PKC-permitted elevation of sarcolemmal K_ATP concentration may explain female-specific resistance to myocardial infarction

Abstract

The female myocardium, relative to that of the male, exhibits sustained resistance to ischaemic tissue injury, a phenomenon termed sex-specific cardioprotection (SSC). SSC is dependent upon the sarcolemmal K_ATP channel (sarcK_ATP), and protein kinase C (PKC). Here we investigate whether PKC-mediated regulation of sarcK_ATP concentration can explain this endogenous form of protection. Hearts from male (M) and female (F) rats were Langendorff-perfused for 30 min prior to either regional ischaemia–reperfusion (I/R), or global ischaemia (GISC). For both protocols, pre-ischaemic blockade of PKC was achieved by chelerythrine (Chel) in male (M + C) and female (F + C) hearts. Additional female hearts underwent sarcK_ATP antagonism during I/R by HMR-1098 (HMR), either alone or in combination with Chel (HMR + Chel). GISC hearts were fractionated to assess cellular distribution of PKCɛ and sarcK_ATP. Sex-specific infarct resistance was apparent under control I/R (F, 23 ± 3%vs. M, 36 ± 4%, P < 0.05) and abolished by Chel (F + C, 36 ± 3%). Female infarct resistance was susceptible to sarcK_ATP blockade (Control, 16 ± 2%vs. HMR, 27 ± 3%), and PKC blockade had no additional effect (HMR + Chel, 26 ± 2%). The prevalence of Kir6.2 and SUR2 was higher in the sarcolemmal fractions of females (Kir6.2: F, 1.24 ± 0.07 vs. M, 1.02 ± 0.06; SUR2: F, 3.16 ± 0.22 vs. M, 2.45 ± 0.09; ratio units), but normalized by Chel (Kir6.2: F, 1.06 ± 0.07 vs. M, 0.99 ± 0.06; SUR2: F, 2.99 ± 0.09 vs. M, 2.82 ± 0.22, M; ratio units). Phosphorylation of sarcolemmal PKCɛ was reduced by Chel (p-PKCɛ/PKCɛ: control, 0.43 ± 0.02; Chel, 0.29 ± 0.01; P < 0.01). We conclude that PKC-mediated regulation of sarcK_ATP may account for the physiologically sustainable dependence of SSC upon both PKC and sarcK_ATP, and that this regulation involves PKC-permitted enrichment of the female sarcolemma with sarcK_ATP. As such, the PKC-sarcK_ATP axis may represent a target for sustainable prophylactic induction of cardioprotection.

Introduction

Ischaemic heart disease (IHD) remains the leader among all causes of mortality in North America, and worldwide. It accounts for 11.8% of deaths in low- to middle-income nations, and 17.3% in high-income nations (WHO, 2004). A variety of factors modulate the burden of IHD, and of these, the biological differences between men and premenopausal women are among the most potent. Across all major risk factors and socio-economic stratifications, premenopausal women experience less than half the IHD-attributable mortality of age-matched men (WHO, 2004).

There is now substantial evidence that, relative to males, females of many species exhibit elevated resistance to myocardial ischaemia–reperfusion (I/R) injury. This form of sex-specific cardioprotection (SSC) manifests across a wide range of experimental paradigms, and in many important indices of I/R injury, such as incidence of arrhythmia (Node et al. 1997; Tsai et al. 2002), infarct-size (Bae & Zhang, 2005; Brown et al. 2005b; Johnson et al. 2006; Chicco et al. 2007), and post-ischaemic functional deficit (Bae & Zhang, 2005; Gabel et al. 2005).

A series of studies published by our laboratory suggest that SSC is dependent on function of the sarcolemmal isoform of the ATP-sensitive potassium channel (sarcK_ATP), but exhibits no sensitivity to blockade of the more heralded mitochondrial channel, mitoK_ATP (Brown et al. 2005a; Johnson et al. 2006; Chicco et al. 2007). These observations place SSC among a small minority of protected phenotypes that do not appear to rely upon function of mitoK_ATP. Importantly, this minority, which includes the phenotype associated with chronic exercise (Brown et al. 2005), is also characterized by sustainable protection against I/R injury.

Protein kinase C (PKC) is a family of intracellular serine/threonine kinases that have also received attention for their ability to induce sustainable myocardial protection (Miyamae et al. 1997; Shi et al. 2004; Inagaki et al. 2005, 2006), so much so that they are among very few targets thought to exhibit genuine therapeutic potential (Inagaki et al. 2006; Budas et al. 2007). In a number of forms of preconditioning including ischaemic preconditioning (IPC), the epsilon isoform (PKCɛ) has been shown to be essential in conferring cardioprotection. Recently, PKCɛ has also been implicated in the mechanism of SSC, both in conjunction with IPC (Shinmura et al. 2008), and in its absence (Bae & Zhang, 2005; Hunter et al. 2007).

In combination, the bodies of literature surrounding the cardioprotective effects of PKC and sarcK_ATP suggest that the two may interact in the pathway underlying sex-specific protection. In principle this concept is permitted by studies indicating that PKC is capable of modulating sarcK_ATP current (I_KATP) and directly interacting with the pore-forming subunit of the channel (Hu et al. 1996; Light et al. 2000; Budas et al. 2007). However, the potential for this interaction to occur in a sex-specific manner remains unknown.

Given that the established paradigm for PKC activation involves translocation from the cytosol to some membrane structure (Newton & Johnson, 1998; Budas et al. 2007), it may be that sex-specific trafficking of PKCɛ to the sarcolemma underlies the dual dependence of SSC upon sarcK_ATP and PKC. Alternatively or concurrently, several investigations have suggested that PKC is involved in controlling the distribution of the essential myocardial sarcK_ATP subunits (Kir6.2 and SUR2A) (Hu et al. 2003; Budas et al. 2004; Garg & Hu, 2007), and it is known that females exhibit higher whole-cell concentrations of those subunits (Brown et al. 2005a,b; Chicco et al. 2007). This would suggest that PKC-dependent sarcK_ATP trafficking, as distinct from modulation of sarcK_ATP unitary conductance, might also contribute to SSC. We proposed two hypotheses to explore these possibilities: (1) the female sarcolemma exhibits greater ischaemia-induced enrichment by PKCɛ and sarcK_ATP than the male sarcolemma, and (2) such enrichment by sarcK_ATP is prevented by pre-ischaemic catalytic blockade of PKC, while enrichment by PKCɛ remains unaffected.

