell dialysis (see
Figs. 1 and 3A). For example, after 10 min of cell dialysis,
a time when the initial increase of ICa,L in
low-Mg2+ dialysates had already been overcome by subsequent
long-lasting depression, ICa,L in 1 µM
[Mg2+]i was still about two times larger than
at 1 mM [Mg2+]i. Only 5 min earlier (see Fig.
1), ICa,L in 1 µM
[Mg2+]i had exceeded that at 1 mM
[Mg2+]i by about three times.
In intact frog cardiomyocytes, submillimolar concentrations of
Mg affected the open probability but
not the conductance of the Ca2+ channel
(31, 33). Channel gating in
Mg2+-depleted cardiomyocytes strongly resembled that of
phoshorylated Ca2+ channels, and an increase in
[Mg2+]i progressively shifted gating from a
"willing" into a "reluctant" mode. Two mechanisms that have
been associated with such a mode shift are Ca2+-induced
inhibition and dephosphorylation of Ca2+ channels (e.g.,
Ref. 13). Because the effect of Mg persisted in the presence of a "death brew" (31),
which prevents phosphorylation by depleting the cell of ATP, the
inhibitory action of Mg appeared to be unrelated to phosphorylation processes and was attributed to the binding of Mg2+ to the regulatory Ca2+-binding site, which
is involved in the Ca2+-induced inhibition of
ICa,L (31). We found no evidence
for a significant contribution of this mechanism to the inhibitory effects of Mg on ICa,L. In
our hands, the inhibitory effects of the ion were occluded by the
inhibition of protein phosphorylation with K252a (see Figs. 4-6);
in contrast, free Ca2+-induced inhibition of
ICa,L persisted in the presence of K252a (34). The recent finding that Ca2+-induced
inhibition of ICa,L likely results from
Ca2+ binding to a channel-associated calmodulin
(36) rather than to a regulatory site at the
Ca2+ channel protein also argues against the hypothesis
that Mg could cause inhibition of the
Ca2+ channel in a Ca2+-analog manner. Although
the ion binds to calmodulin, Mg2+ binding does not appear
to induce the conformational changes required for the interaction of
calmodulin with target peptides (cf. Ref. 17).
An alternative explanation for the inhibitory effects of
Mg is that in Mg-depleted cells,
a considerable number of Ca2+ channels are phosphorylated.
Judging from the differences in ICa,L density
under basal conditions and in the presence of K252a (compare filled
circles in Figs. 3A and 6A), phosphorylated
channels appear to carry >75% of basal ICa,L
in Mg2+-depleted cells but 40% or less at millimolar
[Mg2+]i. With increasing
[Mg2+]i, the contribution of current through
phosphorylated Ca2+channels to whole cell
ICa,L declined either because
Mg selectively inhibits current flowing through
phosphorylated Ca2+ channels or because
Mg reduces the number of phosphorylated channels
that contribute to ICa,L. Our results argue
against an inhibitory action of Mg on current flow
through phosphorylated Ca2+ channels, because, in this
case, the inhibitory effect of Mg2+ should become more
pronounced when Ca2+ channel phosphorylation is stimulated
and a larger number of phosphorylated Ca2+ channels
contributes to whole cell ICa,L. On the
contrary, the stimulation of cAMP-mediated phosphorylation decreased
the sensitivity of ICa,L to inhibition by
Mg (compare Figs. 3A and 9A).
This suggests that Mg activates enzymes that curtail
and/or reverse the phosphate transfer by PKA, thereby reducing the
number of phosphorylated Ca2+ channels contributing to
ICa,L. Although the nature of these enzymes
remains to be determined, there are several likely candidates. For
example, all PDEs are Mg2+ dependent, (e.g., Ref.
9). Because some cAMP-specific PDEs [such as PDE3 (e.g.,
Ref. 18) or PDE4 (20)] have affinities for
Mg2+ in the low micromolar concentration range, the
inhibitory effects of Mg on basal
ICa,L could reflect the
Mg2+-activation of PDEs. Alternatively, the activation of
the Mg2+-dependent PP2C (cf. Ref. 14) could
curtail Ca2+ channel phosphorylation. However, PP2C
activity, at least in cell free preparations, requires millimolar
concentrations of Mg2+ (4), and one would
expect this mechanism to contribute to the reduction of
ICa,L at higher
[Mg2+]i. Indeed, the inhibition of
ICa,L in frog cardiomyocytes by Mg at concentrations between 0.3 and 3 mM appeared
to result either from a direct action of Mg on the
Ca2+ channel or from channel dephosphorylation due to
increasing phosphatase activity (30).
