411
Brief Communication
Reconstitution and Characterization of a
Calcium-Activated Channel From Heart
Joseph A. Hill Jr., Roberto Coronado, and Harold C. Strauss
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
This paper is the first description of a calcium-activated nonspecific cation channel in adult ventricular
muscle. We report gating kinetics and ionic selectivity data from experiments performed at the single
channel level. Calcium activation is described by a gating model wherein two ions are involved in the
reaction. Channel gating exhibited marked voltage dependence. Ionic selectivity experiments indicated
that the channel is cation-selective but unable to discriminate between Na+ and K+. We discuss evidence
that this channel mediates transient inward current in ventricular tissue; thus, the channel may be
involved in afterdepolarization-induced cardiac arrhythmias. (Circulation Research 1988;62:
411-415)
T
he several effects of calcium on cardiac electrical
properties are the subject of intense interest at
present. For example, an oscillatory inward
current exists in calcium-overloaded cardiac muscle
following depolarizing voltage-clamp steps.1"3 At
present, the molecular mechanism that underlies this
transient inward current (LJ remains unidentified. A
substantial body of literature supports the hypothesis
that this current is calcium-induced1-34 and involved in
the genesis of afterdepolarizations and aftercontractions.1-3 In addition, it is known that afterdepolarizations can trigger extra beats in isolated cardiac
preparations5-6 that sometimes develop into runs of
repetitive activity. Thus, 1^ may be an important
pathological mechanism underlying the clinical events
accompanying calcium overload.
Channel opening gives rise to an elementary current
pulse, the size and duration of which can be observed
directly by using single-channel recording techniques.
One such technique, planar bilayer reconstitution,
involves the incorporation of native channel protein
into an artificial membrane with simultaneous measurement of transmembrane current at high gain. Using
this technique, we describe a calcium-activated channel
from cardiac ventricular sarcolemma that would mediate inward current under physiological conditions.
As such, these data support the hypothesis that
channel-mediated current plays a role in !„ and in
From the Departments of Medicine and Pharmacology, Duke
University Medical Center, Durham, North Carolina, and the
Department of Physiology and Molecular Biophysics (R.C.),
Baylor College of Medicine, Houston, Texas.
Supported by National Heart, Lung, and Blood Institute grants
19216and 17670, and National Institute of General Medical Science
grants 5 T32 GM 07171 and 07105.
Address for correspondence: Joseph A. Hill Jr., MD, PhD,
Division of Cardiology, Box 3845, Duke University Medical Center,
Durham, NC 27710.
Received March 26, 1987; accepted August 7, 1987.
the development of afterdepolarizations and triggered
arrhythmias.
Materials and Methods
Sarcolemmal vesicles from adult canine left ventricle were prepared according to the method of Jones.7
Lipid bilayers were prepared by painting a 1:1 (wt/wt)
phospholipid mixture of phosphatidylethanolamine
and phosphatidylserine (Avanti Polar Lipids, Birmingham, Alabama) dissolved in decane (20 mg lipid/ml
decane) across a 300-/xm diameter aperture separating
two aqueous chambers (cis, trans). Typical bilayers
exhibited 200-300 pF capacitance. Resistances were
always greater than 100 Gil. The ventricular vesicles
were incorporated into the artificial membrane by
adding them to the cis chamber under fusion
conditions.8 All aqueous solutions were buffered with
10 mM histidine (pH 7.1). Bath calcium activity was
calculated in CaEGTA solutions using an effective
stability constant of 4.31X106 M"1 (0.1 M ionic
strength, pH 7.1, 25° C).9 All experiments were
performed at room temperature (25-26° C) and under
steady-state conditions. The cis chamber was connected to a voltage source while the trans chamber was
held at virtual ground by a current-to-voltage converter
circuit. Analog data were low-pass filtered (usually — 3
dB at 100 Hz, eight-pole Bessel) and stored on FM tape.
