Research priorities in sarcomeric cardiomyopathies

, The clinical variability in patients with sarcomeric cardiomyopathies is striking: a mutation causes cardiomyopathy in one individual, while the identical mutation is harmless in a family member. Moreover, the clinical phenotype varies ranging from asymmetric hypertrophy to severe dila-tationoftheheart.Identiﬁcationofasinglephenotype-associateddiseasemechanismwouldfacilitatethedesignoftargetedtreatmentsforpatient groups with different clinical phenotypes. However, evidence from both the clinic and basic knowledge of functional and structural properties of thesarcomerearguesagainsta‘onesizeﬁtsall’therapyfortreatmentofoneclinicalphenotype.Meticulousclinicalandbasicstudiesareneededto unraveltheinitialandprogressivechangesinitiatedbysarcomeremutationstobetterunderstandwhymutationsinthesamegenecanleadtosuch opposing phenotypes. Ultimately, we need to design an ‘integrative physiology’ approach to fully realize patient/gene-tailored therapy. Expertise within different research ﬁelds (cardiology, genetics, cellular biology, physiology, and pharmacology) must be joined to link longitudinal clinical studies with mechanistic insights obtained from molecular and functional studies in novel cardiac muscle systems. New animal models, which reﬂect both initial and more advanced stages of sarcomeric cardiomyopathy, will also aid in achieving these goals. Here, we discuss current priorities in clinical and preclinical investigation aimed at increasing our understanding of pathophysiological mechanisms leading from mutation to disease. Such information will provide the basisto improve risk stratiﬁcation and to develop therapies to prevent/rescue cardiac dysfunction and remodelling caused by sarcomere mutations. - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - mutation carriers with a mild phenotype, which are expected to have a dynamic change in clinical phenotype within a relatively brief period of time. Basic studies should investigate myocardial effects of secondary disease modiﬁers and mutation-speciﬁc effects on the heart muscle. PQC, protein quality control; PTM, post-translational modiﬁcation.


Health care problem
Sarcomeric cardiomyopathies, caused by mutations in genes encoding proteins of the sarcomere, constitute one of the most common causes of sudden cardiac death (SCD) in the young and represent major causes for cardiac transplantation. Disease onset typically ranges between 20 and 50 years of age, thus affecting patients in the prime of their life, and earlier onsets represent an important cause of childhood cardiomyopathy. Because of advances in cardiovascular genetics during the last three decades, 1,2 a large number of mutations have been identified in patients with cardiomyopathies previously considered to be idiopathic. Consequently, an increasing number of mutation carriers are being followed. Furthermore, due to increased availability of genotype analysis, at-risk family members can be identified prior to the emergence of a clinical diagnosis. This type of genotype-directed family evaluation is a growing focus of clinical management. However, the use of genetic insights and genetic diagnosis in risk stratification and therapeutic strategies in the clinic remains limited. 3,4 The recently developed model to calculate SCD risk in hypertrophic cardiomyopathy (HCM) includes many clinical variables (e.g. left ventricular wall thickness, left atrial dimension, left ventricular outflow tract gradient, age), 5 but it does not yet include genetics. Current drug therapy is aimed at management of symptom palliation and 'watching and waiting' to see if at-risk mutation carriers develop clinical cardiomyopathy. There is a great need for proactive risk stratification, including genotype, and treatment strategies that will be able to delay and ultimately prevent disease development in mutation carriers. The existence of (rare) neonatal forms of sarcomeric cardiomyopathy that rapidly evolve into systolic heart failure and death within the first year of life 6 -13 reveals a need for molecular diagnosis early after birth or during pregnancy to be able to rapidly give the appropriate treatment. With such advances, genetic discoveries will meet their full potential to change medical practice.
There is impressive clinical variability between affected mutation carriers. 14 Importantly, within the same family, a specific mutation may cause cardiomyopathy and SCD in one relative, while it may appear entirely harmless in another family member. Disease onset and severity differ widely; implying that additional determinants, including genetic variations, environmental and/or toxic disease triggers, and an age-related decline in protective mechanisms (protein quality control systems) 15 -18 may have substantial impact on disease susceptibility and outcomes. Mechanisms governing how mutations cause disease and how secondary disease modifiers impact outcomes are largely unknown. As a result, it remains challenging for cardiologists to perform accurate risk stratification or initiate appropriate preventive therapy for the individual patient, as risk may be underestimated and left undertreated in some individuals, but overestimated in others, who are unnecessarily exposed to drugs or devices with potential side effects and complications. Finally, understanding of human genetic variation is rapidly evolving. Investigation of normal populations has led to the recognition of the prevalence of non-pathogenic sequence variation in genes associated with inherited cardiovascular disease. For example, mutations in genes encoding transmembrane potassium and sodium channels, KCNQ1, KCNH2, and SCN5A, are known to cause long QT syndrome. However, sequence variation in these same genes can also be seen at a low rate in normal population. 19 Similarly, rare sarcomere gene variants were also identified in 11% of the general population. Although almost none of these individuals had clinical manifestations of HCM, the presence of a rare sarcomere variant was associated with increased risk of adverse cardiovascular events. 20 These observations suggest that existing mutation databases are likely 'contaminated' with polymorphisms with little or no clinical relevance, further challenging accurate interpretation of genetic testing.

