Introduction

Sudden cardiac death (SCD) in heart failure (HF) is the leading cause of death in the modern world. Current therapies for preventing SCD due to ventricular tachycardia/fibrillation (VT/VF) are limited. For example, implantable cardioverter-defibrillators (ICDs), the only effective therapy for VT/VF and SCD, are palliative, expensive, and pose several risks. Moreover, many patients suffer from VT/VF and SCD in the early stages of HF before they become eligible to receive an ICD^1–9. Antiarrhythmic drugs, which target ion channels, may confer acute benefits but can also be proarrhythmic, triggering VT/VF. Importantly, most antiarrhythmic drugs worsen survival over the longer term^10,11. Thus, there is a pressing need to identify new antiarrhythmic targets and develop new therapies for SCD.

A growing body of evidence indicates that sympathetic stress-induced hyperactivity of the ryanodine receptor 2 (RyR2), located in the sarcoplasmic reticulum (SR) of cardiac myocytes, causes SR Ca²⁺ leak, which contributes to the pathogenesis of VT/VF and SCD^12–15. During systole, the release of Ca²⁺ by the RyR2 plays a pivotal role in myocyte excitation-contraction coupling. The degree of contractility is determined by the amount of Ca²⁺ in the SR and by the timing and magnitude of Ca²⁺ release. In diseased and/or stressed hearts, protein kinase A (PKA)-mediated hyper-phosphorylation and reactive oxygen species (ROS)-mediated oxidation results in a leaky RyR2^10,11, which allows Ca²⁺ to leak out of SR during diastole. The consequent reduction in SR Ca²⁺ load weakens systolic contraction. The consequent rise in cytosolic Ca²⁺ levels also leads to delayed afterdepolarizations (DADs), which can trigger atrial and ventricular tachyarrhythmias^16–19. Recent evidence suggests that dantrolene, a RyR2 inhibitor, reduces the inducibility of atrial fibrillation in animals with myocardial infarction^20–23. The effect of RyR2 inhibition on ventricular arrhythmias and non-ischemic HF has been unexplored.

We hypothesized that RyR2 inhibition of failing arrhythmogenic hearts reduces sarcoplasmic Ca²⁺ leak and repolarization lability, mitigates VT/VF/SCD and improves contractile function. The present study employs a guinea-pig model that recapitulates key features of human non-ischemic HF including spontaneous episodes of VT/VF resulting in SCD^24–26. We examine the effects of chronic dantrolene treatment on the incidence of VT/VF and SCD incidence as non-ischemic HF develops and progresses. In cross-over studies, we determine whether RyR2 inhibition reverses HF and SCD risk. The findings have important implications for repurposing dantrolene, a clinically-used RyR2 inhibitor, as a potential new therapy for HF and SCD.

Results

1. Inhibition of RyR2 leak with dantrolene therapy prevents VT/VF and SCD

To study the effects of RyR2 leak inhibition on SCD and HF, we leveraged the pressure overload guinea pig model of HF/SCD with spontaneous arrhythmias. Unique to this model is the high incidence of premature ventricular contractions (PVCs) and VT/VF, leading to SCD in about two-thirds of the animals within 4 weeks. Notably, the SCD occurs early on, in the first 2 weeks over the course of cardiac hypertrophy, even before the heart dilates or progresses into failure^24,25. Hence, the primary endpoint of our study was SCD, and the secondary endpoint was cardiac contractility measured by echocardiography at four weeks. To determine if Dantrolene sodium (DS) treatment can be used as a potential therapeutic agent, we treated the HF/SCD animals with and without oral DS in the early stages of heart failure (Fig. 1A). HF animals exhibit frequent PVCs that often trigger sustained and non-sustained VT and VF, causing SCD (Fig. 1B). DS treatment not only drastically reduced the rate of PVCs in early and late stages of HF (Fig. 1C), but also improved survival (Fig. 1D). Compared to the HF group, the HF+DS group had a significantly higher rate of survival (55% HF vs 80% in HF+DS, p<0.05).

Figure 1.

