Poincaré plot analysis of ECG uncovers beneficial effects of omaveloxolone in a mouse model of Friedreich’s ataxia

Arrhythmias

In FA patients, tachycardia and bradycardia arrhythmias manifest due to delayed atrial depolarization and left-ventricular repolarization events.⁶ T-wave abnormalities are the most frequently observed ECG feature, indicating a defect in repolarization.⁵ Table 1 summarizes limited published findings of ECG abnormalities in FA patients, which include Long QT syndrome (LQTS), interatrial block, ST abnormalities, atrial and ventricular ectopy, and deviations in the left and right axes.^7–13 At the mitochondrial and molecular level, insufficient ATP production leads to a cytosolic and metabolic stress.^14,15 In addition, increased production of ROS may impair cardiac excitability by affecting the expression and function of various ion channels and transporters. Consequently, decreased mitochondrial function in FA could contribute to arrhythmias by disrupting the process of excitation-contraction coupling, disturbing intracellular ion homeostasis and membrane excitability. These issues are further complicated by the inherent differences in electrical responses between males and females.¹⁶ The exact mechanisms underlying cardiomyopathy and arrhythmias in FA patients remain poorly understood. Therefore, it is critical to thoroughly interpret fundamental electrical signal propagation in a mouse model that best recapitulates late-stage cardiomyopathy in FA patients to identify and implement appropriate therapeutic interventions.

Omaveloxolone (Skyclarys)

Omaveloxolone (OMAV) was recently approved by the Food and Drug Administration under commercial name Skyclarys as the first treatment option for FA patients. OMAV induces a network of protective antioxidant responsive element proteins regulated by transcription factor, nuclear factor erythroid 2-related factor 2 (Nrf2).¹⁷ FA mouse models showed downstream targets of Nrf2 decreased,^18,19 impacting biosynthesis of glutathione and other transcriptional targets that counteract oxidative and inflammatory stressors. While OMAV showed improved neurological mFARS testing in clinical trials,^20–22 information of electro-cardiac response to OMAV is still lacking.

Animal model and OMAV Treatment

Mouse strains were obtained from The Jackson Laboratory and bred in-house. Strain #028520 (Fxn^flox floxed exon 2) and strain #029100 (Fxn^null::MCK-Cre) were bred to obtain experimental FXN⁻/^flox x MCK-Cre^hemi (FXN-cKO) mice. FXN^+/flox littermates were used as control animals. FXN-cKO mice were treated with 24 mg/kg of OMAV. Due to early mortality in FXN-cKO mice, experiments were performed at 8-weeks of age in control and experimental groups. Mice were housed in a temperature- and humidity-controlled animal facility with a 12-hour light-dark cycle.

Electrocardiogram (ECG)

Mice were placed in a supine position on a 37°C heat pad and Lead I surface electrocardiogram (ECG) was recorded in control, FXN-cKO, and OMAV-treated FXN-cKO mice of either sex under anesthesia with ketamine/xylazine mix (0.01 mL/g IP). A blinded observer processed five and ten-minute recordings using Labchart 8 (ADInstruments) for unbiased annotations. Heart rate variability (HRV) was calculated by a single beat through R-to-R labeling using Labchart 8 algorithms (version 8.1.21 software).

Interpretation ECG Time Intervals

Averaged waveforms: P, PR, QRS, corrected QT (QTc), and RR intervals were calculated as averaged mean values obtained from a 5-minute ECG recording in control, FXN-cKO and OMAV-treated FXN-cKO mice of either sex. Data are shown as mean ± standard deviation. Aggregated frequency distributions: P, QTc, and RR intervals were acquired and analyzed by individual beats or averaged four waves/beat depending on the quality of the signal. A bin size of 1 or 2 was used to reflect the total beats at a specific time interval.

