Reduced cardiac antioxidant defenses mediate increased susceptibility to workload-induced myocardial injury in males with genetic cardiomyopathy

Abstract

Ongoing cardiomyocyte injury is a major mechanism in the progression of heart failure, particularly in dystrophic hearts. Due to the poor regenerative capacity of the adult heart, cardiomyocyte death results in the permanent loss of functional myocardium. Understanding the factors contributing to myocyte injury is essential for the development of effective heart failure therapies. As a model of persistent cardiac injury, we examined mice lacking β-sarcoglycan (β-SG), a key component of the dystrophin glycoprotein complex (DGC). The loss of the sarcoglycan complex markedly compromises sarcolemmal integrity in this β-SG^−/− model. Our studies aim to characterize the mechanisms underlying dramatic sex differences in susceptibility to cardiac injury in β-SG^−/− mice. Male β-SG^−/− hearts display significantly greater myocardial injury and death following isoproterenol-induced cardiac stress than female β-SG^−/− hearts. This protection of females was independent of ovarian hormones. Male β-SG^−/− hearts displayed increased susceptibility to exogenous oxidative stress and were significantly protected by angiotensin II type 1 receptor (AT₁R) antagonism. Increasing general antioxidative defenses or increasing the levels of S-nitrosylation both provided protection to the hearts of β-SG^−/− male mice. Here we demonstrate that increased susceptibility to oxidative damage leads to an AT₁R-mediated amplification of workload-induced myocardial injury in male β-SG^−/− mice. Improving oxidative defenses, specifically by increasing S-nitrosylation, provided protection to the male β-SG^−/− heart from workload-induced injury. These studies describe a unique susceptibility of the male heart to injury and may contribute to the sex differences in other forms of cardiac injury.

Graphical Abstract

INTRODUCTION

Muscular dystrophies are a genetically heterogeneous class of progressive muscle wasting diseases with variable cardiac involvement, including a large and diverse group of Limb-girdle muscular dystrophies (LGMDs). The sarcoglycanopathies are notably among the several LGMDs that feature clinically significant cardiomyopathy (1,2). The four sarcoglycans are important components of the larger dystrophin-glycoprotein complex (DGC), which plays a critical role in maintaining sarcolemmal integrity by serving as the transmembrane node between the intracellular stabilizing protein dystrophin and the extracellular anchoring protein laminin. The absence of β-, δ-, or γ-sarcoglycan results in the depletion of the whole sarcoglycan complex, precipitating childhood onset of pathological effects that include muscle weakness, myocyte necrosis, fibrosis, loss of ambulation, heart failure, and premature death from cardiorespiratory insufficiency (1,3). The loss of β-sarcoglycan appears to be the most devastating, as indicated by the age of onset, progression of the disease, and proportion and severity of cardiac involvement (4).

The β-sarcoglycan-null mouse (β-SG^−/−) is a preclinical model of the autosomal recessive LGMDR4, which displays progressive muscle degeneration and cardiac fibrosis. Earlier work characterizing this mouse model demonstrated total loss of the sarcoglycan complex and the associated protein sarcospan with retention of other DGC components, in agreement with similar observations in patient biopsies (5–7). Cardiac pathology in the β-SG^−/− mouse can be accelerated by increasing stress on the heart to induce acute myocardial injury and replacement fibrosis. The β-adrenergic receptor agonist isoproterenol (Iso) has been successfully used for this purpose by our group and several others, triggering increases in cardiac contractility and heart rate that produce modest injury in healthy hearts and large clusters of injured cardiomyocytes in dystrophic hearts. These areas of myocardial injury are subsequently replaced by fibrotic lesions within a few days, and the fibrotic lesions mature over weeks. In earlier work, we have shown that the hearts of mdx mice, a popular model of Duchenne muscular dystrophy, are highly susceptible to acute Iso-induced injury. Interestingly, this susceptibility is markedly reduced by acute treatment with the angiotensin II type 1 receptor (AT₁R) blocker losartan(8).

In the present study, we aimed to assess the susceptibility of β-SG^−/− hearts to both single and repeating bouts of acute Iso-induced injury. Here we report a large sex-based difference in the magnitude of Iso-induced injury, with male β-SG^−/− mouse hearts developing significantly more injury. We also demonstrate that, prior to injury, male β-SG^−/− hearts have increased susceptibility to oxidative stress and increased macrophage content. Furthermore, we find that blocking AT₁R signaling or increasing oxidative defenses provides a greater benefit to male β-SG^−/− hearts following Iso-induced myocardial injury. The relative resilience of female β-SG^−/− hearts is independent of ovarian hormones and associated with a smaller benefit from AT₁R blockade. The transcriptional response to the absence of β-sarcoglycan is highly divergent between males and females, suggesting a multitude of potential mechanisms for the observed sex-difference.

METHODS

Animals

β-sarcoglycan-null (β-SG^−/−) mice were bred at the University of Minnesota from breeding pairs obtained from the Jackson Laboratory. Mice lacking S-nitrosoglutathione reductase (GSNOR^−/−) were crossed with β-SG^−/− and maintained as a distinct colony(9). Mice were 4–6 months of age during all isoproterenol studies. Both male and female mice were used for these studies. Animals received standard chow, unless noted otherwise, and were housed under a 12:12 light-dark cycle, with lights on at 7:00 AM. All animal procedures were approved by the University of Minnesota Institutional Animal Care and Use Committee and performed in compliance with all relevant laws and regulations.

Isoproterenol Stress Testing

Two forms of the isoproterenol stress test were used in these studies. The single-dose study consisted of a single intraperitoneal (IP) injection of 10 mg/kg of (−)-Isoproterenol hydrochloride (Iso; Sigma #I6504) in sterile saline as previously described(8). Single injections were given between 8:00–9:00 AM and hearts were harvested at timepoints of 30 hours (14:00–15:00 the following day) and 1 month after the injection. The second Iso stress testing protocol was a series of 14 repeating 10 mg/kg doses over 5 days. These doses were given at 8:00, 13:00, and 18:00 on days 1–4. The first two doses were given on day 5, and tissues were harvested at 15:00. Decisions to humanely euthanize animals for their moribund condition were based on repeated observation of reduced spontaneous movement, decreased body temperature, or dehydration.

