Increased Availability and Open Probability of Single
L-Type Calcium Channels From Failing Compared With
Nonfailing Human Ventricle
Frank Schröder, MD; Renate Handrock, PhD; Dirk J. Beuckelmann, MD; Stephan Hirt, MD;
Roger Hullin, MD; Leo Priebe, MD; Robert H.G. Schwinger, MD;
Joachim Weil, MD; Stefan Herzig, MD
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
Background—The role of the L-type calcium channel in human heart failure is unclear, on the basis of previous
whole-cell recordings.
Methods and Results—We investigated the properties of L-type calcium channels in left ventricular myocytes isolated
from nonfailing donor hearts (n516 cells) or failing hearts of transplant recipients with dilated (n59) or ischemic (n57)
cardiomyopathy. The single-channel recording technique was used (70 mmol/L Ba21). Peak average currents were
significantly enhanced in heart failure (38.269.3 fA) versus nonfailing control hearts (13.264.5 fA, P50.02) because
of an elevation of channel availability (55.966.7% versus 26.465.3%, P50.001) and open probability within active
sweeps (7.3661.51% versus 3.1861.33%, P50.04). These differences closely resembled the effects of a cAMPdependent stimulation with 8-Br-cAMP (n511). Kinetic analysis of the slow gating shows that channels from failing
hearts remain available for a longer time, suggesting a defect in the dephosphorylation. Indeed, the phosphatase inhibitor
okadaic acid was unable to stimulate channel activity in myocytes from failing hearts (n55). Expression of calcium
channel subunits was measured by Northern blot analysis. Expression of a1C- and b-subunits was unaltered. Whole-cell
current measurements did not reveal an increase of current density in heart failure.
Conclusions—Individual L-type calcium channels are fundamentally affected in severe human heart failure. This is
probably important for the impairment of cardiac excitation-contraction coupling. (Circulation. 1998;98:969-976.)
Key Words: calcium channels n heart failure n myocytes
C
ardiac excitation-contraction coupling depends on the
function of L-type calcium channels. One may speculate
that calcium channel dysfunction may be involved in the
pathophysiology of heart failure. Numerous studies have
addressed the issue in animal models1–12 or patient material,13–17 mostly by measuring dihydropyridine binding or the
whole-cell calcium current density. The findings are inconsistent, with increases,1,2 decreases,3– 8,13,17 or no change9 –11,14 –16
reported. This may be related to species differences, the
model or severity12 of failure, or the assay used.2 Importantly,
studies on human material revealed a slight reduction in
calcium channel mRNA expression and dihydropyridine
binding sites13 but an unchanged whole-cell current under
both basal15,16 and forskolin-stimulated conditions.15
It is premature to conclude that calcium channel alterations
are irrelevant to human heart failure. The whole-cell current
I is a function of both the number of functional channels N
and their individual properties i (single-channel current amplitude), the open probability (popen, fraction of time spent in
the open state during active sweeps), and the availability
(factive, fraction of active sweeps per number of test pulses),
where I5N3i3popen3factive. Therefore, any incongruence between N and I could in theory be accounted for by alterations
of i, popen, or factive. Because the latter 2 parameters are known
to be modulated physiologically by cAMP-dependent phosphorylation,18 –22 it should be interesting to measure them
under the conditions of heart failure. We recently demonstrated23 that single L-type channel recording is possible in human
ventricular myocytes. Here, we report that heart failure
markedly increases single-channel current because of increased open probability and availability.
Methods
Preparation of Cardiomyocytes
Ventricular myocytes were prepared from failing or nonfailing
hearts. Failing hearts were obtained from patients with end-stage
heart failure caused by ischemic or dilated cardiomyopathy (IC, DC)
who were undergoing transplantation. Nonfailing hearts were from
organ donors who had died of noncardiac causes, whose hearts could
Received January 29, 1998; revision received April 17, 1998; accepted April 22, 1998.
From the Departments of Pharmacology (F.S., R.H., S. Herzig) and Cardiology (D.J.B., L.P., R.H.G.S.), University of Cologne; the Department of
Cardiothoracic Surgery, University of Kiel (S. Hirt); the Department of Cardiology, Ludwig-Maximilians-University, Munich (R.H.); and the Department
of Pharmacology, University of Hamburg (J.W.), Germany.
Correspondence to Stefan Herzig, MD, Department of Pharmacology, University of Cologne, Gleueler Strabe 24, 50931 Cologne, Germany. E-mail
[email protected]
© 1998 American Heart Association, Inc.
969
970
Single L-Type Ca21 Channels in Heart Failure
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
not be transplanted for various technical reasons. According to the
available patient data, there were no significant differences between
the groups regarding age (failing, 57.561.2 years; nonfailing,
51.564.2 years) or sex ('66% male). The medication of transplant
recipients regularly included diuretics, ACE inhibitors, and digitalis
glycosides. Some of the organ donors received intravenous catecholamines (dopamine, norepinephrine) until surgery.
The tissue was placed into ice-cold cardioplegic solution and
shipped to the laboratory within 18 hours. It was placed in preoxygenated solution A (4°C) composed of (mmol/L) NaCl 100, KCl 10,
MgSO4 5, dextrose 20, taurine 50, and MOPS 5 (pH 7.4). After
removal of fat and connective tissue, slices '23230.5 mm were cut
from the free left ventricle and enzymatically digested23 in 10 mL of
solution A containing collagenase (1.5 mg/mL, type CLS 1, Worthington Biochemical Corp), trypsin (1 mg/mL, type III, Sigma
Chemical Co), and BSA (10 mg/mL, Sigma) at 37°C for 40 minutes.