Heart Rate Variability (HRV)

Heart rate variability (HRV) is the measured differences in time intervals between adjacent heartbeats.²³ HRV parameters were used to assess ANS activity in FA mice with and without OMAV treatment. Time parameters examined were the standard deviation of all normal RR intervals (SDRR), the square root of the mean squared between successive differences RR intervals (RMSSD), and the percentage of RR beats longer than 6 ms (pRR6). Frequency parameters examined were low frequency (LF), high frequency (HF) and the LF:HF ratio. The Poincaré plot was utilized to determine the distribution of identities on a scatter plot with R,R_i+1 on the x-axis and R_i+1,R_i+n on the y-axis. Standard descriptors 1 (SD1) and 2 (SD2) indicate the short-term and long-term variability²³, respectively. The oscillation of the signal was calculated as S (area) values.

Western blotting

Total protein was extracted from flash-frozen heart tissues using CellLytic MT Cell Lysis Reagent (Sigma Aldrich, St Louis, MO, USA) supplemented with PhosSTOP phosphatase inhibitor and Complete EDTA-free mini protease inhibitor cocktail tablets (Roche Diagnostics, Indianapolis, IN, USA). Steel beads were added to the samples, then shaken for 30 min at 4°C before being centrifuged at 5,000 rpm for 10 min. Protein concentration in the supernatant was measured using the Pierce BCA Assay (ThermoFisher Scientific Waltham, MA, USA). For western blotting, samples were either detected using Chemi-doc image analyzer by Protein Simple (Biotechne, Minneapolis, MN, USA) or samples were placed onto NuPAGE 4–12% Bis-Tris Gels in NupaGE LDS sample buffer (Invitrogen Inc, MA, USA), transferred to nitrocellulose membranes (BioRad Laboratories, Hercules, CA, USA), imaged on OdysseyCLx imager, and analyzed with Image Studio. Antibodies used: FXN (Abcam, #ab175402) NRK2 (Amsbio, #AMS.AP2791c); GAPDH (Abcam, #ab8245).

Quantitative real-time PCR

RNA was extracted from frozen heart samples using RNeasy Fibrous Tissue Kit (Qiagen, Hilden, Germany) according to manufacturer’s instructions. cDNA was created using iScript cDNA Synthesis kit (BioRad Laboratories, Hercules, CA, USA). Quantitative RT-PCR was performed using PowerUp Syber Green (ThermoFisher Scientific, Walthan, MA, USA) on a QuantStudio 7 Pro Real-Time PCR System (ThermoFisher, Waltham, MA, USA). Data was normalized with ribosomal protein 18S (Rps18) using the delta-delta CT calculation. PCR Primers can be found in Supplemental Table 1.

Histology

Short axis heart cross sections were fixed in 10% neutral buffered formalin, transferred to 70% ethanol, embedded in paraffin, and sliced into 4 μm slices at the UC Davis VMTH Histology Lab. Samples were stained with Picrosirius red to determine cardiac fibrosis. For the total percentage of cardiac fibrosis, we calculated the total percentage of fibrotic area on a montage 10x image from each sample excluding all perivascular fibrosis areas. Imaging was performed under Cytation5, Gen5 software (Agilent, Santa Clara, CA, USA).

Statistical Analysis

All data were analyzed using student’s two-way t-test, one-way ANOVA with Tukey’s multiple comparisons test using Prism version 10.0 (Graph-Pad Software Inc., San Diego, CA) and p<0.05 was considered statistically significant. All data are shown as mean ± SEM.