We herein provide evidence supporting the established role of PKC in the mechanism of SSC, and suggest that while ischaemia does not induce sex-specific sarcolemmal enrichment of PKCɛ, activity of some PKC isoform at the sarcolemma may be required to maintain an elevated sarcolemmal concentration of sarcK_ATP among females.

Methods

Animal model

Adult (3–4 months) male (M, n= 88) and female (F, n= 134) Sprague–Dawley rats were housed in a 12 h–12 h light–dark cycle, and provided standard chow (Harlan) and water ad libitum. All procedures received prior approval from the Institutional Animal Care and Use Committee at the University of Colorado at Boulder, and all were performed in accordance with the guidelines instituted by the American Physiological Society and The Journal of Physiology (Drummond, 2009).

Regional ischaemia–reperfusion

To determine the effect of PKC inhibition on myocardial function and infarct size following ischaemia, hearts from male and female rats were perfused with an isoform non-specific PKC antagonist (chelerythrine, Chel) prior to regional ischaemia–reperfusion. Animals were placed under deep anaesthesia by intraperitoneal injection of sodium pentobarbital (35 mg kg⁻¹) and subsequently killed by midline thoracotomy and rapid heart excision. Hearts were briefly placed in ice-cold saline prior to being suspended and retrograde perfused on the cannulae of a modified Langendorff apparatus, as previously described (Brown et al. 2005a,b;). All then underwent 30 min of baseline perfusion prior to induction of ischaemia by ligation of the left anterior descending coronary artery (LAD). Control hearts received Krebs–Henseleit (KH) buffer (in mm: 117.4 NaCl, 4.7 KCl, 1.9 CaCl₂, 1.2 MgSO₄, 1.2 KH₂PO₄, 5 pyruvate, 11 glucose, 0.5 EDTA, 25 NaHCO₃, and heparin 1200 U l⁻¹) for the entirety of this period. Alternatively, hearts undergoing PKC blockade were exposed to KH buffer containing Chel (2 μm), for the final 20 min of the baseline period. All perfusion buffers were warmed to 37°C, and oxygenated by bubbling with 95% O₂–5% CO₂ throughout I/R.

During the first 10 min of the baseline period, hearts were instrumented (via the aortic valve) with a 3-F pressure-transducing catheter (Millar Instruments, Inc., Houston, TX, USA) for measurement of heart rate and left ventricular performance. Coronary effluent and left ventricular pressure waveforms were measured at 10 min intervals throughout the baseline period, and then at 15 min intervals for the remainder of the protocol, excepting the first 5 min of reperfusion, during which data were collected at post-ischaemic minutes 1, 3 and 5.

Ligation of the LAD was applied to all hearts subsequent to the initial 30 min of baseline perfusion, and was achieved via removable snare. Upon ligation, hearts undergoing PKC blockade were switched to regular KH buffer. Thus, hearts exposed to chelerythrine did not undergo a washout period prior to ischaemia. The ischaemic period continued for 1 h, after which the snare was loosened and reperfusion ensued for a further 2 h.

At the conclusion of reperfusion the hearts were once again ligated, and 100 μl of 0.5% Evans blue dye was injected into the descending perfusate directly proximal to the cannula. Stained hearts were then cut down, and the right ventricle and atria were removed before the left ventricle and septum were sectioned in the transverse plane to yield four slices from apex to base. To differentially stain the zone at risk (ZAR) the slices were incubated in triphenyltetrazolium chloride at 37°C for 10 min. Heart slices were then photographed on both sides using a digital camera (Sony Cybershot DSC-S75, VAD-S70 adapter) and dissection microscope. All images were analysed by digital planimetry using ImageJ software, and by the same trained technician, who was blinded to the origin of the slices. The total area (TA), ZAR and infarcted area (IA) were measured on both sides of each slice. The ZAR was defined as tissue unstained by Evans blue dye, and the infarcted area was the region of the ZAR for which tetrazolium staining was absent. The final TA, ZAR and IA values for each slice were calculated as the mean of values from the two sides, and multiplied by slice mass to arrive at the mass for each zone of the slice. For comparison between groups, ZAR mass was normalized to reflect a percentage of TA mass, and IA mass was normalized to ZAR mass in a similar manner. As such, all ZAR data are presented as a percentage of TA, and all IA data as a percentage of ZAR.

In a separate series of experiments, an identical I/R challenge was applied to female hearts to test the effect of combining PKC blockade and sarcK_ATP blockade on sex-specific infarct resistance. SarcK_ATP antagonism was applied by including an inhibitor specific to the sarcolemmal isoform of K_ATP (HMR-1098; HMR, 30 μm) throughout regional ischaemia and reperfusion. All combinations of these antagonists were performed such that four groups were tested: control (Con), preischaemic chelerythrine-treated (Chel), HMR-treated (HMR), and those treated with both preischaemic chelerythine and HMR (HMR + Chel).

Global ischaemia

To examine the effect of ischaemia, sex and PKC blockade on the cellular distribution of PKCɛ and the myocardial subunits of sarcK_ATP (Kir6.2 and SUR2a), hearts from male and female rats were subjected to a global ischaemia (GISC) protocol followed by quantitative subcellular fractionation. Rats were killed, and their hearts were cannulated and retrograde perfused in a similar manner to those undergoing regional I/R, with the only exception being that they were not instrumented for measurement of left ventricular pressure. These hearts then underwent baseline perfusion identical to that described for regional I/R; each heart received either control KH buffer throughout (− chelerythrine), or KH buffer containing 2 μm chelerythrine during the final 20 min of the baseline period (+ chelerythrine). After baseline perfusion, hearts were subjected to either 30 min of normothermic global no-flow ischaemia (+ ischaemia), or a further 30 min of perfusion by control KH buffer (− ischaemia). At the end of this hour all hearts were cut down for subcellular fractionation.

A small subset of hearts (n= 4) were perfused with the PKC agonist 4β-phorbol 12-myristate 13-acetate (PMA) to confirm that our fractionation and Western blotting techniques are capable of detecting PKCɛ translocation to various membrane structures. This administration was performed in Langendorff mode, and KH containing 100 nm PMA was prepared as described previously (Brooks et al. 1997). After 20 min of control KH perfusion, hearts received PMA for 10 min, and were then returned to control KH for a further 10 min, before being cut down and fractionated.

