Isolated Rabbit Working Heart Function During Progressive Inhibition of Myocardial SERCA ActivityNovelty and Significance
Rationale: The extent to which sarcoplasmic reticulum Ca2+ATPase (SERCA) activity alone determines left ventricular (LV) pump function is unknown.
Objective: To correlate SERCA activity with hemodynamic function of rabbit LV during thapsigargin perfusion.
Methods and Results: Isolated rabbit hearts were perfused in working heart configuration, and LV pump function was assessed using a pressure-volume catheter. Rapid and complete (>95%) inhibition of SERCA was associated with a moderate decrease in cardiac function (to 70%–85% of control). Further decrease in cardiac function to 50%–75% of control occurred over the next ≈30 minutes despite no detectable further inhibition of SERCA activity. Analysis of the 20 seconds prior to pump failure revealed a rapid decrease in end diastolic volume. Intermediate levels of SERCA function (≈50% of control) had only minor hemodynamic effects. Parallel experiments in field-stimulated isolated ventricular cardiomyocytes monitored intracellular Ca2+ and cell shortening. On perfusion with thapsigargin, Ca2+ transient amplitude and cell shortening fell to ≈70% of control followed by increased diastolic Ca2+ concentration and diastolic cell shortening to achieve a new steady state.
Conclusions: The relationship between SERCA activity and LV function in the rabbit is highly nonlinear. In the short term, only moderate effects on LV pump function were observed despite almost complete (>95%) reduction in SERCA activity. The terminal decline of function was associated with sudden sustained increase in diastolic tone comparable to the sustained contraction observed in isolated cardiomyocytes. Secondary increases of intracellular Ca2+ and Na+ following complete SERCA inhibition eventually limit contractile function and precipitate LV pump failure.
Ventricular cardiomyocytes isolated from animal models of cardiac disease and patients with heart failure demonstrate altered expression and function of a number of Ca2+ handling proteins important for excitation-contraction coupling (E-C coupling).1–4 One such protein is the sarcoplasmic/endoplasmic reticulum Ca2+-ATPase pump (SERCA), which plays a pivotal role in both myocardial contraction and relaxation. SERCA functions to pump Ca2+ into the sarcoplasmic reticulum (SR) store. In doing so, it returns the intracellular Ca2+ concentration ([Ca2+]i) to resting levels causing relaxation of the cardiomyocyte and replenishment of the SR Ca2+ store for the next cycle.
To fully understand the impact of SERCA function on the physiology and patho-physiology of cardiac muscle, it is important to quantify the specific contribution of SERCA to whole heart pump function. Transgenic methodology, by allowing selective removal/reduction of one protein, has been used to examine the relationship between cardiac SERCA levels and whole heart function in mice.5–8 However, these chronic reductions of SERCA protein cause compensatory changes in expression of other proteins (eg, L-type Ca2+ channel, sodium/calcium exchanger [NCX], phospholamban) altering the status of E-C coupling and changes in whole heart structure and myofilament sensitivity, all of which prevent a direct assessment of the role of SERCA in isolation. Furthermore, the electrophysiology and E-C coupling process in rodent hearts is distinct in many aspects from that of larger mammals, in particular humans.9,10
The major aim of this study was to quantitatively assess the impact of SERCA activity alone on whole heart mechanical function. We hypothesize that in the complete absence of SR activity (<5%), left ventricular (LV) contractility is insufficient to maintain normal pump function. Acute pharmacological blockade of SERCA activity was achieved using the selective inhibitor thapsigargin (TG) in a rabbit isolated working heart preparation combined with pressure-volume catheter technology to monitor whole heart contractile and hemodynamic function.
The novelty and advantage of this experimental approach is several fold: (1) it uses a species with E-C coupling more similar to that of human, (2) acute inhibition negates the compensatory alterations of transgenics, and (3) there are no changes in vascular resistance (ie, changes in pre/afterloads or centrally-mediated autonomic signaling). Whole heart experiments were performed under near physiological conditions and parallel studies with isolated rabbit cardiomyocytes investigated the cellular mechanisms underlying the response to acute SERCA inhibition with TG. The data demonstrated that SERCA activity was rapidly reduced to below 5% in the first 4 to 5 minutes of perfusion with TG yet working heart function was maintained for up to 60 minutes. Pump function eventually failed due to a sustained increase in diastolic tone with preserved ejection fraction in the absence of arrhythmic events.
An expanded Methods section is provided in the online-only Data Supplement.
