Macrophage-Dependent Interleukin-6-Production and Inhibition of IK Contributes to Acquired QT Prolongation in Lipotoxic Guinea Pig Heart

In the heart, the delayed rectifier K current, IK, composed of the rapid (IKr) and slow (IKs) components contributes prominently to normal cardiac repolarization. In lipotoxicity, chronic elevation of pro-inflammatory cytokines may remodel IK, elevating the risk for ventricular arrythmias and sudden cardiac death. We investigated whether and how the pro-inflammatory interleukin-6 altered IK in the heart, using electrophysiology to evaluate changes in IK in adult guinea pig ventricular myocytes. We found that palmitic acid (a potent inducer of lipotoxicity), induced a rapid (~24 h) and significant increase in IL-6 in RAW264.7 cells. PA-diet fed guinea pigs displayed a severely prolonged QT interval when compared to low-fat diet fed controls. Exposure to isoproterenol induced torsade de pointes, and ventricular fibrillation in lipotoxic guinea pigs. Pre-exposure to IL-6 with the soluble IL-6 receptor produced a profound depression of IKr and IKs densities, prolonged action potential duration, and impaired mitochondrial ATP production. Only with the inhibition of IKr did a proarrhythmic phenotype of IKs depression emerge, manifested as a further prolongation of action potential duration and QT interval. Our data offer unique mechanistic insights with implications for pathological QT interval in patients and vulnerability to fatal arrhythmias.


Introduction
Dietary obesity is a major contributor to the increasing prevalence of cardiovascular diseases worldwide [1]. Obesity may have a direct impact on cardiac electrical activity, suggesting that vulnerability to fatal ventricular arrhythmias and sudden cardiac death (SCD) will remain high. The pathology of obesity-related heart diseases [2][3][4][5][6] is associated with cardiac lipotoxicity (or the abnormal accumulation of free fatty acids, FFAs), and enhanced infiltration of activated macrophages. This macrophage activation triggers an increase in the secretion of pro-inflammatory cytokines including interleukin-6 (IL-6), tumor-necrosis factor-alpha (TNF-α), and interleukin-1 (IL-1β) [7,8]. Preventing such lipotoxic effects may inform rationale development of therapeutic interventions for prevention of ventricular arrhythmias in patients. However, the underlying molecular mechanisms are poorly un-derstood. This objective can be advanced by investigating the effects of pro-inflammatory cytokines on ventricular electrical activity and QT interval.
Previous reports have shown that TNF-α decreases the rapid delayed rectifier channels (I Kr ) and the transient outward current (I to ) is inhibited by IL-6 [9] and IL-1β [10], while IL-1β [11] and IL-6 [12], were shown to increase L-type Ca current density. Whether and how the slow component (I Ks ) of the delayed rectifier K current (I K ) is altered by proinflammatory cytokines is unknown. We hypothesize that cytokines decrease I Ks , thereby elevating the risk for fatal arrhythmias and sudden death in metabolic disorders. Our data reveal an additional mechanism for QT interval prolongation.