A second incubation (30 to 90 minutes, depending on the cell yield
checked at 10-minute intervals) in the presence of collagenase (0.5
mg/mL) and BSA (1 mg/mL) followed. After gravity settling (in
solution A, 15 minutes), cells were placed in solution B containing
(mmol/L) potassium glutamate 50, KCl 40, KH2PO4 20, taurine 20,
KOH 20, MgCl2 3, HEPES 10, EGTA 5, dextrose 10 (pH 7.4, 22°C),
disaggregated, and stored at 4°C ($1 hour) before use. Then, the cell
suspension was incubated (30 to 180 minutes, 22°C) with 10 mmol/L
BAPTA-acetoxymethyl ester to buffer intracellular divalent cations.
Electrophysiological Measurement
Cells were placed in disposable perfusion chambers (3 mL) containing a bath solution of (mmol/L) NaCl 135, KCl 4, MgCl2 1, HEPES
10, CaCl2 2, dextrose 10 (pH 7.4 with NaOH, 21°C to 23°C). Pipettes
(borosilicate glass, 7 to 10 MV) were filled with (mmol/L) BaCl2 70,
sucrose 110, HEPES 10 (pH 7.4 with TEA-OH). Single calcium
channels were recorded in the cell-attached configuration of the
patch-clamp technique. Barium currents were elicited by depolarizing test pulses of 150 ms at 1.66 Hz (see References 21, 23, and 24),
recorded at 10 kHz, and filtered at 2 kHz (23 dB, 4-pole Bessel)
with an Axopatch 200 A amplifier (Axon Instruments). Command
pulses were 120 mV in amplitude (eg, from 2100 to 120 mV or
from 240 to 180 mV, depending on the resting potential of the cell),
with absolute values adjusted to yield single-channel amplitudes
of ' 20.7 nA. This corresponds to a test potential of 120 mV across
the patch membrane, where channel availability is maximal (see
Reference 23, Figure 2). Only the experiments without a shift in
single-channel current amplitude (gauged by amplitude histograms)
were evaluated. PClamp software (version 6.0, Axon Instruments)
was used for acquisition and analysis. 8-Br-cAMP (from Sigma, 0.1
mol/L stock in DMSO) and okadaic acid (NH41 salt, from Calbiochem, 0.1 mmol/L stock in DMSO) were added to the bath as a
30-mL bolus. The final drug concentrations depended on the exact
amount of the bath volume, determined after the experiment. The
final concentrations amounted to 0.8460.04 mmol/L (from 0.6 to
1.1 mmol/L) 8-Br-cAMP and 0.8660.07 mmol/L (0.7 to 1.1 mmol/L)
okadaic acid.
Data Analysis
Experiments were analyzed whenever the channel activity persisted
for at least 72 seconds (120 sweeps) both under control conditions
and after exposure to drug. Linear leak and capacity currents were
digitally subtracted. The availability (fraction of sweeps containing
at least 1 channel opening), the open probability (popen, defined as the
relative occupancy of the open state during active sweeps), and the
peak ensemble average current (ipeak, obtained after optical or
mathematical smoothing) were analyzed from single-channel and
multichannel patches. In the latter case, they were corrected for n, the
number of channels in the patch. n was the maximum current
amplitude observed divided by the unitary current. Peak current was
corrected by division through n. The availability was corrected by
the square root method: (12availabilitycorrected) is the nth root of
(12availabilityuncorrected). The corrected popen was calculated on the
basis of the corrected number of active sweeps, ie, total open time
divided by (n3availabilitycorrected3number of test pulses). Openings
and closures were identified by the half-height criterion. Closed-time
and first-latency analyses were carried out in 1-channel patches only.
First latency was determined by averaging the waiting times between
the beginning of the test pulse and the first opening (if present).
Open-time and closed-time histograms were fitted with a maximumlikelihood estimate (PStat software, Axon Instruments) of log-binned
data. Slow gating was analyzed in experiments (with only 1 channel
in the patch) that contained at least 300 sweeps. The sweep
histograms and probability plots were fitted by least-squares methods. Two-tailed t tests were used for statistical comparisons, with
either the unpaired or paired format as appropriate. Values are given
as mean6SEM.
Whole-Cell Experiments
Cells were isolated as described.16 The bath solution was (mmol/L)
choline chloride 130, HEPES 25, dextrose 22, 4-aminopyridine 4,
CaCl2 2, and MgCl2 1.1, pH 7.4 (with TEA-OH). Peak inward
calcium currents were measured (similar hardware to that for single
channels) at steady state, with 200-ms steps applied every 2 seconds
from a holding potential of 280 to 110 mV (peak of current-voltage
relation). The recording pipette contained (mmol/L) CsCl 140,
HEPES 25, and fura-2 0.05, pH 7.2 (with TEA-OH). Current density
was calculated by dividing peak current through cell capacitance.
Northern Blots
Preparation of poly(A) mRNA and quantification of transcripts for
the calcium channel a1C- and b-subunits by Northern blot analysis
were carried out as previously described.25 Samples were taken from
the left ventricle of the same hearts from which electrophysiological
data were obtained. Calsequestrin expression was used for normalization of the RNA yield, because transcription of this gene is
unaltered in heart failure.13 a1C-mRNA was detected by hybridization
with a 448-base cRNA complementary to a region of the human
cardiac a1C, which includes the IV S6 transmembrane segment.
b-Subunit mRNA was identified by hybridization with a 411-base
cRNA coding for a central core region of the human b-subunit.
