Estrogen regulation of cardiac cAMP-L-type Ca2+ channel pathway modulates sex differences in basal contraction and responses to β2AR-mediated stress in left ventricular apical myocytes

Backgrounds/Aim Male and female hearts have many structural and functional differences. Here, we investigated the role of estrogen (E2) in the mechanisms of sex differences in contraction through the cAMP-L-type Ca2+channel pathway in adult mice left ventricular (LV) apical myocytes at basal and stress state. Methods Isolated LV apical myocytes from male, female (Sham) and ovariectomised mice (OVX) were used to investigate contractility, Ca2+ transients and L-type Ca2+ channel (LTCC) function. The levels of β2AR, intracellular cAMP, phosphodiesterase (PDE 3 and PDE 4), RyR2, PLB, SLN, and SERCA2a were compared among the experimental groups. Results We found that (1) intracellular cAMP, ICaL density, contraction and Ca2+ transient amplitudes were larger in Sham and OVX + E2 myocytes compared to male and OVX. (2) The mRNA expression of PDE 3 and 4 were lower in Sham and OVX + E2 groups compared with male and OVX groups. Treatment of myocytes with IBMX (100 μM) increased contraction and Ca2+ transient amplitude in both sexes and canceled differences between them. (3) β2AR-mediated stress decreased cAMP concentration and peak contraction and Ca2+ transient amplitude only in male and OVX groups but not in Sham or OVX + E2 groups suggesting a cardioprotective role of E2 in female mice. (4) Pretreatment of OVX myocytes with GPR30 antagonist G15 (100 nM) abolished the effects of E2, but ERα and ERβ antagonist ICI 182,780 (1 μM) did not. Moreover, activation of GPR30 with G1 (100 nM) replicated the effects of E2 on cAMP, contraction and Ca2+ transient amplitudes suggesting that the acute effects of E2 were mediated by GPR30 via non-genomic signaling. (5) mRNA expression of RyR2 was higher in myocytes from Sham than those of male while PLB and SLN were higher in male than Sham but no sex differences were observed in the mRNA of SERCA2a. Conclusion Collectively, these results demonstrate that E2 modulates the expression of genes related to the cAMP-LTCC pathway and contributes to sex differences in cardiac contraction and responses to stress. We also show that estrogen confers cardioprotection against cardiac stress by non-genomic acute signaling via GPR30.


Background
Male and female hearts have well-known structural and functional differences [1][2][3]. However, investigations on the mechanisms that underlie these differences have not yielded a consistent conclusion. Differences in βadrenergic receptor (βAR) responses [4], L-type Ca 2+ channel (LTCC) function, sarcoplasmic reticulum (SR) Ca 2+ release dynamics [5], and age [6] have been shown to contribute to sex differences in many animal models. Yet, the cause of these variations remains unsettled, as such, and uncovering them is the focus of this study. It is emerging from prior studies that sex differences in cardiac contraction may be species-specific and/or dependent on multilevel effects of sex hormones [7]. The variation in plasma 17β-estradiol (E2) is known to alter cardiac output [8,9]. In our previous study, we found that E2 influences expression and signaling cascades of β 1 AR and β 2 AR, the two main regulators of cardiac contraction [10,11]. Therefore, we speculate that E2 may be modulating signaling cascades of the cardiac contraction system. This may explain the sex differences observed in cardiac contraction [7].
βARs transmit signals to various Ca 2+ -handling proteins mainly through the Gs/Gi -cAMP-PKA pathways [12]. cAMP-activate PKA, which phosphorylates LTCC leading to entry of Ca 2+ (I CaL ) which triggers SR Ca 2+ release through RyR2. The resultant rise in cytosolic Ca 2+ (Ca 2+ transient) activates myofilaments leading to contraction [13]. For relaxation to occur, Ca 2+ is pumped back to SR by SERCA2a which is regulated by Sarcolipin (SLN) and phospholamban (PLB). Hence, intracellular cAMP level is an important determinant of the activity of LTCCs, RyR2, SERCA2a, SLN, and PLB [12]. Despite the fact that cAMP signals are elicited by both β 1 ARs and β 2 ARs, it is known that β 1 AR is more efficient at increasing cardiac contractility than β 2 AR [14]. Several post-receptor factors affect the progression of βARsinduced signaling to ensure cardiac contraction occurs. One of such factors is the balance between synthesis of cAMP and its breakdown by cyclic-nucleotide phosphodiesterases (PDEs) [15]. PDE2-4 play major roles in cAMP compartmentation in adult cardiomyocytes [16]. Indeed, PDE2 and PDE3 are key determinants of basal cardiac contraction while PDE4 predominate cAMP regulation during β-adrenergic stimulation [15]. Hence, investigating whether there are sex differences in the cardiac expression of PDEs will help to elucidate if it plays a role in contractile differences between sexes.
The role of E2 (through its receptors; ERα, ERβ, and GPR30) and the modulation of β 2 AR in the heart are integral to the mechanisms that orchestrate cardioprotection. In the apex of the left ventricle, β 2 AR outnumbers β 1 AR whereas β 1 AR outnumbers β 2 AR at the base [17]. β 2 AR mediates stress induced by high catecholamine which produces hypocontractility from the mid left ventricle to the apex in Takotsubo cardiomyopathy. E2 on the other hand protects against stress-induced cardiomyopathy [18,19]. Indeed, the loss of E2 in postmenopausal women is thought to account for the high prevalence of cardiovascular diseases in this group compared to men and premenopausal women [20]. But, the role of E2 in sex differences in cardiac contraction is not clear. In this study, we examined the effects of E2 on β 2 AR, cAMP, LTCC, PLB, SERCA2a, RyR2, SLN, PDE3, and PDE4 families) using apical myocytes from adult male and female left ventricle (LV). Cardiac stress was induced by high epinephrine treatment. In identifying the specific E2 receptor (s) mediating the roles of E2, G1 (agonist) and ER antagonists (G15 and ICI 182,780) were used.