Distinct ECG characteristics in human and mouse models

To compare human and mouse ECG abnormalities, we analyzed P-wave, PR segment, QRS complex, RR, and QT intervals (Figure 1A). Waveforms of electrical propagation in the mouse heart can be illustrated in the flow direction (arrows) as depolarized (yellow) and/or repolarized (blue) state in a mouse heart (Figure 1B). As shown in Figure 1B, atrial depolarization, initiated by the sinoatrial (SA) node (pacemaker), causes the P wave. When atrial depolarization is complete, the impulses pause for ~0.1 s at the atrioventricular (AV) node, and then propagate to the atrioventricular (AV) bundle which connects the atria to the ventricles. The bundle branches conduct the impulses through the interventricular septum. The Purkinje fibers depolarize the contractile cells of both ventricles. Ventricular depolarization begins at apex, causing the QRS complex. At this time, atrial repolarization also occurs, which is typically obscured by the large QRS segment and is not visible on ECG. However, the duration of atrial repolarization could be indirectly judged from the PR segment¹⁰ because this segment includes both atrial depolarization and repolarization. When ventricular depolarization is complete, ventricular repolarization begins at apex, causing the T wave. In mice, the morphology of the QRS complex is different from that in humans²⁴ and the S wave is immediately followed by a distinct J wave,^{24, 25} which is associated with early repolarization of ventricles due to extremely high heart rates as compared to humans. The J wave may also exist in humans, but it is barely visible on ECG.^24,25 The variations in the heart rates on ECG are typically reflected by changes in RR intervals.

Figures 1C and 1D show ECGs of FXN-cKO male and female mice. Normal ECG recordings of CTRL mice (black) are shown with a normal sinus rhythm. FXN-cKO mice (red) exhibit extended RR intervals and Torsade de Pointe (TdP) short cycles.²⁶ TdP are features on the ECG that indicate ventricular tachycardias and are typically associated with long QT intervals. Moments before TdP, a short-long sequence of twisting amplitudes around the isoelectric line occurs. Extended RR intervals were more common in FXN-cKO males (Figure 1C), while short-long cycle TdP occurred across both sexes but shown here in female FXN-cKO mice (Figure 1D). These simple observations in Lead I reveal significant electrical propagation abnormalities related to FA. However, some of these changes were not observed with traditional measurement of average time intervals (Figure 1E – 1I), necessitating further investigation when considering variables like sex that can significantly influence findings.

Sexual differences of structural and LQTS channel susceptibility in FXN-cKO mice

As seen in Figure 2A, the FXN-cKO model used in this study involves the excision of exon 2, differing from the alternate FXN-cKO model which excises exon 4 (FXN-cKO^ex4).²⁷ Both models result in loss of the FXN protein, significantly affecting both sexes. As seen in Figure 2B, FXN protein expression was significantly decreased or completely lost (p<0.0001) in both males (p<0.0002) and females (p<0.0001) in FXN-cKO animals compared to CTRL. Using the FXN-cKO^ex4 model, Huang et al.²⁸ found that nicotinamide riboside kinase 2 (NRK2) protein expression, a critical marker of cardiac structural remodeling²⁹, was significantly increased in a grouped sex cohort. Our results show NRK2 was significantly increased in both males (p<0.0003) and females (p<0.0015) compared to CTRL (Figure 2C) but more severely in FXN-cKO males (p<0.0048). NRK2 encoded by gene Nmrk2 is a muscle-specific β1-integrin binding protein which typically has a very low expression in the healthy cardiac muscle, however its expression increased dramatically during cardiac response to stress.³⁰ Emerging evidence indicates that NRK2 regulates cardiac remodeling and failure under stress conditions not only in mice but also in human patients with heart failure.³¹