Cardiac calsequestrin expression was quantified by hybridization
with a 190-base cRNA coding for the carboxy terminus of calsequestrin. Hybridization reactions for all transcripts were done subsequently on the same gels at 42°C.
Results
Single-channel activity of L-type calcium channels is markedly enhanced in failing myocardium compared with nonfailing controls. This is illustrated by original recordings from 2
experiments (Figure 1) and by the corresponding open-time
and closed-time histograms (Figure 2). The increase in
ensemble average current (Figure 1, bottom traces) is due
both to an increased availability and to an increased open
probability. The latter effect is caused predominantly by
shorter closed times, as seen in the histogram analysis (Figure
2). These findings were statistically significant (Figure 3) and
independent of the cause of heart failure. Table 1 presents the
details. It shows that the higher open probability of channels
from failing hearts is due to 3 reasons: a shorter first latency,
a longer mean open time, and a shorter closed time (faster
time constant of the slow component). The unitary current
amplitude i is similar between the 2 groups, which is trivial,
because we adjusted our pulse protocol according to this
parameter (see Methods). Importantly, single-channel conductance, obtained by measuring the amplitudes of fully
resolved openings at 2 different test potentials, is identical
between channels from nonfailing (16.763.2 pS, n56) and
failing (16.861.7 pS, n511) myocardium. This value also
matches our previous findings (16.661.2 pS, n56) in chan-
Schröder et al
September 8, 1998
971
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
Figure 2. Open time (top) and closed time (bottom) histograms
of 2 experiments depicted in Figure 1 (left, channel from nonfailing heart; right, channel from failing heart). Curves were generated with a maximum-likelihood estimate for simple (open times)
or double exponential (closed times). Time constants amounted
to topen50.64 ms, tclosed,fast50.51 ms, and tclosed,slow525.8 ms for
channel from nonfailing heart and topen50.54 ms, tclosed,fast50.34
ms, and tclosed, slow511.4 ms for channel from failing
myocardium.
Figure 1. Comparison of single L-type calcium channels from
nonfailing organ donor heart (left) and terminally failing heart
(right). Top row, Pulse protocol (150-ms steps, amplitude 120
mV, applied every 600 ms). Applied voltages were from 270 to
150 mV (left) and from 2100 to 120 mV (right). Middle, 20 consecutive sweeps for each channel. Bottom rows, Average current of all 240 (left) or 300 (right) sweeps of entire ensembles.
Scale bars520 ms and 2 pA (unitary current traces) or 17.5 fA
(ensemble averages).
nels from failing myocardium (see Reference 23 for detailed
discussion of absolute value) in a potassium-rich solution, for
which the membrane potential is exactly known.
The above-mentioned profile of the enhanced activity of
channels from failing ventricles resembles the pattern of
cAMP-dependent stimulation of cardiac L-type channels
known from animal experiments.18 –22 It was therefore of
interest to compare this profile with the effects of a cAMP
analogue. Our first preliminary attempts to modulate human
cardiac L-type channels by 8-Br-cAMP were fruitless: in
none of 5 technically successful bath applications of the drug
did current increase under depolarizing conditions (see References 20 and 23); therefore, the present study used a
physiological bath solution. This change in condition allowed
us to obtain a stimulation by 8-Br-cAMP, as shown by the
time course of such an experiment (Figure 4). Here, both the
availability (density of bars) and the open probability (bar
height) were markedly elevated in a channel from a nonfailing heart. This is also seen in Figure 5, which shows
representative traces from the same experiment and the
average currents from the entire ensembles. When the results
from all technically successful drug applications were combined (n53 patches from 3 nonfailing hearts, n59 patches
from 8 failing hearts), the peak current was elevated (from
39.6612.9 to 63.8619.4 fA, P,0.05) because of both
effects, but only the increase in open probability (from
7.8462.01% to 9.7262.56%) and not the availability (from
43.868.7% to 57.367.5%) reached statistical significance in
a 2-tailed paired analysis. This was rather unexpected, given
the higher sensitivity22 and the lower variability (eg, see
Figure 3. Statistical analysis of channel behavior
regarding peak current ipeak, open probability
popen, and availability of channels recorded from
cells isolated from nonfailing donor hearts (open
bars, n516 channels from 9 hearts) or failing
hearts (hatched bars, all: n516 experiments from
10 failing hearts, composed of n59 experiments
from 5 hearts with DC and n57 experiments
from 5 hearts with IC). Comparison between
nonfailing and all failing hearts (first 2 bars in
each graph) was done by 2-tailed t test
(*P,0.05). To check for possible differences
depending on cause of failure, ANOVA was performed on nonfailing, DC, and IC data (*P,0.05
by 1-way ANOVA). In post hoc tests, significant
results in ANOVA were due to difference between nonfailing and IC groups (P,0.05 after Bonferroni correction, not depicted).
There were no significant differences between DC and IC.