Animals
Sexually mature, weight-matched wild-type and β 2 ARknock out FVB mice (kindly donated by Professor Daniel Bernstein in (Stanford University, USA)) were used for the experiments. All animals were housed in standard cages at a temperature of 23 ± 2°C and a controlled environment with 12-h/12-h dark-light cycle. Mice were fed on pelleted food and water ad libitum. Adult cardiomyocytes have an advantage over neonatal cardiomyocytes given that stress cardiomyopathy and heart failure occur almost exclusively in the adult population.

Isolation of mice left ventricular apical myocytes
Animal body weight and wet heart weight were measured prior to myocyte isolation. Adult LV apical cardiomyocytes were dissociated using Collagenase B and D (Roche Diagnostics GmbH, Mannheim, Germany) and Protease type XIV (Sigma) as described by Zhou et al. [21], and it was optimized as follows (1) Animals were heparinized and anesthetized using Chloral hydrate and the entire digestion process was performed on the largendorff apparatus, (2) 20 mM taurine was used in all buffer solutions, (3) The hearts were submerged in ice-cold Ca 2+ -free buffer immediately after excision, (4) Another step was added during Ca 2+ re-introduction phase with 0.75 mM Ca 2+ to allow myocytes to acclimatize. (5), Myocytes from the apex of LV were taken from 3 to 4 mm from the very bottom of the heart. The myocytes were stored in a HEPES-buffered solution consisting of (in mM) 1 CaCl 2 , 137 NaCl, 5.4 KCl, 15 Glucose, 1.3 MgSO4, 1.2 NaH 2 PO4, and 20 HEPES, adjusted to pH 7.4 with NaOH at room temperature. Only elongated myocytes with clear striations and quiescent when unpaced were considered healthy for experiments.

Determination of plasma estrogen levels
Adult ovulating female mice were selected by examination of the vaginal opening and vaginal cell typology as surrogate indicators of the underlying changes of the estrous cycle as described previously [22]. Blood was collected after excision of the heart at the time of cardiomyocytes isolation, and serum was stored at − 80°C until the day of assay. Plasma E2 was determined by radiolabeling assay kit (Jiuding Biological Engineering Company, Tianjin, China) according to the manufacturer's instructions.

Mice stress model (β 2 AR-mediated)
A pilot study was performed to determine contractile response at the indicated concentrations of isoproterenol (ISO) (10 − 4 , 10 − 5 , 10 − 6 , 10 − 7 , 10 − 8 M) in male left ventricular myocytes. Briefly, few drops of medium containing myocytes were added to the superfusion chamber on the stage of an inverted microscope (Olympus, Tokyo, Japan). The cardiomyocytes were allowed to settle to the bottom of the chamber after which they were continuously superfused with a HEPES-buffered solution consisting of (in mM) 1 CaCl 2 , 137 NaCl, 5.4 KCl, 15 Glucose, 1.3 MgSO4, 1.2 NaH 2 PO4, and 20 HEPES at 23°C. Myocytes were stimulated at 0.5 Hz via platinum field electrodes in the absence/presence of ISO in HEPES-buffered solution. The contraction amplitude at each concentration of ISO was normalized to the basal amplitude. Peak amplitude shortening was recorded at 1 μM of ISO and hence this concentration was chosen for βAR stimulation throughout the study. Stress was simulated by pre-stimulation of β 2 AR with high epinephrine (Epi). Briefly, myocytes were pre-incubated with 1 μM Epi in HEPES-buffered solution at 23°C. At the duration of 10, 20, 40, and 50 min, their response to ISO stimulation was determined as previously described by Paur et al. [23]. β 2 AR-specific effects were confirmed in β 2 AR-gene knockout (β 2 KO) cardiomyocytes.