Cardiomyopathy phenotypes
Sarcomeric cardiomyopathies are associated with abnormalities in myocardial structure and function. Sarcomere gene defects have emerged as a shared aetiologic element, but genotype-phenotype relations are complex. Sarcomere mutations have been associated with distinct, apparently divergent patterns of ventricular remodelling, including hypertrophy, dilation, non-compaction, and restrictive physiology. HCM is typically characterized by left ventricular hypertrophy (LVH), small LV cavity size, and vigorous LV systolic function. In contrast, dilated cardiomyopathy (DCM) is characterized by an enlarged left ventricle with diminished systolic function. Many attempts have been made to identify a unifying mechanism for the so-called HCM-and DCM-causing mutations, which would explain development of hypertrophy or dilation of the ventricle, respectively. The genotype-phenotype relation analyses became even more complex with the finding that sarcomere mutations also result in restrictive cardiomyopathy (RCM) or left ventricular noncompaction (LVNC). Sarcomeric DCM and LVNC do not have specific phenotypes allowing a clinical distinction from the non-sarcomeric forms. In the case of DCM, for example, forms associated with MHY7, TNNT2, TNNI3, and TPM1 mutations lack the 'red flags' which characterize other genetic aetiologies such as lamin A/C (conduction disturbances) or skeletal muscle involvement (dystrophin complex). If anything, sarcomeric DCM may be suspected based on age at presentation and long-term course. Specifically, presentation early in life, from infancy to adolescence, is not uncommon and associated with severe outcome including SCD and refractory heart failure leading to death or transplantation. This is in sharp contrast with the mild course observed in relatives diagnosed as adults. This clinical profile differs meaningfully from that seen in other genetic causes of DCM, such as that caused by lamin A/C or phospholamban mutations where manifestations typically do not develop until adulthood and are progressive. 21 -25 Of note, subtle abnormalities in systolic function are present in subclinical DCM mutation carriers, despite normal left ventricular size and systolic function. In contrast, impaired relaxation appears to be the predominant early manifestation of sarcomere mutations causing HCM. These findings support the theory that the mutation's intrinsic impact on sarcomere function influences whether a dilated or hypertrophic phenotype develops. 13 The case of LVNC is even more complex, as even its nature as a distinct cardiomyopathy is object of debate. Mutations in MYH7, ACTC, and TNNT2 have been associated to date. 26,27 The most evident gap in current understanding of the phenotypic spectrum of sarcomeric cardiomyopathies, including HCM, DCM, RCM, and LVNC, is that the reasons for development of such different phenotypes from defects in the same genes is largely unknown.
Moreover, the phenotypic spectrum of cardiomyopathies is not limited to modifications in LV morphology or function. For example, atrial dilatation and dysfunction leading to atrial fibrillation (AF) is exceedingly common in both HCM and DCM. AF in turn promotes further dilatation of the atria in a vicious Cycle that is more common and occurs much earlier than in the general population. Although atrial changes have been attributed to the haemodynamic abnormalities and elevated filling pressures intrinsic to cardiomyopathies, the concept of a primary atrial myopathy and its implications for management have not been addressed. Because AF is common and a prominent cause of morbidity in cardiomyopathies, this issue requires further understanding. 28 As well elucidated for HCM, cells other than cardiomyocytes may be directly involved in the disease process, including the coronary arterioles, the interstitial fibrous tissue, and the mitral valve and subvalvar apparatus. Coronary microvascular function is markedly abnormal in both HCM and DCM, although likely due to different mechanisms. 29 In HCM, microvascular dysfunction is largely the result of smooth muscle hyperplasia causing severe reduction in luminal area. Microvascular dysfunction is associated with severe blunting in coronary vascular reserve, the most important cause of ischaemia in patients with HCM, and an important predictor of outcome. 14,30 Ischaemia occurring at the microvascular level is believed to represent the main cause of replacement-type fibrosis, a common finding in patients with cardiomyopathies, whose extent is directly related to the degree of systolic impairment. 31 In the majority of patients with HCM, the mitral leaflets are markedly enlarged, although with normal histological appearance, and show anomalous chordae and papillary muscle insertions. These abnormalities are observed in patients of all ages, including young children, and are thought to be one of the main determinants of anterior systolic motion and dynamic outflow obstruction. 32 The link between the genetic defect in sarcomeric proteins and these features is elusive and often dismissed as being secondary to the myocardial abnormalities. However, other mechanisms may be at play. For example, a single developmental pathway involving the proepicardial organ, i.e. the common progenitor of all these extramyocardial cell types, might provide a simple explanation to such apparent discrepancy and indicate novel targets for phenotype suppression or modification during the early stages of life. 33