Working Heart Set-Up and LV Pressure-Volume Measurement
Hearts were excised from adult male New Zealand White rabbits and cannulated onto an isolated working heart perfusion system via the aorta and pulmonary vein of the left atrium. Perfusion was set in working heart mode (Online Figure I); preload and afterload were set at 6 mm Hg and 60 mm Hg respectively. Aortic flow (AoF) was measured continuously by an in-line ultrasonic flowmeter (Malema) and hearts were paced at 211±5 bpm. A custom-built 3F variable segment pressure-volume catheter (Scisense, Canada) was advanced past the aortic valve into the LV for continuous measurement of hemodynamic function. The catheter allowed selection of volume segments of 11.0, 14.0, 17.0, or 19.0 mm. Parallel conductance was estimated by injection of a bolus of hypertonic saline into the preload chamber of the isolated working heart system.11
Hearts were assigned to 1 of 2 perfusion protocols as shown in Figure 1. Corresponding vehicle control groups (Protocol 1, n=5; Protocol 2, n=7) were produced by perfusion with an equivalent volume of DMSO for time-matched durations.
Protocol 1 (>5 Minutes)
Hearts were perfused with 3 μmol/L TG (n=6) and LV function was continuously recorded until AoF was undetectable (defined as <20 mL/minute on flowmeter; Online Figure II). Hearts were subsequently switched and perfused in nonworking heart mode with (1) nominally Ca2+-free Tyrode's solution containing 50 μmol/L ethylene glycol tetraacetic acid (EGTA) (2 minutes) followed by (2) 3 μmol/L TG (0.5 minutes). These final 2 perfusion steps served to buffer the circulating free [Ca2+] in the tissue to ≈100 to 200 nmol/L (required for the in vitro SERCA activity assay) and to act as a control in the vehicle group for any residual circulating TG in the vasculature at the end of the experiment respectively.
Protocol 2 (<5 Minutes)
For correlation of working heart contractile function with SR Ca2+ uptake, hearts were perfused with 3 μmol/L TG (n=10) for short periods of time (1–4 minutes) and LV function was continuously recorded until dP/dtmax dropped by 5% to 22% of control values.
Immediately at the end of each working heart perfusion protocol, a transmural section of LV free wall was cut and snap frozen in liquid nitrogen. Tissue was homogenized in five volumes of ice cold protease/phosphatase inhibitor buffer and stored at −80°C.
SR Ca2+ Uptake Studies
Total protein content of the whole LV homogenate was assessed using the Bradford protein assay. Oxalate supported SR Ca2+ uptake measurements were performed as described previously12 and converted to [Ca2+]i as detailed in the Online Supplement. SERCA activity was quantified in the hearts with measurable uptake by fitting gradients to the decline in free [Ca2+]i over a 50 nmol/L range at 0.3, 0.5 and 1 μmol/L and expressed relative to those from the time-matched vehicle hearts.
Field Stimulation of Intact Cardiomyocytes With Simultaneous Whole-Cell Epifluorescence and Shortening Measurements
Studies were performed on rabbit LV cardiomyocytes as described previously.13,14 Intact cardiomyocytes were loaded with Fura-2AM (5 μmol/L; Invitrogen) and placed in a bath on an inverted microscope. To parallel working heart conditions, cells were perfused at 37°C with modified Krebs-Henseleit with 2.5 mmol/L [Ca2+] and field stimulated at ≈3.5 Hz. Cells were then perfused with 3 μmol/L TG for variable durations (5–240 seconds) prior to rapid application of 15 mmol/L caffeine. The Fura-2 fluorescence (340 and 380 nm excitation; R340/380 nm) was measured using a spinning wheel spectrophotometer (Cairn Research, UK; sampling rate of 500 Hz) whereas cellular shortening was measured using a video edge detection system (IonOptix, MA; sampling rate of 200 Hz).
Data Analysis and Statistics
All data were analyzed offline and expressed as mean±SEM.
For isolated working heart studies, mean pressure and volume signals were obtained by averaging (1) 10 cardiac cycles in control and 5 minutes post-TG (TG+5 minutes) in Protocol 1 and (2) the final 4 cardiac cycles in the presence of TG at the end of each protocol.