Ovarian Hormone Reduction

The role of ovarian hormones in cardiac injury was investigated by either bilateral removal of the ovaries or induction of chemical ovarian senescence using vinylcyclohexene dioxide (VCD)(10–12). Ovariectomy was performed in 3-week-old female mice as described previously(13,14) with the exception of simple interrupted stitches instead of wound clips for closing the dorsal skin incision. All animals were singly housed for one week during recovery and received post-operative antibiotics.

VCD was purchased from Synquest Laboratories (#2209-1-08) and dissolved at a concentration of 64 mg/ml in sesame oil, as described elsewhere (11). Starting at 1–2 months of age, the mice received daily injections of 160 mg/kg VCD or sesame oil vehicle, in volumes of 35–60 μl adjusted for body weight, for 15 days. Mice were used for studies and sacrificed 40–70 days from the end of their VCD injection course. All VCD-treated mice remained healthy throughout the study. Uterine weight was used to assess the extent of disruption in female sex hormones.

Drug Delivery

Losartan Injection:

As previous described(8), a sterile losartan solution was prepared from a crushed generic losartan tablet dissolved in saline, and delivered as four IP injections at 7:00 (60 mg/kg loading dose one hour prior to Iso injection), 14:00 (20 mg/kg booster), and 19:00 (60 mg/kg) on the first day and a final injection at 7:00 (20 mg/kg) on the following day with tissue harvest at 14:00. This dosing schedule does not interrupt the dark cycle and is based on the half-lives of losartan and its major active metabolite (EXP3174) and is intended to minimize the troughs in drug serum levels between the first dose and the time of sacrifice with minimal disruptions to the mice.

Losartan Oral consumption:

A losartan solution was prepared from a crushed generic tablet and given at a concentration of 600 mg/L in 2.5% dextrose, with an average consumed dose of 80 mg/kg daily. Losartan-containing water was replaced 3 times per week and administered for 2 weeks. During the period of the Iso stress test mice often stop drinking. To address the concern that serum levels would decrease in mice on oral losartan, following the Iso injection they were provided the last three booster doses as described above and given untreated water to drink.

CDDO-Me Administration:

The Nrf2 activator 2-Cyano-3,13-dioxooleana-1,9(11)-dien28-oic methyl ester (CDDO-Me) was dissolved in DMSO and administered daily through 3mg/kg IP injections (15). Mice were pretreated with CDDO-Me or vehicle for 4 days prior to the administration of 10 mg/kg isoproterenol. A fifth dose was provided 8 hours following the isoproterenol administration to ensure that sufficient Nrf2 activation was maintained throughout the response to the isoproterenol-induced injury.

N6022 Administration:

The GSNOR-inhibitor N6022 was obtained from Cayman Chemical (Cat No: 21269). For studies using the multi-dose Iso stress protocol, N6022 was given in a 10% DMSO saline solution in an osmotic pump (Alzet Model: 2002) at a daily dose of 0.5 mg/kg. Pumps were implanted 1 week prior to the first dose of Iso. These pumps last for two weeks and provide a steady flow of N6022 throughout the weeklong pretreatment and during the entirety of the multi-dose protocol. To assess the protection of N6022 against acute Iso-induced injury mice were injected IP with 0.6 mg/kg at 7:00, on hour prior to the injection of Iso. Subsequent doses of 0.2mg/kg at 13:00, 0.6 mg/kg at 19:00, and a final 0.2 mg/kg at 7:00 the following morning. Tissues from these mice were collected at 14:00.

In Vivo Hemodynamic Assay

Left ventricular catheterization was performed as described previously(16–18). Briefly, mice were induced and maintained at a surgical plane of anesthesia with isoflurane. Temperature was maintained at 37°C throughout the procedure. A bilateral thoracotomy was performed to expose the apex of the heart. The pericardium was removed, and a 1.2 Fr pressure-volume catheter was inserted into the left ventricle via an apical stab incision. An intravenous catheter was placed in the external jugular vein and 10% albumin supplemented saline was infused to a final volume of 5 μl/g body weight. This infusion was delivered during a 10-minute equilibration period. At the end of this equilibration period, baseline hemodynamics parameters were collected. Dobutamine was infused (42 μg/kg/min) to assess cardiac reserve. After 5 minutes of dobutamine infusion hemodynamic parameters were collected.

Histopathology

To avoid confounding variables associated with post-mortem changes, only the hearts of mice that survived to their 30-hour or 1-month timepoints were used for quantification of injury area following a single-dose Iso injection.

Histopathology studies were carried out as previously described, using 7 μm slices of unfixed frozen hearts (8). The following reagents were used for IF staining: goat serum for blocking (Jackson ImmunoResearch # 005-000-121, 10%), goat anti-mouse IgG (H+L) secondary antibody (Invitrogen #R37121, 1:200), WGA AlexaFluor 488 conjugate (ThermoFisher, 5μg/ml), and ProLong Gold Antifade Mountant with DAPI (ThermoFisher). After blocking, staining was performed as a single step at room temperature for 1 hour, followed by three 5-minute washes in PBS and cover slip mounting.

The following reagents were used for Sirius Red Fast Green (SRFG) staining: 1.2% picric acid solution (Ricca #R5860000), Direct Red 80 (Sigma #365548), Fast Green FCF (Sigma #F7252), Formula 83 clearing solvent (CBG Biotech F83), and organic mounting medium (CBG Biotech MM83). Slides were fixed for 3 hours in chilled acetone at −20°C before staining, then rehydrated in 70% ethanol followed by two changes of tap water. The rehydrated slides were stained for 25 minutes in rocking SRFG dye solution of picric acid, 0.1% Direct Red 80, and 0.1% Fast Green FCF. The slides were then rinsed in 3 changes of tap water, and dehydrated in 70% ethanol, 100% ethanol, and Formula 83 before applying organic mounting medium and coverslips.