Q-T Interval Is Prolonged in Lipotoxic Hearts
If lipotoxicity is a potent inducer of IL-6 secretion (Figure 1), and APD is the sole determinant of QT interval in guinea pig, we expect to see rapid and progressive changes in QT interval in lipotoxic guinea pigs. To investigate this, adult guinea pigs were exposed to either 1 mM PA-BSA or BSA alone (controls) through the cranial vena cava (CrVc), and right atrium (RA, Figure 3A). Progressive changes in QT interval were determined in ECG measurements conducted over a period of 14 days ( Figure 3A). Figure 3B shows typical ECG traces measured in non-lipotoxic and lipotoxic guinea pigs at day 7. Interestingly, Figure 3C, confirms our prediction. Thus, compared to non-lipotoxic controls, the change in the heart-rate corrected QT interval (∆QT c ) was significantly greater in lipotoxic guinea pigs ( Figure 3C). Additionally, ∆QT c rapidly reached a peak at day 7 and remained sustained for an additional seven days. Thus, on average ∆QT c was 12.2 ± 8.20 ms (or 271 ± 8.23 ms vs. 287.5 ± 3.7 ms, * P < 0.05, n = 7, open circles, Figure 3C), and 62.1 ± 25.6 ms (or 244 ± 22 ms vs. 303.5 ± 3.7 ms, * P < 0.05, n = 8, filled circles, Figure 3C). There were no measurable differences in the heart rate, PR and QRS intervals at baseline or seven-days after exposure to PA-BSA or BSA (Table 1). Collectively the data suggest that IL-6 may contribute to the initiation of QT prolongation in lipotoxic hearts, while IL-1β, IL-1α and INF-β may act to sustain the QT prolonging effect of IL-6.
To examine the propensity for ventricular arrhythmias in conscious and freely moving lipotoxic guinea pigs, radiotelemetry ECG transmitters (TR54PB, MT10B, KAHA Sciences, Auckland, New Zealand) were implanted, and ECG abnormalities were progressively (24 h, Figure 4A) monitored. Figure 4B illustrates typical premature ventricular contractions (PVCs) recorded in adult guinea pigs. PVC number was significantly higher in lipotoxic guinea pigs compared to controls ( Figure 4C). However, adult lipotoxic guinea pigs that were subsequently exposed to isoproterenol (ISO, 0.5 mg/kg) to reproduce sympathetic stimulation, developed more PVCs than controls or lipotoxic guinea pigs that were not exposed to ISO ( Figure 4C). Torsade de pointes (TdP) was readily inducible in ISO-treated lipotoxic guinea pigs ( Figure 4D); in one of these guinea pigs, we detected an episode of pre-excited atrial fibrillation, which degenerated into ventricular fibrillation and ultimately to SCD ( Figure 4E). Telemetered guinea pigs were exposed to lipotoxicity and altered cardiac rhythms were monitored for 24 h (B), Representative images of cardiac rhythms from adult guinea pigs exposed to ISO demonstrating premature ventricular contractions (PVCs). (C), Quantification of PVCs measured in non-lipotoxic, lipotoxic and lipotoxic +ISO adult guinea pigs. Lipotoxic guinea pigs in the absence and presence of ISO displayed significantly more PVCs than age-matched non-lipotoxic controls. TdP (D) and VF (E) in a free-moving lipotoxic guinea pig exposed to ISO. Data are presented as mean ± SEMs. Scale bar: 0.5 mV × 100 ms (B), 0.5 mV × 10 s (D,E). * Statistical significance at P < 0.05.
As a complementary approach, guinea pigs were subjected to short-term (~50 days, Figure 5A), PA-diet and LFD feeding. ECG measurements were conducted to assess relative changes in QT interval. Figures 5B and 5C show that guinea pigs fed either LFD or PA diet displayed similar weights ( Figure 5B) and blood glucose levels ( Figure 5C), demonstrating a separation of lipotoxicity from hyperglycemia. Adult guinea pigs exposed to PA-diet feeding ( Figure 5D) displayed QT c prolongation after 10 days compared to LFD fed controls ( Figure 5E), confirming the QT c prolonging effect of lipotoxicity. On average, PA-diet feeding significantly increased basal QT c interval by 20% (266.4 ± 10.3 vs. 320.8 ± 13.8, n = 12, * P < 0.05, Figure 5F), but remained essentially unchanged in LFD fed controls (275.7 ± 12.6 vs. 277.3 ± 21.9, n = 10, P > 0.05, Figure 5F). Adult guinea pigs were subjected to either PA-diet or low-fat feeding and changes in QT c were assessed at baseline and after 10 days by conducting ECG measurements. Progressive body weight (B) and fasting blood glucose (C) changes relative to time measured in guinea pigs fed with either PA-diet (•) or LFD ( ). (D), Representative ECG traces measured in a PA-diet fed guinea pig at the beginning and after 10 days, highlighting QT c compared to LFD-fed controls (E). Red horizontal lines indicate zero baseline. (F), Computed QT c values measured in guinea pigs fed a PA-diet or LFD. Data are presented as mean ± SEMs. Scale bar: 0.5 mV × 400 ms. * Statistical significance at P < 0.05.
We next determined whether IL-6 + IL-6R exerted an inhibitory effect on I Ks . Figure 6E shows typical whole-cell ventricular I Ks current traces measured in control or untreated adult guinea pig ventricular myocytes. Compared to controls, IL-6 ( Figure 6F) or IL-6 + IL-6R ( Figure 6G) significantly reduced currents at positive potentials to +10 mV ( Figure 6H). Compared to averaged control values (32.8 ± 3.78 pA/pF, n = 7, Figure 6H) measured at +100 mV, I Ks current densities were reduced by 35.1% (or to 21.3 ± 3.00 pA/pF, n = 5, Figure 6H, * P < 0.05) in the presence of IL-6 alone, and by 45.2% (17.98 ± 2.19 pA/pF, n = 7, * P < 0.05) with IL-6 + IL-6R. To determine whether the effects of IL-6 and IL-6 + IL-6R on I Kr and I Ks would alter ventricular electrophysiology, we evaluated a human ventricular cardiomyocyte O'Hara model [18], with modifications [19,20] to incorporate our experimental data for IL-6 alone and IL-6 + IL-6R effects on I Ks (Figure 6H), and previously reported data [9] for I Kr (Table 2). Computer simulation results are summarized in Table 3 and Figure 7. The depression by IL-6 and IL-6 + IL-6R of I Kr densities increased APD 90 by ∆37% and ∆52% (Table 3), respectively, while I Ks inhibition led to a 3% and 4% increase, respectively (Table 3). Moreover, incorporation of the combined inhibitory IL-6 and IL-6 + IL-6R effects on I Kr and I Ks led to a more pronounced effect and increased ∆APD 90 to 45% and 65% (Table 4), for IL-6 and IL-6 + IL-6R, respectively, demonstrating an additive effect of IL-6 on I Kr and I Ks . The pseudo-ECG measured in strand simulations further revealed the implications on QT interval, and of the regional differences and pathological APD prolonging effect of IL-6 alone or in combination with the IL-6R ( Figure 7B). As shown in Figure 7C, QT interval is increased by +36% and +53% (Table 4), for IL-6 and IL-6 + IL-6R, respectively, consistent with the signature high risk proarrhythmic effect that underlie fatal ventricular arrhythmias in patients.   Finally, we examined the effect of IL-6 on cardiomyocyte metabolism. The heart is critically dependent on OXPHOS, and ion channels are ATP-sensitive and exist in a high density in the sarcolemma [21]. We used the Seahorse XFe56 analyzer Cell Mito Stress Assay to measure the mitochondrial oxygen consumption rate profile in human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) acutely (2 h), exposed to IL-6 alone or IL-6 + IL-6R ( Figure 8A). Pre-exposure to IL-6 alone and IL-6 + IL-6R impaired mitochondrial function ( Figure 8B), and severely blunted ATP production, basal and maximal respiration, and spare respiratory capacity ( Figure 8C), consistent with the ability of cytokines to modulate mitochondrial metabolic function [22][23][24], suggests a role for pro-inflammatory cytokines in the pathological effects of lipotoxicity on cardiac cell metabolism. Additionally, proton leak was unchanged, suggesting that the mitochondria is not damaged. , Human-induced pluripotent stem-cell-derived cardiomyocytes (hiPSC-CMs) are exposed to IL-6 alone or IL-6 + IL-6R and mitochondrial stress parameters were analyzed after 2 h (B), IL-6-or IL-R + IL-6R-treated hiPSC-CMs displayed severely depressed oxygen consumption rate (or OCR), compared to untreated controls, indicating impaired metabolism. (C), hiPSC-CMs pre-exposed to IL-6 and IL-6 + IL-6R displayed statistically significant depression in ATP production, basal respiration, and maximal respiration when compared to untreated controls. Nonmitochondrial oxygen consumption respiration and spare respiratory capacity were also severely blunted. Proton leak remained essentially unchanged. Data are presented as mean ± SEMs. * Results with P < 0.05 were considered statistically significant.