We hypothesize that loss of FXN could lead to upregulation of NRK2 (Figure 2D), possibly resulting in changes in ion channel expression involved in development of the long QT syndrome (LQTS). LQTS can result from ion imbalances—Kv7.1 and Kv11.1 are potassium voltage-gated channel subfamily Q member 1 (Kcnq1) and potassium voltage-gated channel subfamily H member 2 (Kcnh2) loss-of-function, respectively, while Nav1.5 involves sodium voltage-gated channel alpha subunit 5 (Scn5a). Indeed, FXN-cKO males Kcnq1 gene expression was significantly decreased (p<0.002), with FXN-cKO females showing a non-significant trend, yet a significant sex difference (p<0.008) in females vs FXN-cKO males (Figure 2E). Kcnh2 gene expression was decreased in both male (p<0.007) and female FXN-cKO mice (p<0.0007) (Figure 2F). However, Scn5a expression was decreased only in FXN-cKO males (p<0.01) (Figure 2G). LQTS in mice differs from human-related genes, with potassium channels being the primary cause in mice³². As shown in Figure 2H, potassium voltage-gated subfamily D member 2 (Kcnd2) gene expression was significantly decreased in both male and female FXN-cKO mice, with a greater reduction in males (p<0.002). Similarly, potassium voltage-gated integrating protein 2 (Kcnip2) was significantly reduced in both sexes, but more so in females (p<0.003). Although not significant, downward trends were observed in other potassium channels known to exhibit LQTS and arrhythmias in mice (Supplemental Figure 1A–D).

These findings highlight that several channels critical for ventricular repolarization are affected in these mice, with the most significant impact on potassium channels. Importantly, the accessory subunit potassium voltage-gated channel subfamily E (KCNE) family and key regulators of Kv7.1 channels, also appear to exhibit sex-specific alterations^33,34, warranting further investigation. Overall, these results indicate that the loss of FXN can result in LQTS, most probably due to structural changes in the heart affecting ion channel expression. Additionally, a pronounced sexual dimorphism was observed and linked to the NRK2 that triggers markers of structural remodeling. Consequently, these alterations impact ion channels, potentially disrupting the electrical conductance in these animals.

Sexual differences of electrical propagation in FXN-cKO mice treated with OMAV

The effects of OMAV on electrical propagation in the heart are unknown. OMAV, Nrf2 signaling pathway inducer, activates Nrf2 by uncapping it from the Keap1 protein as illustrated in Figure 3A. Nrf2 is then able to translocate to the nucleus and transcribe the antioxidant response element (ARE) genes, thereby causing a beneficial effect on the body against oxidative and redox stress.²⁰

To examine electrical activity in the heart, 14 control (gray), 12 FXN-cKO (red) and 13 OMAV-treated FXN-cKO (green) mice were used. Five-minute averaged mean and standard deviation of time intervals from male and female mice were plotted on a bar graph with their respective waveform (Figures 3B–C). Male and female mice showed distinct differences in waveforms and time intervals lengths. FXN-cKO males had significantly increased intervals for QTc (p<0.05) and QRS (p<0.05) segments (Figure 3B), while females showed increased intervals for QRS (p<0.001) and P (p<0.05) only (Figure 3C) indicating an existing sexual dimorphism in these mice. OMAV did not exert any significant effects on averaged ECG parameters (Figure 3B and 3C), however OMAV significantly shortened PR interval (p<0.05) in FXN-cKO females (Figure 3C). These data indicate that FXN-cKO males exhibit problems primarily with early and late ventricular repolarization while FXN-cKO females also experience problems with atrial depolarization in addition to the late ventricular repolarization.

We next determined that aggregated frequency distribution for P-wave (atrial depolarization), QTc (ventricular repolarization) and RR (heartbeat) intervals revealed multi-peak variation in FXN-cKO mice (red) and a shift away from the optimum peak (yellow-shaded box) as compared to control hearts (black) (Figure 3D–F). We found that while there were no changes in the mean value of P-wave duration in FXN-cKO males (Figure 3B), the frequency distribution histogram analysis revealed multiple peaks in P-wave interval variation in FXN-cKO males (red) away from the optimum peak (yellow-shaded box) observed in control mice (black). OMAV treatment (green) decreased the number of events located outside the optimum peak values (Figure 3D, top panel), rescuing the P-wave from a late depolarization effect. A shift to the right on P-wave histogram was noted in FXN-cKO females (red) as compared to control mice (black) indicating an increase in P-wave duration (Figure 3D, bottom panel). OMAV (green) had a dual effect on P-wave histogram distribution in FXN-cKO females: increasing the number of events in the optimal 15 ms range but also increasing the number of events in the 22 ms range outside of the optimal range.

