Abnormal intrastore calcium signaling in chronic
heart failure
Zuzana Kubalova*, Dmitry Terentyev*, Serge Viatchenko-Karpinski*, Yoshinori Nishijima†, Inna Györke*,
Radmila Terentyeva*, Daise N. Q. da Cuñha†, Arun Sridhar‡, David S. Feldman*§, Robert L. Hamlin†,
Cynthia A. Carnes‡¶, and Sandor Györke*¶储
*Department of Physiology and Cell Biology, ¶Davis Heart and Lung Research Institute, †Department of Veterinary Biosciences, §Department of Internal
Medicine, and ‡College of Pharmacy, Ohio State University, Columbus, OH 43210
Edited by Ramon Latorre, Center for Scientific Studies, Valdivia, Chile, and approved August 4, 2005 (received for review May 24, 2005)
Diminished Ca release from the sarcoplasmic reticulum (SR) is an
important contributor to the impaired contractility of the failing
heart. Despite extensive effort, the underlying causes of abnormal
SR Ca release in heart failure (HF) remain unknown. We used a
combination of simultaneous imaging of cytosolic and SR intralu-
minal [Ca] in isolated cardiomyocytes and recordings from single-
ryanodine receptor (RyR) channels reconstituted into lipid bilayers
to investigate alterations in intracellular Ca handling in an exper-
imental model of chronic HF. We found that diastolic free [Ca]
inside the SR was dramatically reduced because of a Ca leak across
the SR membrane, mediated by spontaneous local release events
(Ca sparks), in HF myocytes. Additionally, the magnitudes of
intrastore Ca depletion signals during global and focal Ca release
events were blunted, and [Ca]SR recovery was slowed after global
but not focal Ca release in HF myocytes. At the single-RyR level, the
sensitivity of RyRs to activation by luminal Ca was greatly en-
hanced, providing a molecular mechanism for the maintained
potentiation of Ca sparks (and increased Ca leak) at reduced
intra-SR [Ca] in HF. This work shows that the diminished SR Ca
release characteristic of failing myocardium could be explained by
increased sensitivity of RyRs to luminal Ca, leading to enhanced
spark-mediated SR Ca leak and reduced intra-SR [Ca].
excitation– contraction coupling 兩 ryanodine receptor 兩 sarcoplasmic
reticulum
I n mammalian heart, the major source of Ca required for
contractile activation is the sarcoplasmic reticulum (SR). Ca
sequestered in this organelle can be rapidly released in response
to a small entry of Ca from the extracellular milieu to the cytosol.
This process, termed Ca-induced Ca release, involves specialized
intracellular Ca release channels, the ryanodine receptors
(RyRs), which are clustered in the SR membrane (1, 2). Ca levels
inside the SR determine the functional role of this organelle in
two major ways: firstly, they establish the overall size of the Ca
pool available for release; secondly, SR luminal Ca regulates the
functional state of the RyR Ca release channel. RyR open
probability changes as a direct function of [Ca] at the luminal
side of the channel (3, 4). The responsiveness of RyRs to luminal
Ca seems to be mediated by the auxiliary proteins triadin,
junctin, and calsequestrin (CASQ2), which are coupled to RyRs
at the luminal surface of the SR (5). During the release process,
the increase in cytosolic Ca is accompanied by a simultaneous,
reciprocal decline in intra-SR Ca (6–8). This reduction in [Ca]SR
leads to deactivation or closure of RyRs, contributing to Ca-
induced Ca release termination (9, 10). At the same time,
stimulatory effects of high luminal Ca on RyR channel open
probability are responsible for the activation of a Ca leak
pathway, which plays a role in setting the SR Ca content during
the diastolic phase by leaking excess Ca from the SR (11, 12).
Heart failure (HF) occurs when there is a reduction in cardiac
output that is inadequate to meet the metabolic demands of the
body. The most common defect in HF is impaired contractility
of the ventricles. Evidence suggests that the amount of Ca
14104 –14109 兩 PNAS 兩 September 27, 2005 兩 vol. 102 兩 no. 39
released from the SR into the cytosol is reduced, accounting for,
or contributing to, the reduced contractile force generated by the
failing heart. In a majority of studies, including those in human
cardiomyocytes, reduced SR Ca release has been shown to be
associated with a decrease in the SR Ca content (2, 13–17).
Several explanations have been put forth for the diminished SR
Ca stores in HF. In a number of reports, the reduction of SR Ca
content has been attributed to depressed SR Ca-ATPase func-
tion and兾or enhanced NCX activity during HF (12, 17, 18).