972
Single L-Type Ca21 Channels in Heart Failure
TABLE 1. Comparison of Single-Channel Properties of L-Type
Channels From Nonfailing and Failing Human
Ventricular Myocytes
Parameter
Nonfailing
n
Failing
n
P
ipeak, fA
13.264.5
16
38.269.3
16
0.02
Popen, %
3.1861.33
16
7.3661.51
16
0.04
Availability, %
26.465.3
16
55.966.7
16
0.001
Open time, ms
0.5460.05
16
0.7260.07
16
0.04
topen, ms
0.5060.05
16
0.6960.07
16
0.04
First latency, ms
57.365.2
12
33.364.8
11
0.001
Closed time, ms
9.3361.64
12
6.4361.18
11
tclosed,fast, ms
0.5060.08
10
0.3860.03
11
tclosed,slow, ms
21.863.6
10
13.761.9
11
Fraction tclosed,fast
0.5560.08
10
0.5560.04
11
20.6960.04
16
20.7060.03
16
Amplitude i, nA
0.03
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
For closed times and latency analysis, only patches containing one single
channel were used. The other values were corrected for the number of
channels present. Nonfailing myocytes were from n59 different donor hearts,
and failing myocytes were from n510 failing (n59/5 for DC, n57/5 hearts for
IC) transplant recipient hearts. P values (indicated if ,0.05; all other values
were .0.1) were calculated from an unpaired two-tailed t test.
Reference 24) of phosphorylation-dependent effects on availability compared with open probability in animal experiments. Inspection of the data from all individual experiments
(Figure 6) and separate analysis of results from nonfailing
versus failing tissue (Table 2), however, give a clue for this
phenomenon. Whereas the experiments with cells from nonfailing hearts show a strong increase in current (from
14.1612.4 to 40.9638.6 fA, n53, P5NS), channels from
failing myocardium start off at a very high baseline (at
48.0616.2 fA, n59), as expected (see Table 1), and peak
current can be raised only slightly, to 71.4624.2 fA (P50.07,
2-sided paired t test). In fact, the availability is sometimes
close to its theoretical maximum, and it is clear that these
channels cannot show any further increase after 8-Br-cAMP.
In summary, the same picture emerges when channel activity
from nonfailing versus failing hearts on the one hand and the
effect of cAMP-dependent phosphorylation on the other hand
are compared: the first latency is shorter, the open time
higher, the closed times lower, the open probability higher,
and the peak average current higher both in cells from failing
hearts and after 8-Br-cAMP. This raises the idea that a
Figure 4. Effect of 8-Br-cAMP (0.75 mmol/L) on activity of single channel in ventricular myocyte from nonfailing myocardium.
Open probability (popen) increases after drug addition.
Figure 5. Single consecutive traces from experiment in Figure
4, illustrating mechanism of cAMP-dependent stimulation. Left,
Before 8-Br-cAMP; right, after 8-Br-cAMP. Pulse protocol (top
row) consisted of 150-ms steps from 2100 to 120 mV throughout experiment. Ensemble averages (bottom rows) were calculated from all 540 sweeps before (left) and 600 sweeps after
drug addition. Scale bars520 ms and 2 pA (unitary current
traces) or 46 fA (ensemble averages).
phosphorylation-related mechanism is responsible for the
elevated activity of channels from failing heart.
To examine this idea further, we analyzed the kinetic
properties of the slow gating process, ie, the movement of
channels between an “available” and an “unavailable” state,
as evidenced by the nonrandom occurrence (clustering) of
active and blank sweeps.19,21,23 First, the lifetime of the
available state was estimated by sweep histogram analysis, in
which the distribution of “runs” (ie, series of continuously
active sweeps or continuously blank sweeps) is plotted
against time. For the duration of active runs, we obtained the
rate constants of monoexponential fits (not shown) from the
long-lived recordings with 1 channel in the patch. This value
was significantly (P50.006) decreased in channels from
failing myocardium (0.56260.108 s21, n57) compared with
nonfailing myocardium (1.75760.328 s21, n58). Blank run
duration was biexponential in 4 of 8 cells from nonfailing
tissue, consistent with previous animal data.21 Channels from
failing myocardium, in contrast, revealed a monophasic blank
run distribution. This is also seen in the probability plots
averaged from all these experiments (Figure 7). With the
plain sweep histogram analysis used so far, there are 2
methodological limitations. First, silent transitions between
states occurring during the test pulse interval are missed and
Schröder et al
September 8, 1998
973
Figure 6. Effect of 8-Br-cAMP in all technically successful experiments. Channels
from failing hearts (F) are characterized by
a higher control value before drug and a
smaller drug effect than channels from nonfailing hearts (E). This is especially true for
channel availability.
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
may cause numerical errors in the lifetime estimate. Second,
the existence of a short-lived blank state (obviously present in
nonfailing and presumably also in failing myocardium) will
cause a systematic error (underestimation) of the true lifetime
of the “phosphorylated” available state (see Reference 24).
To circumvent these problems, we used a discrete-time
Markov model.21 This model consists of an available state, A,
linked to 2 nonavailable states, L and S (for long- and
short-lived), and these transitions are taken to be mediated by
a phosphorylation- and a non–phosphorylation-linked event,
respectively. The resulting system S7A7L is thus described
by 4 rate constants. All data sets in Figure 7 were fitted
simultaneously. To reduce the degree of freedom for the fit,
we assumed identical S7A transitions for nonfailing and
failing myocardium (it would be impossible to determine
these rates in the failing data alone). The curves shown in
Figure 7 correspond to the best-fit rate constants, in which
exit from the phosphorylated state (A3 L) occurs at 0.137 s21
in channels from failing myocardium and at 0.798 s21 in
channels from nonfailing myocardium. The reverse phosphorylation reaction L3 A is not too different between the 2
groups (failing, 0.318 s21; nonfailing, 0.269 s21). For both
data sets, the fast rate constants A3 S and S3 A were 0.633
s21 and 2.08 s21, respectively.
