Oxidative Stress and Inflammatory Modulation of Ca2+ Handling in Metabolic HFpEF-Related Left Atrial Cardiomyopathy

Metabolic syndrome-mediated heart failure with preserved ejection fraction (HFpEF) is commonly accompanied by left atrial (LA) cardiomyopathy, significantly affecting morbidity and mortality. We evaluate the role of reactive oxygen species (ROS) and intrinsic inflammation (TNF-α, IL-10) related to dysfunctional Ca2+ homeostasis of LA cardiomyocytes in a rat model of metabolic HFpEF. ZFS-1 obese rats showed features of HFpEF and atrial cardiomyopathy in vivo: increased left ventricular (LV) mass, E/e’ and LA size and preserved LV ejection fraction. In vitro, LA cardiomyocytes exhibited more mitochondrial-fission (MitoTracker) and ROS-production (H2DCF). In wildtype (WT), pro-inflammatory TNF-α impaired cellular Ca2+ homeostasis, while anti-inflammatory IL-10 had no notable effect (confocal microscopy; Fluo-4). In HFpEF, TNF-α had no effect on Ca2+ homeostasis associated with decreased TNF-α receptor expression (western blot). In addition, IL-10 substantially improved Ca2+ release and reuptake, while IL-10 receptor-1 expression was unaltered. Oxidative stress in metabolic syndrome mediated LA cardiomyopathy was increased and anti-inflammatory treatment positively affected dysfunctional Ca2+ homeostasis. Our data indicates, that patients with HFpEF-related LA dysfunction might profit from IL-10 targeted therapy, which should be further explored in preclinical trials.


Introduction
In heart failure (HF) with preserved ejection fraction (HFpEF), patients suffer from typical HF symptoms (i.e., dyspnea, reduced exercise capacity), while presenting normal left ventricular ejection fraction (>50%). HFpEF is increasingly prevalent and currently accounts for approximately half of all HF patients [1]. While pharmacologic treatment options for HF with reduced ejection fraction (HFrEF) evolved notably during last decades, drug therapy for HFpEF remains elusive, despite high morbidity and mortality [2,3]. One reason for failure of large clinical trials in HFpEF may be attributed to the complex and heterogenous etiology of the disease. A broad spectrum of systemic triggers have been have been identified in HFpEF, such as a pro-inflammatory state, arterial and microvascular

Echocardiography
Echocardiography was performed and analyzed as previously described [20] by an experienced observer (N.H.) using a Vevo lab ultrasound system (Visual Sonics, Toronto, ON, Canada).

Cardiomyocyte Isolation
Primary left atrial cardiomyocytes were isolated as described previously [22]. In brief, the rat heart was excised and mounted to a Langendorff-apparatus (self assembly). Pressure of the left atria was manipulated using a novel approach as reported previously [22]. Myocardial tissue was digested using Ca 2+ free solution containing highly purified collagenase. Isolated left atrial cardiomyocytes were readjusted to Ca 2+ in a stepwise manner and placed on laminin-coated dishes for sub-sequent experiments.

Mitochondrial Structure
Left atrial cardiomyocytes were loaded with Mitotracker red, transferred to an LSM 800 laser scanning microscope (Carl Zeiss AG, Oberkochen, Germany) and two-dimensional images acquired (excitation 561 nm, emission: 565-700 nm). Binary images representing the presence and absence of mitochondria were derived using a 2-step Otsu thresholding algorithm [23]. The perimeter of mitochondrial structures was quantified by edge detection. For each cell, the ratio of the total mitochondrial perimeter and the total mitochondrial surface area was calculated as an indicator of mitochondrial fission.

Reactive Oxygen Species (ROS)
Left atrial cardiomyocytes were loaded with H2-DCF and transferred to an LSM 800 laser scanning microscope. A time series of 15 two-dimensional images were recorded (excitation 488 nm, emission: 505-525 nm; interval 2 s). Signal intensity of the initial image was defined as F 0 , production of ROS calculated as ∆F = F − F 0 per image, averaged per image sequence and reported as the respective rate ∆F/(F 0 × t).

