Intra–Sarcoplasmic Reticulum Free [Ca2+] and Buffering in Arrhythmogenic Failing Rabbit Heart
Smaller Ca2+ transients and systolic dysfunction in heart failure (HF) can be largely explained by reduced total sarcoplasmic reticulum (SR) Ca2+ content ([Ca]SRT). However, it is unknown whether low [Ca]SRT is manifest as reduced: (1) intra-SR free [Ca2+] ([Ca2+]SR), (2) intra-SR Ca2+ buffering, or (3) SR volume (as percentage of cell volume). Here we assess these possibilities in a well-characterized rabbit model of nonischemic HF. In HF versus control myocytes, diastolic [Ca2+]SR is similar at 0.1-Hz stimulation, but the increase in both [Ca2+]SR and [Ca]SRT as frequency increases to 1 Hz is blunted in HF. Direct measurement of intra-SR Ca2+ buffering (by simultaneous [Ca2+]SR and [Ca]SRT measurement) showed no change in HF. Diastolic [Ca]SRT changes paralleled [Ca2+]SR, suggesting that SR volume is not appreciably altered in HF. Thus, reduced [Ca]SRT in HF is associated with comparably reduced [Ca2+]SR. Fractional [Ca2+]SR depletion increased progressively with stimulation frequency in control but was blunted in HF (consistent with the blunted force–frequency relationship in HF). By studying a range of [Ca2+]SR, analysis showed that for a given [Ca]SR, fractional SR Ca2+ release was actually higher in HF. For both control and HF myocytes, SR Ca2+ release terminated when [Ca2+]SR dropped to 0.3 to 0.5 mmol/L during systole, consistent with a role for declining [Ca2+]SR in the dynamic shutoff of SR Ca2+ release. We conclude that low total SR Ca2+ content in HF, and reduced SR Ca2+ release, is attributable to reduced [Ca2+]SR, not to alterations in SR volume or Ca2+ buffering capacity.
- excitation–contraction coupling
- heart failure
- sarcoplasmic reticulum
- intra-SR Ca2+ buffering
Defective intracellular Ca2+ handling may play a central pathophysiological role in heart failure (HF).1 Contractile dysfunction in end-stage HF has been attributed to reduced Ca2+ transients, which could be caused by decreased Ca2+ current (ICa), reduced excitation–contraction coupling efficacy between ICa, and sarcoplasmic reticulum (SR) Ca2+ release (excitation–contraction coupling)2,3 or reduced SR Ca2+ load ([Ca]SRT).4–6 Most studies find that peak ICa,L density and fractional SR Ca2+ release are not depressed in HF,4–6 despite lower [Ca]SRT, which by itself would tend to reduce fractional SR Ca2+ release.7–10
Lower [Ca]SRT has been shown in animal HF models and also in human HF.4–6,11–13 This is caused by reduced SR Ca2+-ATPase (SERCA) function, increased Na+–Ca2+ exchange (NCX) function, and increased SR Ca2+ leak (to different extents in different HF models; see the online data supplement at http://circres.ahajournals.org).4–6,11–15 Although these changes in Ca2+ transport cause the reduced [Ca]SRT in HF, the reduced [Ca]SRT could be manifest as (1) reduced free SR Ca2+ ([Ca2+]SR) throughout the SR, which could decrease the driving force for SR Ca2+ release and ryanodine receptor (RyR) activation by luminal Ca2+; (2) reduced intra-SR Ca2+ buffering capacity; and/or (3) reduced volume percent of cell occupied by SR (with normal [Ca2+]SR).16 To distinguish among these possibilities, one needs to directly measure [Ca2+]SR and intra-SR Ca2+ buffering in intact ventricular myocytes in HF versus control, which has not been done previously (even in control myocytes).
The present study aimed to assess the aforementioned manifestations of reduced [Ca]SRT in HF. We measured [Ca2+]SR and intra-SR Ca2+ buffering in a well-characterized rabbit model of nonischemic HF (induced by aortic insufficiency and aortic constriction).5,15,17 These HF rabbits exhibit nonsustained ventricular tachycardias (90%, with 10% incidence of sudden death), a 2-fold functional overexpression of NCX, and myocytes have reductions in contraction, Ca2+ transient amplitude, and [Ca]SRT. Here we measured [Ca2+]SR, [Ca]SRT, and [Ca2+]i simultaneously in ventricular myocytes using the low-affinity Ca2+ indicator fluo-5N (for [Ca2+]SR)18 and fura-2 (for [Ca2+]i). We found that the reduction in [Ca]SRT in HF can be explained by reduced [Ca2+]SR, without substantial changes in either intra-SR buffering or inferred SR volume percentage.