Subcellular fractionation

The left ventricle (LV) was dissected away from the atria, right ventricle and septum for subcellular fractionation, which was performed according to the methods of Fuller et al. (2000). Briefly, each LV was first diced and then incubated in a high salt solution (2 m NaCl, 20 mm Hepes, pH 7.4) for 30 min, to facilitate dissociation of membrane structures from the myofibrillar component. Diced tissue was rinsed and homogenized via a 3 ml glass–glass tissue grinder (Kimble Chase, Vineland, NJ, USA) in 2 ml g⁻¹ buffer A (in mm: 20 Hepes, 2 EDTA, 1 MgCl₂, 250 sucrose, pH 7.4). This whole tissue homogenate was centrifuged at 1000 g to pellet extracellular debris, much of the myofibrillar component, and a sarco-endoplasmic reticulum-enriched fraction. The supernatant was retained and underwent subsequent centrifugations at 10 000 g, 20 000 g, and 100 000 g, from which pellets were collected to contain mitochondria-enriched, sarcolemma-enriched and endosome-enriched fractions, respectively. Each pellet was washed in buffer A, and resuspended in a volume of lysis buffer (20 mm Hepes, 150 mm NaCl, 1% NP-40, 0.01% SDS, 1 mm EDTA) sufficient to arrive at protein concentrations between 5 and 20 μg μl⁻¹. The lysis buffer was supplemented with protease (Complete Mini: Roche, IN, USA) and phosphatase (0.1 mm Na₃VO₄, 10 mm NaF) inhibition. Fractions were immediately frozen in liquid nitrogen upon resuspension.

Co-immunoprecipitation of SUR2

To isolate the species and quantity of SUR2A (the primary myocardial splice variant of SUR2; Seino & Miki, 2003; Flagg & Nichols, 2005) contributing to the mature form of sarcK_ATP, we immunoprecipitated with anti-Kir6.2 (AB5495, Millipore, Billerica, MA, USA) from 200 μg of protein taken from the sarcolemma-enriched fraction. Samples were brought to 1 mg ml⁻¹ in buffer 1 (1% Igepal, 0.1 mm NaV₃VO₄, 150 mm NaCl, 10 mm PBS, pH 7.2) and precleared with 4 μl packed protein A-coupled agarose beads (Millipore), prior to overnight incubation (4°C) with anti-Kir6.2 (2 μg). Immunocomplexes were precipitated with 25 μl packed beads for 2 h (4°C) and recovered by 2 min centrifugation at 1000 g prior to undergoing four washes in buffer 1, and a further four in buffer 3 (in mm: 100 NaCl, 1 EDTA, 0.1 NaV₃VO₄, 10 Tris, pH 7.5). Samples were eluted into 40 μl SDS loading buffer (161-0791, Bio-Rad, Hercules, CA, USA) and this quantity was immediately blotted and probed with anti-SUR2 (H-80, 1: 200, Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) as is otherwise described below.

Western immunoblotting

Western immunoblotting was performed by standard SDS-PAGE, using 4–12% Tris/Bis gels, and the Criterion electrophoresis and blotting systems (Bio-Rad). Fraction samples were loaded at 60 μg per well after protein determination by the Bradford method, and all blots were performed in duplicate. In order to minimize loss of signal over several strip/reprobe cycles, each membrane was cut at 70 kDa prior to any antibody incubation. Serial probes for PKCɛ (C-15, 1: 200, Santa Cruz), p-PKCɛ (Ser 729, 1: 200, Santa Cruz), and the myocardial sodium/calcium exchanger, NCX1 (R3F1, 1: 500, Swant, Bellinzona, Switzerland), were performed on membrane sections including proteins larger than 70 kDa. Between probes, this high molecular weight membrane section was stripped (Restore, Bio-Rad) for 30 min at room temperature. Membrane sections containing proteins below 70 kDa were probed for Kir6.2 (1: 500, Millipore – see the online Supplemental Material for blots characterizing the immunospecificity of this antibody), and the early endosomal marker Rab4 (16) (2167, 1: 1000, Cell Signaling Technology, Inc., Danvers, MA, USA). Fraction enrichment was assessed by probing the various fractions for NCX1, Rab4, the sarco-endoplasmic reticulum marker Calreticulin (1: 1000, Ann Arbor, MI, USA), and the mitochondrial marker Prohibitin 1 (Garg & Hu, 2007) (1: 1000, Cell Signaling). Blots were subsequently incubated in horseradish peroxidase (HRP)-conjugated secondary antibodies (Santa Cruz), and enhanced chemiluminescence (Pierce Biotechnology, Inc., Rockford, IL, USA), before being digitally exposed using the ChemiDoc-It 600 system (UVP, LLC, Upland, CA, USA). All densitometry analyses were performed using ImageJ software. Optical densities reported for the sarcolemma- and endosome-enriched fractions were normalized to NCX1. Optical density reported for p-PKCɛ was normalized to raw PKCɛ density in all circumstances.

Exclusions

Regional I/R

For experiments assessing the sex-specificity of PKC blockade, data were not collected for two animals of the original 36 because the experimenters were unable to successfully perfuse the heart within 90 s of excision. Data from two of the remaining 34 were excluded due to ambiguous staining of the left ventricular sections. Left ventricular function data from a further four animals were excluded because fibrillation at one or more time points prevented statistical treatment by a repeated measures model. Infarct size and morphological data were retained for these functional exclusions, but due to the likelihood of ineffective ligation, all data were excluded for the hearts with ambiguous staining. After all exclusions the sample sizes for infarct analyses were all n= 8, whereas those for all analyses of left ventricular function were: M = 7, F = 8, F + C = 7, M + C = 6.

For experiments assessing the effects of HMR, Chel, and HMR + Chel on female hearts, data were not collected for two of 38 animals due to unsuccessful perfusion. A further four hearts were excluded post hoc due to clotting above the point of ligation (ZAR > 50% total LV mass, n= 1), or due to delayed onset of anaesthesia resulting in impairment of heart rate recovery during baseline perfusion (HR < 150 beats min⁻¹). The final number of animals in each group was then: Con = 8, Chel = 7, HMR = 9, HMR + Chel = 8.

