Metabolic Inhibition Alters Subcellular Calcium Release Patterns in Rat Ventricular Myocytes
http://www.100md.com
循环研究杂志 2005年第3期
the Department of Medicine (G.F., S.T.L., C.M., A.G., J.I.G.), Cardiovascular Research Laboratories, Geffen School of Medicine at UCLA, Los Angeles, Calif
Nora Eccles Harrison CVRTI and Division of Cardiology (J.H.B.B.), University of Utah, Salt Lake City, Utah.
Abstract
Metabolic inhibition (MI) contributes to contractile failure during cardiac ischemia and systolic heart failure, in part due to decreased excitation-contraction (E-C) coupling gain. To investigate the underlying mechanism, we studied subcellular Ca2+ release patterns in whole cell patch clamped rat ventricular myocytes using two-dimensional high-speed laser scanning confocal microscopy. In cells loaded with the Ca2+ buffer EGTA (5 mmol/L) and the fluorescent Ca2+-indicator fluo-3 (1 mmol/L), depolarization from eC40 to 0 mV elicited a striped pattern of Ca2+ release. This pattern represents the simultaneous activation of multiple Ca2+ release sites along transverse-tubules. During inhibition of both oxidative and glycolytic metabolism using carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP, 50 nmol/L) and 2-deoxyglucose (2-DG, 10 mmol/L), there was a decrease in inward Ca2+ current (ICa), the spatially averaged Ca2+ transient, and E-C coupling gain, but no reduction in sarcoplasmic reticulum Ca2+ content. The striped pattern of subcellular Ca2+ release became fractured, or disappeared altogether, corresponding to a marked decrease in the area of the cell exhibiting organized Ca2+ release. There was no significant change in the intensity or kinetics of local Ca2+ release. The mechanism is not fully explained by dephosphorylation of L-type Ca2+ channels, because a similar degree of ICa"rundown" in control cells did NOT result in fracturing of the Ca2+ release pattern. We conclude that metabolic inhibition interferes with E-C coupling by (1) reducing trigger Ca2+, and (2) directly inhibiting sarcoplasmic reticulum Ca2+ release site open probability.
Key Words: excitation-contraction coupling metabolic inhibition heart ischemia calcium
Introduction
Cardiac ischemia is associated with a rapid loss of contractile amplitude, followed by inexcitability and ultimately contracture. Systolic heart failure is also characterized by contractile dysfunction. Defective excitation-contraction (E-C) coupling at the single myocyte level has been implicated as a root cause of contractile dysfunction associated with both myocardial ischemia and systolic heart failure. However, the cellular basis of failed E-C coupling during ischemia and systolic heart failure is not completely understood.
Ca2+ transients are believed to be the summation of many microscopic Ca2+ release events triggered by one or more unitary Ca2+ currents.1 It is possible to study these microscopic release events in single myocytes using rapid (240 to 480 hz) 2-dimensional (2D) laser scanning confocal microscopy (LSCM). This novel technique has revealed that depolarization leads to the recruitment of couplons and therefore Ca2+ release from distinct spark sites along the transverse tubules of the myocyte, eliciting a fused pattern of stripes that run along the z-lines.2
Depressed metabolism is a prominent feature of ischemia3 and systolic heart failure.4 In isolated patch-clamped ventricular myocytes, where application of metabolic inhibitors can reproduce the metabolic stress of ischemia and systolic heart failure, we have previously observed an unexpectedly large reduction in the amplitude of the Ca2+ transient during a voltage clamp, which is out of proportion to the decrease in ICa we also observed, and which is not due to depleted sarcoplasmic reticulum (SR) Ca2+ stores.5 This indicates a reduction in E-C coupling gain, a convenient measure of the efficiency of E-C coupling, and which is defined as the rate of Ca2+ release divided by the amplitude of the Ca2+ current.6 The reasons for this reduction in gain are uncertain.
Because reduced gain may be an important contributor to contractile failure during ischemia and systolic heart failure, we sought to explain the mechanism that underlies reduced gain using rapid 2D LSCM to evaluate the response of local Ca2+ release events to depolarization in whole-cell patch-clamped rat ventricular myocytes exposed to metabolic inhibitors. We found that metabolic inhibition fractures the Ca2+ release pattern, indicating reduced ability of the macroscopic Ca2+ current to recruit local release sites. Ca2+ flux through surviving release sites was unaffected.
Materials and Methods
Enzymatically isolated ventricular myocytes were obtained from adult Sprague-Dawley rats as described previously,7 and studied with the whole-cell patch clamp technique. The Cs+-based pipette solution contained 5 mmol/L EGTA and 1 mmol/L fluo-3 to limit the diffusion distance of free Ca2+ to less than 50 nm8 without reducing SR Ca2+ release. Under these high buffering conditions, spatial and temporal resolutions are determined primarily by the less mobile Ca2+-fluo-3 complex.9 Such an approach has been used successfully to more effectively restrict fluorescence increases to Ca2+ release sites,2,10,11 thus improving the precision of Ca2+ localization with confocal imaging. Song et al12 have used a similar strategy to measure SR Ca2+ flux (JSR) at release site microdomains. Pyruvate, PO4, and ADP were added to the pipette solution as substrates for endogenous synthesis of ATP, which avoids difficulty in interpreting effects of metabolic inhibition in the presence of exogenous ATP.5 Glutathione was included as an antioxidant to reduce the phototoxicity common to laser scanning. A rapid solution exchanger was used to perfuse the myocyte under study and to exchange solutions with a halftime of <100 ms.
Cells were imaged in 2D at high speed (240 frames per second) using a Noran Odyssey XL rapid 2D LSCM (Noran Instruments, Middleton, WI).7 We applied strong depolarizations to 0 mV, where Ca2+ spark recruitment is high (and where sparks tend to fuse), because this is more physiological than weaker depolarizations to negative potentials where Ca2+ release is not uniform. For image analysis, we found it impractical to apply the classical Ca2+ spark detection algorithms, which had been created for line-scan images of infrequent spontaneous sparks, to our 2D image stacks of evoked Ca2+ release. We therefore developed a semiautomated method of detecting and analyzing the characteristics of localized Ca2+ release patterns in 2D during strong depolarizations, so that we could avoid using Ca2+ channel blockers, reduced extracellular Ca2+, or signal averaging.
Z-lines were visually identified by the striped pattern of localized increases in fluorescence during depolarization.2 We then set a threshold that maximized detection of fluorescence along the z-lines on release, while minimizing detection of fluorescence in the "inter-z" space. Pixels exceeding the threshold were used to create a map of the active release sites in both the x and y dimensions, which was applied to the entire image stack. The average fluorescence intensity of mapped pixels was then obtained for each image in the stack collected during the depolarization to determine the time course of local changes in Ca2+ concentration. The total number of mapped pixels was used to calculate the "release area." This operation was performed individually for each voltage clamp depolarization analyzed. The investigator determining the threshold was blinded to the condition (control, rundown, metabolic inhibition). This method minimizes problems caused by changes in baseline fluorescence level due to progressive dye loading, and real effects of metabolic inhibition and rundown on diastolic Ca2+ levels. The averaged Ca2+ release site transients analyzed in this fashion were similar to those obtained by examining rarely observed unfused individual Ca2+ sparks.
For additional details, please refer to the expanded Materials and Methods section in the online data supplement available at http://circres.ahajournals.org.
Results
Ca2+ Release Patterns and Area During Metabolic Inhibition
We exposed patch-clamped myocytes to a low concentration of carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP, 50 nmol/L), to limit oxidative metabolism, in conjunction with a standard concentration of 2-deoxyglucose (2-DG, 10 mmol/L), to inhibit glycolytic metabolism, and depolarized the cells at 1-minute intervals from a holding potential of eC40 mV to a test potential of 0 mV. Each test pulse was preceded by six 100 ms conditioning pulses from eC40 to 0 mV at 1 Hz. We simultaneously recorded 2-dimensional fluo-3 fluorescence confocal images and whole-cell current. Figure 1A displays normalized (F/F0) images, release area maps, and membrane currents from a representative cell under control conditions, and after ICa had decreased by 20% during metabolic inhibition. The displayed images were selected from their respective image stacks at the onset of Ca2+ release. The control image was obtained 10 minutes after forming the whole-cell patch, which was long enough to dialyze the cytoplasm with the fluo-3/EGTA internal solution to steady state. Note the striped appearance of the Ca2+ release pattern under control conditions, consistent with previous studies,2 and which is thought to represent numerous adjacent discrete Ca2+ sparks that fuse together when activated simultaneously. After metabolic inhibition sufficient to reduce ICa by 20%, the Ca2+ release pattern fractured. In rare instances, particularly when the release pattern was slightly fractured at the outset, there was no organized Ca2+ release at all after exposure to metabolic inhibitors.