Microscopy and Image Analysis

All images were collected on a Nikon Eclipse Ni-E upright motorized epifluorescence microscope using NIS-Elements software. The microscope is equipped with a Lumencor SOLA LED engine, Plan Apochromat Lambda objectives, a Nikon DS-Qi1Mc monochrome camera, and a Nikon DS-Fi2 color camera. All image analysis was performed in Fiji(19).

Whole-heart images of acute injury were collected as three fluorescence channels at a resolution of 0.92 μm/pixel. The WGA channel (staining extracellular matrix) was subtracted from the IgG channel to better distinguish intra-myocyte IgG signal. This injury area was then quantified by thresholding the resulting IgG image for total lesion area and normalizing it to total heart section area.

SRFG-stained sections were collected as brightfield whole-heart images at a resolution of 0.85 μm/pixel. Fibrosis was analyzed in these images using the color threshold function. Regions containing Sirius Red-stained collagen were measured as total area having a red hue above a minimum saturation threshold and total heart area was measured based on any hue above the same minimum saturation threshold. Fibrosis was calculated as the Sirius Red area normalized to total heart area.

Langendorff Preparation

Mice were injected with 150 U of heparin and 250 mg/kg of phenobarbital, once a surgical plane of anesthesia was obtained, the heart was quickly excised and placed in ice-cold physiological salt solution (PSS). The aorta was dissected free and mounted on a cannula, the heart was then transferred to the Langendorff apparatus where perfusion of warm oxygenated PSS was initiated. A polypropylene balloon was inserted into the left ventricle through the mitral valve and inflated with degassed water to an end-diastolic pressure of 10 mmHg. The heart was equilibrated for 20 minutes prior to the assessment of baseline contractile properties. Predefined functional cut-offs of a systolic pressure above 90 mmHg and a coronary flow rate between 2 and 6 ml/min were used to remove hearts that were seriously damaged during the isolation procedure; in total 4 hearts were excluded based on these criteria (3 female and 1 male). To assess the contractile reserve isoproterenol was infused into the coronary perfusate to a final concentration of 50 nM. Studies assessing the sensitivity to exogenous ROS were infused with H₂O₂ to the desired concentration. The stock concentration was altered such that less than 10% of perfusate originated from the syringe pump.

RNA-Sequencing

Hearts were harvested from 4-month-old male and female wild type and β-SG^−/− mice, snap frozen in liquid nitrogen, and pulverized into a fine powder. RNA was isolated using RNAeasy kit (Qiagen). Following quality control, samples were sequenced with a depth of 25 million paired 50bp fragments. Sequences were aligned to mouse reference genome using TopHat and transcript abundance identified with Cufflinks. Differential gene expression was calculated using DESeq2 (version1.36.0)(20) with an experimental design of ~ strain + sex + strain:sex. Significance was defined as a P_adj < 0.05 using Benjamini-Hochberg false discovery rate (FDR) to account for multiple testing. All analyses were performed in R (4.2.2)(21).

Non-myocyte Cardiac Cell Isolation

Mice were anesthetized and heparinized as described in the Langendorff protocol. Hearts were perfused with ice-cold PBS to remove blood, then thoroughly minced and placed in digestion buffer (0.1% Collagenase B, 1.0U/ml, Dispase II, 10 mM HEPES, 2.5 mM CaCl₂) at 37°C for 30 minutes. Following trituration of the minced heart it was filtered through a 70 μm. Cells were isolated by centrifugation (1400 rpm at 4°C for 4 minutes) and resuspended into Red Blood Cell lysis buffer. Cells were then pelleted and washed twice with PBS prior to cell counting and mass cytometry analysis.

Mass Cytometry

Following heparinization hearts were isolated and perfused to remove most of the blood within the coronary vasculature. Isolated cells were then released from the myocardium by a combination of enzymatic digestions and mechanical disruption. Single cells were passed through a 40 μm filter and stained using protocols as previously described (22). Samples were acquired using a CyTOF2 cytometer (Fluidigm) and resulting datasets were automatically gated by scripts using methods available in flowCore(23) and openCyto(24). UMAP projections were generated using CATALYST(25). All analysis performed in R(4.2.2)(21).

Statistics

Statistical analyses were performed using R 4.2.2(21). Survival analyses were performed using the survival(26) and survminer(27) packages. When multiple survival groups were compared p values were adjusted using the method of Benjamini-Hochberg as implemented in the pairwise_survdiff method of the survminer package(27). Unless stated otherwise, the myocardial injury data was analyzed in a two-way ANOVA using Bonferroni post-hoc testing to control for multiple comparisons. Statistical outliers were identified using the Grubbs test function of the outliers package(28). Plots were generated using the ggplot2 package(29).

RESULTS

Repeated Elevations in Myocardial Workload Results in Increased Mortality in Male β-SG^−/− Mice but Female β-SG^−/− Mice are Protected.

Sex differences have not been previously described in β-SG^−/− mice. Quantification of the levels of existing fibrosis under baseline conditions failed to identify any significant difference between male and female β-SG^−/− mice. However, repeated stimulation of the β-adrenergic receptor (β-AR) with isoproterenol (Iso) revealed a significant survival difference between male and female β-SG^−/− mice (Fig. 1A). This protocol has been shown to cause significant mortality in male mdx mice (30); however, because of the X-linked nature of the Duchenne muscular dystrophy, for which the mdx mouse is a model, these studies did not include female mice. The repeated isoproterenol dosing implemented here resulted in large areas of necrosis and replacement fibrosis in both male and female β-SG^−/− hearts. These images were not quantified because of the bias introduced by the large differences in survival and the number of doses between the male and female mice. Notably, although female β-SG^−/− mice did have large areas of myocardial injury, they tolerated this injury and survived the entire protocol.

Figure 1. Male β-SG^−/− mice have significantly greater mortality than female β-SG^−/− mice following repeated Iso injections.

