Nitro-fatty acids suppress ischemic ventricular arrhythmias by preserving calcium homeostasis

Nitro-fatty acids are electrophilic anti-inflammatory mediators which are generated during myocardial ischemic injury. Whether these species exert anti-arrhythmic effects in the acute phase of myocardial ischemia has not been investigated so far. Herein, we demonstrate that pretreatment of mice with 9- and 10-nitro-octadec-9-enoic acid (nitro-oleic acid, NO2-OA) significantly reduced the susceptibility to develop acute ventricular tachycardia (VT). Accordingly, epicardial mapping revealed a markedly enhanced homogeneity in ventricular conduction. NO2-OA treatment of isolated cardiomyocytes lowered the number of spontaneous contractions upon adrenergic isoproterenol stimulation and nearly abolished ryanodine receptor type 2 (RyR2)-dependent sarcoplasmic Ca2+ leak. NO2-OA also significantly reduced RyR2-phosphorylation by inhibition of increased CaMKII activity. Thus, NO2-OA might be a novel pharmacological option for the prevention of VT development.


Scientific RepoRtS
| (2020) 10:15319 | https://doi.org/10.1038/s41598-020-71870-6 www.nature.com/scientificreports/ of the membrane potential is enhanced thereby causing inactivation of Na + channels with further slowing of ventricular conduction 7 . As the identification of AMI patients at risk for VT remains difficult, a therapeutic strategy would have to be applied to a broad patient population. Therefore, dietary intervention with endogenously produced antiarrhythmic modulators appears as an attractive therapeutic approach. This concept has been recently supported by the randomized double blind trial REDUCE-IT which demonstrates that prophylactic treatment with icosapentaenoic acid-ethyl (VASCEPA) significantly reduces cardiovascular events, the rate of cardiac arrests, and the number of sudden cardiac deaths (SCD) 8 .
Nitro-oleic acid (9-and 10-nitro-octadec-9-cis-enoic acid; NO 2 -OA) is an endogenously generated unsaturated fatty acid nitroalkene derivative that is formed by oleic acid reaction with nitric oxide and nitrite-derived nitrogen dioxide (•NO 2 ) 9,10 . NO 2 -OA is conferred with an electrophilic β-carbon that rapidly and reversibly reacts with nucleophilic cysteine-, and to a lesser extent histidine, residues via a Michael addition 11 . The covalent nitro adduct is essential for the metabolic effects of NO 2 -OA since native oleic acid is not electrophilic and ineffective in various studies [12][13][14] . Nitro-fatty acids have been identified as potent endogenous anti-inflammatory mediators that affect a number of signaling mediators, transcriptional regulatory proteins and ion channels, including nuclear factor 'kappa-light-chain-enhancer' of activated B-cells (NF-κB), kelch-like ECH-associated protein 1 (Keap1/Nrf2), mitogen-activated protein kinases (MAPK), peroxisome proliferator-activated receptor gamma (PPARγ) and transient receptor potential channels (TRP) channels [15][16][17] . Preclinical studies have demonstrated therapeutic benefit in a variety of disease models such as pulmonary hypertension, nephritis and, of relevance to herein, reperfusion damage after myocardial ischemia (I/R) and a subsequent attenuation of cardiac fibrosis, reduced infarct size and preserved left ventricular function [18][19][20] . Since NO 2 -OA has passed safety clinical phase I trials 21 these results have been translated into two, currently recruiting, clinical phase II trials targeting pulmonary hypertension and primary focal segmental glomerular sclerosis 22,23 .
Very low levels of NO 2 -OA could be detected in human serum-and urine 24 as well as in the myocardium of rodents 19 . Nonetheless, inflammatory stimuli like I/R induce substantial generation of NO 2 -OA up to µM concentrations by enhanced formation of nitrating species 19 . Furthermore, increased endogenous formation of nitro-linoleic acid has been demonstrated in healthy humans after dietary supplementation with conjugated linoleic acid and nitrite or nitrate 25 .
If anti-arrhythmic effects of nitro-fatty acids could indeed be shown, a dietary supplementation or a pure pharmaceutical-based nitro-fatty acid therapeutic strategy in patients at risk for cardiovascular events could evolve as a strategy for prevention of sudden cardiac death.