These changes are expected to facilitate Ca removal from the cell
at the expense of its uptake into the SR and result in under-filled
SR Ca stores in HF. Another potential cause of reduced SR Ca
content is enhanced diastolic leak of Ca via the RyRs (18–20).
Although an enhanced RyR-mediated Ca leak could explain the
reduced SR Ca content observed in HF, direct experimental
support for this possibility in myocytes from failing heart is
lacking. Thus, further studies are needed to determine the
mechanisms of altered SR Ca handing in HF.
This report presents an investigation of subcellular and mo-
lecular features of HF in a canine model of chronic tachypacing-
induced HF (21). To investigate the causes of aberrant SR Ca
function in chronic HF, we performed measurements of myocyte
cytosolic and SR luminal Ca, and conducted recordings from
single-RyR channels. This work shows that the diminished SR Ca
release that is characteristic of failing myocardium could be
explained by increased sensitivity of RyRs to luminal Ca, leading
to enhanced spark-mediated SR Ca leakage and reduced
intra-SR [Ca].
Materials and Methods
Chronic, stable left ventricular dysfunction was induced by right
ventricular apical tachypacing (modified Prevail 8086 pace-
maker, Medtronic, Minneapolis) in a canine model for 13–23
months. Myocytes were isolated by using standard techniques.
Cytosolic and intra-SR [Ca] changes were monitored by using
confocal microscopy, and whole-cell currents were recorded with
the patch clamp technique. Single-channel recordings were
carried out as described in ref. 3. Expression of the major SR Ca
handling proteins was determined by immunoblot analysis. De-
tailed material and methods can be found in Supporting Materials
and Methods, which is published as supporting information on
the PNAS web site.
Results
ICa, Intracellular Ca Transients, and SR Ca Content in Intact Myocytes.
Cell capacitance in HF myocytes was dramatically increased
(126 ⫾ 21 pF, n ⫽ 8, in control, and 214 ⫾ 39 pF, n ⫽ 9 in HF
This paper was submitted directly (Track II) to the PNAS office.
Abbreviations: CASQ2, calsequestrin; HF, heart failure; RyR, ryanodine receptor; SR, sarco-
plasmic reticulum.
储To whom correspondence should be addressed at: Davis Heart and Lung Research Institute,
Ohio State University, 473 West 12th Avenue, Columbus, OH 43210. E-mail: gyorke-1@
medctr.osu.edu.
© 2005 by The National Academy of Sciences of the USA
www.pnas.org兾cgi兾doi兾10.1073兾pnas.0504298102

Fig. 1. HF reduces ICa-induced Ca transients and SR Ca load. (A) Represen-
tative confocal line-scan images of Ca transients (Top) along with their spatial
averages (Middle) and recordings of ICa (Bottom), induced by a depolarizing
step from ⫺50 to 0 mV, in control (Left) and HF (Right) myocytes. The voltage
step is depicted above the images. (B) Averages of Ca transient amplitudes and
peak ICa densities as functions of membrane potential for control and HF
myocytes (P ⬍ 0.05; n ⫽ 8 cells from four control hearts; n ⫽ 9 cells from three
HF hearts). (C) Caffeine-induced Ca transients and INCX in control and HF
myocytes (Left). The average amplitudes (F兾F0) of caffeine-induced Ca tran-
sients (Center) were 2.5 ⫾ 0.2 in control and 1.7 ⫾ 0.3 in HF myocytes. The
integrals of INCX (Right) were 47 ⫾ 5 and 31 ⫾ 7 pC for control and HF myocytes,
respectively (P ⬍ 0.05).
myocytes), consistent with ventricular hypertrophy. Depolariza-
tion-induced ICa and spatially resolved intracellular Ca transients
were measured in cardiomyocytes dialyzed with fluo-3. Fig. 1A
shows representative traces of ICa and confocal line-scan images
along with the spatial average of Ca transients recorded during
a depolarizing step from ⫺50 to 0 mV in myocytes from control
and failing hearts. The amplitude of Ca transients was reduced
and their duration prolonged in HF myocytes. Whereas the
amplitude of current density (ICa) did not change, the time
constant of ICa inactivation was slowed in HF myocytes, appar-
ently because of diminished Ca-dependent inactivation of the
L-type Ca channels (Table 1, which is published as supporting
information on the PNAS web site).
Averaged Ca transient and ICa amplitudes are plotted as a
function of voltage in Fig. 1B. The observed decreases in Ca
transient amplitude in HF myocytes with respect to control were
significant through the whole range of membrane potentials
tested (from ⫺40 to 60 mV). At the same time, ICa density was
preserved, suggesting that the reduced SR Ca release is not
simply due to a reduced Ca trigger in myocytes from failing
hearts.
