Sexual dimorphism in bidirectional SR-mitochondria crosstalk in ventricular cardiomyocytes

Abstract

Calcium transfer into the mitochondrial matrix during sarcoplasmic reticulum (SR) Ca²⁺ release is essential to boost energy production in ventricular cardiomyocytes (VCMs) and match increased metabolic demand. Mitochondria from female hearts exhibit lower mito-[Ca²⁺] and produce less reactive oxygen species (ROS) compared to males, without change in respiration capacity. We hypothesized that in female VCMs, more efficient electron transport chain (ETC) organization into supercomplexes offsets the deficit in mito-Ca²⁺ accumulation, thereby reducing ROS production and stress-induced intracellular Ca²⁺ mishandling. Experiments using mitochondria-targeted biosensors confirmed lower mito-ROS and mito-[Ca²⁺] in female rat VCMs challenged with β-adrenergic agonist isoproterenol compared to males. Biochemical studies revealed decreased mitochondria Ca²⁺ uniporter expression and increased supercomplex assembly in rat and human female ventricular tissues vs male. Importantly, western blot analysis showed higher expression levels of COX7RP, an estrogen-dependent supercomplex assembly factor in female heart tissues vs males. Furthermore, COX7RP was decreased in hearts from aged and ovariectomized female rats. COX7RP overexpression in male VCMs increased mitochondrial supercomplexes, reduced mito-ROS and spontaneous SR Ca²⁺ release in response to ISO. Conversely, shRNA-mediated knockdown of COX7RP in female VCMs reduced supercomplexes and increased mito-ROS, promoting intracellular Ca²⁺ mishandling. Compared to males, mitochondria in female VCMs exhibit higher ETC subunit incorporation into supercomplexes, supporting more efficient electron transport. Such organization coupled to lower levels of mito-[Ca²⁺] limits mito-ROS under stress conditions and lowers propensity to pro-arrhythmic spontaneous SR Ca²⁺ release. We conclude that sexual dimorphism in mito-Ca²⁺ handling and ETC organization may contribute to cardioprotection in healthy premenopausal females.

Keywords: COX7RP; Cardiovascular diseases; Mitochondria; Oxidative stress; Sarcoplasmic reticulum Ca2+ release; Sexual dimorphism.

Fig. 1

Intracellular Ca²⁺ handling is divergent in healthy male and female ventricular myocytes. A Fluo-3 fluorescence (F/F₀) profiles of isoproterenol treated (ISO, 50 nmol/L) rat ventricular cardiomyocytes (VCMs) undergoing 1 Hz pace-pause protocol. Arrows depict field stimulation-evoked Ca²⁺ transients, stars show proarrhythmic diastolic Ca²⁺ waves. B Mean ± SEM Ca²⁺ transient amplitude (ΔF/F₀) and decay rate (s⁻¹), n = 90 male (M) and n = 90 female (F) VCMs; and spontaneous Ca²⁺ wave (SCW) latency (s), n = 62 M and n = 59 F VCMs. N = 11–12 M and N = 11–13 F animals. C Representative traces of caffeine-induced Ca²⁺ transients (10 mmol/L). D Mean ± SEM caffeine transient amplitude (ΔF/F₀) and decay rate (s⁻¹), n = 69 M, n = 75 F VCMs, N = 11 M, N = 13 F animals. *p < 0.05, p values were calculated using two-level random intercept model

Fig. 2

Matrix ROS biosensor MLS-HyPer7 reveals increased mito-ROS in male vs. female ventricular myocytes. A Mitochondrial MLS-Hyper-7 localization in live ventricular cardiomyocytes (VCMs) validated by Mitotracker. B Representative images of VCM infected with MLS-HyPer7, and treated with DTT (5 mmol/L) followed by DTDP (200 µmol/L) to achieve minimum and maximum fluorescence, demonstrating probe sensitivity. C Representative MLS-HyPer-7 recorded in male (M) and female (F) VCMs, respectively. Myocytes were treated with isoproterenol (ISO, 50 nmol/L) and paced at 1 Hz for 5 min. Fluorescence was normalized to minimum (DTT, 5 mmol/L) and maximum (DTDP, 200 µmol/L) fluorescence. D Mean ± SEM MLS-HyPer7 fluorescence, n = 37 M and n = 77 F VCMs, N = 11 M, N = 11 F animals. *p < 0.05, p values were calculated using two-level random intercept model

