OUP user menu

Adjunctive 17β-estradiol administration reduces infarct size by altered expression of canine myocardial connexin43 protein

Tsung-Ming Lee, Mei-Shu Lin, Tsai-Fwu Chou, Chang-Her Tsai, Nen-Chung Chang
DOI: http://dx.doi.org/10.1016/j.cardiores.2004.03.009 109-117 First published online: 1 July 2004


Background: Traffic of potentially harmful cytosolic messengers through gap junctions might cause increased injury during ischemia. The present study was to determine whether the infarct size-reducing effect of adjunctive estradiol administration prior to reperfusion is associated with an attenuated expression of connexin43 at the border of infarction in a canine model. Methods: Experiments were performed in 48 dogs (n=16 each group), assigned to receive either vehicle (control group), 17β-estradiol administered before coronary occlusion (early group), or 3 min before coronary reperfusion following 60-min ischemia (late group). Changes in the amount of phosphorylated connexin43 were measured by Western blot. Results: Infarct size was significantly larger in the control (38±7% of area at risk) than in the supplemented groups (16±6% in the early group; 16±8% in the late group, P<0.0001, both). Reperfusion caused a significant elevation in free radicals as measured by lucigenin-derived chemiluminescence. The rise of free radicals was significantly inhibited in animals treated with estrogen, either early or late. The amount of phosphorylated connexin43 was reduced, as assessed by Western blot in control hearts at the border zone. These changes were significantly enhanced by estrogen administration. The magnitude of infarct size positively correlated with the magnitude of phosphorylated connexin43 expression assessed by Western blot (r=0.83, P<0.0001). Confocal microscopy confirmed the changes of junctional complexes. Conclusions: This result demonstrated that the cardioprotective effect of estrogen as an antioxidant may be associated with the reduced amount of phosphorylated myocardial connexin43 protein.

  • Connexin43
  • Contraction band necrosis
  • Estradiol
  • Myocardial infarction

1. Introduction

During the past few years, myocardial reperfusion therapy, such as primary percutaneous transluminal coronary angioplasty or thrombolytic therapy, has been widely performed in the management of acute myocardial infarction. Although restoration of blood flow arrests the progression of necrosis, paradoxically it is accompanied by functional derangement, including an increase in infarct size [1,2]. Ischemia–reperfusion is thought to increase intracellular calcium concentrations by either increasing inward flux of calcium or inhibiting intracellular calcium sequestration [3]. The increased cytoplasmic calcium concentration has been shown to form contraction band necrosis and extent the infarct size [4]. Jovanovic et al. [5] have shown that intracellular calcium overload may have occurred through extracellular Ca2+ influx during ischemic phase and release of Ca2+ from intracellular stores during reperfusion phase.

Connexin43 (Cx43) is the 43-kDa member of a conserved family of membrane spanning gap junction proteins, and is the principal junctional protein in mammalian myocardium [6]. Gap junctions mediate cell-to-cell movement of Ca2+ ions, which may induce calcium overloading and increase contraction band necrosis during reperfusion [7]. Cx43 is a phosphoprotein [6]. Changes in phosphorylation can affect channel function and properties [8]. An increase in dephosphorylated Cx43 contributed to electrical uncoupling at the gap junction during acute myocardial ischemia [8]. Previous studies have demonstrated that trafficking of potentially harmful cytosolic messengers between ischemic cells and surrounding nonischemic cells might cause increased injury during ischemia [9], leading to myofibrillar hypercontracture and further precipitating cell death [10]. Gap junction uncouplers which induced attenuated amount of Cx43 protein and increased dephosphorylation [11] have been reported to exert a beneficial effect in ischemia–reperfusion models both in the myocardium [12,13] and in the brain [14,15].