Because it is not practical to quantify the number of individual Ca2+ release sites during a full depolarization, we measured a surrogate: the area of the cell exhibiting coordinated Ca2+ release on depolarization. Release area maps were prepared as described in the Methods. The release area maps for the myocyte shown in Figure 1A are presented adjacent to the corresponding normalized image. On average, we found that metabolic inhibition sufficient to decrease ICa by 20% caused a 75±7% decrease in the area of organized Ca2+ release (Figure 1C; P<0.05, n=7). The decrease in release area observed with metabolic inhibition is consistent with a reduction in the number of organized release sites, and the availability of couplons.13 Why, therefore, in the presence of metabolic inhibitors do SR release sites cease to function There are two likely possibilities. The first is that the opening of ryanodine receptors is directly inhibited by metabolic inhibition. The second is that the Ca2+ trigger, ie, L-type Ca2+ channel opening, is partially inhibited. It is possible, of course, that both these processes may explain the loss of release sites.
Ca2+ Release Patterns and Area During Rundown
To determine whether the disintegration of the Ca2+ release pattern and the corresponding decrease in Ca2+ release area were simple consequences of reduced macroscopic ICa, we took advantage of the fact that Ca2+ channels reduce their open probability over time in patch-clamped myocytes, which causes rundown of the macroscopic Ca2+ current,14eC16 despite the presence of Mg2+, EGTA, and cAMP in the patch pipette. We therefore examined the Ca2+ release patterns during ICa rundown. Control cells were depolarized and imaged as described earlier, but in the absence of metabolic inhibitors. The subcellular Ca2+ images, release area maps, and currents from a representative myocytes are shown in Figure 1B. As in Figure 1A, the normalized control depolarization image was taken 10 minutes after forming the whole-cell patch, and again shows the typical striped pattern of Ca2+ release. ICa and subcellular fluorescence were monitored during depolarizations at 1-minute intervals, until the current had decreased by 20% due to rundown. This was the same extent of reduced macroscopic Ca2+ current as during metabolic inhibition. However, in contrast to metabolic inhibition, we found that repeated depolarizations to 0 mV reproducibly activated the same Ca2+ release sites and caused relatively small changes in the Ca2+ release pattern, as shown in the normalized rundown image in Figure 1B. This impression was confirmed by the corresponding Ca2+ release area maps shown adjacent to the normalized images, and by the summary (Figure 1C, n=7). Thus, although both metabolic inhibition and rundown caused similar reductions in ICa, only metabolic inhibition fractured the Ca2+ release pattern and decreased the Ca2+ release area. These findings remain consistent with the general conclusions that we stated.
Calcium Current and Transients During Metabolic Inhibition and Rundown
If the similar decline in macroscopic ICa during metabolic inhibition and rundown is due to similar changes in the microscopic properties of the currents, then the additional fractionation of the Ca2+ release pattern during metabolic inhibition is probably due to loss of functioning couplons as a result of a decline in available ryanodine receptors (RyRs). At a whole-cell level, a reduction in couplons would result in a reduction in the global Ca2+ transient that is out of proportion to the reduction in ICa. Indeed, we have demonstrated a disproportionate reduction in the Ca2+ transient in guinea pig ventricular myocytes exposed to metabolic inhibitors. To confirm that this also occurs in rat cells, we reanalyzed the 2D confocal images obtained during metabolic inhibition to examine the relationship between whole-cell ICa and the whole-cell (spatially averaged) Ca2+ transient. After 5.6±0.8 minutes of metabolic inhibition, there was a 20% reduction in ICa (from 3.69±0.31 to 2.87±0.31 nA, n=7; P<0.001, Figure 1A) and a 39±4% reduction in the amplitude of the spatially averaged Ca2+ transient (Figure 1D). We then compared the Ca2+ transient during rundown. A 20% reduction in ICa during rundown for 4.9±1.1 minutes (ICa reduced from 3.04±0.41 to 2.41±0.35 nA, n=7; P<0.001, Figure 1B) caused only a 23±6% (n=7) reduction in the Ca2+ transient (Figure 1E). Thus the extent of Ca2+ released during metabolic inhibition is reduced well beyond what one would expect in response to a simple reduction in the amplitude of ICa, recapitulating our previous findings in guinea pig using fura-2 as the Ca2+ indicator.5 Similarly, we also found that there was a significantly greater decline in the maximum rate of rise of the spatially averaged whole-cell Ca2+ transient (dF/dtmax) during metabolic inhibition compared with rundown (61±9% versus 33±4% decline, respectively, P<0.05, n=7 each group), despite similar reductions in ICa. The dF/dtmax is a more appropriate measure of release flux than amplitude of the Ca2+ transient when computing E-C coupling gain. When expressed as a gain function (gain=dF/dtmax/ICa), we found that rundown was associated with only a 14±6% decrease in gain, whereas metabolic inhibition was associated with a 49±11% decrease (P<0.05 for rundown versus metabolic inhibition). These results are consistent with a significantly greater reduction in SR Ca2+ release flux during metabolic inhibition than during control rundown. There are two possible explanations for these findings. First, although rundown and metabolic inhibition produce similar changes in magnitude of ICa with a similar time course, the microscopic properties of the two currents may differ fundamentally and in a way that affects triggering to a greater extent in the case of the current recorded during metabolic inhibition. Alternatively, the microscopic and macroscopic properties of the currents during metabolic inhibition and rundown are identical, in which case metabolic inhibition affects some other aspect of the release mechanism, for example the open probability of RyRs. We will discuss these possibilities later (see Discussion).
Characteristics of Ca2+ Release Site Transients
We also considered the possibility that the reduced macroscopic dF/dtmax of the Ca2+ transient produced by metabolic inhibition was caused by reduced Ca2+ flux through individual Ca2+ release sites. Although we cannot easily assess individual release site (spark) Ca2+ flux during a depolarization to 0 mV, we can measure the flux averaged across all active release sites. We therefore measured the average intensity and kinetics of fluo-3 fluorescence from all the active release sites during metabolic inhibition and during rundown. Peak average release site fluorescence intensity declined slightly for both metabolic inhibition (33±4%) and rundown (21±7%), but the extent of decline was similar for both groups (P=0.19 metabolic inhibition versus rundown). Metabolic inhibition and rundown cells also exhibited equivalent reductions in the maximum rate of rise (dF/dtmax) of averaged Ca2+ release site transients (31±8% for metabolic inhibition and 22±15% for rundown, P=0.64), as shown in Figure 2A. Because Ca2+ flux across release sites is related to the time derivative of the rising phase of the Ca2+ transient,17 or proportional to the fluorescence transient under highly Ca2+ buffered conditions,10,12 our results suggest that the flux of Ca2+ across release sites is not significantly altered by metabolic inhibition compared with rundown. Finally, the time to peak of the release transient was unaffected by either metabolic inhibition or rundown (data not shown).
Preservation of SR Ca2+ Store During Metabolic Inhibition
SR Ca2+ load has been shown by some groups to affect Ca2+ spark frequency and E-C coupling gain.18 A reduction in SR Ca2+ load could reduce the amplitude of the Ca2+ transient in response to ICa and could also reduce the flux across the Ca2+ release sites. To assess SR Ca2+ stores, we applied 5 mmol/L caffeine to the bath surrounding the myocyte for 500 ms. This caused a Ca2+ transient whose amplitude reflects the SR Ca2+ store. Similar to our previous work,5 caffeine-induced transients (Figure 2B) were not significantly altered by application of metabolic inhibitors or by rundown (Figure 2C, n=6 each group, P=0.46). Thus, the effects of metabolic inhibition on Ca2+ release could not be attributed to a reduced SR Ca2+ load. It is also notable that RyRs are still capable of responding to caffeine during metabolic inhibition.
Discussion
Previously, we reported that combined inhibition of oxidative and glycolytic metabolism interfered with the ability of the macroscopic Ca2+ current to trigger Ca2+ release from a fully loaded SR in patch-clamped guinea pig ventricular myocytes. The reasons for the reduction in E-C coupling gain were unknown, but we postulated a fundamental defect in the response of RyRs to triggering by Ca2+ entering through Ca2+ channels.