Materials and Methods
Rabbit HF Model and Cardiac Myocyte Isolation
Studies were performed on myocytes isolated as previously described1 from control (n=15) and HF (n=9) adult New Zealand White rabbits of either sex. HF was induced by combined aortic insufficiency and aortic constriction as described5,17 and studied ≈6 months later. Protocols were approved by Animal Studies Committees of Loyola University Chicago and University of Illinois at Chicago. Additional methodological details are in the online data supplement.
Fluorescent Measurement of [Ca2+]SR and In Vivo Calibration
Isolated ventricular myocytes were loaded with 10 μmol/L fluo-5N acetoxymethyl ester (Molecular Probes).18 Experiments were performed at 23°C on an epifluorescence microscope. Fmax was determined for each cell as described.18 After stimulation at 1 Hz with 1 μmol/L isoproterenol, 0.5 mmol/L tetracaine was added to block SR Ca2+ leak and extracellular Na+ was removed to raise [Ca2+]SR via NCX and SERCA. Fluorescence during 10 mmol/L caffeine application depleted [Ca]SRT and was defined as Fmin (or Fcaff), and all fluorescence signals were after Fcaff subtraction. Caffeine-sensitive diastolic fluorescence at 0.5 Hz was defined as F0, and [Ca2+]SR at 0.5 Hz ([Ca2+]SR-rest) was Kd(F0−Fmin)/(Fmax−F0). At other times, [Ca2+]SR=([F/F0]Kd)/(Kd/[Ca2+]SR-rest)−F/F0+1, where Kd was 400 μmol/L.18 Fluorescence signals for data analysis were averaged from 4 transients.
Western Blot Analysis
Left ventricular tissue from control and HF rabbits were subjected to Western blotting15 using primary antibodies to calsequestrin (CSQ) (Santa Cruz Biotechnology) and GAPDH (Chemicon).
Fluorescent Measurements of [Ca2+]i and SR Ca2+ Load [Ca]T
Myocytes were loaded with fura-2 acetoxymethyl ester, and fura-2 was excited at 340±10 and 380±10 nm, and emission (535±20 nm) was recorded at 100 Hz. Background-subtracted fluorescence ratio (R=F340/F380) was converted to [Ca2+]i, and [Ca]SRT was relative to nonmitochondrial cell volume (see the online data supplement).7,8,19
SERCA Ca2+ Flux Analysis
Instantaneous SERCA-mediated Ca2+ flux (JSR) was calculated, where [Ca2+]i and [Ca2+]SR were measured simultaneously: equation
where Kmf and Kmr are Km values for forward and reverse SERCA fluxes, respectively; k is the rate constant for leak flux (Jleak); Vmax is maximum flux rate; and H is Hill coefficient.
For control myocytes, parameters are used as previously described.19,20 In HF, Vmax was reduced by 20% (as in our experiments, see under Results) and k was adjusted to obtain ≈4 μmol/L sec−1 Jleak at diastolic [Ca2+]SR at 0.5 Hz (when [Ca2+]SR≈630 μmol/L; see under Results).13
Alternatively, JSR was calculated using measured [Ca2+]SR and SR buffering properties. Because CSQ has very low Ca2+ affinity, we considered it an instantaneous buffer, such that: [Ca]SRT=[Ca2+]SR+Bmax-SR[Ca2+]SR/([Ca2+]SR+Kd-SR), where Bmax-SR and [Ca2+]SR were measured and Kd-SR=650 μmol/L SR volume (assuming CSQ is the main intra-SR buffer).1,20 Then JSR=d[Ca]SRT/dt was calculated at each time point.
The data are means±SEM. Significant differences were assessed using Student t test at P<0.05.
Characterization of Stimulated and Caffeine-Induced Ca2+ Transients
Figure 1 shows Ca2+ transients in control and HF myocytes. Consistent with prior work using this HF model,5,17 Ca2+ transient amplitude was reduced in HF versus control at 0.5 and 1 Hz stimulation (by 25% and 57%, respectively; Figure 1C), but diastolic [Ca2+]i was unaltered (116±15 nmol/L in HF, 140±17 nmol/L in control). After steady-state pacing at 0.5 and 1 Hz, caffeine-induced Ca2+ transients also showed [Ca]SRT reduction by 24% and 30% in HF versus control (P<0.05; Figure 1B and 1D).
