Dysfunctional survival-signaling and stress-intolerance in aged murine and human myocardium

Abstract

Changes in cytoprotective signaling may influence cardiac aging, and underpin sensitization to ischemic insult and desensitization to ‘anti-ischemic’ therapies. We tested whether age-dependent shifts in ischemia-reperfusion (I-R) tolerance in murine and human myocardium are associated with reduced efficacies and coupling of membrane, cytoplasmic and mitochondrial survival-signaling. Hormesis (exemplified in ischemic preconditioning; IPC) and expression of proteins influencing signaling/stress-resistance were also assessed in mice. Mouse hearts (18 vs. 2–4 mo) and human atrial tissue (75±2 vs. 55±2 yr) exhibited profound age-dependent reductions in I-R tolerance. In mice aging negated cardioprotection via IPC, G-protein coupled receptor (GPCR) agonism (opioid, A₁ and A₃ adenosine receptors) and distal protein kinase C (PKC) activation (4 nM phorbol 12-myristate 13-acetate; PMA). In contrast, p38-mitogen activated protein kinase (p38-MAPK) activation (1 μM anisomycin), mitochondrial ATP-sensitive K⁺ channel (mK_ATP) opening (50 μM diazoxide) and permeability transition pore (mPTP) inhibition (0.2 μM cyclosporin A) retained protective efficacies in older hearts (though failed to eliminate I-R tolerance differences). A similar pattern of change in protective efficacies was observed in human tissue. Murine hearts exhibited molecular changes consistent with altered membrane control (reduced caveolin-3, cholesterol and caveolae), kinase signaling (reduced p70 ribosomal s6 kinase; p70s6K) and stress-resistance (increased G-protein receptor kinase 2, GRK2; glycogen synthase kinase 3β, GSK3β; and cytosolic cytochrome c). In summary, myocardial I-R tolerance declines with age in association with dysfunctional hormesis and transduction of survival signals from GPCRs/PKC to mitochondrial effectors. Differential changes in proteins governing caveolar and mitochondrial function may contribute to signal dysfunction and stress-intolerance.

1. Introduction

An understanding of the molecular basis of cardiac aging, and of the influences of age on myocardial responses to disease and therapy, are increasingly important goals. Ischemic heart disease (IHD) and associated stroke are the lead causes of mortality globally (Fuster & Kelly, 2010), with age a major risk factor in their development. While aged myocardium is most likely to suffer I-R insult (~75% of infarcts occur in those >65 yrs of age), it may possess reduced resistance to I-R injury (Headrick, 1998; Rosenfeldt et al., 1999; Mariani et al., 2000; Headrick et al., 2003; Lesnefsky et al., 2006) compounded by refractoriness to protective intervention (Ferdinandy et al., 2007; Boengler et al., 2009; Peart & Headrick, 2009). These clinically relevant changes could reflect mechanistic determinants of the poorly understood aging process itself: in the ‘green hypothesis’ the expression/functionality of intrinsic ‘detoxification’ systems eliminating molecular damage and governing cellular stress-resistance is forwarded as a primary determinant of aging and longevity (Gems & McElwee, 2005). Intrinsic resistance to diverse stressors is a common (potentially defining) feature of longevity phenotypes (Harper et al., 2006), and this hypothesis is consistent with cytoprotective pathway induction with longevity extension (Shore et al., 2012), and the anti-aging effects of hormesis (Calabrese et al., 2012). Hormesis refers to beneficial biological effects (including improved resistance to injury) arising from moderate sub-lethal stressors such as hypoxia or nutrient deprivation, and is exemplified in cardiac tissue by IPC. The functionality of signaling pathways underpinning cellular stress-resistance and hormesis responses may thus govern aging/longevity, and is also highly relevant to therapeutic manipulation of IHD outcomes.

Hormesis, autophagic damage-management, longevity pathway (eg. mammalian target of rapamycin, mTOR; and insulin-like growth factor, IGF) signaling and cellular stress-resistance are regulated by membrane GPCRs and receptor tyrosine kinases (RTKs), and altered growth-factor/RTK control has been implicated in replicative senescence and cellular aging (Cho & Park, 2005; Yu & Driscoll, 2011). However, there is also evidence of impaired GPCR-dependent stress-signaling in older myocardial tissue (Schulman et al., 2001; Headrick et al., 2003; Peart et al., 2007). The expression of proteins targeted by this signaling, such as GSK3β (Kostyak et al., 2006; Hunter et al., 2007) may also be modified with age, together with mitochondrial determinants of cell survival, including Ca²⁺-sensitive K⁺ channels and the mPTP (Heinen et al., 2008; Zhu et al., 2010, 2013). Such changes may collectively repress cellular resistance and adaptation to stress (a feature of aged phenotypes), and in turn promote the aging process itself (Gems & McElwee, 2005; Calabrese et al., 2012; Shore et al., 2012). Identifying intrinsic protective mechanisms that become dysfunctional or retain efficacy with age can thus unmask mechanistic aspects of biologic aging, and also reveal molecular targets for the manipulation of myocardial resistance to injury/disease.