We [16,17] and others [18] have previously demonstrated that estrogen can provide cardioprotection against infarct sizes in animals undergoing ischemia–reperfusion. However, the cardioprotection has limited the timing of drug administration prior to coronary occlusion in previous studies [16–18]. Because treatment prior to acute myocardial infarction is a virtual impossibility in most clinical situations, there has been a great deal of interest in anti-infarct agents, which do not require pretreatment (adjunctive treatment). Estrogens, especially estriol and 17β-estradiol, which include a phenolic hydroxyl group, have an effective antioxidant action and inhibit lipid peroxidation [19]. We have recently demonstrated that free radicals can modulate the communication of gap junction in hearts [20]. Thus, although there are several reports that estrogen documented cardioprotection in hearts undergoing ischemia–reperfusion, no previous reports examined alterations in Cx43 expression. The aims of the study were (1) to assess the effects on infarct size of adjunctive estrogen administered just prior to coronary reperfusion; (2) to assess whether the cardioprotective effect of estrogen was associated with quantitative and qualitative changes of Cx43 protein expression; and (3) to test whether free radicals act as a mediator of the effect of estrogen on Cx43 protein in an ischemia–reperfusion model. Our results provide evidence that estrogen at physiological concentrations limits myocardial infarction size as an antioxidant, which appears to be associated with the reduced amount of phosphorylated gap junctional protein.

2. Materials and methods

2.1. Preparation

All experiments were conducted on male mongrel dogs, weighing 10–15 kg, preanesthetized with intravenous thiopental at the dose of 5 mg/kg and general anesthesia induced with intravenous pentobarbital sodium. The dogs were intubated and mechanically ventilated (Bennett MA-2) by using room air, delivered at a rate of 20–25 strokes per minute and a tidal volume of 300 ml. The experimental preparation and techniques have been previously described [16]. Fluid replacement, plasma [K+] and [Ca2+], and basic physiological conditions were controlled as described previously [16]. A 22-gauge Teflon catheter was inserted into the great cardiac vein for blood sample collection for measurement of superoxide anion.

Near the base of the heart, the left anterior descending artery proximal to the first diagonal branch was encircled with a 4-0 silk suture. To measure collateral blood flow at baseline and during ischemia, coronary blood flow was detected by intracoronary Doppler flow wire as previously described [16]. All procedures were in accordance with the Guide for the Care and Use of Laboratory Animals (NIH Publication No. 85-23, revised 1996).

2.2. Experimental protocol

The dogs were randomly assigned to one of three groups. All animals were subjected to a 60-min coronary occlusion followed by 120 min of reperfusion. Group I (control) was the control group, and only vehicle (1.0 ml ethanol) was administered prior to the 60-min occlusion. In Group II (early group), animals were treated with intravenous 10 μg/kg of 17β-estradiol (dissolved in 1.0 ml ethanol, Sigma, St. Louis, MO) prior to the 60-min occlusion. In Group III (late group), animals were treated intravenously with 10 μg/kg of 17β-estradiol for 3 min just prior to reperfusion after the 57-min occlusion. The doses were chosen to achieve serum estradiol levels in the range of 200–500 pg/ml, levels equivalent to those in human females during midcycle. Although we excluded dogs with collateral flow >20%, there is evidence that microvascular collaterals may provide blood flow to the periphery of an ischemic region. The drug may have reached the ischemic myocardium via collaterals if the drug is given a fair amount of time before reperfusion. To eliminate the confounding effect of collaterals, we administered estradiol as little as 3 min prior to reperfusion. Because Cx43 undergoes rapid postmortem dephosphorylation within minutes of tissue dissection [8], sham operation (n=3) served as an internal control. Sham dogs underwent the same procedure except the suture was passed under the coronary artery and then removed.

