Abnormal Ca2ⴙ Release, but Normal Ryanodine Receptors,
in Canine and Human Heart Failure
Ming Tao Jiang, Andrew J. Lokuta, Emily F. Farrell, Matthew R. Wolff,
Robert A. Haworth, Héctor H. Valdivia
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
Abstract—Sarcoplasmic reticulum (SR) Ca2⫹ transport proteins, especially ryanodine receptors (RyR) and their accessory
protein FKBP12.6, have been implicated as major players in the pathogenesis of heart failure (HF), but their role remain
controversial. We used the tachycardia-induced canine model of HF and human failing hearts to investigate the density
and major functional properties of RyRs, SERCA2a, and phospholamban (PLB), the main proteins regulating SR Ca2⫹
transport. Intracellular Ca2⫹ is likely to play a role in the contractile dysfunction of HF because the amplitude and
kinetics of the [Ca2⫹]i transient were reduced in HF. Ca2⫹ uptake assays showed 44⫾8% reduction of Vmax in canine HF,
and Western blots demonstrated that this reduction was due to decreased SERCA2a and PLB levels. Human HF showed
a 30⫾5% reduction in SERCA2a, but PLB was unchanged. RyRs from canine and human HF displayed no major
structural or functional differences compared with control. The Po of RyRs was the same for control and HF over the
range of pCa 7 to 4. Subconductance states, which predominate in FKBP12.6-stripped RyRs, were equally frequent in
control and HF channels. An antibody that recognizes phosphorylated RyRs yields equal intensity for control and HF
channels. Further, phosphorylation of RyRs by PKA did not appear to change the RyR/FKBP12.6 association,
suggesting minor ␤-adrenergic stimulation of Ca2⫹ release through this mechanism. These results support a role for SR
in the pathogenesis of HF, with abnormal Ca2⫹ uptake, more than Ca2⫹ release, contributing to the depressed and slow
Ca2⫹ transient characteristic of HF. (Circ Res. 2002;91:●●●-●●●.)
Key Words: heart failure 䡲 ryanodine receptor 䡲 sarcoplasmic reticulum 䡲 protein kinase A
I
t is generally accepted that in ventricular cardiomyocytes,
depolarization of the sarcolemma and T-tubules triggers
contraction by the process of Ca2⫹-induced Ca2⫹ release
(CICR).1,2 Depolarization opens voltage-dependent Ca2⫹
channels/dihydropyridine receptors (DHPR), allowing a
small flux of external Ca2⫹ (the inward Ca2⫹ current, ICa) that
in mature cells is insufficient to elicit a full contraction. The
incoming Ca2⫹, however, opens Ca2⫹ release channels/ryanodine receptors (RyR) of the sarcoplasmic reticulum (SR),
producing a massive Ca2⫹ discharge and raising [Ca2⫹]i to
contracting levels. Synchronous and coordinated contraction
of ventricular cells propels blood into arteries; diastolic
refilling occurs when [Ca2⫹]i is re-sequestered back into the
SR by the Ca2⫹-ATPase and extruded from the cytosol by the
Na⫹-Ca2⫹ exchanger, causing relaxation of the cell.
A prominent characteristic of several animal models of
heart failure (HF) and of the failing human myocardium is a
depression of the rate of contraction and relaxation of
individual cardiomyocytes.3–7 This contractile dysfunction is
at least partly attributed to abnormal intracellular Ca2⫹ cycling, inasmuch as the peak and kinetics of the [Ca2⫹]i
transient appear to be depressed, too. However, despite
extensive research, the exact molecular players and the
mechanisms responsible for these alterations remain controversial. Although most studies agree that SR Ca2⫹ content is
depressed in HF, some investigators attribute this decrease to
abnormal density/function of Ca2⫹ uptake and extrusion
mechanisms (ie, Ca2⫹-ATPase and Na⫹-Ca2⫹ exchanger)6 – 8
and others to abnormal RyR function.9,10
Proposing RyRs as central players in the contractile dysfunction of HF is sound and appealing for several reasons.
First, RyRs are the main (if not the only) gate for Ca2⫹ release
from the SR, and small variations in their activity may
produce large fluctuations in intra-SR Ca2⫹. Second, RyRs
are tightly bound to FKBP12.6, and disruption of the RyR/
FKBP12.6 complex produces unstable channels with high
activity and propensity to leak Ca2⫹ even at rest.11 If PKA
phosphorylation strips RyRs of FKBP12.6, as proposed
recently,9 then the increased catecholamine levels found in
HF would produce hyperphosphorylation of RyRs, potentially resulting in abnormal Ca2⫹ leak from the SR.
In this study, we used the tachycardia-induced canine
model of HF and human failing myocardium to determine
density and function of the major SR Ca2⫹ transport proteins.
Original received April 22, 2002; revision received October 15, 2002; accepted October 15, 2002.
From the Department of Physiology (M.T.J., A.J.L., E.F.F., H.H.V.), Medicine (M.R.W.), and Surgery (R.A.H.), University of Wisconsin Medical
School, Madison, Wis.
Correspondence to Héctor H. Valdivia, MD, PhD, University of Wisconsin, 1300 University Ave, Madison, WI 53706. E-mail
[email protected]
© 2002 American Heart Association, Inc.
Circulation Research is available at http://www.circresaha.org
DOI: 10.1161/01.RES.0000043663.08689.05
1
2
Circulation Research
November 29, 2002
We found that, despite their great potential as regulators of
SR Ca2⫹ content, RyRs remain largely unaffected by the
pathophysiological mechanisms that occur in HF, and that an
abnormal Ca2⫹ uptake, more than Ca2⫹ release, is likely to
explain the blunted Ca2⫹ transient of HF.