For isolated cardiomyocyte studies, mean [Ca2+]i and cell shortening signals were obtained by averaging 6 transients (OriginPro v6.1). In cases of multiple comparisons either repeated-measures ANOVA with Tukey-Kramer post test correction (working heart function and uptake gradient data) or 1-way ANOVA with Dunnett post test correction (working heart function during the final 60 seconds and Ca2+ transient/cell shortening parameters at sequential time points in TG) were used (GraphPad InStat). Values were considered significant when P<0.05.
Effects of Acute SERCA Inhibition on Whole Heart Working Rabbit Heart Function: Protocol 1
The effects of TG perfusion and time-matched vehicle perfusion during Protocol 1 on mechanical function of the working rabbit heart are summarized in Figure 2. Control working heart function was not significantly different between groups prior to treatment with either TG or vehicle solution. Figure 2A(i) shows typical LV pressure (top), volume (middle), and AoF (bottom) recordings from an isolated working heart perfused with 3 μmol/L TG. Panels from left to right depict control, 5 minutes postaddition of TG (TG+5 minutes) and at the point where AoF fell below detectable levels (< 20 mL/minute, end point, denoted by the dashed box). Addition of TG led to an initial rapid decline in LV function over the first 5 minutes of perfusion, after which a pseudo steady state was achieved and maintained for variable periods of time (maximum ≈60 minutes). This was followed by a rapid decline in function before AoF fell to undetectable levels. The mean time for AoF to reach undetectable levels following the TG perfusion in Protocol 1 was 36.3±10.2 minutes. Figure 2A(ii) shows the pressure-volume loops corresponding to the recordings in Figure 2A(i). As perfusion with TG progressed there was a leftward shift of the pressure-volume loop accompanied by a decline in stroke volume. These alterations in volume parameters are summarized in Figure 2A(iii). Consequently, despite the decline in stroke volume following TG perfusion, ejection fraction (EF; Figure 2A[iv]) was not significantly altered between the 3 points (Control: 59.5±2.5%, TG+5 minutes: 54.3±4.6% and end point: 65.9±6.0%; P>0.05). Coronary flow was unaltered throughout the protocol (Figure 2A[v]; Control: 84.6±2.0, TG +5 minutes: 84.9±1.8 and end point: 82.3±1.0 mL/min; P>0.05).
Figure 2B(i–iii) shows indices of contractile function at 5 minutes postaddition (TG+5 minutes) and at end point in TG and vehicle-treated hearts. Perfusion with TG resulted in a significant decline in all contractile parameters at 5 minutes postaddition and at end point. Peak pressure (PP), dP/dtmax, and stroke work (SW) were all significantly reduced 5 minutes postaddition of TG (PP: 108.1±2.3 versus 93.8±2.0 mm Hg; dP/dtmax: 2031±175 versus 1370±48 mm Hg/s; SW: 86.1±4.5 versus 51.8±2.7 mm Hg*mL; control versus TG+5 minutes for all parameters; P<0.05). This decline continued to the end point at which PP had fallen to 80.4±2.2 mm Hg, dP/dtmax to 1010±18 mm Hg/s and SW to 22.5±2.5 mm Hg*mL (P<0.05 control versus end point for all parameters). In vehicle hearts at end point, there was a small but significant decline in dP/dtmax and SW (dP/dtmax: 2152±141 versus 1916±128 mm Hg/s; SW: 68.5±9.9 versus 59.8±9.4 mm Hg*mL; control versus end point; P<0.05).
Working heart diastolic indices are shown in Figure 2C(i–iii). TG perfusion led to significant increases in end diastolic pressure (Figure 2C(i)) and the relaxation constant Tau (Figure 2C(iii)) at 5 minutes postaddition and at the end point of the experiment (end diastolic pressure: 7.7±1.5 versus 12.4±1.1 versus 13.4±1.3 mm Hg; Tau: 27.0±3.1 versus 40.7±6.0 versus 69.8±8.8 ms; SS versus TG+5 minutes versus end point, P<0.05). In addition, TG perfusion led to a significant depression in dP/dtmin after 5 minutes which was more pronounced at end point (SS: -2347±143 versus TG +5 minutes: -1712±134 versus end point: -1341±120 mm Hg/s; P<0.05; Figure 2C(ii)). There were no significant changes in diastolic indices in the vehicle group.
