Myocardial relaxation in human heart failure: Why sarcomere kinetics should be centerstage.

Abstract

Myocardial relaxation is critical for the heart to allow for adequate filling of the ventricles prior to the next contraction. In human heart failure, impairment of myocardial relaxation is a major problem, and impacts most patients suffering from end-stage failure. Furthering our understanding of myocardial relaxation is critical in developing future treatment strategies. This review highlights processes involved in myocardial relaxation, as well as governing processes that modulate myocardial relaxation, with a focus on impairment of myocardium-level relaxation in human end-stage heart failure.

Introduction

Roughly one in nine deaths in the US is due to heart failure, with an annual cost of patient care exceeding 30 billion US$, and those costs are expected to increase to 53 billion US$ by the year 2030(23). Heart failure is a debilitating disease that is often characterized by a heart that is not strong enough to pump sufficient amounts of blood. The weakness of the heart is actually only one of the potential underlying causes of end-stage heart failure, since even in end-stage failure the heart pumps enough blood to keep the patient alive. However, when patients exert themselves by, for instance, climbing stairs, cardiac pathology becomes more evident as seen by the inability of the heart to increase its cardiac output. Over the past decades, we have come to realize that the majority of patients suffer not necessarily from impaired contractile function, but from impaired relaxation, and for a large part of this heart failure patient population this impaired relaxation is the major pathological problem.

Myocardial relaxation is a complex process that involves multiple distinct rate- limiting steps, including calcium transient decline, thin filament de-activation, and crossbridge cycling kinetics/detachment(6). Each of these processes must take place to allow for relaxation to occur. In addition, each of these processes is further critically modulated by mechanical factors such as pre-load, after-load, and the rate and direction in which the sarcomeres move(7, 9, 28). Understanding of this process has proven extremely difficult, since the vast amount of players and processes that determine and regulate myocardial relaxation necessitate an inclusive systems approach to study these processes, as they do not work in isolation. The role of the involved processes is not only complex and interactive, but during different segments of a heartbeat they each exert distinct levels of involvement and regulatory importance. Relaxation of the heart is therefore akin to an emerging systems property, and a failure of this system is a complex process to dissect and understand. The goal of this review is to point out the central role the sarcomere and its components play in the relaxation process, and why they are crucial to further our understanding of human heart failure.

Human versus small rodent myocardium

The study of human myocardial relaxation is complicated by the fact that the kinetics of this process are about 10 times slower than in the most commonly used and powerful model in cardiovascular research, the laboratory mouse(25), and this hampers unambiguous translation of findings from mice to humans. The resting murine heart beats ~10 times faster than a resting human heart, and despite a very similar anatomy of the myocytes, sarcomere, and nearly identical afterload/blood pressure, the vast majority of “players” that are involved in determining myocardial relaxation are much different in the mouse(36). These differences are found at all levels; ion channels, EC- coupling proteins, and myofilament isoforms all significantly differ between mice and humans, as they need to be working in a much different frequency range. Sometimes the molecular differences are very small, but even small differences, for instance in the myosin isoform in small rodents, can have very significant functional consequences(24). The species differences are further complicated by the fact that in the mouse the kinetics of contraction and relaxation are only very modestly modified upon exercise. Whereas the mouse only changes cardiac output by ~20–30%, in the healthy human this change in cardiac output is easily 200–300%, and even substantially more in extreme cases. Thus, although the laboratory mouse has, is, and will advance our knowledge on the molecular and regulatory level of many processes of the heart, it is often unsuitable to study quantifiable determinants of relaxation as they relate to human physiology and disease(25).

