Dysfunctional sarcomeric relaxation in the heart

Abstract

Since cardiac relaxation is commonly impaired in heart failure caused by many different etiologies, identifying druggable targets is a common goal. While many factors contribute to cardiac relaxation, this review focuses on sarcomeric relaxation and dysfunction. Any alteration in how sarcomeric proteins interact can lead to significant shifts in sarcomeric relaxation that may contribute to diastolic dysfunction. Considering examples of sarcomeric dysfunction that have been reported in 3 different pathologies, hypertrophic cardiomyopathy, restrictive cardiomyopathy, and heart failure with preserved ejection fraction, will provide insights into the role sarcomeric dysfunction plays in impaired cardiac relaxation. This will ultimately improve our understanding of sarcomeric physiology and uncover new therapeutic targets.

1. Introduction

Cardiac relaxation has become an important topic of investigation due to increased awareness of heart failure with preserved ejection fraction (HFpEF) and a lack of effective treatments to improve relaxation. While HFpEF is characterized by diastolic dysfunction, it is not the only condition where impaired relaxation occurs^{1, 2}. In fact, diastolic dysfunction has been associated with physiologic aging as well as other types of heart failure (HF) and cardiomyopathies^3–6. Specifically, diastolic dysfunction is defined as decreased ventricular compliance due to impaired relaxation which leads to elevated cardiac pressures and decreased ventricular filling^7–9. Impaired relaxation can be caused by structural, cellular, or metabolic perturbations. While there are multiple factors that contribute to diastolic dysfunction, in this review, we will focus on specific sarcomeric mutations or post-translational regulations that alter relaxation.

Relaxation is mediated by thin filament inactivation and decreased cross-bridge cycling between actin and myosin. These steps are reliant on calcium removal, shifting of the thin filament to inhibit actin’s binding sites, as well as regulation of the myosin heads^10–13. Any mutation or post-translational modification that alters how these proteins interact can lead to significant shifts in relaxation that may contribute to diastolic dysfunction. Experimentally, the use of a specialized apparatus that measures sarcomeric dynamics has elucidated the presence of two phases of relaxation: a linear phase, which reflects the off-rate for cross-bridge cycling, and an exponential phase caused by the sliding of the sarcomeric proteins back to resting sarcomere length^{11, 14, 15} (Figure 1). This technique and others allow assessment of sarcomeric relaxation dynamics and combined with known sarcomeric protein mutations and/or modifications and myocyte and cardiac phenotypes, can elucidate aspects of relaxation regulation. Several recent studies provide insights into specific sarcomeric mutations that cause hypertrophic cardiomyopathy (HCM) that, in addition to leading to a hypercontractile state, are also associated with impaired relaxation. Furthermore, sarcomeric modifications have been implicated in restrictive cardiomyopathy (RCM) and HFpEF, both conditions where cardiac dysfunction is primarily due to impaired relaxation. Considering examples of sarcomeric dysfunction that have been reported in these different conditions that can lead to HF will provide insights into the role sarcomeric dysfunction plays in impaired cardiac relaxation and ultimately, improve our understanding of sarcomeric physiology to uncover new therapeutic targets.

Figure 1:

Representative activation/relaxation traces of myofibrils derived from hiPSC-CMs. The activation, reactivation, and relaxation phases of the trace are each labeled. In the inset box, relaxation is shown in higher resolution, demonstrating that it is biphasic (linear and exponential phases).

2. “Sarcomeric cardiomyopathies”

2.1. Overview HCM and RCM

HCM is a highly studied disease in which more than 1400 mutations in mainly sarcomeric proteins lead to cardiac dysfunction. In fact, 50% of these gene mutations are in β-myosin heavy chain (MHY7) or myosin binding protein C3 (MYBPC3)^16–18. While the HCM phenotype is typified by hypertrophic ventricles, diastolic dysfunction is another important component of HCM. Deciphering the contribution of sarcomeric dysfunction to whole organ relaxation is complicated by the fact that structural changes associated with hypertrophy and increased fibrosis modify ventricular compliance. However, several studies demonstrate that impaired sarcomeric relaxation may also be involved. Along with this, patients with RCM also have severe diastolic dysfunction and a percentage of these patients also have sarcomeric mutations. Several reviews have considered HCM and RCM as a spectrum of a single disease^{18, 19} and this approach may shed light into mechanisms that impact sarcomeric relaxation.