Kubalova et al.
Fig. 2. HF increases Ca spark frequency and reduces both spark amplitude
and SR Ca content in permeabilized myocytes. (A) Representative line-scan
images of Ca sparks recorded in control and HF myocytes. (B and C) Average
frequency (B) and amplitude (C) of sparks (*, P ⬍ 0.01; **, P ⬍ 0.001; n ⫽ 573
events from 43 control cells and 1,416 sparks from 32 HF cells). (D) Represen-
tative recordings of caffeine-induced Ca transients. Average amplitudes of
caffeine-induced Ca transients (⌬F兾F0, mean ⫾ SE) were 2.8 ⫾ 0.2 (n ⫽ 14) and
2.1 ⫾ 0.2 (n ⫽ 9) for control and HF myocytes (P ⬍ 0.05).
PHYSIOLOGY
The SR Ca content was inferred from recordings of Ca transients
and Na兾Ca exchange currents recorded on application of caffeine.
Based on these measurements, the SR Ca content was reduced to
⬇65% of control in HF myocytes (Fig. 1C). Similar changes in
ICa-induced Ca transients and SR Ca content have been described
previously in experimental models of hypertrophy and HF in the
dog (22, 23). The primary focus of the present study was to
determine the factors responsible for the abnormal Ca releasing
and storing capacities of SR in chronic HF.
Ca Sparks and SR Ca Content. To further explore the effects of HF
on the SR Ca release mechanism, we performed measurements
of elementary Ca release events, Ca sparks, in saponin-
permeabilized myocytes (Fig. 2A). The myocytes were bathed in
a fluo-3-containing internal solution at a buffered [Ca] (⬇50
nmol/liter Ca兾0.1 mmol/liter EGTA). In control myocytes, spon-
taneous Ca sparks occurred with an average frequency of ⬇3.0 ⫾
0.3 s⫺1. In myocytes from failing hearts, spark frequency was
increased to 6.6 ⫾ 1.2 (Fig. 2B), whereas event amplitude was
significantly decreased (Fig. 2C). Additionally, although spark
width was not altered there was a small but significant increase
in event duration in HF myocytes compared with controls. These
HF-related changes in Ca sparks were confirmed in intact
myocytes loaded with fluo-3 acetoxymethyl ester (Fig. 8 and
Tables 2 and 3, which are published as supporting information
on the PNAS web site).
Ca spark-mediated leak of Ca from the SR plays an important
role in setting the SR Ca content (11). The observed increase in
frequency of sparks would be expected to result in reduced
sequestration of Ca in the SR in HF myocytes, unless the SR Ca
uptake rate was also increased. The total SR Ca content in the
same two groups of cells was estimated by application of caffeine
(10 mmol兾liter). As shown in Fig. 2D, the amplitude of caffeine-
induced Ca transients was decreased to ⬇75% of control in HF
PNAS 兩 September 27, 2005 兩 vol. 102 兩 no. 39 兩 14105

Fig. 3. HF reduces both local and averaged [Ca] in SR stores. (A) Images of
control and HF permeabilized myocytes with SR-entrapped fluo-5N. (B) En-
larged contour images of subcellular regions illustrating reductions in overall
and spatially localized fluo-5N fluorescence in HF myocytes. (C) Plots of
fluorescence profiles in B at locations indicated by the arrows for control
(black line) and HF (red line) myocytes. Dotted lines represent mean fluores-
cence for control (black) and HF (red) myocytes.
myocytes. These results suggest that Ca spark frequency (and the
associated Ca leak) is increased, potentially accounting for the
reduced SR Ca content in HF myocytes.
Intra-SR [Ca]. To further understand the abnormal SR Ca han-
dling in HF, we performed measurements of intra-SR [Ca] in
permeabilized myocytes. [Ca]SR was monitored with the low
affinity Ca indicator fluo-5N loaded into the SR. Fig. 3 shows
representative two-dimensional (X–Y) images of fluo5N-loaded
control and HF myocytes. Fluo-5N fluorescence intensity, cor-
responding to [Ca]SR, was invariably lower throughout the image
plane in HF myocytes compared with controls (Fig. 3A), indi-
cating that not only focal plane-averaged (Fig. 3 B and C) but also
local diastolic [Ca]SR was reduced in HF myocytes. For simul-
taneous imaging of cytosolic Ca and SR Ca in fluo-5N-loaded
myocytes, rhod-2 was introduced into the bathing solution.