In the present study we test whether age-related intolerance to stress (specifically, clinically relevant I-R) is associated with changes in membrane/GPCR sensitive cytoprotective signaling and hormesis, and emergence of a molecular profile favoring such dysfunction. Responses to the following stimuli were assessed to localize age-dependent changes in survival signaling, and to identify targets for manipulation of I-R outcomes in aged tissue: acute hormesis via IPC; agonism of membrane GPCRs mediating stress-resistance/hormesis (opioid and adenosine A₁ or A₃ receptors); activation of signal kinases (PKC and p38-MAPK) transducing GPCR/IPC responses (Headrick et al., 2003; Peart & Gross, 2006; Fenton et al., 2010); and modulation of mitochondrial effectors governing cell viability (mK_ATP channels and the mPTP). In murine tissue we also tested for shifts in determinants of membrane receptor signaling (caveolin-3, membrane cholesterol and GRK2), survival-kinase signal transduction (protein kinase B or AKT, p70s6K, p38-MAPK, and extracellular signal-regulated kinase 1/2 - ERK1/2), and mitochondrial dysfunction/cell death (GSK3β, GRK2, caspase-3, and cytochrome c).

2. Materials and methods

Investigations conformed to the guidelines of the Animal Ethics Committee of Griffith University. Male C57Bl/6 mice were sourced from the Animal Resources Centre (Canning Vale, WA, Australia). Young (2–4 mo; 24.6±0.5 g body weight) and old (18 mo; 37.1±1.9 g body weight) mice were studied, with hearts removed from terminally anaesthetized animals for Langendorff perfusion. The 18 mo group was studied as an ‘aged’ phenotype, midway between reproductive senescence (12–14 mo in females) and senescence based on median mortality (24–28 mo in C57Bl/6 mice). Onset of myocardial fibrosis is detectable at 18 mo (Willems et al., 2005), consistent with an aged cardiac phenotype. Spontaneous mortality to this age (10–20%) is also equivalent to mortality by 70–75 yrs in human (developed country) populations - the older age group in human tissue analysis here.

2.1. Chemicals

All chemicals and drugs were purchased from Sigma-Aldrich (St. Louis, MO), and antibodies sourced from Cell Signaling Technology (Danvers, MA) and Santa Cruz Biotechnology (Santa Cruz, CA).

2.2. Perfused murine heart model

Mice were anesthetized with sodium pentobarbital (60 mg/kg intraperitoneally), and hearts were excised and perfused in a Langendorff mode with modified Krebs-Henseleit buffer delivered via the aorta at a pressure of 80 mmHg, as described previously (Headrick et al., 2003; Peart et al., 2007). Hearts underwent 20 min of normoxic stabilization at intrinsic heart rates after ventricular balloon placement, followed by 10 min of pacing at 420 beats/min. Hearts were then subjected to 20 min global normothermic ischemia followed by 60 min of aerobic reperfusion. Myocardial cell disruption/death was assayed by quantitating post-ischemic washout of cellular lactate dehydrogenase (LDH). We have previously established that LDH efflux correlates well with infarct size and mechanical dysfunction in this model (Peart & Headrick, 2003). Venous effluent was collected on ice throughout reperfusion before enzymatic determination of total LDH washout (IU/g tissue).

Mouse hearts were either untreated (n=17 for young, 13 for aged), or subjected to IPC (3 × 90 s ischemia interposed with 120 s reperfusion; n=8 for young, 8 for aged) or 10 min pre-ischemic and 5 min post-ischemic treatment with: 10 μM of the broad-spectrum opioid receptor agonist morphine (n=9 for young, 7 for aged); 0.1 μM of the A₁ adenosine receptor agonist N⁶-cyclopentyladenosine (CPA; n=10 for young, 8 for aged); 0.1 μM of the A₃ adenosine receptor agonist 2-chloro-N⁶-(3-iodobenzyl)adenosine-5′-N-methyl-carboxamide (Cl-IB-MECA; n=9 for young, 7 for aged); 4 nM of the PKC activating phorbol ester PMA (n=9 for young, 8 for aged); 1 μM of the p38-MAPK activator anisomycin (n=8 for young, 8 for aged); 50 μM of the putative mK_ATP channel opener diazoxide (n=10 for young, 8 for aged); or 0.2 μM of the mPTP inhibitor cyclosporin A (n=9 for young, 8 for aged). Normoxic time-course experiments were undertaken in young (n=6) and aged (n=5) hearts subjected to 80 min of normoxic perfusion following initial stabilization.