2. Dantrolene therapy improves cardiac function and prevents HF

Cardiac structure and function were preserved in HF animal treated with DS as compared to the HF group. In contrast to the HF group, the HF+DS group did not display a significant dilation of the left ventricle (LV) at the 4-week endpoint (Fig. 2A). Echocardiography measures (Fig. 2B) show that the relative wall thickness (RWT), which is a measure of the LV dimensions and is a marker for adverse events in LV dysfunction, was significantly lower in HF group as compared to control (Fig. 2C, p<0.05). Progressive decline of the RWT causes the end systolic volume (ESV) and the end diastolic volume (EDV) to increase and cardiac output to decrease. Reduced contractility and increased afterload both reduce the ejection fraction and increase end-systolic volumes in failing hearts (Fig. 2D-F). Additionally, EDV increased (not shown) without an increase in the stroke volume. To summarize, chronic DS therapy in HF prevented the thinning of the LV walls and the eventual decline in cardiac contractility.

Figure 2.

3. Dantrolene prevents myocardial structural and molecular damage

Analysis of the myocardial structure showed increase in interstitial fibrosis in the HF group compared to the normal controls. Whereas the fibrotic tissue was twofold higher in the untreated HF group compared to normal hearts, the HF+DS group displayed lower fibrotic tissue thereby confirming that DS therapy was delaying the progression of HF, keeping the myocardial architecture intact (Fig. 3A-B). Interstitial fibrosis is a histological hallmark associated with the progression of HF and can causes lethal arrhythmias due to reentry circuits.

Figure 3.

In our previous report, we noted HF-associated changes in both expression and phospho-proteome of key ion channels, transporters, and excitation contraction-coupling proteins. In particular, hyperphosphorylation of several RyR2 sites were noted in failing hearts when compared to normal²⁴(Supplementary Fig S1). A targeted antioxidant therapy prevented these modifications, pointing to oxidative damage as the primary cause. Therefore, we tested if DS therapy prevented the downstream oxidative damages by quantifying the oxidation profile of RyR2 with and without DS. RyR2 oxidation was significantly reduced in HF+DS group as compared to HF group (Fig. 3C).

4. Dantrolene treatment decreases heart rate and improves chronotropic competency in HF/SCD

As per our experimental protocol, all groups of animals received a daily low dosage isoproterenol bolus for one hour. As anticipated, the heart rate was elevated in response to β-adrenergic stimulation and recovered to resting heart rate 3-4 hours later. We assessed ECG parameters at resting heart rate and during post β-AR stress recovery (Fig. 4A). As expected, to adequately respond to the metabolic demand and increased workload, failing hearts fine-tune the resting heart rate. The HF animals had a significant elevation of the resting heart rate (269 ± 10 bpm) as compared to normal controls (251 ± 5, p<0.005). HF+DS animals, however, demonstrated a significantly lower heart rate compared to the HF group, which was similar to that of normal controls (243 ± 5 bpm, p<0.005) (Fig. 4B). Paired analysis of the heart rate for each subject over 4 weeks show that HF animals experienced an increase in resting heart rate by 20-40 bpm. Whereas HF+DS animals experienced a decreased in heart rate by a minimum of 20-30 bpm. The histogram in Figure 4C depicts the distribution of heart rate over a 24-hour period from the normal, HF, and HF+DS groups, demonstrating the decreased heart rate in the HF+DS group.

Figure 4.

Despite increased resting heart rate, the peak response to transient β-AR stress was much lower in HF than HF+DS group, indicative of the therapeutic role of DS in restoring chronotropic competence in failing hearts. Chronotropic competence describes the capacity of the heart to increase its rate in response to increased demand effectively. The average increase in heart rate during transient stress in HF group (96 ± 5 bpm) was significantly lower (p<0.05) than HF+DS group (114 ± 9 bpm). Furthermore, the heart rate variability (HRV, SDNN) was higher in HF+DS group (9.00 ± 0.21) at rest and post stress recovery (Fig. 4D), similar to normal animals as compared to HF (7.28 ± 0.16; p<0.005). DS treatment not only restored the chronotropic response to stress but also reduced the incidences of VT/VF and SCD. To understand this protective mechanism, we next tested the hypothesis that DS prevents VT/VF and SCD by preventing repolarization abnormalities.