Human myocardium and contractile strength

Several laboratories have used isolated human myocardium in an in vitro/ex vivo setting to assess contraction and relaxation properties. Over the past decades, studies have investigated contractile properties of human myocardium to further understand the working of the human heart and how and to what extent it malfunctions. Because the human population has a much more widespread genotypical and phenotypical variety, the variability in experimental outcome of measured contractile parameters has a significantly larger spread than data obtained in in-bred animal models. Although theoretically this can be easily overcome statistically by simply increasing the n-number of subjects in the study, this has practically proven to be difficult. Working with human myocardium, and especially freshly isolated myocardium that allows for the assessment of dynamic contractile parameters, typically hampers studying large groups of subjects, unless it involves studies conducted over a longer timespan. As a result of the conflict between the need for large n-numbers and the scarceness of available tissue, many past experiments on human myocardium were hampered by a lack of statistical power, sometimes resulting in conflicting results (8, 14–22, 35, 41–43, 45). However, on average these previous studies found no or little difference in developed contractile strength of viable myocardium at rest. A recent study conducted over 6 years (10) on >80 human hearts confirmed that there was indeed no significant difference inactive developed force of viable right ventricular trabeculae between failing and non-failing myocardium at rest. This may seem at odds with our view of heart failure as predominantly a weak heart, but this outcome is actually expected given the fact that the failing heart at rest, and the non-failing heart at rest, produce a very similar cardiac output. Indeed, heart failure typically manifests phenotypically under conditions where cardiac output is elevated, such as with exercise. When human myocardium in vitro is subject to the conditions of exercise, causing an elevation in heart rate, contractile deficits clearly become obvious in the end-stage failing myocardium. Healthy human myocardium increases developed force when heart rate is elevated, failing myocardium either increases less, but more often does not increase at all or even decreases when heart rate is elevated (10, 16, 40).

Impaired relaxation in human myocardium

Although at rest force of contraction is largely unaffected, impaired human myocardial relaxation is a long-known prominent pathological component in the majority of patients suffering from heart failure(13). In most, but not all, past human myocardium studies, slower kinetics of contraction and relaxation have been observed. Again, given the need for large n-numbers, and the generally hard to procure human tissue, some of the past results were hampered by lack of statistical power. Our recent study on 80+ human hearts determined that in end-stage heart failure, at least in the right ventricular myocardium of end-stage failing hearts, a significant deficit in relaxation kinetics is indeed present. This impaired relaxation is present at a resting heart rate, and this impaired relaxation was more prominent in myocardium of ischemic etiological origin compared to failing myocardium from non-ischemic etiology(10).

Determinants of myocardial relaxation

The large amount of players and processes that determine, influence, and regulate myocardial relaxation do not work in isolation, but rather in a complex concerted effort. First and foremost, they are tuned to work together in the heart rate range of the species in question. Although most species will share many of the proteins involved in various aspects of contraction and relaxation, for each species small but impactful changes of these individual proteins allow for their tuning with the other components to work in the desired heart rate range. Although individual changes in a protein, enzyme, and/or a biochemical- or mechanical action can be studied in isolation, understanding their role in the concerted relaxation process necessitates the assessment of their role within this concerted and tuned environment. In isolation, the functional role of a protein cannot be completely understood in relation to its contribution to the ultimate emerging systems property that encompassed myocardial relaxation. From decades of research, we identified several major processes that are involved in the cardiac relaxation process, as they are each necessary for the muscle to relax.

Intracellular calcium decline.

The intracellular calcium level that is increased to initiate contraction needs to decline to allow for relaxation to occur. In humans, about 30% of the calcium transient decline is accomplished by the activity of the Na⁺-Ca²⁺ exchanger, with the remaining 70% of the cycled calcium taken up back into the sarcoplasmic reticulum (SR) by the SR Ca²⁺-ATPase. In sharp contrast, mice almost exclusively depend on SR re-uptake, as transmembrane transport via Na⁺-Ca²⁺ exchanger is limited to only a few % of the total cycled Ca²⁺ ions during a contraction-relaxation cycle(5, 36). There is however not necessarily a given level of calcium that needs to be reached to allow for relaxation, since other factors too play a role in calcium-initiated myofilament (de-)activation. In fact, once the calcium transient has declined below the threshold for further activation, the rate of decline of the calcium transient appears to be no longer a regulatory factor for relaxation. Most experimental work has shown that the quantitative rate of decline of the calcium transient is typically slower than the rate of force relaxation(3, 39). Thus, even though a decline to a certain Ca²⁺ level is essential for relaxation to occur, the rate and regulation of relaxation appears mainly independent of the rate of intracellular calcium decline. Thus, one cannot infer an improved relaxation solely from an increased rate of calcium decline, as they are not directly related.