2.3. Human atrial pectinate trabecular model

Right atrial myocardium was sampled from middle-aged and aged cardiac patients with mean ages of 55±2 (n=10) and 75±2 (n=10) yrs, respectively, and functionally assessed for I-R tolerance as detailed by us previously (Rosenfeldt et al., 1999; Mariani et al., 2000). Resected atrial appendage was acquired from patients undergoing elective coronary artery bypass graft or valve operations with cardiopulmonary bypass at the Alfred Hospital in Melbourne, under approval of the Human Research Ethics Committee for Discarded Tissue. Exclusion criteria included: re-operation; urgent or emergency procedures; current therapy with antioxidants; and recent myocardial infarction (≤6 weeks).

Atrial tissue was dissected under microscopy in Ringer’s solution containing 30 mM 2,3 butanedione monoxime (BDM) to yield 5 pectinate trabeculae (~1 mm diameter, ~7 mm long) from each patient. Muscle strips were connected to a force transducer and held between 2 field-stimulation electrodes in normoxic BDM-free Ringer’s solution (37°C, PO₂ ≥450 mmHg). Strips were initially maintained un-stretched and stimulated at 1 Hz for 30 min, before incremental stretching to lengths yielding maximal force development, and a further 30 min of stabilization. The 5 muscle strips per patient were then assessed as follows: one was assigned to a 60 min normoxic time-course while the remaining 4 strips were subjected to 30 min of simulated ischemia/30 min reperfusion, and were either untreated or pre-treated for 30 min with 10 μM morphine, anisomycin or diazoxide. Serial measures of contractile function were made every 10 min. As described previously (Mariani et al., 2000), the key elements of ischemia - restriction of O₂ and substrate delivery, and accumulation of tissue metabolites - are induced by draining sealed organ baths of oxygenated Ringer’s (thus O₂ and metabolic substrate) and maintaining strips in humidified anoxic gas mix (95% N₂/5% CO₂) at 37°C for 30 min, with stimulation via punctate electrodes. Reperfusion is effected by reintroduction of normoxic Ringer’s solution (restoring O₂ and substrate delivery, washing out accumulated metabolites), with tissue paced at 1 Hz for a 30 min recovery period.

2.4. Protein changes in aging mouse heart

Additional to or facilitating changes in signal transduction, total quantities of proteins influencing cellular stress-resistance may change with age (total protein levels dictating maximal cellular activities, with post-translational events modulating this activity). To test for emergence of a stress-intolerant expression profile, we quantitated protein levels via immunoblot analysis of ventricular lysate samples from young and old hearts (n=6 for each), isolated from anesthetized animals as outlined previously (Peart et al., 2007). We assessed expression of the membrane orchestrator of survival-signaling caveolin-3, together with downstream kinases AKT, p70s6K, ERK1/2 and p38-MAPK implicated in tissue protection. In terms of proteins promoting cellular injury, we assessed expression of GRK2 and GSK3β which both induce mitochondrial dysfunction and cell death (Juhaszova et al., 2009; Brinks et al., 2010; Chen et al., 2013; Zhu et al., 2013), while GRK2 additionally inhibits GPCR signaling.

Lysate samples containing 30 μg of total protein were loaded onto 10% acrylamide gels and electrophoresed at 150 V for 1.5 hrs. Proteins were transferred to polyvinylidene difluoride membranes and blocked in 5% skim milk powder in Tris-buffered saline with 0.05% Tween-20 (TBST) for 60 min. Membranes were incubated with 1:1000 dilutions of primary antibodies overnight at 4°C (from Cell Signaling Technology Inc., Danvers, MA, USA: GRK2, #3982; AKT, #9272; p70s6K, #9202; ERK1/2, #9102; p38-MAPK, #9212; and GSK3β; #9315; from Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA: caveolin-3, sc-5310). Following 3 washes in TBST, membranes were incubated with secondary antibody and visualized on a ChemiDoc XRS system (Bio-Rad, Hercules, CA, USA). Equal loading was confirmed by co-analysis of β-actin or glyceraldehyde 3-phosphate dehydrogenase (GAPDH). For purposes of comparison protein expression levels were normalized to values for young hearts.

