Estrogen but not testosterone preserves myofilament function from doxorubicin-induced cardiotoxicity by reducing oxidative modifications

Abstract

Here, we aimed to explore sex differences and the impact of sex hormones on cardiac contractile properties in doxorubicin (DOX)-induced cardiotoxicity. Male and female Sprague-Dawley rats were subjected to sham surgery or gonadectomy and then treated or untreated with DOX (2 mg/kg) every other week for 10 wk. Estrogen preserved maximum active tension (T_max) with DOX exposure, whereas progesterone and testosterone did not. The effects of sex hormones and DOX correlated with both altered myosin heavy chain isoform expression and myofilament protein oxidation, suggesting both as possible mechanisms. However, acute treatment with oxidative stress (H₂O₂) or a reducing agent (DTT) indicated that the effects on T_max were mediated by reversible myofilament oxidative modifications and not only changes in myosin heavy chain isoforms. There were also sex differences in the DOX impact on myofilament Ca²⁺ sensitivity. DOX increased Ca²⁺ sensitivity in male rats only in the absence of testosterone and in female rats only in the presence of estrogen. Conversely, DOX decreased Ca²⁺ sensitivity in female rats in the absence of estrogen. In most instances, this mechanism was through altered phosphorylation of troponin I at Ser²³/Ser²⁴. However, there was an additional DOX-induced, estrogen-dependent, irreversible (by DTT) mechanism that altered Ca²⁺ sensitivity. Our data demonstrate sex differences in cardiac contractile responses to chronic DOX treatment. We conclude that estrogen protects against chronic DOX treatment in the heart, preserving myofilament function.

NEW & NOTEWORTHY We identified sex differences in cardiotoxic effects of chronic doxorubicin (DOX) exposure on myofilament function. Estrogen, but not testosterone, decreases DOX-induced oxidative modifications on myofilaments to preserve maximum active tension. In rats, DOX exposure increased Ca²⁺ sensitivity in the presence of estrogen but decreased Ca²⁺ sensitivity in the absence of estrogen. In male rats, the DOX-induced shift in Ca²⁺ sensitivity involved troponin I phosphorylation; in female rats, this was through an estrogen-dependent mechanism.

INTRODUCTION

In women, a lower premenopausal risk but an increased postmenopausal risk of cardiovascular disease compared with men suggests differential regulatory and protective actions of female and male sex hormones on the heart (5, 31, 44, 58). Expression of both estrogen (23, 25) and androgen (34) receptors in cardiac tissues of both sexes supports the differential impact of sex hormones. The physiological significance of female and male sex hormones on cardiac contractile function have been well documented (14, 18, 20, 54, 56, 57). While both gonadectomized female and male rats show reductions in maximum cardiac contractile activation, myofilament Ca²⁺ hypersensitivity is only observed in the ovariectomized (OVX) rat heart (54). In contrast, changes in the Ca²⁺ transient, including suppressed amplitude and prolonged decay time, are similarly detected in both female and male rat hearts after sex hormone deprivation (9, 18, 32, 56).

Sex differences in cardioprotective actions have also been reported in many pathological conditions. In myocardial infarction, female sex hormones increase ejection fraction and reduce left ventricular dimension (15), but male sex hormones suppress cardiac contractility as well as the Ca²⁺ transient amplitude (43). Male sex hormones have also been found to increase the cardiac rupture rate and infarct expansion index (16). In an ischemia-reperfusion (I/R) model, female sex hormones reduce the Ca²⁺ transient amplitude (7) but induce recovery of aortic flow, cardiac output, and contractility (28) as well as prevent cardiomyocyte apoptosis (27). On the other hand, male sex hormones increase the Ca²⁺ transient amplitude but desensitize cardiac myofilament Ca²⁺ sensitivity (7).

Besides differences in cardiac physiology and pathology, sex differences in drug-induced cardiotoxicity, especially by certain chemotherapeutic drugs, have also been identified (1, 24, 36, 37). Clinical observations have noted that adult men develop more functional toxicity to antracycline therapy [doxorubicin (DOX)] than age-matched women (6, 39). Conversely, prepubertal girls and postmenopausal women who have low levels of female sex hormones are more sensitive to DOX-induced cardiotoxicity than age-matched male patients with cancer. This notion supports the cardioprotective effect of female sex hormones (13, 29, 30, 46). The sex difference effects on cardiac physiological and pathological actions as well as drug-induced toxicity may be attributable to changes at the levels of myofilament and/or Ca²⁺ handling.