Determination of myocyte shortening and Ca 2+ transients
Measurement of cardiomyocyte contraction and Ca 2+ transients was performed as described before [24]. Briefly, cardiomyocytes were incubated with 1 μM acetoxymethyl ester of Fura-2 (Beyotime Biotechnology, Beijing, China) contained in the HEPES-buffered solution in a dark room at 23 ± 1°C for 25 min and thereafter the cells were washed three times with a fresh HEPES buffer (23°C, pH 7.4) before experimentation. To investigate the effects of 1 nM E2, G1 (100 nM), G15 (100 nM), ICI 182,780 (1 μM), on contraction and Ca 2+ transients, myocytes were co-incubated with these drugs in combinations shown in figures for 1 h at room temperature. For IBMX treatment, myocytes were incubated with IBMX (100 μM in DMSO) for 10 min at room temperature prior to experimentation. Cytosolic Ca 2+ transient (Fura-2 fluorescence) and cell length changes were recorded simultaneously using the IonOptix software (IonOptix Corp., Milton, MA, USA) with 0.5 Hz electrical stimulation at 23°C. Signals from the video edge system were digitized and stored in a computer. At the start of each recording, resting cell length was measured with the edge detector to allow normalization of contraction during analysis. All data were recorded 30 s after stimulation to ensure only stable data were captured. Only rod-shaped cells, without blebs or other visible structural alterations, which responded regularly to stimulation were used for this experiment. At least 10 regularly contracting myocytes were recorded for each group and contraction was presented as fractional shortening normalized to cell length and Ca 2+ transient amplitudes were calculated as the percentage difference between peak amplitude and baseline amplitude. Fluorescence data were analyzed using the IonOptix acquisition software. For determination of myocyte shortening of Fura-2 AM unloaded myocytes, cells were added to the stage of an inverted microscope (Olympus, Tokyo, Japan) and stimulated with 0.5 Hz electrical stimulation at 23°C. Data analysis was performed using Optical Measure software (kindly provided by China's National Defense University of Science and Technology).

Generation of OVX mice and measurement of body and heart weight
Ovariectomy is often used to induce artificial menopause. Weight-matched wild-type and β 2 AR-deficient female mice were randomly assigned to Sham and OVX groups. Mice were anesthetized with Chloral hydrate and ovariectomized after a single midline dorsal skin incision, 3 cm long (OVX group), but for the Sham group, mice were subjected to similar treatments except for the removal of ovaries. Mice were allowed 2-3 weeks to recover from the surgery. The body and wet heart weights were measured at the time of cardiomyocytes isolation ( Table 1).

Patch clamp experiments
Single, Ca 2+ tolerant and rod-shaped myocytes were suitable for whole-cell I CaL measurements. Voltageclamp mode was used in the recording of membrane currents. I CaL recording was performed using PatchMaster (v2x69) software (HEKA Instruments) and EPC10 amplifier. Pipette had tip resistance of 3-5 mΩ when filled with a solution containing the following (in mM) 120 CsCl, 10 EGTA, 5 Na 2 ATP, 10 HEPES (pH 7.2 by CsOH) and was used to form a high-resistance seal to myocyte membrane. A negative pressure was applied to form a whole-cell configuration by rupturing the membrane. I CaL was recorded with an appropriate stimulus protocol. The external bath solution for recording I CaL was (in mM) 133.5 NaCl, 4 CsCl, 1.2 MgCl 2 , 11.1 Glucose, 10 HEPES, 1.8 CaCl 2 (pH 7.4 by NaOH, at 37°C). I CaL currents were filtered at 2 kHz and sampled at 50 kHz using PatchMaster (v2x69) software (HEKA Instruments) and recorded using an EPC10 amplifier. The I CaL was then evoked by voltage pulses of 250 ms between − 40 and + 45 mV in 5 mV increments after inactivation of the I Na following 100-ms voltage steps from the holding potential of − 100 to − 40 mV. All recordings were made immediately after gaining whole cell access to avoid the run-down effect of I CaL . Capacitance measurements were obtained from membrane test parameters. The capacitance of adult male, female, and OVX myocyte capacitance was 163.4 ± 6.7, 147.2 ± 13.1, and 180.0 ± 12.8, respectively. I CaL were normalized to cell capacitance (pA/ pF). Series resistance was compensated 60% to prevent large voltage errors.

cAMP enzyme immunoassay
The level of cAMP levels was determined as described before using ELISA kit (Shanghai Jiang Lai Biological Technology Co., Ltd.). Myocytes from each group were incubated in 96-well plates (1000 cells/well) and with 1 nM E2, G1 (100 nM), G15 (100 nM), and ICI 182,780 (1 μM) in combinations shown in the figures for 1 h in the presence/absence of 1 μM ISO, 1 μM Epi for 40 min in HEPES buffer at room temperature. Thereafter, they were centrifuged to form a pellet (4000 rpm for 10 min) and resuspended in PBS and kept at − 80°C for 14 days. Myocytes cellular membrane rupture was achieved by a freeze-thaw method and intracellular cAMP levels were determined using an ELISA kit. Absorbance was measured at 450 nm by an ELISA plate reader and cAMP levels were calculated from a standard curve (r 2 = 0.96) Body weight and heart weight of male and female wild-type and β 2 AR-knockout mice. There were no significant differences in body weight and heart wet weights between male and female mice. The data are presented as means ± SEM and normalized to the quantity of protein in each sample.