The present study extends our previous work by demonstrating that combined inhibition of oxidative and glycolytic metabolism causes a marked disintegration of the Ca2+ release pattern and a reduced Ca2+ release area in ventricular myocytes. This is due to a reduction in the number of Ca2+ release sites (sparks) recruited by the Ca2+ current during voltage clamp depolarization. The reduced number of release sites cannot be simply attributed to the 20% decline in the macroscopic Ca2+ current, because a similar decline in ICa during rundown had little effect on the Ca2+ release pattern. Furthermore, neither metabolic inhibition nor rundown had a differential effect on the maximum rate of rise of the averaged release site fluorescence transients. Thus, the decrease in macroscopic E-C coupling gain caused by combined inhibition of oxidative and glycolytic metabolism is due to a decrease in release site recruitment by ICa, rather than a reduction in release site flux or intensity. How can we explain the reduced availability of couplons during metabolic inhibition compared with rundown, despite similar changes in the macroscopic Ca2+ current
Ca2+ Current and SR Ca2+ Release During Metabolic Inhibition
There are three possible explanations for the reduction of couplons during metabolic inhibition. First, if microscopic changes in Ca2+ current are identical during rundown and metabolic inhibition, then the additional loss in functioning couplons produced by the application of metabolic inhibitors is very likely due to the loss of functioning RyRs. Alternatively, if microscopic changes in Ca2+ current during rundown are different than those during metabolic inhibition, then the loss of couplons might be explained by differences in triggering. Finally, differences in both microscopic currents and RyR properties may be required to explain the difference in the Ca2+ transients during rundown versus metabolic inhibition, despite the fact that both macroscopic currents are identical.
Rundown is caused by dephosphorylation of Ca2+ channels, which increases Mode 0 gating, the probability that a channel fails to open during depolarization (PNull), and thus the number of blank sweeps during single channel recording.14eC16 There is now evidence that regardless of whether one or a number of L-type Ca2+ channels are required to trigger SR Ca2+ release, a cluster of L-type Ca2+ channels is required to insure adequate coupling fidelity with RyRs.11 Thus, if we assume for simplicity that every opening of an L-type Ca2+ channel triggers a Ca2+ release event, then the probability of generating a spark (PS) is 1-(PNull)N where N equals the number of L-type Ca2+ channels. Therefore, as PNull increases during rundown, PS should decline and produce fractionation in the pattern of Ca2+ release. However, rundown does not produce significant fractionation of the pattern. The most likely explanation for this preserved Ca2+ release is that clusters of L-type Ca2+ channels arranged in a couplon (ie, N) are sufficiently large to compensate for any increase there might be in PNull due to rundown. It is entirely plausible that metabolic inhibition, like rundown, also leads to dephosphorylation of L-type Ca2+ channels. If the mechanisms and time course of Ca2+ current reduction are the same for both rundown and metabolic inhibition, then the fractionation of the Ca2+ release pattern during the latter is most likely caused by a direct effect of metabolic inhibition on RyRs.
If identical reductions in macroscopic Ca2+ currents generate different release patterns without any effect on RyRs, then we must explain the difference in triggering by differences in microscopic properties of the two currents that do not influence the macroscopic amplitude. The metabolic sensitivity of ICa is well known,5,19,20 and ATP generated by glycolysis may have particular regulatory properties on the Ca2+ channel.19 We have already suggested that metabolic inhibition, like rundown, may lead to dephosphorylation of ICa and an increase in PNull. However, to explain the different pattern of Ca2+ release, we would have to postulate that the effect of metabolic inhibition on L-type Ca2+ channel phosphorylation is more extensive than that which occurs during rundown. The effect, therefore, on triggering would be greater resulting in the observed fractionation. However, in that case, we would not expect the macroscopic currents to be of similar amplitude. To maintain similarity in the macroscopic currents, we would have to postulate an increase in channel conductance to compensate for the increase in PNull. However, an increase in channel conductance seems highly unlikely during metabolic inhibition.
If the primary effect of metabolic inhibition on L-type Ca2+ channels is not reduced Po and increased PNull due to dephosphorylation, as it appears to be with rundown, then the most likely alternative is reduced single channel current due to an increase in subconductance states. Subconductances have been described previously in L-type Ca2+ channels,21 although an association with metabolic inhibition has not to our knowledge been previously reported. Because the probability of evoking a Ca2+ spark depends in part on the amplitude of the unitary L-type Ca2+ current,22,23 the development of subconductance states could affect Ca2+ release (although why Ca2+ release should show a greater dependence on channel conductance versus channel number is uncertain). Again, the existence of subconductance states during metabolic inhibition would require some compensating change in the microscopic properties of those currents so that the macroscopic currents are the same during rundown and metabolic inhibition. These issues clearly require further investigations of microscopic currents, which are beyond the scope of this study.
We cannot absolutely exclude the possibility that the high concentrations of EGTA and fluo-3 in the pipette solution interfered with alterations in E-C coupling introduced by metabolic inhibition. However, because the conditions were similar for both metabolic inhibition and rundown, this seems an unlikely explanation for the differential response of the Ca2+ release pattern to these two conditions.
Role of RyR Activity
RyR openings are required to generate Ca2+ sparks. Both rundown and metabolic inhibition caused a 20% reduction in trigger Ca2+, but only metabolic inhibition resulted in a reduction in Ca2+ spark activation. The most straightforward explanation for the marked reduction in Ca2+ release site probability is a decrease in the Ca2+ dependence of RyR Po. This makes sense on theoretical grounds, because the open probability of RyRs is ATP dependent.24 Elevations in Mg2+ and H+ may also contribute to inhibition of Ca2+ release from RyRs during metabolic inhibition and ischemia25eC29 if these ions escape control by patch electrode dialysis. Our results are also consistent with those of Overend et al,30 who demonstrated that metabolic inhibition with cyanide and 2-DG reduces the tendency of Ca2+ challenged myocytes to develop spontaneous Ca2+ wave activity and eliminates spontaneous spark activity at diastolic potentials. However, our experiments indicate that some RyRs can still be activated by ICa, as well as caffeine, at least at an early stage of metabolic inhibition before the onset of rigor.
Role of SR Ca2+ Load and NCX Activity
Spark probability is influenced not only by RyR behavior, but also by SR Ca2+ load18 and sodium-calcium exchange (NCX).7 Consistent with our prior studies,5 we found no evidence for any significant change in the SR Ca2+ store due to metabolic inhibition (Figure 2B and 2C). Preservation of SR Ca2+ during metabolic inhibition and hypoxia has been observed by other investigators.30,31 An intact SR excludes changes in spark probability due to alterations in SR Ca2+. The intact SR Ca2+ load also accounts for the preserved kinetics of Ca2+ release from surviving release sites. The paradox of maintained SR Ca2+ load during metabolic stress is most readily explained by a decrease in the fraction of SR Ca2+ released during depolarization, owing to defective E-C coupling.30,32 This allows for an energy-compromised SERCA to maintain the SR Ca2+ load. Reduced uptake of Ca2+ by mitochondria during application of FCCP might also support increased uptake by the SR.
Although we have shown that NCX can alter spark probability,7 we did not observe any effect of combined oxidative and glycolytic inhibition on NCX activity in patch clamped guinea pig ventricular myocytes.5 However, it is known that NCX is regulated by ATP, likely via PIP2,33 and reverse-mode NCX may be sensitive to hypoxia.34 Whether NCX is less sensitive than other E-C coupling proteins to changes in ATP levels, or whether other counter-regulatory events take place during metabolic inhibition is uncertain. Further studies will be required to determine whether metabolic inhibition causes changes in NCX activity that are masked by our experimental conditions.
Implications for E-C Coupling
Our explanation for the effect of metabolic inhibition and rundown on E-C coupling is consistent with the idea that a couplon not only contains numerous RyRs, but also a cluster of L-type Ca2+ channels as well.11 A cluster of L-type Ca2+ channels would provide a safety factor for E-C coupling. It has been proposed that synchronization of release sites can affect the whole-cell Ca2+ transient and the Ca2+ myofilament interaction.12 We did not observe any loss of synchronization of Ca2+ release during metabolic inhibition, although asynchronous Ca2+ release has been described in a model of postinfarction cardiac failure.35 It is unlikely that the high Ca2+ buffering we used obscured loss of synchronization, because Song et al12 used a similar buffering strategy to study Ca2+ release site synchronization.
Implications for Clinical Conditions Characterized by Metabolic Stress, eg, Ischemia, Reperfusion, and Failure
To study E-C coupling in detail, it is essential to use isolated heart cells. This means that we must simulate the metabolic stress of ischemia and heart failure by applying metabolic inhibitors. Metabolic stress is but one aspect of the complex milieu of ischemia and heart failure. Our conclusions and their relevance to ischemic heart disease and systolic heart failure must therefore be appropriately guarded.