Later, the signal was sampled and stored on magnetic
disk for digital analysis. Linear regression was performed according to the method of least squares.
Abbreviations used include: [X], concentration of
species X; KOAc, potassium acetate; EGTA, ethylene
glycol-bis03-amino-ethylether)A^,Af,Af',Af'-tetraacetic
acid; pCa, -log([Ca]).
Results
Gating Kinetics
A representative continuous record of single channel
current is illustrated in Figure 1A. As can be seen,
Circulation Research
412
urn
500 TO
56 ms
22 ms
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
200 ms
rooms
9 ms
29 ms
31
3
8
Vol 62, No 2, February 1988
tions of open channel duration or closed state duration
were fitted satisfactorily only by the sum of two
exponential components (Panel B).
Single channel open probability is modulated by cis
calcium activity (Figure 2). Representative recordings
selected from data points depicted in Panel B are shown
in Panel A. Open probability ranges from 8% at 0.2 /xM
Ca2+ in the cis chamber to 99% at 10 mM Ca2+; this
effect is readily reversible upon the addition of EGTA.
Popen, a function of cis calcium concentration defined
as cumulative open time divided by the sum of
cumulative open and closed times [0/(0 + C)], varied
approximately between unity and 0 over the pCa range
of 6 to 7 (Panel B). This observation fits well with the
experimentally observed values of intracellular calcium in ventricular myocytes under physiological
conditions (approximately 100 nM diastolic Ca2+)10
and in states of calcium overload (1 piM and
above)."13 We have, however, observed shifts in the
open state probability between different experiments
(cf., Figure 2A and Figure 3) as have other
investigators14'5 working with calcium-activated potassium channels.
The sigmoidal curve superimposed on the data in
Figure 2B is that predicted by a model wherein two
calcium ions bind to the channel. A fit to the data was
obtained by plotting 1/Popen versus [Ca2+]~2 (Panel C).
The slope of the fitted line was 3.1 X 10"" M2 and is
a measure of the equilibrium dissociation constant, Kd.
In all experiments (n = 9), the calcium-sensitive site
was oriented toward the cis chamber.
Methfessel and Boheim14 have proposed an activation/blockade model to describe the gating kinetics
observed in calcium-activated channels from skeletal
muscle:
Ca2
R
50 ms
100 ms
FIGURE 1. Single channel gating kinetics. Panel A, Continuous analog recording of single channel current, Vhold= + 10
mV,cis[KCl] = 73mM,tnns[KCl]=50mM,
[Ca!+]<]00nM,
open = up. In the record from which the illustration was taken,
mean open time was 93 msec. Panel B, Cumulative probability
distributions for open and closed states calculatedfrom a record
with 1,590 openings. Data were recorded at —40 mV in same
solutions as in Panel A. Probability distributions were calculated according to the following protocol: current amplitude
histograms were calculated, and means of approximately
normally distributed current samples were identified. Transition
discriminators for channel opening and closing were assigned
midway between the two means. The sum of two exponential
functions is superimposed on each histogram with time constants
(ms) as depicted.
gating occurs in bursts of activity separated by relatively long periods of silence. The open state (displayed
up) exhibits flickery kinetics as it is interrupted by
numerous short-lived closures. In all cases, distribu-
Ca2
R - C a < — > R*-Ca
R*-Ca,(l)
where R is the channel receptor, and the asterisk
signifies the open state. This linear scheme is a hybrid
of two models, viz., an activation model in which a
calcium ion binds to the closed channel allowing it to
open (R <---> R - C a < - - > R*-Ca) and a
"blockade" model in which a second calcium ion
stabilizes the open configuration (R —Ca <—>
R * - C a < — > R*-Ca 2 ).