Stimulatory effects of
Mg.
At very low [Mg2+]i (<20 µM), the
predominant action of Mg on
ICa,L was stimulatory (see Fig. 3), suggesting
that the underlying mechanism(s) have an even higher affinity for
Mg than the inhibitory mechanism(s), which prevail
at higher [Mg2+]i. Considering that the
increase in [Mg2+]i from 1 to 17 µM was
accompanied by a similar (~3-4 pA) increase in
ICa,L density after 10 and 30 min of cell
dialysis (see Fig. 3A), the sensitivity of Ca2+
channels to stimulation by Mg appears to be
preserved during cell dialysis and unrelated to the overall density of
ICa,L, which declined considerably.
The stimulatory action of Mg was abolished in cells
treated with the broad-spectrum protein kinase inhibitor K252a (see
Fig. 6). This strongly suggests that the stimulation of
ICa,L requires protein kinase activity and
likely results from interactions of Mg with proteins
involved in phosphorylation/dephosphorylation. The occlusion of
Mg-induced stimulation after maximal stimulation of
cAMP-mediated phosphorylation (see Fig. 9) indicates that stimulatory
low Mg increases the cAMP-mediated phosphorylation
of Ca2+ channels. In line with this hypothesis is that
stimulation by Mg was associated with a leftward shift in the voltage dependence of ICa,L at 17 µM (see Fig. 2), which reflects that the stimulatory action of
Mg was potential dependent and most pronounced at
negative membrane potentials (see Fig. 3B), where
cAMP-mediated phosphorylation has the largest impact on the open
probability of Ca2+ channels (e.g., Ref. 13).
Also in agreement with this hypothesis is that the
Mg- induced stimulation of
ICa,L was reversed by
Mg-dependent inhibitory mechanisms, which
appear to curtail or reverse the cAMP-mediated phosphorylation of the
Ca2+ channel (see above). The nature of the
Mg-dependent proteins involved in the stimulation of
ICa,L remains to be determined. One possible
target for Mg is AC. Mg2+ is required
for the synthesis of cAMP at the catalytic site of the enzyme and
stimulates the basal enzymatic activity of cardiac AC in cell-free
preparations (21, 23, 28); however, these processes occur
at much higher concentrations than the concentrations of
Mg that caused a sizable stimulation of
ICa,L. For example, Steinberg et al.
(28) reported an apparent activation constant for
Mg of 1.5 mM. Hence, one has to assume that either
the affinity for Mg2+ of AC is considerably reduced in
cell-free preparations or that another Mg2+-dependent
process increases the cAMP-mediated stimulation of Ca2+
channels. The possibility that AC activity is stimulated by a Mg2+-induced activation of Gs protein is
unlikely because the activation of G proteins in GTP-containing
dialysates decreased ICa,L density at 17 µM
[Mg2+]i (see Fig. 3).
Modulation of basal Ca2+ channel
activity by cAMP and Mg
.
Considering that ICa,L densities under basal
conditions were noticeably higher than in K252a-treated myocytes
(compare Figs. 3 and 6), one has to assume that under basal conditions,
phosphorylation processes maintain Ca2+ channel activity
and Ca2+ entry at an elevated level. An important
consequence of this arrangement is that it enables heart cells not only
to increase but also to decrease Ca2+ entry when receiving
appropriate extrinsic signals. The gain heart cells provide for
extrinsic and, in particular, for inhibitory signals appears to be
strongly dependent on [Mg2+]i. For example,
in Mg2+-depleted cells (1-10 µM
[Mg2+]i), extrinsic signals that reduce the
activity of cardiac kinases and/or promote dephosphorylation can
diminish Ca2+ entry up to threefold (compare Figs.