Results
Susceptibility to VT upon right ventricular stimulation. Susceptibility to VT induction 20 min after ligation of the left anterior descending artery (LAD) was drastically reduced in animals pretreated with NO 2 -OA (representative ECG recordings shown in Fig. 1A). Thus, the probability of VT induction, calculated as the percentage of induced VT episodes per stimulation maneuvers (Fig. 1B), as well as the total number of inducible VT episodes (Fig. 1C) were both significantly lower in NO 2 -OA pretreated animals. Additionally, the total time of VT episodes was reduced after NO 2 -OA-pretreament (Fig. 1D) indicating that VT episodes in NO 2 -OA pretreated ischemic hearts were rapidly self-terminating.

Homogeneity of conduction.
Epicardial mapping studies within the peri-ischemic region (positioning of the multi-electrode array (MEA) is shown in Fig. 2A) clearly revealed a more homogeneous conduction pattern in NO 2 -OA pretreated-compared to vehicle treated animals (representative conduction maps are shown in Fig. 2B). Of interest, only weak and therefore unevaluable electrical signals were detectable within the core ischemic region (data not shown). The mean electrical latency, measured apico-septally next to the ischemic core region 26 , was significantly elevated after ischemia induction, indicating a reduction of electrical conduction velocity. Although latency was numerically lower in NO 2 -OA pretreated hearts as compared to controls, it failed to reach statistical significance (Fig. 2C). However, the index of inhomogeneity was significantly lower upon administration of NO 2 -OA, indicating a reduction in ventricular conduction blocks (Fig. 2D) 27 .
To investigate potential side effects of NO 2 -OA on action potential (AP) development, we performed patchclamp analysis of isolated adult cardiomyocytes under baseline conditions. The treatment with NO 2 -OA revealed no differences in the AP morphology regarding resting potential (Supplemental Fig. 1A ca 2+ transient analyses under isoproterenol treatment. Given that catecholamine induced disturbances in Ca 2+ homeostasis are a major contributor to arrhythmia development in acute ischemia 3 , we investigated cytosolic Ca 2+ transients and arrhythmic events in isolated adult cardiomyocytes by ratiometric fluorescence imaging (IonOptix Corp) 28 . The number of arrhythmic events which appeared aside from a basic 1 Hz pacing stimulus (e.g. caused by early after depolarizations (EAD) and delayed after depolarizations (DAD)), was increased after beta-adrenergic stimulation with isoproterenol (Iso) 3 . Crucially, this increase was prevented by NO 2 -OA treatment (Fig. 3A). Further investigation of single cellular calcium transients revealed that the time to Ca 2+ peak concentration upon Iso challenge (Fig. 3B) was reduced after NO 2 -OA treatment indicating a faster cytosolic Ca 2+ influx. Furthermore, an increased peak height of the Ca 2+ transient (Fig. 3C) was noted in both Iso treated groups as compared to vehicle treatment indicating elevated cytosolic Ca 2+ levels. Accordingly, adrenergic stimulation led to a faster cytosolic Ca 2+ clearance, as indicated by the reduced time to reach 50% or 90% of the maximum cellular Ca 2+ concentration relative to baseline levels (Fig. 3D www.nature.com/scientificreports/ NO 2 -OA treatment of isolated adult cardiomyocytes under baseline conditions furthermore did not influence the characteristics of Ca 2+ transients and intracellular Ca 2+ levels (Supplementary Figure S2).
Given the reduced appearance of spontaneous arrhythmic events upon NO 2 -OA treatment after Iso stimulation, we next investigated RyR2-dependent sarcoplasmic calcium leak as a potential pro-arrhythmic molecular mechanism. ca 2+ leak analyses under isoproterenol treatment. Analyses of Ca 2+ levels in isolated cardiomyocytes showed spontaneous Ca 2+ peaks after RyR2 unblocking upon Iso stimulation (between the red and black cross in Fig. 4A), whereas almost no arrhythmic events could be observed upon additional NO 2 -OA treatment (Fig. 4A, B). Iso treatment resulted in an increased SR-dependent Ca 2+ leak which was estimated as the difference between the Fura-2 ratio recorded at the end of the 0 Na + /0 Ca 2+ Tyrode perfusion (black cross) and at the end of the RyR2 inhibitor tetracaine treatment (red cross) (Fig. 4C). Final caffeine treatment revealed that the SR Ca 2+ load was not significantly elevated after additional NO 2 -OA treatment (Fig. 4D).
Increased SR Ca 2+ levels have been closely correlated with increased SR Ca 2+ leak by RyR2-phosphorylation 29,30 . Of interest, caffeine treatment of short-time paced cardiomyocytes (1 Hz) demonstrated equal SR Ca 2+ loads in vehicle-and NO 2 -OA treated cells. SR Ca 2+ load was significantly elevated after Iso stimulation, an effect which was slightly, but not significantly, attenuated upon additional NO 2 -OA treatment (Supplemental Figure S3). ca 2+ sensitivity. We further investigated the overall influence of NO 2 -OA on cytosolic Ca 2+ levels and arrhythmia development upon increasing Ca 2+ concentrations (representative Ca 2+ transients are shown in Fig. 5A). Loading of the cells with Ca 2+ concentrations ranging from 0.25 to 2 mM enhanced systolic calcium transients (Fig. 5B) as well as diastolic Ca 2+ levels, the latter of which were not affected by additional NO 2 -OA www.nature.com/scientificreports/ treatment (Fig. 5C). Of importance, NO 2 -OA treatment completely inhibited Iso induced arrhythmic Ca 2+ release (Fig. 5D).