Global ischaemia

Tissue and data for four of 96 animals were not collected because the experimenters were unable to successfully perfuse the heart within 90 s of excision. Fractions for a further four animals were not retained due to contamination during homogenization. As a result the number of hearts contributing to sarcolemmal analyses were as follows: n= 10 for all male groups, n= 14 for both female groups not receiving Chel, and n= 10 for both female groups receiving Chel. Low protein yield for the endosomal fractions reduced the number of analysable fractions to (n= 6) for all groups.

Statistical analyses

Morphological data were analysed for sex differences by Student's t test. Time-series data for each index of ventricular performance (LV minimum pressure, LV maximum pressure, LV developed pressure, +dP/dt, −dP/dt, and coronary flow) were first separated into the three phases of the I/R protocol (baseline, ischaemia and reperfusion). Data for each phase were then treated by repeated measures ANOVA. Infarct area and zone at risk for each component were assessed by 2 (sex) × 2 (chelerythrine treatment) ANOVA. Changes in total PKCɛ, p-PKCɛ/total PKCɛ, Kir6.2, SUR2, and the various fraction markers were assessed within each fraction by 2 (sex) × 2 (chelerythrine treatment) × 2 (ischaemia) factorial ANOVA. Where warranted, factorial ANOVA was followed by specific between group contrasts using Student's t test (one-tailed for the effect of sex on infarct size – a confirmatory test, two-tailed for all other exploratory comparisons). One-way ANOVA was used to determine differences in infarct size for experiments involving HMR and Chel. Dunnet's post hoc analysis was used to assess pairwise differences, with Control specified as the reference group. An α level of 0.05 was adopted to indicate significant differences for all analyses, whereas an α level of 0.10 was selected to be indicative of statistical trends (Curran-Everett & Benos, 2004). All summary data are presented as means ±s.e.m. For appropriate pair-wise comparisons, P-values are presented in absolute form, whereas global P-values (ANOVA) are presented relative to α= 0.05.

Results

Morphology

Age-matched female rats were lighter and exhibited lower left ventricular mass than their male counterparts. These observations were true in absolute terms and when body and left ventricular masses were expressed relative to tibia length (all P < 0.05). Females also exhibited higher adrenal masses, both absolute and relative, and lower splenic mass (all P < 0.05; Table 1). All of these findings are consistent with our previous observations (Brown et al. 2007).

Table 1.

Morphology of male and female rats at the time of sacrifice

	Male	Female
Body mass (g)	293 ± 8	202 ± 4**
Left ventricle (mg)	647 ± 23	466 ± 12**
Left adrenal (mg)	24.0 ± 0.8	30.4 ± 0.8**
Right adrenal (mg)	22.6 ± 0.7	28.6 ± 0.7**
Spleen (mg)	707 ± 22	618 ± 17**
Tibia (mm)	34.2 ± 0.6	33.3 ± 0.2**

Data are means ±s.e.m.

^**P < 0.01.

Infarct size

A significant sex × chelerythrine interaction indicated that males and females responded differently to pre-ischaemic blockade of PKC (P < 0.05). This interaction was due to two pair-wise differences (Fig. 1A). Females not treated with chelerythrine exhibited smaller infarcts than males, with a relative reduction of 37% (M = 36 ± 4%vs. F = 23 ± 3%: P < 0.05), which agrees well with previous work from our laboratory (Brown et al. 2005b; Johnson et al. 2006; Chicco et al. 2007), and others (Song et al. 2003; Bae & Zhang, 2005). This resistance was completely abolished by pre-ischaemic administration of chelerythrine (F + C = 35 ± 3%, F vs. F + C: P < 0.01), suggesting that one or more isoforms of PKC are involved in the effector mechanism underlying the infarct resistance characteristic of SSC. These effects were not associated with any group differences in the size of the ischaemic zone (Fig. 1B).

Figure 1. Infarct area (% ZAR; A) and zone at risk (% Total LV area; B) among male and female rats exposed to control buffer (filled bars) or pre-ischaemic chelerythrine (open bars).

Sample size was constant across all groups at n= 8. *P < 0.05 compared to male (control). **P < 0.05 compared to female (control).

Figure 2 shows that HMR-1098 alone caused an expansion of infarct size among females (HMR = 27 ± 3%, Con = 16 ± 2%: P < 0.05), and chelerythrine alone elicited a similar trend to support the difference presented in Fig. 1 (Chel = 22 ± 2%, P < 0.1). The combination of HMR and Chel did not result in expansion of infarct size beyond that achieved by HMR alone (HMR + Chel = 26 ± 2%), suggesting that PKC and sarcK_ATP are involved in the same pathway underlying SSC. These effects were again independent of differences in the size of the zone at risk across the four groups (Fig. 2B).

Figure 2. Infarct area (% ZAR; A) and zone at risk (% Total LV area; B) among female hearts exposed to control buffer (open bars, Con), pre-ischaemic chelerythrine (light grey bars, Chel), HMR-1098 (dark grey bars, HMR), and combined chelerythrine and HMR-1098 (black bars, HMR + Chel).

*P < 0.05 vs. Con, #P < 0.1 vs. Con.

Coronary flow

No differences in coronary flow attributable to sex or chelerythrine treatment were apparent during any portion of the I/R protocol (all P > 0.05), suggesting that the effects of PKC blockade did not impact coronary vasomotor function and were limited to the myocardium itself. These results held for analyses of the decrement in flow upon ligation, and the hyperaemic response upon reperfusion (Fig. 3), the latter being a circumstance in which any differences in vasomotor function would be expected to surface (all P > 0.05). These results also support the assertion that ligation was equally effective across all groups studied.

Figure 3.

Changes in coronary flow (ml min⁻¹ g⁻¹) due to ligation (A) and reperfusion (B) among male and female rats exposed to control buffer (filled bars) or pre-ischaemic chelerythrine (open bars)

Left ventricular performance

The regional ischaemia applied here was not sufficient to elicit a significant effect of sex or either experimental treatment on any index of left ventricular performance, during any period of the I/R protocol (Figs 4–6).

Figure 4. Left ventricular developed pressure (mmHg), during 1 h/2 h regional I/R, by sex and chelerythrine.