Clinical studies
A clear-cut separation of patient groups on the basis of type of remodelling turns out to be unrealistic, as current clinical classification lacks adequate precision and specificity. HCM patients presenting with heart failure may have evolved a phenotype that resembles DCM or may manifest profound restrictive physiology. As such, their classification may differ depending on the predominant stage of disease at the time of clinical evaluation. In addition, in-depth understanding about how cardiomyopathy progresses and how sarcomere mutations lead to disease is currently limited. The development of end-stage remodelled myocardium represents the final stage of a long complex pathway from the initial genetic and biophysical insult ( Figure 1). Human studies have largely focused on the late stages of cardiomyopathy and therefore do not account for the confounding influence of disease itself on phenotypic progression or genotype-phenotype correlations. As such, prior studies have been importantly hindered in their efforts to identify primary disease mechanisms.

From mutation to cardiomyopathy: characterizing disease evolution
Identifying the genetic basis of disease is a prerequisite for developing 'personalized' medicine-rational, mechanism-based therapy. However, substantial challenges must be overcome to be able to initiate clinical trials and begin to realize this goal. We must first gain greater knowledge about the steps that lead from mutation to disease. Understanding the link between genotype and phenotype requires longitudinal clinical studies, starting with sarcomere mutation carriers early in life, before obvious disease has developed ( Figure 1). The potential power of this approach was recently demonstrated by Ho and colleagues, 34 who detected evidence for an early profibrotic state in HCM patients before the onset of hypertrophy or detectable fibrosis. Likewise, diastolic dysfunction has been reported in human HCM mutation carriers and transgenic mouse models before onset of the disease phenotype. 2,35 -37 Crilley et al. 38 reported a reduction in the cardiac PCr-to-ATP ratio, a measure of energetic status, in mutation carriers with or without LVH. Reduced myocardial efficiency at the pre-hypertrophic stage of HCM was confirmed using PET and CMR studies to assess myocardial oxygen consumption and external work, respectively. 39,40 Other changes in the early stage of cardiomyopathy include crypts 41 -44 and extracellular volume expansion assessed by cardiac MRI. 45 At this early stage, the cardiac sarcomere mutation represents the single pathogenic constant. Presumably, during ageing, additional disease modifiers ( Figure 1) initiate detrimental mutation-related changes in the heart. Longitudinal studies that begin early in life, anchored on genotype, are an ideal starting point for research on genotype-phenotype relations. Such studies will help to answer fundamental questions regarding penetrance, expressivity, and clinical outcomes. This is a highly ambitious effort that requires careful evaluation of a large number of families with affected and non-affected carriers as well as age-matched 'control' partners over long periods of time. Close collaboration is needed between all key investigators involved in sarcomeric cardiomyopathy, including geneticists, cardiologists, paediatrician cardiologists, and physiologists. Well-designed clinical studies will reveal disease mechanisms (epigenetic and environmental) that may modify the occurrence and progression of cardiomyopathy, which subsequently need to be investigated in basic studies (animal models, in vitro cell systems) to provide proof that these factors indeed enhance or delay onset and progression of disease ( Figure 1).
In addition to longitudinal clinical studies, unbiased statistical approaches, utilizing techniques of machine learning or cluster analysis, may prove fruitful in trying to better identify previously unrecognized patterns in diseases with complex, overlapping phenotypes. Such studies may provide new insights to clinical outcomes and risk prediction, as well as genotype-phenotype correlations.