The effects of metabolic inhibition on Ca2+ release occurred before major reductions in ICa, or the development of rigor/contracture. This suggests that metabolic stress could contribute to the early contractile dysfunction of ischemia, currently attributed to mechanisms such as the "garden-hose effect,"36 and shortening of the action potential due to activation of IKATP.31 The development of defective E-C coupling before other major abnormalities typically associated with ATP depletion also reinforces the importance of this aspect of Ca2+ regulation to control of contractile force in systolic heart failure.
Acknowledgments
This work was supported by NIH R29 HL51129, NIH R01 HL70828, AHA Western States Affiliate GIA 9950748 years, the Laubisch Endowment for Cardiovascular Research, and The Ferman Foundation. We acknowledge the helpful comments of James N. Weiss and the technical assistance of Tan Duong.
References
Cannell MB, Cheng H, Lederer WJ. The control of calcium release in heart muscle. Science. 1995; 268: 1045eC1049.
Cleemann L, Wang W, Morad M. Two-dimensional confocal images of organization, density, and gating of focal Ca2+ release sites in rat cardiac myocytes. Proc Natl Acad Sci U S A. 1998; 95: 10984eC10989.
Rovetto MJ, Whitmer JT, Neely JR. Comparison of the effects of anoxia and whole heart ischemia on carbohydrate utilization in isolated working rat hearts. Circ Res. 1973; 32: 699eC711.
Liao R, Nascimben L, Friedrich J, Gwathmey JK, Ingwall JS. Decreased energy reserve in an animal model of dilated cardiomyopathy: relationship to contractile performance. Circ Res. 1996; 78: 893eC902.
Goldhaber JI, Parker JM, Weiss JN. Mechanisms of excitation-contraction coupling failure during metabolic inhibition in guinea-pig ventricular myocytes. J Physiol. 1991; 443: 371eC386.
Wier WG, Egan TM, Lopez-Lopez JR, Balke CW. Local control of excitation-contraction coupling in rat heart cells. J Physiol. 1994; 474: 463eC471.
Goldhaber JI, Lamp ST, Walter DO, Garfinkel A, Fukumoto GH, Weiss JN. Local regulation of the threshold for calcium sparks in rat ventricular myocytes: role of sodium-calcium exchange. J Physiol. 1999; 520: 431eC438.
Adachi-Akahane S, Cleemann L, Morad M. Cross-signaling between L-type Ca2+ channels and ryanodine receptors in rat ventricular myocytes. J Gen Physiol. 1996; 108: 435eC454.
Blatter LA, Wier WG. Intracellular diffusion, binding, and compartmentalization of the fluorescent calcium indicators indo-1 and fura-2. Biophys J. 1990; 58: 1491eC1499.
Sham JS, Song LS, Chen Y, Deng LH, Stern MD, Lakatta EG, Cheng H. Termination of Ca2+ release by a local inactivation of ryanodine receptors in cardiac myocytes. Proc Natl Acad Sci U S A. 1998; 95: 15096eC15101.
Inoue M, Bridge JH. Ca2+ sparks in rabbit ventricular myocytes evoked by action potentials: involvement of clusters of L-type Ca2+ channels. Circ Res. 2003; 92: 532eC538.
Song LS, Sham JS, Stern MD, Lakatta EG, Cheng H. Direct measurement of SR release flux by tracking ‘Ca2+ spikes’ in rat cardiac myocytes. J Physiol. 1998; 512: 677eC691.
Franzini-Armstrong C, Protasi F, Ramesh V. Shape, size, and distribution of Ca2+ release units and couplons in skeletal and cardiac muscles. Biophys J. 1999; 77: 1528eC1539.
Yue DT, Herzig S, Marban E. Beta-adrenergic stimulation of calcium channels occurs by potentiation of high-activity gating modes. Proc Natl Acad Sci U S A. 1990; 87: 753eC757.
Ono K, Fozzard HA. Phosphorylation restores activity of L-type calcium channels after rundown in inside-out patches from rabbit cardiac cells. J Physiol. 1992; 454: 673eC688.
Erxleben C, Gomez-Alegria C, Darden T, Mori Y, Birnbaumer L, Armstrong DL. Modulation of cardiac Ca(V)1.2 channels by dihydropyridine and phosphatase inhibitor requires Ser-1142 in the domain III pore loop. Proc Natl Acad Sci U S A. 2003; 100: 2929eC2934.
Sipido KR, Wier WG. Flux of Ca2+ across the sarcoplasmic reticulum of guinea-pig cardiac cells during excitation-contraction coupling. J Physiol. 1991; 435: 605eC630.
Lukyanenko V, Gyrke I, Gyrke S. Regulation of calcium release by calcium inside the sarcoplasmic reticulum in ventricular myocytes. Pflugers Archiv. 1996; 432: 1047eC1054.
Losito VA, Tsushima RG, Diaz RJ, Wilson GJ, Backx PH. Preferential regulation of rabbit cardiac L-type Ca2+ current by glycolytic derived ATP via a direct allosteric pathway. J Physiol. 1998; 511: 67eC78.
Hool LC. Hypoxia increases the sensitivity of the L-type Ca2+ current to beta-adrenergic receptor stimulation via a C2 region-containing protein kinase C isoform. Circ Res. 2000; 87: 1164eC1171.
Lacerda AE, Brown AM. Nonmodal gating of cardiac calcium channels as revealed by dihydropyridines. J Gen Physiol. 1989; 93: 1243eC1273.
Lopez-Lopez JR, Shacklock PS, Balke CW, Wier WG. Local calcium transients triggered by single L-type calcium channel currents in cardiac cells. Science. 1995; 268: 1042eC1045.
Santana LF, Cheng H, Gomez AM, Cannell MB, Lederer WJ. Relation between the sarcolemmal Ca2+ current and Ca2+ sparks and local control theories for cardiac excitation- contraction coupling. Circ Res. 1996; 78: 166eC171.
Meissner G. Ryanodine receptor/Ca2+ release channels and their regulation by endogenous effectors. Annu Rev Physiol. 1994; 56: 485eC508.
Elliott AC, Smith GL, Eisner DA, Allen DG. Metabolic changes during ischaemia and their role in contractile failure in isolated ferret hearts. J Physiol. 1992; 454: 467eC490.
O‘Neill SC, Eisner DA. pH-dependent and -independent effects inhibit Ca2+-induced Ca2+ release during metabolic blockade in rat ventricular myocytes. J Physiol. 2003; 550: 413eC418.
Xu L, Mann G, Meissner G. Regulation of cardiac Ca2+ release channel (ryanodine receptor) by Ca2+, H+, Mg2+, and adenine nucleotides under normal and simulated ischemic conditions. Circ Res. 1996; 79: 1100eC1109.
Headrick JP, Willis RJ. Cytosolic free magnesium in stimulated, hypoxic, and underperfused rat heart. J Mol Cell Cardiol. 1991; 23: 991eC999.
Murphy E, Steenbergen C, Levy LA, Raju B, London RE. Cytosolic free magnesium levels in ischemic rat heart. J Biol Chem. 1989; 264: 5622eC5627.
Overend CL, Eisner DA, O’Neill SC. Altered cardiac sarcoplasmic reticulum function of intact myocytes of rat ventricle during metabolic inhibition. Circ Res. 2001; 88: 181eC187.
Stern MD, Silverman HS, Houser SR, Josephson RA, Capogrossi MC, Nichols CG, Lederer JW, Lakatta EG. Anoxic contractile failure in rat heart myocytes is caused by failure of intracellular calcium release due to alteration of the action potential. Proc Natl Acad Sci U S A. 1988; 85: 6954eC6958.
Eisner DA, Trafford AW, Diaz ME, Overend CL, O‘Neill SC. The control of Ca release from the cardiac sarcoplasmic reticulum: regulation versus autoregulation. Cardiovasc Res. 1998; 38: 589eC604.
Hilgemann DW, Ball R. Regulation of cardiac Na-Ca exchange and KATP potassium channels by PIP2. Science. 1996; 273: 956eC959.
Mochizuki S, MacLeod KT. Effects of hypoxia and metabolic inhibition on increases in intracellular Ca2+ concentration induced by Na+/Ca2+ exchange in isolated guinea-pig cardiac myocytes. J Mol Cell Cardiol. 1997; 29: 2979eC2987.
Litwin SE, Zhang D, Bridge JH. Dyssynchronous Ca2+ sparks in myocytes from infarcted hearts. Circ Res. 2000; 87: 1040eC1047.