The rate of [Ca2+]i decline during caffeine application depends mainly on NCX-mediated Ca2+ extrusion, whereas twitch [Ca2+]i decline is more SERCA dependent in rabbit myocytes.17,21 The mean time constant (τ) of NCX-mediated [Ca]i decline (τcaff=τNCX) was faster in HF versus control (1.3±0.1 versus 2.0±0.1 s) after 0.5 Hz of stimulation (Figure 1E). However, twitch τ (τtwitch) was similar (636±54 ms in HF versus 670±60 ms in control). If twitch [Ca]i decline is attributed to parallel SERCA and NCX function and k is the rate constant (k=1/τ), then ktwitch=kSR+kNCX.17 In HF myocytes, ktwitch=1.70±0.09 sec−1 and kNCX=0.81±0.07 sec−1, whereas in control myocytes, ktwitch=1.69±0.16 sec−1 and kNCX=0.54±0.03 sec−1. This indicates a 23% reduction in SERCA function (kSR=0.89 versus 1.15 sec−1, HF versus control) but a 50% increase in NCX function. Therefore, enhanced NCX activity in HF compensates for SERCA to allow unaltered rate of twitch [Ca2+]i decline overall (whereas both effects contribute to reduced [Ca]SRT).13 Figure 1F shows the [Ca2+]i dependence of transport, indicating that (as a function of [Ca2+]i) overall Ca2+ removal (by SERCA+NCX) was comparable in control and HF, but SR and NCX transport were more evenly balanced in HF, consistent with previous studies in this HF model.17
Decreased [Ca2+]SR in HF Myocytes
To test whether [Ca2+]SR is decreased in HF, we measured [Ca2+]i and [Ca2+]SR simultaneously using fura-2 and fluo-5N.18 Figure 2A shows typical fluorescence changes during twitches and caffeine-induced Ca2+ transients. When [Ca2+]i transiently rises (top), fluo-5N fluorescence transiently decreases (bottom). Fmin is taken as Fcaff, whereas F0 is the caffeine-sensitive diastolic fluo-5N fluorescence at 0.5 Hz (Fd-0.5Hz). To calibrate fluo-5N signals, maximal fluorescence (Fmax) was measured in situ from the same myocyte (Figure 2B). Briefly, isoproterenol was first added at 1 Hz (which by itself caused fluorescence to approach Fmax in control myocytes; compare Figure 2A and 2B). Then RyR Ca2+ release was inhibited by 0.5 mmol/L tetracaine, and finally Na+-free solution was applied. This protocol pushes SERCA toward its thermodynamic limit.18 Figure 2B (top) shows that tetracaine initially blocked SR Ca2+ release, but massive SR Ca2+ loading induced by Na+-free solution allowed some SR Ca2+ release (see Figure 2B, top). Despite this, intra-SR fluorescence (F/F0; Figure 2B, bottom) did not further increase, indicating saturation of intra-SR fluo-5N (taken as Fmax, after correcting for fluorescence quench by tetracaine).
Fmax was measured in each HF and control myocyte (1.76±0.15×F0 in HF, 1.54±0.09×F0 in control; P=0.26). The higher Fmax/F0 in HF myocytes directly indicated that at 0.5 Hz, intra-SR fluo-5N was less saturated in HF versus control myocytes (57% versus 65%) and that [Ca2+]SR was lower in HF (regardless of absolute calibrations).
Figure 3A shows normalized representative cytosolic and intra-SR Ca2+ transients (with SR signal inverted). When the raw signals were superimposed, kinetic differences were apparent (left). However, when both signals were calibrated in micromoles per liter, there was less disparity in kinetics. This is likely attributable to the fura-2 signal being fairly linear versus [Ca]i (except near the Ca2+ transient peak) and fluo-5N being in the nonlinear range (above Kd) for most of the excursion (and least so at the [Ca2+]SR nadir). Although some discrepancy might be expected, calibrated signals are required for such comparisons.
Figure 3B and 3C illustrates calibrated [Ca2+]SR signals at 2 pacing rates in control and HF myocytes. In control, doubling frequency increased both diastolic [Ca2+]SR and the extent of [Ca2+]SR depletion, but in HF, neither diastolic [Ca2+]SR nor depletion significantly changed (and we saw no distinct net diastolic leak phase in steady-state [Ca2+]SR traces). Figure 4A shows how diastolic [Ca2+]SR and [Ca]SRT were simultaneously measured using diastolic fluo-5N and caffeine-induced fura-2 transients. To test whether fluo-5N significantly affects SR Ca2+ load (as an additional intra-SR buffer), we compared the [Ca]SRT in myocytes with both fura-2 plus fluo-5N or fura-2 alone (Figure 4A, bottom). Fluo-5N did not affect [Ca]SRT assessed by fura-2 and, therefore, was not a major SR Ca2+ buffer (consistent with our recent estimate using this approach22 that intra-SR [fluo-5N] is ≈50 μmol/L or ≈1% of expected endogenous SR Ca2+-binding sites).
Figure 4B shows that at very low frequency (0.1 Hz), diastolic [Ca2+]SR was not different between control and HF (≈0.48 mmol/L). At 0.5 Hz, control [Ca2+]SR nearly doubled (to 0.93±0.17 mmol/L), whereas in HF, [Ca2+]SR only increased 31% (to 0.63±0.07 mmol/L; P=0.03 versus control). At 1 Hz, the more blunted rise in [Ca2+]SR in HF was still apparent (0.79±0.06 mmol/L in HF versus 1.3±0.13 mmol/L in control; P=0.03). Figure 4B also shows that changes in diastolic [Ca]SRT (blue bars) closely paralleled the changes in [Ca2+]SR (no significant difference at 0.1 Hz but blunted frequency-dependent increase in [Ca]SRT in HF). This parallel behavior between [Ca2+]SR and [Ca]SRT and similar depression in HF (Figure 4B) already suggests that reduced [Ca2+]SR is sufficient to explain the reduced [Ca]SRT in HF.