Materials and Methods
Induction of Heart Failure
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
All animal studies were performed in accordance to the guidelines in
the NIH Guide for the care and Use of Laboratory Animals DHHS
publication No. [NIH] 85-23 (revised 1985), and approved by
institutional review board. The protocol for induction of heart failure
is detailed elsewhere.12 A total of 18 control (sham-operated) and 14
HF dogs were used in this study. The studies involving human hearts
were approved by the Institutional Review Committee of the
University of Wisconsin and were conducted with informed consent
from the patients. Human ventricular cells and myocardium (for
isolation of cell organelles) were obtained from explanted hearts of
6 nonfailing (control) and 19 failing patients with ischemic or dilated
cardiomyopathic failure (ICM and DCM, respectively). Echocardiographic and intracardiac pressure measurements in both canine and
human failing hearts showed overt signs of congestive HF including
LV dilation and systolic dysfunction.
Isolation of Ventricular Myocytes and
Measurement of [Ca2ⴙ]i Transients
Single ventricular myocytes were obtained by enzymatic digestion
using the procedure of Haworth et al.13 Cells were suspended in
Krebs-Henseleit HEPES medium and kept at room temperature until
used, usually within 8 hours. For measurements of intracellular Ca2⫹,
myocytes were perfused with a modified Tyrode solution containing
(in mmol/L) NaCl 138, MgCl2 1, KCl 4.4, dextrose 11, CaCl2 1,
HEPES 12, pH 7.4. Cells were incubated at room temperature in
Tyrode solution with 5 ␮mol/L Fluo-4AM for 30 minutes. Confocal
images (BioRad MR-1) were recorded with the scan line oriented
along the long axis of the cell. Fluo-4 was excited at 488 nm, with
emitted fluorescence measured at 515 nm. Ca2⫹ transients were
reconstructed by stacking 512 consecutive line scans and performing
a time-intensity plot using a software program running in IDL 5.4
(written by Dr Ana M. Gómez). [Ca2⫹]i was calculated using a
pseudoratio method assuming a Kd and resting [Ca2⫹]i of 1.1 ␮mol/L
and 150 nmol/L, respectively.
Preparation of SR-Enriched Microsomes
SR-enriched membranes were isolated from canine and human left
ventricle by differential centrifugation as described in Lokuta et al.14
Ventricular tissue was flash-frozen in liquid N2 immediately after
explantation. In some experiments, NaF (20 mmol/L) and okadaic
acid (1 ␮mol/L) were added to the homogenization medium to
prevent protein dephosphorylation that might occur during
homogenization.
45
Ca2ⴙ Uptake Measurements
ATP-dependent, oxalate-facilitated 45Ca2⫹ uptake by SR microsomes
was determined by the rapid filtration technique, as described15 (see
also online data supplement).
trophoresis (SDS-PAGE) in 4 to 20% polyacrylamide precast minigels (BioRad). Phosphorylated PLB and RyR were detected by
autoradiography, and the gel slices containing both proteins were cut
and counted by liquid scintillation spectrophotometry.
Western Blot Assays
SR microsomes or whole homogenates were dissolved in sample
buffer at 37°C for 10 minutes and electrophoresed in SDS gels as
described above for transfer onto nitrocellulose membranes, as
described.16 Membranes were probed with primary antibodies
against SERCA2a, PLB, and RyRs (Affinity Bioreagents Inc.), or
against FKBP12 (polyclonal antibody that recognizes FKBP12.6 and
FKBP12, kindly provided by Dr Angela Dulhunty), or against
phosphorylated RyR (P-S2809, Badrilla), diluted 1:2500 (FKBP12
and P-S2809) or 1:5000 (all others) in phosphate-buffered saline
(PBS). After washing, membranes were incubated with secondary
antibodies (IgG) conjugated to horseradish peroxidase (1:30 000
dilution in PBS). Protein-antibody reactions were detected by chemiluminescence using Kodak X-Omat films. The relative amount of
proteins on the blots was determined by densitometric analysis using
a HP 3c laser scanner and the program SigmaGel. Protein standards
were included in each blot to normalize the densitometric data to a
known amount of protein loaded.
[3H]Ryanodine Binding Assay
[3H]Ryanodine binding was performed as described.14 –17 The incubation medium contained 1 mol/L KCl, 20 mmol/L MOPS (pH 7.2),
60 to 120 ␮g of cardiac microsomes or 120 to 240 ␮g of homogenates, 10 ␮mmol/L CaCl2 (total volume 100 ␮L), and [3H]ryanodine
(0.5 to 20 nmol/L). The incubation lasted 90 minutes at 36°C. The
Ca2⫹ dependence of [3H]ryanodine binding was determined in total
homogenates in medium containing 20 mmol/L MOPS (pH 7.2),
150 mmol/L KCl, 3 mmol/L ␤ , ␥ -methylene-adenosine 5⬘triphosphate (AMP-PCP, a nonhydrolyzable ATP analog), 1 mmol/L
EGTA, 2.8 to 3 mmol MgCl2 (final free [Mg2⫹] 0.55 mmol/L), and
CaCl2 to give a range of free [Ca2⫹] from pCa 7 to pCa 4.
Single-Channel Recording of RyRs and Activation
by Fast Ca2ⴙ Steps
RyRs in SR vesicles were incorporated into planar lipid bilayers and
recorded as described.14 –17 The trans and the cis chamber (luminal
and cytoplasmic side, respectively, 600-␮L each) contained
300 mmol/L cesium methanesulfonate and 10 mmol/L HEPES-Na
(pH 7.2). Rapid exchange of solutions at the cytosolic side of the
channel (Figure 5) was achieved as described in the online data
supplement.
Data Analysis
All data are presented as mean⫾SEM and differences are considered
significant with P⬍0.05. Ca2⫹ transients and SR protein phosphorylation were compared with t tests. Ca2⫹ transport and its kinetic data
(Vmax, K0.5, and Hill coefficients), effects of HF on Po of RyRs at
constant and transient [Ca2⫹] were analyzed with two-way ANOVA
for repeated measurements, followed by post hoc analysis with t
tests.
Results
2ⴙ
Phosphorylation of SR Proteins by PKA
SR microsomes were phosphorylated at 37°C in medium containing
(in mmol/L) 50 Tris-maleate (pH 6.8), 120 KCl, 2 MgCl2, 10 NaF,
1 [␥-32P]ATP (100 to 300 cpm/pmol), 1 ␮mol/L of okadaic acid, the
catalytic subunit of PKA (40 U), and 2 ␮mol/L thapsigargin.
