‘(De-)sensitization’ vs. ‘Uncoupling’: what drives cardiomyopathies in the thin filament?

Roughly half of all familial hypertrophic cardiomyopathy (HCM) cases can be linked to a sarcomere protein mutation, raising the possibility that the causative molecular pathways can be understood and corrected through pharmacotherapy. Nowhere is this more true than in the thin filament, for which all major components have been recombinantly expressed and extensively characterized by biophysical methods. It is generally accepted that for thin filament proteins, HCM-associated mutations increase the calcium sensitivity of the cardiac sarcomere. The prevailing theory is that increased calcium sensitivity causes hyper-activation of cardiac muscle, predisposing it to diastolic dysfunction and abnormal hypertrophy. Following similar reasoning, it has been postulated that familial dilated cardiomyopathy (DCM) mutations in the thin filament lead to decreased calcium sensitivity. However, decreased sensitivity has not been consistently demonstrated, causing some to seek alternative explanations. One interesting observation that has emerged in some studies is that thin filaments carrying DCM mutations are resistant to the desensitizing effect of protein kinase A (PKA) phosphorylation.1 In normal healthy individuals, sympathetic β-adrenergic stimulation drives PKA phosphorylation of cardiac troponin I Ser22/232,3 (among many other targets), and this speeds up sarcomeric calcium release and cardiac muscle relaxation. Messer and Marston have suggested that the lack of response to troponin I Ser22/23 phosphorylation, termed ‘uncoupling’, is a hallmark of familial DCM that also contributes to its pathogenesis. Could uncoupling be a feature central to HCM as well? Using in vitro motility assay (IVMA) to study mutations involving actin, tropomyosin, and the troponin complex, Papadaki et al.4 demonstrate a striking lack of response to troponin I phosphorylation in every mutant examined (5 DCM, 3 HCM). They further show that uncoupling can be reproduced in wild-type thin filaments through the action of calcium sensitizer compounds, EMD 57 033 and bepridil. Finally, and most remarkably, they demonstrate that the calcium-desensitizing compound, EGCG, can reverse the uncoupling defect in every mutant. These findings could have major implications for the treatment of cardiomyopathies. The key to understanding both calcium sensitivity and uncoupling is cardiac troponin C (cTnC), the calcium sensor of the thin filament. cTnC consists of two globular domains connected by a flexible linker. The structural C-terminal domain (cCTnC) contains two high affinity (Kd 3–40 nM) Ca2+ binding sites that are always occupied by Ca2+ or Mg2+,5 stabilizing an open conformation that remains anchored to the rest of the troponin complex throughout the cardiac cycle.6 In contrast, the regulatory N-terminal domain (cNTnC) has only a single low affinity site. Ca2+ binding during systole partially stabilizes an open conformation that is fully stabilized by the switch region of cardiac troponin I (cTnI148–158).7 cNTnC-cTnI148–158 binding sequesters adjacent inhibitory segments (cTnI135–147 and cTnI160–209) from actin, stabilizing troponin–tropomyosin to a position that promotes actin–myosin interaction and thus muscle contraction. The calcium sensitivity of cardiac muscle reflects not only the calcium binding affinity of cNTnC, but also the balance between contraction and relaxation for the entire sarcomere. Thin filament activation releases cTnI148–158 switch peptide from actin and makes it more readily available for binding cNTnC, shifting the conformational equilibrium of cNTnC from the closed form (Ca2+Kd ∼ 5 µM) towards the open form (Ca2+Kd < 0.7 µM).8 Thus, calcium sensitivity depends on the balance between the closed state and the cTnI148–158-bound open state, which can be promoted by downstream factors that favour contraction (e.g. strong actin–myosin interaction), small molecule troponin activators (that act as calcium sensitizers),9 and optimal orientation of the cNTnC domain through electrostatic interaction with cTnI19–37.10 Thus, the interaction between cNTnC and cTnI19–37, which is attenuated by phosphorylation of cTnI Ser22/23, is only one of several mechanisms that impact calcium sensitivity. Mutations that specifically interfere with this interaction or with phosphorylation itself (like cTnI R20C) would be expected to give rise to uncoupling. However, it is uncertain why the loss of this particular mechanism should be a feature universal to HCM or DCM. I propose that the pervasive uncoupling observed by Papadaki et al. results from an alternative mechanism of calcium-dependent thin filament activation (Figure 1). The phosphorylation-sensitive cNTnC-cTnI19–37 interaction requires the adjacent cTnI39–60 segment to be anchored to the cCTnC domain. Displacement of cTnI39–60 from cCTnC would have a three-fold effect: (i) uncoupling; (ii) cCTnC becomes able to bind the cTnI inhibitory region, cTnI135–147,11 allowing it to activate the thin filament at low calcium concentrations (the measured range of calcium affinities in the current study is Kd 46–260 nM, too tight for even maximally activated cNTnC); and (iii) binding sites for bepridil12 and EMD 5703313 become exposed on cCTnC. In fact, EMD 57 033 will not bind to troponin C unless cTnI39–60 is displaced, whereas EGCG is able to bind in the presence of cTnI39–60.14 Two alternative mechanisms for calcium-dependent thin filament activation. In the conventional mechanism (on the left side), cTnC is anchored to the rest of the troponin complex via the troponin I-T arm. Binding of the open state of cNTnC to the switch region of troponin I, cTnI148–158, activates the thin filament. The calcium sensitivity of the sarcomere is determined by the balance between cNTnC open and closed states. The equilibrium can be shifted towards the activated state by HCM-causing mutations, dephosphorylation of cTnI Ser22/23, or binding of calcium sensitizers like DFBP-o to cNTnC. Conversely, it can be shifted towards the inactivated state by DCM-causing mutations, phosphorylation of cTnI Ser22/23, or calcium desensitizers like W7. In the alternative activation mechanism (on the right side of the figure), cTnC is dissociated from the I-T arm, allowing cCTnC to bind directly to the inhibitory region, cTnI135–147, to activate the thin filament. (In contrast to cNTnC, cCTnC is always in an open conformation when bound to calcium. Binding of cTnI135–147 to cCTnC would be expected to enhance its calcium affinity, though a Kd value has not been published.) The compounds used by Papadaki et al., EGCG, bepridil, and EMD 57033, all bind to cCTnC in the dissociated state. The alternative cCTnC-cTnI135–147 interaction is weaker and less effective at activating the thin filament than the canonical cNTnC-cTnI148–158 interaction.15 However, its relative contribution to calcium sensitivity could be increased by (i) HCM mutations that destabilize the inactive state of the thin filament, favouring the release of cTnI135–209 from actin, (ii) DCM mutations that destabilize the activated state to make the cNTnC-cTnI148–158 interaction less probable, or (iii) compounds that promote the cCTnC-cTnI135–147 interaction. This model provides a possible explanation for why a decrease in calcium sensitivity can be difficult to observe for DCM-associated mutations: it is obscured by the overriding effect of cCTnC-cTnI135–147-based thin filament activation. It is unknown whether free cTnC exists in vivo, but its presence may be enhanced in reconstituted systems like IVMA, perhaps due to stoichiometric excess or incomplete incorporation into troponin complex. Studies are needed to assess for its presence and influence in a wide range of experimental setups. Purging dissociated cTnC may lead to measurements of calcium sensitivity that are more in keeping with expectations. On the other hand, uncoupling may end up being a more sensitive marker for the destabilizing effects of cardiomyopathy-causing mutations, even if it does not occur in vivo. Further studies using native myofilaments are needed to clarify this point.16