Similarly, the QTc interval histograms exhibited multi-peak variations (red) from the optimum peak range (black) in both FXN-cKO males and females (Figure 3E). OMAV treatment efficiently shifted the majority of QTc events back toward the optimal interval of ~150 ms in FXN-cKO males (top panel, green), however two distinct peaks in QTc intervals were observed in OMAV-treated females (bottom panel, green).

Heartbeat (RR variation) frequency distribution histograms (Figure 3F) revealed that the vast majority of RR intervals in FXN-cKO males was shifted to the left (top panel, red) indicating the prevalence of tachycardia (or fast heartbeats) in FXN-cKO males. The number of short heartbeats (RR intervals) was decreased in OMAV-treated FXN-cKO males while the RR events above 350 ms range were not affected. In contrast, no major peaks were identified on RR histograms in FXN-cKO females (bottom panel, red) with RR intervals spreading between 147 and 448 ms. Altogether, these data indicate a significant sex-specific disturbances in ECG signals and sex-specific response to OMAV treatment in FXN-cKO mice.

OMAV enhances early repolarization by restoring J-wave in FXN-cKO males via Kcnd3 and Atp1a1 gene expression

As we noted above (Figure 1A), mouse ECG differs from human by clear appearance of J-wave on ECG tracings. Figure 4A shows the J-wave was absent in 80% FXN-cKO male mice and reduced in females, but OMAV treatment restored the J-wave in 60% of FXN-cKO males. In contrast, females showed no significant change with OMAV treatment. Normal sinus rhythm was observed (Figure 4B), suggesting OMAV’s potential to enhance repolarization abnormalities, possibly by influencing transient outward potassium ion currents (I_to). In Figure 4C, gene expression indicated that Kcnd3, which is responsible for the early transient repolarization I_to channel, was significantly reduced (p<0.0077) in FXN-cKO males, while females maintained their gene expression levels. Notably, males and females exhibited significantly opposing trends (p<0.0001), with FXN-cKO females showing an upward trend. Additionally, OMAV-treated female mice displayed a nearly significant (p<0.0523) increase compared to FXN-cKO females. Importantly, a significant increase in Kcnd3 expression (p<0.0001) was found in OMAV-treated FXN-cKO females compared to FXN-cKO treated males. In Figure 4D, the gene expression of Atp1a1, which encodes Na⁺/K⁺ ATPase, revealed contrasting trends between FXN-cKO and CTRL groups, with Atp1a1 expression levels increasing in FXN-cKO males and decreasing in FXN-cKO females (both non-significant). However, a significant difference was found between FXN-cKO male and female mice (p<0.001) when comparing Atp1a1 gene expression. Although both sexes of the OMAV-treated FXN-cKO mice showed a downward trend, a significant difference remained between OMAV-treated FXN-cKO males and OMAV-treated FXN-cKO females (p<0.001).

The J-wave’s return in FXN-cKO mice suggests OMAV’s role in mitigating ST-depression, although it may exacerbate ST-elevation in FA patients (Figure 4E). Structural changes have the capacity to change ion-channel properties (see Figure 2). Therefore, we performed Picrosirius red staining on cardiac tissue of OMAV-treated and untreated FXN-cKO and CTRL animals of either sex to visualize fibrotic areas (Figure 4F). We found that the percent of the fibrotic area was significantly increased 75.3% in FXN-cKO male mice compared to CTRL male mice and FXN-cKO female mice showed a 61% increase compared CTRL female mice (Figure 4G). OMAV show no protection against fibrotic formation in cardiac tissue in both male and female FXN-cKO groups (Figure 4G). Overall, the data show that despite OMAV had no protection against development of fibrotic tissue, cardiac early repolarization function was improved in FXN-cKO hearts by OMAV. Controversy over the J-wave in mice and whether it translates to human electrical signal propagation still remain²⁴, however, abnormalities in the J-wave should be considered when treating FA mice with therapeutics.