Ca 2+ Imaging
Left atrial cardiomyocytes were incubated with Tyrode solution in the presence or absence of TNF-α (10 nM) or IL-10 (10 nM) at 37 • C for 30 min. Sub-sequently, cells were loaded with Fluo-4 for another 30 min at 37 • C (in the continuous presence of the respective cytokine). The cells were then washed twice with Tyrode solution and transferred to an LSM 800 laser scanning microscope. Experiments were conducted at 37 • C at steady state following approximately 1 min of electric stimulation at 1 Hz. Ca 2+ transients (CaT) were recorded for 10 s and averaged prior to analysis. Time-dependent changes of cytosolic [Ca 2+ ] during electric pacing are reported as ∆F/F 0 . Tau of decay (mono-exponential fit) is reported as a mesuare of Ca 2+ reuptake. For early Ca 2+ release (ER) site analysis: F50 was defined as 50% of the CaT amplitude and the corresponding time-to-F50 (TF50) calculated as an indicator of ER. Scan lines along the longitudinal axis were grouped into 1 µm intervals, indicating active couplons [24,25]. ER was defined to be smaller than the average TF50 of the control group (<15.0 ms) and an ER site was defined to be an active couplon with ER events in at least two out of six consecutive stimulation cycles. The fraction of ER events ER sites in six consecutive cycles was quantified as the probability of ER.

Statistical Analysis
Results are shown as mean ± standard error. Statistical tests and p-values are reported in the figure legend. A p-value below 0.05 was assumed to be of statistical significance.

ZSF-1 Obese Rats Show Distinct Features of Metabolic HFpEF-Related Left Atrial Cardiomyopathy
Left ventricular function was assessed in vivo with echocardiography ( Figure 1A) and left ventricular mass weighed post-mortem. Left ventricular ejection fraction remained preserved ( Figure 1B), while left ventricular mass was increased ( Figure 1C). Combined mitral valve and tissue Doppler analysis yielded an increased E/e' ratio in HFpEF, a measure of diastolic dysfunction ( Figure 1D). Left atria were severely enlarged in HFpEF, indicating advanced adverse remodeling ( Figure 1E). Antioxidants 2020, 9, x FOR PEER REVIEW 5 of 13

HFpEF Shows Increased Mitochondrial Fission and Production of ROS
To investigate the contribution of mitochondrial function to the pathology of HFpEF-related atrial cardiomyopathy, left atrial cardiomyocytes were isolated and mitochondria examined in vitro. Structural analysis of mitochondria revealed a reduced perimeter to area ratio in HFpEF, indicating enhanced mitochondrial fission (Figure 2A,B). The rate of oxidation was measured to evaluate mitochondrial function. In accordance with increased mitochondrial fission, left atrial cardiomyocytes in HFpEF also showed an increased production of ROS ( Figure 2C,D).

Ca 2+ Signaling during Excitation-Contraction Coupling in HFpEF Is Not Affected by TNF-α, but Benefits from Exposure to IL-10
In order to assess inflammatory modulation of Ca 2+ cycling in HFpEF, CaT of left atrial cardiomyocytes were recorded after exposure to TNF-α and its anti-inflammatory counterpart IL-10 ( Figure 3A). In WT, TNF-α led to a significant reduction of CaT amplitudes ( Figure 3B), which was linked to a deceleration of Ca 2+ removal ( Figure 3C). Anti-inflammatory treatment with IL-10 did not show an effect on CaT in WT ( Figure 3B,C). Interestingly, a diametral effect of IL-10 could be observed in HFpEF. IL-10 mediated an increased CaT amplitude ( Figure 3D), linked to an accelerated Ca 2+ removal ( Figure 3E), while TNF-α did not have an effect on CaT ( Figure 3D,E). Neither TNF-α, nor IL-10 had an effect on Ca 2+ release in WT ( Figure 4A). In HFpEF however,

IL-10 Induces Recruitment and Synchronization of Early Ca 2+ Release Sites in HFpEF
Neither TNF-α, nor IL-10 had an effect on Ca 2+ release in WT ( Figure 4A). In HFpEF however, IL-10 significantly accelerated Ca 2+ release ( Figure 4B). An in-depth analysis of early Ca 2+ release (ER) sites was conducted to evaluate the origin of improved Ca 2+ release ( Figure 4C). The positive effect was found to be mediated by an improved recruitment of ER sites ( Figure 4D), rather than an increased overall probability of ER ( Figure 4E). In addition, IL-10 synchronized ER sites ( Figure 4F).

TNF-α Receptor Expression is Reduced, IL-10 Receptor Expression Partially Maintained in HFpEF.
To further explore the contrasting effects of TNF-α and IL-10 on Ca 2+ signaling during excitation-

TNF-α Receptor Expression Is Reduced, IL-10 Receptor Expression Partially Maintained in HFpEF
To further explore the contrasting effects of TNF-α and IL-10 on Ca 2+ signaling during excitation-contraction coupling in HFpEF, protein expression of their respective receptors was quantified by Western Blot analysis. Left atrial tissue in HFpEF showed a significant down regulation of TNF-α receptors type 1 and type 2 ( Figure 5A-C). IL-10 receptor expression was partially maintained in HFpEF: type 1 expression was unchanged, while type 2 was significantly lower (Figure 5D-F).