Myocytes incubated in a solution containing 50 nmol兾liter Ca
and 50 ␮mol兾liter EGTA exhibited periodic spontaneous Ca
waves (Fig. 4 A and B). Additionally, Ca sparks similar to those
recorded with fluo-3 could be readily observed, again more
often in HF myocytes than in control myocytes (Fig. 4A). During
Ca release, rhod-2 fluorescence transiently increases, whereas
fluo-5N fluorescence transiently decreases, thus forming mirror
images of Ca waves and sparks in the cytosolic and luminal
compartments, respectively. In both control and HF cells, Ca
waves arise with the same frequency of ⬇0.25 s⫺1 (Fig. 4A). At
the same time, the amplitude of the cytosolically measured Ca
waves was reduced in HF myocytes (Fig. 4B and Table 4, which
14106 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0504298102
is published as supporting information on the PNAS web site),
consistent with the blunted depolarization- and caffeine-induced
Ca transients in HF myocytes. Additionally, the velocity of Ca
wave propagation was reduced in HF myocytes as indexed by the
prolonged slope of the Ca wave front on the myocyte line-scan
images (Fig. 4A). Similar HF-induced changes in properties of
Ca waves were observed in intact myocytes loaded with fluo-3
acetoxymethyl ester (Fig. 8 and Table 5, which are published as
supporting information on the PNAS web site).
Strikingly, the SR-entrapped fluo-5N revealed a dramatic
reduction in the basal [Ca]SR in HF myocytes compared with
controls (Fig. 4 A–C and Table 6, which is published as sup-
porting information on the PNAS web site). Additionally, the
amplitudes of both the global and focal Ca depletion signals (i.e.,
during Ca waves and sparks, respectively) were diminished in HF
cells compared with controls (Figs. 4 A and B and 5 A and B).
The amplitude of the Ca wave depletion signal constituted
⬇45% of the total reduction of fluo-5N fluorescence in caffeine,
indicative of partial SR depletion during Ca wave-associated Ca
release in both control and HF myocytes (10 mmol兾liter, FCaf;
Fig. 4C). Recovery kinetics of [Ca]SR for Ca wave-associated Ca
release were significantly slower in HF myocytes, consistent with
a compromised ability of the SR to sequester and retain Ca in
HF myocytes (Fig. 4C). However, [Ca]SR recovery after focal Ca
release was similar in HF and control myocytes (Fig. 5B). The
differential effects of HF on [Ca]SR recovery after large scale vs.
spatially limited SR Ca release suggests different SR refilling
mechanisms after global and localized Ca release events (24).
Collectively, these results showed that SR luminal [Ca] is re-
duced in HF myocytes. They also demonstrated that generation
of spontaneous Ca waves and sparks in HF myocytes occurs at
abnormally low intra-SR Ca levels, thereby accounting for, or
contributing to, the reduced sequestration of Ca in the SR in HF
myocytes. The increased occurrence of spontaneous release
events at reduced SR Ca content in HF myocytes is reminiscent
of the effects of the RyR-sensitizing agent caffeine (11, 25) and
consistent with potential changes in the functional activity of
RyRs in failing hearts.
Single-RyR Channel Activity. To directly compare the sensitivities
of RyR channels from normal and failing hearts to luminal Ca,
we performed single-RyR channel recordings using the planar
lipid bilayer technique. SR microsomes were isolated from
normal and failing dog hearts by using differential centrifuga-
tions, vesicles were incorporated into planar lipid bilayers, and
single-RyR channels were measured by using Cs⫹ as the charge
carrier in the presence of cytosolic Mg2⫹ and ATP. Represen-
tative single-channel recordings at varying luminal Ca concen-
trations are shown in Fig. 6A, whereas Fig. 6B plots averaged
RyR open probability (Po) as a function of trans [Ca] for control
and HF channels. As demonstrated previously in RyRs from
normal dog hearts (3, 4), increasing trans [Ca] from the micro-
molar to the millimolar range resulted in increased channel
activity (EC50 ⬃ 0.6 mmol兾liter). In RyRs from failing hearts, the
luminal Ca sensitivity was profoundly shifted toward lower
concentrations compared with controls such that the channel
was substantially activated at trans [Ca] as low as 20 ␮mol兾liter
(EC50 ⬃ 5 ␮mol兾liter; Fig. 6). In HF myocytes, reductions in
luminal Ca below 1 ␮mol兾liter were required to abolish luminal
Ca activation. These results showed that RyR sensitivity to
luminal Ca is greatly enhanced in HF.