Clinical trials of disease modification and prevention
As our understanding of disease pathogenesis improves, we will be able to develop novel treatments intended to delay or attenuate phenotypic evolution and, ultimately, to prevent the emergence of disease all together. Such strategies will target key early steps in disease evolution, identified by basic investigation. This approach is highly appealing, as it seems both more feasible and more desirable to treat to delay or prevent disease, rather than trying to reverse or rescue severe changes associated with entrenched, late-stage disease. Indeed, animal studies have suggested that disease-modifying treatment was ineffective if started after a clinical phenotype developed. 46,47 However, clinical trials to test these potential new therapies face unique and daunting challenges. For example, although the pathogenic sarcomere mutation is present at birth, overt disease may not develop for decades and in some cases may not develop at all or be associated with only very mild clinical consequences. Given the great variability in disease onset and course, who should be treated and when should treatment be started? Since mutations are present throughout life, will lifelong therapy be needed to inhibit phenotypic progression or emergence? Moreover, although sarcomeric cardiomyopathies can be associated with devastating outcomes, the event rate in the overall patient population is relatively low, especially if mutation carriers without clinical disease are included. As a result, trials will require treating large cohorts followed over long periods of time to demonstrate treatment benefit as reflected by hard outcomes, such as mortality or development of clinically overt disease. Achieving the scope and scale of such studies will require monumental effort given the relative rarity of sarcomeric cardiomyopathies.
Owing to these challenges, traditional approaches to clinical trial design may not be appropriate or successful. To allow initial test-of-concept trials to proceed over manageable time frames and with feasible cohort sizes, we will need to identify surrogate endpoints that accurately reflect disease progression and treatment benefit, and that are more likely to show dynamic change over shorter periods of time. Identifying additional robust, quantitative early phenotypes that reflect a continuum of disease progression from normal controls to patients with overt disease will be important to advance these efforts. Such phenotypes could function as surrogate endpoints in trials. Effective treatment would potentially show that early-stage mutation carriers become more 'normal' for the metrics under study.
Next, choosing which individuals to target and how long to treat them requires careful consideration. Although it is appealing to consider directing therapy towards the youngest mutation carriers without any phenotypic expression, such trials will be exceedingly difficult. At this stage, disease progression is likely to be extremely subtle and play out over many years. As a result, events are unlikely to accrue in the placebo-treated cohort, challenging the ability to demonstrate treatment benefit in the cohort receiving intervention during trials taking place over ,10 years. It may be best to focus initial trials of disease modification on mutation carriers with mild phenotypic expression ( Figure 1) and refine patient selection as more knowledge is obtained regarding both early phenotypes and pathways governing disease evolution.
Finally, more innovative and sophisticated statistical analysis plans will also be needed to be guard against false-negative trial results, without excessively increasing the risk for false-positive results. 48 Experience gained from such initial trials of disease-modifying therapies will set the stage for more definitive trials to follow. Ideally, we will be able to determine accurate predictors of both adverse outcomes and imminent development of clinically overt disease. Therapy could then be targeted to those at highest risk who would derive greatest benefit from potentially lifelong treatment. Potential novel ways to address the challenges of trials to modify and prevent HCM are explored in the ongoing VANISH trial (Valsartan for Attenuating Disease Evolution in Early Sarcomeric HCM; NCT01912534).

Bench and preclinical studies
Apart from clinical studies, more advanced basic methodologies are crucial to establish the mechanistic links between mutation and the pathogenic molecular and cellular mechanisms underlying the cardiomyopathy phenotype. As a first step in understanding the pathophysiology of sarcomeric cardiomyopathies, it is essential to assess the functional effects of the mutant sarcomeric proteins per se. Until now, substantial data originated from experiments with engineered mouse models, gene transfer in cardiomyocytes or engineered heart tissue, or human recombinant proteins (reviewed in this issue). 49 -54 Based on these in vitro and transgenic rodent studies, it has been proposed that HCM gene mutations cause enhanced contractile function (hypercontractility), while DCM-associated mutations reduce function (hypocontractility). However, based on the complex functional and structural properties of the affected sarcomeric proteins, it is unlikely that mutations in different sarcomeric proteins would cause cardiomyopathy via a single disease mechanism. Myocardial dysfunction will depend on mutation expression, mutation location within the gene, age-related protein isoform background and secondary disease-related protein modifications. Below we highlight the mutation-related changes in cardiomyopathies which warrant future preclinical research (Figure 1).