The example trace of AoF for the complete time course in TG (Figure 3A) demonstrates the rapid drop in LV function, which immediately preceded pump failure in Protocol 1 TG treated hearts. Closer inspection of the 60 seconds prior to cessation of AoF dissected the changes in functional parameters during this phase with improved resolution (Figure 3B). AoF was significantly reduced at 20 and 10 seconds prior to end point and thereafter fell precipitously toward undetectable levels. Over the course of the final minute, pressure indices remained unchanged until end point at which PP, dP/dtmax, and dP/dtmin were significantly reduced (P<0.05 versus −60-second timepoint). End diastolic volume, end systolic volume, and SW were significantly reduced at −10 seconds and end point (P<0.05 versus −60-second timepoint) thus confirming the rapid time course of the cessation of working heart function in TG-treated hearts where changes in volume appeared to precede changes in pressure.
SERCA Activity in Working Hearts Treated With TG
In order to determine SERCA activity at the end of working heart experiments subject to either Protocol 1 (>5 minutes) or Protocol 2 (<5 minutes), a section of LV free wall was immediately taken from the whole heart, snap frozen, and homogenized. Homogenates (4 mg/mL total protein) were then subject to a cuvette-based oxalate-facilitated Ca2+ uptake assay. Vehicle-treated hearts subject to either protocol consistently demonstrated a rapid Ca2+ uptake component (mean maximum uptake rate and dissociation constant were 2.96±0.25 nmoles/mg total protein/minute and 243±16.3 nmol/L, respectively, which are within the range of previously published values15–17; Online Figure III). Assay sensitivity was verified (Online Figure IV) and the potential for unbound TG contamination and regional variation were controlled for (Online Figure V & VI).
SERCA Activity in Working Hearts Treated With TG: Protocol 1
All vehicle-treated hearts demonstrated rapid Ca2+ uptake as depicted by the decline in free [Ca2+]i in the example control traces in Figure 4A (i&ii, black traces), which became faster in time course with increasing total protein concentration in the cuvette (data not shown). In contrast however, there was no detectable Ca2+ uptake from any Protocol 1 TG-treated samples (Figure 4A, gray trace).
SERCA Activity in Working Hearts Treated With TG: Protocol 2
Ten hearts were subject to Protocol 2 and perfused with TG for between 1 to 4 minutes; vehicle hearts were perfused with DMSO and time-matched (n=7). Assessment of SERCA activity revealed 5 of these TG-treated hearts had a measurable uptake component (Figure 4A[ii], gray trace) and 5 hearts did not. SERCA activity was quantified in hearts with measurable uptake as shown in Figure 4A(ii) inset. Comparison of the gradients of decline between these 3 points in each heart showed no significant differences in any of the 5 hearts indicating no alteration in dissociation constant of SERCA following exposure to TG.
Correlation of function with the degree of SERCA activity was performed by examining a range of functional parameters immediately prior to the end of the experiment in TG-treated hearts with and without measureable Ca2+ uptake (Figure 4B). Of these parameters, mean PP, developed pressure, and dP/dtmax were significantly different between the 2 groups (uptake versus no uptake; PP: 97.6±0.6 versus 94.6±1.0, P<0.05; developed pressure: 97.1±1.0 versus 92.0±2.0, P<0.05; dP/dtmax 91.0±1.5 versus 83.3±1.34, P<0.001; all % control). The most sensitive functional parameter to TG was dP/dtmax, which showed no overlap in function between hearts with measurable uptake (circles) and those with undetectable uptake (squares; Figure 4B). Linear regression analysis of a plot of individual dP/dtmax values against SERCA activity (expressed relative to vehicle; Figure 4C) revealed that SERCA activity fell below undetectable levels once dP/dtmax had fallen below 84% of control function.
