Introduction
In hypertrophic cardiomyopathy (HCM), energetic deficits precede the manifestation of septal hypertrophy. In previous work, we revealed that increased energetic demand at the myofilaments in HCM mouse models oxidizes mitochondrial NAD(P)H. Since its regeneration is not sufficiently matched by mitochondrial Ca²⁺ uptake, this causes elevated emission of reactive oxygen species (ROS) and arrhythmias. Here, we investigated whether mitochondrial ROS further compromises energetic regeneration through the creatine kinase (CK) system, both at mitochondrial (Mt-CK) and myofilamental (M-CK) sites.

Methods & Results
Human HCM patient tissue (n=22-30) was compared to non-failing donor controls (n=8-14) using western blots and colorimetric CK activity assay, where M-CK protein was found oxidized and with decreased expression (58±3% of control), but also hypoactive (47±5% of control). Oxidation was confirmed via OxICAT labeling mass spectrometry, which identified two oxidized cysteine sites (Cys146 and Cys254), essential for kinase activity function. Mouse cardiomyocytes (at 5 Hz and isoproterenol) treated with the Ca²⁺ sensitizer EMD reproduced oxidation of mitochondrial NAD(P)H, elevated H₂O₂ emission, and deactivated M-CK activity (by 28.1±2.1%) compared to vehicle. All these effects were prevented when repeated in mice with mitochondrial overexpression of catalase. Protein oxidation of Mt-CK via mass spec identified Cys317 in human HCM samples oxidized, an essential site for enzyme activity and protein assembly. Unlike dimeric M-CK, Mt-CK is an octamere and speculated to hold additional structural function within the mitochondrial intermembrane space. Using the CK specific inhibitor DNFB, we noted that Mt-CK inhibition does not only affect its enzymatic activity but also the stability of its superstructure. Mitochondrial O₂ respirometry with simultaneous H₂O₂ recording (Amplex UltraRed) was performed with pyruvate/malate, saturating ADP levels (1mM) and DNFB (1 and 20µM). In the presence of DNFB, O₂ consumption was not compromised and native PAGE protein analysis revealed no alterations to the composition of the respirasome. Most important, DNFB stimulated a large H₂O₂ generation in cardiac mitochondria at 1 µM concentration and even higher at 20 µM. To validate upstream formation of superoxide, we measured the effects of DNFB using electroparamagnetic resonance and observed substantial increases of superoxide (~37%). To rule out unspecific effects of DNFB, cardiac mitochondria were compared to liver mitochondria, which have very low expression of Mt-CK. H₂O₂ and superoxide formation was almost negligible in liver mitochondria with DNFB.

Conclusion
Hypercontractility of myofilaments induces a Ca²⁺ mismatch in mitochondria, causing mitochondrial oxidation of NAD(P)H and ROS emission, which in turn oxidizes and affects the activity of both M-CK and Mt-CK. Furthermore, disruption of Mt-CK leads to additional ROS via electron leakage. Together, these results suggest a futile feed-forward mechanism of energetic and redox mismatch in HCM that likely aggravates diastolic dysfunction and – according to previous results of our group – account for arrhythmias through providing a trigger (redox) and substrate (energy depletion) for arrhythmias.