Scheme 1 allowed us to make several predictions for
experimentally observable variables. First, the model
predicts that except under limiting conditions (very
high or very low [Ca2+]), two exponential components
will be required to describe the open and closed state
probability distributions (Figure IB). Second, the
model predicts that the ratio of mean open time to mean
closed time (equilibrium constant, K) will be a function
of calcium concentration raised to an exponent reflecting the number of ions that bind to the channel. A
best-fit line obtained when these data are plotted as
log(K) versus log([Ca2+]) had a slope of 2.5 (data not
shown). Third, scheme 1 predicts mat under limiting
conditions of calcium activity, mean open time is a
linear function of [Ca2+] and that mean closed time is
Hill et al Cardiac Calcium-Activated Channel
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
20
40
(TIIDCT 10EI2)
1 / [CaT2
413
-6.4
-62
LOG [Co]
FIGURE 2. Popen is a function o/cis calcium concentration. Panel A, Representative records are shown from data points in Panel
B. Popen [defined as cumulative open time divided by the sum of cumulative open and closed times, 0/(0 + C)] is as listed; symmetrical
[KOAc] = 100 mM, cis [Co2*] = 10 mM, 1.6 (JLM, 0.4 yM, 0.25 \xM, and0.2 fxM, trans [Cd*]<100 nM (open = up). Panel B, Popen
is plotted versus — logdCa2*]). All recordings were made at —40 mV holding potential. The continuous curve is the transformed
equation from the analysis in Panel C. Panel C, 1/Popen versus IllCa2*]2. Coefficient of correlation frj of the fitted line is 0.988.
Panel D, Open and closed state durations are functions of cis [Ca2*]. • , log mean open time versus logdCa2*]). The slope of the
best fit line is 0.93, x = 0.93; D, log mean closed time versus logdCa1*]). The slope of the best fit line is -1.5, r = 0.90.
a linear function of l/[Ca 2+ ] (Figure 2D). In all cases,
our observations were consistent with scheme 1.
Channel gating was strongly voltage-dependent
(Figure 3). In this experiment, cis and trans calcium
concentrations are symmetrical and represent base line
calcium for solutions in our laboratory in which neither
calcium nor chelator is added. We have subsequently
measured this calcium concentration as less than 100
nM. Single channel open probability varied from
approximately 0 to 1 over the voltage range - 60 mV
to 0 mV under these conditions. The sigmoidal
activation curve depicted is a Boltzmann relation obtained as a fit to the logit transformation ln[(l - Popen)/
Popen] versus voltage. The slope of the fitted line is
interpreted to indicate that —2.5 gating charges move
from trans to cis upon channel opening. The v intercept
obtained by this method indicates that - 2 . 2 kcal/mol
"nonelectrical" free energy are required for channel
opening. That this nonelectrical component is less than 0
confirms that the open state is favored at 0 mV.
Ionic Selectivity
Figure 4 depicts single channel current as a function
of holding voltage. Each data point represents the mean
of at least three (and usually four to seven) amplitude
measurements. Single channel current was ohmic over
the voltage range - 80 to + 20 mV. In the presence of
a KC1 activity gradient across the bilayer of 1.4, the
interpolated reversal potential was —7 mV, which
agrees well with the Nernst potential for K + ( - 9 mV)
(Panel A). In the presence of increasingly asymmetric
KC1 solutions, die single channel current-voltage
relation shifts toward more negative potentials. In each
case, the experimentally determined reversal potential
agreed with the Nernst potential for a K + electrode
within 3 mV. Thus, we conclude that the channel
selects strongly for cations over anions.
We have also measured the reversal potential under
biionic conditions for Na + andK + (Figure4B). Afitted
line is superimposed on the data that has an x intercept
of + 2 mV and a slope of 120 pS. By assuming ionic
independence, we may invoke the Goldman-HodgkinKatz equation for biionic conditions to calculate
relative ionic permeabil ities. In so doing, it is clear that
these data are consistent with a channel that is
completely nonselective between Na + and K + ions.