2.2. Super relaxed state and relaxation

Myosin II serves as the motor unit of the sarcomere, hydrolyzing ATP to drive muscle contraction²⁰ and is comprised of six individual proteins: 2 myosin heavy chains, either MYH6 or MYH7, 2 calcium binding regulatory myosin light chains (RLC), and two essential myosin light chains (ELC). Mutations in myosin heavy chain 7 (MYH7) are the most common cause of cardiomyopathy, while both RLC and ELC mutations have also been associated with HCM²¹. X-ray diffraction experiments have demonstrated that myosin heads exist in distinct structural conformations. In the past several years use of the fluorescent ATP analog mant-ATP has enabled measurement of single-nucleotide turnover by myosin ATPase, allowing calculation of the proportion of myosin heads in the disordered relaxed state (DRX) versus the super relaxed state (SRX) in tissue samples as well as in isolated cells^{22, 23}. In SRX, the two myosin heads fold together in a conformation that places them far from the thin filament and in this position, they hydrolyze ATP at a slower rate^24–27 (Figure 2). Recent studies in animal and cell-based models demonstrate that mutations that modify SRX are associated with diastolic dysfunction. Future studies investigating how the proportion of myosin heads available to interact with actin alters the off-rate of the cross-bridge cycle may lead to insights into specific sarcomeric mutations are associated with changes in sarcomeric relaxation. While the research conducted in this space to date has yielded interesting and important insights, there are several important caveats to keep in mind when considering these studies. These include 1) this is an emerging field of study and while altered SRX-DRX equilibrium may contribute to disease phenotypes, dysregulation of SRX-DRX has not been shown to be directly causative of disease phenotypes; 2) current techniques are largely qualitative; and 3) there is still debate in the field regarding the physical intermediary states of myosin structure.

Figure 2:

Role of myosin states in cardiomyopathies. A) Structural models of the open or active and closed or interacting heads motif (IHM) of myosin. The IHM is structurally related to the myosin super relaxed state (SRX) although the exact interplay of these two related states is still being determined. Adapted from Anderson et al., PNAS, 2018 ¹⁷. B) Diagram displaying the effects of relative levels of SRX versus active or disordered relaxed state (DRX) of myosin on cardiac function and disease. The greater the proportion of myosin heads in SRX, the lower the cost of contraction, which is largely associated with restrictive cardiomyopathies. By contrast, the greater the percentage of myosin heads in DRX, the greater the energy cost of contraction, which is associated with hypertrophic cardiomyopathies. Figure created with BioRender.