The extent of [Ca2+]SR depletion (in micromoles per liter) during the twitch also increased progressively with frequency in control myocytes (Figure 4C) but was significantly blunted in HF. The same applied for the fractional [Ca2+]SR depletion (Figure 4D). This is consistent with the less positive force–frequency relationships typical of human HF.1,11
Although positive feedback in Ca2+-induced Ca2+ release might be expected to empty the SR, Figure 4E shows that SR Ca2+ release terminated when [Ca2+]SR dropped to 0.3 to 0.5 mmol/L during systole in both nonfailing and HF myocytes. Clearly, complete SR Ca2+ depletion is not required to terminate release. A possible mechanism of release termination is that as luminal [Ca2+] declines, approaching 0.3 to 0.5 mmol/L, release turns off because of the regulatory effect of [Ca2+]SR on RyR gating.7,8,23–25 However, Figure 4E also shows that the minimum twitch [Ca2+]SR increased with frequency in control myocytes (P=0.0025, 1-way ANOVA). Thus, there was not a strict [Ca2+]SR at which release shuts off. Possibly the higher [Ca2+]SR driving force and initial rate of Ca2+ release may raise cleft [Ca2+] to a higher level, where cytosolic Ca2+-dependent RyR inactivation may also contribute to release termination (before [Ca2+]SR reaches 0.3 mmol/L).18 The unaltered minimum twitch [Ca2+]SR in HF cells (Figure 4E) may thus relate to the limited increase in SR Ca2+ release at higher frequency. At 0.1 Hz in control and all 3 frequencies in HF, release terminated at approximately the same [Ca2+]SR. Thus, there may be unaltered regulation of RyR by [Ca2+]SR in HF.
Figure 4F shows SR Ca2+ release duration (time to minimum [Ca2+]SR). In control, release duration was nearly constant for all frequencies. However, in HF, release was significantly prolonged versus control (especially at low frequency). Speculatively, the slower Ca2+ release in HF (attributable to lower [Ca2+]SR) might prolong the time required to reach the shutoff value of [Ca2+]SR, or altered release kinetics (eg, via Ca2+/camodulin-dependent kinase [CaMK]II) may prolong release.26
[Ca2+]SR Dependence of SR Ca2+ Release
Figure 5A assesses how [Ca2+]SR influences SR Ca2+ release in control and HF myocytes. At low [Ca2+]SR, Ca2+ release is very small (in either control or HF myocytes). The control data were fit with a Hill curve, where fractional SR Ca2+ release increased to a maximum (82%) as a function of [Ca2+]SR with half-maximal activation (Kact) of 1.2 mmol/L and Hill coefficient 2.2. Using the same maximal fractional release and Hill coefficient, the Kact for HF myocytes was lower (762 μmol/L). This suggests that for a given [Ca2+]SR (or [Ca]SRT), fractional SR Ca2+ release is actually higher in HF than in control.
Figure 5B plots the fractional SR Ca2+ release (percentage decline in [Ca2+]SR), which increases with increasing [Ca2+]SR. This is evidence for activation (or sensitization) of release by [Ca2+]SR. Moreover, the simplest fit linear regression was steeper for HF myocytes. Additionally, the dotted curves are from the Hill functions in Figure 5A (as fractional SR Ca2+ release). Clearly, the release process is sensitized in HF. This may explain the relatively unaltered fractional SR Ca2+ release in HF versus control in Figure 4D at 0.5 to 1 Hz, despite the significantly lower [Ca2+]SR and [Ca]SRT shown in Figure 4B (see Discussion).
Unaltered Intra-SR Ca2+ Buffering in HF Myocytes
To test whether HF alters intra-SR Ca2+ buffering, [Ca]SRT was varied by pacing frequency, and diastolic [Ca2+]SR and [Ca]SRT were both measured (as in Figure 4A). Figure 6A shows that [Ca]SRT as a function of [Ca2+]SR was similar in HF versus control myocytes. Note that cytosolic Ca2+ buffering was also unaltered in this HF model,5 as well as in human HF and a canine HF model.6,27 Here we estimated intra-SR Ca2+ buffering Bmax using [Ca]SRT and [Ca2+]SR (see Materials and Methods). This yielded similar Bmax in HF and control myocytes (135.8±9.4 versus 125.8±11.5 μmol/L cytosol, respectively; P=0.5). We also compared CSQ expression level (Figure 6B) and found no significant difference between the 2 groups (consistent with our functional results). Thus, the concentration of total effective intra-SR Ca2+-binding sites was unaltered in HF myocytes.