(A) Survival curve showing mortality in male and female β-SG^−/− and wild type mice following repeated injections of 10 mg/kg Iso. n = 6 C57BL/6 male, 6 C57BL/6 female, 23 β-SG^−/− male, and 27 β-SG^−/− females. (B) Survival curve following ovariectomy (OVX) or sham surgery and repeated injections of 10 mg/kg Iso in female and male β-SG^−/− mice. N = 8 β-SG^−/− sham male, 15 sham β-SG^−/− female, and 19 OVX β-SG^−/− female.* Indicates P < 0.05 calculated by pairwise comparisons with correction for multiple testing.

There are many reports of estrogen providing various cardioprotective effects to females. To investigate this possibility, we performed ovariectomy or sham surgeries in 3-week-old mice. By removing the ovaries at a young age prior to the onset of estrous cycles, we aimed to limit any potential long-lasting effects of prior estrogen exposure. Interestingly, the protection of female β-SG^−/− mice was completely intact in the absence of ovarian derived hormones. In contrast, β-SG^−/− males that underwent a sham surgical procedure did even worse with 100% mortality by the 5^th injection, suggesting that the surgical procedure, months prior, made the males even more susceptible to death following repeated increases in workload. The female sham mice fared slightly but not significantly worse than the ovariectomized mice, underscoring the independence of the female protection from ovarian hormones (Fig. 1B).

Hemodynamic Response to β-Adrenergic Receptor Activation Reveals Few Sex Differences.

To evaluate the hemodynamic response to β-AR stimulation we used cardiac catheterization to assess cardiac function (Fig. 2A). Both male and female β-SG^−/− mice responded to dobutamine infusion with significant increases in heart rate and maximal rate of pressure development (dP/dt_Max), indicating β-AR responsiveness remains intact and was of equal magnitude in both sexes (Fig. 2 B and C). Ejection fraction was also significantly increased with dobutamine in both sexes (Fig. 2D). Load-independent measures of contractility provide a more focused picture of myocardial function, with preload recruitable stroke work in female β-SG^−/− mice having a robust increase in left ventricular contractility following β-AR stimulation. In contrast, male β-SG^−/− mice had no significant increase in myocardial contractility following β-AR activation (Fig. 2E). Overall, these data demonstrate that female β-SG^−/− mice have similar, if not greater, contractile response to β-AR activation.

Figure 2. Hemodynamic responses to β-adrenergic receptor stimulation are similar between male and female β-SG^−/− mice.

(A) In vivo cardiac catheterization confirms that female β-SG^−/− mice have equivalent, if not greater, response to β-adrenergic receptor activation by dobutamine. The chronotropic (B) and ionotropic (C-E) cardiac functional measures are presented at baseline (BL) and after dobutamine infusion (Dob). The bars show the mean ± SEM. Each point is an individual heart, 6 β-SG^−/− males and 6 β-SG^−/− females; note: one female β-SG^−/− mouse was lost to a procedural complication prior to the collection of dobutamine data. * p < 0.05 vs. baseline female values; † - p < 0.05 vs. male baseline values calculated by two-way ANOVA with correction for multiple comparisons.

A Single Injection of Isoproterenol Reveals Male Sensitivity to Injury that is Mediated by Angiotensin II

To further investigate the acute occurrence of myocardial injury, damage induced by a single injurious β-AR activation bout was measured 30 hours following a single bolus injection of 10 mg/kg isoproterenol. This dose results in 3–4 hours of elevated heart rate and contractility(31). In previous studies we demonstrated that myocardial injury was evident at 8 hours, but the extent of the injury peaks 30 hours following Iso injection(8). Intracellular IgG staining was used to identify the disruption of the sarcolemmal integrity in cardiac myocytes. Extensive myocardial damage was evident throughout the heart of male β-SG^−/− mice 30 hours after isoproterenol (Fig. 3A and B). This injury was evenly distributed throughout the right and left ventricles and did not appear more prevalent in endocardial myocytes relative to epicardial cells, suggesting that ischemia is not likely to play a major role. Most injured myocytes were a part of large areas of myocardial damage, with single isolated injured myocytes being less common. Female β-SG^−/− hearts showed significantly less injury compared to their male counterparts.

Figure 3. Acute and chronic myocardial injury following increased workload reduced by losartan.

(A) Myocardial injury 30 hours following a single injection of 10 mg/kg isoproterenol without and with acute losartan pretreatment in male and female β-SG^−/− hearts. N = 16 β-SG^−/− male Iso, 20 β-SG^−/− male Iso + losartan, 15 β-SG^−/− female Iso, 18 β-SG^−/− female Iso + losartan. (B) Representative images of data in panel A, showing IgG-positive lesions in heart cross-sections. (C) Fibrosis 1 month following a single injection of isoproterenol, without and with losartan treatment in male and female β-SG^−/− hearts. N= 12 β-SG^−/− male-Iso, 10 β-SG^−/− male-Iso + losartan, 10 β-SG^−/− female-Iso, and 8 β-SG^−/− female-Iso + losartan. (D) Representative images of data in panel C, showing Sirius Red-stained fibrosis. Graphs are mean ± SEM. Each point is an individual heart.

* p <0.05 vs. untreated females; † p <0.05 vs. untreated males calculated by two-way ANOVA with correction for multiple comparisons. Bar represents 500 μm.

To determine the fate of these injured myocytes, we measured the fibrosis present in hearts 1 month following the isoproterenol-induced injury. Male β-SG^−/− hearts showed extensive fibrosis throughout the myocardium (Fig. 3C and D). Female β-SG^−/− hearts also displayed fibrosis but at levels significantly lower than observed in males, consistent with lower injury at 30 hours. However, when taking the amount of initial injury into account, male hearts appeared to have greater eventual fibrosis. In male β-SG^−/− hearts, 24.1 ± 1.8 % of the myocardium was damaged at 30 hours corresponded to 21.2 ± 2.7 % fibrosis at 1 month, reflecting a 12% contraction in lesion area. Female β-SG^−/− hearts displayed damage in 13.0 ± 2.3 % of their myocardium 30 hours after Iso and fibrosis in 9.6 ± 1.6 % of the heart after 1 month, showing a lesion area reduction of 26%. This discrepancy highlights additional potential differences in post-injury cardiac remodeling in male vs. female β-SG^−/− hearts.