Discussion
Diet related lipotoxicity is a critical contributor to arrhythmias in patients with obesity and related pathologies; however, the underlying molecular mechanisms are unknown. The goal of the present study is to investigate the role for hyperinflammation as a key event in lipotoxicity-induced ventricular electrophysiology remodeling in guinea pig. PA induced significant secretions of distinct pro-inflammatory cytokines (IL-6, IL-1β, IL-1α, and INF-β), in RAW264.7 macrophage cells. PA had a more potent effect on IL-6 production which, in turn, exerted a profound and adverse modulation (APD prolongation, triggered EADs and spontaneous beats, and inhibition of I Kr and I Ks densities) of ventricular electrical activity and function. Short-term PA-diet feeding in guinea pig separates hyperlipidemia from significant weight gain and hyperglycemia, induces QT prolongation and promotes cardiac arrythmias (PVCs, TdP and VT). Our data suggests that lipotoxicity-dependent IL-6-production and the subsequent inhibition of I Kr and I Ks underlies adverse ventricular electrical remodeling prior to fatal ventricular arrhythmias acquired in obesity and associated disease pathologies.
Macrophages are associated with heightened immune responses to infectious pathogens and tissue damage in obesity [25] and diabetes [26,27]. In metabolic disorders, the enlargement of adipocytes triggers the generation of saturated FFAs through lipolysis [28]. Chronic accumulation of lipid droplets within the myocardium is associated with increased secretion of pro-inflammatory cytokines (IL-6, TNFα, IL-1β) [7,8] and ventricular electrical dysfunction [8].
In cardiomyocytes, repolarizing I K (I Kr and I Ks ) function is an important contributor to mechanisms (APD prolongation) [9,29] that predispose to ventricular arrhythmias. Therefore, pathological depression of I K would be expected to delay repolarization and contribute to ventricular arrythmias. The potential role for impaired ion channel mechanisms in inflammation-related arrhythmias is highlighted by previous studies in patients with inflammatory disorders [30][31][32].
The electrophysiological effects of cytokines on some voltage-gated channels, including I Kr , have previously been studied by us [9] and others [10,11,[33][34][35][36]. However, a paucity of studies in the extant literature has directly assessed the modulation of I Ks in the heart by pathological cytokine levels. As shown in this study, IL-6 alone or in combination with the IL-6R inhibited I Ks density, and this inhibition increased vulnerability to pathological QT prolongation (Figure 7). Our finding suggests that combined prevention of pathological inhibition of I Kr and I Ks , particularly in lipotoxicity-induced hyperinflammation has the potential to be anti-arrhythmic by shortening APD and reducing fatal ventricular arrhythmias ( Figure 7) in patients that show vulnerability to cardiac events. Therefore, it would also be valuable to explore the link between IL-6 + IL-6R, I Ks and ventricular arrhythmias.
In this study, the β-adrenergic receptor agonist ISO increased the number of PVCs, induced TdP and ventricular tachycardia. Furthermore, IL-6 + IL-6R decreased mitochondrial metabolism and, more importantly, ATP production in hiPSC-CMs. Our data suggest that IL-6 may depress I Ks channel function by limiting available ATP, critical for PKAdependent phosphorylation of I Ks channel (KCNQ1) subunits, activation of I Ks [37,38], and the regulation by β-adrenergic stimulation. This mechanistic insight is likely to have important implications for predicting the effects of pathological changes in cytokines on I Ks function, and cardiac repolarization especially during exertion.
In conclusion, our data imply that a cytokine-mediated decrease in I Ks may contribute to high-risk QT prolongation that underlie life-threatening cardiac events. It is intriguing to speculate that dietary interventions that either prevent (1) cardiac and systemic lipid accumulation (anti-lipotoxic drugs promote lipid storage in adipose stores or decrease systemic lipid levels); (2) pathological decreases in I Kr and I Ks channel function (cellular mediators that enhance channel opening) may be anti-arrhythmic, and therefore beneficial to obese patients that display vulnerability to arrhythmogenesis and fatal arrhythmias.