Active caspase-3 and cytosolic cytochrome c content were assayed as measures of apoptotic potential. Caspase-3 activity was measured via a commercial kit (Clontech Laboratories Inc., Mountain View, CA, USA) for fluorometric detection of 7-amino-4-trifluoromethyl coumarin (AFC) cleaved from a synthetic DEVD (Asp-Glu-Val-Asp) substrate. Relative caspase-3 activities determined from fluorescence change/mg protein were normalized to values for young hearts. Cytochrome c was assayed in crude cytosolic fractions from ventricular lysate samples using an immunoassay kit (R&D Systems, Minneapolis MN), with levels expressed per mg protein. Finally, total cholesterol content was assayed in crude membrane fractions from cardiac lysates using an Invitrogen Amplex Red Cholesterol kit (Life Technologies Australia Pty Ltd, VIC, Australia). Fluorescence change was measured on a microplate reader (Tecan Australia Pty Ltd, VIC, Australia) with excitation and detection at 560 nm and 590 nm, respectively. Cholesterol content per mg protein was determined from standard curves acquired with each analysis, and data normalized to values for young hearts.

To further explore effects of age on the membrane compartment, immunofluorescence and electron microscopy (EM) were employed to examine membrane ultrastructure, caveolin-3 expression/localization, and caveolar density. For EM, tissue from young and aged mice was fixed with 2.5% glutaraldehyde in 0.1 M cacodylate buffer for 2 h, post-fixed in 1% OsO4 in 0.1 M cacodylate buffer (1 h), and embedded in LX-112 (Ladd Research, Williston, VT, USA), as outlined recently (Fridolfsson et al., 2012). Sections were stained in uranyl acetate and lead citrate and assessed via EM (Jeol 1200 EX-II; Jeol Ltd., Akishima, Japan). For immunohistochemistry, fixed tissue was sectioned and processed as previously described (Head et al., 2005).

2.5. Statistical analysis

All data are expressed as means ± the standard error of the mean (SEM). Age and treatment effects were statistically identified via 2-way ANOVA (employing repeated-measures for functional responses assessed in Fig. 1), with a post-hoc Newman-Keuls test employed for specific comparisons. Significance in all tests was accepted for P<0.05.

Fig. 1.

Effects of I-R on contractile dysfunction and cell damage in young to senescent murine (A and B) and human (C) myocardium. Functional outcomes and LDH efflux (inset to A) were assessed in Langendorff hearts subjected to 20 min ischemia/60 min reperfusion. Final diastolic and developed pressures are shown for normoxic hearts (time-course non-ischemic controls) from young and aged mice (open and closed squares, respectively). Functional outcomes were assessed in human trabeculae subjected to 30 min ischemia and 30 min or longer reperfusion (C). Recovery of LV end-diastolic pressure (mmHg) and LV developed pressure (mmHg) is shown, together with post-ischemic LDH efflux (see panel within A). Stability of function in normoxic control (non-ischemic) tissue is also shown for the 55 and 75 yr old groups (open and closed squares, respectively). * P<0.05 vs. Young. Values are reported as mean ± SEM (murine: n=13–17 per group; human: n=10 per group).

3. Results

3.1. Age-dependence of I-R injury in murine and human myocardium

Heart mass increased by ~50% with age, while a modest rise in the heart:body weight ratio to 18 mo was not statistically significant (Table 1). Normoxic contractile function did not vary substantially between young and aged hearts (Table 1), although we observed a significant 10% fall in spontaneous beating rate and an insignificant trend towards reduced dP/dt (by ~15%, suggestive of emerging hypo-contractility). Despite comparable mechanical function, normoxic coronary perfusion rate was significantly reduced in older myocardium (Table 1). Stability of contractile function over 80 min of normoxic perfusion was also comparable in young and aged hearts, with no change in diastolic pressure (Fig. 1A) and similar maintenance of force development (86±6% and 83±7% of baseline force, respectively; Fig. 1B). A 20 min ischemic insult resulted in sustained contractile dysfunction throughout 60 min of reperfusion (Fig. 1A & 1B), in association with significant cellular damage as indicated by LDH efflux (Fig. 1A inset). Contractile dysfunction and cellular injury were markedly exaggerated in aged vs. young hearts.

Table 1.

Baseline cardiac function in perfused hearts from young and aged mice.

Group	Heart Wt (mg)	Heart:Body Weight	EDP (mmHg)	LVDP (mmHg)	Heart Rate (beat/min)	+dP/dt (mmHg/s)	−dP/dt (mmHg/s)	Coronary Flow (ml/min/g)
Young (n=17)	122±5	4.87±0.12	4±2	135±4	385±10	5876±367	−4573±413	23.0±1.6
Aged (n=13)	184±7*	5.08±0.10	3±1	146±8	344±11*	5118±395	−3677±248	16.9±1.2*

Heart rate is the spontaneous rate prior to commencement of pacing at 420 bpm. All other values were recorded immediately prior to onset of ischemia. EDP, left ventricular end-diastolic pressure; LVDP, left ventricular developed pressure; + and −dP/dt, peak differentials of LV pressure development and relaxation, respectively.

^*P<0.05 vs. Young. Values are reported as mean ± SEM (n=13–17 per group).