Levels of SR Proteins. We performed immunoblot analyses to
determine the contents of the major SR Ca handling proteins
cardiac RyR (RyR2), CASQ2, SR CaATPase (SERCA2a),
phospholamban (PLB), triadin, and junctin in cell lysates from
control and failing hearts (Fig. 7). We did not detect changes in
the expression of CASQ2, triadin, junctin, SERCA2a, or PLB,
Kubalova et al.
Fig. 4. Basal and Ca-wave-associated cytosolic and intra-SR [Ca] is altered in HF. (A) Representative line-scan images of rhod-2 (Upper) and fluo-5N (Lower)
fluorescence in control (Left) and HF (Right) myocytes. (B) Time-dependent profiles of a cytosolic (Upper) and intra-SR (Lower) Ca waves in control (Left) and HF
(Right) myocytes. (C) Representative time-dependent profiles of intra-SR Ca waves followed by caffeine-induced intra-SR Ca transients in control and HF myocytes
(black and red trace, respectively). (Inset) comparison of time course of [Ca]SR recovery after Ca wave-associated Ca release.
whereas RyR2 content was significantly reduced in homogenates
from failing hearts. Therefore, altered stoichiometry between
RyR and other proteins of the Ca release channel complex could
contribute to the abnormal Ca handling in HF.
Discussion
We investigated alterations in intracellular Ca handling in
chronic HF using a combination of simultaneous imaging of
cytosolic and SR intraluminal [Ca] in isolated cardiomyocytes
and recordings from single-RyR channels reconstituted into
lipid bilayers. We found that diastolic [Ca] inside the SR was
dramatically reduced because of increased Ca spark-mediated
Ca leak across the SR membrane in HF myocytes. At the
single-RyR level, the sensitivity of RyRs to activation by luminal
Ca was greatly enhanced, providing a molecular mechanism for
the maintained potentiation of Ca sparks (and increased Ca
leak) despite a reduced intra-SR [Ca] in HF. These results show
that defective intrastore Ca signaling is involved in the patho-
genesis of HF.
Defective RyR Luminal Ca Regulation as a Cause of Abnormal Ca
Handling in HF. [Ca] inside the SR is an important determinant of
the functional activity of the RyR Ca release channels (5, 26).
Normally, RyR luminal Ca regulation appears to play a stabi-
lizing role by countering the intrinsic positive feedback of
Ca-induced Ca release during the release process. Specifically,
dissociation of Ca from the luminal sensing sites is involved in
termination of SR Ca release, permitting cardiac relaxation (9,
10, 27). Additionally, RyRs sensitive to [Ca]SR provide a luminal
Ca-dependent leak pathway and play a role in regulating SR Ca
content in cardiac muscle (11, 12). Presumably, this mechanism
Kubalova et al.
PHYSIOLOGY
operates to prevent SR Ca overload by releasing excess Ca from
the store (11).
Our finding that, in HF, RyRs become hypersensitive to
luminal Ca helps to reconcile some apparently contradictory
results and improves our understanding of abnormal Ca handling
in the failing heart. Previous studies in normal myocytes have
shown that the frequency of Ca sparks and the incidence of Ca
waves are increased when the SR Ca load is elevated, whereas
reduced SR Ca content is accompanied by reduced Ca spark and
Ca wave occurrence. Yet in HF, an increased frequency of sparks
is associated with reduced SR Ca content (ref. 23 and this work).
Our demonstration of enhanced RyR activity at a given luminal
[Ca] provides a logical explanation for these seemingly discrep-
ant results. Moreover, our finding may explain the paradoxically
increased incidence of arrhythmogenic Ca waves and delayed
after depolarizations in various models of HF, despite the
reduced SR Ca content (28–30). Of note, we did observe
premature ventricular depolarizations in some of the HF dogs,
consistent with an arrhythmogenic response to impaired calcium
regulation during HF (Supporting Materials and Methods).
Comparison with Previous HF Models. Whereas the observed re-
duction and prolongation of SR Ca release are hallmarks of
failing myocardium (2, 14–16), the effects of HF on other EC
parameters, including ICa and SR Ca load, vary widely between
different species and models. Our finding that alterations of SR
Ca release occurred in the absence of significant changes in ICa
density but in parallel with a decrease in SR Ca content is
consistent with results obtained previously in shorter term (3–6
weeks) canine models of tachypacing-induced HF (22, 31).