Negative controls were prepared by adding 10 ␮mol/L peptide
inhibitor of PKA (PKI, Sigma). The reaction (50 ␮L) was initiated
by the addition of [␥-32P]ATP and stopped after 3 minutes with 12.5
␮L of sample buffer (4⫻) consisting of 0.25 mol/L Tris-HCl (pH
6.8), 8% SDS, 50% glycerol, and 0.4 mmol/L DTT. Aliquots (40 ␮g)
of SR membranes were subjected to SDS-polyacrylamide gel elec-
Ca
Transients in Control and HF Myocytes
We used line-scan imaging to measure intracellular Ca2⫹
transients in Fluo 4 –loaded control and HF canine ventricular
myocytes. Cells were field-stimulated with a 10-ms, 30-V
pulse applied at 1 Hz. Figure 1A shows stacked line-scan
Ca2⫹ images in ventricular myocytes from control and HF
dogs. Control cells showed a rapid and homogeneous increase
in [Ca2⫹]I, whereas the majority of HF cells displayed
attenuated and slower [Ca2⫹]i elevations. Figure 1B plots the
integrated pixel intensity of the line-scan image versus time
Jiang et al
Abnormal Ca2ⴙ Release in Heart Failure
3
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
Figure 1. Ca2⫹ transients in canine control and HF cells. Top, Line scan (2-ms) images were stacked horizontally to determine the time
course of the [Ca2⫹]i transient in control and HF cells. Scan line was aligned along the longitudinal axis of the cell. Distance is therefore
displayed on the y-axis (0.2 ␮m/pixel). Bottom, Line scan images (Ca2⫹ transients) are presented as plots of integrated pixel intensity
versus time. Cells of the HF group had Ca2⫹ transients with smaller amplitude and slower kinetics.
and represents the Ca2⫹ transient for each stimulation in the
control (blue line) and HF (green line) cells. Both, the
amplitude and the kinetics of the [Ca2⫹]i transient were
reduced in HF cells. Peak [Ca2⫹] was 0.91⫾0.012 ␮mol/L in
control and 0.60⫾0.018 ␮mol/L in HF cells (n⫽34 and 32,
respectively; P⬍0.05). Time to peak was 47.3⫾4.1 ms in
control and 98.2⫾12 ms in HF cells. Time to 50% decay of
the [Ca2⫹]i transient was 625⫾36 ms in control and 861⫾64
ms in HF. Thus, HF cells display marked abnormalities in the
[Ca2⫹]i transient. These results are in agreement with previous
studies that find that the contractile dysfunction of HF may
not be solely attributed to defects of contractile proteins3– 8,18;
abnormalities in mechanisms that regulate [Ca2⫹]i may also
contribute significantly to the depressed contraction and
relaxation.
Alterations in SR Ca2ⴙ Uptake and PKA Stimulation
Figure 2A shows ATP-supported and oxalate-facilitated
45
Ca2⫹ uptake by SR-enriched microsomes from control and
HF dogs, in the absence and the presence of 4 ␮g/mL of the
catalytic subunit of PKA. Compared with control, both the
initial rate and the magnitude of Ca2⫹ uptake were lower in
HF SR at various [Ca2⫹]. Figure 2B shows that PKA
stimulated Ca2⫹ uptake in control and HF SR, especially at
low [Ca2⫹], but the extent of stimulation was lower in the HF
group (P⬍0.05 at 0.17 to 1.4 ␮mol/L Ca2⫹). Kinetic analysis
indicated that Vmax for 45Ca2⫹ uptake was depressed by 45%
(P⬍0.05) in HF compared with control, with no significant
changes in K0.5 for Ca2⫹ or n, the Hill coefficients. In control
SR, PKA stimulated Ca2⫹ uptake, but did not significantly
alter K0.5 for Ca2⫹ or Vmax. In HF SR, PKA reduced K0.5
Figure 2. Ca2⫹ uptake by control and HF SR. A, ATP-driven 45Ca2⫹ uptake into SR vesicles was measured for 3 minutes at 37°C in the
absence and the presence of the catalytic subunit of PKA and the indicated [Ca2⫹]. Rate of 45Ca2⫹ uptake is lower in HF at all [Ca2⫹]
(P⬍0.05 by two-way ANOVA), and PKA stimulated the rate of 45Ca2⫹ uptake at all [Ca2⫹] in both control and HF (P⬍0.05 to 0.001 with
paired t test after ANOVA). ANOVA indicated the effect of PKA was more pronounced in control than in HF at 0.17, 0.32, and 1.4
␮mol/L [Ca2⫹] (P⬍0.05). B, PKA-stimulated 45Ca2⫹ uptake (%). *Percent stimulation differs significantly between control and HF
(P⬍0.05, ANOVA followed by t-test, n⫽8)
4
Circulation Research
November 29, 2002
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
Figure 3. SERCA2a and phospholamban density in canine HF. A, Western blots of SERCA2a and PLB in control and HF samples. In
this and the remaining figures, each C and HF indicate a different control and HF sample, respectively. B, Cumulative data from 8 to 10
separate samples for each group. Density of SERCA2a and PLB were both depressed in HF (P⬍0.05) with no apparent alteration in
stoichiometry. Data are presented as mean⫾SE, n⫽10.
without changing significantly the Vmax. These results suggest
that the global Ca2⫹ alterations observed in isolated HF cells
(Figure 1) may be due to reduced rate of SR Ca2⫹ transport.
SR Ca2ⴙ Pump and Phospholamban Density in HF
We compared the density of the SR Ca 2⫹ -ATPase
(SERCA2a) and its regulatory protein, phospholamban
(PLB), in control and HF. Figure 3A is a Western blot of
SR-enriched microsomes from control and failing dog hearts
probed with antibodies against SERCA2a and PLB. We first
measured the intensity of 18 SERCA2a and 12 PLB bands
from SR of control dogs and normalized each number to
100%; the intensity of the corresponding bands from the same
number of HF samples was then compared against control.
