Ventricular myocytes were isolated from a redox-dead transgenic mouse model in which a redox-sensitive Cysteine was made resistant to oxidative activation by knock-in replacement with Serine (PKARIαCys17Ser knock-in mouse model, KI). The point mutation of Cys17Ser was introduced into exon 1 of the Prkar1a gene as described previously (19). Ventricular myocytes from wild-type littermates were used as control. All animal procedures were performed in accordance with the Guide for the Care and Use of Laboratory Animals and approved by the Institutional Animal Care and Use Committee.

The isolation of ventricular cardiomyocytes was performed as previously reported (23) using adult (12–15 weeks) mice of both sexes in equal shares. Hearts were mounted on a Langendorff perfusion apparatus and were retrogradely perfused for 7–9 min, starting with a calcium-free solution containing (in mmol/L): NaCl 113, KCl 4.7, KH₂PO₄ 0.6, Na₂HPO₄ 0.6, MgSO₄ 1.2, phenol-red 0.032, NaHCO₃ 12, KHCO₃ 10, HEPES 10, taurine 30, BDM (2,3 butanedione monoxime) 10, and glucose 5.5 (at 37°C, pH 7.4). After 4 min, trypsin 0.6% (Thermo Fisher Scientific Inc. of Waltham, MA, USA), 7.5 mg/ml Liberase™ (Roche Diagnostics, Mannheim, Germany), and 0.125 mmol/L CaCl₂ were added and the hearts were perfused until they became flaccid. After enzymatic digestion, the ventricular myocardium was dissected from the atrium and mechanical dissociation of the heart was performed in a solution containing 10% bovine calf serum (BCS; Sigma-Aldrich, St. Louis, MO, USA). For immediate measurements, a gradual Ca²⁺ reintroduction was performed from 0.1 to 0.8 mmol/L. The isolated cardiomyocytes were plated on laminin-coated glass coverslips. Cardiomyocytes were allowed to settle for 15 min at room temperature to allow cell adhesion for the following experiments.

ATX-II, which was used to increase I_NaL, was purchased from Abcam (ab141870, Cambridge, Great Britain). The sources of the fluorescent dyes and antibodies are given in the following corresponding sections. If not indicated otherwise, Standard Tyrode's solution was used as the experimental solution, containing (in mmol/L) NaCl 140, KCl 4, HEPES 5, MgCl₂ 1, glucose 10, and CaCl₂ 1. The pH value was adjusted to 7.40 at 37°C (referred to as “normal Tyrode's solution”, NT).

Ruptured-patch whole-cell voltage-clamp was used to measure I_NaL and I_Ca as previously reported (23–25). The myocytes were mounted on the stage of a microscope (Nikon TE2000-U). They were incubated for 15 min with the respective bath solution before measurements were undertaken. Notably, I_NaL and I_Ca measurements were performed within the same time limits as the Ca²⁺ imaging experiments (i.e., following at least 10 min of ATX-II exposure). The experimental groups were additionally treated with 1 nmol/L ATX-II. Cardiomyocytes typically achieved a seal greater than 1 GΩ and a resistance <10 MΩ after rupture. For I_NaL measurements, the microelectrodes (2–3 MΩ) were filled with (in mmol/L) the following: CsCl 95, Cs-glutamate 40, NaCl 10, MgCl₂ 0.92, Mg-ATP 5, Li-GTP 0.3, HEPES 5, niflumic acid 0.03, nifedipine 0.02, strophanthidin 0.004, EGTA 1, and CaCl₂ 0.36 [free (Ca²⁺)_i, 100 nmol/L, pH 7.2, CsOH]. The bath solution contained (in mmol/L) the following: NaCl 135, tetramethylammonium chloride 5, CsCl 4, MgCl₂ 2, glucose 10, and HEPES 10 (pH 7.4, CsOH). To measure I_NaL, the cardiomyocytes were held at −120 mV, and I_NaL was elicited using a train of pulses to −35 mV (1,000 ms duration, 10 pulses, basic cycle length (BCL) 2 s). Recordings were initiated 3–4 min after rupture and were performed at room temperature. The measured current was integrated (between 100 and 500 ms) and normalized to membrane capacitance. For I_Ca measurements, the microelectrodes (2–3 MΩ) were filled with (in mmol/L) the following: CsCl 86, Cs-glutamate 40, MgCl₂ 0.92, Mg-ATP 5, Li-GTP 0.3, HEPES 10, EGTA 5, and CaCl₂ 1.8 [free (Ca²⁺)_i 100 nmol/L, pH 7.2, CsOH]. The bath solution contained (in mmol/L) the following: NaCl 140, CsCl 4, MgCl₂ 1, glucose 10, HEPES 10, and CaCl₂ 1 (pH 7.4, CsOH). The signals were filtered with 2.9 and 10 kHz Bessel filters and recorded with an EPC10 amplifier (HEKA Elektronik). The recordings were started 2–3 min after rupture and conducted at room temperature. I_Ca analysis was performed using LabChart 10 (ADInstruments) and determined by subtracting the steady-state current from the peak I_Ca current. Subsequently, the amplitude was normalized to cell capacitance. For analysis of I_Ca inactivation kinetics, time constants τ_slow (tau slow) and τ_fast (tau fast) were calculated by biexponential fitting of I_Ca currents using ClampFit.