In our previous studies, we demonstrated that the Iso-induced injury was mediated to a large extent by signaling through the angiotensin II type 1 receptor (AT₁R) in mdx mice. To determine if this observation extends to β-SG^−/− mice, we pre-treated mice with the AT₁R antagonist losartan at least 1 hour prior to administering Iso. Blockade of AT₁R signaling significantly reduced the levels of injury observed in male β-SG^−/− hearts, but the more modest effect in female β-SG^−/− hearts did not reach statistical significance (p=0.1, Fig. 3A). Continuation of losartan treatment for a month following isoproterenol-induced injury resulted in lower levels of fibrosis in male β-SG^−/− relative to untreated males. Fibrosis was not significantly different in female β-SG^−/− mice, consistently with the non-significant difference in acute injury.

Injury Resulting from a Single Dose of Isoproterenol is Not Dependent on Ovarian Hormones

To assess the potential effect of ovarian hormones on the relative protection from injury observed in female β-SG^−/− mice, mice were ovariectomized at 3 weeks of age and given a single dose of isoproterenol at 16 weeks of age. There was no significant difference in the myocardial injury level between sham and ovariectomized mice 30 hours following the isoproterenol injection (Fig. 4A). Interestingly, mice that underwent prior surgical procedure, either sham or OVX, had significantly less fibrosis at 1 month following injury (Fig. 4B).

Figure 4. Ovarian removal or senescence has no impact on workload-induced injury in β-SG^−/− females.

Levels of workload-induced injury 30 hours following Iso (A) and cardiac fibrosis 1 month after Iso (B) in female β-SG^−/− mice lacking ovarian hormones through ovariectomy (OVX) or chemical-induced ovarian senescence (VCD). Untreated injury levels initially shown in Fig 3 are reflected using horizontal purple (female) and blue (male) lines. (Sham:8, OVX:7, vehicle (Veh):5, and VCD:11). * p <0.05 vs. untreated females, † p <0.05 vs. untreated males. (C) Both OVX and VCD treatments resulted in significant uterine atrophy. Graphs are mean ± SEM. Each point is an individual animal. Uterine weight data from 88 control, 14 OVX, and 21 VCD treated β-SG^−/− females. ‡ p < 0.05 vs. untreated controls, § p <0.05 vs. OVX by one-way ANOVA with correction for multiple comparisons.

To evaluate the role of ovarian hormones in the absence of surgical intervention, the ovarian follicular toxin 4-vinylcyclohexene diepoxide (VCD) was used to induce premature ovarian senescence. This treatment resulted in a significant decline in uterine weight relative to untreated and sham mice but still greater than that observed with complete ovariectomy (Fig. 4C). Following isoproterenol administration, VCD-treated mice had levels of acute injury and chronic fibrosis that were not different from vehicle treated controls, reaffirming the ovarian hormone-independence of the resilience of the female heart to injury.

In preparing for this study, we evaluated the role of ovarian cycling as a potential contributor to the response to isoproterenol-induced injury. It was discovered through vaginal morphology and cytology that only half of the female mice housed in standard ventilated microisolator cages were cycling. In contrast, females housed in non-ventilated cages in the same room as non-ventilated male cages showed signs of universally preserved ovarian cycling. To ensure that ovarian cycles were not a confounding variable in these studies, all mice were housed in non-ventilated cages in a room containing both sexes, and a portion of male nesting material was placed in all female cages. Using vaginal cytology to identify estrous cycle phases, we found no phase-dependent differences in injury among female mice, however the process of collecting samples for cytology was correlated with significantly more injury, likely resulting from increased handling stress. Therefore, we did not include cycle-dependent measures of injury in our global analysis.

Male β-SG^−/− Hearts Have Increased Sensitivity to Workload-Induced Damage and Exogenous Oxidative Stress

To evaluate the intrinsic susceptibility of the β-SG^−/− heart to workload-induced damage, hearts were mounted on a Langendorff perfusion apparatus. This removes confounding variables associated with innervation and changes in coronary perfusion that may occur with in vivo administration of isoproterenol. Following an equilibration period, 50 nM isoproterenol was administered into the coronary circulation. There were no differences in the contractile function at baseline or following isoproterenol administration between male and female hearts (Fig. 5A). Interestingly, despite normal contractile response, male β-SG^−/− hearts displayed greater levels of creatine kinase release, indicative of myocardial injury (Fig. 5B).

Figure 5. Immediate workload-induced injury and susceptibility to oxidative damage in isolated male β-SG^−/− hearts.

(A) Similar contractile baseline (BL) and response to isoproterenol (Iso) infusion in isolated perfused hearts from male and female β-SG^−/− mice. (B) Release of creatine kinase (CK), indicating a loss of cardiomyocyte membrane integrity, at baseline (BL) and in the first 5 minutes of Iso stimulation from male and female β-SG^−/− hearts. Data from 9 β-SG^−/− female and 12 β-SG^−/− male hearts at baseline and 4 β-SG^−/− female [after one removed as a statistical outlier] and 6 β-SG^−/− male hearts following Iso treatment. (C) Decline in contractile function (rate pressure product; RPP) associated with the addition of hydrogen peroxide (H₂O₂) directly into the coronary perfusate in male and female β-SG^−/−hearts (data from 6 female and 14 male hearts). Graphs show mean ± SEM. Each point is an individual heart. * p< 0.05 vs. baseline females, † p < 0.05 vs. baseline males, ‡ p < 0.05 vs. isoproterenol infused females, ▽ p<0.05 vs. males at the same H₂O₂ concentration using two-way ANOVA with corrections for multiple comparisons.