Fig. 3

Absence of sex‐related differences in mitochondrial counts and morphology in rat heart. A Mitochondrial density assessed by transmission electron microscopy (TEM) in the left ventricular cardiac tissue from male (M) and female (F) rats. Original magnification is × 3200 and the white bars represent 2 μm scale. Heart were fixed with 2.5% glutaraldehyde solution then processed for transmission electron microscopy. B–F Mitochondrial morphometric parameters, including mito area (B), perimeter (C), aspect ratio (D), form factor and density (% of the cell area, F), in male (M) and female (F) rat hearts. The analysis was performed randomly from 60 M (30 images), 60 F (21 images) cardiac mitochondria from three animals per group and were analyzed by three-level random intercept model

Fig. 4

Anti-oxidant defenses in male and female left ventricular cardiac tissues. A, B Representative western blots and pooled data for mitochondrial superoxide dismutase Mn-SOD and peroxiredoxin 5 (PRDX5) normalized optical density from male (M) and female (F) rat left ventricular (LV) tissue samples. GAPDH and Hsp60 was used as loading control. Mean ± SEM, N = 4 M and N = 4 F rat LV samples; *p < 0.05, Student’s t-test. C Total tissue GSH, reduced GSH, oxidized GSH (GSSG), and the ratio of GSH/GSSG was measured from male and female rat ventricles. Mean + SEM, N = 7 M and N = 6 F rat, *p < 0.05, Student’s t-test

Fig. 5

Mitochondrial ROS scavenging reduces spontaneous Ca²⁺ release is male ventricular myocytes. A Fluo-3 fluorescence (F/F₀) profiles of isoproterenol treated (ISO, 50 nmol/L) rat ventricular cardiomyocytes (VCMs) undergoing 1 Hz pace-pause protocol. Arrows depict field stimulation-evoked Ca²⁺ transients, stars show proarrhythmic diastolic Ca²⁺ waves. Pretreatment with MitoTEMPO, a mitochondria-specific ROS scavenger (20 mmol/L, 8 min) reduces spontaneous Ca²⁺ waves. B Mean ± SEM Ca²⁺ transient amplitude (ΔF/F₀) and decay rate (s⁻¹), n = 26 male (M) and n = 37 MitoTEMPO treated male (M + MT) VCMs; and spontaneous Ca²⁺ wave (SCW) latency (s), n = 23 M and n = 25 M + MT VCMs. N = 4 M and N = 4 M + MT animals. C Representative traces of caffeine-induced Ca²⁺ transients (10 mmol/L). D Mean ± SEM caffeine transient amplitude (ΔF/F₀) and decay rate (s⁻¹), n = 11 M, n = 15 M + MT VCMs, N = 4 M, N = 4 M + MT animals. *p < 0.05, p values were calculated using two-level random intercept model

Fig. 6

Female ventricular myocytes have reduced Ca²⁺ levels in mitochondria matrix vs. males. A Representative confocal images of a ventricular myocyte (VCM) infected with matrix-targeted mtRCamp1h (Kd ~ 1.3 mmol/L) and mito-GFP, with a merged figure demonstrating correct probe localization. B Representative time course of mtRCamp1h fluorescence recorded in male (M) and female (F) VCMs. Myocytes were treated with isoproterenol (ISO, 50 nmol/L) for 3 min before pacing at 1 Hz for 5 min. C Mean ± SEM baseline matrix [Ca²⁺], and peak matrix [Ca²⁺], time to peak matrix [Ca²⁺], and decay rate during pacing, n = 13 M and n = 15 F VMs. N = 8 M and 8 F animals. *p < 0.05, p values were calculated using two-level random intercept model. D, E Representative western blots and pooled data for COX7RP normalized optical density in rat and human left ventricular (LV) tissue samples. Mean ± SEM, N = 8 M and N = 8 F rat; and N = 4 M and N = 4 F human samples. *p < 0.05, Student’s t-test