Koretsune Y, Corretti MC, Kusuoka H, Marban E. Mechanism of early contractile failure: inexcitability, metabolic accumulation, or vascular collapse Circ Res. 1991; 68: 255eC262.(Gary H. Fukumoto, Scott T)
Nora Eccles Harrison CVRTI and Division of Cardiology (J.H.B.B.), University of Utah, Salt Lake City, Utah.
Abstract
Metabolic inhibition (MI) contributes to contractile failure during cardiac ischemia and systolic heart failure, in part due to decreased excitation-contraction (E-C) coupling gain. To investigate the underlying mechanism, we studied subcellular Ca2+ release patterns in whole cell patch clamped rat ventricular myocytes using two-dimensional high-speed laser scanning confocal microscopy. In cells loaded with the Ca2+ buffer EGTA (5 mmol/L) and the fluorescent Ca2+-indicator fluo-3 (1 mmol/L), depolarization from eC40 to 0 mV elicited a striped pattern of Ca2+ release. This pattern represents the simultaneous activation of multiple Ca2+ release sites along transverse-tubules. During inhibition of both oxidative and glycolytic metabolism using carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP, 50 nmol/L) and 2-deoxyglucose (2-DG, 10 mmol/L), there was a decrease in inward Ca2+ current (ICa), the spatially averaged Ca2+ transient, and E-C coupling gain, but no reduction in sarcoplasmic reticulum Ca2+ content. The striped pattern of subcellular Ca2+ release became fractured, or disappeared altogether, corresponding to a marked decrease in the area of the cell exhibiting organized Ca2+ release. There was no significant change in the intensity or kinetics of local Ca2+ release. The mechanism is not fully explained by dephosphorylation of L-type Ca2+ channels, because a similar degree of ICa"rundown" in control cells did NOT result in fracturing of the Ca2+ release pattern. We conclude that metabolic inhibition interferes with E-C coupling by (1) reducing trigger Ca2+, and (2) directly inhibiting sarcoplasmic reticulum Ca2+ release site open probability.
Key Words: excitation-contraction coupling metabolic inhibition heart ischemia calcium
Introduction
Cardiac ischemia is associated with a rapid loss of contractile amplitude, followed by inexcitability and ultimately contracture. Systolic heart failure is also characterized by contractile dysfunction. Defective excitation-contraction (E-C) coupling at the single myocyte level has been implicated as a root cause of contractile dysfunction associated with both myocardial ischemia and systolic heart failure. However, the cellular basis of failed E-C coupling during ischemia and systolic heart failure is not completely understood.
Ca2+ transients are believed to be the summation of many microscopic Ca2+ release events triggered by one or more unitary Ca2+ currents.1 It is possible to study these microscopic release events in single myocytes using rapid (240 to 480 hz) 2-dimensional (2D) laser scanning confocal microscopy (LSCM). This novel technique has revealed that depolarization leads to the recruitment of couplons and therefore Ca2+ release from distinct spark sites along the transverse tubules of the myocyte, eliciting a fused pattern of stripes that run along the z-lines.2
Depressed metabolism is a prominent feature of ischemia3 and systolic heart failure.4 In isolated patch-clamped ventricular myocytes, where application of metabolic inhibitors can reproduce the metabolic stress of ischemia and systolic heart failure, we have previously observed an unexpectedly large reduction in the amplitude of the Ca2+ transient during a voltage clamp, which is out of proportion to the decrease in ICa we also observed, and which is not due to depleted sarcoplasmic reticulum (SR) Ca2+ stores.5 This indicates a reduction in E-C coupling gain, a convenient measure of the efficiency of E-C coupling, and which is defined as the rate of Ca2+ release divided by the amplitude of the Ca2+ current.6 The reasons for this reduction in gain are uncertain.
Because reduced gain may be an important contributor to contractile failure during ischemia and systolic heart failure, we sought to explain the mechanism that underlies reduced gain using rapid 2D LSCM to evaluate the response of local Ca2+ release events to depolarization in whole-cell patch-clamped rat ventricular myocytes exposed to metabolic inhibitors. We found that metabolic inhibition fractures the Ca2+ release pattern, indicating reduced ability of the macroscopic Ca2+ current to recruit local release sites. Ca2+ flux through surviving release sites was unaffected.
Materials and Methods
Enzymatically isolated ventricular myocytes were obtained from adult Sprague-Dawley rats as described previously,7 and studied with the whole-cell patch clamp technique. The Cs+-based pipette solution contained 5 mmol/L EGTA and 1 mmol/L fluo-3 to limit the diffusion distance of free Ca2+ to less than 50 nm8 without reducing SR Ca2+ release. Under these high buffering conditions, spatial and temporal resolutions are determined primarily by the less mobile Ca2+-fluo-3 complex.9 Such an approach has been used successfully to more effectively restrict fluorescence increases to Ca2+ release sites,2,10,11 thus improving the precision of Ca2+ localization with confocal imaging. Song et al12 have used a similar strategy to measure SR Ca2+ flux (JSR) at release site microdomains. Pyruvate, PO4, and ADP were added to the pipette solution as substrates for endogenous synthesis of ATP, which avoids difficulty in interpreting effects of metabolic inhibition in the presence of exogenous ATP.5 Glutathione was included as an antioxidant to reduce the phototoxicity common to laser scanning. A rapid solution exchanger was used to perfuse the myocyte under study and to exchange solutions with a halftime of <100 ms.
Cells were imaged in 2D at high speed (240 frames per second) using a Noran Odyssey XL rapid 2D LSCM (Noran Instruments, Middleton, WI).7 We applied strong depolarizations to 0 mV, where Ca2+ spark recruitment is high (and where sparks tend to fuse), because this is more physiological than weaker depolarizations to negative potentials where Ca2+ release is not uniform. For image analysis, we found it impractical to apply the classical Ca2+ spark detection algorithms, which had been created for line-scan images of infrequent spontaneous sparks, to our 2D image stacks of evoked Ca2+ release. We therefore developed a semiautomated method of detecting and analyzing the characteristics of localized Ca2+ release patterns in 2D during strong depolarizations, so that we could avoid using Ca2+ channel blockers, reduced extracellular Ca2+, or signal averaging.
Z-lines were visually identified by the striped pattern of localized increases in fluorescence during depolarization.2 We then set a threshold that maximized detection of fluorescence along the z-lines on release, while minimizing detection of fluorescence in the "inter-z" space. Pixels exceeding the threshold were used to create a map of the active release sites in both the x and y dimensions, which was applied to the entire image stack. The average fluorescence intensity of mapped pixels was then obtained for each image in the stack collected during the depolarization to determine the time course of local changes in Ca2+ concentration. The total number of mapped pixels was used to calculate the "release area." This operation was performed individually for each voltage clamp depolarization analyzed. The investigator determining the threshold was blinded to the condition (control, rundown, metabolic inhibition). This method minimizes problems caused by changes in baseline fluorescence level due to progressive dye loading, and real effects of metabolic inhibition and rundown on diastolic Ca2+ levels. The averaged Ca2+ release site transients analyzed in this fashion were similar to those obtained by examining rarely observed unfused individual Ca2+ sparks.
For additional details, please refer to the expanded Materials and Methods section in the online data supplement available at http://circres.ahajournals.org.
Results
Ca2+ Release Patterns and Area During Metabolic Inhibition
We exposed patch-clamped myocytes to a low concentration of carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP, 50 nmol/L), to limit oxidative metabolism, in conjunction with a standard concentration of 2-deoxyglucose (2-DG, 10 mmol/L), to inhibit glycolytic metabolism, and depolarized the cells at 1-minute intervals from a holding potential of eC40 mV to a test potential of 0 mV. Each test pulse was preceded by six 100 ms conditioning pulses from eC40 to 0 mV at 1 Hz. We simultaneously recorded 2-dimensional fluo-3 fluorescence confocal images and whole-cell current. Figure 1A displays normalized (F/F0) images, release area maps, and membrane currents from a representative cell under control conditions, and after ICa had decreased by 20% during metabolic inhibition. The displayed images were selected from their respective image stacks at the onset of Ca2+ release. The control image was obtained 10 minutes after forming the whole-cell patch, which was long enough to dialyze the cytoplasm with the fluo-3/EGTA internal solution to steady state. Note the striped appearance of the Ca2+ release pattern under control conditions, consistent with previous studies,2 and which is thought to represent numerous adjacent discrete Ca2+ sparks that fuse together when activated simultaneously. After metabolic inhibition sufficient to reduce ICa by 20%, the Ca2+ release pattern fractured. In rare instances, particularly when the release pattern was slightly fractured at the outset, there was no organized Ca2+ release at all after exposure to metabolic inhibitors.