This finding suggests that the enhanced availability, indicative of a higher steady-state level of phosphorylation in
channels from failing myocardium, is primarily due to an
impaired dephosphorylation reaction, with little or no change
in the phosphorylation rate constant. Because channel availability is controlled primarily by a type 1 protein phosphatase
in animals21,24 and because this phosphatase is itself controlled
by cAMP-dependent phosphorylation,21,26 we investigated
whether this regulation is still present in channels from failing
myocardium. Interestingly, the active run durations remained
unaffected by 8-Br-cAMP (rate constants changed from
0.61860.168 s21 to 0.65260.215 s21, P5NS) in those 4
single-channel experiments in cells from failing hearts in
which the availability was elevated by the drug. Kinetically,
the stimulation was due to a decrease of the blank run lifetime
(because the rate constants tended to increase from
0.73560.273 s21 to 1.05560.212 s21, P50.08, n54), suggesting a pure effect on the phosphorylation reaction. In
contrast, the 2 single-channel patches from nonfailing myocardium revealed the known21 dual response, ie, a .2.5-fold
change of both constants, namely, a decrease in the blank run
lifetime and an increase in the active run lifetime. These
findings suggest that a type 1 protein phosphatase, which
normally controls L-type calcium channel availability, is
downregulated in heart failure.
To further test the role of protein phosphatases for channel
activity, we applied the membrane-permeant phosphatase inhibitor okadaic acid to n55 patches (1 single-channel, 4 multichannel) from failing myocardium. No significant stimulatory effect
was found for peak current (from 35.3614.6 to 40.8619.3 fA),
open probability (from 8.565.1% to 9.265.3%), or availability
(from 45.868.9% to 47.5610.5%). This is in striking contrast to
the profound stimulation we found previously in guinea pig
myocardium under identical conditions24 and to the effect of
okadaic acid on a single channel recorded from a nonfailing
heart (peak current was raised from 4.5 to 16.5 fA, open
probability from 0.8 to 3.1%, and availability from 32% to 41%).
These findings further support the idea that a downregulation of
channel dephosphorylation is the reason why channels from
failing myocardium are more active.
To examine the idea of whether the number of functional
channels is also affected by heart failure, we subjected tissue
samples of the same hearts as studied at the single-channel
level to a whole-cell study and to a Northern blot analysis. In
whole-cell recordings, the cell capacitance (nonfailing,
196626 pF, n54 cells from 3 hearts; failing, 192626 pF,
n56 cells from 4 hearts) and the current density (nonfailing,
5.161.5 pA/pF; failing, 2.460.6 pA/pF) were not significantly altered. Expression of mRNA for the calcium channel
subunits a1C and b was measured in 18 sufficiently large
tissue specimens from 13 hearts (5 nonfailing and 8 failing),
and successful RNA preparation and subsequent hybridization reactions with the probes for a1C- and b-subunits took
place in 13 cases (4 specimens from 4 different nonfailing
hearts and 9 specimens from 7 different failing hearts). There
were only subtle, insignificant changes of expression of the
a1C- and b-subunits (Figure 8). However, there was a large
scatter, especially for the values from nonfailing hearts. The
ratio of b-subunit mRNA over a1C-subunit mRNA was
significantly reduced in heart failure (nonfailing, 8.1063.20
compared with 3.1660.59 in failing hearts), possibly indicating an altered subunit composition. In summary, these data
indicate that the profound changes seen at the single-channel
level are not reflected in similar alterations of overall current
density or in clear reciprocal changes of mRNA expression.
Discussion
The main finding of our study is a dramatically increased
activity of L-type calcium channels in human heart failure. At
974
Single L-Type Ca21 Channels in Heart Failure
TABLE 2.
Effect of 8-Br-cAMP on Single-Channel Behavior
Parameter and
Group
Control
8-Br-cAMP
n
P
Open time, ms
NF
0.6360.13
0.6860.08
F
0.7860.12
0.8460.09
3
9
NF1F
0.7460.09
0.8060.09
12
NF
0.6060.15
0.6660.08
3
F
0.7460.12
0.7460.09
9
NF1F
0.7160.10
0.7260.07
12
topen, ms
First latency, ms
NF
55.3
36.5
F
31.767.3
27.768.1
2
6
NF1F
37.668.7
29.967.4
8
6.20
4.86
2
0.05
Closed time, ms
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
NF
F
5.8862.00
5.2761.96
6
0.03
NF1F
5.9661.45
5.1661.44
8
0.005
tclosed,fast, ms
NF
0.42
0.36
1
F
0.3560.05
0.2960.03
6
NF1F
0.3660.04
0.3160.03
7
0.09
tclosed,slow, ms
NF
10.5
8.2
1
F
12.563.3
12.163.3
6
NF1F
12.262.7
11.562.8
7
0.64
1
Fraction tclosed,fast
NF
0.51
F
0.5960.04
0.5960.07
6
NF1F
0.5860.04
0.6060.05
7
NF
20.7560.06
20.7460.05
3
F
20.7260.04
20.7260.04
9
NF1F
20.7360.03
20.7360.03
12
i amplitude, nA
Figure 7. Analysis of slow gating of channels from nonfailing
(left) and failing (right) hearts. Only those experiments with 1 single channel and a long recording time ($300 sweeps) were
used (n58 for nonfailing, n57 for failing). Length of runs of consecutive active (top) or blank (bottom) sweeps was counted;
probability of active or blank to be at least x seconds long was
calculated and then averaged. Calculated curves were generated with a discrete-time Markov model, as explained in detail
in text. Data show that duration of active state is prolonged in
channels from failing hearts, whereas overall duration of blank
states is shortened.
the number of functional channels, N, must be largely
reduced in heart failure, ie, by a factor of 2- or 3-fold.