Quantitative real-time PCR-RNA isolation and cDNA synthesis
Isolated male and female myocytes were used to isolate RNA using the TRIzol® reagent (Invitrogen Co., Carlsbad, CA, USA). The purity and concentration of total RNA (200 ng) were measured using NanoDrop 1000 (Thermo Scientific). Freshly isolated OVX myocytes were incubated with 1 nM E2 after 1 h of acclimatization for 5 h and before RNA isolation. cDNA was synthesized using ReverTra Ace qPCR RT kit (Vazyme Biotech, Nanjing, China) by reverse transcription of total RNA. The intron-spanning primers for all genes were synthesized by Sangon Biotech (Shanghai, China) Co., Ltd. All primers are shown in Table 1. Analysis of the cDNA was done in duplicate using 0.2 μmol/L specific primers ( Table 2) and 1× LightCycler 480 SYBR green I Master (Roche Applied Science, Germany) in a total volume of 10 μl. GAPDH was used as reference gene. PCR was performed using a Light Cycle 480 (Roche Applied Science, Germany) using the following thermal procedure 5 min at 95°C followed by 40 cycles of 10 s at 95°C, 30 s at 60°C , a melting curve of 15 s at 95°C, 60 s at 60°C and cooling for 30 s at 4°C. Analysis of the results was performed using the LightCycler 480 software (version 1.5, Roche Applied Science, Germany). The relative levels of mRNA were analyzed using the 2 ΔΔCt method. Quantitative mRNA level of all genes was normalized to the housekeeping gene GAPDH which were similar in all groups.

Statistical analysis
The value n denotes the number of mice/cells used in each experiment. All data analyses were represented as means ± SEM. Statistical analysis and construction of figures were performed with GraphPad Prism 5.01 (San Diego, CA, USA). Statistical significance was estimated by one-way ANOVA followed by Student's t-test to determine group differences. Spearman rank correlation test was performed to determine the relationship between plasma E2 and contraction amplitude shortening. Differences between groups were considered significant for P < 0.05.

E2 modulates sex differences in contraction and Ca 2+ transient amplitude in LV apical myocytes
Experiments were performed to investigate sex differences in left ventricular apical myocytes contraction and to determine the effect of E2 on contractility in FVB mice. At basal state, peak contractile amplitudes were significantly larger in Sham than in male and OVX myocytes (Fig. 1a). 1 h pretreatment with 1 nM E2 reversed the effects of OVX. 1 nM E2 was chosen as it is close to  Fig. 1c. To test the correlation between plasma E2 and contractile amplitude, we matched plasma E2 levels with myocyte cell shortening. Spearman's correlation analysis revealed that contraction amplitudes were positively and significantly correlated with plasma E2. (r = 0.8584; P < 0.05 (Fig. 1d). Collectively, these results show that contractility may be influenced by plasma E2 and that E2 may contribute to larger contraction amplitudes in female left ventricular apical myocytes compared to male. Since the effects of E2 observed here were acute, it is likely that they were mediated by non-genomic signaling. To determine the subtype of E2 receptor that mediated these effects, myocytes were pretreated with 1 nM E2, 100 nM G1, 100 nM G15, and 1 μM ICI 182,780 for 1 h. As shown in Fig. 1 (e and f ), pretreatment of myocytes with ERα and ERβ antagonist did not abolish the effects of E2, but GPR30 antagonist (G15) did. Moreover, activation of GPR30 with G1 replicated the effects of E2 on contraction amplitude and Ca 2+ transient amplitude suggesting that the acute effects of E2 were mediated by GPR30 via non-genomic signaling.
Sex difference in I CaL density of LV apical myocytes is modulated by E2 Cardiomyocytes were voltage-clamped using the protocol shown by the inset in Fig. 2a. Cell capacitance (pF) as a measurement of the membrane surface area was 163.4 ± 6.7, 147.2 ± 13.1, and 180.0 ± 12.8 pF for male, sham, and OVX, respectively (Table 1). Peak I CaL density was higher in Sham myocytes compared to male and OVX, but E2 increased I CaL density and restored it to normal levels. (Fig. 2b). These findings show that E2 increases I CaL density and may be responsible for higher I CaL density in female myocytes. To further examine whether the expression of LTCC at mRNA and protein differs between sexes and whether its expression is influenced by E2, we performed western blotting and real-time qPCR on isolated ventricular myocytes. Although the acute effects of E2 on functional parameters were tested after incubation for one hour, considering that the most mRNAs have a half-life of several hours, the effect of E2 on gene expression was tested after incubation for 5 h. LTCC mRNA and protein levels were markedly higher in sham compared to male and OVX myocytes. (Fig. 2 c and d). Incubation of OVX myocytes with 1 nM E2 for 5 h increased the expression of LTCC and reversed the effects of ovariectomy. Taken together, these data suggest that the sex differences in I CaL density are due to differences in LTCC expression and may be modulated by E2.
Basal cAMP levels are higher in sham compared to male and OVX myocytes The cAMP-PKA-LTCC pathway regulates cardiomyocyte contraction. To elucidate whether sex differences in contraction and Ca 2+ transient amplitude and I CaL density could be explained by intracellular cAMP levels between male and female myocytes, we measured the basal level of cAMP using ELISA kit. The results are shown in Fig. 3a. Sham myocytes had higher cAMP levels compared to male and OVX. But OVX myocytes incubated with 1 nM E2 for 1 h had higher cAMP concentration compared to OVX. Pretreatment of myocytes with 1 nM E2, 100 nM G1, 100 nM G15, and 1 μM ICI 182,780 for 1 h revealed that ERα and ERβ antagonist (ICI 182,780) did not abolish the effects of E2 on cAMP level, but GPR30 antagonist (G15) did. Moreover, activation of GPR30 with G1 produced similar effects as E2 on cAMP level implying that the effects of E2 were mediated by GPR30 via non-genomic signaling. These results suggest that E2 increases cAMP in female myocytes. We then postulated that differential cAMP breakdown could underlie the sex differences in cAMP levels. To test this possibility, male and female myocytes were treated with non-selective PDE inhibitor, 3-isobutyl-1-methylxanthine (IBMX) 100 μM in DMSO for 10 min since PDE is responsible for cAMP degradation. The effects of IBMX were examined functionally by measuring contraction and Ca 2+ transient amplitude. As shown in Fig. 3b-d, IBMX increased contraction and Ca 2+ transient amplitude in both groups and eliminated the differences between them. These findings imply that cAMP breakdown is different between male and female mice myocytes. We further determined the mRNA levels of phosphodiesterases PDE 3A, 3B, 4A, 4B, and 4D in male and female myocytes and the effect of E2 on PDE expression. We found that the mRNA expression of all PDE subgroups was higher in male and OVX compared with Sham and OVX + E2 groups, respectively (Fig. 4). Collectively these observations indicated that (1) lower cAMP levels in male mice may be due to high cAMP We also compared the expression of other proteins involved in regulation of intracellular Ca 2+ transients between sexes, and whether they are regulated by E2. mRNA levels of PLB, SLN, SERCA2a, and RyR2 were determined by real-time qPCR. Figure 5a shows that RyR2 expression was higher in sham myocytes and OVX + E2 compared with male and OVX myocytes, respectively. These data suggest that E2 may be responsible for the sex difference in RyR2 levels. High levels of RyR2 are associated with high Ca 2+ release from sarcoplasmic reticulum (SR). There was no significant difference in SERCA2a mRNA levels between male and sham (Fig. 5b). However, OVX decreased SERCA2a mRNA level while E2 reversed this effect. SLN and PLB regulate cardiomyocyte contraction through their inhibitory effects on SERCA2a and they are targeted by cAMP-PKA. As shown in Fig. 5c, both SLN and PLB were lower in Sham compared to male. Interestingly, OVX increased PLB while it decreased SLN compared to Sham. The effects of OVX were reversed by E2 treatment for 5 h.