Figure 3B shows that SERCA2a and PLB were significantly
reduced in HF (65⫾7% and 68⫾10% of control, respectively;
P⬍0.05). Thus, a reduction in SERCA2a protein density and
its regulator PLB in HF may be an underlying cause for the
decreased 45Ca2⫹ uptake in SR microsomes (Figure 2A) and
conceivably, for the prolongation of the [Ca2⫹]i transient
observed in intact cells (Figure 1B).
Density and Function of RyRs in Canine HF
We first conducted Western blot analysis and [3H]ryanodine
binding experiments to measure the density of RyRs in
control and HF samples. Neither Western blots (Figure 4A)
nor saturation binding experiments with [3H]ryanodine (Figure 4B) yielded different density values for control and HF
samples. The intensity of the RyR band was 1.0⫾0.10 in
control and 1.07⫾0.11 in HF (n⫽14 samples each). Bmax, the
maximal density of [3H]ryanodine binding sites, was
0.37⫾0.07 and 0.39⫾0.09 pmol/mg protein in control and
HF microsomes, respectively (n⫽13).
We also examined whether Ca2⫹-dependent activation of
[3H]ryanodine binding differs between control and HF. Binding was conducted in the presence of Mg2⫹ and AMP-PCP (a
nonhydrolyzable analogue of ATP), two cytosolic ligands of
RyRs that modulate the RyR response to Ca2⫹. Figure 4C
shows that Ca2⫹ was equally effective in activating RyRs
from control and HF homogenates. Both data sets could be
fitted with the same line. For the range of [Ca2⫹] tested, the
activation curve was sigmoidal with a half-maximal effective
Ca2⫹ concentration (K0.5) ⫽1.8 ␮mol/L. Thus, the Ca2⫹
sensitivity of RyRs does not appear altered in HF (see also
Figure 5).
We reconstituted SR microsomes into planar lipid bilayers
to directly determine the basal activity and the Ca2⫹ sensitivity of RyRs from control and failing hearts (Figure 5). Single
RyRs were stimulated by fast and calibrated elevations of cis
(cytosolic) [Ca2⫹]. Ca2⫹ was elevated from 0.1 to 100 ␮mol/L
with a rapid (␶ ⬇15 ms) microinjection system that exchanged the solution in the vicinity of the RyR (0.5 ␮L)
without altering substantially the composition of the bath
solution (600 ␮L). The Ca2⫹ step produced a sudden increase
of channel activity (Figures 5A and 5B), but then the activity
decayed even though [Ca2⫹] remained elevated (Figure 5D).
The ensemble current generated by summing sweeps of
single channel currents (Figure 5C) revealed that the probability of the channel being open, Po, was high immediately
after the injection, and then slowly decayed to a new
steady-state ⬇2 seconds after the injection (RyR adaptation).17,19 An exponential fit of the ensemble current showed
Jiang et al
Abnormal Ca2ⴙ Release in Heart Failure
5
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
Figure 4. Density and Ca2⫹-sensitivity of RyR in control and failing hearts. A, Western blots of RyR in control and HF canine SR. RyRs
were probed with a specific antibody as described in Materials and Methods section. B, Equilibrium [3H]ryanodine binding curves show
no difference between control and HF SR. C, Ca2⫹ dependence of [3H]ryanodine binding in control and HF. SR-enriched microsomes
were incubated with 7 nmol/L [3H]ryanodine and the indicated [Ca2⫹] at 37°C for 90 minutes. Ca2⫹-dependent binding was similar in
control and HF samples. Thus, both Bmax and K0.5 for Ca2⫹ were identical (n⫽8 dogs of each group).
that the time constant of activation (␶on⫽20 ms, not shown),
the peak Po (0.94⫾0.08), and the time constant of decay
(␶decay⫽1.4⫾0.21 seconds) were identical in control and HF
RyRs. Thus, the basal activity and the capacity of RyRs to
respond to rapid and sustained Ca2⫹ stimuli appears intact in
RyRs from canine failing hearts.
Changes in Human Heart Failure
We also tested for changes in density of SR Ca2⫹ transport
proteins in human HF. Myocardial homogenates and SR
microsomes were obtained from freshly explanted hearts of
12 ICM and 7 DCM patients and compared with the same
from 6 nonfailing (control) individuals whose heart could not
be matched for transplantation. As with the canine model
above, samples from control and HF groups were run sideby-side for a direct comparison of protein expression. Western blots revealed that RyR expression tended to be lower in
HF, but the difference did not reach statistical significance
(control⫽1.0⫾0.089, n⫽5, and HF 0.87⫾0.082, n⫽19;
P⬍0.2) (Figure 6A). Unlike the canine model of HF, there
was no difference in PLB expression, either for the high
molecular weight form (H) (1.0⫾0.085 and 0.97⫾0.13 for
control and HF, respectively) or the low molecular weight
form (L) (1.0⫾0.080 and 0.96⫾0.076 for control and HF,
respectively). However, like human HF, canine HF showed a
significant reduction of SERCA2a (1.0⫾0.067 in control
versus 0.70⫾0.050 in HF; P⬍0.002) (Figure 6B).
Human RyRs from control and HF SR were also quantified
with [3H]ryanodine binding assays and reconstituted in lipid
bilayers to assess their gating kinetics and Po. Like in the
canine model of HF, neither density nor function of RyRs
were different in human HF. Bmax was 0.34⫾0.05 (control
n⫽5), 0.315⫾0.089 (ICM n⫽12), and 0.359⫾0.033 (DCM
n⫽7) (not shown). Figures 7A and 7B show that, when
reconstituted in lipid bilayers, both groups of RyRs had the
same level of activity in the presence of steady concentrations
of Ca2⫹; furthermore, neither population of RyR channels
showed a significant frequency of subconducting states, a
characteristic that would have been found if FKBP12.6 were
dissociated from RyRs.11,20 The current-voltage relationship
for the full-conducting level (⬎96% of all events) was linear
and had a slope of 750 pS for both types of channels (Figure
6
Circulation Research
November 29, 2002
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
Figure 5. Single-channel activity of dog control and HF RyRs. SR-embedded RyRs were fused to lipid bilayers and activated by rapid
steps of Ca2⫹ (0.13100 ␮mol/L) applied to the cis (cytoplasmic) side of the channel. A, Two traces of representative channel activity
from a control RyR or (B) from a HF RyR. Openings are represented as upward deflections of the baseline current. Holding potential⫽30 mV. C, Po of individual channels was calculated from ensemble currents generated using 24 and 28 traces from control and HF
channels, respectively (n⫽ 6 channels from each group). D, Time course and magnitude of the Ca2⫹ step used for activation of RyRs.