Thin filament de-activation

A second factor in relaxation is that the thin filament activation needs to cease in order to prevent additional myosin head binding to the thin filament(31). The calcium dissociation from Troponin-C is governed by the concentration of Ca²⁺ in the myofilament matrix, and by the dissociation constant of TnC for Ca²⁺ ions. It is often assumed that the calcium concentration reflects the amount of TnC that has calcium bound, i.e. it is assumed that calcium binding and release of TnC is in a rapid equilibrium. That is however not the case, the association and dissociation rates of calcium for TnC are highly dependent on the “level of organization”. In isolation, i.e. in solution, the TnC calcium kinetics are extremely fast, but in the cellular setting, with all other proteins present, these rates are much, much slower, to the extend that they can be quite similar to the rate of force relaxation, and thus can become rate-limiting(11). This implies that even when the calcium concentration is very low, a substantial amount of Ca²⁺ can still be bound to TnC. Moreover, TnC-Ca²⁺ dissociation is indirectly modulated by the contraction itself. When the ventricle ejects blood, sarcomere length decreases, and with the decrease in sarcomere length a decrease in calcium sensitivity is induced(1, 12). The length-dependent activation, which is for a large part due to calcium sensitization of the myofilaments is now reversed, causing an ejection- dependent deactivation. Thus, TnC-Ca²⁺ dissociation kinetics can play a major role in determining the rate of relaxation.

Cross-bridge detachment

A third “must-happen” factor for myocardial force to decline is that cross-bridges have to detach from the thin filament. The detachment rate of cross-bridges is generally seen as the rate-limiting step of the cross-bridge cycle. Most of the data on crossbridge cycling has been obtained from perturbations from a steady-state activation, often in skinned muscles, fibers, or myofibrils. In the beating heart, no steady-state contraction is reached, and since cycling behavior is generally dependent on the level of activation, the cycling behavior of cross-bridges in intact, contracting myocardium is still only partially resolved and understood. Assessment, or approximation of cross-bridge cycling behavior in human myocardium has been done mainly in skinned preparations, typically at cold temperatures. Extrapolation of such data to body temperature, as well as recent studies on intact human myocardium in which the rate of tension redevelopment was assessed(35), shows that the cycling rate of cross-bridges in human myocardium is rather slow, with time constants in the same range as TnC calcium dissociation, and relaxation of force. This implies that a single given crossbridge likely only cycles once or twice during a heart-beat, and that the overall detachment rate of cross-bridges could encompass a critical rate limiting step, specifically in late relaxation where relaxation of force is often mono-exponential.

Sarcomere lengthening

A fourth important modulator of relaxation is found not in the proteins that regulate the above processes, but in the direction and amplitude the sarcomeric proteins moves. When a sarcomere is stretched near or right after the end of systole, either by a single stretch(9), by application of an external vibration(28, 29), or by stretch by a neighboring sarcomere(47), relaxation has been shown to greatly accelerate. A recent study by Chung and coworkers(9) showed that in load-clamped twitch contractions, it was end systolic strain rate, and not solely afterload, that accelerated the myocardial relaxation by accelerating the detachment of the cross-bridges. Earlier studies showed that application of a low amplitude sinusoidal vibration at the peak of an isometric contraction greatly accelerated relaxation(28), and this acceleration impact could be observed throughout the relaxation process(29). In general, a lengthening sarcomere likely gives the myosin head a much lesser chance of binding to the actin binding site than in an isometric or shortening sarcomere. This “direction-dependent detachment” phenomenon also helps explain why slow isovolumic relaxation correlates with impaired relaxation during early filling in patients with diastolic dysfunction(50). As soon as the myocardium start to relax, and sarcomeres lengthen, it would provide a mechanical way of cessation of additional cross-bridge formation, despite a potentially activated thin filament. If this lengthening does not happen, or happens less or slower, cross-bridges can still attached, and/or remain attached longer, impairing filling. In addition to an absolute change in sarcomere length, the prevailing absolute sarcomere length can too impact the myocardial relaxation rate. The regulatory protein MyBP-C is located on a distinct part of the thick filament. Since MyBP-C has been shown to modulate the cross-bridge cycling kinetics(32, 34, 44) and is one of the few proteins that can change contraction-relaxation coupling (26, 27), not only the amount and direction of relengthening is important, but likely also the specific prevailing sarcomere length can impact myocardial relaxation.