In the present study, we aimed to study sex differences and the protective effect of sex hormones on cardiac contractile responses to chronic DOX treatment. We used gonadectomized female and male rats with and without hormone supplementation for 10 wk. Changes in cardiac contractile functions were analyzed using skinned fiber and myocyte preparations. Our results demonstrate the potential for female sex hormones, especially estrogen, to protect against DOX-induced depression of myofilament function, whereas testosterone does not.

MATERIALS AND METHODS

Animal experiments.

Female Sprague-Dawley rats weighing between 180 and 200 g (7–8 wk old) and male Sprague-Dawley rats weighing between 250 and 300 g (7–8 wk old) were housed in a plastic shoebox cage under a 12:12 light-dark cycle with controlled temperature and humidity. They were fed ad libitum with laboratory rat chow (C. P.) during the whole period of the experiment. Female and male rat hearts were examined under physiological and chronic DOX treatment. To investigate the impact of sex hormones as a key modulator underlying the differences in responses, female rats were subdivided into the following four groups: 1) sham-operated (sham) control, 2) OVX, 3) OVX with estrogen (E) supplementation, and 4) OVX with progesterone (P) supplementation. Male rats were subdivided into the following three groups: 1) sham control, 2) orchidectomy (ORX), and 3) ORX with testosterone supplementation. The animal experimental protocol was approved by the Experimental Animal Committee of the Faculty of Science of Mahidol University in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (2011).

Hormone replacement and DOX injection.

The hormone supplementation groups of OVX rats were subcutaneously injected with either 5 µg of 17β-estradiol or 1 mg of progesterone using corn oil as a solvent in a fixed volume of 0.1 ml for 3 times/wk for 10 wk starting 2 days after the operation, as previously described (42). An equal volume of corn oil was injected in the control group of female rats. In the supplementation group of male rats, ORX rats were subcutaneously injected with testosterone propionate in ethyl oleate (2 mg/kg body wt) for 3 times/wk adjusted from 2.5 mg/kg used in a previous study (56). Ethyl oleate was injected in the control groups. The groups with DOX treatment received an intraperitoneal injection of 2.0 mg/kg body wt every other week in 10 wk. Thus, the accumulative dose of DOX was 10 mg/kg body wt (8, 17) throughout the period of study. In a vehicle-treated group, rats received normal saline solution in the same volume as that injected in DOX-treated groups.

Measurement of myofilament Ca²⁺ activation.

Left ventricular papillary muscle was placed in ice-cold high relaxing buffer containing 10 mM EGTA, 2 mM free Mg²⁺, 5 mM MgATP, 79.2 mM KCl, 12 mM creatine phosphate, 20 mM MOPS (pH 7.0) (ionic strength: 0.15 M), 2.5 µg/ml pepstatin A, 1 µg/ml leupeptin, and 50 µM PMSF. Small muscle bundles were dissected for force measurements as previously described (55). Briefly, stripped papillary fibers were skinned in high relaxing buffer containing 1% Triton X-100 for 1 h at 25°C. The skinned fiber bundle was attached using aluminum T clips at one end to a displacement generator and at the other end to a force transducer (KG-7). Active tension was measured at a fixed sarcomere length of 2.2 μm in a solution containing various Ca²⁺ concentrations ranging from pCa 7.0 to 4.5 at pH 7.0 and 20°C (26). For the H₂O₂ treatment experiment, a new skinned fiber bundle was directly introduced to 5 mM H₂O₂ in the active buffer containing various Ca²⁺ concentrations ranging from pCa 7.0 to 4.5. The cross-sectional area of the fiber bundle was calculated based on an elliptical model. pCa-active tension relationships were fit to a modified Hill equation using nonlinear least-squares regression analysis (GraphPad Prism, version 5) to derive maximum tension (T_max), half-maximal activating Ca²⁺ concentration (pCa₅₀; an index representing myofilament Ca²⁺ responsiveness), and the Hill coefficient.

Measurement of myofilament Ca²⁺ activation under DTT treatment.