Previous Northern blot experiments as well as some dihydropyridine binding studies indeed showed a reduced channel
expression in human heart failure.13 We were unable to
confirm a reduced expression at least of a1C-subunit mRNA,
possibly because of the small number of samples and the
large scatter. In future studies, the number of functional
Eight single-channel patches and 4 double-channel patches were recorded
from cells of 11 hearts (NF1F), 3 nonfailing (NF), and 8 failing (F; 4 with DC,
4 with IC). P values (given for P,0.1) indicate significance of the effect of
8-Br-cAMP (two-sided paired t test).
first glance, this effect is in apparent conflict with data from
the literature suggesting unchanged or depressed calcium
currents, at least in the majority of studies. It must be
remembered, however, that none of the previous investigations, in animals or humans, have addressed the question at
the single-channel level. As delineated in the introduction, the
whole-cell current I is defined as I 5 N3i3popen3factive. An
unchanged value for I may reflect no change at all or any sort
of reciprocal alterations in the other terms. We find no
evidence for a change in the permeation properties, which
determine the size of the single-channel amplitude. However,
both popen and factive are increased. We were able to confirm the
previous finding of unaltered15,16 or perhaps slightly reduced17
whole-cell current density I in heart failure. This means that
Figure 8. Northern blot analysis of human cardiac tissue samples. Left, Radiolabeled antisense probes detected transcripts of
8-kb size for a1C-subunit, 5.6-kb for b-subunit, and 2.6-kb for
calsequestrin in nonfailing and failing myocardium. Each lane
contained 10 mg of RNA. Exposure time for autoradiography
was 120 hours (a1C), 72 hours (b), and 16 hours (calsequestrin).
Right, Hybridization signals were analyzed densitometrically and
normalized for calsequestrin signal of respective sample. No
significant changes were seen.
Schröder et al
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
channels might be addressed by simultaneous measurements
of whole-cell and single-channel currents.
Considering the mechanism of the elevated activity of
channels from failing hearts, the similarity with the effects of
a cAMP derivative suggests a phosphorylation-dependent
mechanism. Although chronic effects of the standard medical
treatment with diuretics, ACE inhibitors, and digitalis cannot
be ruled out in this study, we do not believe that acute
b-adrenergic effects arising in vivo from the noradrenergic
tone in heart failure were conserved in our experiments. The
tissue and cells had been washed 12 times, the time between
explantation and our measurements ranged from 4 to .24
hours, and cells from nonfailing hearts of donors treated with
intravenous catecholamines did not show such increased
activity. However, it seems counterintuitive to attribute the
elevated single-channel activity to an increased baseline of
phosphorylation, given the presumed deficit in PKAmediated protein phosphorylation in failing human myocardium.27,28 One may speculate that the channel behavior
indicates, instead of increased phosphorylation state, an
uncoupling from (dephosphorylated) inhibitory subunits.
Both the a1C-19 and the b-subunits30 are target proteins for
cAMP-dependent phosphorylation of cardiac calcium channels, and we found a reduced relative abundance of b-subunit
mRNA. Alternatively, an increased phosphorylation state of
the channel may also result from a decrease in the dephosphorylation rate. Our data provide a kinetic picture at the
single-channel level. Here, the results on slow gating are
entirely compatible with an increased channel phosphorylation state despite unaltered kinase activity: the prolonged
dwell time of the available state probably reflects a reduction
in dephosphorylation rate, and a reduced type 1 protein
phosphatase activity could be responsible. Accordingly, inhibition of protein phosphatases by okadaic acid was ineffective in channels from failing myocardium. Interestingly,
protein phosphatase activity in tissue samples from failing
human heart is increased in sarcoplasmic reticulum membrane preparations31 but not in homogenates from whole
tissue,32 suggesting an altered subcellular distribution of
phosphatases.
The alterations in the rapid gating parameters also deserve
consideration: an increased open probability is the singlechannel manifestation of voltage- or frequency-induced potentiation (facilitation) of the channel.24,33 Our findings may
thus be reflected at the whole-cell level by the altered
high-frequency–induced facilitation of calcium current recently reported for failing human heart.34 If pertinent to
physiological conditions, this type of change in singlechannel gating, when associated with altered deactivation
properties,20 may cause proarrhythmic effects by early
afterdepolarizations.35
Important questions remain unresolved at present. What is
the activity of calcium channels located in the T tubules and
closely coupled to ryanodine receptors and EC coupling? In
the cell-attached mode, only superficial channels are seen,
and channels in T tubules may behave differently. This would
explain the results of Gomez and coworkers,11 who showed an
impaired coupling between calcium currents and calcium
release sparks from the sarcoplasmic reticulum in a rat model.
September 8, 1998
975
Single-channel data under their conditions would be interesting. It is also unclear whether the increase in single-channel
activity is a primary or a secondary event. Sarcoplasmic
reticulum proteins are dysfunctional in heart failure (see
Reference 36), and reciprocal regulation of calcium current
by expression of ryanodine receptors has been found.37 It is
therefore feasible that the increased activity of superficial
channels just compensates for a reduced channel expression
in the T tubules. A relative lack of b-subunit could contribute
to such an altered distribution of channels.38 Alternatively, a
reduced calcium channel expression could also be primary,
because of the increased noradrenergic tone in heart failure,
as shown in cell culture.39 The increased activity of remaining
channels would then be a secondary compensatory
phenomenon.