This observation suggests that the channel conducts
current that reverses near 0 mV under physiological
Circulation Research
414
1.0 -0.8 • •
0.6 - •
0.4-•
0.2--
-80
-60
-40
-20
VOLTAGE/mV
FIGURE 3. Popen is a function of holding potential. Symmetrical [KOAc] = 100 mM, symmetrical [Ca'+]<100 nM. The
sigmoid curve is a Boltzmann relation calculated as described
in the text.
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
conditions. We interpret the conductance (120 pS) to
represent a first approximation (neglecting temperature
effects) to the channel's conductance in a physiological
environment.
At present, we are unable to quantify relative
calcium permeability. However, the experiment shown
in Figure 2 allowed us to estimate relative Ca2+
permeability with respect to K + at less than 0.2. Single
channel current increased less than 0.7 pA (5 %) at — 40
mV over the pCa range of 2 to 7 in the cis chamber;
from this, we conclude that the channel exhibited
relatively little calcium permeability.
-JO
0
VOLTAGE /HV
•?0
VOLT»CE/oV
0
JO
FIGURE 4. Steady-state current-voltage (I-V) relations,
[Ca!*]<]00 nM. Panel A, I-V curves recorded under the
following conditions (cis [KCl]// trans [KCl], ratio of activities): 731150 mM, 1.4 (*), 961150 mM, 1.8 (*). 3191150 mM,
5.4 (O). In every case, standard error for each data point was
less than 0.2 pA (Panels A and B); Panel B, Biionic potential
experiment performed under the following conditions: cis 700
mM NaOAc, trans 700 mM KOAc.
Vol 62, No 2, February 1988
Discussion
Calcium-Activated Channels in Other Tissues
This report represents the first description of single
calcium-activated channels from working adult myocardium. Colquhoun et al16 have described a calciumactivated channel from cultured neonatal rat heart that
exhibited some of these same characteristics except that
the single channel conductance was markedly lower,
viz., 30-40 pS in the presence of physiological salt
solutions. In addition, they observed voltageindependent gating and calcium sensitivity in the 1-10
(xM range. We speculate such discrepancies may be due
to species and tissue development differences. One
report has been published on single channel measurements of a calcium-activated potassium-selective channel in bovine cardiac Purkinje fibers." Calciumactivated channels have also been described from
skeletal" and cardiac" sarcoplasmic reticulum. The
sarcoplasmic reticulum channels, however, differ from
the one described here in several respects: 1) they select
for divalent cations over monovalents by a factor of
10,1S 2) are weakly voltage-dependent," 3) display
open lifetime kinetics that are independent of calcium
concentration,20 4) have a calcium binding stoichiometry of 0.3 (R. Coronado, personal communication)
instead of 2 (Figure 2C), and 5) are inhibited by cis
calcium concentrations above 1 mM21 unlike activity as
shown in Figure 2A.
Most of the biophysical measurements we report
here are similar to those reported by others studying
calcium-activated channels.22 Specifically, channel gating displays bursting kinetics, a characteristic feature
of calcium-activated potassium channels from skeletal
muscle1415-23 as well as calcium-activated nonspecific
cation channels from cultured neonatal rat hearts."
Indeed, it was the distinctive appearance of the open
state that led us to investigate the possibility of
calcium-dependent gating. Second, large conductance
is characteristic of calcium-activated channels,22 with
a few notable exceptions, such as the calcium-activated
channel of neonatal rat heart (see above).16 Marked
voltage dependence of gating and an agonist-receptor
stoichiometry of 2 : 1 are features of calcium-activated
channels in skeletal muscle.1323 In fact, the binding of
two agonist ions for channel activation is in keeping
with the general properties of many ligand-gated
channels, e.g., calcium-activated channels, I5J3 acetylcholine receptor channels,24 and glycine and
y-aminobutyric acid receptors.23 Finally, the calcium
dependence of mean open time and mean closed time
is similar to that observed in calcium-activated potassium channels from cultured skeletal muscle studied
with the patch-clamp technique14 or reconstituted into
bilayers.23
Significance
We were able to reconstitute this channel into
bilayers sporadically; this suggests that the channel
may be present in the preparation in relatively small
numbers. Further, the fact that the cardiac channel has
never been observed with the patch-clamp technique
415
Hill et al Cardiac Calcium-Activated Channel
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
suggests that this channel may be localized to clefts in
the sarcolemmal membrane as in skeletal muscle.