E2 modulates sex responses to stress in LV apical myocytes
On the basis of β 2 AR signal trafficking, we developed an acute in vitro stress model in male LV cardiomyocytes according to a protocol established previously [23]. A pilot experiment was performed to determine the β-adrenergic receptor (βAR)-response to isoproterenol (ISO), a non-specific β 1 AR/β 2 AR agonist. As shown in Fig. 6a, incremental increases in ISO concentration elicited corresponding increases in myocyte shortening e PDE3B mRNA level in sham, male, OVX, and OVX + E2. (n = 3 mice). * P < 0.05, ** P < 0.01, and *** P < 0.001. All data are presented as mean ± S.E.M amplitude. Optimal amplitude shortening was observed at 1 μM ISO and this concentration was chosen for subsequent βAR stimulation. Stress was induced by incubating myocytes with epinephrine (Epi), which has a higher affinity for β 2 AR. Contractile responses to stress were evaluated at 20, 40, and 50 min using ISO. As shown in Fig. 6b, contractility was significantly decreased at 40 min of Epi incubation compared to the control group. The decline in contractility tended to recover at 50 min showing a recovery phase. The β 2 AR-mediated stress was confirmed using β 2 AR-gene knockout (β 2 KO) cardiomyocytes. The effects of stress on contraction and Ca 2+ transient amplitude were eliminated by β 2 AR gene knockout ( Fig. 6c and d). To test if E2 modulates sex differences in stress responses, apical LV myocytes from male, Sham, OVX, and OVX + E2 (myocytes pretreated with E2 after isolation for 1 h) were incubated with Epi for 40 min and their response to ISO challenge was examined. As shown in Fig. 7 a and b, there were no significant sex differences in contraction and Ca 2+ amplitude elicited by ISO, but OVX myocytes had higher amplitudes compared to Sham implying that OVX increased sensitivity to ISO. During stress, contraction amplitude and peak Ca 2+ transient were decreased only in wild-type (Wt) male and OVX groups but not in Wt Sham myocytes ( Fig. 7a  and b). Pre-treatment of OVX myocytes with 1 nM E2 and G1 (100 nM) increased contraction and Ca 2+ transient amplitude and restored the response to ISO. Moreover, pretreatment of myocytes with GPR30 antagonist (100 nM, G15) eliminated the effects of E2, but pretreatment with ERα and ERβ antagonist (1 μM, ICI 182,780) did not abolish the effects of E2. These datasets imply that GPR30 mediated the acute effects of E2 via non-genomic signaling. We further tested whether intracellular cAMP concentration was altered during stress. Similarly, we found that stress decreased cAMP concentration only in Wt male and OVX groups but not in Wt Sham myocytes. Moreover, 1 nM E2 and 100 nM G1 reversed the effects of OVX (Fig. 7c). Treatment of myocytes with G15 abolished the effect of E2 but co-treatment with CI 182,780 did not. These findings suggest that the β 2 AR-induced stress decreased cAMP thereby reducing contraction and Ca 2+ transient amplitude. In addition, Sham myocytes were resistant to stress which may be attributed to E2 protection.
E2 compensates for the loss of function resulting from a β 2 AR-gene knockout in female myocyte As further illustrated in Fig. 8a, I CaL density was smaller in β 2 AR KO Sham compared to Wt Sham myocytes, but it was not significantly different between Wt OVX and β 2 AR KO OVX myocytes. Similarly, LTCC mRNA level was decreased in β 2 AR KO Sham myocytes compared to c Relative PLB and SLN mRNA levels normalized to GAPDH mRNA in sham, male, OVX, and OVX + E2. (n = 3 mice). * P < 0.05, ** P < 0.01, and *** P < 0.001. All data are presented as mean ± S.E.M Wt Sham (Fig. 8b). Ovariectomy decreased LTCC mRNA level and eliminated the differences between wild-type Sham and β 2 KO Sham myocytes. These results indicate that the decrease of LTCC expression and I CaL density in female β 2 KO myocytes was depended on the E2 environment. Furthermore, we found that although there were no sex differences in mRNA level of β 2 AR, OVX decreased it while E2 abolished this effect (Fig. 8c). These results show that the effects of β 2 AR on LTCC may be influenced by E2.