3A and 6A). At 17 µM
[Mg2+]i, the gain for inhibitory signals is
even higher due to the prevailing stimulatory effects of
Mg2+ but decreases at higher
[Mg2+]i, where the inhibitory effects of the
ion dominate so that at 1 mM [Mg2+]i,
inhibitory extrinsic signals can reduce ICa,L by
not more than ~50%. Our results suggest that these effects of
Mg2+ reflect changes in the cAMP-mediated phosphorylation
of Ca2+ channels (see above) and that they originate from
interactions of Mg2+ with proteins involved in cAMP
signaling rather than from direct interactions of the ion with the
Ca2+ channel protein. Important questions that arise from
these observations are whether heart cells can indeed tune their
sensitivity to regulation by sympathetic and parasympathetic stimuli
and whether they utilize Mg2+ for this purpose.
Although the Mg2+-induced changes in
ICa,L density can be reasonably well explained
with Mg2+-induced changes in the activity of AC, PDEs, and
possibly PP2C, one has to keep in mind that G proteins, which regulate
the activity of most of these enzymes, are also Mg2+
dependent. In the GTP-free dialysates we used in this study, G protein
activity is likely discouraged. In GTP-containing dialysates, ICa,L density was similar at 1 mM
[Mg2+]i but significantly reduced at 17 µM
[Mg2+]i (see Fig. 3). Considering that at low
[Mg2+]i the presence of GTP facilitates
mainly the activation of Gi and Go proteins,
which have a higher affinity for Mg2+ than Gs
(6), a possible explanation for the reduction of
ICa,L at low [Mg2+]i
is that Gi-mediated inhibition of AC reduces
Ca2+ channel phosphorylation. Indeed, Gi
protein activation with acetylcholine caused a similar reduction of
ICa,L density (data not shown).
Effects of Mg on
unphosphorylated Ca2+ channels.
Several actions of Mg2+ are expected to affect the current
through unphosphorylated Ca2+ channels. Thorough single
channel studies (cf. Ref. 11) show that Mg2+
binds to cation-binding sites in the conducting pore of the
Ca2+ channel, exhibiting characteristics of both a weak
blocker and a weak permeator. We found the currents through
unphosphorylated Ca2+ channels to be remarkably
insensitive to the regulatory effects of
Mg. When the phosphorylation-related effects of Mg were suppressed, the most significant effect of Mg was a
progressive leftward shift of the
ICa,L-voltage relation with increasing
[Mg2+]i with minimal alteration of the shape
of the ICa,L-voltage relation (see Fig. 5). This
argues against significant direct channel block by
Mg, which should reduce ICa,L
in a potential-dependent manner and, under the experimental conditions used here, most prominently at positive membrane potentials. A possible
explanation for the effects of Mg on
unphosphorylated Ca2+ channels is that
Mg screens intracellular negative membrane charges,
thereby reducing the voltage drop across the membrane that is
experienced by the Ca2+ channel. This would also explain
why Mg caused a similar leftward shift of
Vmax at [Mg2+]i >100
µM under basal conditions (see Fig. 2) but not in cAMP-loaded cells
(see Fig. 8), where Ca2+, entering the cells in much higher
amounts (compare Figs. 3 and 9), can be expected to occupy a large
fraction of the negative sites at the intracellular side of the
membrane, thus either preventing the binding of Mg2+ or
rendering additional binding of Mg2+ inconsequential.
In conclusion, changes in [Mg2+]i affect
cardiac ICa,L in a complex fashion, comprising
both stimulatory and inhibitory effects. Our results strongly suggest
that these effects originate from interactions of Mg2+ with
enzymes involved in the regulation of protein
phosphorylation/dephosphorylation rather than from direct interactions
of Mg2+ with the Ca2+ channel protein.
The stimulatory effects of Mg on
ICa,L appear to increase the cAMP-mediated
phosphorylation of the Ca2+ channel. The inhibitory effects
of Mg seem to curtail the cAMP-mediated
phosphorylation of the Ca2+ channel, possibly by activating
PDEs and/or protein phosphatases.
An important question is the physiological relevance of the regulatory
effects of submillimolar concentration of free Mg on
cardiac ICa,L. There is considerable evidence
supporting the hypothesis that [Mg2+]i may
change under physiological conditions via Mg2+ transport
across the sarcolemma and/or via intracellular redistribution (e.g.,
Ref. 16). For example, 
-agonists and other
interventions that increase cAMP cause a large Mg2+ efflux
from cells, whereas the activation of protein kinase C induces
Mg2+ uptake in the same order of magnitude (e.g., Refs.