5. Dantrolene prevents repolarization abnormalities, normalizes the QT variability, and reduces T-wave heterogeneity

Detailed analysis of the ECG features revealed how inhibition of the RyR2 leak with DS prevents VT/VF and SCD. QT variability and prolongation are known to have prognostic significance in failing hearts. While prolonged QTc is considered a risk factor for arrhythmias, QT heterogeneity as an add-on is a surrogate for lethal VT/VF²⁷. The HF group displayed prolonged QTc, increased QT variability and repolarization abnormalities (Fig. 5A). The guinea pig model of HF exhibits prolonged QTc, as shown in our previous studies^24,25. The QTc of HF+DS group was significantly shorter than HF group and was similar to normal animals at resting heart rate (HF vs HF+DS;175.59 ± 4 vs 155.07 ± 10 msec, p<0.05). DS treatment not only prevented QTc prolongation (Fig. 5B), but also significantly reduced QT variability. Figure 5A also illustrates the temporal dispersion of repolarization by superimposing T waves from a 24- hour period of ECG recording at week four. QT dispersion was reduced in the DS-treated group, thereby imparting greater electrical stability to the heart. The QT variability at resting heart rate and post stress recovery are lower in DS treated HF animals (Fig. 5C-D) (HF vs HF+DS; -0.35 ± 0.02 vs -0.58 ± 0.02, p<0.001). Inhibiting RyR2 with DS reduced the dispersion of repolarization/QT variability, demonstrating a central role of diastolic calcium in increasing cardiac repolarization lability, and terminating triggered activity.

Figure 5.

6. Crossover studies. Dantrolene after HF development mitigates VT/VF and prevent SCD

A crossover study was designed where (i) only vehicle was administered until the endpoint i.e., week 4 (HF) (ii) DS was administered at the onset of aortic constriction until chronic contractile dysfunction is evident in the HF group (usually 3 weeks post banding; HF+/-DS) or (iii) after HF had already developed (3 weeks post banding; HF-/+DS group) (Fig. 6A). HF animals were randomized and were either treated with DS after chronic HF had developed or at the time of aortic banding. In the HF+/-DS group, none of the animals experienced SCD, including the period when DS therapy was stopped (Fig. 6B). Serial echocardiography shows that the FS% in the HF group declines over a period of 4 weeks. DS prevented decline in cardiac function. Interestingly, HF+/-DS group, the fractional shortening does not decline significantly after stopping DS therapy. This could possibly be an effect of the washout period (Fig. 6C FS; pre vs post: 43.35 ± 0.45 vs 53.10 ± 3.16, p: ns). In contrast, the HF-/+DS groups experienced SCD similar to that of HF before DS therapy was started. However, after crossover to DS treatment, no further SCD occurred. Notably, HF-/+DS group, which showed poor cardiac function at week 3 was noted to quickly recover and normalize FS% (Fig. 6C, FS%; pre vs post: 32.43 ± 3.48 vs 43.10 ± 0.87).

Figure 6.

Figure 7.

We next examined the ECG parameters (Fig. 6E-F) in HF-/+DS group at week 1 (post banding, initial stages of HF), Week 3 (chronic HF) and Week 4 (after starting DS therapy). PVC burden significantly increases in the HF-/+DS group pre-DS therapy but declines sharply post therapy. We observed the opposite effect in the HF+/-DS group, where PVC burden increases after DS therapy is stopped (Supplementary Fig. S2). The resting heart rate during chronic HF is significantly higher than pre-HF. After starting DS therapy in chronic failing hearts, we observed a decrease in resting HR to pre-HF levels. Similar trends were seen in the post stress recovery period. Similarly, HRV improved significantly after the start of DS. The QTc interval was shortened after DS treatment as compared to during pre or chronic HF. The QT variability was also reduced after starting DS. Taken together, RyR2 inhibition with DS is effective both in early and late stages of HF, preventing VT/VF and SCD and improving cardiac contractile function.

Discussion

The novel findings of this study are: (1) Inhibition of RyR2 hyperactivity with DS abolishes VT/VF and SCD by normalizing repolarization abnormalities, QTc variability and shortening prolonged QTc in non-ischemic HF; (2) dantrolene improves β-adrenergic signaling and chronotropic competency, thereby improving LV contractility; and (3) inhibition of RyR2 with DS not only prevents the progression of HF but also reverses chronic heart failure. These findings support the idea that the coupling between chronic sympathetic activity and oxidative stress makes RyR2 hyperactive. ROS increases the channel activity by decreasing the threshold for store overload-induced Ca²⁺ release ^28–30. This causes dispersion of calcium transients, reduces repolarization lability leading to spontaneous ectopic beats and VT/VF/SCD in HF^31,32.