Discussion
Left atrial cardiomyocytes exhibited an increased production of ROS in a rat model of metabolic HFpEF-related atrial cardiomyopathy, which was associated with enhanced mitochondrial fission. TNF-α severely impaired Ca 2+ uptake and CaT release amplitudes only in WT, yet it had no effect on left atrial cardiomyocytes in HFpEF. However, CaT amplitudes and Ca 2+ uptake were improved by anti-inflammatory treatment with IL-10. In addition, IL-10 enhanced recruitment of ER sites in HFpEF, thus accelerating and synchronizing Ca 2+ release. Left atrial cardiomyocytes in HFpEF showed reduced expression of TNF-α receptors, while expression of IL-10 receptor type 1 remained intact.
Abnormal Ca 2+ /Na + handling and oxidative stress are important, closely intertwined pathophysiological hallmarks of HF [26,27]. Dysfunctional excitation-contraction coupling demands an increased energy production, which leads to oxidative stress once the buffering capacity of the antioxidant defense system has been exceeded. In addition, an altered cellular Ca 2+ /Na + homeostasis may directly affect mitochondrial [Ca 2+ ] uptake and regulate mitochondrial metabolism [28]. In turn, ROS can trigger Ca 2+ -mediated cellular arrhythmias, i.e., via abnormal Ca 2+ release from the sarcoplasmic reticulum (resulting in delayed after depolarizations [29]), or via stimulation of L-type Ca 2+ current (early after depolarizations) [30]. In this context, the discovery of an increased ROS production of left atrial cardiomyocytes in the setting of metabolic syndrome-related HFpEF ( Figure  2) is of particular importance. The prevalence of atrial fibrillation in HFpEF has been reported as high as 39% (12% higher than HFrEF) [31] and ROS may serve as a promising target to reduce the burden