HRV analysis uncovers beneficial effects of OMAV against arrhythmic events

Our HRV analysis of RR intervals over time demonstrated that FXN-cKO mice exhibited significantly increased RMSSD (p<0.05) and SDRR (p<0.05) in the combined sex cohorts as compared to control mice with no change in pRR6 values. No sex differences were noted in HRV time or frequency domain parameters when these combined cohorts were split into male and female groups (Supplemental Figure S2). Notably, OMAV treatment in FXN-cKO mice (green) significantly reduced both SDRR (p<0.05) and RMSSD (p<0.05) scores, indicating its beneficial effect on heart rhythm stability.

Figure 5A shows the differences in high and low variation. In Figure 5B, non-linear Poincaré plot analysis revealed distinct differences between control (black dots) and FXN-cKO animals (red dots). Control mice showed a clustered ellipse of RR intervals within an optimal range (~300 ms) while FXN-cKO mice displayed multi-clustering, indicating arrhythmias. FXN-cKO males shifted clusters left (Figure 5B), suggesting tachycardia, whereas females shifted right (Figure 5C), indicating bradycardia³⁵. Although no sex differences were noted in HRV SD1, SD2, or S domain parameters (Supplemental Figure S2), quantitative measures demonstrated that grouped FXN-cKO mice exhibited increased SD1, SD2 (Figure 5D), and S area (Figure 5E), reflecting varied arrhythmic patterns not evident from mean RR interval analysis.

Importantly, OMAV (green) treatment prevented multi-clustering in FXN-cKO mice (Figure 5B and 5C), reducing grouped SD1, SD2, and the S area, an effect not seen by traditional analysis methods (Figure 5D–5F). This highlights the potential of Poincaré plots as a superior, quantifiable method for assessing ECG activity in conditions with mitochondrial deficits. As shown in Figure 5G, OMAV showed little impact on averaged data but effectively averted disparate arrhythmias. We encourage others to adopt Poincaré plots or other advanced quantitative and qualitative ECG analysis techniques for better insights into therapeutic interventions for treating FA.

Sexual dimorphism exists in FA electrical propagation

The data presented in Figures 2, 3 and 4 reveal the presence of sexual dimorphism in ECG characteristics in FXN-cKO mice. Our group reported a sexual dimorphism in FXN-cKO mice³⁶, which supports our current findings related to electrical propagation. We are the first to report significant sex differences in electrical conductance within the FXN-cKO model. Prior studies on FA mouse models either focused on single sex groups or did not specify the sex of the animals used.^{27, 28}

The FXN-cKO mouse model we used in our study is similar to the FXN-cKO^ex4 model, which involves the removal of exon 4. Huang et al. found that the FXN-cKO^ex4 exhibits a critical stress response that leads to cardiac remodeling initiated by phosphorylated eIF2a activity and downstream markers like NRK2²⁸. In our work, we distinguished our cohorts by sex. We discovered both sexes experienced structural remodeling as indicated by NRK2 expression, with males showing a significantly greater increase compared to CTRL and FXN-cKO females. Furthermore, we discovered that ion channels expression was significantly affected by FXN loss.

Our findings align with clinical data showing that female FA patients are more likely to have normal ECG results (p<0.004) compared to male FA patients. The study by Schadt et al.²⁰ found more pronounced left (27% vs. 13%, p=0.005) and right (23% vs. 5%, p<0.001) ventricular hypertrophy in male FA patients compared to females. Additionally, left axis deviation (LAD) was more frequently observed in the ECGs of male FA patients than in females (8% vs. <1%, p=0.005). LAD indicates a shift in the QRS complex depolarization direction to the left (between 0° and −30°), often resulting from conduction abnormalities such as left bundle branch block, left anterior fascicular block, and left ventricular hypertrophy.