DTT treatment was performed in single skinned cells. Briefly, left ventricular tissue was placed in ice-cold high relaxing buffer containing 1 mM DTT and homogenate to generate single cell fragments as previously described (51). Cells were next placed in ice-cold high relaxing buffer containing 1% Triton X-100 for 30 min at 25°C (55). The skinned cells continued to stay in ice-cold high relaxing buffer with 1 mM DTT, and active force was then measured by attaching one end to a displacement generator and at the other end to a force transducer. Active tension was measured at a fixed sarcomere length of 2.2 μm in a solution containing various Ca²⁺ concentrations ranging from pCa 7.0 to 4.5 at room temperature. Isometric tension was recorded on a chart recorder. Isometric tension measurements were plotted as a function of pCa and fit by a nonlinear least-squares regression analysis to the Hill equation using GraphPad Prism 5. From this fitted curve, we derived the T_max, pCa₅₀, and the Hill coefficient.

OxyBlot procedure.

Oxidized myofilament proteins were determined using Western blot analysis and anti-dinitrophenyl (DNP) antibodies. The protein carbonyl side chains were derivatized to DNP by reaction with 2,4-dinitrophenylhydrazine. Samples were run on 12% SDS-PAGE at 4°C with dinitrophenylated protein molecular weight standards (Intergen). Proteins were transferred onto 0.45-µm pore size polyvinylidene difluoride membranes. Membranes were blocked and incubated with specific anti-DNP (1:100, OxyBlot kit, Millipore). Blots were washed, incubated with peroxidase-labeled anti-rabbit IgGs (1:300 dilution), and developed using a chemiluminescence detection system. The ratio between densitometric values of the OxyBlot bands and those of the corresponding bands stained with Coomassie blue was used as an index of oxidized proteins.

Immunoblot analysis.

The ratio of phosphorylated troponin I (pTnI) to total troponin I (TnI) was evaluated using immunoblot near-infrared fluorescence analysis. Myofilament proteins were purified, run on 10% SDS-PAGE, and then transferred to the nitrocellulose membrane. The membrane was blocked for 1 h in Odyssey blocking buffer and incubated overnight with specific antibodies to pTnI (Ser²³/Ser²⁴). Protein quantification was performed using a Li-Cor Odyssey automated infrared imaging system according to the manufacturer’s instructions. Quantification of pTnI and TnI was accomplished using Image Studio (version 4.0). Myosin heavy chain (MHC) isoforms were separated electrophoretically with a 6.5% SDS-polyacrylamide gel as previously described (35). After samples had been stained with Coomassie blue dye, the relative amount of α-MHC to total MHC was analyzed.

Statistical analysis.

Data are presented as means ± SE. Two-way ANOVA was used to determine the global effects of sex and DOX-induced cardiotoxicity followed by a Student-Newman-Keuls post hoc test (GraphPad PRISM7). A paired t-test was applied to study direct effects of H₂O₂ on the tension development of cardiac myofilaments. Values of P < 0.05 were considered as a significant difference between and among groups.

RESULTS

Sex differences in DOX-induced changes in myofilament function.

The in vivo characteristics of the female and male groups are shown in Table 1 and Table 2, respectively. Of the male rats treated with DOX, 80% had ascites, enlarged and pale livers, swollen kidneys, and spleens, and half suffered from diarrhea. Interestingly, these symptoms only occurred in 15% of female rats treated with DOX. While there was a protective effect of estrogen on body weight after DOX exposure, neither progesterone nor testosterone had any protective effect. However, there was no change in the percentage of heart weight normalized to body weight in rats treated with DOX compared with control rats. While all DOX-untreated groups of both sexes survived well throughout the term of study, DOX treatment decreased the percent survival in both groups (Fig. 1). DOX treatment induced later onset of mortality in female rats than in male rats. In addition, mortality in DOX-exposed groups of female OVX rats and female OVX rats supplemented with progesterone was ~25%, whereas ~40% mortality was observed in all male groups.

Table 1.