For the assessment of cytosolic Ca²⁺, isolated cardiomyocytes were loaded with 10 µmol/L Fura-2-AM (Thermo Fisher Scientific Inc. of Waltham, MA, USA) in the presence of 0.02% (w/v) pluronic acid (Molecular Probes, Eugene, OR, USA) for 15 min at room temperature in darkness (23, 25). Afterward, the cells were incubated for another 15 min in NT to ensure complete deesterification of intracellular Fura-2. After deesterification, Fura-2-loaded cardiomyocytes were mounted on the stage of a Nikon Eclipse TE200-U inverted epifluorescence microscope. Fura-2 was alternately excited at 340 ± 5 and 380 ± 5 nm (250 pairs per second, HyperSwitch system, IonOptix Corp., Westwood, MA, USA) and emitted fluorescence was collected at 510 ± 20 nm. Excitation light was provided by a 75 W xenon arc lamp (Ushio, Japan). During measurements, the cardiomyocytes were continuously transilluminated with red light (>650 nm), stimulated at a frequency of 0.5 Hz, and superfused with experimental solution (with vehicle control or ATX-II, respectively) at a flow rate of 50 ml/h at 37°C. To estimate SR Ca²⁺ content, Ca²⁺ transient amplitude was determined after rapid caffeine application (10 mmol/L, Sigma-Aldrich) during field stimulation pause. After subtraction of background fluorescence, cytosolic Ca²⁺ levels were assessed as the 340/380 nm fluorescence ratio (F₃₄₀/F₃₈₀). Simultaneously with Ca²⁺ measurements, sarcomere length and fractional shortening of the myocytes were recorded using a sarcomere length detection system (MyoCam, IonOptix Corporation, Westwood, MA, USA). Contractions and Ca²⁺ transients were averaged over about 10 beats for the analysis of all parameters. Recorded data were analyzed with the software IONWizard (IonOptix Corp.).