Because it is not practical to quantify the number of individual Ca2+ release sites during a full depolarization, we measured a surrogate: the area of the cell exhibiting coordinated Ca2+ release on depolarization. Release area maps were prepared as described in the Methods. The release area maps for the myocyte shown in Figure 1A are presented adjacent to the corresponding normalized image. On average, we found that metabolic inhibition sufficient to decrease ICa by 20% caused a 75±7% decrease in the area of organized Ca2+ release (Figure 1C; P<0.05, n=7). The decrease in release area observed with metabolic inhibition is consistent with a reduction in the number of organized release sites, and the availability of couplons.13 Why, therefore, in the presence of metabolic inhibitors do SR release sites cease to function There are two likely possibilities. The first is that the opening of ryanodine receptors is directly inhibited by metabolic inhibition. The second is that the Ca2+ trigger, ie, L-type Ca2+ channel opening, is partially inhibited. It is possible, of course, that both these processes may explain the loss of release sites.
Ca2+ Release Patterns and Area During Rundown
To determine whether the disintegration of the Ca2+ release pattern and the corresponding decrease in Ca2+ release area were simple consequences of reduced macroscopic ICa, we took advantage of the fact that Ca2+ channels reduce their open probability over time in patch-clamped myocytes, which causes rundown of the macroscopic Ca2+ current,14eC16 despite the presence of Mg2+, EGTA, and cAMP in the patch pipette. We therefore examined the Ca2+ release patterns during ICa rundown. Control cells were depolarized and imaged as described earlier, but in the absence of metabolic inhibitors. The subcellular Ca2+ images, release area maps, and currents from a representative myocytes are shown in Figure 1B. As in Figure 1A, the normalized control depolarization image was taken 10 minutes after forming the whole-cell patch, and again shows the typical striped pattern of Ca2+ release. ICa and subcellular fluorescence were monitored during depolarizations at 1-minute intervals, until the current had decreased by 20% due to rundown. This was the same extent of reduced macroscopic Ca2+ current as during metabolic inhibition. However, in contrast to metabolic inhibition, we found that repeated depolarizations to 0 mV reproducibly activated the same Ca2+ release sites and caused relatively small changes in the Ca2+ release pattern, as shown in the normalized rundown image in Figure 1B. This impression was confirmed by the corresponding Ca2+ release area maps shown adjacent to the normalized images, and by the summary (Figure 1C, n=7). Thus, although both metabolic inhibition and rundown caused similar reductions in ICa, only metabolic inhibition fractured the Ca2+ release pattern and decreased the Ca2+ release area. These findings remain consistent with the general conclusions that we stated.
Calcium Current and Transients During Metabolic Inhibition and Rundown
If the similar decline in macroscopic ICa during metabolic inhibition and rundown is due to similar changes in the microscopic properties of the currents, then the additional fractionation of the Ca2+ release pattern during metabolic inhibition is probably due to loss of functioning couplons as a result of a decline in available ryanodine receptors (RyRs). At a whole-cell level, a reduction in couplons would result in a reduction in the global Ca2+ transient that is out of proportion to the reduction in ICa. Indeed, we have demonstrated a disproportionate reduction in the Ca2+ transient in guinea pig ventricular myocytes exposed to metabolic inhibitors. To confirm that this also occurs in rat cells, we reanalyzed the 2D confocal images obtained during metabolic inhibition to examine the relationship between whole-cell ICa and the whole-cell (spatially averaged) Ca2+ transient. After 5.6±0.8 minutes of metabolic inhibition, there was a 20% reduction in ICa (from 3.69±0.31 to 2.87±0.31 nA, n=7; P<0.001, Figure 1A) and a 39±4% reduction in the amplitude of the spatially averaged Ca2+ transient (Figure 1D). We then compared the Ca2+ transient during rundown. A 20% reduction in ICa during rundown for 4.9±1.1 minutes (ICa reduced from 3.04±0.41 to 2.41±0.35 nA, n=7; P<0.001, Figure 1B) caused only a 23±6% (n=7) reduction in the Ca2+ transient (Figure 1E). Thus the extent of Ca2+ released during metabolic inhibition is reduced well beyond what one would expect in response to a simple reduction in the amplitude of ICa, recapitulating our previous findings in guinea pig using fura-2 as the Ca2+ indicator.5 Similarly, we also found that there was a significantly greater decline in the maximum rate of rise of the spatially averaged whole-cell Ca2+ transient (dF/dtmax) during metabolic inhibition compared with rundown (61±9% versus 33±4% decline, respectively, P<0.05, n=7 each group), despite similar reductions in ICa. The dF/dtmax is a more appropriate measure of release flux than amplitude of the Ca2+ transient when computing E-C coupling gain. When expressed as a gain function (gain=dF/dtmax/ICa), we found that rundown was associated with only a 14±6% decrease in gain, whereas metabolic inhibition was associated with a 49±11% decrease (P<0.05 for rundown versus metabolic inhibition). These results are consistent with a significantly greater reduction in SR Ca2+ release flux during metabolic inhibition than during control rundown. There are two possible explanations for these findings. First, although rundown and metabolic inhibition produce similar changes in magnitude of ICa with a similar time course, the microscopic properties of the two currents may differ fundamentally and in a way that affects triggering to a greater extent in the case of the current recorded during metabolic inhibition. Alternatively, the microscopic and macroscopic properties of the currents during metabolic inhibition and rundown are identical, in which case metabolic inhibition affects some other aspect of the release mechanism, for example the open probability of RyRs. We will discuss these possibilities later (see Discussion).
Characteristics of Ca2+ Release Site Transients
We also considered the possibility that the reduced macroscopic dF/dtmax of the Ca2+ transient produced by metabolic inhibition was caused by reduced Ca2+ flux through individual Ca2+ release sites. Although we cannot easily assess individual release site (spark) Ca2+ flux during a depolarization to 0 mV, we can measure the flux averaged across all active release sites. We therefore measured the average intensity and kinetics of fluo-3 fluorescence from all the active release sites during metabolic inhibition and during rundown. Peak average release site fluorescence intensity declined slightly for both metabolic inhibition (33±4%) and rundown (21±7%), but the extent of decline was similar for both groups (P=0.19 metabolic inhibition versus rundown). Metabolic inhibition and rundown cells also exhibited equivalent reductions in the maximum rate of rise (dF/dtmax) of averaged Ca2+ release site transients (31±8% for metabolic inhibition and 22±15% for rundown, P=0.64), as shown in Figure 2A. Because Ca2+ flux across release sites is related to the time derivative of the rising phase of the Ca2+ transient,17 or proportional to the fluorescence transient under highly Ca2+ buffered conditions,10,12 our results suggest that the flux of Ca2+ across release sites is not significantly altered by metabolic inhibition compared with rundown. Finally, the time to peak of the release transient was unaffected by either metabolic inhibition or rundown (data not shown).
Preservation of SR Ca2+ Store During Metabolic Inhibition
SR Ca2+ load has been shown by some groups to affect Ca2+ spark frequency and E-C coupling gain.18 A reduction in SR Ca2+ load could reduce the amplitude of the Ca2+ transient in response to ICa and could also reduce the flux across the Ca2+ release sites. To assess SR Ca2+ stores, we applied 5 mmol/L caffeine to the bath surrounding the myocyte for 500 ms. This caused a Ca2+ transient whose amplitude reflects the SR Ca2+ store. Similar to our previous work,5 caffeine-induced transients (Figure 2B) were not significantly altered by application of metabolic inhibitors or by rundown (Figure 2C, n=6 each group, P=0.46). Thus, the effects of metabolic inhibition on Ca2+ release could not be attributed to a reduced SR Ca2+ load. It is also notable that RyRs are still capable of responding to caffeine during metabolic inhibition.
Discussion
Previously, we reported that combined inhibition of oxidative and glycolytic metabolism interfered with the ability of the macroscopic Ca2+ current to trigger Ca2+ release from a fully loaded SR in patch-clamped guinea pig ventricular myocytes. The reasons for the reduction in E-C coupling gain were unknown, but we postulated a fundamental defect in the response of RyRs to triggering by Ca2+ entering through Ca2+ channels.