Isolated human atrial trabecular strips from 55 and 75 yr age groups did not differ in terms of size or baseline contractile and relaxation properties (Table 2). The functional stability of normoxic trabeculae over a 90 min period was also similar for the 55 and 75 yr age groups (90±9% and 87±10% of basal force development, respectively; Fig. 1C). Atrial tissue subjected to simulated ischemia failed to recover fully during a subsequent 30 min reperfusion period, with relative functional recovery in old tissue approximately half that for younger trabeculae (Fig. 1C).

Table 2.

Baseline properties of human atrial tissue segments from 55 and 75 yr old cohorts.

Age	Weight (mg)	Lmax (mm)	CSA (mm2)	DT (mg)	RT (mg)	tPT (ms)	tR50% (ms)	CD (ms)
55±2 (n=10)	6.8±1.2	7.4±1.1	0.8±0.1	1147±150	294±77	90±13	96±11	187±11
75±2 (n=10)	7.0±1.1	7.8±1.1	0.9±0.2	1215±151	357±81	95±8	101±11	196±11

Tissue dimensions are provided (weights and cross-sectional areas, CSA) together with baseline contractile parameters assessed immediately prior to simulated ischemia (or parallel normoxic control conditions). Contractile parameters measured and reported include: L_max, the tissue length yield maximal force or tension development; DF, developed tension (peak tension minus resting tension); RT, resting tension; tPT, time taken to develop peak tension per contraction; tR50%, time taken to reach 50% relaxation post-contraction; CD, contraction duration.

^*P<0.05 vs. 55 yrs. Values are reported as mean ± SEM (n=10 per group).

3.2. Age-dependence of protective stimuli in murine myocardium

Induction of IPC together with pre-ischemic agonism of opioid and adenosine A₁ and A₃ GPCRs significantly enhanced I-R tolerance in young hearts, whereas these stimuli all failed to modify outcomes in aged hearts (Fig. 2). Protection in young hearts was evident as improvements in diastolic function (Fig. 2A), pressure development (absolute and % of pre-ischemia; Figs. 2B and C) and LDH efflux (Fig. 2D). The relative protective efficacies of these receptor-dependent stimuli within young hearts were comparable. The cardioprotective effects of distal signaling kinase activation varied with age (Fig. 3). As for GPCR agonism, pre-ischemic treatment with the PKC activator PMA markedly improved post-ischemic outcomes in young but not aged hearts (Fig. 3). In contrast, p38-MAPK activation with anisomycin was effective in both ages, though the effect in older hearts was not as robust as in younger tissue. Targeted modulation of mitochondrial end-effectors proved effective in both young and aged mouse hearts, though responses appeared moderately reduced with age (Fig. 4). The putative mK_ATP channel opener diazoxide improved contractile recoveries by 30–50% in young and aged hearts, while the mPTP inhibitor cyclosporin A exerted slightly less improvement in outcomes. Since mK_ATP channel activation and mPTP inhibition are protective in aged hearts, failed responsiveness to PKC activation (Fig. 3) likely originates upstream of these mitochondrial effectors. The impaired efficacy of PMA, and of cyclosporin A, did not appear dose-dependent: trials of higher concentrations (10 nM PMA, 2 μM cyclosporin A; n=4–5) failed to further improve outcomes (data not shown). Despite profound protection via mK_ATP activation and mPTP inhibition in both age groups, myocardial ischemic tolerance was not normalized across ages with a substantial 2-fold difference in cellular damage persisting in aged vs. young hearts (Fig. 4).

Fig. 2.

Protective efficacies of IPC and GPCR stimuli in young (open bars) and aged (shaded bars) murine hearts. Effects of IPC, and opioid (morphine) and A₁ and A₃ adenosine receptor agonism (CPA and Cl-IB-MECA, respectively) are shown on: A) post-ischemic recovery of LV end-diastolic pressure; B) post-ischemic recovery of LV developed pressure; C) % post-ischemic recovery of LV developed pressure; and D) post-ischemic efflux of LDH. † P<0.05 vs. Young; * P<0.05 vs. Untreated. Values are reported as mean ± SEM (n=7–17 per group).

Fig. 3.

Protective efficacies of survival kinase stimuli in young (open bars) and aged (shaded bars) murine hearts. Effects of PKC (PMA) and p38-MAPK (anisomycin) activation are shown on: A) post-ischemic recovery of LV end-diastolic pressure; B) post-ischemic recovery of LV developed pressure; C) % post-ischemic recovery of LV developed pressure; and D) post-ischemic efflux of LDH. † P<0.05 vs. Young; * P<0.05 vs. Untreated. Values are reported as mean ± SEM (n=8–17 per group).

Fig. 4.