In summary, our findings emphasize an important role of
the L-type calcium channel in the pathophysiology of human
cardiac failure. Its markedly increased single-channel activity
indicates that it is not an innocent bystander in heart failure.
Acknowledgments
This study was supported by the Deutsche Forschungsgemeinschaft
(He 1578 6 –1). We gratefully acknowledge Elke Hippauf for
excellent technical help and Ursula Kreuzberg for contributing to
some experiments and analyses.
References
1. Wagner JA, Reynolds IJ, Weisman HF, Dudeck P, Weisfeldt ML, Snyder
SH. Calcium antagonist receptors in cardiomyopathic hamster: selective
increases in heart, muscle, brain. Science. 1986;232:515–518.
2. Keung EC, Toll L, Ellis M, Jensen RA. L-type cardiac calcium channels
in doxorubicin cardiomyopathy in rats: morphological, biochemical, and
functional correlations. J Clin Invest. 1991;87:2108 –2113.
3. Rossner KL. Calcium current in congestive heart failure of hamster
cardiomyopathy. Am J Physiol. 1991;260:H1179 –H1186.
4. Gopalakrishnan M, Triggle DJ, Rutledge A, Kwon YW, Bauer JA, Fung
HL. Regulation of K1 and Ca21 channels in experimental cardiac failure.
Am J Physiol. 1991;261:H1979 –H1987.
5. Gengo PJ, Sabbah HN, Steffen RP, Sharpe JK, Kono T, Stein PD,
Goldstein S. Myocardial beta adrenoceptor and voltage sensitive calcium
channel changes in a canine model of chronic heart failure. J Mol Cell
Cardiol. 1992;24:1361–1369.
6. Ming Z, Nordin C, Siri F, Aronson RS. Reduced calcium current density
in single myocytes isolated from hypertrophied failing guinea pig hearts.
J Mol Cell Cardiol. 1994;26:1133–1143.
7. Colston JT, Kumar P, Chambers JP, Freeman GL. Altered sarcolemmal
calcium channel density and Ca21-pump ATPase activity in tachycardia
heart failure. Cell Calcium. 1994;16:349 –356.
8. Li GR, Ferrier GR, Howlett SE. Calcium currents in ventricular myocytes
of prehypertrophic cardiomyopathic hamsters. Am J Physiol. 1995;268:
H999 –H1005.
9. Sen L, Smith TW. T-type Ca21 channels are abnormal in genetically
determined cardiomyopathic hamster hearts. Circ Res. 1994;75:149 –155.
10. Kääb S, Nuss HB, Chiamvimonvat N, O’Rourke B, Pak PH, Kass DA,
Marban E, Tomaselli GF. Ionic mechanism of action potential prolongation in ventricular myocytes from dogs with pacing-induced heart
failure. Circ Res. 1996;78:262–273.
11. Gomez AM, Valdivia HH, Cheng H, Lederer MR, Santana LF, Cannell
MB, McCune SA, Altschuld RA, Lederer WJ. Defective excitationcontraction coupling in experimental cardiac hypertrophy and heart
failure. Science. 1997;276:800 – 806.
12. Gruver EJ, Glass MG, Marsh JD, Gwathmey JK. An animal model of
dilated cardiomyopathy: characterization of dihydropyridine receptors
and contractile performance. Am J Physiol. 1993;265:H1704 –H1711.
13. Takahashi T, Allen PD, Lacro RV, Marks AR, Dennis AR, Schoen FJ,
Grossmann W, Marsh JD, Izumo S. Expression of dihydropyridine
receptor (Ca21 channel) and calsequestrin genes in the myocardium of
patients with end-stage heart failure. J Clin Invest. 1992;90:927–935.
976
Single L-Type Ca21 Channels in Heart Failure
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
14. Schmidt U, Schwinger RH, Böhm S, Überfuhr P, Kreuzer E, Reichart B,
Meyer L, Erdmann E, Böhm M. Evidence for an interaction of halothane
with the L-type Ca21 channel in human myocardium. Anesthesiology.
1993;79:332–339.
15. Mewes T, Ravens U. L-type calcium currents of human myocytes from
ventricle of non-failing and failing hearts and from atrium. J Mol Cell
Cardiol. 1994;26:1307–1320.
16. Beuckelmann DJ, Näbauer M, Erdmann E. Intracellular calcium handling
in isolated ventricular myocytes from patients with terminal heart failure.
Circulation. 1992;85:1046 –1055.
17. Ouadid H, Albat B, Nargeot J. Calcium currents in diseased human
cardiac cells. J Cardiovasc Pharmacol. 1995;25:282–291.
18. Cachelin AB, de Peyer JE, Kokubun S, Reuter H. Ca21 channel modulation by 8-bromocyclic AMP in cultured heart cells. Nature. 1983;304:
462– 464.
19. Ochi R, Kawashima Y. Modulation of slow gating process of calcium
channels by isoprenaline in guinea-pig ventricular cells. J Physiol. 1990;
424:187–204.
20. Yue DT, Herzig S, Marban E. b-Adrenergic stimulation of calcium
channels occurs by potentiation of high-activity gating modes. Proc Natl
Acad Sci U S A. 1990;87:753–757.