Fig. 7

Intracellular Ca²⁺ handling is similar in male and female ventricular myocytes from aged (22 month old) hearts. A Fluo-4 fluorescence (F/F₀) profiles of isoproterenol treated (ISO, 50 nmol/L) rat ventricular myocytes (VCMs) paced at 1 Hz. B Mean ± SEM Ca²⁺ transient amplitude (ΔF/F₀) and decay rate (s⁻¹), n = 23 male (M) and n = 25 female (F) VCMs; proportion of cells displaying Ca²⁺ waves following 2 Hz pacing and distribution of latency to Ca²⁺ waves (s), n = 33 for M and n = 33 for F, N = 4 M and F animals. C Representative traces of caffeine-induced Ca²⁺ transients. D Mean ± SEM caffeine transient amplitude (ΔF/F₀) and decay rate (s⁻¹), n = 7 M and F VMs, N = 4 M and F animals. p values were calculated using two-level random intercept model except where indicated. E Aged male (M) and female (F) VCMs exhibit similar mito-ROS levels, while sexual dimorphism in matrix [Ca²⁺] preserved with aging. VCMs were passed at 1 Hz for 5 min in the presence of 50 nmol/L ISO, Mean ± SEM, N = 2–4 M, 2–4 F, n = 6–20, *p < 0.05, p values were calculated using one way ANOVA

Fig. 8

Oxygen consumption is similar in male and female working heart preparations. A Picture of rat working heart preparation. B Representative left ventricular (LVP), aortic and atrial pressure traces (mmHg) and LVP dP/dt of isolated rat heart in working mode. C–E Quantitation of male and female isolated rat heart myocardial O₂ consumption N = 3 males and N = 3 females. p values were calculated using Student’s t-test

Fig. 9

Female left ventricular tissues demonstrate increased mitochondrial supercomplex formation and expression of COX7RP compared to males. A, B Representative BN-PAGE images of ETC complexes from male (M) and female (F) rat (A) and human (B) left ventricular (LV) tissue samples. Antibodies used: Complex I—anti-NDUFA9; Complex III—anti-UQCRFS1; Complex IV—anti-COX IV all from Abcam. Hsp60 was used as loading control (inset, Anti-Hsp60 from Cell Signaling). Supercomplexes (SCs) are indicated by black bars. Mean ± SEM, N = 8 M and N = 8 F rat LV samples; and N = 6 M and N = 4 F human LV samples, *p < 0.05, Student’s t-test. C, D Representative western blots and pooled data for COX7RP normalized optical density in rat (C) and human (D) LV tissue samples. Hsp60 was used as loading control (inset, Anti-Hsp60 from Cell Signaling). Mean ± SEM, N = 8 M and N = 8 F rat samples; and N = 4 M and N = 4 F human samples, *p < 0.05, Student’s t-test

Fig. 10

Reduced COX7RP levels in LV heart tissues from old female donors and ovariectomized female rats. A, B Representative western blots and pooled data for COX7RP and MCU normalized optical density in left ventricular (LV) tissue samples from young (less than 35 year old, N = 10) and old (more than 60 years old, N = 10) female donor hearts. Mean ± SEM, *p < 0.05, Student’s t-test. C, D Representative western blots and pooled data for COX7RP and MCU normalized optical density in left ventricular (LV) tissue samples Sham and ovariectomised (OVX) rats. Hsp60 was used as loading control (Cell Signaling). E Representative BN-PAGE images of ETC complexes from rat LV tissue samples. Antibodies used: Complex I—anti-NDUFA9; Complex III—anti-UQCRFS1; Complex IV—anti-COX IV; all from Abcam. Hsp60 was used as loading control (inset, Anti-Hsp60 from Cell Signaling). Supercomplexes (SCs) are indicated by black bar. F Pooled normalized optical density data for (E). Mean ± SEM, N = 5 Sham and N = 7 OVX, *p < 0.05, Student’s t-test

