Alterations in excitation-contraction coupling in chronically ischemic or hibernating myocardium

Abstract

In coronary artery disease, areas subtended by a severely stenotic artery or by collateral vessels can develop chronic contractile dysfunction in the absence of necrosis. This dysfunction is thought to be adaptive to the reduced flow reserve and can be reversible upon revascularization, hence the term ‘hibernating’ myocardium. In the present report, the underlying cellular mechanisms were studied in a pig with severe stenosis in the left circumflex coronary artery, resulting in hibernation in the distal myocardium.

After six weeks, single cardiomyocytes were isolated enzymatically from the hibernating region (HIB) and their properties compared with those of cardiomyocytes from the same area in matched control pigs (CTRL). The amplitude of cell shortening during field stimulation (1 Hz) was reduced in HIB versus CTRL; the accompanying Ca²⁺ transients were only modestly reduced. In whole cell recording, prolongation of action potential was observed in HIB. When this difference was excluded by using depolarizing steps of fixed duration in both HIB and CTRL, the Ca²⁺ transients in HIB myocytes were reduced compared with CTRL. There was also a decrease in peak L-type Ca²⁺ current in HIB. In intact cells, increasing the available Ca²⁺ for contraction did not correct the contractile deficit in HIB, suggesting alterations to the myofilaments.

In conclusion, in this pig model for hibernating myocardium, intrinsic remodelling of the myocytes with a unique profile of excitation-contraction coupling is demonstrated. Along with the changes observed in myocytes from the border zone of a myocardial infarction, or in the remote area, this specific phenotype adds to the diversity and complexity of the remodelling processes in ischemic cardiomyopathy.

Ischemic heart disease is the most common underlying cause of chronic heart failure, currently a major chronic disease with significant morbidity and mortality. Contractile dysfunction in ischemic cardiomyopathy is the result of the loss of myocardium through infarction, additional myocyte loss through apoptosis and also of the presence of viable myocardium with reduced function because of chronic and/or repeated ischemia, the so-called ‘hibernating’ myocardium (1,2). Hibernation, introduced as a term by Rahimtoola (3), describes a state of decreased contractility in the presence of a critical flow limitation despite preserved viability. This state appears to be reversible upon revascularization. The concept of hibernation has received widespread attention and has had important implications in clinical practice (4–7). Methods have been developed to detect viability in areas with reduced contraction that would benefit from revascularization. Numerous clinical studies (8,9) have demonstrated that the revascularization of hibernating myocardium improves ventricular function and reduces mortality.

Several animal models have been developed to study the pathophysiology of hibernation (10–14). These studies have defined and confirmed the concept of a reduction in contractile function in response to critical coronary stenosis. There remains some debate, however, regarding the nature and magnitude of flow reduction that induces a state of hibernation. The concept of a significantly reduced baseline resting flow is not shared by all authors, but there is a general consensus that there is a severely reduced coronary flow reserve (reviewed by Heusch [15] and by Camici and Rimoldi [16]).

The cellular mechanisms underlying contractile dysfunction have been studied in human biopsies and animal models. Cellular dedifferentiation with abnormal mitochondria, myolysis and glycogen accumulation has been observed in cells within the hibernating area (17–20). In a pig model of chronic hibernation, sarcoplasmic/endoplasmic reticulum Ca²⁺-ATPase, phospholamban and ryanodine receptor were downregulated (21). In the model of short-term hibernation, a reduced contractile response to increasing external Ca²⁺ was reported, suggesting involvement of the myofilaments (22). The intrinsic properties of myocytes from hibernating myocardium should reflect the actual remodelling and altered expression of proteins relevant to contractile function. To study these intrinsic properties, cardiomyocytes must be examined in vitro, without the confounding influence of alterations in mechanical loading or the extracellular millieu. There are no specific data on cellular remodelling in human hibernating myocardium because studies have been mostly restricted to myocytes isolated at the time of transplantation from end-stage failing hearts without specification of the area from which they were sampled. To study specific changes in hibernation, animal models are currently the better option. Nevertheless, there are only limited data available so far on intrinsic myocyte functional remodelling.