Assuming that SR=3.5% of total cell volume, then Bmax is 2.34 and 2.53 mmol/L SR in HF and control, respectively. This is close to our previous estimation of Bmax of 2.7 mmol/L SR volume.20
So far, we measured decreased diastolic [Ca2+]SR and [Ca]SRT and unaltered SR buffering in HF. We did not directly assess SR volume percentage in HF, but it is unlikely to be a major factor for 2 reasons. First, neither [Ca2+]SR nor [Ca]SRT was altered in HF versus control at 0.1 Hz. If SR volume percentage was reduced in HF, [Ca2+]SR should have increased for the same [Ca]SRT. Second, we found parallel reductions in [Ca]SRT and [Ca2+]SR at 0.5 and 1 Hz in HF (Figure 4B). Based on similar logic, if reduced SR volume percentage of SR was the main cause of reduced [Ca]SRT in HF, then [Ca]SR should not have been reduced in HF.
SERCA Ca2+ Flux Analysis
To further test our measurements and interpretations, we analyzed SR Ca2+ flux 3 ways in control and HF myocytes, using our measured mean data and some parameters from our previous studies.13,20 First, we calculated SERCA flux assuming JSR1=JTwitch−JNCX and JRel=0, using the cytosolic Ca2+ transient decline (converted to total [Ca]i).19 By integrating Ca2+ removal fluxes through the SR pump and NCX versus time, we calculated how these systems competed dynamically and simultaneously during a normal twitch. Figure 7A shows the fractions of Ca2+ transported by SERCA and NCX were 66% and 34%, respectively, in control myocytes, which agrees with our previous work.21
Figure 7B shows JSR1 along with 2 other flux analysis methods. For the second method, JSR2 was calculated using both [Ca2+]i and the simultaneously measured [Ca2+]SR inserted into Equation 1 (see Materials and Methods). The time course of integrated JSR2 matched JSR1 rather well. The third method depends solely on the measured [Ca2+]SR and intra-SR buffering (as in Figure 6A). Thus, JSR3 is a simple function of [Ca2+]SR, Kd-SR, and Bmax-SR. Although the JSR3 integral rose a bit faster than for JSR1 and JSR2 for control myocytes, they had similar kinetics in HF myocytes. These same 3 methods are shown for HF myocytes (Figure 7C). Based on all 3 methods, steady-state Ca2+ removal by SERCA was decreased by ≈25% in HF, which is consistent with ≈23% reduction in SERCA pump function in this HF model (Figure 1F here and our prior work).5,17 Overall, it is remarkable that 3 different analytical methods to determine JSR (which use different assumptions and data) lead to very similar curves. This gives us more confidence in the reliability of our quantitative SR Ca2+ flux analyses and our conclusion that reduced [Ca2+]SR (not intra-SR Ca2+ buffering or volume percentage) is responsible for the decreased [Ca]SRT in HF.
In the present study, we compared [Ca2+]SR and intra-SR buffering in a well-characterized rabbit model of nonischemic HF versus control, using simultaneous measurement of [Ca2+]i and [Ca2+]SR. We found that (1) [Ca2+]SR was lower in HF versus control during both 0.5 and 1 Hz of stimulation, which paralleled changes in [Ca]SRT (and Δ[Ca2+]i); (2) intra-SR Ca2+ buffering capacity was unchanged in HF; and (3) lower [Ca]SRT in HF was mainly attributable to reduced [Ca2+]SR.
[Ca]SRT and [Ca2+]SR in HF: Comparison With Previous Study
There is some consensus that [Ca]SRT is reduced in HF and that altered function of SERCA, NCX, and RyR-mediated leak are responsible (see the online data supplement for mechanistic discussion).1,13.17 Here we measured, for the first time, how reduced [Ca]SRT is manifested, whether it is attributable to reduced [Ca2+]SR, intra-SR buffering, or the volume percentage of SR in HF myocytes.16 This is an important distinction, because it could guide novel therapeutic strategies. Among these possibilities, reduced [Ca2+]SR is particularly critical, because [Ca2+]SR determines the driving force for SR Ca2+ release, total SR Ca2+ content, fractional Ca2+ release, and regulation of RyR open probability.
Kubalova et al14 used fluo-5N to probe [Ca2+]SR in a canine tachypacing-induced HF model. They found that fluo-5N fluorescence was lower in HF myocytes, but they did not measure Fmax or otherwise calibrate the signal. This is a crucial limitation because fluo-5N is a single-wavelength indicator with a fluorescence that is affected by its concentration, distribution, and compartmentalization (eg, differential indicator loading), and its fluorescence is a nonlinear function of [Ca2+]SR in the range of interest.18 Here we provide the first direct on-line measurements of [Ca2+]SR in HF myocytes.