Catecholamine-induced CaMKII activity and RyR2 phosphorylation. Given its importance in
RyR2 activity and dysfunction, we next investigated the effect of NO 2 -OA treatment on CaMKII activity by performing a substrate binding assay using GST-HDAC4 419-670, which contains a CaMKII activity-dependent binding site 31 . CaMKII activity of Iso stimulated isolated cardiomyocytes was significantly reduced after NO 2 -OA treatment (Fig. 6A). In accordance, NO 2 -OA markedly diminished the Iso induced increase of pro-arrhythmic RyR2 phosphorylation on Ser2814 32 as revealed by immunoblotting (Fig. 6B). CaMKII-mediated Thr17phosphorylation of PLN, which enhances sarcoplasmic reticulum (SR) Ca 2+ -ATPase (SERCA2a) activity 33 , was increased upon Iso treatment and not further affected by NO 2 -OA treatment (Fig. 6C).

Discussion
Herein, we show that the electrophilic fatty acid nitroalkene NO 2 -OA prevents induction of VT after AMI in vivo by attenuating CaMKII dependent RyR2 leak. These results are of significant clinical relevance since sudden cardiac death related to AMI is most frequently attributed to the occurrence of sustained VT or ventricular fibrillation (VF) 34 . Although the absolute rates of VT and VF have declined during the era of percutaneous coronary intervention, patients suffering from VT or VF prior to revascularization still have a significantly impaired outcome and are at a higher risk for stent thrombosis. These data may underestimate the true incidence, since prehospital SCD before fibrinolysis or any other interventions were not included in this analysis 35 . In the present study, we modeled a clinical scenario of myocardial ischemia by LAD ligation and demonstrated that intraperitoneal injection of NO 2 -OA 20 min before he ischemic episode sustainably reduced ventricular arrhythmic events specifically in the acute phase after AMI. The prospect of using NO 2 -OA as a therapeutic option for the treatment of acute VT after AMI is underlined by further beneficial effects of this fatty acid nitroalkenes in various cardiovascular diseases. As we and others have demonstrated, the application of NO 2 -OA preserves left ventricular ejection fraction and reduces infarct size in a mouse model of AMI 19,36 . Moreover, NO 2 -OA also protects from angiotensin II induced atrial fibrillation 12 .
Mechanistically, the occurrence of ventricular arrhythmic events during the acute phase of AMI is attributed to development of proarrhythmic substrates with subsequent formation of reentry circuits 3 . The concomitantly enhanced sympathetic tone leads to cellular Ca 2+ overload, resulting in a disturbed Ca 2+ homeostasis, accompanied by diastolic Ca 2+ leak further increasing the vulnerability for VT 37 . On a cellular level, alterations in Ca 2+ transients have been linked to the emergence of alternans and afterdepolarizations which may trigger VTs 4 , as well as to reduced sodium channel availability resulting in ventricular conduction slowing. This increases the wavelength of reentry circuits thereby increasing the vulnerability to VTs 38-40 . www.nature.com/scientificreports/ To investigate the effects of NO 2 -OA on cellular Ca 2+ transients and sarcoplasmic Ca 2+ load, both important in arrhythmia development in AMI, we performed analyses of electrically stimulated isolated cardiomyocytes. We did not detect major changes in both Ca 2+ transient characteristics and in SR Ca 2+ load upon NO 2 -OA treatment.
We quantified PLN phosphorylation, a protein regulating cytosolic Ca 2+ clearance into the SR by SERCA2a 5 , at its CaMKII specific phosphorylation site Thr17 under Iso stimulation. Iso increased PLN phosphorylation but no further change was detectable upon treatment with NO 2 -OA. This indicates that NO 2 -OA does not influence cytosolic Ca 2+ clearance by CaMKII mediated phosphorylation of PLN on this specific site, although we could not fully exclude potential other target sites or phosphorylation mechanisms being modified by NO 2 -OA.