EXCITATION-CONTRACTION COUPLING IN CELLS FROM HIBERNATING MYOCARDIUM

In our institution (Katholieke Universiteit Leuven, Belgium), a nonsurgical pig model was developed to study different substrates of ischemic cardiomyopathy (23–25). The introduction of a copper-coated stent into the circumflex coronary artery induced intima proliferation, leading to a critical stenosis and the development of hibernating myocardium in the posterior wall. This comprised a reduction in regional contractile function as identified using tissue Doppler imaging and strain rate analysis, with a typical biphasic response to dobutamine infusion (26): increased contraction at low dose, but reduced contraction at higher (greater than 10 μg/kg) dose, consistent with viability and with the presence of a severe reduction of coronary flow reserve.

Intact single cardiomyocytes were enzymatically isolated from the dysfunctional region (HIB) and their properties were compared with those of myocytes isolated from matched control pigs (CTRL) from the same region to avoid confounding factors regarding regional heterogeneity (25). The isolated myocytes were studied for size, morphology and functional properties.

The myocytes from HIB were, on average, larger than the myocytes from CTRL. When stimulated at a rate of 1 Hz, the contraction amplitude was reduced in HIB and associated with a significant slowing of contraction rate (Figure 1A). The differences in contraction amplitude persisted after the myocytes were patched with a microelectrode and perfused with a solution containing 5 mM Mg-ATP. The amplitude of the increase in intracellular Ca²⁺ ([Ca²⁺]_i) ([Ca²⁺]_i transient) during contraction was not significantly different between HIB and CTRL, but action potentials were longer in HIB. When possible differences in action potential duration were eliminated by studying the [Ca²⁺]_i transient in voltage clamp mode with a similar duration of depolarization, the amplitude of the [Ca²⁺]_i transient was reduced in HIB compared with CTRL. The amplitude of the inward L-type Ca²⁺ current was also reduced (Figure 1B). On the other hand, no differences in the function or expression of the sarcolemmal Na⁺/Ca²⁺ exchanger could be detected and no reduction in the expression of the sarcoplasmic Ca²⁺ pump (sarcoplasmic/endoplasmic reticulum Ca²⁺-ATPase) was observed. Consistent with this, no reduction in the sarcoplasmic reticulum Ca²⁺ content was seen.

Figure 1).

Excitation-contraction coupling in myocytes. A Typical example of a time course of unloaded cell shortening recorded from a matched control (CTRL) (left panel) and from a myocyte from the dysfunctional myocardium of the pig model (HIB) (right panel) during steady state stimulation at 1 Hz. Cell shortening (ΔL) is normalized to cell length (L₀). B Pooled data of cell shortening (ΔL/L₀) and time to peak shortening (TtPS) in CTRL (21 cells from eight pigs) and HIB (18 cells from six pigs) cells. *Indicates P<0.05. I_CaL L-type Ca²⁺ current

Although the differences in amplitude of the [Ca²⁺]_i transient were not very pronounced, it was found that the Ca²⁺ response curve of normal pig myocytes was very steep and, therefore, small changes in [Ca²⁺]_i could have a profound impact on contractile function. However, if a reduction in the amplitude of the [Ca²⁺]_i transient is the only factor responsible for the reduced contractile amplitude, then increases in the [Ca²⁺]_i transient (such as those induced by the rapid application of 10 mM caffeine) should rescue the contractile deficit; however, this was not the case. This observation thus indicates that the observed changes in Ca²⁺ handling may not have been the only factor responsible for the contractile dysfunction.