ECG interpretations correlate sex to chamber-specific differences in FA mice

FA research focuses mainly on LV malfunction in FA cohorts in a combined sex cohort.⁷ Women generally experience longer QT intervals than men, and it is suggested that the effect of female hormones on these currents mediates the differences in the duration of the QT interval.³⁷ These findings are similar in mice, where QT durations were also observed to be dissimilar between sex.³⁸ To our surprise, we found that QT interval in FA male mice were more afflicted than female mice under traditional methods (Figures 3B and 3E). Late-stage male mice presented with significant extended QTc interval. FA female mice showed more impairment in atria (P and PR intervals) yet retained their QTc interval length compared to controls (Figure 3C–E). These results indicate that sex is correlated to chamber-specific differences in electrical propagation of FA mice. Indeed, these major differences stresses the importance of evaluating electrophysiology in a sex-stratified manner.

J-wave depression in FXN-cKO mice suggests early repolarization abnormalities in FA

J-wave depression observed in FXN-cKO mice suggests the presence of repolarization abnormalities typical of FA patients. The appearance of the J-wave is linked to early repolarization in mice, indicating that FXN-cKO males experience impairments in both early and delayed ventricular repolarization, whereas FXN-cKO females retain their J-waveform. In FXN-cKO males, the most severe cardiac phenotypes are associated with repolarization deficits, linked to prolonged QT intervals and the disappearance of the J-wave on their ECG readings. Interestingly, FXN-cKO females maintain a reduced J-wave without any treatment, possibly indicating a hormonal effect on electrical conductance³⁹. Treatment with OMAV effectively restored the J-wave phenotype in FXN-cKO males (Figure 4A), reestablishing repolarization homeostasis as shown in the QTc histograms (Figure 3E, top). In humans, the J-point signifies the transition from the QRS complex to the ST-segment near the isoelectric line²⁴. Deviations from this line can lead to J-deflections, ST-elevation, and positive T-waves, which are indicative of early repolarization syndrome and Brugada syndrome. Although there are similarities in ECG propagation between mice and humans, as noted by Boukens et al., the impaired repolarization dynamics in mice do not necessarily reflect the same characteristics in human hearts. However, various idiopathic forms of J-wave appearance have been documented in both mice⁴⁰ and patients with heart failure⁴¹. FA patients often exhibit inverted T-waves, indicating repolarization abnormalities and J-point or ST abnormalities.

The J-wave is mediated by the I_to current, which can heavily influence downstream potassium channels involved in repolarization and other ionic currents such as inward sodium current (I_Na) via voltage-gated sodium channels and calcium current (I_Ca) via L-type Ca²⁺ channels. The Kcnd2 and Kcnd3 gene is responsible for the transient early repolarization and was found to be significantly impacted in FXN-cKO mice, especially with treatment (Figure 4C). OMAV prevents cell death, which helps cells retain their integrity and function, averting mitophagy, thereby maintaining electrophysiological properties.²¹ This protective effect may explain the increased gene expression observed only in female mice. However, excessive Kcnd3 expression may trigger frequent early transient repolarization, potentially causing ion-disruption and leading to a reduced J-wave phenotype in FXN-cKO females. Additionally, gene (Atp1a1) expression of Na⁺/K⁺ ATPase, which is responsible for Na⁺ release by ATPase, was significantly decreased in untreated and OMAV-treated FXN-cKO females. In contrast, OMAV-treated FXN-cKO males showed decreased Atp1a1 expression to levels similar to controls. These results suggest that impaired Kcnd3 and Atp1a1 gene expression is likely responsible for the observed sexual dimorphism in J-wave depression during electrical propagation, however, more investigation into their regulators like KCNE subunits, I_to, and Na⁺/K⁺ ATPase protein expression is needed in mouse and human samples to confirm these claims. Moreover, I_to ionic currents can be influenced by hormonal, autonomic factors, and therapies. In FA, the ST segment often exhibits fluctuating patterns, likely influenced by the autonomic nervous system (ANS).^10,42 It is crucial to characterize J-wave patterns in FXN-cKO mice since ectopic early and late heartbeats can affect subsequent beats and potentially lead to cardiac arrhythmias.¹⁵ Understanding these factors is essential when considering therapeutic options for patients with FA.