Previous studies have demonstrated that isolated male cardiomyocytes have increased susceptibility to exogenous oxidative stress(32). To examine potential sex differences in the vulnerability to oxidative injury of myocytes within the intact myocardium, exogenous hydrogen peroxide was added to the coronary perfusate of untreated β-SG^−/− hearts. Male β-SG^−/− hearts displayed significantly greater reductions in contractile function at lower concentrations of this exogenous oxidative stress relative to female hearts (Fig. 5C). The relative hypoxia present in the isolated heart does represent an additional oxidative stress but this occurs equally in both sexes.

Significant sex differences in the transcriptional response to muscular dystrophy.

To evaluate transcriptional changes that may underlie the observed sex difference in susceptibility to injury in dystrophic hearts, we performed a bulk RNA-seq analysis of wild type and β-SG^−/− hearts from both sexes (Supplemental Figure and Tables). Despite sharing identical genetic backgrounds, the transcriptional response to the absence of β-sarcoglycan was highly divergent between males and females. Compared to sex-matched wildtype hearts, female β-SG^−/− hearts had 135 significantly altered transcripts and male β-SG^−/− hearts had 79 differentially expressed transcripts versus male wild type hearts. However, only 41 of these differentially expressed transcripts were different in the hearts of both sexes of dystrophic mice. Among the transcripts increased in male β-SG^−/− hearts were a variety of transcripts indicating an increased inflammatory environment. These included genes related to immune cell infiltration (e.g. CXCL14 and CCL7), matricellular transcripts (eg. tenascin C and osteopontin), and modulators of inflammatory mediators (eg. Prostoglandin D2 Synthase and CCL7, the latter of which is a monocyte attractant). Notably, female β-SG^−/− hearts also have up regulations of cytokines, relative to wildtype females, but they are distinct from those in males and include increased expression of CCL8 and CCL9. Together these transcriptional changes suggested that increased inflammation might contribute to the increased susceptibility of male β-SG^−/− hearts to injury.

Sex Differences in Cardiac Immune Cell Populations Before and After Workload-Induced Injury.

To evaluate the hypothesis that immune cell activity is linked to sex-dependent differenced in β-SG^−/− cardiac injury, we determined the identity of immune cells isolated from the β-SG^−/− heart before and after injury in male and female mice. Previous studies have demonstrated a large increase in immune cells following workload-induced myocardial injury in the dystrophic heart(8). It was hypothesized that differences in this immune response could contribute to the sex differences observed in this study. Using Cytometry by Time of Flight (CyTOF), non-myocyte heart cells from male and female β-SG^−/− mice were characterized before and 30 hours after workload-induced injury. The number of live cells isolated from β-SG^−/− hearts prior to injury was 1.38±0.16 × 10⁶ and 0.90 ± 0.16 × 10⁶ cells per heart for males and females, respectively (Fig. 6A). Thirty hours after workload-induced injury, male β-SG^−/− hearts yielded 5.40 ± 0.72 × 10⁶ cells per heart, while female β-SG^−/− hearts yielded 4.11 ± 1.1 × 10⁶ cells. Under baseline conditions, 24 ± 2% of male and 24 ± 1% of female live cells isolated from β-SG^−/− hearts were CD45⁺ immune cells (Fig. 6B). Myocardial injury significantly increased the percentage of immune cells in both male (51 ± 3% of non-myocyte cells) and female (45 ± 3% of non-myocyte cells) β-SG^−/− hearts (Fig. 6B).

Figure 6. Cellular response to workload-induced injury.

Mass cytometry data collected before (BL) and 30 hours after a single isoproterenol injection (Iso) in male and female β-SG^−/− hearts. (A) Total number of non-myocyte cells isolated from β-SG^−/− hearts before and after injury. (B) The percentage of live cells that were CD45⁺ immune cells. (C) Proportions of each cell type at baseline and 30 hours following isoproterenol injection. Bars show mean and SEM, and individual points reflect independent samples; BL data from 6 male and 6 female hearts and Iso data from 7 male and 7 female hearts. Analysis by two-way ANOVA with corrections for multiple comparisons. * p < 0.05, ** p<0.01, *** p<0.001 vs. baseline; † p <0.05 vs. female.

The relative proportions of the immune cells in the β-SG^−/− heart prior to injury demonstrated sex-differences. The proportions of B and T cells were significantly reduced in male β-SG^−/− hearts relative to female hearts (Fig. 6C). The reduction in these cells of the adaptive immune system was largely countered by an increase in CD64⁺ macrophages in male β-SG^−/− hearts. The proportion of macrophages in both male and female β-SG^−/− hearts was significantly increased 30 hours after injury. Neutrophils (polymorphonuclear; PMN) were also significantly increased following injury in both male and female β-SG^−/− hearts. The increased proportion of macrophages and PMNs in male hearts led to a relative decline in the proportion of other immune cell populations (Fig. 6C), although the absolute number of cells in the heart from each of these cell populations was significantly increased following injury.

Given the abundance of CD64⁺ macrophages in male β-SG^−/− hearts, a more in-depth analysis of these cells was performed. A non-biased clustering of CD64⁺ cells revealed that cardiac macrophages in the β-SG^−/− heart can be grouped into eight major groups based on their level of expression of major histocompatibility complex class II (MHC-II), the adhesion receptor F4–80 and the chemokine receptor CX3CR1 (Fig. 7A and B). The relative proportion of these cell population shifts 30 hours following workload induced injury. Prior to injury, the cardiac macrophages are split with ≈50% expressing high levels of MHC-II and 50% expressing low levels of MHC-II. Following myocardial injury, this macrophage population shifts significantly toward higher levels of the more inflammatory MHC-II^lo macrophages, which account for ≈75% of all macrophages in the injured heart. The small population of F4–80^lo cells also expands significantly with injury, likely representing monocytes (Fig. 7C and D). In concurrence with the shift toward more MHC-II^lo macrophages, the number of pro-inflammatory Ly6C⁺ macrophages significantly increased 30 hours following injury. There were also significant declines in more pro-reparative CD206⁺, CD115⁺, and Tim-4⁺ macrophages in the injured heart. Interestingly, only female hearts had significant increases in CD11c⁺ macrophages following injury (Fig. 7E). In summary, there were no significant sex differences in the nature of the macrophages in the dystrophic heart prior to injury, however the greater number of these macrophages in the male heart may contribute to exacerbated injury response. Workload-induced injury in the dystrophic heart resulted in a shift toward a more pro-inflammatory macrophage population.