Protective efficacies of mitochondria-dependent stimuli in young (open bars) and aged (shaded bars) murine hearts. Effects of mK_ATP channel activation (diazoxide) and mPTP inhibition (cyclosporin A) are shown on: A) post-ischemic recovery of LV end-diastolic pressure; B) post-ischemic recovery of LV developed pressure; C) % post-ischemic recovery of LV developed pressure; and D) post-ischemic efflux of LDH. † P<0.05 vs. Young; * P<0.05 vs. Untreated. Values are reported as mean ± SEM (n=8–17 per group).

3.3. Age-dependence of protective stimuli in human myocardium

Similar to observations in murine hearts, opioid receptor agonism with morphine failed to protect older human atrial tissue while exerting significant benefit in younger tissue (Fig. 5). Moreover, the p38-MAPK activator anisomycin and the mK_ATP opener diazoxide both improved recoveries in young and aged human tissue (Fig. 5), as observed in murine studies. Thus, as for perfused murine hearts, GPCR activation was ineffective in older human atrial tissue while p38 and mitochondrial effector modulation conferred benefit across ages. Moreover, as in mouse hearts the age-dependent difference in human atrial ischemic tolerance was reduced but not eliminated by mK_ATP (or p38-MAPK) activation (Fig. 5).

Fig. 5.

Functional recoveries from 30 min I-R in human atrial trabeculae preparations from 55±2 and 75±2 yrs cohorts. Data are shown for the % recovery of baseline normoxic (pre-ischemic) developed force in the absence and presence of pretreatment with a GPCR stimulus (10 μM morphine), a protein kinase activator (10 μM anisomycin) and a mitochondrially targeted stimulus (10 μM diazoxide). † P<0.05 vs. younger age group; * P<0.05 vs. Untreated. Values are reported as mean ± SEM (n=10 per group).

3.4. Age-dependent shifts in myocardial molecular makeup

Analysis of protein expression in murine hearts (Table 3) revealed that older tissue expresses less caveolin-3 (a key orchestrator of membrane receptor signaling) and p70s6K (a pro-survival signal kinase that phospho-inhibits GSK3β), and increased levels of GRK2 (an inhibitor of GPCR signaling that promotes mitochondrial injury/cell death) and GSK3β (which also promotes mitochondrial dysfunction and cell death). On the other hand, levels of the survival kinases AKT, ERK1/2 and p38-MAPK did not differ significantly between young and aged hearts (Table 3). Cytosolic cytochrome c content was increased in aged hearts, whereas active caspase-3 content was comparable in both groups. Membrane cholesterol content, critical to caveolar formation and control, was also significantly reduced in aged tissue (Table 3).

Table 3.

Stress-intolerant molecular profile in aged vs. young murine hearts.

	SARCOLEMMAL		PRO-SURVIVAL				PRO-INJURY		PRO-APOPTOSIS
Group	CAV-3	CHOL	AKT	p70s6K	ERK	p38	GRK2	GSK3β	CASP-3	Cyt-C
Young (n=6)	100±3	100±4	100±8	100±6	100±9	100±12	100±11	100±8	100±25	740±24
Aged (n=6)	75±5*	77±5*	110±9	78±7*	106±13	109±8	162±14*	142±10*	110±19	973±38*

Myocardial protein expression and membrane cholesterol levels normalized to values in young murine hearts. Cytosolic cytochrome c levels are expressed as ng/μg protein. CAV-3, caveolin-3; CHOL, cholesterol; AKT, protein kinase B; p70s6K, p70 ribosomal protein S6 kinase; ERK, extracellular signal-regulated kinase 1/2; p38, p38-mitogen activated protein kinase; GRK2, G-protein coupled receptor kinase 2; GSK3β, glycogen synthase kinase 3β; CASP-3, caspase-3; Cyt-C, cytochrome c.

^*P<0.05 vs. Young. Values are reported as means ± SEM; n=6 per group.

Consistent with altered cholesterol and caveolin-3 levels, immunohistochemical and EM assessment revealed significant alterations in sarcolemmal architecture in aged vs. young hearts (Fig. 6). As shown in prior work (Head et al., 2005), in young myocardium caveolin-3 is expressed in a punctate pattern within the sarcolemma and in transverse striations running across the cell interior (Fig. 6A). This defined pattern was not evident in aged tissue, which exhibited reduced total caveolin-3 expression and a reduction in structural membrane invaginations consistent with caveolar depletion (Fig. 6B).

Fig. 6.

Aging alters the membrane, reducing caveolae and caveolin-3 (Cav-3) expression. A) Immunohistochemistry of young and aged heart tissue for caveolin-3. Large inset is higher zoom of smaller inset. Aged hearts show reduced Cav-3. B) Immunoblots confirming reduced Cav-3 protein expression, together with evidence of reduced caveolae formation (assessed by EM) in aged relative to young hearts.