Similar to a recent report by Song et al. (23) in which a canine
model of left ventricular hypertrophy due to chronic pressure
PNAS 兩 September 27, 2005 兩 vol. 102 兩 no. 39 兩 14107
Fig. 5. Simultaneous measurements of focal cytosolic and intra-SR Ca during
spontaneous Ca sparks. (A) Representative line-scan images of rhod-2 (Upper)
and fluo-5N (Lower) fluorescence in control (Left) and HF (Right) myocytes. (B)
Time courses of the local cytosolic and intra-SR Ca signals. (Inset) Comparison
of time course of [Ca]SR recovery after spark-associated Ca release in control
(black) and HF (red) myocytes. Fig. 6. RyR sensitivity to luminal [Ca] is increased in failing hearts. (A) Repre-
sentative traces of current recordings in RyR channels from control (Left) and HF
(Right) samples at different trans Ca concentrations (as indicated above the
overload was used, our experiments showed an increase in the traces). Single-channel activities were determined at a ⫹40-mV holding poten-
frequency of sparks in HF myocytes. In the study by Song et al., tial. Experimental solutions contained 350 mmol兾liter CsCH3SO3, 0.07 mmol兾liter
the increase in the Ca spark rate was rationalized by proposing CaCl2, 3 mmol兾liter MgATP, and 20 mmol兾liter Hepes (pH 7.4) on the cytosolic (cis)
that local regions with elevated SR Ca content exist in HF side of the bilayer, and 20 mmol兾liter CsCH3SO3, 0.0001–2 mmol兾liter CaCl2, 0 –1
myocytes that escape detection by the caffeine method used to mmol兾liter EGTA, and 20 mmol兾liter Hepes (pH 7.4) on the luminal (trans) side of
measure total SR Ca content. However, our direct [Ca]SR the bilayer. Channel openings are shown as upward deflections from the closed
level. (B) Open probability (Po) of RyR channels from control and failing hearts as
measurements demonstrated that [Ca] was lowered throughout
a function of trans [Ca]. Data are presented as means ⫾ SEM for three to eight
the entire SR network, and that the increased Ca spark rate was control and two to four failing hearts. Continuous lines were obtained by fitting
a result of enhanced sensitivity of RyRs to luminal Ca in the data with sigmoidal functions.
myocytes from failing hearts. It is unknown whether calcium
dysregulation and the underlying mechanisms differ between left
ventricular hypertrophy and HF. RyR modulation by luminal Ca appears to involve auxiliary
In previous studies measuring RyR function in lipid bilayers,
proteins such as CASQ2, triadin, and junctin that are removed
no differences in channel Po between preparations from control
by RyR purification (5) (Fig. 9, which is published as supporting
and failing canine hearts were reported (31). This result is in
contrast to the altered RyR function that we observed. We
suggest that the substantially longer duration of HF in our study
may contribute to this difference. Alternatively, it is possible that
the experimental conditions may underlie this discrepancy. We
note that in our experiments, differences in channel behavior
were observed only in the presence of cytosolic MgATP (un-
published results). This observation would be consistent with
previous reports that the effects of luminal Ca on RyR activity
rely on the presence of an allosteric activating ligand (ATP,
caffeine, sulmazole) (3, 4).

Molecular Mechanisms of Abnormal RyR Function in HF. The specific

molecular causes for altered RyR function in failing hearts are
not known. Marks and coworkers (19) have suggested that RyR
phosphorylation by PKA results in dissociation of FKBP12.6
from RyRs, rendering the channels hyperactive with associated
Ca leak in HF. Other laboratories, however, have found no
evidence for phosphorylation-induced detachment of FKBP12.6 Fig. 7. Expression of Ca handling proteins in normal and failing hearts.
from RyRs in normal and diseased hearts (31–34). Thus, further Shown are representative immunoblots of RyR2, CASQ2, SERCA2a, PLB, junc-
studies are needed to determine whether the abnormal RyR tin, and triadin. Averaged optical density of RyR2 from failing hearts was
modulation by luminal Ca in HF is caused by an altered reduced to 52% of control (n ⫽ 7; P ⬍ 0.05), whereas there were no significant
interaction with FKBP12.6. changes in expression levels of other proteins (n from 6 to 10).

14108 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0504298102 Kubalova et al.

information on the PNAS web site); alterations in the expression
or function of these regulatory proteins may contribute to our
observations. Although CASQ2 levels do not change in various
models of HF (ref. 2 and this work), CASQ2 glycosylation is
dramatically altered in HF, indicative of a potential defect in
junctional SR trafficking and Ca release complex assembly (35).
There is little information regarding the expression, localization,
and function of triadin and junctin in failing hearts. In the
present study, the levels of these proteins did not change in HF.
However, when considering the decrease in RyR abundance to
approximately half of control values, the ratios of both triadin
and junctin to RyRs increased ⬇2-fold in failing hearts. This
altered stoichiometry of triadin and junctin to RyRs could
contribute to increased RyR function in HF, because both
triadin and junctin appear to act on RyRs by increasing RyR
activity (5, 36).