Ca²⁺ spark frequency was analyzed using a laser scanning confocal microscope (Zeiss LSM 700 Pascal; Göttingen, Germany). Isolated cardiomyocytes were loaded with 10 µmol/L Fluo-4-AM (Thermo Fisher Scientific Inc. of Waltham, MA, USA) in the presence of 0.02% (w/v) pluronic acid (Molecular Probes, Eugene, OR, USA) for 15 min at room temperature in darkness (23). Afterward, the solution was replaced by Tyrode's solution. The cardiomyocytes were left to incubate for further 15 min for complete deesterification until the Fluo-4-loaded cardiomyocytes were placed on the stage of a laser scanning confocal microscope. Measurements were conducted in the line-scan mode (10,000 lines per scan). The cardiomyocytes were excited at 488 nm using an argon laser and emission was collected at 505–530 nm through a long-pass emission filter under continuous superfusion with experimental solution (with vehicle or 1 nmol/L ATX-II) and field stimulation at 0.5 Hz at 37°C. Ca²⁺ sparks were detected and quantified using SparkMaster with manual detection of sparks. Ca²⁺ spark frequency (CaSpF) was calculated and normalized to scanning interval and cell width. Only cells that showed stimulated Ca²⁺ transients were included in the evaluation to avoid misinterpretation by an elevated Ca²⁺ spark frequency in non-viable cells. To assess mitochondrial ROS formation, isolated ventricular cardiomyocytes were loaded with 5 µmol/L MitoSox Red (Thermo Fisher Scientific Inc. of Waltham, MA, USA) in presence of 0.02% (w/v) pluronic acid (Molecular Probes, Eugene, OR, USA) for 15 min at 37°C. The cardiomyocytes were excited at 488 nm using an argon laser and fluorescence emission was collected at 610 nm through a long-pass emission filter. Once every minute, frame scans were acquired and MitoSox Red fluorescence emission was normalized to the initial fluorescence (expressed as F/F₀). Likewise, cytosolic ROS formation was assessed by excitation of CellRox Orange-loaded cardiomyocytes at 555 nm while fluorescence emission was collected at 570 nm (Thermo Fisher Scientific Inc. of Waltham, MA, USA, incubation for 30 min with 5 µmol/L at 37°C). Please note that mitochondrial ROS formation as depicted in Figure 1D was assessed in independent sets of experiments in WT vs. KI myocytes with higher initial MitoSox fluorescence in KI cells at baseline (1,665 ± 188 a.u.) as compared with WT (1,004 ± 52 a.u.).

For Western blot analysis, the cardiomyocytes were treated with vehicle or ATX-II (1 nmol/L) and incubated for 15 min at room temperature. The harvested cardiomyocytes were frozen immediately. The proteins were denatured for 5 min at 95°C under non-reducing conditions for PKARIα dimer and PKARIα monomer. To assess pSer2809 RyR2, pSer2814 RyR2, RyR2, pSer16 phospholamban (PLB), and PLB, the proteins were denatured for 30 min at 37°C in presence of 10% β-mercaptoethanol. They were denatured for 5 min at 95°C in presence of 10% β-mercaptoethanol for assessment of pThr287 CaMKII and CaMKIIδ expression. Following denaturation, the proteins were separated on 5% (RyR2, pSer2809 RyR2, pSer2814 RyR2), 8% (PKARIα dimer, PKARIα monomer, pThr287 CaMKII, and CaMKIIδ), or 12.5% (PLB, pSer16 PLB) SDS-polyacrylamide gels, transferred to a nitrocellulose membrane and incubated with the following primary antibodies: rabbit polyclonal anti-pThr287 CaMKII (1:1,000, PhosphoSolutions, Aurora, CO, USA), rabbit polyclonal anti-CaMKIIδ (1:10.000, Thermo Fisher Scientific Inc. of Waltham, MA, USA), mouse monoclonal anti-PKARIα (1:1,000, BD Biosciences, Heidelberg, Germany), rabbit polyclonal anti-RyR2 antibody (1:10.000, Sigma-Aldrich, St. Louis, MO, USA), rabbit polyclonal anti-pSer2809 RyR2 antibody (1: 1,000, Badrilla, Leeds, United Kingdom), rabbit polyclonal anti-pSer2814 RyR2 antibody (1:1,000, Badrilla, Leeds, United Kingdom), mouse monoclonal anti-PLB (1:10.000, Thermo Fisher Scientific Inc. of Waltham, MA, USA), rabbit polyclonal anti-pSer16 PLB (1:500, Badrilla, Leeds, United Kingdom), and mouse monoclonal anti-GAPDH (1:10.000, Sigma-Aldrich, St. Louis, MO, USA) at 4°C overnight. Secondary antibodies were horseradish peroxidase (HRP)-conjugated sheep anti-mouse and donkey anti-rabbit IgG (1:10.000, GE Healthcare, Chicago, Illinois, USA) and incubated for 1 h at room temperature. For chemiluminescent detection, Immobilon™ Western Chemiluminescent HRP Substrate (Millipore) was used. The values were normalized to GAPDH and WT vehicle values later.