2.3. Measurements of infarct size

At the end of the protocol, the coronary artery was clamped, and 4 ml of a solution of methyl blue dye (Sigma) was injected into the cardiac apex to define the ischemic region at risk of necrosis (uncolored by the blue dye) as described previously [16,17]. The deeply anesthetized dog was then sacrificed by an injection of potassium chloride. The heart was sliced transversely into six to eight sections, and the slices were photographed to record the ischemic areas (uncolored by the blue dye) and the nonischemic, normal areas (perfused blue) in each slice. After a 10-min incubation in a 1% solution of buffered triphenyltetrazolium chloride (TTC), the slices were again photographed to record the necrotic regions (unstained by TTC) and the noninfarcted regions (stained red by TTC). Later the photographic slides were projected and traced. The areas of ischemic and normally perfused regions and the areas of necrotic and nonnecrotic regions were measured on the tracings by computerized planimetry (Image Pro Plus, California) as described previously [16].

2.4. Histology analysis

Extensive histological samples were taken from each transverse section, processed by conventional methods, and stained with hematoxylin and eosin, and Masson for contraction band. A pathologist who was unaware of the treatment protocol examined the samples for microscopic evidence of contraction band necrosis on random fields at a magnification of 400×. In each case the results from the section with the highest number of bands was used. The histological severity of contraction band necrosis was used to grade injury on a scale of 0–3 for the number of contraction bands: 0 (absent), 1 (mild), 2 (moderate) and 3 (severe) as described previously [16]. Interobserver variability for grading contraction band necrosis was determined by having a second observer (TML) re-analyze staining from the original preparations analyzed by the first observer. The calculated coefficients of inter- and intraobserver variations were less than 9% and 5%, respectively.

2.5. Western blot analysis

After staining with TTC, samples of the left ventricle from the border zone were cut transmurally to include all layers from the epi- to the endocardium, were frozen rapidly in liquid nitrogen, and stored at −80 °C until use. To distinguish the border zone, histopathological analysis was performed in sections adjacent to those used for Western blot and confocal microscopy as described previously [20]. A representative TTC- and Masson-stained section of myocardium is shown from a control dog in Fig. 1. Samples were homogenized with a kinametic polytron blender in 100 mM Tris–HCl, pH 7.4, supplemented with 20 mmol/l EDTA, 1 mg/ml pepstatin A, 1 mg/ml antipain, and 1 mmol/l benzamidin. Homogenates were centrifuged at 10,000×g for 30 min to pellet the particulate fractions. The supernatant protein concentration was determined with the BCA protein assay reagent kit (Pierce). Twenty microgram protein was separated by 8% SDS-PAGE and electrotransferred onto a nitrocellulose membrane. After incubation with rabbit polyclonal anti-Cx43 antibodies (Zymed, South San Francisco, CA), the nitrocellulose membrane was then rinsed with a blocking solution and incubated for 2 h at room temperature. The antibody is raised against a segment of the third cytoplasmic domain of rat Cx43 for detection of both phosphorylated and non-phosphorylated forms. Antigen–antibody complexes were detected with 5-bromo-4-chloro-3-indolyl-phosphate and nitroblue tetrazolium chloride (Sigma). Prestained low molecular weight markers were used to identify the electrophoretic mobility of Cx43. Films were volume-integrated within the linear range of the exposure using a scanning densitometer. Experiments were replicated three times and results expressed as the mean value.

Fig. 1

(A) A representative TTC-stained section of myocardium from the ischemia-reperfused left anterior descending artery region from a dog treated with vehicle, showing infarction at the endocardial layer (white color, from 5 to 8 o'clock). (B) Masson-stained micrograph taken from box in A. Note the prevalence of contraction band necrosis, graded as 3.

2.6. Confocal microscopy

In order to investigate the spatial distribution and quantification of Cx43, analysis of confocal microscopy was performed on left ventricular muscle from the border zone. Hearts were snap-frozen in liquid nitrogen, embedded in OCT compound (Tissue-Tek), and cryosections were performed at a thickness of 5 μm. The slides containing the sectioned tissues were rehydated in 0.01% sodium bicarbonate at pH 7.4. Tissues were incubated with Chemicon polyclonal anti-Cx43 antibodies to detect both phosphorylated and non-phosphorylated forms at dilution 1:200 in 5% non-fat milk in PBS for 2 h at 37 °C. We also used a Zymed monoclonal antibody to bind selectively to non-phosphorylated Cx43. The second antibody was monoclonal goat anti-mouse IgG conjugated to fluorescein isothiocyanate (Sigma), at 1:50 dilution in PBS containing 0.5% BSA for 1 h. The sections were washed three times with PBS and mounted in Dako fluorescent mounting medium. Primary antibody was omitted in negative controls run in parallel. Immunolabelled sections were examined through the use of confocal laser scanning microscopy as described previously [20].