The present study extends our previous work by demonstrating that combined inhibition of oxidative and glycolytic metabolism causes a marked disintegration of the Ca2+ release pattern and a reduced Ca2+ release area in ventricular myocytes. This is due to a reduction in the number of Ca2+ release sites (sparks) recruited by the Ca2+ current during voltage clamp depolarization. The reduced number of release sites cannot be simply attributed to the 20% decline in the macroscopic Ca2+ current, because a similar decline in ICa during rundown had little effect on the Ca2+ release pattern. Furthermore, neither metabolic inhibition nor rundown had a differential effect on the maximum rate of rise of the averaged release site fluorescence transients. Thus, the decrease in macroscopic E-C coupling gain caused by combined inhibition of oxidative and glycolytic metabolism is due to a decrease in release site recruitment by ICa, rather than a reduction in release site flux or intensity. How can we explain the reduced availability of couplons during metabolic inhibition compared with rundown, despite similar changes in the macroscopic Ca2+ current
Ca2+ Current and SR Ca2+ Release During Metabolic Inhibition
There are three possible explanations for the reduction of couplons during metabolic inhibition. First, if microscopic changes in Ca2+ current are identical during rundown and metabolic inhibition, then the additional loss in functioning couplons produced by the application of metabolic inhibitors is very likely due to the loss of functioning RyRs. Alternatively, if microscopic changes in Ca2+ current during rundown are different than those during metabolic inhibition, then the loss of couplons might be explained by differences in triggering. Finally, differences in both microscopic currents and RyR properties may be required to explain the difference in the Ca2+ transients during rundown versus metabolic inhibition, despite the fact that both macroscopic currents are identical.
Rundown is caused by dephosphorylation of Ca2+ channels, which increases Mode 0 gating, the probability that a channel fails to open during depolarization (PNull), and thus the number of blank sweeps during single channel recording.14eC16 There is now evidence that regardless of whether one or a number of L-type Ca2+ channels are required to trigger SR Ca2+ release, a cluster of L-type Ca2+ channels is required to insure adequate coupling fidelity with RyRs.11 Thus, if we assume for simplicity that every opening of an L-type Ca2+ channel triggers a Ca2+ release event, then the probability of generating a spark (PS) is 1-(PNull)N where N equals the number of L-type Ca2+ channels. Therefore, as PNull increases during rundown, PS should decline and produce fractionation in the pattern of Ca2+ release. However, rundown does not produce significant fractionation of the pattern. The most likely explanation for this preserved Ca2+ release is that clusters of L-type Ca2+ channels arranged in a couplon (ie, N) are sufficiently large to compensate for any increase there might be in PNull due to rundown. It is entirely plausible that metabolic inhibition, like rundown, also leads to dephosphorylation of L-type Ca2+ channels. If the mechanisms and time course of Ca2+ current reduction are the same for both rundown and metabolic inhibition, then the fractionation of the Ca2+ release pattern during the latter is most likely caused by a direct effect of metabolic inhibition on RyRs.
If identical reductions in macroscopic Ca2+ currents generate different release patterns without any effect on RyRs, then we must explain the difference in triggering by differences in microscopic properties of the two currents that do not influence the macroscopic amplitude. The metabolic sensitivity of ICa is well known,5,19,20 and ATP generated by glycolysis may have particular regulatory properties on the Ca2+ channel.19 We have already suggested that metabolic inhibition, like rundown, may lead to dephosphorylation of ICa and an increase in PNull. However, to explain the different pattern of Ca2+ release, we would have to postulate that the effect of metabolic inhibition on L-type Ca2+ channel phosphorylation is more extensive than that which occurs during rundown. The effect, therefore, on triggering would be greater resulting in the observed fractionation. However, in that case, we would not expect the macroscopic currents to be of similar amplitude. To maintain similarity in the macroscopic currents, we would have to postulate an increase in channel conductance to compensate for the increase in PNull. However, an increase in channel conductance seems highly unlikely during metabolic inhibition.
If the primary effect of metabolic inhibition on L-type Ca2+ channels is not reduced Po and increased PNull due to dephosphorylation, as it appears to be with rundown, then the most likely alternative is reduced single channel current due to an increase in subconductance states. Subconductances have been described previously in L-type Ca2+ channels,21 although an association with metabolic inhibition has not to our knowledge been previously reported. Because the probability of evoking a Ca2+ spark depends in part on the amplitude of the unitary L-type Ca2+ current,22,23 the development of subconductance states could affect Ca2+ release (although why Ca2+ release should show a greater dependence on channel conductance versus channel number is uncertain). Again, the existence of subconductance states during metabolic inhibition would require some compensating change in the microscopic properties of those currents so that the macroscopic currents are the same during rundown and metabolic inhibition. These issues clearly require further investigations of microscopic currents, which are beyond the scope of this study.
We cannot absolutely exclude the possibility that the high concentrations of EGTA and fluo-3 in the pipette solution interfered with alterations in E-C coupling introduced by metabolic inhibition. However, because the conditions were similar for both metabolic inhibition and rundown, this seems an unlikely explanation for the differential response of the Ca2+ release pattern to these two conditions.
Role of RyR Activity
RyR openings are required to generate Ca2+ sparks. Both rundown and metabolic inhibition caused a 20% reduction in trigger Ca2+, but only metabolic inhibition resulted in a reduction in Ca2+ spark activation. The most straightforward explanation for the marked reduction in Ca2+ release site probability is a decrease in the Ca2+ dependence of RyR Po. This makes sense on theoretical grounds, because the open probability of RyRs is ATP dependent.24 Elevations in Mg2+ and H+ may also contribute to inhibition of Ca2+ release from RyRs during metabolic inhibition and ischemia25eC29 if these ions escape control by patch electrode dialysis. Our results are also consistent with those of Overend et al,30 who demonstrated that metabolic inhibition with cyanide and 2-DG reduces the tendency of Ca2+ challenged myocytes to develop spontaneous Ca2+ wave activity and eliminates spontaneous spark activity at diastolic potentials. However, our experiments indicate that some RyRs can still be activated by ICa, as well as caffeine, at least at an early stage of metabolic inhibition before the onset of rigor.
Role of SR Ca2+ Load and NCX Activity
Spark probability is influenced not only by RyR behavior, but also by SR Ca2+ load18 and sodium-calcium exchange (NCX).7 Consistent with our prior studies,5 we found no evidence for any significant change in the SR Ca2+ store due to metabolic inhibition (Figure 2B and 2C). Preservation of SR Ca2+ during metabolic inhibition and hypoxia has been observed by other investigators.30,31 An intact SR excludes changes in spark probability due to alterations in SR Ca2+. The intact SR Ca2+ load also accounts for the preserved kinetics of Ca2+ release from surviving release sites. The paradox of maintained SR Ca2+ load during metabolic stress is most readily explained by a decrease in the fraction of SR Ca2+ released during depolarization, owing to defective E-C coupling.30,32 This allows for an energy-compromised SERCA to maintain the SR Ca2+ load. Reduced uptake of Ca2+ by mitochondria during application of FCCP might also support increased uptake by the SR.
Although we have shown that NCX can alter spark probability,7 we did not observe any effect of combined oxidative and glycolytic inhibition on NCX activity in patch clamped guinea pig ventricular myocytes.5 However, it is known that NCX is regulated by ATP, likely via PIP2,33 and reverse-mode NCX may be sensitive to hypoxia.34 Whether NCX is less sensitive than other E-C coupling proteins to changes in ATP levels, or whether other counter-regulatory events take place during metabolic inhibition is uncertain. Further studies will be required to determine whether metabolic inhibition causes changes in NCX activity that are masked by our experimental conditions.
Implications for E-C Coupling
Our explanation for the effect of metabolic inhibition and rundown on E-C coupling is consistent with the idea that a couplon not only contains numerous RyRs, but also a cluster of L-type Ca2+ channels as well.11 A cluster of L-type Ca2+ channels would provide a safety factor for E-C coupling. It has been proposed that synchronization of release sites can affect the whole-cell Ca2+ transient and the Ca2+ myofilament interaction.12 We did not observe any loss of synchronization of Ca2+ release during metabolic inhibition, although asynchronous Ca2+ release has been described in a model of postinfarction cardiac failure.35 It is unlikely that the high Ca2+ buffering we used obscured loss of synchronization, because Song et al12 used a similar buffering strategy to study Ca2+ release site synchronization.
Implications for Clinical Conditions Characterized by Metabolic Stress, eg, Ischemia, Reperfusion, and Failure
To study E-C coupling in detail, it is essential to use isolated heart cells. This means that we must simulate the metabolic stress of ischemia and heart failure by applying metabolic inhibitors. Metabolic stress is but one aspect of the complex milieu of ischemia and heart failure. Our conclusions and their relevance to ischemic heart disease and systolic heart failure must therefore be appropriately guarded.