Body, heart, uterine, and lung weights from sham control and OVX rats without and with estrogen or progesterone supplementation of the DOX-untreated/treated rat set

	Without DOX				With DOX
Parameters	Sham	OVX	OVX + E	OVX + P	Sham	OVX	OVX + E	OVX + P
Body weight, g	272 ± 3	370 ± 9*	240 ± 5	367 ± 7*	270 ± 4	320 ± 7*	236 ± 5	323 ± 6*
Heart weight, g	1.04 ± 0.03	1.33 ± 0.03*	0.92 ± 0.03	1.31 ± 0.04*	1.05 ± 0.04	1.19 ± 0.04	0.93 ± 0.03	1.17 ± 0.03
Heart weight/body weight, %	0.38 ± 0.01	0.36 ± 0.01	0.38 ± 0.01	0.36 ± 0.01	0.39 ± 0.01	0.37 ± 0.02	0.39 ± 0.01	0.36 ± 0.01
Uterine weight, g	0.55 ± 0.05	0.15 ± 0.01*	0.57 ± 0.03	0.14 ± 0.01*	0.55 ± 0.03	0.13 ± 0.01*	0.57 ± 0.03	0.14 ± 0.01*
Lung dry/wet weight, %	19.3 ± 0.4	20.7 ± 0.8	19.3 ± 0.7	19.8 ± 0.4	20.1 ± 0.4	20.4 ± 0.4	20.1 ± 0.5	20.7 ± 0.4

Values are means ± SE; n = 7–8 rats/group. DOX, doxorubicin; sham, sham operated; E, estrogen; P, progesterone; OVX, ovarectomized.

^*Significantly different (P < 0.05) from the sham control group of the same treatment using a Student-Newman-Keuls test after ANOVA.

Table 2.

Body, heart, uterine, and lung weights from sham control and ORX rats without and with testosterone supplementation of the DOX-untreated/treated rat set

	Without DOX			With DOX
Parameters	Sham	ORX	ORX + T	Sham	ORX	ORX + T
Body weight, g	498 ± 11	482 ± 9	482 ± 11	436 ± 8†	424 ± 9†	422 ± 11†
Heart weight, g	1.73 ± 0.04	1.69 ± 0.06*	1.69 ± 0.03	1.45 ± 0.02†	1.41 ± 0.03†	1.38 ± 0.02†
Heart weight/body weight, %	0.35 ± 0.01	0.35 ± 0.01	0.35 ± 0.01	0.33 ± 0.01	0.33 ± 0.01	0.33 ± 0.01
Seminal vesicular weight, g	0.72 ± 0.01	0.09 ± 0.01*	0.73 ± 0.02	0.40 ± 0.01†	0.04 ± 0.01*	0.53 ± 0.04†
Lung dry/wet weight, %	21.5 ± 0.4	21.2 ± 0.3	21.1 ± 0.3	21.4 ± 0.1	21.9 ± 0.3	21.6 ± 0.3

Values are means ± SE; n = 8 rats/group. DOX, doxorubicin; sham, sham operated; T, testosterone; ORX, orchidectomized.

^*P < 0.05 from the sham control group of the same treatment;

^†P < 0.05 from the DOX-untreated sham group using a Student-Newman-Keuls test after ANOVA.

Fig. 1.

Effects of sex hormone deprivation and doxorubicin (DOX) treatment on percent survival. A: percent survival of female DOX-treated sham-operated (sham) control and ovarectomized (OVX) rats with and without estrogen (E)/progesterone (P) supplementation. B: percent survival of male DOX-treated sham control and orchidectomized (ORX) rats with and without testosterone (T) supplementation. *P < 0.05 from the DOX-treated sham control group by Kaplan-Meier analysis.

To elucidate sex differences in cardiac contractile function, active tension-Ca²⁺ relationships were examined in skinned papillary fibers. In female rats, T_max was not significantly reduced by DOX exposure, although after OVX (which reduced T_max by 31% on its own), DOX decreased T_max by 55% (Fig. 2A). This suggests that estrogen protects against DOX-induced depression of T_max. This was not the case in male rats, where the effect of ORX, DOX, and ORX + DOX on T_max were all similar (40%, 49%, and 42% decreases, respectively; Fig. 2B). The protective effect of estrogen, but not testosterone, on DOX-induced changes in T_max was further supported by treatment with sex hormones after gonadectomy. Reintroduction of estrogen was able to recover T_max (Fig. 2C), whereas testosterone was not (Fig. 2D). Summary data for T_max are shown in Fig. 2, E and F, and also show that progesterone in female rats, like testosterone in male rats, had no protective effect.

Fig. 2.