See online data supplement for details.
7E). Although there were some detectable subconducting
states (Figures 7A and 7B, arrow in the expanded traces), they
were also present in control RyRs and were not frequent
enough to skew the Gaussian distribution of open events in
the current histograms (Figures 7C and 7D). Instead, subconducting states were abundant when control RyRs were subjected to trypsin digestion (Figure 7F). Addition of trypsin
(0.3 U/mL) to the cis (cytoplasmic) side of the channel first
abolished the fast, full conducting openings, increased mean
open time, and induced fractional openings before causing an
almost complete blockade of channel activity. Hence, subconducting states appear to be relatively rare occurrences in
RyR gating unless channels are subjected to strenuous conditions, eg, limited proteolysis or FKBP12.6 detachment.
RyR Phosphorylation and FK506-Binding Protein
in Heart Failure
Marx et al9 postulated that FKBP12.6 dissociates from RyRs
on phosphorylation, and that this reaction is increased in HF.
We first determined whether PKA phosphorylation of RyRs
causes dissociation of FKBP12.6. Figure 8A, top panel, is an
autoradiogram showing that the phosphorylation protocol
using the catalytic subunit of PKA effectively phosphorylates
control canine RyRs. An ⬇450-kDa band corresponding to
RyRs incorporates [32P-␥]ATP in a saturable and timedependent manner. Phosphorylation is prevented by including
10 ␮mol/L of PKI peptide in the phosphorylation cocktail
(not shown), indicating that the reaction is specific for PKA.
Using this protocol, we back-phosphorylated control and HF
canine RyRs side by side (Figure 8A, bottom panel) to
determine their level of phosphorylation in vivo. Although
some control and HF samples markedly deviated from the
average back-phosphorylation level, the pooled data from
n⫽12 control and 10 HF samples (Figure 8D) yielded no
significant difference. Figure 8B is a Western blot of control
dog and human ventricular homogenates probed with a
polyclonal antibody that recognizes the FKBP12.6 and
FKBP12 isoforms. The homogenate was incubated with the
cocktail of Figure 8A in the absence and the presence
(⫹PKA) of PKA and immediately fractionated in soluble
(Cyt) and membranous (Membr) components to test whether
phosphorylation dissociates either immunophilin from
Jiang et al
Abnormal Ca2ⴙ Release in Heart Failure
7
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
Figure 6. Density of SR Ca2⫹ transport proteins in human control and HF. A, Western blots of RyRs, SERCA2a, and the high (H) and
low (L) molecular weights of PLB from ventricular homogenates. B, Pooled data from 8 control and 12 HF samples. Intensity values of
control samples were normalized to 1.0.
Figure 7. Single-channel activity of human RyRs. A and B, Two-second traces of representative single channel activity from control
and HF RyRs. Activating [Ca2⫹] was nominally free and measured at ⬇5 ␮mol/L. The last trace is an expansion of the indicated segment above, chosen to display the different subconductance states observed (arrows). Scale bar⫽500 ms (x-axis) and 30 pA (y-axis). C
and D, Current histograms constructed with data from 14 control and 16 HF channels from 4 different preparations each. Arrows point
to subconducting states. E, Current-voltage relation for human control and HF channels. Points could be fitted with the same line,
which had a slope conductance⫽750 pS (n⫽4). F, Control canine RyR exhibits subconducting states (middle) before blockade of channel activity by trypsin (bottom). Recording conditions as in previous figures. HP⫽30 mV.
8
Circulation Research
November 29, 2002
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
Figure 8. Phosphorylation of RyRs and FKBP12.x sedimentation. A, Autoradiograms showing the time course of PKA phosphorylation
of RyRs, conducted as described in Materials and Methods section, and back-phosphorylation of 4 control and 4 HF canine SR. B,
Western blots of FKBP12.6 and FKBP12, recognized by the same antibody and separated based on gel retention. Ventricular homogenates of control dogs and humans were phosphorylated with PKA as in (A) and immediately spun at 12 000g for 10 minutes. Aliquots
of supernatant (Cyt) and pellet (Membr) (80 ␮g each) were run in SDS-PAGE, transferred to nitrocellulose membranes and blotted with
FKBP12.x antibody. C, Top, Western blots of homogenates (60 ␮g/lane) of rat hearts stimulated without (⫺) or with isoproterenol
immediately before homogenization. Homogenates were probed with the P-S2809 antibody that recognizes the phosphorylated form of
the RyR only. Bottom, Four control and 4 HF RyRs from dog SR run side-by-side. D, Pooled results from back-phosphorylation and
Western blots experiments. Data from n⫽8 control samples was normalized to 1.0; HF samples (n⫽10) were then compared against
control.
membrane-associated proteins (RyRs). Figure 8B shows that
the distribution pattern of both, FKBP12.6 and FKBP12, is
independent of RyR phosphorylation. Although the
FKBP12.6/FKBP12 ratio differs between dog and human
cardiomyocytes, in agreement with previous reports,21 neither
immunophilin dissociates from membrane-bound proteins
after phosphorylation. Lastly, we used an antibody that
recognizes the phosphorylated form of RyRs only (phosphorylated serine 2809) for a direct assessment of phosphorylation
levels in HF. In rat hearts perfused with isoproterenol (100
nmol/L for 10 minutes) before homogenization, the intensity
of the RyR band is 4-fold higher than in hearts without
isoproterenol treatment (Figure 8C), indicating specificity of
the antibody. Control and HF RyRs from dog homogenates
(Figure 8C) show variable levels of native phosphorylation,
but again, the pooled data from n⫽8 control and 10 HF
samples (Figure 8D) yields no significant difference. Thus,
there was no structural or functional evidence that phosphorylation removes FK506-binding protein from RyRs and that
this process is exaggerated in HF.