Using a combination of in vivo (HCM mutant mouse models) and in vitro (mutant human induced pluripotent stem cell derived cardiomyocytes (hiPSC-CMs)) approaches, Toepfer et al. characterized the pathogenic myosin mutations R403Q, V606M, and R719W. Each mutation decreased the number of myosin heads in SRX and caused prolonged relaxation compared to cells/animals expressing WT myosin²⁸. These authors also reported that mavacamten (a myosin inhibitor) normalized the proportion of myosin heads in SRX and relaxation duration. Another study demonstrated prolonged cellular relaxation in hiPSC-CMs from patients with HCM²⁹. Complementary studies also demonstrate that the number of interacting myosin heads is increased in the R403Q mutation³⁰. Even though cellular relaxation was prolonged in this study, isolated and permeabilized myofibrils from a patient with the R403Q mutation relax more quickly than do WT myofibrils³¹; similar findings have been observed in myofibrils from HCM hearts without known mutations³². These findings suggest that cross-bridge off-rate is faster in HCM-inducing mutations that increase DRX (likely due to weaker interaction with actin). Moreover, this discrepancy between cellular and myofibril relaxation highlights a key difference in sarcomeric mechanical interactions and overall cellular function. The MYH7 mutation R403Q increases DRX; therefore, more myosin heads are available to form cross-bridges³⁰. However, myofibrils expressing this mutation have a faster off-rate (the slow or linear phase of relaxation)³¹. In contrast, the prolonged cellular relaxation reported by Toepfer et al. may be due to altered Ca²⁺ signaling, which manifests in intact cells but is experimentally controlled in ex vivo myofibril preparations. Another important consideration is the energy cost in the HCM-causing sarcomeric mutations. Vitale et al. discusses how a faster off-rate reflects a higher energy cost, which at the cellular level, may lead to an energy deficit and thereby alter relaxation³³. In line with this, another HCM-causing mutation in troponin T (K280N) has been shown to increase contractile energy cost and increase off-rate kinetics³⁴. Even though this mutation may not be representative of HCM as a whole as it is a homozygous mutation reported in a single patient, studying sarcomere mechanics can provide insights to common factors that contribute to HCM phenotypes.

MYBPC3 mutations, which represent a significant portion of HCM, have been reported to decrease the number of myosin heads in SRX^{35, 36}, while increasing cellular relaxation time³⁷. In addition to mutations, a recent study demonstrated that altered phosphorylation of MYBPC3 can also alter SRX where phosphorylation of MYBPC3 increased DRX³⁸. While the previous study did not measure relaxation kinetics, another study by Rosas et al (2019) reported that phosphorylation of MYBPC3 is decreased with ageing in mice and expression of mutant MYBPC3 with serine to aspartate (phosphormimetic) substitution, could attenuate age-induced cardiac hypertrophy and dysfunction while improving relaxation. In contrast, expression of serine to alanine (phosph-null) MYBPC3 largely caused the opposite changes, exacerbating age-induce dysfunction. These changes occurred independent of age-induced fibrosis ³⁹. Taken together, these studies implicate a role for MYBPC – associated changes in SRX with altered relaxation.

Along with MHY7 and MYBPC2 mutations, certain mutations in the myosin light chains also alter SRX and cause HCM. Myofibrils isolated from mice expressing the A57G mutation in myosin ventricular essential light chain (MYL3) show significantly lower levels of SRX^{40, 41}. Echocardiography also revealed prolonged isovolumetric relaxation time, indicating diastolic dysfunction⁴¹. Similar investigations of the role of posttranslational modifications in age induced or disease induced relaxation dysfunction will likely be an area of continued interest.

A recent study by Tobacman and Cammarato leveraged this evolving catalog of mutations and demonstrated that pathogenetic troponin mutations disrupt contact with tropomyosin or actin. Mutated regions normally prevent contraction at low calcium by sterically hindering myosin; therefore, mutations in these regions reduce this steric hinderance and increase calcium sensitivity and prolonged relaxation⁴².

As well as considering how sarcomeric mutations in HCM contribute to diastolic dysfunction, studying sarcomeric mutations in RCM may elucidate protein targets to modify relaxation. Sarcomeric mutations that induce diastolic dysfunction associated with RCM can provide insights into targets that modify relaxation. A specific mutation in myosin essential light chain (ELC, E143K) identified in patients with RCM has been shown to increase the number of myosin heads in SRX⁴¹. Interestingly, this is in contrast to mutations in HCM that increase DRX and are associated with diastolic dysfunction^{28, 40}. Once again, this highlights that relaxation kinetics are determined by a complex interaction involving cross-bridge cycling rate and non-sarcomeric mechanisms, as opposed to the number of myosin heads interacting with actin in isolation.