Data are presented as means ± standard error of the mean (SEM). Statistical analyses were performed using unpaired Student's t-test, one-way ANOVA, and two-way ANOVA for repeated measurements (RM) with the Holm–Sidak post-hoc test and Mantel–Cox test for survival. Values of p < 0.05 were considered as statistically significant. The graphs were created using Sigma Plot 12 and Graph Pad Prism 8.0.1.

In a first step, we wanted to exclude the possibility that the knock-in replacement of Cys17 with Serine in the regulatory subunit of PKARIα might lead to a different responsiveness of I_NaL to ATX-II in KI cells. Therefore, we exposed isolated ventricular cardiomyocytes of both genotypes to 1 nmol/L ATX-II. As depicted in the original measurements in Figure 1A, ATX-II led to an approximately threefold increase of I_NaL in both WT and KI cells (mean values of the I_NaL integral are shown in Figure 1B). As previously reported by Viatchenko-Karpinski et al. (17), this increase in I_NaL was associated with significant cytosolic (Figure 1C) and mitochondrial ROS production (Figure 1D) in both genotypes in our model as well.

In a next step, we aimed to investigate whether oxidatively activated PKARIα is required for the ATX-II-dependent and PKA-mediated activation of intracellular Ca²⁺ handling that has been previously described by Eiringhaus et al. (9). As shown in Figures 2A,B, ATX-II caused a fairly immediate and time-dependent increase in the amplitude of the systolic Ca²⁺ transient that reached its maximum about 6–10 min after infusion of ATX-II. However, this pronounced, approximately threefold increase in systolic Ca²⁺ transient amplitude was largely comparable between WT and KI cells, which indicates that oxidatively activated PKARIα is functionally not required here (p = 0.368 for interaction, see Figure 2B). In line with this, a positive inotropic effect in terms of time-dependently increased fractional myocytes shortening was comparably observed in both WT and KI cells (Figure 2C). Further, rapid caffeine application was used to assess SR Ca²⁺ content in WT and KI myocytes (Figures 3A,B). Original traces in Figure 3A show that SR Ca²⁺ content was significantly increased by ATX-II treatment in both WT and KI myocytes without major differences between genotypes. The mean values in Figure 3B illustrate dose-dependency of SR Ca²⁺ load following ATX-II treatment in both genotypes without significant differences between WT and KI cells (p = 0.873 for interaction). The ATX-II-dependent rise in SR Ca²⁺ content was further associated with a time-dependent increase in the frequency of diastolic Ca²⁺ spark events as a measure of diastolic Ca²⁺ leakage from the SR (see Figure 3C, original confocal line-scans are shown in Figure 3D). Equivalent to Ca²⁺ transient amplitudes and SR Ca²⁺ content, this ATX-II-dependent increase in Ca²⁺ spark frequency did not differ between WT and KI cells (p = 0.765, see Figures 3C,D), further pointing to the fact that activated Ca²⁺ handling does not require oxidized PKARIα. On the protein level, an unchanged phosphorylation status of Serine 2809 at the RyR2 was observed upon ATX-II exposure that was again not different between genotypes (Ser2809 is the PKA phosphorylation site at the RyR2, see Table 1 and Supplementary Figures S1A and SB). By contrast, increased CaMKII activity in terms of increased autophosphorylation at Threonine 287 was clearly present upon ATX-II, and accompanied by comparable CaMKII-dependent hyperphosphorylation of Serine 2814 at RyR2 in both genotypes (see Supplementary Material Figures S1 and SC–F).