Characterization of Ca2+ Transients, Cell Shortening, and SR Ca2+ Content in Single Rabbit Cardiomyocytes Perfused With TG
To assess the effects of TG on [Ca2+]i handling and cell shortening under equivalent conditions to working heart experiments (2.5 mmol/L [Ca2+], ≈3.5 Hz, 37°C), isolated rabbit cardiomyocytes were field stimulated and perfused with TG for varying durations. At the end of TG perfusion, rapid application of caffeine was used to assess SR Ca2+ content. Figure 5A shows typical sections of trace obtained in the solutions and at time points indicated above. The cell in Figure 5A(i) was exposed to TG for 40 seconds, which was followed by a significant response to the subsequent caffeine bolus, whereas the cell in Figure 5A(ii) was exposed to TG for 180 seconds, which resulted in no response to caffeine. TG perfusion times were varied between 5 to 240 seconds before caffeine application (n=12 cells from 6 hearts). A subset of cells were perfused with a comparable percentage of DMSO (0.09%) for time-matched durations to obtain vehicle SR contents. Analysis of the mean peak and diastolic [Ca2+]i (Figure 5B[i]) and mean cell shortening (Figure 5B[ii]) transients at set time points during perfusion with TG was performed as described in methods. [Ca2+]i transient peak dropped sharply within the first 40 seconds of TG perfusion before slowly increasing thereafter until the 240-second time point, whereas Ca2+ transient diastolic levels continued to steadily rise throughout exposure to TG (Figure 5B[i]). Diastolic [Ca2+]i was significantly increased at 220 and 240 seconds post-TG (P<0.05 versus control pre-TG). A comparable pattern of change was seen in peak and diastolic cell shortening during perfusion with TG. The amplitude of the caffeine-induced Ca2+ transient was taken as a measure of SR Ca2+ content both in free [Ca2+]i and also converted to total [Ca2+]i using previously published cytoplasmic buffer characteristics.15 All cells perfused with TG for less than 150 seconds (n=6) demonstrated SR Ca2+ release on caffeine application; all those perfused for 150 seconds or longer had no response to caffeine (n=6). A plot of total SR Ca2+ content against TG perfusion time revealed that SR Ca2+ content declined exponentially with increasing exposure time to TG (Figure 5C[i]). As can be seen in Figure 5C(ii) there was a hyperbolic relationship (exponent=3.6) between the mean amplitude of the Ca2+ transient immediately before caffeine application and the total SR Ca2+ content. This relationship suggested that total SR Ca2+ content is reduced to zero when Ca2+ amplitude falls to 52% of control. A similar relationship was demonstrated for cell shortening (data not shown). The decay of the Ca2+ transient is predominantly due to SERCA, accounting for ≈70% of Ca2+ removal in the rabbit.18 Therefore a measure of the contribution of SERCA to the decay of the Ca2+ transient was taken using the gradient of decline of the total [Ca2+]i transient over a 10 μmol/L range expressed relative to the control value (Figure 5C[iii]). This plot was fit with an exponential decay, which suggested that the SERCA activity decayed rapidly within the first 50 seconds of TG perfusion reaching a plateau at approximately 60 seconds post-TG.
The averaged free [Ca2+]i transients in control, 70 seconds post-TG and 240 seconds post-TG, shown in Figure 5D(i–iii), demonstrate the propensity for a rightwards shift in the peak with increasing durations of TG treatment. Mean time to rise and time to fall were calculated as an indicator of the alteration in Ca2+ transient kinetics to TG under the conditions used in working hearts (Figure 5D[iv]). Both the time to rise and time to fall were significantly altered at the 70-second time point in TG (rise time increased to 185.7±18.3%, whereas fall time decreased to 87.4±2.3% both of control; P<0.05 control versus 70 seconds, n=10 at 70 second time point). At these high (physiological) stimulation rates, TG had no significant effect on the half time of decay of the Ca2+ transient.
The main objective of this study was to assess the impact of SERCA activity on cardiac mechanical function in the absence of changes in pre- or afterload on the heart or any major compensatory changes that can occur over the longer term following SERCA downregulation. A secondary consideration was to use a heart where the relative contribution of the SR Ca2+ release to E-C coupling was similar to that of the human heart.9,10 These aims were achieved by applying a specific inhibitor of SERCA19 (3 μmol/L TG) to a rabbit working heart preparation. Initially, SERCA inhibition led to a relatively rapid (within the first 5 minutes) but modest decrease in cardiac function (to ≈70%–85% of pre-TG levels) across a range of functional parameters (PP, dP/dtmax, dP/dtmin). A further decrease in cardiac function to 50% to 75% of pre-TG levels occurred over the next ≈30 minutes despite no detectable SERCA activity in these hearts. Parallel experiments of [Ca2+]i and cell shortening in isolated rabbit cells suggest that a substantial Ca2+ transient remained after acute SR inhibition accompanied by increased diastolic Ca2+ concentration and diastolic cell shortening. The data suggest that the mechanical deficit associated with an acute yet substantial reduction in SERCA activity (>95%) is insufficient to prevent cardiac pump function directly.