Moreover, only a portion of sarcomeric mutations associated with HCM or RCM are associated with modified SRX/DRX equilibrium. There are many other mechanisms by which relaxation can be impaired. An intriguing concept postulates that several HCM-causing mutations in cTnI or cTnC perturb the interaction within the troponin complex and rather than affecting calcium binding affinity, rather impact function due to in ability to respond to protein-kinase A signaling⁴³. One of the most well-documented mediator of myofibril relaxation is phosphorylation of cTnI at Serine23/24⁴⁴. Increased phosphorylation at this site speeds up relaxation. Cheng et al (2016) reported on an HCM causing mutation in cTnI P82S that in a rodent model of mutation had a blunted myofibril relaxation response to PKA treatment, demonstrating that in myofibrils expressing this mutation, phosphorylation at cTnI 23/24 does not induce faster relaxation⁴⁵. Another example of a mutation that leads to RCM but does not affect SRX/DRX is the TNNT2 (cardiac troponin T) mutation R94C. The cTnT R94C was identified in a single family and can also lead to RCM⁴⁶. This mutation leads to increased calcium sensitivity such that the thin filament is active at lower calcium concentrations⁴⁶. By destabilizing tropomyosin at low calcium states, this leads to less inhibition of myosin binding. Another important consideration in the instance of this mutation, which is also common in other types of familial myopathies, is that while the children who express the R94C TNNT2 mutation in this study have a severe phenotype, the parental carrier of the mutation displayed mild phenotype. This discordance demonstrates that the mechanisms contributing to the cardiac phenotype are complex, and that other modifiers likely influence function.

These examples of sarcomeric dysfunction leading to impaired relaxation in multiple forms of cardiac dysfunction highlight both the central importance of relaxation to causing cardiovascular disease, as well as the importance of considering the impact of mutations in individual proteins on the complex interactions of the entire sarcomere.

3. Diastolic Dysfunction and HFpEF

3.1. Overview

In contrast to HCM and RCM, known sarcomeric mutations have not been documented to cause HFpEF. Rather, recent studies have suggested the involvement of differential post-translational modifications of sarcomeric proteins inducing functional changes contributing to diastolic dysfunction.

3.2. Dysfunction of sarcomeric relaxation

The heterogeneity of HFpEF phenotypes has been well documented⁴⁷. In a recent report⁴⁸ patients with HFpEF with high blood pressure and hypertrophy were separated from patients diagnosed with obesity and diabetes. This study demonstrated that passive stiffness and maximal tension were significantly higher in cardiomyocytes isolated from right ventricular biopsies from patients with high blood pressure compared to patients with obesity and diabetes⁴⁸. While not defining specific sarcomeric differences underlying these changes, this study demonstrates that the etiology of HFpEF is critical in understanding how sarcomeric changes modify whole heart function. This is also evident in the myriad animal models of HFpEF or diastolic dysfunction, where different models mimic only certain aspects of HFpEF, and only specific models display sarcomeric modifications that regulate diastolic function. Potentially, these sarcomeric modifications could be due to a compensatory response to other ventricular, energetic, or inflammatory changes occurring in HFpEF, in a similar fashion to the differences observed between myofibril and cell-based relaxation discussed above. Determining which models have prolonged relaxation or changes in passive tension, on both the level of the cell and the myofibril, will have important implications for understanding how sarcomeric proteins are modified and if these modifications contribute to disease.

Myofibril relaxation is prolonged in the Dahl-Salt Sensitive (DSS) rat model of diastolic dysfunction, and this is prevented by treatment with an inhibitor of class I/II histone deacetylases (HDACs)⁴⁹. Moreover, in this study, directly treating myofibrils with p300 (inducing acetylation) or HDAC2 (decreasing acetylation) caused faster or prolonged relaxation, respectively. Identification of a lysine residue on cardiac troponin I that is differentially acetylated in this model demonstrated that acetylation of cTnI causes faster relaxation and decreased calcium sensitivity⁵⁰. In line with these studies, a recent report out of the Houser laboratory utilizing a young feline model of diastolic dysfunction suggest modified acetylation of sarcomeric proteins may regulate diastolic dysfunction⁵¹. In this study, young male cats underwent mild thoracic aortic banding, causing diastolic dysfunction with prolonged myofibril relaxation that was ameliorated by treatment with an HDAC inhibitor. Similar to the aforementioned DSS rat study, specific sarcomeric proteins that were acetylation by HDAC treatment have yet to be determined. However, sarcomeric proteins that affect relaxation may be inferred from mutations that induce impaired relaxation. Potentially, post-translational modifications of particular residues within those proteins, while causing less severe phenotypes, lead to impaired relaxation.