Since ATX-II is known to have proarrhythmogenic effects (26), we also analyzed cellular arrhythmias upon ATX-II treatment in our model [occurring as “non-triggered events” (24, 27, 28) as illustrated in Figure 4A, i.e., as spontaneous Ca²⁺ elevations during Ca²⁺ transient decay] to test whether oxidized PKARIα might exert a functional effect in that regard (see Figures 4A,B). As an overview parameter, we assessed the total proportion of arrhythmic cells in each minute at a certain point in time of the measurement (Figure 4B). By doing so, we observed that 78% of WT myocytes and 57% of KI myocytes developed cellular arrhythmias in terms of non-triggered events (as depicted in Figure 4A) within a time frame of 10 min of treatment with ATX-II, which, however, was not different between groups (p = 0.755 for interaction). Likewise, we did not observe a difference with respect to the ATX-II-dependently increased occurrence of non-stimulated Ca²⁺ waves between genotypes as assessed by confocal microscopy (p = 0.346, data not shown). This lack of change with respect to non-triggered arrhythmogenic events in case of absent oxidative PKARIα activation was mimicked by a similar delay in diastolic Ca²⁺ elimination that was comparably present in both genotypes upon ATX-II, but not different between WT and KI cells (p = 0.664 for interaction, see Figure 4C). In our in vitro model, PKA-specific phosphorylation of PLB at Serine 16 was not different upon acute ATX-II treatment as compared to the control and did not alter between genotypes (see Table 1).

We have previously reported that oxidized PKARIα exerts a stimulating effect on the L-type-mediated Ca²⁺ current (I_Ca) when acutely stimulated with angiotensin II (AngII) as well as in a long-term setting in vivo succeeding transverse aortic constriction (TAC) (23). Since I_Ca greatly contributes to the amplitude of the systolic Ca²⁺ transient, the relevance of I_Ca for the acute activation of Ca²⁺ handling upon ATX-II was investigated in a next step. As shown in Figures 5A,B and in Supplementary Material Figure S2, ATX-II did not alter the magnitude (Figure 5B) or the inactivation properties (Supplementary Material Figure S2) of I_Ca in WT myocytes. In contrast to our previous findings, we also failed to observe a functional effect on I_Ca in KI cells despite clearly increased ROS formation upon ATX-II (compare Figures 1C,D). Notably, we did not observe increased oxidation (i.e., dimerization) of PKARIα upon ATX-II in WT cells at 1 or 2 mmol/L Ca²⁺ (while no dimerization of PKARIα was observed in KI cells as expected, see Figures 5C,D). Hence, oxidation of PKARIα appears not to be required for the ATX-II-mediated acute activation of Ca²⁺ handling despite a clearly present pro-oxidant intracellular milieu.

To induce an increase of I_NaL, which typically occurs in heart failure, we treated mouse cardiomyocytes from WT and PKARIα KI mice with ATX-II. An increased I_NaL in heart failure is known to be associated with arrhythmias (29) and impaired Na ⁺-Ca²⁺ handling (4, 6). However, Na⁺ channels in ventricular myocytes are influenced by numerous factors. Among others, I_NaL is enhanced by CaMKII (30) and can be progressively induced by ROS (31, 32). Most importantly in regard to our study, protein kinase A is also involved in the regulation of Na⁺ channels (33, 34) and the modulation of I_NaL (35). Hence, PKA activation may potentially induce a positive feedback mechanism and further enhance I_NaL. To exclude the possibility that treatment with equal concentrations of ATX-II affects ion currents in genotypes differently, we assessed the I_NaL in cardiomyocytes of both genotypes under control conditions and upon ATX-II treatment. Treatment with 1 nmol/L ATX-II resulted in an approximately threefold increase in I_NaL without significant differences between genotypes. Hence, it could be assumed that initial conditions were equal in both genotypes and that I_NaL is not subject to oxidatively-activated PKARIα (oxPKA)-dependent regulation upon acute ATX-II exposure.