Effects of sex hormone deprivation and doxorubicin (DOX) treatment on skinned papillary function. A: mean active tension as a function of pCa and fitted curves from skinned left ventricular (LV) fibers from female sham-operated (sham) and ovarectomized (OVX) rats with and without DOX treatment. B: mean active tension as a function of pCa and fitted curves from skinned LV fibers from male sham and orchidectomized (ORX) rats with and without DOX treatment. C: mean active tension as a function of pCa and fitted curves from skinned LV fibers from female sham, OVX + DOX, and OVX + DOX groups with estrogen treatment (OVX + DOX + E₂). D: mean active tension as a function of pCa and fitted curves from skinned LV fibers from male sham, ORX + DOX, and ORX + DOX groups with testosterone treatment (ORX + DOX + T). E and F: summary data for maximum active tension (T_max). G and H: summary data for Ca²⁺ sensitivity (pCa₅₀). Data are means ± SE of 10–16 fibers from 7–8 hearts/group. *P < 0.05 from the sham control groups of the same treatment; #P < 0.05 from the DOX-untreated sham groups using a Student-Newman-Keuls test after two-way ANOVA.

In addition, there were sex differences in DOX-induced effects on Ca²⁺ sensitivity in rat hearts. In female rats, DOX and OVX both increased Ca²⁺ sensitivity independently (Fig. 2G); however, in combination, OVX and DOX treatment had no effect on Ca²⁺ sensitivity. Thus, the DOX-induced increase in Ca²⁺ sensitivity is dependent on the presence of either estrogen or progesterone. Male rats showed the opposite, where DOX only caused an increase in Ca²⁺ sensitivity in the absence of the sex hormone (Fig. 2H).

From these results, we identified sex differences in the impact of DOX exposure on myofilament function, specifically T_max and Ca²⁺ sensitivity. We explored the mechanisms of these differences independently, focusing first on T_max and then on Ca²⁺ sensitivity.

Mechanisms of sex differences in the DOX-induced decrease in T_max.

A shift in the MHC isoform from the predominant α-isoform toward more β-isoforms (the slow isoform) may be one mechanism underlying the suppressed T_max in both hormone deprivation and DOX treatment conditions. Female sex hormone deprivation and DOX treatment significantly decreased the ratio of α-MHC to β-MHC in ventricular tissue compared with control animals (Fig. 3A). Estrogen prevents the isoform switching of MHC and preserves the suppressive effect of female sex hormone deficiency and DOX-induced toxicity on T_max development of the heart. Male sex hormone deprivation also decreased the ratio of α-MHC to β-MHC, which could be prevented by testosterone supplementation. However, the protective effect of testosterone on MHC isoform switching disappeared under DOX treatment (Fig. 3B). The correlation between T_max and MHC isoform switching (compare Fig. 2, E and A, with Fig. 2, F and B) suggests a role for this mechanism.

Fig. 3.

Effects of sex hormone deprivation and DOX treatment on expression of myosin heavy chain (MHC) isoforms. A: relative amount of α-MHC as a percentage of total (α+β) MHC of left ventricular homogenates from each female group. Inset: example gel for each of the eight groups. B: same as in A but for male groups. E, estrogen; P, progesterone; T, testosterone; OVX, ovarectomized; ORX, orchidectomized. Data are means ± SE from 4–5 hearts/group. *P < 0.05 from the sham control group of the same treatment; #P < 0.05 from the doxorubicin (DOX)-untreated sham group using a Student-Newman-Keuls test after two-way ANOVA.

However, DOX induces significant oxidative stress as well, so we also examined the oxidation level of the myofilaments using OxyBlot analysis. Myofilament oxidation was higher in OVX and DOX-exposed OVX rats (Fig. 4, A and B), a result that implies a higher level of oxidative stress in both groups. Progesterone did not reduce oxidative modifications in DOX-exposed OVX rats. In contrast, estrogen reduced myofilament protein oxidation in both OVX and DOX-exposed OVX groups. The oxidation level was not changed in male groups without DOX treatment (Fig. 4, C and D). However, the myofilament oxidation level was enhanced in DOX-treated male groups. These findings indicate that testosterone does not prevent oxidative modifications of myofilament proteins, whereas estrogen protects against myofilament oxidative modifications. Therefore, similar to MHC isoform shifting, these findings also show a correlation between T_max and oxidative modifications of myofilament proteins.

Fig. 4.