Discussion
There is agreement that the depressed cardiac performance
characteristic of HF is likely to result, at least in part, from
abnormal [Ca2⫹]i cycling, but a more controversial issue is the
level of involvement of specific molecules transporters of
Ca2⫹. Although there is substantial evidence suggesting that
upregulation of the sarcolemmal Na⫹-Ca2⫹ exchanger may
depress the [Ca2⫹]i transient in human failing cells and in
animal models of HF, the involvement of the SR has yet to be
defined,21 especially in regard to specific Ca2⫹ transporters,
ie, RyRs, SERCA2a, and PLB. In the present study, we used
the fast-paced canine model of HF and human failing myocardium to carry out a detailed analysis of the density and
main function of SR Ca2⫹ transporters. Previous studies have
measured density of SERCA2a and PLB in canine HF,3
Na⫹-Ca2⫹ exchanger function in human failing cells,7 and
RyR activity in both canine and human HF.9 Our results do
not invalidate previous conclusions; however, if we accept
that the depressed SR Ca2⫹ load observed in HF partly
underlies the blunted [Ca2⫹]i transient, they are only congru-
Jiang et al
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
ent with the idea that abnormal Ca2⫹ uptake, not Ca2⫹ release,
contributes to the contractile dysfunction of HF.
Recently, it was reported that RyRs play a fundamental
role in the abnormal Ca2⫹ cycling that occurs in HF.9
According to this study, PKA phosphorylation of RyRs
dissociates FKBP12.6; as catecholamine levels are increased
in HF,22 PKA would produce hyperphosphorylation of RyRs
and dissociation of FKBP12.6, increasing the channel’s
sensitivity to Ca2⫹ and causing diastolic SR Ca2⫹ leak.9
Upregulation of Na⫹-Ca2⫹ exchange6 – 8 may then extrude the
Ca2⫹ leak. Thus, the overall effect of RyR hyperactivity
would be a depressed SR Ca2⫹ load, which alone could
explain the reduced amplitude and kinetics of the [Ca2⫹]i
transient in HF. As intelligible as it sounds, we failed to detect
key observations that support this scheme. Ca2⫹ sensitivity of
RyRs was not different between control and HF groups
(Figures 4C and 5). Subconductance states, predominant in
FKBP12.6-stripped RyRs,11,20 were equally frequent in control and HF channels (Figures 5 and 7). Phosphorylation of
RyRs by PKA did not appear to dissociate FKBP12.6 (or
FKBP12) from SR-embedded RyRs (Figure 8). Finally, the
effects of rapamycin on single channels and [3H]ryanodine
binding were the same for control and HF RyRs (not shown,
see online Figures 1 and 2 available online at http://www.
circresaha.org in the data supplement), suggesting that
FKBP12.6 remains associated to both groups of channels.
These crucial differences cannot be explained on variability
in models of HF, as Marx et al9 and the present study both
used the tachycardia canine model and failing human myocardium. However, a potential source of discrepancy is that
Marx et al9 used solubilized and immunoprecipitated RyRs
for phosphorylation and single channel recordings, whereas
we used SR-embedded RyRs. It is possible that immunoprecipitation concentrates proteins such as phosphatases, kinases, or proteases and favor reactions that would not
normally occur in more intact preparations. It may be argued
also that we failed to observe hyperphosphorylation of RyRs
and the resultant subconductance states because the SR
microsomes might have been dephosphorylated during isolation, but the presence or the absence of phosphatase inhibitors
(NaF and okadaic acid) in the isolation media did not change
our results (not shown). Also, if abnormal dissociation of
phosphatases would have occurred in HF, as proposed,9 we
would not have detected the same level of RyR-associated
FKBP12.6 in control and HF SR.
␤-Adrenergic receptor (␤-AR) activation produces important inotropic and chronotropic effects on cardiac contraction
via activation of GTP-binding protein (Gs). Gs activates
adenylate cyclase and increases cAMP levels which, in turn,
activate PKA. PKA phosphorylation of sarcolemmal proteins
(particularly DHPR) and SR proteins increase SR Ca2⫹ uptake
and release, and myofibrils (troponin I and protein C)
decrease their Ca2⫹ sensitivity to relax at a faster rate.23 PLB
and the RyR are the main proteins of the SR phosphorylated
by PKA. It is conceivable then, that the PKA-induced
increase of Ca2⫹ uptake and release result from phosphorylation of PLB and RyR, respectively. However, sustained
stimulation of RyRs causes transient increases in Ca2⫹ release
only,24 because the resultant decrease in SR Ca2⫹ load exerts
Abnormal Ca2ⴙ Release in Heart Failure
9
a negative feedback on RyRs, thus attenuating their activity
despite persistent presence of the activator. Furthermore, a
recent study determined that phosphorylation of RyRs is
functionally inconsequential for Ca2⫹ release if SR Ca2⫹
levels remain constant.25 From this perspective, it is improbable that phosphorylation of RyRs alone may account for the
sustained anomalies of Ca2⫹ cycling and the persistently
reduced levels of SR Ca2⫹ observed in HF. Although there is
consensus that the etiology of the contractile dysfunction of
HF is multifactorial, phosphorylation of RyRs as pivotal
cause of depressed SR Ca2⫹ load is unlikely given the
self-correcting mechanisms on Ca2⫹ release that are triggered
by Ca2⫹ inside the SR.
In summary, we found that the abnormal [Ca2⫹]i transients
underlying contractile dysfunction in HF are at least partly
due to depressed SR function. Because RyRs appear structurally and functionally normal in HF, we conclude that
downregulation of SERCA2a function in HF directly leads to
depressed SR Ca2⫹ load and indirectly to reduced Ca2⫹
release. Our results suggest that therapeutic interventions
directed at increasing SR Ca2⫹ uptake may prove more
beneficial than those than nonselectively increase e-c coupling gain.