Cardiac Pump Function After Substantial Reductions of SERCA Activity
Cardiac pump function was measured using a pressure-volume catheter in combination with the rabbit working heart. Use of these devices ensured sensitive measurement of various systolic and diastolic functional parameters, which were then correlated with SERCA activity measurements on the same heart. This is in contrast to previous studies where TG and other SERCA inhibitors have been used on isovolumetric ex vivo heart preparations20,21 and where vascular effects of TG have been demonstrated.22 Since coronary flow was unaltered during the protocol in our preparation, the direct effect of reduced SERCA activity on working heart mechanical function could be assessed. When comparing the current working heart model to previous studies on transgenic mice,6 the relationship between cardiac contractility and SERCA activity shows a similar nonlinear relationship, ie, in the transgenic studies, SERCA reductions of up to ≈50% have small effects on function, whereas reductions of SERCA of up to ≈90% have significant effects on contractility yet mechanical pump function can be maintained. In the current working heart model, despite having no detectable SERCA activity or compensatory changes in protein expression, the decrease in mechanical function over the first 5 minutes was also initially limited with a 30% decrease in dP/dtmax, 28% decrease in dP/dtmin, and a 13% decrease in PP (Figure 2). Thereafter, cardiac function in the working heart model slowly diminished over the next ≈30 minutes by an additional 18% (dP/dtmax), 15% (dP/dtmin), and 12% (PP), at which point the heart failed to produce sufficient aortic flow. Although the current study examined the effects of SERCA inhibition alone to whole heart function under near physiological conditions the response under conditions of varying rates and loads requires future examination.
Effect of Progressive Thapsigargin Exposure on Isolated Cardiomyocytes
Experiments on single cells from the rabbit at stimulus rates comparable to working heart experiments showed that the Ca2+ transient amplitude and the associated cell shortening was reduced to ≈60% of control after ≈60 seconds in the TG solution (Figure 5B[i&ii]). The rate of Ca2+ transient decay was reduced to a steady level after a similar time scale (Figure 5C[iii]), yet caffeine induced Ca2+ release indicated that ≈40% of the SR content remained (Figure 5C[i]). Continued perfusion with TG resulted in the loss of Ca2+ from the SR without any significant change in the Ca2+ transient characteristics. This suggests that the SR contribution to E-C coupling was not present when the SR Ca2+ content fell below ≈40% of control; this is consistent with previous measurements.23–25 The remaining SR Ca2+ content slowly declined to undetectable levels in the subsequent ≈90 seconds (Figure 5C[i]) thus showing a similar time-course to SR depletion as previously reported for rabbit cells perfused with TG.26 Plotting the range of values of SR total Ca2+ content and Ca2+ transient amplitude against the Ca2+ transient amplitude suggested a hyperbolic relationship between SR content and the SR contribution to the Ca2+ transient amplitude (Ca2+-transientSR α [SR-content]3; Figure 5C[ii]). This steep relationship has been observed by other groups and predicts that when the SR Ca2+ content decreases to ≈50% of control values, the SR component of the Ca2+ transient will be only ≈8% of normal. In rabbit myocardium, the remaining Ca2+ transient (≈50% of control values) resulted from the large influx via L-type Ca2+ channel. The involvement of the SR in E-C coupling appeared to be negligible after ≈60 seconds in TG, yet diastolic [Ca2+] and diastolic cell shortening continued to change over the subsequent 200 s. The reasons for these progressive changes in diastolic [Ca2+] are unclear; one possibility is that the slow rise of diastolic [Ca2+] is caused by a slow increase in intracellular Na+ concentration ([Na+]i) observed during SERCA inhibition,20 this would raise diastolic [Ca2+] via NCX. Additional experiments on isolated single cells confirmed that following addition of TG, [Na+]i increased significantly (by ≈3.4 mmol/L, Online Figure VII). This increase is comparable to the measurements made by Maier et al (1997) in whole hearts.20 On cessation of stimulation, [Na+]i fell rapidly to below pre-TG values (12.0±1.0 versus 5.7±2.8 mmol/L, TG +240 seconds, ≈3.5Hz versus TG +340 seconds, 0Hz, P<0.05) as would be expected due to sarcolemmal extrusion processes.