Effects of sex hormone deprivation and doxorubicin (DOX) treatment on carbonylation levels of cardiac myofilament proteins. A: representative OxyBlot (left) and Coomassie-stained (right) gels demonstrating carbonylation of various myofilament proteins in left ventricular homogenates from each group of female rats. B: summary data showing the relative amount of carbonylated proteins to total protein from each group of female rats. C and D: same as in A and B but for male rat groups. MW, molecular weight standard; E, estrogen; P, progesterone; T, testosterone; OVX, ovarectomized; ORX, orchidectomized; MHC, myosin heavy chain; MyBP, myosin-binding protein; TnT, troponin T; Tm, tropomyosin; MLC, myosin light chain. Data are means ± SE from 4–5 hearts/group. *P < 0.05 from the sham control group of the same treatment; #P < 0.05 from the DOX-untreated sham group using a Student-Newman-Keuls test after two-way ANOVA.

We next wanted to determine the contribution of each of these mechanisms (MHC isoform shifting or myofilament oxidative modifications) on the observed functional impact of gonadectomy and DOX on T_max. To assess whether the impairment of myofilament function is due to a direct action of reactive oxygen species (ROS), we measured myofilament tension-Ca²⁺ relationships with and without 5 mM H₂O₂. As shown in Fig. 5, A and B, T_max was decreased by H₂O₂ exposure in both female and male rats without DOX treatment. There was no further decrease in T_max in both OVX and ORX groups with and without DOX treatment. This implies that oxidative modifications of myofilament proteins resulting in decreased T_max had already occurred before the addition of H₂O₂ in those groups.

Fig. 5.

Direct effects of H₂O₂ and DTT on maximum active contraction in rat hearts. A: maximum tension (T_max) in skinned left ventricular fibers in the absence and presence of 5 mM H₂O₂ from each group of female rats. B: same as in A but for the male rat groups. DOX, doxorubicin; E, estrogen; P, progesterone; T, testosterone; OVX, ovarectomized; ORX, orchidectomized. Data are means ± SE of 10–16 fibers from 7–8 hearts/group. *P < 0.05 from the H₂O₂-untreated group; #P < 0.05 from the sham control group of the same treatment; $P < 0.05 from the DOX-untreated sham group using a Student-Newman-Keuls test after two-way ANOVA. C: T_max in skinned cardiac cells from sham groups of female and male rats in the absence/presence of 1 mM DTT treatment. Data are means ± SE of 15–16 single skinned cells from 5 hearts/group. *P < 0.05 from the DOX-untreated sham control group using a Student-Newman-Keuls test after two-way ANOVA.

To further confirm whether the observed decreases in T_max were caused by oxidative modifications of myofilament proteins, we exposed single skinned myocytes to 1 mM DTT to determine whether the effects on T_max could be reversed by a reducing agent. The suppression of T_max from 13.9 ± 0.7 mN/mm² in the male sham control group to 6.6 ± 0.3 mN/mm² in the DOX-treated group could be completely reversed by DTT (14.4 ± 0.7 mN/mm²; Fig. 5C). As female rats were already protected from oxidative modifications, DTT showed no additional effect.

Together, these results suggest that the observed changes in T_max induced by DOX exposure, sex hormone deprivation, or a combination are directly due to oxidative modifications of myofilament proteins.

Mechanisms of sex differences in DOX induces effects on myofilament Ca²⁺ sensitivity.

Interestingly, H₂O₂ (Fig. 6, A and B) and DTT (Fig. 6C) did not alter Ca²⁺ sensitivity compared with control fibers in both male and female groups. This suggests that the alteration in myofilament Ca²⁺ sensitivity induced by DOX treatment are caused by other modifications of myofilament proteins.

Fig. 6.

Effects of H₂O₂ and DTT on myofilament Ca²⁺ sensitivity in rat hearts. A: pCa₅₀ in skinned left ventricular fibers in the absence and presence of 5 mM H₂O₂ from each group of female rats. B: same as in A but for the male rat groups. DOX, doxorubicin; E, estrogen; P, progesterone; T, testosterone; OVX, ovarectomized; ORX, orchidectomized. Data are means ± SE; n = 12–16 fibers from 7–8 hearts/group. #P < 0.05 from the DOX-untreated sham group using a Student-Newman-Keuls test after two-way ANOVA. C: pCa₅₀ of skinned cardiac cells from sham groups of female and male rats with and without 1 mM DTT. Data are means ± SE of 15–16 cells from 5 hearts/group. *P < 0.05 from the sham control group of the same treatment using a Student-Newman-Keuls test after two-way ANOVA.