M, filled squares; F, filled circles; M + C, open squares; F + C, open circles. Data are means ±s.e.m.

Rates of left ventricular pressure development and decay (mmHg s⁻¹), during regional 1 h/2 h I/R, by sex and chelerythrine.

M, filled squares; F, filled circles; M + C, open squares; F + C, open circles. Data are means ±s.e.m.

Figure 5. Left ventricular minimum pressure (mmHg), during 1 h/2 h regional I/R, by sex and chelerythrine.

M, filled squares; F, filled circles; M + C, open squares; F + C, open circles. Data are means ±s.e.m.

Cellular fractionation and fraction enrichment

The proportions of whole-tissue protein retained in the fractions separated at 1000 g, 10 000 g, 20 000 g and 100 000 g were 60 ± 9%, 22 ± 2%, 2.9 ± 0.5% and 2.1 ± 0.4%, respectively. Leaving approximately 13% soluble above 100 000 g. Figure 7 indicates enrichment of the 1000 g fraction with calreticulin and the 10 000 g fraction with PHB1, and NCX1 appears exclusively in fractions precipitated at 20 000 g and above. Collectively, these data suggest efficient dissociation of the sarcolemmal protein pool from the major intracellular protein depots, and measurable dissociation of the mitochondrial compartment from the sarco-endoplasmic reticulum. The 20 000 g and 100 000 g fractions are further characterized by their relative content of the early endosomal marker Rab4. Figure 8 shows that Rab4 was substantially (27.2%, P < 0.0001) enriched in the 100 000 g fraction, indicating that this fraction is representative of the endosomal pool involved in sarcolemmal protein recycling, and distinguished from the 20 000 g sarcolemma-enriched fraction. Importantly, the enrichment characteristics of the various fractions for the above markers did not differ by sex, ischaemia, or chelerythrine treatment (all P > 0.05).

Figure 7. Representative immunoreactivity for myocardial sodium calcium exchanger (NCX1), calreticulin and prohibitin 1, across all collected fractions.

All three blots were transferred from the same polyacrylamide gel before being cut and probed with appropriate primary antibodies. Cellular fractions are: whole tissue homogenate, Whole; 1000 g pellet; 10 000 g pellet; 20 000 g pellet; 100 000 g pellet; female, F; male, M.

Figure 8. Mean immunoreactivity for Rab4 in the sarcolemma-enriched (open bars) and endosome-enriched (filled bars) fractions.

Data are presented as mean optical density in ratio to background. *P < 0.01 versus sarcolemma-enriched.

Cellular distribution of PKCɛ, Kir6.2 and SUR2

To first confirm the utility of our fractionation technique in assessing changes in localization of PKCɛ, a set of experiments were performed to test whether PKC activation by β-phorbol ester (PMA, 100 nm; a known stimulus for PKC isozyme translocation; Newton & Johnson, 1998; Budas et al. 2007) could induce measurable enrichment of PKCɛ in the various membrane fractions. Figure 9 shows that PMA enriches all membrane fractions with PKCɛ, but does not alter whole-homogenate PKCɛ.

Figure 9. Western immunoblot for total PKCɛ with or without PMA pretreatment (100 nm) across all cellular fractions.

All densitometry values are normalized to the control value for the corresponding sex within the same cellular fraction. Cellular fractions are presented in terms of their respective enrichment: whole tissue homogenate, Whole; sarcoplasmic reticulum-enriched, S/ER, mitochondria-enriched, Mito.; sarcolemma-enriched, Sarc.; endosome-enriched, EV. Male, filled bars; female, open bars.

In the context of our primary hypothesis, that the sarcolemmal fraction is sex-specifically enriched with PKCɛ, we did not observe any evidence of differential enrichment or trafficking of PKCɛ between males and females in any cellular fraction (all P > 0.05). However, we did observe a chelerythrine-dependent decrease in sarcolemmal p-PKCɛ/total PKCɛ among both males and females (control = 0.43 ± 0.02; Chel = 0.29 ± 0.01; P < 0.01; Fig. 10).

Figure 10. Mean Western immunoblot data for p-PKCɛ/total PKCɛ among males (filled bars) and females (open bars) within the sarcolemma-enriched fraction.

Data are presented as optical density (OD) ratio, ±s.e.m. **Main effect of chelerythrine, P < 0.01.

The relative concentration of Kir6.2 did exhibit sex specificity in the sarcolemmal fraction, where in the absence of chelerythrine pretreatment it was 13% higher among females not exposed to GISC (P < 0.05). Pretreatment with 2 μm chelerythrine removed this sex difference whether or not the hearts underwent ischaemic challenge (Fig. 11A). Collectively these differences were observed as a main effect of sex and a sex × chelerythrine interaction, and importantly, all differences for Kir6.2 occurred in the absence of any differences for NCX1 (Fig. 11B). Endosome-enriched Kir6.2 concentrations exhibit a main effect of sex due largely to increased Kir6.2 concentration in response to PKC blockade. Figure 12A shows that chelerythrine resulted in an average difference between sexes of approximately 23%, which was significant as a main effect of sex (P < 0.05), and exhibited a trend for sex × chelerythrine interaction (P < 0.1). These data suggest that chelerythrine may have caused sex-specific internalization of Kir6.2 among females. When Kir6.2 immunoreactivity is adjusted for the difference in protein content between the sarcolemma- and endosome-enriched fractions, the reduction in female sarcolemmal Kir6.2 attributable to chelerythrine is capable of accounting for approximately 77% of the corresponding increase in endosomal Kir6.2.

Figure 11. Mean immunoreactivity for Kir6.2 (A) and NCX1 (B) among males (filled bars) and females (open bars) within the sarcolemma-enriched fraction.

Kir6.2 data are presented as optical density (OD) ratio to NCX1, and NCX1 data are presented as raw optical density, both ±s.e.m. *P < 0.05 vs. female control, #P < 0.05 vs. male control.

Figure 12. Mean immunoreactivity for Kir6.2 (A) and NCX1 (B) among males (filled bars) and females (open bars) within the endosome-enriched fraction.

Kir6.2 data are presented as optical density (OD) ratio to NCX1, and NCX1 data are presented as raw optical density, both ±s.e.m. #ANOVA main effect (Sex, P < 0.05), †drug × sex interaction trend (P < 0.1), *P < 0.05 vs. male (same condition).