25 and 26). Although it is presently unclear to which
extent these fluxes change the cytoplasmic concentration of free
Mg2+, it can be assumed that they cause considerable
changes in the concentration of free Mg2+ in the
diffusion-restricted submembrane space, where the enzymes controlling
cAMP signaling to the Ca2+ channel appear to be located in
microdomains within macromolecular distances of the Ca2+
channel pore (cf. Ref. 35). Our results suggest that the
activity of these enzymes enhances Ca2+ channel activity
and Ca2+ entry under basal condition and that small changes
in the concentration of Mg alter the activity of
some of these enzymes sufficiently to cause profound changes in
ICa,L. This suggests that Mg2+ may
be the third player in a signaling network that interrelates the
systems regulating the concentrations of Ca2+ and cAMP in
heart cells.
| |
ACKNOWLEDGEMENTS |
We thank Darren Cole for technical assistance and Brian Hoyt for
computer support.
| |
FOOTNOTES |
This work was supported by grants from the Heart and Stroke Foundations
of Nova Scotia (to D. J. Pelzer) and New Brunswick (to S. Pelzer).
Address for reprint requests and other correspondence: D. J. Pelzer, Dept. of Physiology and Biophysics, Dalhousie University, Faculty of Medicine, Sir Charles Tupper Medical Bldg., Halifax, Nova
Scotia, Canada B3H 4H7 (E-mail: dpelzer{at}is.dal.ca).
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received 4 August 2000; accepted in final form 21 June 2001.
| |
REFERENCES |
1.
Agus, ZS,
Kelepouris E,
Dukes I,
and
Morad M.
Cytosolic magnesium modulates calcium channel activity in mammalian ventricular cells.
Am J Physiol Cell Physiol
256:
C452-C455,
1989[Abstract/Free Full Text].
2.
Agus, ZS,
and
Morad M.
Modulation of cardiac ion channels by magnesium.
Annu Rev Physiol
53:
299-307,
1991[ISI][Medline].
3.
Buri, A,
Chen S,
Fry CH,
Illner H,
Kickenweiz E,
McGuigan JAS,
Noble D,
Powell T,
and
Twist VW.
The regulation of intracellular Mg2+ in guinea-pig heart, studied with Mg2+-selective microelectrodes and fluorochromes.
Exp Physiol
78:
221-223,
1993[Abstract].
4.
Cohen, P,
Klumpp S,
and
Schelling DL.
An improved procedure for identifying and quantitating protein phosphatases in mammalian tissues.
FEBS Lett
250:
596-600,
1989[ISI][Medline].
5.
Fabiato, A,
and
Fabiato F.
Effects of magnesium on contractile activation of skinned cardiac cells.
J Physiol (Lond)
249:
497-517,
1975[Abstract/Free Full Text].
6.
Gilman, AG.
G proteins: transducers of receptor-generated signals.
Annu Rev Biochem
56:
615-649,
1987[ISI][Medline].
7.
Hamill, OP,
Marty A,
Neher E,
Sakmann B,
and
Sigworth FJ.
Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches.
Pflügers Arch
391:
85-100,
1981[ISI][Medline].
8.
Howarth, FC,
and
Levi JA.
Internal free magnesium modulates the voltage dependence of contraction and Ca transient in rabbit ventricular myocytes.
Pflügers Arch
435:
687-698,
1998[ISI][Medline].
9.
Jost, JP,
and
Rickenberg HV.
Cyclic AMP.
Annu Rev Biochem
40:
741-774,
1971[ISI].
10.
Kaspar, SP,
and
Pelzer DJ.
Modulation by stimulation rate of basal and cAMP-elevated Ca2+ channel current in guinea pig ventricular cardiomyocytes.
J Gen Physiol
106:
175-201,
1995[Abstract/Free Full Text].
11.
Kuo, CC,
and
Hess PJ.
Block of the L-type Ca2+ channel pore by external and internal Mg2+ in rat phaeochromocytoma cells.
J Physiol (Lond)
466:
683-706,
1993[Abstract/Free Full Text].
12.
Liu, J,
Lai ZF,
Wang XD,
Tokutomi N,
and
Nishi K.
Inhibition of sodium current by chloride channel blocker 4,4'-diisothiocyanatostilbene-2,2'-disulfonic acid (DIDS) in guinea pig cardiac ventricular cells.
J C
TOP
ABSTRACT
INTRODUCTION