Changes in Ca2+ transient kinetics were also observed, the most pronounced effect being the increase in the time to peak of the Ca2+ transient to 185% of control after ≈60 seconds perfusion with TG (Figure 5D[iv]). This time-course is consistent with the loss of rapid SR Ca2+ release and the loss of associated rapid SERCA-mediated reuptake. In the absence of SR activity, intracellular [Ca2+] would continue to increase even after the L-type Ca2+ channel current had decayed due to Ca2+ entry via NCX during the plateau phase of the action potential. Repolarization of the action potential would drive NCX into the forward mode (Ca2+ extrusion) abruptly reducing [Ca2+]i. Thus, as reported previously, time to peak [Ca2+]i in the absence of SERCA activity is determined by duration of the action potential.27
Causes of Failure of Pump Function in TG
All working heart preparations subject to Protocol 1 eventually failed to produce an aortic flow as a result of sustained perfusion with TG. Arrhythmias were not seen at this end point, instead sinus rhythm was maintained and pump function fell precipitously over a period of the final 60 seconds (Figure 3A, dashed box). SERCA activity was undetectable in preparations with impaired but significant working heart function and in tissue immediately after pump failure. In the current form, the SERCA assay would fail to register significant uptake when SERCA activity is <5% of the control value (Online Figure IV). The failure to detect differences in SERCA uptake between working and nonworking myocardium could have 2 explanations: (1) the transition to complete working heart failure occurs with a reduction in SERCA activity to below <5% (ie, the resolution of our measurements) or (2) an event secondary to complete SERCA inhibition is responsible for failure. Examination of the working heart end diastolic volumes at the end point indicated a significantly smaller end diastolic volume just prior to failure and a significant increase in the relaxation time (Figure 2A[iii] & 2C[iii], respectively). These data suggest that pump failure was associated with a sustained diastolic tone, which limited stroke volume and precipitated failure of the working heart.
Measurements of cell shortening and [Ca2+]i did not show an equivalent failure of E-C coupling during progressive TG perfusion. Instead a new steady state was achieved after ≈2 minutes of perfusion with TG and contractility was maintained for as long as measurements were recorded (25–30 minutes, data not shown), which is comparable to the mean time to end point of working heart function (36 minutes) in Protocol 1. The maintenance of Ca2+ transients and cell shortening over this time course is also consistent with the presence of a measureable degree of contractility of the working heart at the point of pump failure (Figure 2A [i] dashed box). During this time, diastolic cell length decreased indicating the development of a sustained contraction, which may be equivalent to the sustained contraction observed in whole hearts. This suggests that the pump fails as a result of slow sustained increase in diastolic tone that develops after complete SERCA inhibition and that the steady-state cellular contractile activity achieved in the complete absence of SERCA is insufficient to maintain pump function in the long term. In line with the current study, previous studies on isolated rat heart using an acute block of SERCA20 and in mouse heart after SERCA knock-out28 have also reported a sustained contracture and raised diastolic [Ca2+]i that is associated with a raised [Na+]i level. As suggested by both groups, the cause of raised [Na+]i is enhanced Na+ influx via NCX that occurs as a result of the increased Ca2+ efflux required to balance enhanced Ca2+ entry via the L-type Ca2+ channel after SERCA block. The increased [Na+]i would in turn reduce the effectiveness of Ca2+ efflux via the NCX mechanism leading to a further increase in diastolic [Ca2+]i and stimulation of NCX activity until a new steady state had been achieved. In addition, raised efflux of Ca2+ from mitochondria in response to the elevated [Na+]i may also contribute to the accumulation of [Ca2+]i.29 The slow secondary rise of [Ca2+]i, [Na+]i, and the consequent raised diastolic tone appear to be the major limiting factors to the functioning of the working heart.
This study shows that the relationship between SERCA activity and LV pump function in the rabbit is highly nonlinear. Intermediate levels of SERCA function (≈50% of control levels) have only minor hemodynamic effects. In the short-term, only moderate detrimental effects on LV pump function were observed despite the almost complete absence of SERCA activity. Rapid decline of function was associated with sudden sustained increase in diastolic tone comparable to the sustained contraction observed in isolated cardiomyocytes. In conclusion, following complete inhibition of SERCA activity, secondary increases of intracellular Ca2+ and Na+ limit contractile function and precipitate eventual LV pump failure.
Sources of Funding
This work was supported by Heart Research U.K., Medical Research Scotland and European Community's Seventh Framework Programme FP7/2007–2013 under grant agreement n° FP7-HEALTH-2009-single-stage 241526. The equipment was funded by Tenovus Scotland and the Royal Society.
In March 2012, the average time from submission to first decision for all original research papers submitted to Circulation Research was 13.2 days.
The online-only Data Supplement is available with this article at http://circres.ahajournals.org/lookup/suppl/doi:10.1161/CIRCRESAHA.111.262337/-/DC1.