2.7. Superoxide anion assays

Serial in vivo blood samples from the great cardiac vein were withdrawn at baseline (15 min after stabilization), 15 min after 17β-estradiol administration (early group as baseline), 30 min after coronary ligation, immediately before coronary reperfusion, 1, 2, 3, 4, 30 and 60 min after reperfusion, and at the termination of reperfusion (2 h). The measurement of blood lucigenin-derived chemiluminescence (LDCL) was similar to that described previously [20]. In preparation for LDCL, a 0.2 ml sample was added in a stainless steel cell, 0.5 ml of 5 μM lucigenin (bis-N-methy-lacridinium nitrate, Sigma) was injected into the cell, and the chemiluminescence was continuously measured for a total of 300 s. The assay was performed in duplicate for each timed point and was expressed as chemiluminescence counts/10 s.

2.8. Laboratory measurements

17β-Estradiol concentrations were quantified by enzyme-linked immunoassay (Diagnostic Products, Los Angeles, CA) before and at the end of the study. The detection limit was 20 pg/ml for estradiol. If blood results were below the limit of detectability of the test, the lower limit of detection was recorded.

2.9. Exclusion criteria

Animals were omitted from analysis: (1) if such severe hypotension was observed that the experiment could not be continued successfully for the duration of the protocol; (2) if intractable ventricular fibrillation occurred or antiarrhythmic agents were needed to correct arrhythmia; or (3) presence of heart worms. Because of influence of collateral circulation on infarct sizes [21], we excluded collateral flow >20% of baseline coronary blood flow to make our study animals homogeneous. Dogs with ventricular fibrillation during reperfusion were resuscitated and converted to a stable rhythm by internal electric shocks (3×10 W). The low energy did not result in more cell necrosis [22].

2.10. Statistical analysis

The continuous variables are expressed as mean±S.D. Differences were tested for hemodynamic variables, infarct size and area at risk (AAR) among the three groups by nonparametric Kruskal–Wallis statistic analysis followed by a Newman–Keuls post hoc test. Densitometric quantification of signal intensity was performed for Western analysis. Background measurements of signal intensity were subtracted individually from each lane. Spearman rank correlation analysis was used to determine the relationship between the parameters and infarct size as a percent of the left ventricle. A value of P<0.05 was considered to be significant.

3. Results

There were low baseline estrogen concentrations in the control group (Table 1). Intravenous injection of 10 μg/kg of 17β-estradiol, either prior to coronary occlusion or prior to reperfusion, resulted in similar increases of plasma estradiol concentration equivalent to those in human females during midcycle. Concentrations of arterial blood gas, calcium, sodium and potassium were stable throughout the study.

View this table:
Table 1

Changes of hemodynamics, coronary blood flow and estradiol at different fixed times among controls, early and late groups