However, in order to elucidate further its physiological
role, it will be necessary to demonstrate its function
directly using either patch clamping or high-affinity
ligand binding techniques. In planning such experiments, it is a useful exercise to estimate channel density.
We have estimated the channel's conductance in a
physiological environment at 120 pS (see above). In
addition, we have assumed 100 nA of transient inward
current in a Purkinje fiber strand of radius 100 pirn and
length 1,000 pm (see Cannell and Lederer,3 Figure 1).
Finally, true membrane surface area was taken to be 10
times the apparent surface area of a smooth envelope
of membrane.26 Macroscopic current I is given by
I = nip, where n is number of channels, i is single
channel current, and p is channel opening probability
(taken to be 1 at 1 /i.M calcium and — 50 mV, Figure
1 A). Further, i = vy, where v is electrochemical driving
force (taken to be 50 mV), and y is single channel
conductance. With these assumptions, we have estimated the channel density to be 1 per 200 /urn2. Thus,
we predict that if this channel mediates 1^, then it will
be sparsely distributed in the cell membrane;
Cannell and Lederer3 have presented evidence that
the transient inward current in sheep Purkinje fibers is
dependent on channel current (although Na-Ca exchange-mediated current or current from other electrogenic mechanisms may also be involved). The
putative transient inward channel must exhibit calciumdependent gating over a physiological range of intracellular calcium, conduct inward current at diastolic
membrane potentials, and select for cations over
anions. Our measurements of calcium sensitivity,
P Ni : PK approaching 1, and chloride impermeability are
consistent with these predictions. Further, the steep
voltage dependence we observed may explain the
characteristic curvature in the macroscopic currentvoltage relation.3 We have not tried to measure Ca2+
current under the same conditions used by these
investigators (replacing trans monovalent cations with
Ca 2+ ); we do report, however, absence of substantial
calcium permeability in the presence of monovalent
cations. This observation is consistent with studies of
calcium-activated channels from neuroblastoma.27 For
these reasons, we propose that this channel may
contribute to transient inward current in canine ventricular muscle and may be involved in the development
of triggered arrhythmias in calcium-overloaded tissue.
Acknowledgments
We gratefully acknowledge Dr. Jonathan Lederer for
his critical reading of the manuscript. We thank Joseph
C. Greenfield Jr. for continued support and S. Webb
and B. Hill for editorial assistance.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
References
1. Kass RS, Lederer WJ, Tsien RW, Weingart R: Role of calcium
ions in transient inward currents and aftercon tract ions induced
by strophanthidin in cardiac purkinje fibres. J Physiol
1978^281:187-208
2. Kass RS, Tsien RW, Weingart R: Ionic basis of transient inward
27.