Acknowledgments
This work was supported by grants from the NIH HL-55438 and
PO1-HL47053 (to H.H.V.) and HL-61534 (to R.A.H., M.R.W., and
H.H.V.) and by AHA grant 0060443Z (to M.T.J.). We wish to thank
Larry Whitesell for technical support in preparing the canine heart
failure model.
References
1. Fabiato A. Time and calcium dependence of activation and inactivation of
calcium-induced release of calcium from the sarcoplasmic reticulum of a
skinned canine cardiac Purkinje cell. J Gen Physiol. 1985;85:247–289.
2. Bers D. Excitation Contraction Coupling and Cardiac Contractile Force.
2nd ed. The Netherlands: Kluwer Academic Publishers; 2000.
3. O’Rouke B, Kass DA, Tomaselli GF, Kaab S, Tunin R, Marban E.
Mechanisms of altered excitation-contraction coupling in canine
tachycardia-induced heart failure, I: experimental studies. Circ Res. 1999;
84:562–570.
4. Beuckelman DJ, Nabauer M, Erdman E. Intracellular calcium handling in
isolated ventricular myocytes from patients with terminal heart failure.
Circulation. 1992;85:1046 –1055.
5. Gómez AM, Valdivia HH, Cheng H, Lederer M, 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.
6. Pieske B, Maier LS, Bers DM, Hasenfus G. Ca2⫹-handling and sarcoplasmic reticulum Ca2⫹ content in isolated failing and nonfailing human
myocardium. Circ Res. 1999;85:38 – 46.
7. Dipla K, Mattiello JA, Margulies KB, Jeevanandam V, Houser SR. The
sarcoplasmic reticulum and the Na⫹-Ca2⫹ exchanger both contribute to
the Ca2⫹ transient of failing human ventricular myocytes. Circ Res.
1999;84:435– 444.
8. Hobai IA, O’Rourke B. Enhanced Ca2⫹-activated Na⫹-Ca2⫹ exchange
activity in canine pacing-induced heart failure. Circ Res. 2000;87:
690 – 698.
9. Marx SO, Reiken S, Hisamatsu Y, Jayaraman T, Burkhoff D, Rosemblit
N, Marks AR. PKA phosphorylation dissociates FKBP12.6 from the
calcium release channel (ryanodine receptor): defective regulation in
failing hearts. Cell. 2000;101:365–376.
10. Ono K, Yano M, Ohkusa T, Kohno M, Hisaoka T, Tanigawa T,
Kobayashi S, Kohno M, Matsuzaki M. Altered interaction of FKBP12.6
with ryanodine receptor as a cause of abnormal Ca2⫹ release in heart
failure. Cardiovasc Res. 2000;48:323–331.
10
Circulation Research
November 29, 2002
11. Xiao R-P, Valdivia HH, Bogdanov K, Valdivia C, Lakatta EG, Cheng H.
The immunophilin FK506-binding protein modulates Ca2⫹ release
channel closure in rat heart. J Physiol. 1997;500:343–354.
12. Wolff MR, Whitesell LF, Moss RL. Calcium sensitivity of isometric
tension is increased in canine experimental heart failure. Circ Res. 1995;
76:781–789.
13. Haworth RA, Hunter DR, Berkoff HA. The isolation of Ca2⫹-resistant
myocytes from the adult rat. J Mol Cell Cardiol. 1980;12:715–723.
14. Lokuta AJ, Rogers TB, Lederer WJ, Valdivia HH. Modulation of cardiac
ryanodine receptors of swine and rabbit by a phosphorylation-dephosphorylation mechanism. J Physiol (Lond). 1995;487:609 – 622.
15. Jiang MT, Narayanan N. Effect of aging on phosphorylation of phospholamban and calcium transport in rat cardiac sarcoplasmic reticulum. Mech
Ageing Dev. 1990;54:87–101.
16. Lokuta AJ, Meyers MB, Sander PR, Fishman G, and Valdivia HH.
Modulation of cardiac ryanodine receptors by sorcin. J Biol Chem. 1997;
272:25333–38.
17. Valdivia HH, Kaplan JL, Ellis-Davies GCR, Lederer WJ. Rapid adaptation of cardiac ryanodine receptors: modulation by Mg2⫹ and phosphorylation. Science. 1995;267:1997–2000.
18. Phillips RM, Narayan P, Gomez AM, Dilly K, Jones LR, Lederer WJ,
Altschuld RA. Sarcoplasmic reticulum in heart failure: central player or
bystander? Cardiovasc Res. 1998;37:346 –351.
19. Gyorke S, Fill M. Ryanodine receptor adaptation: control mechanism of
Ca2⫹-induced Ca2⫹ release in heart. Science. 1993;260:807– 809.
20. Kaftan E, Marks AR, Ehrlich BE. Effects of rapamycin on ryanodine
receptor/Ca2⫹-release channels from cardiac muscle. Circ Res. 1996;78:
990 –997.
21. Jeyakumar LH, Ballester L, Cheng DS, McIntyre JO, Chang P, Olivey
HE, Rollins-Smith L, Barnett JV, Murray K, Xin HB, Fleischer S. FKBP
binding characteristics of cardiac microsomes from diverse vertebrates.
Biochem Biophys Res Commun. 2001;281:979 –986.
22. Brodde OE. ␤1- and ␤2-adrenoceptors in the human heart: properties,
function, and alterations in chronic heart failure. Pharmacol Rev. 1991;
43:203–242.
23. Bers DM. Cardiac excitation-contraction coupling. Nature. 2002;415:
198 –205.
24. Trafford AW, Diaz ME, O’Neill SC, Eisner DA. Integrative analysis of
calcium signalling in cardiac muscle. Front Biosci. 2002;7:D843– 852.
25. Li Y, Kranias EG, Mignery GA, Bers DM. Protein kinase A phosphorylation of the ryanodine receptor does not affect calcium sparks in mouse
ventricular myocytes. Circ Res. 2002;90:309 –316.