Control (n=16)Early (n=16)Late (n=16)
Heart rate (bpm)182±21175±25180±20
Mean blood pressure (mm Hg)124±16131±17125±15
RPP (×102, bpm×mm Hg)294±58304±67301±53
LVEDP (mm Hg)9±49±37±4
Coronary blood flow (ml/min)37±536±540±4
Estradiol (pg/ml)41±944±648±8
Occlusion, 30 min
Heart rate (bpm)189±20179±29185±16
Mean blood pressure (mm Hg)112±16127±20106±19
RPP (×102, bpm×mm Hg)252±57305±45311±51
LVEDP (mm Hg)19±4*15±718±7*
Coronary blood flow (ml/min)3±2*2±3*2±3*
Reperfusion, 30 min
Heart rate (bpm)178±22179±23184±12
Mean blood pressure (mm Hg)101±23122±25118±26
RPP (×102, bpm×mm Hg)240±55301±64301±50
LVEDP (mm Hg)21±3*17±7*15±7
Coronary blood flow (ml/min)42±739±943±11
Reperfusion, 120 min
Heart rate (bpm)184±19179±19187±13
Mean blood pressure (mm Hg)107±24126±13124±14
RPP (×102, bpm×mm Hg)267±65275±42292±33
LVEDP (mm Hg)21±5*14±816±5*
Coronary blood flow (ml/min)37±938±1142±8
Estradiol (pg/ml)46±11426±57*541±44*
  • Data are expressed as mean±S.D. Early group: 17β-estradiol was given intravenously 15 min before coronary occlusion; Late group: 17β-estradiol was given 3 min before coronary reperfusion following 60-min ischemia. LVEDP: left ventricular end-diastolic pressure; RPP: rate pressure product.

  • * P<0.05 compared with respective baseline.

  • P<0.05 compared with the control group.

3.1. Mortality and exclusions

A total of 65 animals were randomly assigned in the study. Nine animals in the control group were excluded: four because of collaterals, four for intractable ventricular fibrillation during reperfusion, and one for hypotension. Five animals were excluded in the early group: four for collaterals and one for hypotension. Three animals were excluded in the late group for collaterals >20%. The remaining dogs were assigned to each group of 16.

3.2. Hemodynamic variables

The hemodynamic data are summarized in Table 1. There were no significant differences in heart rate, mean aortic pressure, left ventricular end-diastolic pressure, and rate-pressure product. Baseline coronary blood flow measured with intracoronary Doppler flow wire was not significantly different among the three groups. The amount of decrease after coronary occlusion was within the same range among the three groups. Blood flow of the left anterior descending artery among the three groups was similar during coronary occlusion, suggesting that collateral flow to the ischemic region was not altered by estrogen treatment.

3.3. Superoxide anion

Dogs in the control group generated a significantly higher concentration of superoxide anion than those in estrogen-treated animals in response to ischemia–reperfusion. LDCL increased significantly at the time of 30 min after myocardial ischemia. It peaked at 2 min after reperfusion, then decreased but remained above the control level 2 h after reperfusion. Peak superoxide anion production was 4742±614 counts/10 s in control group, compared with 1798±298 and 2004±342 counts/10 s in the early and late groups, respectively (both P<0.05, Fig. 2).

Fig. 2

Line graphs showing the relative amount of superoxide anion assessed by lucigenin-derived chemiluminescence during different phases of the study. Data are expressed as mean±S.D. *P<0.05 vs. controls. P<0.05 vs. baseline data.

3.4. Infarct size

There were no differences in body weight or heart weight among the groups. There was no significant difference in AAR expressed as a percentage of the left ventricle among the three groups, indicating a comparable degree of ischemic risk (Fig. 3). After 1 h of coronary artery occlusion followed by 2 h of reperfusion infarct size in control animals averaged 38±7% of AAR, compared with 16±6% of AAR in the early group and 16±8% of AAR in the late group (both P<0.0001 vs. controls). There were no statistically significant differences of infarct size between early and late groups.

Fig. 3

Effects of vehicle (control), early administration of estrogen before ischemia (early group), and late administration of estrogen prior to reperfusion (late group) on the area at risk (AAR), indexed to left ventricle and necrosis area indexed to AAR and to the left ventricle (LV). Treatment with estradiol, either early or late, resulted in a significant reduction in infarction compared with control group. Data are expressed as mean±S.D. *P<0.05 vs. controls.

The correlation coefficient between infarct size and AAR was 0.56 (P<0.05) in the control animals. Treatment with 17β-estradiol, either early or late, resulted in infarct sizes smaller than predicted from AAR.