4. Discussion

Given prevalence of heart disease in older individuals, it is critical we understand the influence of age on myocardial responses to injurious insult and protective therapies (and unravel poorly understood mechanisms of cell aging). Myocardial I-R tolerance was markedly reduced in older rodent and human tissue, in agreement with reports of age-related changes in stress-resistance in animal (Headrick, 1998; Headrick et al., 2003; Lesnefsky et al., 2006) and human tissue (Rosenfeldt et al., 1999; Mariani et al., 2000). This is consistent with the view of declining stress-resistance as a hallmark of cellular aging, however the genesis of this decline remains obscure. Reduced myocardial I-R tolerance is manifest as worsened ionic and energetic imbalances (Headrick, 1998), oxidative-stress (Liu et al., 2004; Oudot et al., 2006), apoptotic activation (Azhar et al., 1999), mitophagy dysregulation (Dutta et al., 2012) and mitochondrial dysfunction (Pepe et al., 1999; Lesnefsky et al., 2006; Boengler et al., 2007; Marzetti et al., 2013). Importantly, these processes are all modulated by cytoprotective signaling paths, whose functionality could influence the aging process itself (Gems & McElwee, 2005; Shore et al., 2012). The current study shows that I-R intolerance in aged tissue is associated with repression of hormesis and cytoprotective signaling, with impaired transduction of signals from a modified membrane environment to mitochondrial effectors (Fig. 7). This dysfunctional stress-intolerant phenotype may involve differential shifts in determinants of caveolar control (caveolin-3, membrane cholesterol), GPCR signaling (GRK2), pro-survival signal transduction (p70s6K), and mitochondria-dependent cell death (GSK3β, GRK2, cytosolic cytochrome c). These changes may underpin the negative effects of age on myocardial ischemic injury and responsiveness to therapeutic intervention, and potentially promote myocardial aging.

Fig. 7.

Schematic of myocardial protective signaling, highlighting effects of age (up/down arrows: increased/reduced responsiveness or expression; sideways arrows: no change). Multiple membrane receptors (G-protein coupled and growth factor receptors) are activated by local ligands to initiate pro-survival signaling during stress, involving diverse 2nd messengers and kinase signaling (select kinases are shown, with specific interactions between kinase paths, the precise order of signaling events, and mechanisms coupling kinases to mitochondria yet to be fully elucidated). Age negatively impacts the membrane signal compartment and local aspects of kinase signaling, while mK_ATP channels and the mPTP retain functionality (localizing dysfunction between GPCRs/PKC and p38-MAPK/mitochondria). Impaired stress-resistance and survival signaling may involve shifts in determinants of membrane signaling (caveolin-3, cholesterol), GPCR function (GRK2), kinase signaling (p70s6K, PKC), mitochondrial injury (GSK3β, GRK2) and apoptosis (cytochrome c).

Abbreviations: AKT, protein kinase B; CAV-3, Caveolin-3; Cyt c, cytochrome c; eNOS, endothelial nitric oxide synthase; ERK1/2, extracellular signal-regulated kinase 1/2; GRK2, G-protein coupled receptor kinase 2; GSK3β, glycogen synthase kinase 3β; MEK, MAPK/ERK kinase; mK_ATP, mitochondrial ATP-gated K⁺ channels; mPTP, mitochondrial permeability transition pore; mTOR, mechanistic target of rapamycin; p38, p38-mitogen activated protein kinase; p70s6K, p70 ribosomal protein S6 kinase; PI3K, phosphoinositide 3-kinase; PKC, protein kinase c; PKG, protein kinase G; PLC, phospholipase C; PLD, phospholipase D; RhoA, Ras homolog family member A; RTK, receptor tyrosine kinase.

4.1. Impaired cytoprotective signaling from membrane to mitochondria

Protective interventions to limit I-R injury with myocardial infarction or surgery must be efficacious in older hearts. Unfortunately, aging may induce dysfunction within the pathways targeted by experimental therapies, with evidence of shifts in multiple protective signaling networks (Tani et al., 2001; Headrick et al., 2003; Boengler et al., 2007; Peart et al., 2007; Boengler et al., 2009; Peart & Headrick, 2009). Such changes may contribute to disparities between moderate benefit in clinical trials of conditioning stimuli (in middle-aged/aged patients) vs. profound protection in experimental (typically young) animal models (Peart & Headrick, 2009; Ludman et al., 2010). The age-dependent pattern of cardioprotective efficacies and the molecular profile of older myocardium support impaired initiation and transduction of protective signals from cell membrane to mitochondria (Fig. 7).