Figure 7. Analysis of macrophage phenotype.

(A) UMAP analysis of macrophages (CD64⁺ cells) isolated from male and female β-SG^−/− hearts at baseline (BL) and 30 hours after isoproterenol (Iso) injection (B). Quantification of MHC-II^hi (C) and MHC-II^lo (D) macrophage populations based on F4–80 and CX3CR1 expression. (E) Relative abundance of macrophages positive for Ly6C, CD206, CD115, Tim-4, and CD11c. Bars show mean and SEM, and individual points reflect independent samples.

* p<0.05, ** p<0.01, *** p<0.001 vs. baseline, † p<0.05 vs. Iso treated female. Heart numbers and statistical analyses as described in Fig. 6.

Augmentation of Antioxidant Defenses in Male β-SG Null Mice Decreases Workload-Induced Myocardial Injury and Death

The observation that β-SG^−/− male hearts have increased susceptibility to oxidative damage, the involvement of signaling through AT₁R, and the greater presence of macrophages suggested that increasing the antioxidant defenses in the male β-SG^−/− heart may offer a functional benefit. Treatment with CDDO-Me activates the transcription factor nuclear factor-erythroid factor 2-related factor 2 (Nrf2) which activates the transcription of many antioxidant defenses (15,33,34). Increasing cardiac workload in male β-SG^−/− mice pretreated with CDDO-Me induced significantly less acute myocardial injury relative to vehicle-treated controls (Fig. 8A and B). Pre-treatment of female β-SG^−/− mice with CDDO-Me had no significant effect on the degree of acute injury induced by increasing cardiac workload.

Figure 8. Increasing anti-oxidant defenses protects the male dystrophic heart from workload-induced injury.

(A) Reduction in male β-SG^−/− myocardial injury following activation of Nrf2 by CDDO-Me male β-SG ^−/− mice are significantly protected from Iso-induced myocardial injury. DMSO data from 3 female and 10 male hearts; CDDO data from 9 female and 12 male hearts. Significance tested by T-test within sexes. (B) The GSNOR inhibitor N6022 significantly protect male β-SG ^−/− mice from Iso-induced injury, bar represents 500 μm. (C and D) Reduction GSNOR provides protection of male β-SG^−/−, during repeated Iso injections. Survival data from 8 wild type (WT), 20 gsnor^+/−, and 15 gsnor^−/− females and 10 WT, 20 gsnor^+/−, and 18 gsnor^−/− males; statistical analysis as in Fig. 1. * p <0.05 vs. DMSO-treated male β-SG^−/−-null mice. § p < 0.05 β-SG^−/−/gsnor^+/− vs. β-SG^−/−/gsnor^−/−. # p < 0.05 β-SG^−/−/gsnor^−/− vs. WT.

Cysteine residues are critical for protein function but can also be targeted by reactive oxygen species, leading to irreversible oxidation. S-nitrosylation (SNO), which is the reversible addition of a nitric oxide (NO) moiety to cysteine residues, provides a level of protection against oxidative protein damage(35). Protein SNO levels are regulated by denitrosylase enzymes, including S-nitrosoglutathione reductase (GSNOR). Descreased GSNOR activity leads to an increase in protein SNO levels. The small molecule GSNOR inhibitor N6022 has been shown to provide benefit in ischemia-reperfusion(9,36). In initial studies, isoproterenol administration resulted in significantly less injury in N6022-treated male and female β-SG^−/− mice compared to untreated mice of the same sex. Chronic administration of N6022 by osmotic pump also provided a small (p=0.049) survival benefit observed in β-SG^−/− males treated with N6022. The analysis of chronic administration of N6022 is complicated by a large increase in the mortality of female β-SG^−/− mice receiving the same dose of N6022. However, further studies with N6022 were halted by inconsistent effects between lots of N6022 from the same and different suppliers, thus the data is not shown here.

Given the inconsistency of small molecule manipulation of the S-Nitrosylation signaling axis, we crossed whole body GSNOR knockout mice(9)(gsnor^−/−) to β-SG^−/− and assessed the potential protection offered by permanent inhibition of GSNOR activity. Any reduction in GSNOR activity introduced increased mortality in female β-SG^−/− mice following repeated Iso dosing (Fig. 8C). In male β-SG^−/− mice the complete absence of GSNOR had no survival benefit. In contrast male β-SG^−/− mice heterozygous for GSNOR had a significant survival benefit relative to the full GSNOR knockout mice (Fig. 8D). These results suggest that partial inhibition of GSNOR may have therapeutic value. The observation that female β-SG^−/− mice are not protected by reductions in GSNOR activity suggests that differences in myocardial S-nitrosylation is an important mediator of the sex-differences observed in β-SG^−/− mice under high-workload conditions.

DISCUSSION

Estrogen-dependent health differences have been observed in patients and animal models representing many health conditions, yet the protection of the female β-SG^−/− heart from workload-induced myocardial injury is independent of ovarian hormones. This is in marked contrast to the sex differences observed in ischemia reperfusion(37–39) and cardiac hypertrophy(40,41) which depend on estrogen. The mechanism underlying this cardiac resilience and survival despite deficits in sarcolemmal integrity is unknown, however the results of this study suggest that increased susceptibility of male hearts to oxidative damage and alterations in the immune response to cardiac injury may play a role.