Discussion
In this study, we investigated sex differences in basal contractility and responses to catecholamine-induced stress in LV apical myocytes. We found that (1) I CaL density, contraction and Ca 2+ transient amplitudes were larger in Sham compared to male and OVX myocytes at basal state. E2 reversed the effects of OVX on these parameters. (2) Basal cAMP concentration was higher in Sham myocytes compared to male and OVX; this was due to a lower cAMP breakdown in Sham myocytes which had lower mRNA levels of PDE 3A, 3B, 4A, 4B, and 4D compared with male and OVX. The effects of OVX on PDE expression were abolished by E2 treatment. Moreover, inhibition of cAMP breakdown by 100 μM IBMX increased contraction and Ca 2+ transient amplitude in both sexes and canceled differences between them. (3) In stress state, cAMP concentration, contraction amplitude, and peak Ca 2+ transient were decreased in male and OVX groups but not in Sham myocytes. Treatment of OVX myocytes with E2 in increased cAMP levels, contraction, and Ca 2+ transient amplitude.
(4) Pretreatment of myocytes with ERα and ERβ antagonist ICI 182,780 (1 μM) did not abolish the effects of E2, but GPR30 antagonist G15 (100 nM) did. Moreover, activation of GPR30 with G1 (100 nM) replicated the effects of E2 on cAMP, contraction and Ca 2+ transient amplitudes suggesting that the acute effects of E2 were mediated by GPR30 via non-genomic signaling. Collectively, these results show that E2 plays a key role in sex contractile differences in LV apical myocytes.
Male and female hearts have well-known contractile differences. Here, differences in basal contraction (in absence of agonist) and response to stress were observed. A summary of previous studies comparing sex differences is presented in Table 3. Similar to some studies [26][27][28], female myocytes had a larger contraction and Ca 2+ transient amplitudes compared to male myocytes (Fig. 1a-c). In contrast, Farrell et al. [5] found that female rat myocytes had a smaller contraction and Ca 2+ c Effects of stress on contraction amplitude abolished by β 2 AR knockout (β 2 KO) n = 10 cells/4 mice. d Effects of stress on Ca 2+ transient amplitude were abolished by β 2 KO, n = 10 cells/4 mice. * P < 0.05, ** P < 0.01, and *** P < 0.001. All data are presented as mean ± S.E.M transient amplitudes than males. This discrepancy may be explained by the experimental model (whole ventricular myocytes versus or LV apical myocytes) or species differences. In addition, in their study, rats were age-matched. But in this study, mice were weightmatched, and therefore, it is likely that male mice were younger than female considering that the growth rate of female mice is much slower than males [29]. Hence, as reported by Howlett et al., contraction in aged female is higher compared to younger male [6]. Other likely causes of the discrepancy are the experimental conditions such as (1) contraction and Ca 2+ transient recorded at 37°C versus 23°C used in this study. Temperature would affect the activity of ion channels of E-C coupling [30]. (2) In their study, contractions and Ca 2+ transients were activated simultaneously with 250-ms test steps to different potentials. Here, field stimulation with action potential were used. These two methods were found to produce different level of fractional shortening, being higher in stimulation by voltage-clamp steps [5]. Equally interesting is that Parks et al. [31] found that Ca 2+ transient was smaller in female than in male C57BL/6 mice with no differences in I CaL density between the sexes. The discrepancy observed here may arise from several factors. In their experiments, global ventricular myocytes were investigated while in our study, only apical myocytes of the left ventricle were studied. This presents a major source of discrepancy considering that apical versus basal variations in I CaL density have been reported in rabbit hearts [32]. In addition, differential distribution of βAR receptors in the left ventricle (i.e. β 2 AR > β 1 AR at the apex and β 1 AR > β 2 AR at the base [17]) implies that myocytes from these regions may respond differently to stimulation. However, it remains to be determined whether there are sex differences in βAR distribution in the left ventricle regions.
Short-term OVX (21 days) decreased the contractions and Ca 2+ transients and pretreatment of myocytes with E2 abolished these effects implying that they were mediated by E2. Moreover, pretreatment of OVX myocytes with GPR30 antagonist G15 (100 nM) abolished the effects of E2, whereas ERα and ERβ antagonist ICI 182,780 (1 μM) did not. Treatment of OVX myocytes Corresponding Ca 2+ transient amplitude in male, sham, OVX, and OVX + E in the presence or absence of estrogen receptor agonists and antagonists (n = 10 cells/7 mice). c Effect of ISO-stimulation (control) and Epi (stress) on cAMP in the presence or absence of estrogen receptor agonists and antagonists (n = 10 cells/7 mice). * P < 0.05, ** P < 0.01, && P < 0.01 vs control OVX. All data are presented as mean ± S.E.M with G1 (100 nM) replicated the effects of E2 on contraction and Ca 