The effects of metabolic inhibition on Ca2+ release occurred before major reductions in ICa, or the development of rigor/contracture. This suggests that metabolic stress could contribute to the early contractile dysfunction of ischemia, currently attributed to mechanisms such as the "garden-hose effect,"36 and shortening of the action potential due to activation of IKATP.31 The development of defective E-C coupling before other major abnormalities typically associated with ATP depletion also reinforces the importance of this aspect of Ca2+ regulation to control of contractile force in systolic heart failure.
Acknowledgments
This work was supported by NIH R29 HL51129, NIH R01 HL70828, AHA Western States Affiliate GIA 9950748 years, the Laubisch Endowment for Cardiovascular Research, and The Ferman Foundation. We acknowledge the helpful comments of James N. Weiss and the technical assistance of Tan Duong.
References
Cannell MB, Cheng H, Lederer WJ. The control of calcium release in heart muscle. Science. 1995; 268: 1045eC1049.
Cleemann L, Wang W, Morad M. Two-dimensional confocal images of organization, density, and gating of focal Ca2+ release sites in rat cardiac myocytes. Proc Natl Acad Sci U S A. 1998; 95: 10984eC10989.
Rovetto MJ, Whitmer JT, Neely JR. Comparison of the effects of anoxia and whole heart ischemia on carbohydrate utilization in isolated working rat hearts. Circ Res. 1973; 32: 699eC711.
Liao R, Nascimben L, Friedrich J, Gwathmey JK, Ingwall JS. Decreased energy reserve in an animal model of dilated cardiomyopathy: relationship to contractile performance. Circ Res. 1996; 78: 893eC902.
Goldhaber JI, Parker JM, Weiss JN. Mechanisms of excitation-contraction coupling failure during metabolic inhibition in guinea-pig ventricular myocytes. J Physiol. 1991; 443: 371eC386.
Wier WG, Egan TM, Lopez-Lopez JR, Balke CW. Local control of excitation-contraction coupling in rat heart cells. J Physiol. 1994; 474: 463eC471.
Goldhaber JI, Lamp ST, Walter DO, Garfinkel A, Fukumoto GH, Weiss JN. Local regulation of the threshold for calcium sparks in rat ventricular myocytes: role of sodium-calcium exchange. J Physiol. 1999; 520: 431eC438.
Adachi-Akahane S, Cleemann L, Morad M. Cross-signaling between L-type Ca2+ channels and ryanodine receptors in rat ventricular myocytes. J Gen Physiol. 1996; 108: 435eC454.
Blatter LA, Wier WG. Intracellular diffusion, binding, and compartmentalization of the fluorescent calcium indicators indo-1 and fura-2. Biophys J. 1990; 58: 1491eC1499.
Sham JS, Song LS, Chen Y, Deng LH, Stern MD, Lakatta EG, Cheng H. Termination of Ca2+ release by a local inactivation of ryanodine receptors in cardiac myocytes. Proc Natl Acad Sci U S A. 1998; 95: 15096eC15101.
Inoue M, Bridge JH. Ca2+ sparks in rabbit ventricular myocytes evoked by action potentials: involvement of clusters of L-type Ca2+ channels. Circ Res. 2003; 92: 532eC538.
Song LS, Sham JS, Stern MD, Lakatta EG, Cheng H. Direct measurement of SR release flux by tracking ‘Ca2+ spikes’ in rat cardiac myocytes. J Physiol. 1998; 512: 677eC691.
Franzini-Armstrong C, Protasi F, Ramesh V. Shape, size, and distribution of Ca2+ release units and couplons in skeletal and cardiac muscles. Biophys J. 1999; 77: 1528eC1539.
Yue DT, Herzig S, Marban E. Beta-adrenergic stimulation of calcium channels occurs by potentiation of high-activity gating modes. Proc Natl Acad Sci U S A. 1990; 87: 753eC757.
Ono K, Fozzard HA. Phosphorylation restores activity of L-type calcium channels after rundown in inside-out patches from rabbit cardiac cells. J Physiol. 1992; 454: 673eC688.
Erxleben C, Gomez-Alegria C, Darden T, Mori Y, Birnbaumer L, Armstrong DL. Modulation of cardiac Ca(V)1.2 channels by dihydropyridine and phosphatase inhibitor requires Ser-1142 in the domain III pore loop. Proc Natl Acad Sci U S A. 2003; 100: 2929eC2934.
Sipido KR, Wier WG. Flux of Ca2+ across the sarcoplasmic reticulum of guinea-pig cardiac cells during excitation-contraction coupling. J Physiol. 1991; 435: 605eC630.
Lukyanenko V, Gyrke I, Gyrke S. Regulation of calcium release by calcium inside the sarcoplasmic reticulum in ventricular myocytes. Pflugers Archiv. 1996; 432: 1047eC1054.
Losito VA, Tsushima RG, Diaz RJ, Wilson GJ, Backx PH. Preferential regulation of rabbit cardiac L-type Ca2+ current by glycolytic derived ATP via a direct allosteric pathway. J Physiol. 1998; 511: 67eC78.
Hool LC. Hypoxia increases the sensitivity of the L-type Ca2+ current to beta-adrenergic receptor stimulation via a C2 region-containing protein kinase C isoform. Circ Res. 2000; 87: 1164eC1171.
Lacerda AE, Brown AM. Nonmodal gating of cardiac calcium channels as revealed by dihydropyridines. J Gen Physiol. 1989; 93: 1243eC1273.
Lopez-Lopez JR, Shacklock PS, Balke CW, Wier WG. Local calcium transients triggered by single L-type calcium channel currents in cardiac cells. Science. 1995; 268: 1042eC1045.
Santana LF, Cheng H, Gomez AM, Cannell MB, Lederer WJ. Relation between the sarcolemmal Ca2+ current and Ca2+ sparks and local control theories for cardiac excitation- contraction coupling. Circ Res. 1996; 78: 166eC171.
Meissner G. Ryanodine receptor/Ca2+ release channels and their regulation by endogenous effectors. Annu Rev Physiol. 1994; 56: 485eC508.
Elliott AC, Smith GL, Eisner DA, Allen DG. Metabolic changes during ischaemia and their role in contractile failure in isolated ferret hearts. J Physiol. 1992; 454: 467eC490.
O‘Neill SC, Eisner DA. pH-dependent and -independent effects inhibit Ca2+-induced Ca2+ release during metabolic blockade in rat ventricular myocytes. J Physiol. 2003; 550: 413eC418.
Xu L, Mann G, Meissner G. Regulation of cardiac Ca2+ release channel (ryanodine receptor) by Ca2+, H+, Mg2+, and adenine nucleotides under normal and simulated ischemic conditions. Circ Res. 1996; 79: 1100eC1109.
Headrick JP, Willis RJ. Cytosolic free magnesium in stimulated, hypoxic, and underperfused rat heart. J Mol Cell Cardiol. 1991; 23: 991eC999.
Murphy E, Steenbergen C, Levy LA, Raju B, London RE. Cytosolic free magnesium levels in ischemic rat heart. J Biol Chem. 1989; 264: 5622eC5627.
Overend CL, Eisner DA, O’Neill SC. Altered cardiac sarcoplasmic reticulum function of intact myocytes of rat ventricle during metabolic inhibition. Circ Res. 2001; 88: 181eC187.
Stern MD, Silverman HS, Houser SR, Josephson RA, Capogrossi MC, Nichols CG, Lederer JW, Lakatta EG. Anoxic contractile failure in rat heart myocytes is caused by failure of intracellular calcium release due to alteration of the action potential. Proc Natl Acad Sci U S A. 1988; 85: 6954eC6958.
Eisner DA, Trafford AW, Diaz ME, Overend CL, O‘Neill SC. The control of Ca release from the cardiac sarcoplasmic reticulum: regulation versus autoregulation. Cardiovasc Res. 1998; 38: 589eC604.
Hilgemann DW, Ball R. Regulation of cardiac Na-Ca exchange and KATP potassium channels by PIP2. Science. 1996; 273: 956eC959.
Mochizuki S, MacLeod KT. Effects of hypoxia and metabolic inhibition on increases in intracellular Ca2+ concentration induced by Na+/Ca2+ exchange in isolated guinea-pig cardiac myocytes. J Mol Cell Cardiol. 1997; 29: 2979eC2987.
Litwin SE, Zhang D, Bridge JH. Dyssynchronous Ca2+ sparks in myocytes from infarcted hearts. Circ Res. 2000; 87: 1040eC1047.
Koretsune Y, Corretti MC, Kusuoka H, Marban E. Mechanism of early contractile failure: inexcitability, metabolic accumulation, or vascular collapse Circ Res. 1991; 68: 255eC262.(Gary H. Fukumoto, Scott T)