current induced by strophanthidin in cardiac purkinje fibres. J
Physiol 1978;281:209-226
Cannell MB, Lederer WJ: The arrhythmogenic L, current in the
absence of electrogenic sodium-calcium exchange in sheep
cardiac purkinje fibres. J Physiol 1986;374:201-219
Eisner DA, Vaughan-Jones RD: Do calcium-activated potassium channels exist in the heart? Cell Calcium 1983;4:371-386
Ferrier GR: Digitalis arrhythmias: Role of oscillatory afterpotentials. Prog Cardiovasc Dis 1977; 19:459-474
Rosen MR, Reder RF: Does triggered activity have a role in the
genesis of cardiac arrhythmias? AnnlntM ed 1981;94:794-801
Jones LR: Methods in Enzymology. (in press)
Miller C: Voltage-gated cation conductance channel from
fragmented sarcoplasmic reticulum. Steady state electrical
properties. J Membr Biol 1978;40:1-23
Martell AE, Smith RM: Critical Stability Constants. New
York, Plenum Press, 1974
Wier WG, Cannell MB, Berlin JR, Marban E, Lederer WJ:
Cellular and subcellular heterogeneity of [Ca2+], in single heart
cells revealed by fura-2. Science 1987;235:325-328
Cannell MB, Wier WG, Berlin JR, Marban E, Lederer WJ:
Free intracellular calcium in normal and calcium-overloaded
rat heart cells: Digital imaging fluorescent microscopy using
fura-2 (abstract). Biophys J 1986;49:466a
Yue DT, Marban E, Wier WG: Relationship between force and
intracellular [Ca1+] in tetanized mammalian heart muscle. /
Gen Physiol 1986;87:223-242
Valdeolmillos M, Eisner DA: The effects of ryanodine on
calcium-overloaded sheep cardiac purkinje fibers. Circ Res
1985^6:452^56
Methfessel C, Boheim G: The gating of single calciumdependent potassium channels is described by an activation/
blockade mechanism. Biophys Struct Mech 1982;9:35-60
Barrett JN, Magleby KL, Pallotta BL: Properties of single
calcium-activated potassium channels in cultured rat muscle.
J Physiol 1982;331:211-230
Colquhoun D, Neher E, Reuter H, Stevens CF: Inward current
channels activated by intracellular Ca in cultured cardiac cells.
Nature 1981 ;294:752-754
Callewaert G, Vereecke J, Carmeliet E: Existence of a
calcium-dependent potassium channel in the membrane of cow
cardiac purkinje cells. Pflugers Arch 1986;406:424-426
Smith JS, Coronado R, Meissner G: Sarcoplasmic reticulum
contains adenine nucleotide-activated calcium channels. Nature 1985 ;316:446-^49
Rousseau E, Smith JS, Henderson JS, Meissner G: Single
channel and "Ca2+ flux measurements of the cardiac sarcoplasmic reticulum calcium channel. Biophys J 1986;50:10091014
Smith JS, Coronado R, Meissner G: Single channel measurements of the calcium release channel from skeletal muscle
sarcoplasmic reticulum: Activation by Ca2+ and ATP and
modulation by Mg^+. J Gen Physiol 1986;88:573-588
Meissner G, Darling E.EvelethJ: Kinetics of rapid Ca2+release
by sarcoplasmic reticulum. Effects of Ca2+, Mg2+, and adenine
nucleotides. Biochemistry 1986;25:236-244
Schwarz W, Passow H: Ca2+-activated K* channels in erythrocytes and excitable eel Is. Annu Rev Physiol 1983;45:359-374
Moczydlowski E, Latorre R: Gating kinetics of Ca2+-activated
K+ channels from rat muscle incorporated into planar lipid
bilayers. J Gen Physiol 1983;82:511-542
Changeux J-P, Devillers-Thiery A, Chemouilli P: Acetylcholine receptor An allosteric protein. Science 1984;225:
1335-1345
Sakmann B, Hamill OP, Borman J: Activation of chloride
channels by putative inhibitory transmitters in spinal cord
neurons (abstract). Pflugers Arch 1982;392:R19
Mobley BA, Page E: The surface area of sheep cardiac purkinje
fibres. J Physiol 1972^220:547-563
Yellen G: Single Ca2+-activated non-selective cation channels
in neuroblastoma. Nature 1982;296:357-359
KEY WORDS • reconstitution • calcium-activated channel
• triggered arrhythmia
Reconstitution and characterization of a calcium-activated channel from heart.
J A Hill, Jr, R Coronado and H C Strauss
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
Circ Res. 1988;62:411-415
doi: 10.1161/01.RES.62.2.411
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1988 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/62/2/411
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Circulation Research 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 Research is online at:
http://circres.ahajournals.org//subscriptions/