An additional factor contributing to contractile dysfunction could be a reduction in the response of the myofilaments to Ca²⁺ as previously described in a short-term model for hibernation (22). This hypothesis can be tested by examining the Ca²⁺ response of skinned cardiomyocytes (ie, single myocytes with permeabilized membranes) which allow direct activation of the myofilaments by external Ca²⁺ and measurement of the resultant steady state force development (27). Preliminary data indicate that there is a reduction in the potential of maximal force development, although the half-maximal activation of the myofilaments occurred at the same Ca²⁺ level (28). A reduction in maximal force development could be related to the previously described myolysis in hibernating myocardium (19). Although histology can not directly quantify myolysis, light microscopy images clearly show glycogen accumulation which, in itself, will already reduce the volume density of the myofilaments (Figure 2). Further studies to quantify this are still underway in our model. Another explanation for the reduced maximal force delopment could be the degradation of troponin I, as proposed for acutely stunned myocardium in small animal models (29), even if this could not be confirmed in a pig model of myocardial stunning (30). Troponin I and potential degradation products were analyzed using immunoblotting, and no significant differences between samples from HIB and CTRL were detected (25). However, it cannot be excluded that small but functionally important differences are below the detection level of this technique. A significant switch in isoform expression of the major contractile proteins could not be detected using one-dimensional gel electrophoresis (Figure 3); however, this is being investigated in more detail.

Figure 2).

The histology of hibernating myocardium. A Periodic acid-Schiff staining of a transmural section of myocardium (original magnification ×100) shows the absence of necrosis and positive staining for glycogen predominantly in the subendocardial and midmyocardial layer; shown enlarged in the right panel (original magnification ×400). B The glycogen and positive staining is removed after amylase treatment

Figure 3).

Isoform composition of contractile proteins. Silver-stained gel of contractile proteins in a sample from matched control pigs (lane 1) and from myocytes from the dysfunctional myocardium of the pig model (lane 2). Protein bands include: beta-myosin heavy chain (β-MHC); myosin binding protein C (MyBP-C); troponin T (TnT); tropomyosin (TM); troponin I (TnI); and myosin light chain (MLC) 1 and 2. All indicated proteins have been previously identified with a specific monoclonal antibody in Western immunoblotting

Currently unexplained is the profound reduction in the rate of contraction observed during unloaded shortening of the isolated myocytes. Such a slowing was not observed in the skinned myocytes and, thus, probably results from alterations in excitation-contraction coupling in the intact myocytes. The reduction in Ca²⁺ current could, perhaps, contribute to this. Additionally, a reduction in the density of the T-tubular structures could lead to dyssynchrony of Ca²⁺ release, as shown in myocytes in primary culture (31). Possible alterations in transverse-tubular structures in HIB are currently being investigated.

SUMMARY AND PERSPECTIVES

Our studies describe for the first time the intrinsic properties of myocytes from hibernating myocardium. We demonstrate that cellular remodelling is a major factor contributing to contractile dysfunction of the hibernating myocardium, and that reduced contractility in vivo is not only related to the local environment.

The observed alterations in excitation-contraction coupling are of unique signature. In contrast with models of heart failure (32,33), we saw only modest changes in Ca²⁺ handling, although there was a significant reduction in L-type Ca²⁺ current. We could not detect changes in the activity of the Na⁺/Ca²⁺ exchanger, as opposed to what has been observed in tachycardia-induced failure (34). Contractile dysfunction is also apparently related to a reduction in the maximal myofilament response, which may, in turn, be due to a reduction in myofilament density in hibernating myocardium.

Recently, Canty et al (35) reported an increased incidence of sudden death in their pig model of chronic hibernation. In humans as well, increased mortality observed with viable hibernating myocardium may, in part, be due to sudden (and presumed arrhythmic) death. Our observation of a prolonged action potential indicates that there is electrical remodelling on the cellular level, which may promote arrhythmias. The underlying changes in ion channels remain to be identified.

CONCLUSIONS

Hibernating myocardium has a unique remodelling profile. Along with the changes observed in myocytes from the border zone of a myocardial infarction, or in the remote area, this specific phenotype adds to the diversity and complexity of the remodelling processes in ischemic cardiomyopathy. This diversity could relate to specific signalling pathways being activated with regional variations in mechanical load, metabolic demands and perfusion. Identifying these pathways and mechanisms is one of the challenges ahead.