Finally, abnormal RyR function in HF could be due to
acquired defects in the channel protein caused by the altered
intracellular environment in HF. HF is accompanied by in-
creased formation of metabolic by-products, including reactive
oxygen and nitrogen species and aldehydes, with potentially
deleterious effects on intracellular proteins (37). These com-
pounds are capable of reacting with lysine, cysteine, tyrosine,
and histidine residues, resulting in long-term or permanent
modifications in the RyR structure. Because RyR is a long-lived
protein (38), it is particularly vulnerable to such acquired
chemical modifications. Of note, mutations of RyR have been
reported to sensitize RyR to luminal Ca (39). It is conceivable
that posttranslational modifications in the protein structure can
lead to similar functional changes. Regardless of its specific
biochemical cause, abnormal activation of RyR by luminal Ca is
1. Fabiato, A. (1985) J. Gen. Physiol. 85, 247–289.
2. Bers, D. M. (2001) in Excitation-Contraction Coupling and Cardiac Contractile
Force (Kluwer Academic, Dordrecht, The Netherlands), 2nd Ed.
3. Györke, I. & Györke, S. (1998) Biophys. J. 75, 2801–2810.
4. Ching, L. L., Williams, A. J. & Sitsapesan, R. (2000) Circ. Res. 87, 201–206.
5. Györke, I., Hester, N., Jones, L. R. & Györke, S. (2004) Biophys. J. 86,
2121–2128.
6. Shannon, T. R., Guo, T. & Bers, D. M. (2003) Circ Res. 93, 40–45.
7. Kubalova, Z., Györke, I., Terentyeva, R., Viatchenko-Karpinski, S., Terentyev,
D., Williams, S. C. & Györke, S. (2004) J. Physiol. 561, 515–524.
8. Brochet, D. X., Yang, D., Di Maio, A., Lederer, W. J., Franzini-Armstrong, C.
& Cheng, H. (2005) Proc. Natl. Acad. Sci. USA 102, 3099–3104.
9. Terentyev, D., Viatchenko-Karpinski, S., Valdivia, H. H., Escobar, A. L. &
Györke, S. (2002) Circ. Res. 91, 414–420.
10. Sobie, E. A., Dilly, K. W., dos Santos Cruz, J., Lederer, W. J. & Jafri, M. S.
(2002) Biophys. J. 83, 59–78.
11. Lukyanenko, V., Viatchenko-Karpinski, S., Smirnov, A., Wiesner, T.F. &
Györke, S. (2001) Biophys. J. 81, 785–798.
12. Shannon, T. R., Ginsburg, K. S. & Bers, D. M. (2002) Circ Res. 91, 594–600.
13. Lindner, M., Erdmann, E. & Beuckelmann, D. J. (1998) J. Mol. Cell. Cardiol.
30, 743–749.
14. O’Rourke, B., Kass, D. A., Tomaselli, G. F., Kääb, S., Tunin, R. & Marbán, E.
(1999) Circ. Res. 84, 562–570.
15. Houser, S. R., Piacentino, V., III, & Weisser, J. (2000) J. Mol. Cell. Cardiol. 32,
1595–1607.
16. Hasenfuss, G. & Pieske, B. (2002) J. Mol. Cell. Cardiol. 34, 951–969.
17. Houser, S. R. & Margulies, K. B. (2003) Circ. Res. 92, 350–358.
18. Shannon, T. R., Pogwizd, S. M. & Bers, D. M. (2003) Circ. Res. 93, 592–594.
19. Marx, S. O., Reiken, S., Hisamatsu, Y., Jayaraman, T., Burkhoff, D., Rosemblit,
N. & Marks, A. R. (2000) Cell 101, 365–376.
20. Bers, D. M., Eisner, D. A. & Valdivia, H. H. (2003) Circ. Res. 93, 487–490.
21. Nishijima, Y., Feldman, D. S., Bonagura, J., Ozkanlar, Y., Jenkins, P. J.,
Lacombe, V. A., Abraham, W. T., Hamlin, R. L. & Carnes, C. A. (2005)
J. Cardiac Failure, in press.
Kubalova et al.
an important factor in the pathogenesis of HF. Determining the
precise nature of RyR defects in HF will require an understand-
ing of the molecular basis of RyR regulation by luminal Ca. The
putative subcellular mechanisms revealed in the present study
should be considered in conjunction with global phenomena
such as desynchronization of myocardial contraction that may
also contribute to impaired cardiac function in this model of
HF (21).