The prevalence of SUR2 after co-immunoprecipitation with anti-Kir6.2, suggests a similarly elevated (29%) prevalence of mature sarcK_ATP in the sarcolemma-enriched fractions of non-ischaemic females (P < 0.05). This difference was also eliminated by chelerythrine (Fig. 13).

Figure 13. Mean immunoreactivity for SUR2 after co-immunoprecipitation by anti-Kir6.2, among males (filled bars) and females (open bars) within the sarcolemma-enriched fraction.

Data are presented as mean optical densities normalized to background, ±s.e.m. *P < 0.05 vs. male (same condition).

Discussion

Data collected by several independent groups demonstrate that the C-kinases play a central role in sex-specific cardioprotection, and the majority of those data suggest that PKCɛ is the predominant protective isozyme (Bae & Zhang, 2005; Hunter et al. 2007; Shinmura et al. 2008). Here we further demonstrate that SSC requires function of the PKC family, and elucidate several novel characteristics of the mechanism by which PKC may exert its protective effects in a sex-specific manner. First, abrogation of SSC by generalized catalytic blockade of the PKC family does not sum with abrogation due to blockade of sarcK_ATP. This suggests that, in the context of SSC, some PKC isozyme plays a role in the same mechanistic pathway as sarcK_ATP. Second, bulk translocation of the PKC isozyme thought to be most important in SSC (PKCɛ) to the mitochondria, sarcolemma, sarco-endoplasmic reticulum, or endosomal vesicles does not appear to exhibit sex specificity in response to 30 min of global ischaemia, and is therefore unlikely to underlie SSC. Third, catalytic blockade of PKC results in decreased PKCɛ phosphorylation at the sarcolemma, but this effect does not exhibit sex specificity. Fourth, this decrease in phosphorylation of sarcolemmal PKCɛ is associated with reciprocal changes in the sarcolemmal and endosomal concentrations of the essential sarcK_ATP subunit Kir6.2, and removal of elevated female sarcolemmal concentration of mature sarcK_ATP. Collectively, these data suggest that the reliance of SSC upon function of PKC may result from PKC-mediated maintenance of elevated sarcolemmal concentrations of sarcK_ATP among females.

Previous work has shown that the female myocardium is intrinsically resistant to I/R tissue injury as measured by infarction, and in some circumstances, post-ischaemic functional recovery (Node et al. 1997; Tsai et al. 2002; Bae & Zhang, 2005; Brown et al. 2005b; Gabel et al. 2005). As observed currently and previously (Bae & Zhang, 2005), this intrinsic cardioprotection appears to be PKC dependent. In the current study, we observed sex-specific infarct-resistance of a magnitude (37% lower infarct size among females) very similar to that apparent in the broader literature (Song et al. 2003; Bae & Zhang, 2005; Brown et al. 2005b; Johnson et al. 2006; Chicco et al. 2007). Importantly, while the infarct-resistance characteristic of females is subject to blockade of both PKC (a 28–34% expansion of infarct size by Chel alone, Figs 1 and 2) and sarcK_ATP (68% expansion of infarct size by HMR alone) those effects do not sum when the antagonists are applied in combination (63% expansion of infarct size by HMR + Chel). To our knowledge these are the first data to suggest that sarcK_ATP and PKC have requisite roles in the same pathway of SSC. Indeed, our observation that pre-ischaemic PKC blockade alone results in an infarct size expansion nearly identical in magnitude to the difference between males and females, while HMR expands infarct size much further, supports a model in which PKC exerts an effect upstream of sarcK_ATP that is responsible for precipitating SSC. This expansion of female infarct size by HMR-1098, beyond the level attributable to SSC, is not unique to this investigation but has been a consistent feature of our studies of sarcK_ATP (Brown et al. 2005a; Johnson et al. 2006; Chicco et al. 2007). As such, the observed redundancy of PKC and sarcK_ATP blockade among females suggests that PKC may be the physiological modulator of sarcK_ATP responsible for permitting sex-specific cardioprotection. Further, activation of this PKC-sarcK_ATP axis not only provides an explanation for how the female myocardium is afforded resistance to I/R challenge, but also, how it may simultaneously be exposed to a widely observed deficit in basal repolarization reserve (Shimoni, 1999; Kurokawa et al. 2008, 2009). Given that 17β-oestradiol causes acute impairment of unchallenged cardiac repolarization (Kurokawa et al. 2008, 2009), and that sarcK_ATP conductance is primarily determined by cellular metabolic status, an intriguing hypothesis is that female sex hormones act to shift myocardial repolarization towards a more metabolically sensitive phenotype. Certainly this model requires further empirical support, but may be capable of integrating the findings from a variety of investigations.

Unlike some previous studies (Bae & Zhang, 2005; Gabel et al. 2005; Willems et al. 2005) we did not observe any sex-specific protection against post-ischaemic functional impairment. Regarding this discrepancy, it would appear that manifestation of SSC-associated cardiac functional preservation is highly dependent on the form of I/R challenge chosen for investigation. To our knowledge, all studies that have demonstrated improved post-ischaemic function among females have also employed a global ischaemic stress (Bae & Zhang, 2005; Gabel et al. 2005; Willems et al. 2005; Wang et al. 2006; Hunter et al. 2007), whereas those employing regional ischaemia have shown improved infarct-resistance among females in the absence of improved functional recovery (Lee et al. 2000; Tsai et al. 2002; Booth et al. 2005; Brown et al. 2005a,b; Chicco et al. 2007). Here, and in most of our studies, we have chosen to apply regional rather than global ischaemia because it better represents the physiological conditions of human myocardial infarction. Given that regional ischaemia involves substantially less of the ventricular myocardium than global ischaemia (∼38% of the LV myocardium in our studies, see Fig. 1B), it is probably a less powerful method for inducing left ventricular functional deficit.