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
Abnormal Ca2+ Release, but Normal Ryanodine Receptors, in Canine and Human Heart
Failure
Ming Tao Jiang, Andrew J. Lokuta, Emily F. Farrell, Matthew R. Wolff, Robert A. Haworth and
Héctor H. Valdivia
Circ Res. published online October 24, 2002;
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2002 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/early/2002/10/24/01.RES.0000043663.08689.05.citation
Data Supplement (unedited) at:
http://circres.ahajournals.org/content/suppl/2002/11/17/91.11.1015.DC1
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/
Online Supplement to Jiang, MT et al. Ms #4204/R2
Materials and Methods:
45
Ca2+ uptake measurements.
ATP-dependent, oxalate-facilitated 45Ca2+ uptake by SR microsomes was determined by
the rapid filtration technique, as described1. Briefly, the reaction medium (0.25 mL total volume)
contained (in mmol/L): 50 Tris-maleate (pH 6.8), 5 MgCl2, 2.5 Na2ATP, 120 KCl, 5 K2C2O4
(potassium oxalate), 5 NaN3, 0.1 EGTA, varying concentrations of
45
CaCl2 (free [Ca2+] 0.17 to
4.3 µmol/L), 10 µmol/L ruthenium red (RyR blocker) with and without 40~80 units of the catalytic
subunit of PKA. After preincubation at 37o C for 3 min, freshly thawed SR microsomes (60~90
µg) were added to start the reaction. After 3 min, a 200-µl aliquot was then filtered under mild
vacuum using Millipore filters (pore size 0.45 µm) and washed with 5 ml of ice-cold 10 mmol/L
Tris-maleate (pH 6.8).
The filters were dried and the radioactivity retained counted on a
Beckman LS6500 liquid scintillation counter.
The reaction was linear within 3 min in the
presence of ruthenium red. The initial rate of active Ca2+ uptake was expressed as nmol.mg1
.min-1. Kinetic analysis (Vmax, K0.5 and Hill coefficient) of Ca2+ uptake by Ca2+-ATPase was
done with the Hill equation using the computer program Origin 6.0.
Activation of single RyRs by rapid solution exchange.
Activation of RyRs by fast applications of Ca2+ was accomplished as follows: a 100-µl, 0.5
mm i.d. Teflon needle is introduced into the cis chamber and positioned with a micromanipulator
∼50 µm directly in front of the bilayer aperture (Fig. 1).
A piezo-electric circuit in the
microprocessor (Micropump 1, World Precision Instruments) activates the plunger of the syringe
in step increments to infuse (or withdraw) solutions. The rate of injection is 40 µl s-1, which
corresponds to 12.5 ms for a complete exchange of solution in front of the bilayer (0.5 µl). Faster
solution exchanges may be attained by reducing the bilayer-needle distance or by using syringes
of higher capacity, however, increasing the flux rate increases the probability of breaking the
bilayer membrane. Activation of the piezo-electric circuit charges the bilayer membrane
transiently, allowing recording of spike currents that mark the duration of the injection. In our
configuration, these currents are of reverse polarity to the single channel current and do not
interfere with the analysis of channel kinetics (see traces in Fig. 5 and the expanded traces of Fig.
1 in this supplement). Once the injection is complete, modulators near the bilayer surface remain
constant for several seconds, and then diffuse into the bulk solution until resting conditions are
reestablished. The volume injected (typically 1 µl) produces negligible changes in the composition
of the 600-µl bulk solution, thus allowing injection of modulators several times during the course of
a single experiment. After about 10 injections, the bulk solution is replaced fresh by perfusing the
chamber with a peristaltic pump. [Ca2+] is measured in parallel experiments, by filling the bilayer
aperture with a Ca2+-ionophore resin, as described in Valdivia et al2.
Fig. 1. Main components of the fast injection setup (not to scale). Rapid activation of RyRs is
achieved by exchanging the solution directly in front of the channel with solutions containing Ca2+
and/or other agonists. (B) Single RyR channel activity elicited by a Ca2+ step from 0.1 µM to 100
µM. The downward deflection of the baseline current marks the duration of the pulse. The time
course of the Ca2+ signal was obtained in parallel experiments where the bilayer aperture was
converted into a Ca2+-selective electrode, as done for photolysis of caged Ca2+.
The current trace of Fig. 2 shows that, while the duration of the pulse is 50 ms, the actual
time required for solution exchange does not exceed ∼15 ms. The onset and duration of solution
exchange is signaled by evident shifts of the baseline current (panel A, marked by the vertical
dotted lines), caused by capacitive currents as Ca2+ ions screen the bilayer surface charge and
change the surface potential. The voltage readout of the Ca2+ electrode (panel B) indicates that ≥
95% of the targeted Ca2+ step occurs during this time period, in this particular case, Ca2+ was
jumped from 0.1 → 50 µM in 5 ms and from 0.1 → 96 µM Ca2+ in 15 ms. There is a ∼30 ms time
delay between activation of the piezo-electric circuit (beginning of the current spikes) and the
capacitive current that marks exchange of the solution in contact with the bilayer. This delay likely
corresponds to a flow delay and a mixing/diffusion delay, as observed by Laver & Curtis3 with a
similar set-up.
Fig. 2. Time course of the Ca2+ step produced by
rapid solution exchange as shown in Fig. 1. (A)
Current records during the Ca2+ jump. The
downward, spike-like deflections of the baseline
current mark the duration of the injection (50 ms).
The bilayer potential was clamped at –30 mV.
The vertical dotted lines mark the duration of the
capacitive current, which is shorter than the
duration of the pulse and indicates the time in
which the solution surrounding the channel is
exchanged. (B) The voltage readout from the
Ca2+-selective electrode indicates that ∼95% of
the Ca2+ jump occurs during the capacitive
current.
References:
1. Jiang MT, Narayanan N. Effect of aging on phosphorylation of phospholamban and
calcium transport in rat cardiac sarcoplasmic reticulum. Mech Ageing Dev 1990; 54:87101.
2. Valdivia HH, Kaplan JL, Ellis-Davies GCR, Lederer WJ. Rapid adaptation of cardiac
ryanodine receptors: modulation by Mg2+ and phosphorylation. Science 1995; 267:19972000.
3. Laver DR, and Curtis BA. Response of ryanodine receptor channels to Ca2+ steps
produced by rapid solution exchange. Biophys. J. 1996; 71:732-741.