HRV reveals autonomic nervous system (ANS) dysfunction in FXN-cKO mice

Cardiomyopathy and arrhythmias are presented as the leading cause of death in FA,⁴³ therefore, the ANS should be investigated to delineate the differences in PNS tone from SNS excitation. HRV analysis is quick and reliable approach to examine the balance between SNS and PNS tone on the heart rate. The SNS effect on heart rate is mediated by epinephrine and norepinephrine, released from sympathetic nerve fibers that act on the sinoatrial and atrioventricular nodes and increase the rate of cardiac contraction leading to facilitating conduction at the atrio-ventricular node. In contrast, parasympathetic mediator acetylcholine, released by the parasympathetic nerve fibers, acts on the sinoatrial and atrioventricular nodes to decrease heart rate. It is important to investigate beat-to-beat variability of the entire ECG signal as opposed to the averaged ECG signal to verify these claims.

OMAV attenuates irregular idiopathic arrhythmias, but tachycardia and bradycardia still remain

We detected abnormal ECG function in FXN-cKO mice and further confirmed ANS trends with OMAV-treatment using Poincaré plots (Figure 5). These late-stage cardiomyopathy FA mice showed major qualitative differences in Poincaré plot analysis, where RR beats showed multiple clusters. In OMAV-treated FA mice, traditional quantitative measures showed no changes (Figure 3). However, Poincaré plot analysis showed that OMAV-treatment mostly exhibited a single ellipse, indicating that OMAV may prevent idiosyncratic arrhythmias from occurring in late-stage hearts. In addition, Poincaré plots for OMAV-treated mice shifted away from the optimal range, indicating that tachycardia or bradycardia arrhythmias still persisted regardless of attenuating arrhythmic activity.³⁵ Further, we show that variability of heartbeats, which is considered an indicator for heart failure,²³ was significantly decreased with OMAV treatment (Figures 5 and Supplemental Figure S2).

The Nrf2 pathway is not sufficiently understood^17–19 and a shift in electrical activity can change the dynamics of the heart rhythm, directly implicating the ANS. Reports on ANS activity in FA patients have been scarce with most publications dating before the millennia. Lees et al.⁴⁴ have summarized them and conveyed the importance for further investigating ANS in FA patients. ANS dysfunction is widely known to contribute to LV dysfunction in cardiomyopathy.⁴⁵ Treatment with OMAV could potentially ameliorate cardiac muscle through Nrf2 expression, while indirectly complicating ANS function. The ANS is therefore imperative to investigate when examining electrical activity and therapeutics in FA.

Poincaré Plot HRV Analysis as an Essential Standard for ECG Interpretation in FA

Traditional ECG interpretation methods have limitations. As illustrated in Figure 3, averaged means and aggregated frequency distributions fail to capture all data from ECGs. These methods struggle with analyzing data from unique animal models that exhibit varied arrhythmias. FA is a mitochondrial disease that affects each cardiac chamber differently, thus, averaged and aggregated data might overlook critical aspects of the heart’s response to FXN loss.

The Poincaré plot is well-documented for quantifying signal variability. It is designed to analyze autonomic nervous system (ANS) function using scatter plots and it offers many advantages.²³ The Poincaré plot analysis has proven to be an invaluable tool in uncovering nuances in ECG data that traditional methods might overlook, particularly in the context of FA. Furthermore, therapeutic responses of Poincaré plot findings are significant, offering a path to tailor treatments like OMAV for more effective management of FA-related cardiac abnormalities. This approach not only improves our understanding of cardiac physiology in disease conditions but also paves the way for more personalized medical interventions. Moving forward, integrating advanced data analysis methods with simple techniques will be crucial in improving our diagnostic and therapeutic strategies for FA.