In view of findings from other groups that sarcK_ATP activity can be increased by PKCɛ (Hu et al. 1996; Light et al. 2000), and given that we have previously demonstrated that SSC is eliminated by the blockade of sarcK_ATP but not mitoK_ATP (Brown et al. 2005a; Johnson et al. 2006; Chicco et al. 2007), we sought to determine if sarcolemmal enrichment by either PKCɛ or sarcK_ATP is sex dependent. First and foremost, we found that the sarcolemmal and endosomal concentrations of the pore-forming sarcK_ATP subunit, Kir6.2, exhibit reciprocal and sex-specific changes after pre-ischaemic administration of chelerythrine. These effects, and the parallel sarcolemmal changes in anti-Kir6.2 immunoprecipitated SUR2, suggest that constitutive PKC-dependent processes act to maintain an elevated sarcolemmal concentration of mature sarcK_ATP in females. This finding offers key insight into the manner by which the PKC enzyme family may interact with sarcK_ATP in the mechanism underlying SSC.

Several previous studies have investigated the possibility that trafficking of sarcK_ATP is involved in cardioprotected phenotypes (Hu et al. 2003; Budas et al. 2004; Garg & Hu, 2007; Sukhodub et al. 2007). Most notably, Budas et al. (2004) used a variety of techniques to show that trafficking of sarcK_ATP to the sarcolemma occurs at the onset of an ischaemic bout subsequent to hypoxic preconditioning (HPC), but not in the absence of prior HPC. Furthermore, general blockade of protein trafficking abolished the protection afforded by HPC in that study. In a subsequent study, the same group has shown that AMP-activated protein kinase may be required for anterograde translocation in response to ischaemic preconditioning (Sukhodub et al. 2007). The remaining two investigations have directly assessed the effect of PKC activity on the cellular distribution of myocardial K_ATP subunits (Hu et al. 2003; Garg & Hu, 2007). Hu et al. (2003) describe a biphasic effect of the PKC agonist, PMA, characterized by a sharp early (5 min after treatment) induction of I_KATP and subsequent delayed decay (30 min after treatment). They also present data to suggest that this delayed decay occurs after exposure to adenosine, which is known to activate PKC during ischaemic preconditioning (Cohen et al. 2000). The authors interpret this decay as indicative of a counter-regulatory internalization of sarcK_ATP that may protect against the detrimental effects of prolonged induction of I_KATP. Garg & Hu (2007) collected data to suggest that recombinant Kir6.2 is trafficked to the mitochondria in response to identical pharmacological activation by PMA. These articles have provided valuable insight into how PKC may act to control K_ATP subunit traffic in both heterologous and native cell models. However, for several reasons the pharmacological activation applied by both studies prevents their findings from being easily interpreted in terms of endogenous PKC activity. Chief among these is the likely divergence in isoform specificity of the PKC responses to phorbol ester and adenosine, as compared to ischaemia (Tsouka et al. 2002; Vasara et al. 2003). It has been recognized by a number of investigators that phorbol ester activation is unlikely to closely represent physiological control of the C-kinases (Ahmed et al. 1993; Kazanietz et al. 1995; Brooks et al. 1997; Dorit & Kazanietz, 1999), and diversity in action of the various PKC isoforms is well appreciated in the context of myocardial I/R injury (Budas, 2007). To date this isoform specificity has not been described for the interaction of PKC and sarcK_ATP, although it is certainly possible that it contributes to the distinction between our current observations and those of Hu et al. (2003) and Garg & Hu (2007). Additionally or concurrently, it has previously been shown that phorbol ester activation causes ubiquitylation and degradation of PKC detectable 30 min after treatment (Lu et al. 1998). In light of this, an alternative interpretation of the delayed diminution of I_KATP observed by Hu et al. (2003) is that the primary action of PKC is to cause enrichment of the sarcolemmal membrane with sarcK_ATP, and decay of this effect occurs secondary to degradation of PKC. Regarding the observations of Garg & Hu (2007) it should also be noted that a large number of investigations, using a variety of experimental models, have attempted to isolate endogenous Kir6 from the mitochondria. To date no study has been capable of sequencing any subspecies of the Kir6 family from mitochondrial isolates, although a broad range of studies have associated mitochondrial anti-Kir6 immunoreactivity with the electrophysiological manifestations of an ATP-sensitive potassium channel in the mitochondria (Lacza et al. 2003; Jiang et al. 2006). Furthermore, a recent detailed analysis suggests that the mitochondrial immunoreactivity associated with anti-Kir6.1 antibodies is largely non-specific (Foster et al. 2008). For these reasons we consider the current data to be the most physiologically representative of those available, and suggest that PKC-mediated maintenance of sarcolemmal sarcK_ATP concentration is a plausible means by which SSC relies upon function of both sarcK_ATP and PKC.

The second novel finding of this study is that PKCɛ does not exhibit sex-specific enrichment at the sarcolemma, or in any other cellular fraction. To our knowledge this is the only investigation to have directly addressed the question of whether PKCɛ enrichment of these depots occurs differently between the sexes. To date only three studies have focused on the role of PKCɛ in SSC (Bae & Zhang, 2005; Hunter et al. 2007; Shinmura et al. 2008). Of these, the two that have investigated the prevalence of PKCɛ at discrete cellular loci have also employed ovariectomy to dissect the mechanisms underlying the protective effects of female sex hormones (Hunter et al. 2007; Shinmura et al. 2008). Hunter et al. (2007) associated changes in mitochondrial, nuclear and cytosolic PKCɛ prevalence among females (due to increasing age, ovariectomy and I/R) with corresponding alterations in resistance to I/R injury. While the data presented here do not directly comment on their primary finding, which was related to the effects of ageing and ovariectomy, a secondary finding of their study was that I/R challenge among hearts from young intact females decreases mitochondrial PKCɛ. In the second study, Shinmura et al. (2008) examined the interaction of ovariectomy and oestradiol rescue with ischaemic preconditioning (IPC). They observed that IPC caused an opposite effect to that observed by Hunter et al. (2007) in that it increased total PKCɛ in the membranous fraction isolated from intact young females. This effect was removed by ovariectomy, and rescued by exogenous oestrogen supplementation. Of the several potential reasons for these differing results, one that is capable of explaining the findings of both studies is the different subcellular preparations employed by each. The membranous fraction collected by Shinmura et al. (2008) was relatively crude in that it included mitochondrial, sarcolemmal and microsomal constituents, as well as other non-nuclear membrane components. Hunter et al. (2007) focused on a more refined mitochondrial fraction, unlikely to include marked contributions from the sarcolemmal and endosomal protein pools. This difference in cellular fractionation may be responsible for the disparate findings of those two studies, and if so, it is implied that trafficking of PKCɛ to non-mitochondrial components of the cruder membranous fraction must have occurred, and that this process may be associated with SSC. Pursuing this logic, a primary objective of the current study was to delineate the contributions of non-mitochondrial components within the membranous fraction collected by Hunter et al. (2007), with a particular focus on PKCɛ enrichment at the sarcolemma and endosomal vesicles. The data presented here suggest that ischaemic challenge does not sex-specifically alter the prevalence of PKCɛ at the sarcolemma (Fig. 11) or endosomal pool (not shown). As such, PKCɛ enrichment at these two depots is unlikely to explain disagreement between the two previous studies. We suggest that other key differences between the studies, such as the severity of the applied ischaemic stress (Cohen et al. 2000; Budas et al. 2007), may explain the equivocal observations.