We next tested whether ATX-II-dependently increased Na⁺ influx into the cell led to increased ROS formation in our setting as it was proposed previously (10). Various studies have shown that altered Na⁺ and Ca²⁺ handling in cardiomyocytes can lead to increased mitochondrial ROS production. Interestingly, there are several mechanisms leading to increased mitochondrial ROS production that may be relevant in our model, as follows: Increased cytosolic Na⁺ levels have been shown to reduce mitochondrial Ca²⁺ accumulation by promoting Ca²⁺ efflux through the mitochondrial Na⁺-Ca²⁺-exchanger (mNCE) (10, 11). Kohlhaas et al. demonstrated that decreased mitochondrial Ca²⁺ levels were associated with increased mitochondrial ROS production (10). In addition, it has also been suggested that enhanced oxidative phosphorylation during beta-adrenergic stimulation (that is needed to meet the increased energy demand) leads to increased electron leakage and thus mitochondrial ROS formation (36). In our present study, we found significantly increased mitochondrial as well as cytosolic ROS formation in both WT and KI myocytes following acute ATX-II exposure in the face of a comparable positive inotropic effect in both groups. We therefore interpret increased ROS formation in our model as the result of at least two mechanisms, namely, (i) cytosolic Na⁺/Ca²⁺ overload leading to mitochondrial ROS formation and (ii) enhanced oxidative phosphorylation in the mitochondria as a consequence of positive inotropy. Regardless of the exact mechanism, we clearly observed increased ROS production in WT and KI cells following ATX-II treatment so that we could largely exclude the possibility that ATX-II might have failed to increase ROS formation in one of the genotypes.

We have previously found that oxidized PKARIα is an important factor for the maintenance of functional Ca²⁺ handling in the setting of enhanced oxidative stress by sustaining I_Ca (23). Hence, we now sought to determine the impact of oxidatively activated PKARIα on Ca²⁺ handling in the setting of acutely increased I_NaL with consecutively enhanced oxidative stress. In line with previous studies (9), we observed a positive inotropic effect upon ATX-II in terms of significantly increased Ca²⁺ transients and increased fractional shortening. In addition, we found an impeded diastolic decay of the Ca²⁺ transient that was paralleled by a time-dependent increase in arrhythmogenic non-triggered Ca²⁺ elevation events during Ca²⁺ transient decay, which altogether resulted in a delay of diastolic Ca²⁺ elimination. However, these effects were equally present in both genotypes, which speaks against a crucial involvement of oxidized PKARIα.

Likewise, ATX-II treatment induced diastolic SR Ca²⁺ leakage in both genotypes to a similar extent. An increase in SR Ca²⁺ leakage following enhanced I_NaL is generally in accordance with previous studies. Various factors such as activation of CaMKII (24) and direct ROS-dependent regulation of Ca²⁺ spark frequency (37) have been shown to modify this effect. Fischer et al. investigated the effect of I_NaL on the induction of SR Ca²⁺ leakage in atrial murine myocytes and uncovered that it was due to the activation of both CaMKII and PKA (38). By contrast, Wagner et al. ruled out a Nox2-dependent (and therefore redox-related) influence of PKA on SR Ca²⁺ leak and concluded that the leak was CaMKII-mediated (39), which is in accordance with another study from our group in which we have observed that acutely increased oxidative stress induces SR Ca²⁺ leakage independent of RIα dimer formation (23). Since we did not observe increased phosphorylation of Ser-2809 (which is the PKA-dependent phosphorylation site at the RyR2) here, it can be assumed that the increase in Ca²⁺ spark frequency upon ATX-II-related ROS generation as observed in our model is likely also not PKA dependent. Moreover, this finding indicates again that redox-activated PKA is not a major driver of SR Ca²⁺ leak in cardiac myocytes. Instead, in our experimental setting, ATX-II-activated CaMKII in WT and KI cells as well resulted in subsequently hyperphosphorylated RyR2 at the CaMKII-dependent phosphorylation site 2814 in both genotypes, which points to the fact that the induction of the ATX-II/ROS-mediated SR Ca²⁺ leak is primarily driven by increased CaMKII activity. In addition, ATX-II-dependent SR Ca²⁺ leakage may be a consequence of a sensitization of the RyR2 to the increased Ca²⁺ load of the SR (which increases the driving force for Ca²⁺ release for a given Ca²⁺ load of the SR that was also present as increased systolic fractional Ca²⁺ release).