21. Herzig S, Patil P, Neumann J, Staschen CM, Yue DT. Mechanisms of
b-adrenergic stimulation of cardiac Ca21 channels revealed by discrete-time
Markov analysis of slow gating. Biophys J. 1993;65:1599–1612.
22. Hirano Y, Suzuki K, Yamawake N, Hiraoka M. Multiple kinetic effects
of b-adrenergic stimulation on single cardiac L-type Ca channels. Am J
Physiol. 1994;266:C1714 –C1721.
23. Handrock R, Schröder F, Hirt S, Haverich A, Mittmann C, Herzig S.
Single-channel properties of L-type calcium channels from failing human
ventricle. Cardiovasc Res. 1998;37:445– 455.
24. Wiechen K, Yue DT, Herzig S. Two distinct functional effects of protein
phosphatase inhibitors on guinea-pig cardiac L-type Ca21 channels.
J Physiol. 1995;484:583–592.
25. Biel M, Hullin R, Freudner S, Singer D, Dascal N, Flockerzi V, Hofmann
F. Tissue-specific expression of high-voltage activated dihydropyridinesensitive L-type calcium channels. Eur J Biochem. 1991;200:81– 88.
26. Neumann J, Gupta RC, Schmitz W, Scholz H, Nairn AC, Watanabe AM.
Evidence for isoproterenol-induced phosphorylation of phosphatase inhibitor-1 in the intact heart. Circ Res. 1991;69:1450 –1457.
27. Schwinger RHG, Böhm M, Schmidt U, Karczewski P, Bavendiek U,
Flesch M, Krause EG, Erdmann E. Unchanged protein levels of SERCA
II and phospholamban but reduced Ca21 uptake and Ca21-ATPase activity
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
of cardiac sarcoplasmic reticulum from dilated cardiomyopathy patients
compared with patients with nonfailing hearts. Circulation. 1995;92:
3220 –3228.
Bartel S, Stein B, Eschenhagen T, Mende U, Neumann J, Schmitz W,
Krause EG, Karczewski P, Scholz H. Protein phosphorylation in isolated
trabeculae from nonfailing and failing human hearts. Mol Cell Biochem.
1996;157:171–179.
Gao T, Yatani A, DellÁcqua ML, Sako H, Green SA, Dascal N, Scott JD,
Hosey MM. cAMP-dependent regulation of cardiac L-type Ca21 channels
requires membrane targeting of PKA and phosphorylation of channel
subunits. Neuron. 1997;19:185–196.
Haase H, Bartel S, Karczewski P, Morano I, Krause EG. In-vivo phosphorylation of the cardiac L-type calcium channel beta-subunit in
response to catecholamines. Mol Cell Biochem. 1996;163–164:99 –106.
Neumann J, Eschenhagen T, Jones LR, Linck B, Schmitz W, Scholz H,
Zimmermann N. Increased expression of cardiac phosphatases in patients
with end-stage heart failure. J Mol Cell Cardiol. 1997;29:265–272.
Linck B, Boknik P, Knapp J, Müller FU, Neumann J, Schmitz W,
Vahlensieck U. Effects of cantharidin on force of contraction and phosphatase activity in nonfailing and failing human hearts. Br J Pharmacol.
1996;119:545–550.
Pietrobon D, Hess P. Novel mechanism of voltage-dependent gating in
L-type calcium channels. Nature. 1990;346:651– 655.
Piot C, Lemaire S, Albat B, Seguin J, Nargeot J, Richard S. High
frequency-induced upregulation of human cardiac calcium currents. Circulation. 1996;93:120 –128.
January CT, Riddle JM. Early afterdepolarizations: mechanism of
induction and block: a role for L-type Ca21 current. Circ Res. 1989;64:
977–990.
Arai M, Matsui H, Periasamy M. Sarcoplasmic reticulum gene expression
in cardiac hypertrophy and heart failure. Circ Res. 1994;74:555–564.
Nakai J, Dirksen RT, Nguyen HT, Pessah IN, Beam KG, Allen PD.
Enhanced dihydropyridine receptor channel activity in the presence of
ryanodine receptor. Nature. 1996;380:72–75.
Beurg M, Sukhareva M, Strube C, Powers PA, Gregg RG, Coronado R.
Recovery of Ca21 current, charge movements, and Ca21 transients in
myotubes deficient in dihydropyridine receptor b1 subunit transfected
with b1 cDNA. Biophys J. 1997;73:807– 818.
Maki T, Gruver EJ, Davidoff AJ, Izzo N, Toupin D, Colucci W, Marks
AR, Marsh JD. Regulation of calcium channel expression in neonatal
myocytes by catecholamines. J Clin Invest. 1996;97:656 – 663.
Increased Availability and Open Probability of Single L-Type Calcium Channels From
Failing Compared With Nonfailing Human Ventricle
Frank Schröder, Renate Handrock, Dirk J. Beuckelmann, Stephan Hirt, Roger Hullin, Leo
Priebe, Robert H. G. Schwinger, Joachim Weil and Stefan Herzig
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
Circulation. 1998;98:969-976
doi: 10.1161/01.CIR.98.10.969
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1998 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circ.ahajournals.org/content/98/10/969
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial
Office. Once the online version of the published article for which permission is being requested is located,
click Request Permissions in the middle column of the Web page under Services. Further information about
this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation is online at:
http://circ.ahajournals.org//subscriptions/