Fig. 11

COX7RP modulates oxygen consumption in H9C2 myoblasts. A Representative western blot demonstrating efficiency of COX7RP overexpression and COX7RP shRNA constructs in H9C2 cells infected with adenoviruses for 48 h. Hsp60 and GAPDH were used as loading controls. B Representative Seahorse Analyzer recordings of oxygen consumption in H9C2 cells in control cells and cells expressing COX7RP or COX7RP shRNA. C Pooled respirometry data, Mean ± SEM, n ≥ 18, N = 3 independent experiments, *p < 0.05, one way ANOVA

Fig. 12

COX7RP expression levels regulate formation of mitochondrial supercomplexes in rat ventricular myocytes. A, B Transfection of rat ventricular cardiomyocytes with Ad-COX7RP-FLAG (male (M)) and with shRNA COX7RP (female (F)) alters expression levels of COX7RP protein 48 h after infection (10 MOI). Hsp60 was used as loading control. B Pooled data for (A), Mean ± SEM, N = 4 M and N = 5 F, *p < 0.05, Student’s t-test. C Representative BN-page experiments demonstrating that COX7RP overexpression increases formation of multimolecular ETC supercomplexes in Ms, while COX7RP shRNA decreases it in Fs. Antibodies used: Complex I—anti-NDUFA9; Complex III—anti-UQCRFS1; Complex IV—anti-COX IV; anti-GAPDH all from Abcam. Hsp60 was used as loading control (inset, Anti-Hsp60 from Cell Signaling). Supercomplexes (SCs) are indicated by black bars. D Pooled normalized optical density data for (C) mean ± SEM, N = 4 M and N = 5 F animals, *p < 0.05, Student’s t-test

Fig. 13

Altering COX7RP expression levels does not affect MCU expression and mito-Ca²⁺ uptake in male and female rat ventricular myocytes. A, B Transfection of rat ventricular cardiomyocytes (VCMs) with Ad-COX7RP-FLAG (male (M)) and with shRNA COX7RP (female (F)) does not change expression levels of MCU 48 h after infection (10 MOI). Hsp60 was used as loading control. Mean ± SEM, N = 4 M and N = 3 F, *p < 0.05, Student’s t-test. C, D Adenovirus-mediated changes in COX7RP expression levels do not affect mitochondrial Ca²⁺ uptake measured using mtRCamp1h Ca²⁺ biosensor in M and F VCMs. Mean ± SEM, n = 9 M and n = 20 M + COX, N = 4 M; n = 27 F and n = 21 F + shC and N = 3 F, *p < 0.05, p values were calculated using two-level random intercept model

Fig. 14

COX7RP overexpression reduces spontaneous SR Ca²⁺ release and mito-ROS in male ventricular myocytes under β-adrenergic stimulation. Conversely, COX7RP knock-down increases Ca²⁺ waves and mito-ROS in females. A, B Representative Ca²⁺ traces (A) and pooled data (B) for spontaneous Ca²⁺ waves (SCW) latency in male (M), female (F), male overexpressing COX7RP (M + COX), and female expressing shRNA COX7RP (F + shC) ventricular cardiomyocytes (VCMs) exposed to 50 nmol/L isoproterenol (ISO, 5 min) after cessation of field stimulation (1 Hz). Mean ± SEM, n = 58–64, N = 6–9 preparations, *p < 0.05, p values were calculated using two-level random intercept model. C Pooled data for normalized fluorescence of mitochondria matrix ROS biosensor MLS-HyPer-7. Mean ± SEM, n = 29–42 VCMs, N = 6–9 animals, *p < 0.05, p values were calculated using two-level random intercept model. D Adenovirus–mediated expression of COX7RP in males and shRNA COX7RP in females alter RyR2 oxidation levels in rat VMs exposed to isoproterenol (ISO, 50 nmol/L) paced at 1 Hz for 5 min cultured for 48 h after infection (10 MOI). Immunoprecipitated RyR2s from M and F cultured VCM samples were probed with Anti-DNP antibody to detect amount of oxidized cysteines. E Pooled DNP optical density normalized to corresponding RyR2 signal. Mean ± SEM, N = 3 M, N = 4 F. *p < 0.05, Student’s t-test