We show that in HF myocytes diastolic [Ca2+]SR and twitch Δ[Ca2+]SR were reduced in HF myocytes during stimulation at 0.5 to 1 Hz. This could explain why in HF there is reduced [Ca]SRT, smaller twitch Ca2+ transients and contractions, and a blunted force–frequency (especially at higher heart rates).1 We suspect that CaMKII-dependent enhancement of SR Ca2+ leak might be critical in limiting [Ca]SRT in the force–frequency relationship in HF (see the online data supplement).
This decreased [Ca2+]SR in HF is expected to reduce fractional SR Ca2+ release because of the steep relationship between [Ca]SRT and fractional SR Ca2+ release7–10 and sensitivity of RyR gating to [Ca2+]SR in bilayer studies.28 Because small reductions in [Ca]SRT can substantially reduce fractional SR Ca2+ release, it was surprising that fractional release was only modestly changed in HF (Figure 4D). However, this is consistent with analysis of SR Ca2+ release in Figure 5, where for a given [Ca2+]SR, there was a higher fractional SR Ca2+ release in HF. It is also in line with less direct estimates of fractional SR Ca2+ release in this rabbit HF model and in human HF myocytes, which are unchanged despite 40% to 50% reduction in [Ca]SRT.5,6 We suspect that this sensitization may be partly because there is substantial upregulation and activation of CaMKII and RyR phosphorylation in HF,15,29,30 which can enhance fractional SR Ca2+ release for a given [Ca]SRT and Ca2+ current trigger.31 Despite this sensitization of release, the amount of SR Ca2+ release was still significantly reduced in HF (Figures 1A and 1C and 4B and 4C). That is, reduced [Ca2+]SR tends to lower fractional SR Ca2+ release in HF, but sensitization increases this fraction back toward control (despite persistently low [Ca2+]SR and Ca2+ transient amplitudes).
Regulation of RyR gating by [Ca2+]SR may also contribute to the shutoff of SR Ca2+ release during excitation–contraction coupling.23 That is, in all HF cells and control cells at 0.1 Hz, release terminated at [Ca2+]SR of 0.3 to 0.4 mmol/L (Figure 4E), regardless of what [Ca2+]SR was before release (Figure 4A). Thus, at this [Ca2+]SR, the RyR may be effectively deactivated (or even inactivated). Indeed, increased exogenous low-affinity Ca2+ buffer in the SR (which slows [Ca2+]SR decline) prolongs release time.23,24 At our larger initial [Ca2+]SR and SR Ca2+ release (control myocytes at 0.5 to 1 Hz), release terminated before [Ca2+]SR fell to 0.3 to 0.4 mmol/L. This might be a manifestation of Ca2+-dependent cytosolic inactivation. That is, with more SR Ca2+ release, local [Ca2+]i-dependent inactivation could become more important, terminating release before [Ca2+]SR falls to 0.3 to 0.4 mmol/L.
Intra-SR Buffering in HF
Ca2+ is heavily buffered in the SR, and CSQ is the most prominent SR Ca2+ buffer (with a Ca2+ affinity of ≈0.6 mmol/L).1 If reduced intra-SR buffer capacity was responsible for reduced [Ca]SRT in HF, one might expect the relatively unaltered fractional SR Ca2+ release observed in HF (as above). Indeed, there is evidence suggesting that CSQ inhibits RyR opening (via triadin and junctin) and that Ca2+ binding to CSQ may relieve that inhibitory effect.24,25,28 Thus, if CSQ had been reduced (but [Ca2+]SR normal), we might have anticipated enhanced fractional SR Ca2+ release or even arrhythmogenic spontaneous SR Ca2+ release.25,32–34 However, intra-SR Ca2+ buffering was unaltered in HF, and this was consistent with our finding of unaltered CSQ expression (as seen in other HF studies1).
Intra-SR Ca2+ buffering properties in intact cardiac myocytes have been inferred only indirectly20 but have not previously been measured in situ (in any cardiac myocytes). Here, we provide the first direct measurement of SR Ca2+ buffering in intact myocytes by simultaneously measuring [Ca2+]SR and [Ca]SRT. We found no difference of intra-SR Ca2+ buffering properties or CSQ expression levels in HF versus control myocytes.
It is notable that for normal diastolic [Ca2+]SR of 1 to 1.5 mmol/L, only ≈50% to 60% of the SR Ca2+ content is bound to the endogenous buffer, with 40% to 50% as free [Ca2+]. This may explain why complete knockout of cardiac CSQ did not result in drastic cardiac systolic dysfunction and that a 50% increase in SR volume provided substantial functional compensation.34 It also emphasizes that bound and free [Ca2+] in the SR are quantitatively comparable, a point not widely appreciated. Arrhythmogenesis seen in CSQ knockout mice may partly reflect the inhibitory regulatory role of CSQ on RyR gating (as above).