Given the current findings, we propose that chelerythrine may remove sex-specific enrichment of sarcolemmal sarcK_ATP by preventing PKC from inhibiting constitutive internalization of the channel complex. Chelerythrine is an isoform non-specific catalytic inhibitor of PKC (Herbert et al. 1990; Chmura et al. 2000), and activation of the epsilon isozyme requires two autophosphorylations including that at Ser729 (Newton & Johnson, 1998). Thus, we interpret the observed decrease in sarcolemmal p-PKCɛ/total PKCɛ to be representative of general catalytic inhibition of all PKC isozymes present at the sarcolemma (Fig. 14B). Combining this finding with the previous observation that females exhibit elevated whole-cell Kir6.2, and SUR2 (Ranki et al. 2001; Brown et al. 2005a,b;), we propose that an active sarcolemmal population of some combination of PKC isozymes plays a permissive role in SSC by allowing that higher whole-cell sarcK_ATP concentration to be represented at the sarcolemma. While there are few data available to describe how PKC may act to permit this reflection of elevated whole-cell sarcK_ATP at the sarcolemma, the reciprocal nature of the changes in Kir6.2 at the sarcolemma and endosomal pool observed here suggest that internalization of the channel complex may be sex-specifically inhibited by PKC (Fig. 14). In combination with the observations of Sukhodub et al. (2007) that AMPK is involved in anterograde translocation of sarcK_ATP, the possible interaction of these two key enzymes of metabolic control in regulation of the sarcolemmal sarcK_ATP concentration presents an intriguing avenue for future work.

Figure 14. Schematic representation of putative PKCɛ modulation of sarcolemmal and endosomal prevalence of sarcK_ATP in the female myocardium.

Described in the presence (A) and absence (B) of chelerythrine. 1, elevated whole-cell de novo sarcK_ATP arrives at the sarcolemma via canonical transport of integral membrane proteins. 2, if unimpeded constitutive internalization of Kir6.2 and SUR2a limits sarcolemmal prevalence of sarcK_ATP among females. 3, inhibition of (2) by active PKC allows enrichment of the sarcolemma with sarcK_ATP which is disinhibited by pre-ischaemic administration of chelerythrine (4).

Limitations

Perhaps the greatest limitation of our data in supporting the above model is that they only describe the sarcolemmal enrichment of the epsilon isoform of PKC. It is certainly possible that the female sarcolemma exhibits enrichment for some other isoform of PKC, which may be responsible for the observed alterations in sarcolemmal sarcK_ATP enrichment, and abrogation of SSC by chelerythrine. Thus, we recommend subsequent investigations to better understand the isozyme specificity of the regulatory interaction between PKC and sarcK_ATP described herein.

Summary and conclusions

This investigation reiterates the established requisite role of PKC in the myocardial infarct-resistance characteristic of female rats, and extends understanding of that role by providing three important and novel findings.

1. Combining blockade of PKC and sarcK_ATP does not result in additive expansion of infarct size among females, indicating that PKC and sarcK_ATP are likely to act in-series in a mechanism underlying SSC. This finding permits a novel model of SSC, in which PKC-mediated regulation of sarcK_ATP explains much of the protection apparent in females.

2. Female-specific sarcolemmal enrichment with sarcK_ATP requires function of one or more PKC isozymes, is associated with the relative phosphorylation of sarcolemmal PKCɛ, and is therefore likely to be involved in the mechanism by which PKC permits SSC in a sarcK_ATP-dependent manner.

3. Generalized catalytic blockade of PKC results in a female-specific increase in concentration of sarcK_ATP in a cellular fraction enriched for the early endosomal marker, Rab4. This suggests that PKC activity may prevent internalization of sarcK_ATP in females, and in doing so permit females’ higher whole-cell concentration of sarcK_ATP to be realized at the sarcolemma.

Collectively these findings suggest that the PKC-sarcK_ATP axis may be a viable target for strategies to induce chronic myocardial resistance to ischaemia–reperfusion injury. We recommend further investigations to describe isoform specificity and temporal characteristics of PKC-dependent sarcolemmal enrichment by sarcK_ATP, particularly as it relates to the different phases of I/R, and possible interaction of these effects with ischaemic preconditioning and AMPK.

Glossary

Abbreviations

Chel

chelerythrine;

GISC

global ischaemia;

HMR

HMR-1098;

I/R

ischaemia–reperfusion;

IA

infarcted area;

I_KATP

ATP-sensitive potassium current;

LAD

left anterior descending coronary artery;

LV

left ventricle;

NCX1

myocardial sodium–calcium exchanger;

PHB1

prohibitin 1;

PKC

protein kinase C;

PKCɛ

PKC epsilon isoform;

PMA

4β-phorbol 12-myristate 13-acetate;

sarcK_ATP

sarcolemmal ATP-sensitive potassium channel;

SSC

sex-specific cardioprotection;

TA

total area;

ZAR

zone at risk.

Author contributions

A.G.E.: experimental conception and design, data collection and analysis, and manuscript preparation, revision and submission. M.L.R.: data collection and analysis, and manuscript revision and approval. R.A.G.: Data collection and analysis, and manuscript revision and approval. D.K.Z.: data collection and analysis, and manuscript revision and approval. J.M.L.: data collection and analysis, and manuscript revision and approval. J.C.B.: data collection and analysis, and manuscript revision and approval. A.J.C.: experimental conception and design, data collection and analysis, and manuscript revision and approval. R.L.M. experimental conception and design, and manuscript preparation, revision and approval.