Despite increased diastolic SR Ca²⁺ loss, however, SR Ca²⁺ content was still significantly elevated after treatment with ATX-II as compared with untreated control cardiomyocytes in both groups, which points to the fact that SR Ca²⁺ loading prevailed over SR Ca²⁺ loss in our model. Enhanced SR Ca²⁺ loading may be due to increased SERCA2a activity. In that regard, Eiringhaus et al. observed maintained, and upon CaMKII inhibition even increased, SR Ca²⁺ load despite I_NaL-related diastolic SR Ca²⁺ loss, which they attributed to increased SERCA2a activity due to PKA-mediated PLB phosphorylation (9). However, in our model, we did not find a functional acceleration of SR reuptake [that can be approximated by Ca²⁺ transient decay kinetics in mouse myocytes (40)] nor evidence for PKA-dependent hyperphosphorylation of PLB (at Ser-16). This lack in ATX-II-dependent and PKA-mediated activation of central target structures of intracellular Ca²⁺ handling in our model may be explained by slightly different incubation strategies with ATX-II resulting in different exposure times to ATX-II (e.g., we did not add ATX-II to the fluorescent dye during incubation here). By contrast, we do not believe that this difference is due to different experimental Ca²⁺ concentrations, because we still failed to observe increased RIα dimerization following ATX-II even upon elevated Ca²⁺ concentrations as used by Eiringhaus et al. (compare Figures 5C,D). Conversely, since we observed clearly activated Ca²⁺ handling, we believe that PKA-dependent phosphorylation of PLB and RyR2 are at least not crucially involved in the very acute effects of ATX-II-dependent Ca²⁺ activation as observed here. Instead, in our murine model, diastolic Ca²⁺ overload and increased Ca²⁺ transient amplitudes are presumably a consequence of increased NCX activity in its reverse mode leading to severe cytosolic Ca²⁺ overload, since ATX-II does not enhance transsarcolemmal Ca²⁺ influx via the LTCC (which would be an alternative mechanism of cytosolic Ca²⁺ overload, see the following). While this effect may be different in larger animals and humans (41), our observation would be in line with previous reports that have demonstrated that inhibition of NCX significantly attenuates intracellular Ca²⁺ overload and also reduces arrhythmias upon increased I_NaL (30, 42). In contrast to our previous study (23), in which oxidized PKARIα maintained I_Ca upon oxidative stress, we failed to detect any functional effect of ATX-II treatment on the amplitude or the inactivation properties of I_Ca in both WT and KI cells here (while I_Ca was slightly but significantly enhanced in KI vs. WT upon ATX-II treatment in the face of a generally lower I_Ca amplitude in WT cells). Our observation that ATX-II-related ROS formation did not result in PKARIα oxidation in WT cells further suggests that this unexpected lack of functional effect on I_Ca is rather a result of absent oxidative PKARIα activation in comparison with a potentially differential regulation of I_Ca by oxidized PKARIα in the context of ATX-II.

We acknowledge that our present study raises many new questions. Most importantly, it remains unclear why elevated cytosolic and mitochondrial ROS as acutely induced by ATX-II do not oxidize PKARIα, yet we believe that this aspect is beyond the scope of this study. Nevertheless, there is no question that this finding is in contrast to earlier studies in which an increased dimerization of the RIα subunit was clearly induced by various forms of oxidative stress such as by chronic pressure overload (23) or by pharmacologically induced oxidative stress using 1-nitrosocyclohexalycetate (NCA) (43). Interestingly, preliminary and as yet unpublished data by Simon et al. suggest that endogenous ROS indeed do not necessarily oxidize PKARIα in the heart, which appears to be correct for ROS as induced by NADPH oxidase 2 and 4 (NOX2 and NOX4), xanthine oxidase (XO), and the mitochondria (44). Instead, nitric oxide might turn out to be the relevant source for PKARIα dimer formation in the heart. Hence, our results may be at least hypothesis-generating in a way that it supports the notion that different stimuli and sources of ROS (or even NO) might differentially facilitate the dimerization of the PKARIα subunit. Conversely, we cannot rule out that chronically elevated I_NaL (such as in HF or in experimental models leading to HF such as TAC or myocardial infarction) would lead to PKARIα oxidation with consecutive functional consequences in intracellular Ca²⁺ handling and should therefore be subsequently investigated.