Sarcoplasmic Reticulum Volume in HF
In HF, myocyte volume is increased, but if SR volume fails to keep up with cell size, then the volume percentage of SR could reduce [Ca]SRT (in cellular and myofilament volume terms) even with unaltered [Ca2+]SR.16 One previous study reported unchanged SR volume percentage in a hypertrophic rat model.35 Here, we did not directly measure SR volume, but we inferred that SR volume percentage was not substantially altered, based on other directly measured values. First, both diastolic [Ca2+]SR and [Ca]SRT decreased in parallel in HF (Figure 4B), suggesting that a decreased volume percentage would not be required to explain the reduced free [Ca]SRT. Second, intra-SR Ca2+ buffering properties were unaltered in HF (along with parallel reduction in [Ca2+]SR and [Ca]SRT). If SR volume percentage had gone down 30% (as did [Ca]SRT) in HF, the [Ca2+]SR should have been essentially unaltered in HF. This was clearly not the case (Figure 4B). We cannot rule out small SR volume percentage changes in HF, but detailed electron microscopic stereological studies would be required to directly test that.
Relevance to Human Heart Failure: Clinical Implications
Reduced [Ca]SRT in HF is a major factor in systolic dysfunction. Because we now know that the reduced [Ca]SRT is mainly attributable to reduced [Ca2+]SR (at least in this HF model), it has key implications for therapeutic strategies. First, low [Ca2+]SR in HF means that the driving force for SR Ca2+ release ([Ca2+]SR−[Ca2+]i) is lower, limiting SR Ca2+ release rate throughout the cycle. Second, the low [Ca2+]SR tends to decrease the fractional SR Ca2+ release (attributable to the regulatory effect on RyR). This would further limit SR Ca2+ release during systole. Notably, there may be adaptive changes that compensate for this effect (eg, RyR phosphorylation and sensitization to ICa trigger).5,6,15 However, these adaptations may not really benefit systolic function because of intrinsic autoregulation (where fractional release is enhanced, but [Ca]SRT is consequently further reduced),10,15 and the enhanced diastolic SR Ca2+ leak concomitant with sensitized RyRs can be arrhythmogenic. For example, CaMKII-dependent RyR phosphorylation can enhance fractional SR Ca2+ release31 but also enhances diastolic SR Ca2+ release events (sparks) and arrhythmogenic Ca2+ waves.26 Third, it is the reduced [Ca2+]SR that limits [Ca]SRT in HF. Thus, if SERCA pumps can restore normal [Ca2+]SR, then they will also restore [Ca]SRT. That is, the capacity is there, if net Ca2+ uptake can be enhanced. In summary, we found that a decrease in [Ca2+]SR is largely responsible for the decrease in [Ca]SRT in HF myocytes. This knowledge will be valuable in understanding how to improve SR Ca2+ cycling in the failing heart.
We thank Jaime O’Brien for technical support and Drs K. S. Ginsburg and S. Huke for advice.
Sources of Funding
This work was supported by NIH grants HL-30077 and HL-80101 (to D.M.B.) and HL-46929 and HL-73966 (to S.M.P.) and an American Heart Association predoctoral fellowship (T.G.).
Original received September 11, 2006; resubmission received March 15, 2007; revised resubmission received July 24, 2007; accepted August 7, 2007.
Bers DM. Excitation-Contraction Coupling and Cardiac Contractile Force. 2nd ed. Dordrecht, The Netherlands: Kluwer Academic Press; 2001.
Esposito G, Santana LF, Dilly K, Cruz JD, Mao L, Lederer WJ, Rockman HA. Cellular and functional defects in a mouse model of heart failure. Am J Physiol Heart Circ Physiol. 2000; 279: H3101–H3112.
Gomez AM, Valdivia HH, Cheng H, Lederer MR, Santana LF, Cannell MB, McCune SA, Altschuld RA, Lederer WJ. Defective excitation-contraction coupling in experimental cardiac hypertrophy and heart failure. Science. 1997; 276: 800–806.
Hobai IA, O’Rourke B. Decreased sarcoplasmic reticulum calcium content is responsible for defective excitation-contraction coupling in canine heart failure. Circulation. 2001; 103: 1577–1584.
Pogwizd SM, Schlotthauer K, Li L, Yuan W, Bers DM. Arrhythmogenesis and contractile dysfunction in heart failure: roles of sodium-calcium exchange, inward rectifier potassium current, and residual beta-adrenergic responsiveness. Circ Res. 2001; 88: 1159–1167.
Piacentino V 3rd, Weber CR, Chen X, Weisser-Thomas J, Margulies KB, Bers DM, Houser SR. Cellular basis of abnormal calcium transients of failing human ventricular myocytes. Circ Res. 2003; 92: 651–658.