To further examine the underlying molecular effects of NO 2 -OA-mediated VT prevention in detail, we measured RyR2-dependent Ca 2+ leak from the SR in isolated adult cardiomyocytes treated with Iso. We observed an increased number of arrhythmic events (e.g. EADs and DADs) upon Iso treatment that was markedly reduced by NO 2 -OA. Therefore, we hypothesized that NO 2 -OA beneficially influences CaMKII-dependent modulation of Ca 2+ -homeostasis. Indeed, CaMKII activity and subsequent RyR2 phosphorylation on the critical serine residue (Ser2814), which has been linked to CaMKII-mediated Ca 2+ leak in cardiomyopathy and catecholamine treatment 41 , was significantly attenuated by NO 2 -OA in Iso-stressed cardiomyocytes (Fig. 6B). These results are in accordance with recent data demonstrating that increased RyR2 phosphorylation and thereby altered Ca 2+ sensitivity may lead to local Ca 2+ waves, subsequently depolarizing the membrane potential and forming DADs. These local events furthermore trigger Ca 2+ waves in adjacent cells and create propagating arrhythmic events 42 .
Speculating on underlying biochemical mechanisms, it is noted that CaMKII activity and subsequent RyR2 phosphorylation is regulated by S-nitrosylation of specific cysteines (e.g. Cys273/ Cys290) 43 . Previous work shows that NO 2 -OA could act as a NO-donor for S-nitrosylation via a modified Nef reaction. This minor and controversial reaction, as opposed to the signaling responses induced by the post-translational alkylation of protein thiols, occurs in very low yields and only yields NO in aqueous buffered solutions when no protein is present 44 . Thus, it is anticipated that under pathological conditions, NO 2 -OA acts via a reversible nitroalkylation of functionally-significant cysteines 11,19 . The further investigation of critical protein targets in this model of cardioprotection will be the subject of further studies.
As this study was performed in a small animal model, a number of limitations may apply when trying to translate the present findings to human pathology. The mechanisms underlying VT in the setting of AMI are the result of multiple alterations, e.g. loss of viable myocardium with loss of cell-to-cell contact, changes in cell www.nature.com/scientificreports/ membrane potential, ion composition, and other factors predisposing for sudden cardiac death which are only partly mirrored by our experimental setting 3 . Excessive catecholamine stimulation may be just one of multiple pathways leading to VT development during AMI modulated by NO 2 -OA. However, if other mechanisms, e.g., those responsive to acidosis and cell hypoxia, are altered by NO 2 -OA remains to be investigated.
In summary, we show for the first time that, by inhibiting CaMKII, NO 2 -OA modulates a critical pathway of electrical remodeling following AMI and thus has potential as a pre-emptive anti-arrhythmic strategy for patients at risk for developing AMI.

conclusions and perspectives
Our current data reveal that NO 2 -OA rapidly stabilizes myocellular Ca 2+ homeostasis by inhibition of CaMKII activity in the setting of AMI (Fig. 7). Crucially, herein we show for the first time that, by modulation of this pathway, NO 2 -OA markedly suppresses the susceptibility to ventricular arrhythmias.
There are limited treatment options for prevention of acute arrhythmic tachycardias within the first 72 h after AMI, especially considering the amount of time to reach therapeutic levels for available drugs 45 . In this regard, the long-term dietary supplementation of NO 2 -OA or its precursors, might be a therapeutic option to protect AMI patients against subsequent VT / SCD.

