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

Dr Hwang’s recent letter concerning our paper raises some very interesting points about the molecular mechanism of familial cardiomyopathy (HCM and DCM) due to mutations in the proteins of the thin filament and the key role of troponin.

In principle such mutations could cause defects in force generation or force transmission but so far all the mutations in thin filament proteins that have been investigated in detail indicate abnormal force production linked to abnormal Ca²⁺-regulation. Troponin, in concert with tropomyosin and actin comprises the Ca²⁺-dependent switch of muscle and the properties of this switch are modulated by PKA phosphorylating TnI, therefore troponin abnormalities are likely to play a central role in cardiomyopathies. The case has been made that HCM-causing mutations increase myofilament Ca²⁺- sensitivity and this is sufficient to trigger the symptoms of HCM. On the other hand DCM-causing mutations uncouple the normal relationship between Ca²⁺-sensitivity and PKA-dependent phosphorylation of troponin I (TnI) and this has been shown to blunt responses to adrenergic stimulation and predispose the heart to systolic dysfunction under stress ^{1, 2}.

Our recently published work highlights the importance of troponin still more, since we have demonstrated that small molecule Ca²⁺-sensitisers (EMD57033 and Bepridil) can mimic the uncoupling effect of mutations, whilst the Ca²⁺- desensitiser epigallocatechin-3-gallate (EGCG) can restore the modulation of Ca²⁺-sensitivity by TnI phosphorylation to thin filaments containing DCM or HCM mutations that uncouple this relationship ³. All of these reagents act by binding to troponin; thus understanding how Ca²⁺ and phosphorylation control contractility through troponin is the key to determining how thin filament cardiomyopathies are initiated

There are severe practical problems in linking troponin structure with function because the most important parts of the complex are intrinsically disordered and cannot be defined in terms of a static structure. These include the cardiac-specific N-terminus of cTnI (amino acids 1-30) that contains the PKA phosphorylation sites (Serines 22 and 23), the so-called switch peptide of cTnI (148-164), that plays a key role in the Ca²⁺-dependent switching on and off of the thin filament, and the C-terminus of troponin T (cTnT₃ 280-298) whose function is not known although it is likely to be located close to the regulatory region of troponin and may interact with the other two peptides. Protein nuclear magnetic resonance has been used to determinethe structure of the missing peptides and their binary complexes with troponin C (TnC) in several states ^4–9. When TnI is unphosphorylated there is a weak ionic bond between the N terminal peptide and the regulatory Ca²⁺-binding EF hand of TnC. When Ser 22 and 23 are phosphorylated the binding is further weakened. We have proposed that the unphosphorylated state can also be disrupted by mutations or other alterations in any component of the thin filament or by small molecules resulting in uncoupling. However, whilst it can account for all our experimental observations, we do not believe that this hypothesis can be tested until the structure and dynamics of the whole troponin structure is known. This question is now being addressed by molecular dynamics simulations, but many practical and technical problems need to be solved before a believable structural and dynamic description of the regulatory elements of troponin becomes possible.

In his letter to the editors Dr Hwang presents a model based on his extensive experience of investigating troponin by protein nuclear magnetic resonance protocols ⁶ that is proposed to be able to explain the structural basis of the results we obtained. In this model an alternative mechanism of Ca²⁺- dependent activation is proposed in addition to the current consensus mechanism that accounts for Ca²⁺-switching of muscle and its modulation by cTnI phosphorylation ( see for instance Solaro et al. 2008¹⁰). It is proposed that the helical segment of cTnI (39-60), downstream of the phosphorylatable peptide, that normally interacts strongly with the C-terminal domain of TnC can becomes dissociated, thus allowing a different interaction between the two domains of TnC that would have a higher Ca²⁺-sensitivity and not involve cTnI 1-30, therefore it would be uncoupled. The rationale for proposing this is the observation that we measure uncoupling and the effects of EMD57033 and EGCG in the in vitro motility assay, where the Ca²⁺-sensitivity of filament activation is very high (EC₅₀ 0.1-0.3 μM) whereas the Ca²⁺-sensitivity of contractility in skinned cardiac muscle in the phosphorylated and unphosphorylated states is about ten times lower.

However, there seems no justification for a two-mechanism model. As described in our papers, Ca²⁺-regulation of thin filaments occurs normally when measured by the in vitro motility assay and all the effects of phosphorylation, mutants, Ca²⁺ sensitisers and desensitisers and, especially uncoupling and re-coupling, on Ca²⁺-sensitivity were consistently observed using the IVMA technique and, in the case of the small molecules, effects were shown to be fully reversible. In these studies Ca²⁺-sensitivity ranged between 0.04 and 0.2 μM so there is no reason to invoke two mechanisms based on different Ca²⁺-sensitivity ranges. In fact it would be very hard indeed to explain how drugs such as EGCG could reverse uncoupling if they were acting through a mechanism separate from the one that modulates Ca²⁺- sensitivity by phosphorylation or mutations. The issue of absolute Ca²⁺- sensitivity, and even the experimental system studied is irrelevant since we have clearly shown that exactly the same modulation of Ca²⁺-sensitivity of contraction by phosphorylation, drugs and mutations and the recoupling due to EGCG can be demonstrated in myofibrils, where the Ca²⁺- sensitivity is in the 0.4-2 μM range ^{3, 11}

What we undoubtedly agree upon is the central role of troponin abnormalities in cardiomyopathy and the need to know much more about the structure and dynamics of the intrinsically disordered regulatory segments of the molecule. Not only is this necessary to provide a molecular mechanism for the physiological effects of TnI phosphorylation and mutations in thin filaments that cause cardiomyopathy, it is also essential in order to understand how small molecules can interact with this system. ¹⁴ This is clinically relevant since Ca²⁺-sensitisers acting on TnC have been investigated as inotropic drugs and our discovery of the ability of EGCG and similar compounds to reverse the uncoupling effect of DCM-causing mutations (and some HCM-causing mutations) points to potential therapies for familial cardiomyopathies.