Mutation expression and gene dosage
Sarcomeric cardiomyopathies are mainly inherited in an autosomaldominant fashion, with the exception of some X-linked mutations, indicating that the presence of the mutation on one of the two alleles is sufficient to cause the disease phenotype. Therefore, expression of both wild-type and mutant alleles is basically expected to result in the incorporation of both wild-type and mutant proteins in the sarcomere. However, it is important to emphasize that the expression of the mutation is regulated at different levels by quality control mechanisms (such as the non-sense-mediated mRNA decay, ubiquitin-proteasome system, and/or the autophagy-lysosomal pathway) 17,55,56 that could reduce or increase production of mutant proteins in comparison to wild-type. This is particularly the case for truncating MYBPC3 mutations, which result in reduced mutant mRNAs and/or proteins, leading to haploinsufficiency of the proteins in the sarcomere. 57 -59 A large variability in the level of wild-type cardiac myosin-binding protein-C (cMyBP-C) protein was found in tissue from HCM patients, 59 -61 which may correlate with the heterogeneity of the disease phenotype severity. In addition, heterozygous mutations could be subjected to allelic imbalance, which could affect disease severity. For example, higher levels of mutant mRNA appear to contribute to a more severe cardiac phenotype. 62,63 Infants carrying homozygous or compound heterozygous truncating MYBPC3 mutations are expected to have low levels or complete absence of mutant cMyBP-C in the heart. As a consequence, these infants present with cardiomyopathy in the neonatal period and rapidly develop heart failure and die within the first year of life. 6,8 -12 Similarly, many transgenic rodent models exhibit gene dosage dependence on the phenotype severity. For example, homozygous Mybpc3targeted knock-in mice developed an earlier and more severe disease phenotype than their heterozygous littermates. 35,64,65 Similarly, a transgene dose-dependent increase in Ca 2+ sensitivity of force development and concomitant decrease in relaxation parameters have been reported in transgenic mice expressing the HCM-associated E180G Tpm1 mutation. 66 In addition, mice expressing high levels of the HCM-associated cTnT D160E Tnnt2 mutation showed a further decrease in relaxation capacity and exacerbated disturbances in Ca 2+ transients compared with mice with low mutant cTnT D160E levels. 67 These data are supported by recent molecular-based therapeutic approaches. Allele-specific RNA interference in heterozygous Myh6 mutant R403Q HCM mice showed that a reduction of 25% of the mutant allele expression was enough to prevent HCM development. 68 Adeno-associated viral-mediated Mybpc3 gene transfer in Mybpc3-targeted knock-in mice showed replacement of the endogenous mutant by wild-type cMyBP-C protein and prevention of LVH and cardiac dysfunction in a dose-dependent manner. 69 Thus, a body of evidence from the clinic and transgenic mouse models indicate that mutation dose contributes to the onset and severity of cardiomyopathy. However, little information is available on the dose needed to perturb sarcomere function and structure of the heart muscle as many in vitro studies focused on functional consequences of mutant protein in the absence of wild-type protein. A recent study based on troponin exchange experiments revealed that ,50% of poison peptide is sufficient to perturb sarcomere function in human cardiomyocytes obtained from patients with a homozygous TNNT2 mutation or a heterozygous TNNI3 mutation. 70