Guinea Pig Ventricular Myocytes
Primary ventricular myocytes were isolated and cultured as previously described [39,40]. Briefly, adult male and female Hartley guinea pig hearts were excised and Langendorff perfused with Tyrode solution containing (in mM): 118 NaCl, 4.8 KCl, 1 CaCl 2 , 10 Glucose, 1.25 MgSO 4 , 1.25 K 2 HPO 4 (pH = 7.4) for 5 min. Ventricular myocytes were isolated by enzymatic digestion in Ca 2+ -free Tyrode solution containing Collagenase B (final concentration, 0.6 mg/mL; Boehringer Mannheim, Indianapolis, IN, USA) for an additional 6 min. The heart was subsequently perfused with high-K solution containing (in mM): 70 KOH, 50 L-glutamic acid (potassium salt), 40 KCl, 10 Taurine, 2 MgCl 2 , 10 Glucose, 10 HEPES, 5 EGTA, and 1% albumin (pH 7.4, with KOH) for 10 min. The digested heart tissue was placed in fresh high-K solution, minced into smaller pieces and triturated several times to dissociate the cells. The cell suspension was filtered through a mesh strainer and allowed to settle for 15-20 min. The pellet was resuspended in 10% M199 media and plated on laminin-coated coverslips. Cells were patched 6-8 h after plating.