2+ transient amplitudes suggesting that the acute effects of E2 were mediated by GPR30 via non-genomic signaling. In contrast, in guinea pigs, long-term OVX increased Ca 2+ transient [33]. One possible mechanism to explain this discrepancy is that loss of E2 may have time-dependent effects on the transcription of genes involved in E-C coupling. Nevertheless, they found that OVX decreased fractional shortening in vivo which agrees to our in vitro results. It was previously shown that plasma E2 varies with the estrous cycle in murine [22]. Therefore, we explored the correlation between myocyte contraction and plasma E2 using Spearman's correlation test. As illustrated in Fig. 1, fractional shortening was positively correlated with plasma E2, suggesting that E2 may positively affect contractility. Although we did not confirm whether plasma E2 correlates with intracellular E2 levels in cardiomyocytes, E2  has been shown to affect heart autonomic functions such as heart rate [8]. In addition, James et al. found that I CaL varies with estrous cycle in guinea pigs [34]. Hence, it is likely that plasma E2 affects contractility. LTCC plays an important role in the regulation of contraction and intracellular Ca 2+ . The size of I CaL is proportional to the Ca 2+ transient amplitude [13], but studies have been inconsistent regarding the size of I CaL density between sexes. Hence, we further investigated whether sex differences in contraction and Ca 2+ transient were due to different levels of I CaL density between sexes. Consistent with previous studies, our data demonstrated that I CaL density was larger in Sham myocytes compared to male [27,35,36], suggesting that the larger contraction in female compared to male may be caused by larger I CaL density in female although other studies [5,6,[37][38][39] did not find sex differences in I CaL density. This discrepancy could be due to inter-strain/species differences or I CaL recording conditions (see Table 3). OVX decreased I CaL density which was reversed by E2 (Fig. 2b). This implies that E2 modulated sex differences in I CaL density. Moreover, we performed western blot and Real-time qPCR on LV apical cardiomyocytes from between sexes. As illustrated in Fig. 2 c and d, LTCC protein and mRNA levels were higher in Sham myocytes than in male myocytes suggesting that high I CaL in Sham compared to male was due to the high number of LTCC. Previously, it was found that long-term OVX (8 months) decreased LTCC protein and I CaL density in mice [41]. However, in this study, short-term OVX (14-21 days) decreased LTCC mRNA, protein, and I CaL density, which were reversed by 1 nM E2 treatment. In agreement with these findings, treatment of female cardiomyocytes derived from induced pluripotent stem cells (iPS-CM) with 1 nM E2 increased LTCC mRNA level and I CaL [42]. Similarly, the expression of LTCC protein in rat heart ventricle was found to be higher in female than male [43]. Recently, using H9C2 and rat myocytes, Yang et al. demonstrated that treatment with E2 enhanced I CaL and Cav1.2α1C expression through plasma membrane-bound ER. This finding is in agreement with our study [44].
The cAMP-PKA-LTCC pathway is an important component of cardiomyocyte contraction. To determine whether there are sex differences in intracellular cAMP levels, LV apical myocytes were subjected to cAMP ELISA test. As shown in Fig. 3a, Sham myocytes had higher levels of cAMP compared to male. OVX decreased cAMP while E2 reversed these effects. Our results contrast with those reported by Parks et al. [45]. In their study, contraction and Ca 2+ transient amplitude were higher in OVX than Sham and but no differences were found in basal cAMP between Sham and OVX ventricular myocytes. The differences noted here may be partly due to the type of myocytes used i.e. global ventricular myocytes versus left ventricular apical myocytes in our study. In addition, they investigated the long-term effects of OVX (8 months) which may be different from the short-term OVX (21 days) effects in this study. Taken together, the cAMP results supported the findings that contraction and Ca 2+ transient were higher in Sham than in male myocytes.
We then postulated that differential cAMP breakdown may be responsible for the sex differences in cAMP levels. The effects of non-selective PDE inhibitor (IBMX) were determined functionally in terms of contraction and Ca 2+ transient amplitude. As shown in Fig. 3b-d, IBMX increased contraction and Ca 2+ transient amplitude in male and female myocytes and eliminated the sex differences. These findings imply that cAMP breakdown is different between sexes at the basal state. Further analysis of mRNA level of PDEs (PDE 3A, 3B, 4A, 4B, and 4D [46]) revealed that mRNA levels of these PDEs were higher in male and OVX LV apical myocytes than Sham and OVX + E2 groups (Fig. 4). Collectively, these observations imply that (1) lower cAMP levels in male were due to high cAMP breakdown by PDE compared to female and (2) E2 modulates the mRNA level of PDE. These observations differ from those reported by Parks et al. who found that long-term OVX increased expression of only PDE4A [45]. They also found that cAMP was lower in female due to higher PDE4B than male [31] and no sex differences were found for PDE3 and 4 families. These discrepancies are possibly due to model differences described above.