Oxidative stress has been centrally implicated in HF. We and others have shown that the ROS dynamics affect the functioning of RyR2^30,33–36. Increased energy demand and workload due to impaired contractile function and downregulation of antioxidant enzymes lead to high ROS accumulation³⁷. This is followed by an increase in use and shortage of reducing equivalents like NADPH/NADH, that further weakens the antioxidant defense^24,37. Our previous studies have confirmed that targeting mROS in failing hearts prevents ventricular fibrillation (VT/VF) and sudden cardiac death (SCD), and reverses heart failure. We see that mROS brings about several changes in protein function including oxidation³⁴, phosphorylation^14,38–41, S-nitrosylation^33,36,42, and remodeling of the contractile machinery. Notably, the RyRs are in proximity to and tightly coupled with the mitochondrial network. They are readily modified by ROS^43–46, via direct oxidation of thiol groups of cysteine residues^44–46 as well as ROS-mediated calcium-independent activation of CaMKII⁴⁷, increasing SR calcium leak^48–51 to promote VT⁵², and increasing the channel activity by decreasing the threshold for store overload-induced Ca²⁺ release ^28–30.

In normal physiology, transient stimulation of the β-adrenergic receptors prompts a positive inotropic response to meet the physiological demands. This includes opening of L-type calcium channels (LTCCs) and the release of cytosolic calcium from the SR, triggering contraction through the activation of ryanodine receptors. Tight regulation of Ca²⁺ release from and reuptake into the sarcoplasmic reticulum is required for proper excitation-contraction coupling. However, in pathology, chronic activation of β- adrenergic receptors is maladaptive, increases ROS which remodels ion channels, modifies proteins, and worsens Ca²⁺ leak from ryanodine receptors; these factors result in EADs or DADs that depose the heart to VT/VF, SCD or injure myocardium; they may also impede contractile machinery. During diastole, the RyR2 still releases a small volume of Ca²⁺ known as the leak. When this leak is continuous, like in failing hearts, the SR Ca²⁺ is depleted and diastolic Ca²⁺ is elevated, which weakens contractile force. At the same time, SR’s capacity to retain the Ca²⁺ is compromised. With the efflux of Na via the NCX, the exchanger imports Ca²⁺, but this influx doesn’t not replenish SR Ca²⁺ load, but rather contributes to the diastolic Ca^{2+ 53–56}. The increased Na load also depolarizes membrane potential and reduces conduction velocity, thus increasing anisotropy^57,58. Additionally, intracellular [Na⁺] accumulation at higher heart rates decreases Ca²⁺ extrusion by NCX, which may now function in its reverse mode^59,60.

In the GP model of non-ischemic HF and SCD, we found that not only SERCA and RyR2 protein expression was downregulated, but the proteins were also hyperphosphorylated. Treatment with an mROS scavenger prevented these alterations from occurring²⁴. Because ROS is a secondary messenger, a systemic antioxidant therapy may suppress cellular signaling. A target like RyR2, which is downstream of ROS and directly involved in Ca²⁺ handling, could be of potential clinical interest. Post-translational modifications by oxidation of the RyR2 during the progression of HF causes leaky RyR2, which decreases SR Ca²⁺ stores and elevated levels of diastolic calcium. Decreased SR Ca²⁺ stores, leads to decreased Ca²⁺ transient amplitude and duration during systole.

Our results show that chronic treatment with dantrolene in HF/SCD animal prevents lethal VT/VF, prevents HF, and delays the onset of cardiac decompensation in pressure overload. The effects of dantrolene on RyR2 were reflected by a decrease in oxidation of the RyR2 at the molecular level, a decrease in PVC burden, and an improvement in survival at the functional level in vivo. Pressure overload-induced structural alterations like increased fibrosis⁶¹. and progressive contractile dysfunction, thinning of ventricular walls and ventricular dilation were successfully ameliorated by chronic dantrolene treatment. Our previous work confirmed that sham controls (sham HF+ isoproterenol challenge in the absence of pressure overload) did not significantly increase fibrosis, although there was a trend towards more^25,26. Although no mechanistic links between SR calcium leak and fibrosis has been explained, an association between reduced SR Ca²⁺ and fibrosis has been previously reported^34,61,62. When dantrolene treatment was started after chronic HF was established, the contractile function and electrical properties of the heart started to recover and reverse.