Preparation of Bovine Serum Albumin Conjugated FFA Solutions
Palmitic acid (PA) stock solution was prepared as previously described [14].

Low-Fat Diet and Palmitic-Acid Diet Feeding in Guinea Pig
Guinea pigs (male/female; 200-250 g) were purchased from Charles River Laboratories (Wilmington, MA, USA). Control guinea pigs were fed, ad libitum, a low-fat diet (LFD, Research Diets Inc., New Brunswick, NJ, USA) containing (in kcal%): 10 fat, 70 carbohydrates, 20 protein, and 2300 corn starch. PA diet group was fed diet (in which most of the soybean, was replaced with 315 kcal% palm oil), containing 10% of its kcal from fat, 70% from carbohydrates, and 20% from protein. The PA rich diet contained saturated and unsaturated free fatty acids (FFA), which provided 48.4 and 36.8% of the fat-derived calories. Guinea pigs were fed LFD or palm oil rich diet for a duration of 50 days (~7 weeks), while monitoring temporal changes in weight and blood glucose every 10 days. Blood was taken from the paw and analyzed using an Accucheck glucometer (Roche, Indianapolis, IN, USA).

Preparation of hiPSC-CMs
The hiPSC-CMs (iCell) were purchased from Cellular Dynamics International (CDI cells, Madison, WI, USA). Cells were maintained in media according to the manufacturer's protocol and instructions. For Seahorse XF Cell Mito Stress Test assay, single hiPSC-CMs were plated onto 96 well Agilent Seahorse XF Cell Culture Microplate, and allowed to recover at 37 • C in a CO 2 incubator for about 10 days before recordings. Agilent Seahorse XF Cell Mito Stress Test was applied to hiPSC-CMs and oxygen consumption rate (OCR) was measured as a function of time according the manufacturer's instructions. Cells we exposed to IL-6 (25 ng/mL) alone or in combination with IL-6R (25 ng/mL) for 2 h and six mitochondrial respiration parameters were determined: nonmitochondrial oxygen consumption, respiration, basal respiration, maximal respiration, proton leak, ATP production, and spare respiratory capacity.

Electrophysiology
Whole-cell membrane currents were recorded in cardiomyocytes using an EPC-10 patch clamp amplifier (HEKA Electronics) controlled by PatchMaster software (HEKA) [17,41], or Axopatch-200B amplifier (Axon Instruments, Inc., Burlingame, CA, USA) [9]. Pipette resistance was typically 1.5-2 MΩ when filled with internal solution. I-V curves were generated from a family of step depolarizations (−40 to +100 mV in 10-mV steps for 2 s from a holding potential of −50 mV), followed by a repolarizing step to −40 mV for 2 s to obtain tail currents. Currents were sampled at 20 kHz and filtered at 5 or 10 kHz. Traces were acquired at a repetition interval of 10 s. Cell capacitance or cell size (in pF) was compensated and measured using the built-in compensation unit of the amplifier. Action potential waveforms were continuously recorded from ventricular myocytes in current clamp mode by passing depolarizing currents for 20 ms at a subthreshold (1.5×) intensity in 10 s interval for~2 min. Ventricular myocytes were pre-treated (2 h) with IL-6 (20 ng/mL) alone or in combination with IL-6R (25 ng/mL) before experimentation.