Conclusions
Using a canine model of chronic HF, we showed that the
diminished SR Ca release characteristic of failing myocardium is
at least partially due to increased sensitivity of RyRs to luminal
Ca, leading to enhanced spark-mediated SR Ca leak and reduced
intra-SR [Ca]. The biochemical causes for the abnormal RyR
luminal Ca-dependent gating behavior are not known. They
could involve disrupted protein–protein interactions within the
RyR complex, altered covalent modification of RyRs such as
phosphorylation, or acquired defects caused by reactive intra-
cellular metabolites. Sensitization of RyRs to luminal Ca in HF
could reflect an adaptive transformation, allowing the SR to
operate at reduced SR Ca loads (and reduced energy costs) in
the ATP-starved failing myocardium. Alternatively, the RyR
sensitization may play a pathophysiologic or maladaptive role in
HF. More studies are needed to determine the nature of the
altered intrastore control of Ca release in HF. This defect in RyR
function may provide a novel therapeutic target to treat con-
tractile dysfunction in HF.
This work was supported by the American Heart Association (S.V.-K.
and C.A.C.) and National Institutes of Health Grants HL074045 and
HL063043 (to S.G.).
PHYSIOLOGY
22. Hobai, I. A. & O’Rourke, B. (2001) Circulation 103, 1577–1584.
23. Song, L. S., Pi, Y., Guatimosim, S., Kim, S. J., Yatani, A., Kudej, R. K., Zhang,
Q., Hittinger, L., Ghaleh, B., Cheng, H., et al. (2005) Circ. Res., in press.
24. DelPrincipe, F., Egger, M. & Niggli, E. (1999) Nat. Cell Biol. 1, 323–329.
25. Trafford, A. W., Sibbring, G. C., Diaz, M. E. & Eisner, D. A. (2000) Cell
Calcium 28, 269–276.
26. Györke, S., Györke, I., Lukyanenko, V., Terentyev, D., Viatchenko-Karpinski,
S. & Wiesner, T. F. (2002) Front. Biosci. 7, d1454–d1463.
27. Terentyev, D., Viatchenko-Karpinski, S., Györke, I., Volpe, P., Williams, S. C.
& Györke, S. (2003) Proc. Natl. Acad. Sci. USA 100, 11759–11764.
28. Tomaselli, G. F., Beuckelmann, D. J., Calkins, H. G., Berger, R. D., Kessler,
P. D., Lawrence, J. H., Kass, D., Feldman, A. M. & Marbán, E. (1994)
Circulation 90, 2534–2539.
29. Pogwizd, S. M., Schlotthauer, K., Li, L., Yuan, W. & Bers, D. M. (2001) Circ.
Res. 88, 1159–1167.
30. Pogwizd, S. M. & Bers, D. M. (2004) Trends Cardiovasc. Med. 14, 61–66.
31. Jiang, M. T., Lokuta, A. J., Farrell, E. F., Wolff, M. R., Haworth, R. A. &
Valdivia, H. H. (2002) Circ. Res. 91, 1015–1022.
32. Stange, M., Xu, L., Balshaw, D., Yamaguchi, N. & Meissner, G. (2003) J. Biol.
Chem. 278, 51693–51702.
33. Xiao, B., Sutherland, C., Walsh, M. P. & Chen, S. R. (2004) Circ. Res. 94,
487–495.
34. Xiao, B., Jiang, M. T., Zhao, M., Yang, D., Sutherland, C., Lai, F. A., Walsh,
M. P., Warltier, D. C., Cheng, H. & Chen, S. R. (2005) Circ. Res. 96, 847–855.
35. Kiarash, A., Kelly, C. E., Phinney, B. S., Valdivia, H. H., Abrams, J. & Cala,
S. E. (2004) Cardiovasc. Res. 63, 264–272.
36. Terentyev, D., Cala, S. E., Houle, T. D., Viatchenko-Karpinski, S., Györke, I.,
Terentyeva, R., Williams, S. C. & Györke, S. (2005) Circ. Res. 96, 651–658.
37. Uchida, K. (2000) Free Radical Biol. Med. 28, 1685–1696.
38. Ferrington, D. A., Krainev, A. G. & Bigelow, D. J. (1998) J. Biol. Chem. 273,
5885–5891.
39. Jiang, D., Xiao, B., Yang, D., Wang, R., Choi, P., Zhang, L., Cheng, H. & Chen,
S. R. (2004) Proc. Natl. Acad. Sci. USA 101, 13062–13067.
PNAS 兩 September 27, 2005 兩 vol. 102 兩 no. 39 兩 14109