Mutation location
Apart from mutant protein dose, mutant protein conformation, which depends on location of the mutation in the gene, represents a likely pathomechanism underlying variable onset and phenotypic expression of cardiomyopathies. Maass et al. 71 compared two HCM-associated Tnnt2 transgenic mouse models that express a missense or a truncating mutation, i.e. R92Q and cTnT trunc , and found that remodelling was mutation dependent. Fibrosis was only seen in R92Q mice, which also showed a hypertrophic response to prolonged adrenergic stimulation, while cTnT trunc mice did not. Likewise, S532P and F764L mutations in a-myosin heavy chain (Myh6) are both known to cause DCM, but their effect on hypertrophy and contractile function differs considerably. 72 While hypertrophy was noted in F764L mice, the heart weights of S532P mice did not differ from non-transgenic mice. Moreover, cardiomyocytes from F764L mice produced significantly less force than those from S532P hearts. 72 Recent data suggest mutation-specific perturbations in sarcomere function in human HCM samples. 40,73 -75 In particular, mutations in the head domain of myosin heavy chain (MyHC) reduce the contractile strength of sarcomeres. 73,74 Moreover, a genespecific increase in energetic cost of contraction was found both in vivo in HCM patients and in vitro in human HCM samples, being more severe in MYH7 than in MYBPC3 mutation carriers. 40 The increase in tension cost in cardiac muscle strips of manifest HCM patients was explained primarily by a reduction in maximal force-generating capacity. Likewise, at the pre-hypertrophic phase of HCM, the decrease in myocardial efficiency was largely explained by a reduction in cardiac work rather than an increase in oxygen consumption. In contrast to the abovementioned assumption that HCM mutations enhance contractile function, these recent studies in human HCM indicate that hypocontractile sarcomeres may represent a primary abnormality, in particular in cardiomyopathy caused by certain MYH7 mutations. The apparent difference in functional effects may be explained by the different HCM mutations investigated. The recent studies emphasize the need to establish the toxic effects of mutants on structure and function of the heart muscle with respect to mutation location and protein dose.

Protein isoform background
Effects of mutant proteins may be more or less severe depending on protein isoform background in the heart. The relevance of protein background for mutant-related effects has been demonstrated by the species-dependent MyHC isoform composition present in the heart. As indicated in the present issue, 54 many studies in transgenic rodents have been performed on the Myh6 background, while human ventricles predominantly express the slow b-MyHC (MYH7). Functional and structural changes were different in mice expressing the first identified R403Q mutation in a Myh6 or Myh7 background. 76,77 In addition, it has also been shown that a genetically engineered ato b-MyHC shift mice in the contest of TNNT2 mutations could partially rescue observed energetic effects, 78,79 further evidence that sarcomeric isoform composition (known to be altered in the context of pathogenic remodelling in humans) may play a greater role in modulating ventricular remodelling in sarcomeric cardiomyopathies. A recent study directly demonstrated the effects of switching troponin isoforms in an established transgenic mouse model of Tpm1-associated DCM. 80 These results suggest that age-dependent isoform switching of sarcomeric proteins may directly modulate ventricular structure and function in a transient fashion.

Transient changes in post-translational protein modifications
Apart from the impact of isoform composition on disease outcome, the cardiomyopathic phenotype is not static and changes during the course of disease development as a result of secondary post-translational disease-modifying processes. An important post-translational modification (PTM) that changes during disease development is protein phosphorylation due to altered expression and activity changes of the myocardial kinases and phosphatases. An optimal balance between kinase and phosphatase activities is required to regulate cellular processes via targeted phosphorylation of proteins involved in cell signalling and contractility. During disease onset and progression, changes occur in protein phosphorylation due to an imbalance between kinase and phosphatase activities, which subsequently underlie structural and functional changes of the heart muscle. One of the most described perturbations is the down-regulation and desensitization of the b-adrenergic receptors in heart failure. At the level of the sarcomeres, reduced b-adrenergic-mediated phosphorylation by protein kinase A (PKA) is associated with reduced phosphorylation of myofilament proteins such as cardiac troponin I and cMyBP-C. A reduction in myofilament protein phosphorylation has been associated with an increased myofilament Ca 2+ sensitivity. Increased Ca 2+ sensitivity compared with healthy sarcomeres has been proposed as a hallmark of HCM, opposite to reduced myofilament Ca 2+ sensitivity in DCM. Recent studies in human HCM samples with mutations in thick filament proteins, MyHC and cMyBP-C, indicate that high myofilament Ca 2+ sensitivity is mostly due to reduced myofilament protein phosphorylation compared with non-failing myocardium rather than the mutation itself. 70 The secondary disease-related high myofilament Ca 2+ sensitivity may be a target of treatment in patients with manifest HCM, but most likely is not the initial trigger of cardiomyopathy onset. Likewise, PTMs induced by increased oxidative stress 81 may contribute to cardiac dysfunction and disease progression.