Electrocardiogram (ECG)
Surface ECG was recorded using a Dual Animal BioAmp amplifier PowerLab (LabChart 8/s, AD instruments, Colorado Springs, CO, USA), and analysis system (LabChart v8.1.2, AD instruments, Colorado Springs, CO, USA). Guinea pigs were placed on a warm pad and subjected to anesthetic inhalation, using a table-top isoflurane (3-4%) vaporizer (Harvard Apparatus, Holliston, MA, USA). After the guinea pig was slightly anesthetized, it was removed, and a mask was used to maintain anesthesia with 1-2% isoflurane (mix of isoflurane and 700 mL O 2 /minute). Anesthesia depth from isoflurane was monitored by respiratory rate and toe pinch response. Electrodes were positioned on the sole of each guinea pig foot. Electrical signals were recorded for 1 min at 1200 Hz, stored on a computer hard disk and analyzed off-line using the average of five representative consecutive beats. Tracings were analyzed for QT c interval, heart rate, PR interval, and QRS duration. QT c interval was calculated by Bazett's formula where QT c = QT/ √ RR.

Telemetry
Radiotelemetry ECG guinea pig pressure and biopotential telemeter (SN:11934) transmitters (TR54PB; ZC Analytics, Denver, CO, USA) were implanted under isoflurane anesthesia (2-3% isoflurane in 100% oxygen at 1 L/min). To reduce procedure induced inflammation all animals were given buprenorphine-HCl (0.1 mg/kg, SQ) and carprofen (10 mg/kg/day, SQ) a 1/2 h prior to the time of anesthesia induction. The probes were placed in the abdominal cavity and core body temperature was maintained at 37 • C throughout the procedure, using a heating blanket. The telemetry probes were placed on top of the exposed intestines and the two ECG leads were inserted through the abdominal wall muscles and subcutaneously tunneled in a lead II configuration. The quality of ECG signal and parameters were evaluated during the procedure. Once the quality of the ECG signal was satisfactory the distal ends of the leads were secured to the underlying tissue and the skin incision was closed with 5-0 vicryl and sealed. After the procedure, the guinea pigs were allowed to recover at 37 • C and were monitored until they regained their toe-pinch reflex,~5-10 min after the end of the anesthesia. Guinea pigs were allowed to recover for 24 h post-surgery before experimentation.

Computer Simulation
We used computer simulations to study the effect of IL-6 alone or in combination with IL-6R in human ventricular myocytes. For this purpose, we used the O'Hara et al. model (ORd) [18], with the modifications proposed by Mora et al. in the fast sodium [19] current and by Dutta et al. in the conductance of several ionic currents [20], to simulate the membrane electrical activity of human ventricular myocytes in basal conditions. The conductance of I Kr and I Ks were modified based on the changes measured in experiments with guinea pig ventricular myocytes pre-treated with IL-6 alone or a combination of IL-6+IL-6R. Single cell and heterogeneous 1D strand simulations, as well as pseudo-ECG recordings, were performed by using the ELVIRA software [42], with a time step of 0.01 ms. Unicellular models were stabilized for 60 min and an additional stabilization of 100 pulses was carried out in strand simulations, at a basic cycle length of 1000 ms. The heterogeneous 1D strand was comprised of 165 cells distributed in 60 endocardial, 45 midmyocardial and 60 epicardial cells, similar to prior simulation studies [18,43], and the action potential (AP) propagation was solved by using the monodomain formalism. Pseudo-ECGs were computed using the large volume conductor approximation [44], as a result of the propagation of the AP from the endocardium to the epicardium region in the strand, and were calculated at a virtual electrode located 2 cm from the epicardium [18,44]. The QT interval in the pseudo-ECG was defined as the time between QRS onset and the end of the T wave (cross by 0).

Data and Statistical Analyses
The conditions of each individual group of experiments were blinded to experimenter. Electrophysiological data were analyzed off-line using built in functions in Fitmaster (HEKA), Clampfit and Origin software. Current amplitudes (in pA) were divided by cell size (in pF) and expressed as current densities (pA/pF). Data are reported as means ± S.E.M. Statistical differences were determined using one-way ANOVA with Bonferroni post-hoc analysis or two-tailed unpaired t test for comparisons between groups and considered significant at P < 0.05.