Another important mediator of diastolic dysfunction that needs to be considered is titin. Titin is a sarcomeric protein that is one of the key regulators of cellular passive stiffness and relaxation. In fact, decreased phosphorylation of titin in the ZSF1 obese rat model of HFpEF was associated with increased passive tension⁵². Another study that explored diastolic dysfunction in several different models (aging, hypertension, and metabolic syndrome) demonstrated that nicotinamide treatment decreases passive tension in ZSF1 obese rats possibly through deacetylating titin⁵³. The results of this study highlight the challenge of determining true mechanism of action in HFpEF and treatment. In this study, nicotinamide decreased systemic blood pressure and fibrosis. To determine the mechanisms of sarcomeric relaxation, this case as with others, it will be critical to tease out the effects of compensatory such as decreasing the stiffness of the extracellular matrix.

A recent mouse model of HFpEF where mice exposed to a high fat diet and L-NAME demonstrated diastolic dysfunction and impaired cellular shortening; however, it is unclear if sarcomeric proteins are modified in this model⁵⁴. In another mouse model, diastolic dysfunction was induced by exposing animals to deoxycorticosterone acetate and inducing mild kidney dysfunction⁵⁵. Contrary to what was observed in the studies of diastolic dysfunction summarized above, myofibril relaxation was not prolonged⁵⁵. In this study, increased ventricular stiffness was due to increased intracellular fibrosis. As discussed above, considering how changes in extracellular matrix stiffness and sarcomeric relaxation dynamics interact is important to consider. Dissecting differences in the underlying factors contributing to diastolic dysfunction in these models may provide critical insights into mechanisms of sarcomeric involvement and modifications that induce relaxation dysfunction.

3. Conclusions

It is clear we cannot consider sarcomeric function or dysfunction in isolation. By collectively studying mutations in sarcomeric proteins that induce particular phenotypes, as well as differential post-modifications that lead to altered relaxation, we will be able to identify critical residues in these proteins that are essential for contraction and relaxation. It must also be noted that many studies defining how mutations impact function have been conducted on patient tissue collected at end-stage failure. As mentioned above, given the significant structural, pressure, and drug-induced changes that occur in these hearts, it is difficult to draw conclusions about primary mechanism. In HCM with diastolic dysfunction, many sarcomeric mutations decrease SRX, yet do not reduce the myosin off-rate, as would be expected for prolonged relaxation – in fact, investigation into sarcomeric dynamics demonstrate that several mutations increase the off-rate and the energy cost of crossbridge cycling. While a subset of HCM hearts demonstrate quite consistent changes in sarcomeric protein regulation⁵⁶, modulation of SRX appears to be mutation specific^{28, 36}. It is apparent, therefore, that relaxation dysfunction is due to a combination of sarcomeric dysfunction with other factors, leading to modified energetics and calcium handling. Therefore, both sarcomeric and non-sarcomeric changes to induce diastolic dysfunction. Recent progress has been made in characterizing relaxation on the single cell level, including in stem cell derived cardiomyocytes⁵⁷, as well as in cardiac organoids^{58, 59}, both of which are amenable to genetic and pharmacological manipulation. Combining these approaches with mass spectrometry, allowing simultaneous characterization of expression and posttranslational modification of sarcomeric proteins with relaxation function⁶⁰ may allow for significant progress to be made in the coming years.