Materials and methods
An extended description of the methods can be found within the Supplemental Material.  www.nature.com/scientificreports/

Right ventricular stimulation. A detailed stimulation protocol can be found within the Supplemental
Material. In short, right ventricular stimulation was performed 20 min after induction of myocardial ischemia, while the mouse was kept under anesthesia, according to a standardized protocol that initially included a programmed ventricular stimulation with a fixed S1S1 interval to whose last S1-impulse a short-coupled additional stimulus was applied. After 10 s of recovery, automated burst stimulation was performed according to the protocol. VT were defined as a series of repetitive ventricular ectopic beats lasting for > 200 ms 26 .
In vivo electrophysiological mapping. In brief, directly after electrophysiological investigation the heart was exposed by thoracotomy. A 32-electrode microelectrode array (MEA, Multichannel Systems, Reutlingen, Germany) was positioned on the epicardial surface of the left ventricle apico-septally of the peri-ischemic region as described previously 26 . Field potentials were recorded using a 128-channel, computer-assisted recording system (Multichannel Systems). Data was evaluated according to Lammers et al. 46 . In short, inter-electrode conduction latencies (reciprocal conduction velocity) were calculated for all neighboring electrodes. Of each neighboring quadruplet of electrodes, the largest latency was taken and plotted as a phase map. From this phase map the mean latency of conduction was calculated. The variation coefficient of the phase map was calculated (Percentile (P): P 5-95 /P 50 ) to receive the index of inhomogeneity as velocity independent factor of conduction inhomogeneity. Phase maps were calculated using custom-programmed software (Excel).
Isolation of adult ventricular cardiomyocytes. Ventricular myocytes were obtained from 8 to 12 week old male wild-type mice (FVB/N) as described previously 28 .   www.nature.com/scientificreports/ Assessment of Ca 2+ transients. All experiments were performed at 37 °C within 6 h after cell isolation, as described previously using ratiometric Ca 2+ imaging with Fura-2 dye (IonOptix) 28 . For Ca 2+ transient analyses the cells were incubated with either 0.1% EtOH as control, 10 µM Iso in 0.1% EtOH or 10 µM Iso and 5 µM NO 2 -OA in 0.1% EtOH for 10 min. Calcium transients were recorded under electrically stimulated biphasic field pulses (20 V, 4 ms) at a frequency of 1 Hz for 1 min.
SR Ca 2+ leak and load were measured according to a modified protocol 28,48 . Fura-2 loaded ventricular cardiomyocytes were incubated with 10 µM Iso for 7 min and stimulated for 3 min at 1 Hz, 20 V, 4 ms until cellular Ca 2+ transients reached a steady state followed by a burst stimulation with 4 Hz for 30 s. Directly after the last pulse the pacing was stopped and the normal Tyrode solution was substituted by a 0 Na + /0 Ca 2+ Tyrode supplemented with 10 mmol/l EGTA and 1 mmol/l of the RyR2 inhibitor tetracaine in which Na + was replaced by Li + . This condition allowed measuring intracellular Ca 2+ levels in a closed system without trans-sarcolemmal Ca 2+ fluxes and prevents SR Ca 2+ leak into the cytoplasm. After 40 s recording time the solution was switched back to 0 Na + /0 Ca 2+ Tyrode without tetracaine for another 40 s to unblock the ryanodine receptors and allow potential Ca 2+ leak. SR Ca 2+ leak was estimated as the difference between the Fura-2 ratio recorded at the end of the 0 Na + /0 Ca 2+ Tyrode perfusion with and without tetracaine. At the end of the protocol, 10 mM caffeine was applied to evaluate the total SR Ca 2+ content.
Ca 2+ sensitivity: Fura-2 loaded ventricular cardiomyocytes were incubated with Tyrode solution containing 0.25 mM, 0.5 mM, 1 mM or 2 mM of Ca 2+ together with either 10 µM Iso in 0.1% EtOH or 10 µM Iso and 5 µM NO 2 -OA in 0.1% EtOH. For each calcium concentration cells were paced at 1 Hz for 30 s following a 20 s nonpacing period in which pro-arrhythmic calcium release events was counted. Ca 2+ transient height was determined within the last paced Ca 2+ transient before the non-paced period.  www.nature.com/scientificreports/