We further compared the expression of PLB, SLN, SERCA2a, RyR2 in male, Sham and OVX myocytes at the gene level. Figure 5 shows that RyR2 mRNA expression was higher in sham myocytes than male while both SLN and PLB were higher in male compared to Sham. OVX decreased RyR2 and SLN mRNA level but it increased PLB expression and the effects of OVX were reversed by E2. These data indicate that sex differences exist in the expression of PLB, SLN and RyR2, and their expression in female myocytes is regulated by E2. High levels of RyR2 are associated with high Ca 2+ release from sarcoplasmic reticulum [47]. Consequently, a lower level of RyR2 in male myocytes may cause smaller SR Ca 2+ release and hence smaller Ca 2+ transient and contraction amplitude compared to female. Both SLN and PLB are phosphorylated by PKA [48]. The higher SLN and PLB in male myocytes compared to Sham (Fig. 5c) would result in strong superinhibition of SERCA2 and eventually suppress cardiac contractility [49]. Interestingly, no sex differences were observed in the expression of SERCA2a Fig. 5b). But OVX decreased SERCA2a mRNA level compared to Sham. Furthermore, OVX increased PLB while it decreased SLN compared to Sham. These effects of OVX were reversed by 1 nM E2 treatment. Even though the phosphorylation state of these proteins was not examined, their activity is partly dependent on their expression level. Taken together, these results further suggest that low SLN and PLB, high RyR2 in female than male may contribute to a higher contraction in female myocytes compared to male.
Stress-induced cardiomyopathy is predominant among aged postmenopausal women compared to men and premenopausal women [50]. In a previous study [23], a 20-min exposure of male rat cardiomyocytes to Epi decreased contractile response to ISO due to signal trafficking from β 2 AR-Gs to β 2 AR-Gi. Here, we established the stress model in left ventricular apical mice myocytes. The results showed that the β 2 AR-mediated decline in contraction was observed at 40 min (Fig. 6b) and was absent in β 2 KO myocytes. The difference in the time frame of Epi effects may be explained by differential βAR coupling to Gs and Gi as reported in different species [51]. This study also revealed sex-specific responses to stress. Epi treatment decreased contraction and Ca 2+ transient amplitude only in Wt male and OVX myocytes but not in Wt sham and OVX + E2 myocytes (Fig. 7a). Furthermore, stress has similar effects on cAMP levels (Fig. 7c). Our data implied that E2 acted through GPR30 to suppress the effects of stress which is in line with our previous study [52]. These findings show that female myocytes with normal E2 levels are protected against β 2 AR-mediated stress. This is consistent with a previous study [19], in which E2 modulated cardiac responses to emotional stress. Figure 8 a and b shows that lack of ovarian hormones due to OVX influenced the effects of β 2 AR on the LTTC mRNA and I CaL density. β 2 KO Sham myocytes had decreased I CaL density and LTCC mRNA level compared to wild-type female. Interestingly, OVX decreased both LTCC mRNA level and I CaL density and eliminated the differences between wild-type Sham and β 2 KO Sham myocytes. These results show that estrogen compensates for the loss of function resulting from β 2 AR-gene knockout in female LV apical myocytes. Furthermore, OVX decreased β 2 AR mRNA while E2 normalized its expression (Fig. 8c). However, it should be noted that OVX leads to the loss of all ovarian hormones, besides estrogen. These findings are in line with our previous report that β 2 AR-mediated cardioprotection was depended on the E2 status [53].

Conclusion
This study shows, for the first time, that sex contractile differences exist in LV apical myocytes which are mediated by E2 through modulation of gene transcription in the cAMP-PKA-LTCC pathway. We also reveal that E2 confers cardioprotection against β 2 AR-derived signals during cardiac stress. The differential cAMP breakdown in left ventricular apical myocytes (between sexes) contributes to differences in contraction and responses to stress. Equally important are the sex differences in the expression of RyR2, PLB and SLN. The data presented here have revealed new concepts that are likely to be important for additional investigations into the roles of β 2 AR and E2 (and its receptors) in cardiac contraction. Given the current findings at single myocytes level, we are currently carrying out further studies at tissue and organ levels to validate these results and compare contraction in basal left ventricular myocytes. It will be interesting to determine the transcriptional mechanism by which E2 influenced the expression of the genes studied in this work. Nonetheless, this study provides important findings on the role of estrogen in cardiac physiology and sex-dependent cardiomyopathies.