Myofilament Ca 21 sensitivity, disease pathogenesis, and arrhythmias
Myofilament Ca 2+ sensitivity is universally increased in sarcomeric cardiomyopathies associated with cardiac hypertrophy. As discussed above, the Ca 2+ sensitization is often due to altered phosphorylation patterns of sarcomeres, but can also be a primary consequence as shown for Tnnt2 mutations. Upon myofilament Ca 2+ sensitization, altered intracellular Ca 2+ homeostasis and slowed muscle relaxation are expected, which will contribute to diastolic dysfunction regardless of the underlying molecular mechanism. 82 Although the molecular mechanisms of how increased Ca 2+ sensitivity can cause cardiac hypertrophy and contributes to HCM pathogenesis remains unclear, proof-of-concept studies showed that hypertrophy was prevented and relaxation enhanced when myofilaments were genetically desensitized in mice expressing the HCM-associated Tm180 Tpm1 mutation. 83 Even in the absence of cardiac hypertrophy, increased myofilament Ca 2+ sensitivity generated acutely by drugs or chronically by Tnnt2 mutations has been shown to cause susceptibility to ventricular arrhythmias in mice. 84 Two independent arrhythmia mechanisms have been proposed: Ca 2+ -sensitizing Tnnt2 mutations were shown to increase the cytosolic Ca 2+ -binding affinity (Ca 2+ buffering), Ca 2+ -dependent action potential remodelling and promote pause-dependent Ca 2+triggered arrhythmias. 85,86 Myofilament Ca 2+ sensitization was also linked to focal energy deprivation and re-entry arrhythmias during stress, but the underlying cause is unclear and is currently investigated. 87 Although remains to be tested if primary or acquired myofilament Ca 2+sensitization differ, normalizing myofilament Ca 2+ sensitivity may be beneficial on several levels: (i) normalizing it early may prevent disease progression and (ii) if that opportunity has been missed, it may still be effective to prevent lethal ventricular arrhythmias. Whether Ca 2+desensitizing drugs will be useful therapeutically needs to be explored. In addition, studies are warranted to investigate the underlying mechanisms in large animal models and humans to validate the proposed mechanisms and identify therapeutic targets. 88 On the other hand, thin filament mutations that reduce myofilament Ca 2+ sensitivity have been associated with a primary DCM phenotype in humans, 88,90 which has been reproduced in mouse models. 90 -92 Normalizing reduced Ca 2+ sensitivity either using transgenic approaches or with the Ca 2+ sensitizer levosimendan was effective in preventing the development of DCM in mice expressing a human TNNT2 mutation. Hence, future basic and clinical studies are warranted to better identify how and to what extent abnormal myofilament Ca 2+ sensitivity contributes to disease pathogenesis and arrhythmia risk (i.e. using iPSC models 53 and large animal models 54 ), and whether it can be used to risk stratify mutation carriers (e.g. to test whether human carriers of Ca 2+ -sensitizing mutations exhibit an increased risk for pause-triggered ectopic ventricular beats that are predicted by studies in mice 84 ). If confirmed by such studies, targeting altered myofilament Ca 2+ sensitivity, although possibly only a downstream consequence of the disease mutation, would nevertheless be an attractive, mutation-independent treatment strategy in sarcomeric cardiomyopathy.

Future
The critical gap in our knowledge is insight in the way a mutation becomes toxic. The large clinical variability implies that additional factors eventually determine whether a mutation becomes effective and injures the heart. Disease severity will depend on expression of the mutant allele, disease-modifying genes, cardiac load, and additional non-genetic disease modifiers including epigenetic modifications, miRNA, PTMs of proteins, an age-related decline in protein quality control, and environmental disease triggers. This complex interaction makes it difficult to determine the main pathophysiological mechanism.
Previous paradigms need to be tested properly in disease model systems, which take into account expression levels of mutant proteins, disease progression, and, as indicated in the paper by Duncker and colleagues, 54 the protein isoform content as present in the human heart. Observations from rodent studies need to be tested properly in a human-like model systems (iPSC-derived disease modelling in cardiac myocytes and engineered heart tissue) as described by Eschenhagen et al. 53 to test applicability for the human situation. Ultimately, identification of disease modifiers that underlie the complex pathophysiology of sarcomeric cardiomyopathy may serve as rational therapeutic targets.