3.5. Histological analysis

Macroscopically, control dogs exhibited confluent infarctions, whereas infarcts in the early and late estrogen-treated groups were unevenly distributed, interspersed with islands of viable myocardium. The number of areas of necrosis in the section with the highest patches of necrosis was measured after TTC staining. The number of patches of necrosis was 2±2, 6±3 and 7±5 in control, early and late-treated groups (both P<0.05 vs. control). Histological analysis in the control revealed infarcts composed almost exclusively of contraction band necrosis. The severity of contraction band necrosis was significantly higher in the control compared with the treated groups (2.6±0.5 vs. 1.3±0.5 in the early group and 1.3±0.5 in the late group, P<0.0001, both).

3.6. Cx43 western analysis

Fig. 4 shows a representative blot and quantitative results in which Cx43 band intensities were normalized to the value measured from sham operation. Two predominant forms of Cx43 were detected: one nonphosphorylated form (Cx43-NP; 41 kDa) and the other phosphorylated species (Cx43-P; 43 kDa). Western analysis derived from the border zone revealed that Cx43 band pattern is modified qualitatively in response to ischemia–reperfusion and estrogen administration. Densitometric analysis of immunoblots revealed similar total amount of Cx43 signals and a decreased intensity of Cx43-P form in tissues undergoing ischemia–reperfusion. Thus, ischemia–reperfusion is associated with progressively decreased phosphorylation of Cx43 in the border zone. The quantitative changes of Cx43-P were significantly lower in groups treated with estrogen, either early or late compared with data from the control group.

Fig. 4

(A) Western blot analysis of Cx43 showing the effect of estradiol on immunorecognition of Cx43 in homogenates of the left ventricle from the border zone after 60-min ischemia and 2-h reperfusion in canine heart. Ischemia–reperfusion was associated with marked loss of phosphorylated Cx43 (43-kDa band) and a corresponding increase in dephosphorylated Cx43 (41-kDa band). A significantly reduced phosphorylation had taken place in the groups treated with either early or late administration of 17β-estradiol compared with control. (B) Densitometric quantification of blot band intensities for relative Cx43 normalized to a sham group (mean±S.D). Total amount of Cx43 signal (∀) revealed no change throughout the ischemia–reperfusion, indicating progressively reduced Cx43-P (!). *P<0.05 compared with sham; P<0.05 compared with control group.

Quantitative analysis showed a significant relationship between the amount of phosphorylated Cx43 expression assessed by Western blot analysis and contraction band necrosis or infarct size (r=0.73, P<0.0001 for contraction band necrosis, and r=0.83, P<0.0001 for infarct size, Fig. 5).

Fig. 5

Graph shows correlation between the amount of phosphorylated Cx43 (Cx43-P) protein assessed by Western blot analysis and infarct size.

3.7. Confocal microscopy

Western blot data were confirmed by confocal microscopic data analysis. Qualitative immunofluorescence analysis was performed in the border zone. In the sham group, sections stained with the Chemicon polyclonal Cx43 antibody produced intense punctate labelling primarily at contacts between cardiomyocytes (Fig. 6A). In contrast, virtually no signal was detected in sham stained with the Zymed monoclonal antibody (Fig. 6B). After ischemia–reperfusion in the control, there was a loss of the Chemicon polyclonal antibody immunoreactivity (Fig. 6C) with a corresponding increase in signal of the Zymed monoclonal antibody (Fig. 6D). These alterations in the Chemicon polyclonal antibody immunostaining were higher in animals treated with estradiol (early group in Fig. 6E, late group in Fig. 6G), consistent with Western blot shown in Fig. 4.

Fig. 6

Representative confocal images of structural features of myocytes in the sham group (A, B), the border region in the control group (C, D), early group (E, F), and late group (G, H) after 60-min ischemia and 120-min reperfusion. Sections were stained with either a Chemicon polyclonal antibody for detection of both phosphorylated and non-phosphorylated (total) Cx43 or a Zymed monocloncal antibody for detection of non-phosphorylated (NP) Cx43. The amount of the Chemicon polyclonal antibody is significantly decreased in groups treated with estradiol, either early or late compared with the control group. I: infarction area. Bar=80 μm.