Regulation of relaxation

In the human heart, the rate of relaxation is modulated to allow the heart to perform over a wide range of heart rates, with relaxation rates accelerating up to 30–45% from rest to peak heart rate. Such modulation barely occurs in the mouse, where relaxation changes by only 5–10% from rest to peak heart rate(25). The 3 most prominent regulators of contractile strength are volume (sarcomere length), heart rate (frequency), and β-adrenergic stimulation (the flight-fright-fight response). Each of these three regulatory mechanisms impact not only contractile strength, but also impact the kinetics of contraction and relaxation. First, with increased volume (i.e. increase muscle or sarcomere length), relaxation becomes slower(2). This is mainly due to myofilament properties, as calcium transients change only slightly, and more so, in the opposite direction; the calcium transients decline slightly accelerated when sarcomeres are stretched, while force relaxation slowed down(2, 39). This later finding again argues against the rate of calcium transient decline being a prominent determinant of relaxation, since calcium decline and force decline are uncoupled. The slower force decline with increased sarcomere length is predominantly attributed to increase myofilament calcium sensitivity(l), and may additionally involve post-translational modification of myofilament proteins(38). Second, when heart rate increases, relaxation accelerates(26). The calcium transients both gain in amplitude and accelerate in decline rate, while in addition the myofilaments desensitize to activator calcium(48), allowing for faster relaxation to occur, despite the increased intracellular diastolic calcium levels. Rise of intracellular calcium with frequency show that diastolic calcium levels at high rates can even exceed systolic levels at low rates(46), necessitating a substantial frequency-dependent myofilament desensitization to maintain force development(48, 49). Third, β-stimulation causes a similar impact as the elevation of heart rate alone (note that heart rate elevation is mainly due to an increase in β- stimulation), with additional phosphorylation targets activated that overall increase contractile strength, and aid in acceleration of relaxation(33).

Myofilament properties, the sarcomere takes stage

The collective proteins, processes, and regulatory parameters in the myocardium discussed above govern the relaxation of force development as an integrated system. One could view the myocardial relaxation process as an emerging systems property, with no clear single molecular target, pathway, process, or regulator that is the main driver. That said, the proteins and processes that are most likely to be the major players have in common that they are an integral part of the sarcomere. Although EC- coupling is a critical part of the initiation of contraction, the determinants of kinetics of contraction, and those of relaxation, are located and regulated in and by the sarcomere as a whole(4, 30). Troponin-C as the critical Ca²⁺-sensor does not work in isolation to activate the thin filament; an integrated process that involves many other proteins feed back and co-determine the apparent association and dissociation constants for calcium(11). In addition, cross-bridge cycling kinetics play a major part in the kinetic regulation of the heart(37). Lastly, the external and internal loads that the sarcomere both carries and generates resulting in sarcomere length changes impacts TnC-Ca²⁺ binding as well as cross-bridge cycling properties. In human heart failure, the kinetic rate of contraction and relaxation are prominently involved in the manifestation of dysfunction, even at rest(10), possibly more often so than a weak contraction, and a further understanding of the sarcomere as an integrated system will be needed to strategize sarcomere-based treatment for this sarcomere-based dysfunction.