It is notable that both male and female β-SG^−/− mice have equal chronotropic response to β-adrenergic receptor activation. Similarly, both sexes display significant increases in load-dependent measures of contractility (dP/dt_Max, systolic pressure, ejection fraction, etc.). However, male β-SG^−/− mice display a marked attenuation of the load-independent contractile response, consistent with the beginning of contractile dysfunction within minutes of β-adrenergic receptor stimulation. In the Langendorff preparation, both sexes respond to β-adrenergic receptor activation; however, male hearts have significantly greater creatine kinase release, again consistent with greater acute injury in response to increased workload. The extent of the myocardial necrosis induced by increased workload in male β-SG^−/− mice is evident by the extensive IgG incorporation within myocytes a day following stimulation. Like the mdx mouse(42), much of workload-induced injury in male β-SG^−/− mice is mediated through angiotensin II type 1 receptors. Transcriptional analysis of wild type hearts reveals a relatively small number of sex-based expression changes. The number of genes with sex differences in expression is increased in dystrophic hearts. At the age of tissue collection, the hearts of dystrophic mice have small elevation in cardiac fibrosis prior to any workload challenge(43). This moderate level of disease is associated with many transcriptional changes, with many genes significantly down or up regulated. Most striking is highly divergent response between males and females in response to muscular dystrophy. Only 20% of the altered genes are shared by β-SG^−/− males and females. Like the transcriptional sex-differences in humans(44,45), the sex-differences observed in dystrophic mouse hearts are relatively modest in size. However, small changes in expression levels can have important physiological effects. Most of the genes differentially regulated between males and females did not segregate uniformly into specific biological processes.

Under baseline conditions male β-SG^−/− hearts have relatively more macrophages and fewer B- and T-cells compared to female β-SG^−/− hearts. Characterization of the macrophage populations reveal few sex-differences that might explain the dramatic difference in workload-induced injury. An important observation from these studies is that despite having nearly twice as much myocardial injury (Fig. 3), after 30 hours male β-SG^−/− hearts have the same number of CD45⁺ cells as seen in females (Fig. 7). This suggests the accumulation of immune cells is a limiting factor of the immune response 30 hours following injury. Interestingly, despite extensive myocardial necrosis, the number of neutrophils is relatively modest compared to the high number of neutrophils recruited into the ischemic heart (46–48). This may result from differences in the nature of the myocardial injury in the dystrophic heart. In ischemic heart disease there is nearly universal cell death of all cell types in the affected areas, in the dystrophic heart the acute cell death is limited to the cardiomyocytes. The cell type specific injury and the intact blood supply likely alter both the inflammatory signals and the ability of cells to respond through local proliferation or recruitment from the circulation.

The significant protection provided to male hearts by the AT₁R antagonist losartan is particularly interesting as angiotensin II administration is sufficient to cause a significant impact on the immune cell composition of the heart. Exogenous angiotensin II results in an increase of a complex mixture of pro-inflammatory MHC-II^lo/Ly6C⁻, likely derived from local expansion of resident macrophages, and an influx of Ly6C+ cells with a diversity of cellular markers. Furthermore, angiotensin II infusion resulted in a decline in the pro-reparative MHC-II^hi macrophages within the heart (49). These previously documented effects of angiotensin II occurred after two days of exposure, while we observed similar changes 30 hours following workload-induced injury in the hearts of both male and female β-SG^−/−. Interestingly, females have lower levels of injury and also have a significantly greater number of CD11c⁺ cells, but the importance of this cell population in limiting myocardial injury is unknown. At present it is not possible to determine if the relevant angiotensin II signaling occurs directly within immune cells or if these immune cells are drawn to the heart by actions of angiotensin II within the myocardium. During development CD206⁺ macrophages are pro-angiogenic(50), suggesting that these cells may have a reparative function; however, if this is the case in adult hearts remains to be seen. The decline of CD206⁺ in response to injury does suggest that these cells are not a significant contributor to the inflammatory reaction to injury in the dystrophic heart.

The increased susceptibility of male β-SG^−/− hearts to oxidative damage indicates that ROS may have an important role in the pathogenesis of workload-induced injury in the male dystrophic heart. The protection provided to male β-SG^−/− following Nrf2 activation further supports a role for ROS in the pathophysiology of the cardiomyopathy in male β-SG^−/− hearts. The addition of nitric oxide moieties to cysteine residues through S-nitrosylation has been demonstrated to provide protection against ischemia reperfusion injury (51–53). The regulation of protein S-nitrosylation is dependent on the enzyme S-nitrosoglutathione reductase (GSNOR). In the present study, partial inhibition of GSNOR provided protection against workload-induced myocardial injury in male β-SG^−/− hearts. In contrast, female β-SG^−/− hearts showed no meaningful benefit from GSNOR inhibition and may have been harmed by it. It is notable that complete absence of GSNOR activity fails to protect male β-SG^−/− hearts from workload-induced injury.

Here we present a mechanism of sex differences in susceptibility of the myocardium to injury in a genetic cardiomyopathy. In contrast to many previously described sex differences, the relative protection of females is completely independent of ovarian sex hormones. There are questions remaining regarding the underlying cause of increased susceptibility of male β-SG^−/− hearts to exogenous oxidative damage, but the observation is important given that oxidative stress is a common feature of many types of cardiac pathology. The role of this limited tolerance of oxidative stress in acquired heart disease remains uncertain, but the observation that dystrophin is disrupted in heart failure(54,55) and the strong correlation of serum cardiac troponins, markers of myocyte injury, with poor outcomes in heart failure(56,57) suggest that the pathophysiological mechanisms driving dystrophic cardiomyopathies may also contribute to other forms of heart failure.

HIGHLIGHTS.

• Male mice with a dystrophic mutation are more susceptible to myocardial injury.

• Female protection from dystrophic injury is not mediated by ovarian hormones.

• Males and females have distinct transcriptional responses to dystrophic mutations.

• Dystrophic males are more susceptible to exogenous oxidative stress.

• Improving oxidative defenses protects males, but not females.

Data Availability:

Raw RNASeq data are available with GEO accession number GSE249946.