Non-standard Abbreviations and Acronyms
- aortic flow
- intracellular Ca2+ concentration
- left ventricle
- intracellular Na+ concentration
- sodium/calcium exchanger
- peak pressure
- sarco-endoplasmic reticulum Ca2+ ATPase
- sarcoplasmic reticulum
- stroke work
- Received December 7, 2011.
- Revision received April 23, 2012.
- Accepted April 26, 2012.
- © 2012 American Heart Association, Inc.
- Pogwizd SM,
- Qi M,
- Yuan W,
- Samarel AM,
- Bers DM
- Chen X,
- Piacentino V III.,
- Furukawa S,
- Goldman B,
- Margulies KB,
- Houser SR
- Del Monte F,
- Harding SE,
- Schmidt U,
- Matsui T,
- Kang ZB,
- Dec GW,
- Gwathmey JK,
- Rosenzweig A,
- Hajjar RJ
- Periasamy M,
- Reed TD,
- Liu LH,
- Ji Y,
- Loukianov E,
- Paul RJ,
- Nieman ML,
- Riddle T,
- Duffy JJ,
- Doetschman T,
- Lorenz JN,
- Shull GE
- Ver Heyen M,
- Heymans S,
- Antoons G,
- Reed T,
- Periasamy M,
- Awede B,
- Lebacq J,
- Vangheluwe P,
- Dewerchin M,
- Collen D,
- Sipido K,
- Carmeliet P,
- Wuytack F
- Ji Y,
- Lalli MJ,
- Babu GJ,
- Xu Y,
- Kirkpatrick DL,
- Liu LH,
- Chiamvimonvat N,
- Walsh RA,
- Shull GE,
- Periasamy M
- Pieske B,
- Maier LS,
- Bers DM,
- Hasenfuss G
- Loughrey CM,
- Smith GL,
- MacEachern KE
- Seidler T,
- Loughrey CM,
- Zibrova D,
- Kettlewell S,
- Teucher N,
- Kogler H,
- Hasenfuss G,
- Smith GL
- Elliott EB,
- Hasumi H,
- Otani N,
- Matsuda T,
- Matsuda R,
- Kaneko N,
- Smith GL,
- Loughrey CM
- Loughrey CM,
- Otani N,
- Seidler T,
- Craig MA,
- Matsuda R,
- Kaneko N,
- Smith GL
- Sagara Y,
- Inesi G
- Bassani JW,
- Bassani RA,
- Bers DM
- Liu T,
- O'Rourke B
Novelty and Significance
What Is Known?
The function of the sarco-endoplasmic reticulum Ca2+ATPase (SERCA) in heart muscle is to ensure rapid removal of intracellular Ca2+ from the cytosol during diastole, thus allowing the heart to relax rapidly.
After a transgenic knockout of SERCA, the adult mouse heart tolerates very low levels of SERCA; the tolerance may be due to compensatory changes in the expression and function of other Ca2+ handling proteins.
The transgenic mouse hearts fail only when SERCA levels are <10% of normal, but the cause of contractile failure is not clear.
What New Information Does This Article Contribute?
This study shows the consequences of rapid (within minutes) pharmacological inhibition of SERCA on the pump function of isolated rabbit hearts before any compensatory changes in expression of other Ca2+ handling proteins.
Even in the absence of compensatory changes, the ventricle fails as a working pump only when SERCA activity is <5% of normal.
The failure of pump function is the result of a sustained increase in diastolic tone as a consequence of raised diastolic Ca2+ concentration secondary to a raised intracellular [Na+] and not due to a slower rate of relaxation.
The extent to which SERCA activity alone determines left ventricular (LV) pump function is unknown. We show that substantial downregulation of SERCA alone does not dramatically affect LV pump function in the rabbit myocardium, which has excitation-contraction coupling characteristics similar to humans. Secondary increases of intracellular Ca2+ and Na+ following complete SERCA inhibition eventually limit contractile function resulting in a rise in diastolic tone and LV pump failure. Intermediate levels of SERCA, comparable to those seen in failing myocardium, have minimal hemodynamic consequences. Contrary to current thinking, this work indicates that the myocardium can tolerate substantial reductions in SERCA activity before the effects of this deficiency are evident in ventricular pump function. Eventually failure occurs due to raised diastolic tone and not due to reduced rate of relaxation. This work challenges the current view of SERCA that suggests that the activity of this protein is critical for cardiac pump function because of its direct effects on excitation-contraction coupling. The implication is that improvements of human cardiac function that are a consequence of moderate increases in SERCA activity may be caused by indirect effects of long-term changes in the calcium transient.