Myofilament protein phosphorylation is a well-established modulator of Ca²⁺ sensitivity, and so we investigated this mechanism in myofilament protein fractions of heart tissue from each group. We used Western blot analysis to probe TnI phosphorylation directly. In female rats, we observed an increase in Ser²³/Ser²⁴ TnI phosphorylation in the OVX + DOX group; in male rats, we observed increases in DOX and DOX + ORX + testosterone groups (Fig. 7, A and B). There were no changes in TnI phosphorylation status in the absence of DOX exposure. Thus, the TnI phosphorylation status explains the effects of DOX and ORX in male rats but only the effect of DOX in female rats in the absence of estrogen. There is therefore an additional, estrogen-dependent mechanism by which DOX alters Ca²⁺ sensitivity in female rats.

Fig. 7.

Effects of sex hormone deprivation and doxorubicin (DOX) treatment on cardiac troponin I (TnI) phosphorylation. A: quantification and representative immunoblot (above) of phosphorylated TnI at Ser²³/Ser²⁴ (pS23/24 TnI) and total TnI in female groups. B: same as in A but for the male groups. p/T TnI, phosphorylated to total TnI; E, estrogen; P, progesterone; T, testosterone; OVX, ovarectomized; ORX, orchidectomized. Data are means ± SE from 5 hearts/group. *P < 0.05 from sham control group of the same treatment; #P < 0.05 from the DOX-untreated sham group using a Student-Newman-Keuls test after two-way ANOVA.

DISCUSSION

In the present study, we revealed sex differences in cardiac contractile responses to chronic DOX-induced cardiotoxicity. First, specifically, estrogen, but not testosterone or progesterone, protects against DOX-induced depression of myofilament maximum Ca²⁺-activated tension. Second, while these changes correlated with shifting MHC isoforms, the primary mechanism by which estrogen protects against DOX-induced depressed T_max is by inhibiting or reversing oxidative modifications of myofilament proteins. Third, DOX increased Ca²⁺ sensitivity in female rats only in the presence of estrogen and progesterone but increased Ca²⁺ sensitivity in male rats only in the absence of testosterone. Finally, DOX alters Ca²⁺ sensitivity in a phospho-Ser²³/phospho-Ser²⁴ TnI-dependent manner in male rats but only in the absence of estrogen in female rats (the mechanism is phosphorylation independent in the presence of estrogen).

In agreement with previous studies from our group and others (18, 20, 52, 54, 56), a reduction in T_max was observed with both male and female sex hormone deprivation. However, with chronic DOX treatment, only estrogen could maintain T_max. Thus, in male rats, T_max decreased by 40% after 10 wk of DOX treatment, in agreement with a previous study (8). A proposed mechanism for this decrease in T_max is a change in the relative levels of α-MHC and β-isoform of MHC (α-MHC and β-MHC, respectively) (14, 40, 52, 57), which also decreases the rate of force redevelopment (k_tr) (19). However, our data here suggest that MHC isoform shifting is only one component, and we propose an alternative or additional mechanism: oxidative modifications of myofilament proteins in OVX female and male groups with DOX treatment.

It is unclear, however, whether the protective effects of estrogen on T_max occur by abrogating the oxidative stress component of DOX exposure or by reversing the oxidative modifications of the proteins themselves. Metabolism of quinone-containing compounds like DOX generally leads to generation of free radicals in the microsomal NADPH-oxidase system (50). Due to the quinone moiety in the tetracyclic aglycone molecule, DOX can form many free radicals by cycling from quinone to semiquinone and then back to quinone, leading to the production of superoxide anions from oxygen. DOX has been reported to increase the peroxidation of polyunsaturated fatty acids within membranes as well as free oxygen radical activity. Estrogen has also been revealed as an effective antioxidant in fatty acid and cholesterol peroxidation (4) as well as in maintaining glutathione peroxidase, catalase, and SOD levels (38). Thus, estrogen might then be able to attenuate oxidative stress-induced cardiotoxicity directly.

The redox-dependent changes in contractile properties are determined by the source and type of the oxidant species, the level of oxidative stress, and the residues modified in individual myofilament proteins (48). There are many types of oxidative modifications of sarcomeric proteins induced by I/R or end-state human heart failure, including the reaction of a cysteine thiolate anion with H₂O₂ (S-sulfhydration), with nitric oxide (S-nitrosylation), or with GSH (S-glutathionylation). Moreover, ROS and peroxynitrite (ONOO⁻) can promote protein carbonylation, the addition of a carbonyl group to Lys, Arg, or Pro residues (11, 49). However, few of these have been studied with DOX exposure.