Comprehensive analysis of the electrophysiological properties of the HF/SCD group shows that with chronic HF decrease in heart rate variability, QT prolongation, increase in dispersion of repolarization. An increase in QT variability is due to the reduction of repolarization lability. Reduction of repolarization lability is associated with increased susceptibility of arrhythmogenesis in patients with prolonged QT and/or structural heart disease. With high cytoplasmic Ca²⁺ load, the inward sarcolemmal NCX current might significantly reduce the function of repolarization lability in calcium overloads and induce early (EADs) and delayed after-depolarization (DADs). Subsequently, the vulnerability to arrhythmias is increased. These local Ca²⁺ handling abnormalities are reflected in beat-to-beat changes in ECG⁶³ T wave morphology and lability causing augmented QT variability^64,65. Dantrolene prevents SR Ca²⁺ leak from RyR2 and normalize Ca²⁺handling, suggesting that SR Ca²⁺ leak is central to Ca²⁺ induced triggered activity. Chronic dantrolene therapy enhanced chronotropic competence in response to increased metabolic demand in addition to correcting ventricular electrophysiological abnormalities, even when DS therapy started after HF had already been established.

Because the electrophysiological, Ca²⁺ handling, autonomic, metabolic, and immune systems of the guinea pig are similar to that of humans, the findings in this study may be more relevant to humans as compared to other small animal models of SCD. For example, the ventricular action potential has a rapid upstroke and long plateau, defined by the balance of inward L-type Ca²⁺ current (Cacna1c) and repolarizing currents carried by the inward rectifier K channel (Kcnj2), and the rapid (Kcnh2) and slow (Kcnq1/Kcne1) components of the delayed rectifier K current, as well as the long cellular action potential plateau morphology that contributes to the prolonged QT interval of human heart failure. Excitation-contraction (EC) coupling displays a 60:35 balance between SR Ca²⁺ reuptake and Na/ Ca²⁺ exchange for Ca²⁺ removal on each beat, as well as a 3:1 β to α myosin heavy chain isoform ratio^66,67. These properties are remarkably similar to the adult human heart and essential for studying how cardiac remodeling contributes to contractile dysfunction and VT/VF as HF develops and progresses.

Conclusion and future directions

Damage caused by oxidative stress alters the proteomic profile of the cell, including altering expression levels as well as the structure and function of EC-coupling proteins. This study demonstrates that inhibition of Ca²⁺ leak via RyR2 pauses the transition to heart failure by reducing cardiac fibrosis, increasing repolarization lability, and decreasing QT heterogeneity, improving chronotropic competency, ultimately improving LV contractility, and mitigating lethal VT/VF. Given that dantrolene is available for clinical use, future studies should evaluate its efficacy in humans with heart failure.

Footnotes

Author contributions

BCK and SD jointly conceived and designed this study. PJ and SE performed the experiments. SD, DD and PJ performed data analyses and interpreted the results. The manuscript has been read, revised, and approved by all authors.

Sources of funding

This work was supported by American Heart Association Grants AHA19TPA34850151(to SD) and AHA19SFRN34830019 (to BCK), NHLBI/National Institutes of Health Grants NIH 4R00HK130662 (to DD), NIH DP2HL157941 (to DD) and Department of Defense Grants DOD-W81XWH-19-1-0640 (to SD), DOD-W81XWH-20-1-0701 (to SD and DD).

Abbreviations

• DS: Dantrolene Sodium

• EDV: End Diastolic Volume

• ESV: End Systolic Volume

• FS: Fractional Shortening

• HF: Heart Failure

• HRV: Heart Rate Variability

• ICD: Implantable Cardioverter-Defibrillators

• LV: Left Ventricle

• PTM: Post-Translational Modifications

• ROS: Reactive Oxygen Species

• RWT: Relative Wall Thickness

• RyR2: Ryanodine Receptor 2

• SCD: Sudden Cardiac Death

• VT/VF: Ventricular Tachycardia/Ventricular Fibrillation

• β-AR: β-adrenergic receptors

Supplemental Figure

Figure S1.

Figure S2.

Methods

Animal model

All animal work followed IACUC-approved protocols at the respective institutions. A pressure overload model of HF and SCD was surgically generated with ascending aortic constriction (AC) and a daily bolus of low dose isoproterenol (2mg/kg/day) for β-adrenergic challenge. The surgical procedure and animal model have been previously described in detail²⁴. Briefly, animals were anesthetized with isoflurane then intubated. Ascending AC was accomplished by banding the aorta with a 18 mm diameter ligation clip. Throughout the anesthetic event, animals were continuously monitored for vital signs including ECG, pulse oximetry and temperature. Sham animals received a thoracotomy without aortic banding. Daily isoproterenol bolus was delivered via a programmable iPRECIO® pump (Primetech Corp., Tokyo, Japan) implanted in the peritoneal cavity. Animals were randomized to either vehicle or DS. The DS (25 mg/Kg, Patterson Vet catalog# 07-893-7963) was administered via the drinking water. All animals received an ECG transmitter implant (Data Sciences International) in the abdominal cavity and the leads were secured in a lead II axis arrangement. All devices and surgical equipment were gas sterilized (ethylene oxide) or autoclaved. A small subgroup of animals were subjected to a randomized crossover study. The following treatment group were studied:

1. Normal Control

2. HF (AC + once daily β-adrenergic challenge till end point)

3. HF+DS (HF animals received DS therapy till endpoint)

4. HF+DS (+/-) (HF animals received DS therapy till week 3; DS was stopped after week 3)

5. HF+DS (-/+) (HF animals without DS therapy till week 3; DS therapy was started during week 3 and continued till endpoint)

Personnel involved in data collection and analysis were blinded to the treatment and non-treatment groups. All groups received similar incisions and thus, could not be distinguished based on these interventions. Each animal was assigned a unique computer-generated numeric ID. Only male animals were used for this study because of heterogeneity in phenotype of female animals. The progression of HF in female guinea pigs was delayed and the incidence of VT/VF/SCD was distinct from males. Females and ovariectomized females will be the focus of a follow-up study.

Echocardiography

Echocardiography was performed on conscious animals using a L16-4HE Linear Array transducer (5.4 MHz-13.5 MHz) with a Mindray M9 Vet Ultrasound System in a blinded manner. Long axis views were used to obtain two-dimensionally directed M-mode images. Echocardiography measurements from M-mode images were used to determine fractional shortening, wall thickness, cardiac function, volumes, and ejection fraction of the different treatment groups. Three consecutive cardiac cycles by the leading edge-to-leading edge method were used to obtain the measurements. LV end-diastolic and end-systolic dimensions, and LV end-diastolic posterior wall thickness, were measured from the M-mode images, and left ventricular fractional shortening was calculated by the software.

Electrocardiogram (ECG) analysis

Continuous serial ECG recordings were collected using the Data Sciences International, St. Paul MN, telemetry system. Ponemah software was used for data acquisition. Data was exported to and analyzed in a custom developed MATLAB software (MathWorks, Inc., Natick, MA). ECG tracings were reviewed to quantify PVC, VT/VF burden and to adjudicate if death occurred due to SCD. ECG features were extracted from the recordings using the custom MATLAB algorithms and parameters were compared between the various treatment groups.

RyR2 immunoprecipitation and oxidation assay from guinea pig left ventricle tissue

Snap frozen guinea pig LV tissue was lysed using RIPA lysis buffer (Sigma, cat# R0278), supplemented with phosphatase (Roche, cat# 4906837001) and protease inhibitors (Roche, cat# 4693124001). RyR2 was immunoprecipitated from tissue lysate using anti-RyR2 antibody (Invitrogen Cat# MA3916, 5μL) in 0.5 mL RIPA buffer overnight at 4 °C as previously described ⁶⁸. Samples were incubated with Protein A/G Plus agarose beads (Santa Cruz, cat# sc-2003) for 1 h at 4°C and washed three times with RIPA buffer. To determine the oxidation status of RyR2, the Carbonyl Content Assay Kit (Abcam, cat #ab126287) was used, whereby carbonyl groups of immunoprecipitated RyR2 were derivatized to 2,4 dinitrophenylhydrazone (DNP) by reaction with 2,4 dinitrophenylhydrazine which can be quantified using a spectrophotometer at 375nm.

Masson’s Trichrome (MT) staining

Connective and fibrotic tissues are stained blue, nuclei are stained deep brown to black, and the cytoplasm is stained red with MT staining. LV tissue was collected from HF animals with and without DS treatment and fixed in 10% neutral buffered formalin. The samples were sectioned and stained at the Translational Pathology Shared Resource at Vanderbilt University.

Statistical analysis

All animal groups were age and weight matched male and randomized to study groups. We use Stata 17 (Stata Corp LP, College Station, TX) for statistical analyses. The Kolmogorov-Smirnov test was used to evaluate normality of distribution. For normal distributions, continuous variables were compared for 2 groups using unpaired t test and ≥3 groups using ANOVA. Non-parametric tests were used for data not normally distributed. Kaplan-Meier analysis was used to compare survival over time between groups. Bonferroni’s correction was applied when a sample had multiple time points/treatment measurements. A p value of 0.05 was considered significant and is denoted with an asterisk (*) in the figures.