Discussion
Left atrial cardiomyocytes exhibited an increased production of ROS in a rat model of metabolic HFpEF-related atrial cardiomyopathy, which was associated with enhanced mitochondrial fission. TNF-α severely impaired Ca 2+ uptake and CaT release amplitudes only in WT, yet it had no effect on left atrial cardiomyocytes in HFpEF. However, CaT amplitudes and Ca 2+ uptake were improved by anti-inflammatory treatment with IL-10. In addition, IL-10 enhanced recruitment of ER sites in HFpEF, thus accelerating and synchronizing Ca 2+ release. Left atrial cardiomyocytes in HFpEF showed reduced expression of TNF-α receptors, while expression of IL-10 receptor type 1 remained intact.
Abnormal Ca 2+ /Na + handling and oxidative stress are important, closely intertwined pathophysiological hallmarks of HF [26,27]. Dysfunctional excitation-contraction coupling demands an increased energy production, which leads to oxidative stress once the buffering capacity of the antioxidant defense system has been exceeded. In addition, an altered cellular Ca 2+ /Na + homeostasis may directly affect mitochondrial [Ca 2+ ] uptake and regulate mitochondrial metabolism [28]. In turn, ROS can trigger Ca 2+ -mediated cellular arrhythmias, i.e., via abnormal Ca 2+ release from the sarcoplasmic reticulum (resulting in delayed after depolarizations [29]), or via stimulation of l-type Ca 2+ current (early after depolarizations) [30]. In this context, the discovery of an increased ROS production of left atrial cardiomyocytes in the setting of metabolic syndrome-related HFpEF (Figure 2) is of particular importance. The prevalence of atrial fibrillation in HFpEF has been reported as high as 39% (12% higher than HFrEF) [31] and ROS may serve as a promising target to reduce the burden of atrial electrical dysfunction (i.e., arrhythmias) in HFpEF.
Our experimental data indicates that Ca 2+ handing of left atrial cardiomyocytes in HFpEF may be impaired associated with an upregulation of inflammatory pathways ( Figure 6). In WT, left atrial cardiomyocytes responded to TNF-α with decelerated Ca 2+ removal and a decreased CaT amplitude (Figure 3), in accordance with previous reports of TNF-α induced downregulation of sarcoplasmic/endoplasmic reticulum calcium ATPase (SERCA) 2a [32]. In HFpEF however, TNF-α did not affect Ca 2+ handling. However, left atrial cardiomyocytes responded with an accelerated Ca 2+ uptake and an increased CaT amplitude in response to IL-10. IL-10 directly attenuates TNF-α induced nuclear factor 'kappa-light-chain-enhancer' of activated B-cells (NFκB) pathway activation by upregulating extracellular signal-regulated kinase 1/2 (ERK 1/2) mitogen-activated protein kinase (MAPK) phosphorylation [33], which has also been reported as effectively countering TNF-α induced oxidative stress in cardiomyocytes [34]. In addition to improving Ca 2+ removal and CaT amplitudes, IL-10 also facilitated a notable acceleration and synchronization of Ca 2+ release kinetics ( Figure 4). did not affect Ca handling. However, left atrial cardiomyocytes responded with an accelerated Ca uptake and an increased CaT amplitude in response to IL-10. IL-10 directly attenuates TNF-α induced nuclear factor 'kappa-light-chain-enhancer' of activated B-cells (NFκB) pathway activation by upregulating extracellular signal-regulated kinase 1/2 (ERK 1/2) mitogen-activated protein kinase (MAPK) phosphorylation [33], which has also been reported as effectively countering TNF-α induced oxidative stress in cardiomyocytes [34]. In addition to improving Ca 2+ removal and CaT amplitudes, IL-10 also facilitated a notable acceleration and synchronization of Ca 2+ release kinetics (Figure 4).
From these findings, two plausible approaches to anti-inflammatory treatment of metabolic HFpEF-related atrial cardiomyopathies can be derived: antagonizing TNF-α (i.e., lowering circulating cytokines, receptor blockage at effector site) or agonizing IL-10 (i.e., exogenous substitution). Large clinicals trials evaluating TNF-α inhibitors in HF (Randomised Etanercept Worldwide Evaluation (RENEWAL) [35], Anti-TNF Therapy Against Congestive Heart Failure (ATTACH) [36]) have yielded dissatisfactory results and further testing of TNF-α inhibitors in patients seems currently unlikely. The lacking efficacy of TNF-α inhibitors can be partly supported by the experimental data of this study: at the time of disease manifestation, cardiomyocytes have already entered an 'inflamed state' through NFκB activation and have reduced TNF-α receptor expression ( Figure 5), thus removing the beneficial effector site of lower TNF-α levels. However, antiinflammatory therapy with IL-10 for the treatment of HF seems plausible and has not previously been tested in clinical trials. Expression of IL-10 receptor type 1 was maintained and a satisfactory attenuation of Ca 2+ handling could be observed upon IL-10 treatment. Previous studies with patients suffering from Crohn's disease have reported subcutaneous therapy with recombinant IL-10 to be safe in humans [37]. However, further animal studies (i.e., isolated hearts, gene knockouts) need to be performed in order to confirm the present findings before translational in vivo studies can be performed.  From these findings, two plausible approaches to anti-inflammatory treatment of metabolic HFpEF-related atrial cardiomyopathies can be derived: antagonizing TNF-α (i.e., lowering circulating cytokines, receptor blockage at effector site) or agonizing IL-10 (i.e., exogenous substitution). Large clinicals trials evaluating TNF-α inhibitors in HF (Randomised Etanercept Worldwide Evaluation (RENEWAL) [35], Anti-TNF Therapy Against Congestive Heart Failure (ATTACH) [36]) have yielded dissatisfactory results and further testing of TNF-α inhibitors in patients seems currently unlikely. The lacking efficacy of TNF-α inhibitors can be partly supported by the experimental data of this study: at the time of disease manifestation, cardiomyocytes have already entered an 'inflamed state' through NFκB activation and have reduced TNF-α receptor expression ( Figure 5), thus removing the beneficial effector site of lower TNF-α levels. However, anti-inflammatory therapy with IL-10 for the treatment of HF seems plausible and has not previously been tested in clinical trials. Expression of IL-10 receptor type 1 was maintained and a satisfactory attenuation of Ca 2+ handling could be observed upon IL-10 treatment. Previous studies with patients suffering from Crohn's disease have reported subcutaneous therapy with recombinant IL-10 to be safe in humans [37]. However, further animal studies (i.e., isolated hearts, gene knockouts) need to be performed in order to confirm the present findings before translational in vivo studies can be performed.

Conclusions
Cardiomyocytes in metabolic HFpEF-related atrial cardiomyopathy exhibit mitochondrial fission, increased oxidative stress and dysfunctional Ca 2+ handling. Anti-inflammatory treatment with IL-10 facilitates an attenuation of Ca 2+ signaling of left atrial cardiomyocytes in HFpEF. IL-10 substitution might pose a viable treatment strategy for atrial dysfunction (arrhythmias, contractile dysfunction) in the setting of metabolic syndrome-related HFpEF and its efficacy should be further evaluated in preclinical trials.