4. Discussions

Our present results clearly show that despite the late administration of estrogen prior to coronary reperfusion, the drug effectively reduced infarct size after myocardial ischemia and was as effective when it was given just prior to reperfusion as it was when present throughout ischemia and reperfusion. Thus, the infarct-limited effect of estrogen is through direct protection during coronary reperfusion. Altered expression of Cx43 protein by a change in the reduced phosphorylated state was accumulated in the border zone of myocardial infarction especially in dogs treated with estrogen, either early or late. Cardioprotective effects of estrogen could result from modification of the gap junction protein, leading to an inhibition of functional propagation of intercellular signaling and less formation of contraction band necrosis. Thus, we demonstrated for the first time that beneficial effects of estrogen on limiting infarct size at physiological concentrations appear to be associated with altered expression of myocardial Cx43 protein in the infarct border.

4.1. Mechanisms of cardioprotective effects of 17β-estradiol

The results of this study show that estrogen exerts its cardioprotective effects by attenuated formation of myocardial phosphorylated Cx43 protein assessed by Western analysis and decreased formation of superoxide anion assessed by LDCL. Some myocytes which are reversibly injured at the time that coronary artery occlusion is terminated undergo irreversible cell damage during reperfusion [23]. Infarct-induced spreading waves invade the ischemic zone and their severity might contribute to secondary expansion of ischemic lesions [24], which are salvageable by agents such as gap junction blockers. If left untreated, aberrant [Ca2+] signaling in tissue surrounding the evolving infarct could traverse gap junctions to neighboring cells and contribute to the extent of contraction band necrosis. Thus, myocytes did not die randomly but, rather, in groups. The notion was further corroborated by our observation that control dogs exhibited confluent infarctions, whereas infarcts in the estrogen-treated groups were patchily distributed, interspersed with islands of viable myocardium.

Our results here showed qualitative (reduced phosphorylated) changes of Cx43 protein during ischemia–reperfusion in groups treated with estrogen. Phosphorylation of Cx43 is an important checkpoint for gap junction function. It has been shown that gap junction uncoupling occurs during acute ischemia [25]. However, the rapid restoration of blood flow after reperfusion results in reopening of gap junctions in surviving cells [26]. Reopening of gap junctions in the presence of abnormally high intracellular concentration between dying and healthy cardiomyocytes induced calcium influx and propagation of contraction band necrosis. Transient gap junction blockade during the first minutes of reperfusion until the cell recovers Ca2+ control has been shown to prevent hypercontracture formation [14]. The time sequences of gap junction function may help explain, at least in part, reperfusion injury and were compatible with our finding that estrogen provided cardioprotective effects mainly through the reperfusion phase. These findings of decreased contraction band necrosis were consistent with our speculation that the major protection of estradiol is due to inhibition of the development of calcium overload by blocking gap junction functions.

The specific signal transduction pathways responsible for the effects of estrogen on Cx43 are not known. Cx43 is not static but undergo a continual process of formation and removal with a short half-life of 1–1.5 h [27]. Estrogen may alter gap junctions through genomic or nongenomic pathways. The presence of putative estrogen response element sites at the promoter region of Cx43 gene suggests that estrogen may increase Cx43 through a genomic pathway mediated by the nuclear estrogen receptor [28]. However, the inhibition of gap junctions detected in the late group occurred rapidly, implying that a nongenomic pathway is more critical such as free radicals. The signaling pathways to mediate free radical-induced Cx43 expression are a complex process involving alteration in levels of mitogen activated protein (MAP) kinase [29] and cAMP [30]. Free radicals have been identified as a factor that mediates Cx43 functions in rat hepatocytes [31]. Aikawa et al. [32] have demonstrated that free radicals activate MAP kinase pathways in rat cardiomyocytes. Cx43 is a target of the MAP kinase signaling pathway in vivo and the activation of MAP kinase signaling cascade increased Cx43 phosphorylation [32]. Thus, estradiol may induce attenuated expression of Cx43 protein by reducing activated MAP kinase through decreased formation of free radicals. Second, free radicals have been shown to reduce the cAMP levels [30]. cAMP agonists have been shown to enhance gap junction protein expression [33]. Li et al. [34] have shown that estradiol decreases the formation of cAMP. Thus, attenuated increase of estradiol-induced cAMP levels resulted in inhibition of Cx43 trafficking from intracellular storage sites [35], compatible with our finding. Although we suggested that estradiol may play an important role in MAP kinase activity and cAMP levels, which in turn altered expression of Cx43 protein possibly by attenuated formation of free radicals, the specific signal transduction pathways by which estradiol modulates Cx43 remain speculative.