Trafford AW, Diaz ME, Negretti N, Eisner DA. Enhanced Ca2+ current and decreased Ca2+ efflux restore sarcoplasmic reticulum Ca2+ content after depletion. Circ Res. 1997; 81: 477–484.
Pieske B, Maier LS, Bers DM, Hasenfuss G. Ca2+ handling and sarcoplasmic reticulum Ca2+ content in isolated failing and nonfailing human myocardium. Circ Res. 1999; 85: 38–46.
O’Rourke 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.
Shannon TR, Pogwizd SM, Bers DM. Elevated sarcoplasmic reticulum Ca2+ leak in intact ventricular myocytes from rabbits in heart failure. Circ Res. 2003; 93: 592–594.
Kubalova Z, Terentyev D, Viatchenko-Karpinski S, Nishijima Y, Gyorke I, Terentyeva R, da Cunha DN, Sridhar A, Feldman DS, Hamlin RL, Carnes CA, Gyorke S. Abnormal intrastore calcium signaling in chronic heart failure. Proc Natl Acad Sci U S A. 2005; 102: 14104–14109.
Ai X, Curran JW, Shannon TR, Bers DM, Pogwizd SM. Ca2+/calmodulin-dependent protein kinase modulates cardiac ryanodine receptor phosphorylation and sarcoplasmic reticulum Ca2+ leak in heart failure. Circ Res. 2005; 97: 1314–1322.
Venetucci L, Trafford AW, Eisner DA. Illuminating sarcoplasmic reticulum calcium. Circ Res. 2003; 93: 4–5.
Pogwizd SM, Qi M, Yuan W, Samarel AM, Bers DM. Upregulation of Na+/Ca2+ exchanger expression and function in an arrhythmogenic rabbit model of heart failure. Circ Res. 1999; 85: 1009–1019.
Shannon TR, Guo T, Bers DM. Ca2+ scraps: local depletions of free [Ca2+] in cardiac sarcoplasmic reticulum during contractions leave substantial Ca2+ reserve. Circ Res. 2003; 93: 40–45.
Shannon TR, Ginsburg KS, Bers DM. Quantitative assessment of the SR Ca2+ leak-load relationship. Circ Res. 2002; 91: 594–600.
Wu X, Bers DM. Sarcoplasmic reticulum and nuclear envelope are one highly interconnected Ca2+ store throughout cardiac myocyte. Circ Res. 2006; 99: 283–291.
Terentyev D, Viatchenko-Karpinski S, Valdivia HH, Escobar AL, Gyorke S. Luminal Ca2+ controls termination and refractory behavior of Ca2+-induced Ca2+ release in cardiac myocytes. Circ Res. 2002; 91: 414–420.
Terentyev D, Viatchenko-Karpinski S, Gyorke I, Volpe P, Williams SC, Gyorke S. Calsequestrin determines the functional size and stability of cardiac intracellular calcium stores: mechanism for hereditary arrhythmia. Proc Natl Acad Sci U S A. 2003; 100: 11759–11764.
Terentyev D, Nori A, Santoro M, Viatchenko-Karpinski S, Kubalova Z, Gyorke I, Terentyeva R, Vedamoorthyrao S, Blom NA, Valle G, Napolitano C, Williams SC, Volpe P, Priori SG, Gyorke S. Abnormal interactions of calsequestrin with the ryanodine receptor calcium release channel complex linked to exercise-induced sudden cardiac death. Circ Res. 2006; 98: 1151–1158.
Guo T, Zhang T, Mestril R, Bers DM. Ca2+/calmodulin-dependent protein kinase II phosphorylation of ryanodine receptor does affect calcium sparks in mouse ventricular myocytes. Circ Res. 2006; 99: 398–406.
Hobai IA, O’Rourke B. Enhanced Ca2+-activated Na+-Ca2+ exchange activity in canine pacing-induced heart failure. Circ Res. 2000; 87: 690–698.
Hoch B, Meyer R, Hetzer R, Krause EG, Karczewski P. Identification and expression of delta-isoforms of the multifunctional Ca2+/calmodulin-dependent protein kinase in failing and nonfailing human myocardium. Circ Res. 1999; 84: 713–721.
Viatchenko-Karpinski S, Terentyev D, Gyorke I, Terentyeva R, Volpe P, Priori SG, Napolitano C, Nori A, Williams SC, Gyorke S. Abnormal calcium signaling and sudden cardiac death associated with mutation of calsequestrin. Circ Res. 2004; 94: 471–477.
Knollmann BC, Chopra N, Hlaing T, Akin B, Yang T, Ettensohn K, Knollmann BE, Horton KD, Weissman NJ, Holinstat I, Zhang W, Roden DM, Jones LR, Franzini-Armstrong C, Pfeifer K. Casq2 deletion causes sarcoplasmic reticulum volume increase, premature Ca release, and catecholaminergic polymorphic ventricular tachycardia. J Clin Invest. 2006; 116: 2510–2520.