There are several oxidative modifications that have been shown to affect T_max, which may be involved here. A study in skinned rat cardiac trabeculae found that superoxide anion (O₂⁻) (33) and H₂O₂ (2) suppress T_max without altering Ca²⁺ sensitivity, similar to what we found here. S-nitrosocysteine can also significantly decrease maximal isometric force, although also with a concomitant decrease myofilament Ca²⁺ sensitivity (21). Conversely, nitroxyl (HNO) increased force production and Ca²⁺ sensitivity by inducing the formation of dimers between MHC and Cys⁸¹ in myosin light chain-1 (22). S-nitrosylation of α-tropomyosin at Cys¹⁹⁰ (in the overlap region) in human heart failure yields an interchain disulfide cross-link and alters its flexibility and interferes with its interaction with other thin filament proteins (10, 12, 53). These previous findings suggest that we are observing an oxidative modification that reversibly impacts T_max but does not interfere with protein phosphorylation, such as S-glutathionylation of cardiac myosin-binding protein C (cMyBP-C), which attenuates cMyBP-C phosphorylation by protein kinases (47).

In addition to sex differences in T_max, we observed differences in myofilament Ca²⁺ sensitivity, although the effect was complicated. DOX reduced Ca²⁺ sensitivity in OVX rat hearts concurrent with an increase in TnI Ser²³/Ser²⁴ phosphorylation. However, estrogen enhanced Ca²⁺ sensitivity in response to DOX-induced oxidative stress independently of any change in TnI phosphorylation. This suggests an estrogen-dependent, DOX-induced, irreversible (by DTT) oxidative modification that increases Ca²⁺ sensitivity. Previous studies have identified specific oxidant species and oxidative modifications that increase Ca²⁺ sensitivity that could be responsible for the changes observed here. For example, they are MHC and Cys⁸¹ in myosin light chain-1 dimer formation (induced by HNO) (22), tropomyosin oxidation at Cys¹⁹⁰ detected in female mice early postmyocardial infarction (3), and S-glutathionylation of cMYBP-C at Cys⁴⁷⁹, Cys⁶²⁷, and Cys⁶⁵⁵ (identified in female mice treated with GSSG) (41). Moreover, sex-dependent myocardial S-nitrosothiol proteomes after I/R have recently been reported. Estrogen increases endothelial nitric oxide synthase expression and phosphorylation, resulting in increased S-nitrosylated proteins in female hearts, which likely contribute to sex-dependent cardioprotection (45). These studies suggest the possibility of female sex hormones modifying some aspects of the contractile machinery in a very specific manner that promotes myofilament Ca²⁺ sensitivity.

It is also possible that the increase in Ca²⁺ sensitivity with DOX in the presence of estrogen is a compensatory mechanism to offset the decrease in T_max. While this would help to maintain force production, it could also provoke arrhythmias and diastolic dysfunction.

In conclusion, the results from the present study reveal sex differences in the cardiac contractile response to chronic DOX treatment. Our results support a cardioprotective role for estrogen but not testosterone. These findings expand our understanding of the impact of sex hormones on cardiac contractile function and provide a basis for further preventive and/or therapeutic approaches to cardiac damage induced by DOX.

GRANTS

This work was supported by grants from Mahidol University (to J. Wattanapermpool), National Heart, Lung, and Blood Institute Grants HL-62426 (to P. P. de Tombe) and HL-136737 (to J. A. Kirk), Royal Golden Jubilee PhD Program Grant PHD/0233/2553 (to C. Rattanasopa and J. Wattanapermpool), and American Heart Association Grant 14SDG20380148 (to J. A. Kirk).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

C.R., J.A.K., and J.W. conceived and designed research; C.R. and M.P. performed experiments; C.R. and J.A.K. analyzed data; C.R., J.A.K., P.P.d.T., and J.W. interpreted results of experiments; C.R. and J.A.K. prepared figures; C.R. and J.A.K. drafted manuscript; C.R., J.A.K., T.B.-I., M.P., P.P.d.T., and J.W. edited and revised manuscript; J.A.K. and J.W. approved final version of manuscript.