4.2. Clinical implications

Although the structure of Cx43 is known, its importance for heart function remained unclear except for its role in synchronization of electrical activation and contraction of cardiac myocytes [26]. The hypothesis that Cx43 disruption plays an important role in limiting infarction extension is extremely attractive. Cell-to-cell communication could allow propagation of hypercontracture of severely injured cells to neighboring cells that could have survived. A small probability of cell-to-cell transmission could have a major impact on infarct size. The hypothesis was supported by the fact that Cx43-deficient mice develop smaller infarct size than wild-type mice after infarction [36]. Gap junction blockers such as octanol or heptanol have been shown to have cardioprotective effects in experimental ischemia–reperfusion models; however, the drugs cannot be used clinically due to their toxic effects such as impaired contractile function and arrhythmogenic properties. Estrogen has been used clinically and has few side effects. Because of the relatively few side effects of estrogen and its protective function even when started after the onset of ischemia, it may be an agent that could be administered very early to patients who present with acute myocardial infarction.

4.3. Study limitations

There were several limitations of this study. First, only male dogs were included in this study, which excluded the concentration effect of estrogen during the menstruation cycle in females. The effects of acute administration of estrogen in females need more clinical studies to confirm the benefits although we have demonstrated in canine hearts that estrogen limits myocardial infarction size resulting from coronary artery occlusion and reperfusion in a dose-dependent manner, irrespective of gender difference [16]. Second, although we have demonstrated qualitative changes of Cx43 protein expression in response to estradiol administration, they do not necessarily correlate with functional changes. We did not show reduced intracellular calcium overload either by buffering the cytosolic Ca2+ concentration via patch clamp, or by directly measuring the cytosolic Ca2+ concentration by means of fluorescent emission. However, the techniques are impossible to perform in an in vivo study. The severity of contraction band necrosis, an indirect method to assess the calcium overload, was evaluated by histological analysis, which showed a close relation between the amount of Cx43 expression and the severity of contraction band necrosis. Third, the reperfusion model used differs from the clinical setting. Sudden relief of occlusion by snare instead of a more gradual reperfusion in actual clinical situations may have influenced the ensuing amount of free radicals. The clinical benefits of such a reduction in infarct size should be tested in future clinical trials.

4.4. Conclusions

Our results provide evidence that late administration of estrogen, prior to coronary reperfusion, reduced infarct size after myocardial ischemia as effectively as estrogen administered when given before coronary ischemia. The beneficial effects of estrogen at physiological concentrations appear to be associated with the reduced amount of phosphorylated myocardial connexin43 protein. This provides perspective for analysis of the pathophysiology and treatment of acute myocardial infarction.


This work was supported by the grant of Chi-Mei Medical Center (CMFHT 9201, CMFHR9203, CMFHR9303, and CMFHR9307). We thank Chi-Ren Lue, In-Ping Cheng and Li-Lan Chien for their excellent technical assistance, Wen-Mei Fu, PhD, for confocal microscopic analysis, and Wang-Ru Chen, Ya-Chan Kuo, and Yu-Wei Chao, for Western blot analysis.


  • Time for primary review 25 days


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
  33. [33]
  34. [34]
  35. [35]
  36. [36]
View Abstract