Membrane- and GPCR-dependent cardioprotection

While changes in growth-factor/RTK signaling have been implicated in aging and senescence of non-cardiac cells (Cho & Park, 2005; Yu & Driscoll, 2011), the protective functionality of GPCR signaling may also decline in older myocardium (Schulman et al., 2001; Headrick et al., 2003; Peart et al., 2007; Boengler et al., 2009). We show aged hearts are insensitive to agonism of opioid and adenosine receptors mediating cardiac conditioning responses (hormesis) and stress-resistance. This is congruent with impairment of acute IPC in older mouse hearts. Whether initiated by stress and endogenous ligands (IPC) or via exogenously applied agonists, GPCR-dependent cytoprotection is clearly impaired in older tissue. These findings are consistent with age-dependent repression of pre- (Schulman et al., 2001; Przyklenk et al., 2003; Boengler et al., 2007) and post-conditioning responses (Przyklenk et al., 2008) mediated by these GPCR families, and indicate that the anti-aging influences of hormesis (Calabrese et al., 2012) may wane with age.

Insensitivity to IPC and multiple GPCRs supports a broad-spectrum dysfunction that could arise locally, via altered membrane micro-domain control of receptor signaling (Fridolfsson et al., 2012, 2013), and/or distally within convergent signaling pathways (Boengler et al., 2009). The former possibility is supported by reductions in membrane cholesterol and caveolin-3 expression in older hearts, the latter in agreement with the findings of Kawabe et al. (2009) in old rat hearts. Moreover, sarcolemmal ultrastructure is substantially modified with age, with altered localization of caveolin-3 and a paucity of caveolae (Fig. 6). These changes are predicted to impair caveolar control of transmembrane signaling and stress-resistance (Head et al., 2005; Tsutsumi et al., 2008; Fridolfsson et al., 2012, 2013). Caveolae are also important in cell-tension homeostasis and responses to mechanical stress (Sinha et al., 2011), and caveolin-3 represses cardiac myoblast growth (Fujita et al., 2007), and myocardial hypertrophy (Horikawa et al., 2011) that is a feature of the aged phenotype. Age-dependent reductions in caveolin-3 and caveolae may thus not only impact stress-signaling but also alter responses to mechanical loading and facilitate age-dependent hypertrophy. These reductions in myocardial caveolin-3 and caveolae (Figs. 6–7), together with evidence of caveolin-dependent inhibition of neurodegeneration and aging in vivo (Head et al., 2010), and caveolin-dependent senescence in replicating cells in vitro (Cho & Park, 2005), support a potentially overarching role for caveolins/caveolae in governing cellular aging, albeit in a tissue-specific manner.

Survival-kinase dependent protection

Broad-spectrum dysfunction in cytoprotection may also arise within common signaling downstream of membrane receptors. Pro-survival kinases such as PKC, phosphoinositide 3-kinase (PI3K), AKT, MAPKs and protein kinase G transduce signals from GPCRs to targets such as GSK3β, mitochondrial K_ATP channels and the mPTP. PKC is critical in opioid (Peart & Gross, 2006) and adenosine receptor responses (Headrick et al., 2003; Fenton et al., 2010) and IPC (Boengler et al. 2009), and induces mitochondrial protection. However, as for IPC and GPCR agonism, PKC activation fails to limit I-R injury in aged hearts. Age-related changes in PKC signaling have been identified in rat hearts, including altered expression/regulation of PKC isoforms (Takayama et al., 2001) and protection via PMA (Tani et al., 2001). Przyklenk et al. (2003) also report that age modifies the PKC-dependence of conditioning stimuli in rabbits. Age-dependent dysfunction in cardiac PKC signaling is thus conserved across species, and likely participates in impaired protective signal transduction.

Cardiac 38-MAPK signaling may also contribute to protection via opioid and adenosine GPCRs (Peart & Gross, 2006; Fenton et al., 2010), limiting mitochondrial damage and apoptosis (Whittaker et al., 2009). Beneficial effects of pre-ischemic p38-MAPK activation in both young and older murine and human myocardium agrees with prior findings (Hassouna et al., 2006; Peart et al., 2007), and localizes age-dependent signal dysfunction upstream of p38-MAPK. These observations also support acute MAPK manipulation as a potential protective strategy in aged hearts.

Mitochondria-dependent protection

Preservation of mitochondrial function is central to cellular stress-resistance, while dysfunction is considered important in cell aging (Marzetti et al., 2013). Preconditioning, protective GPCRs, and PKC and p38-MAPK activities converge on mK_ATP- and mPTP-dependent control of mitochondrial function. Beneficial effects of diazoxide, which activates mK_ATP channels to increase mitochondrial reactive oxygen species (ROS) signaling and inhibit the mPTP (Pain et al., 2000; Facundo et al., 2007), supports preserved functionality of mK_ATP channels in aging human and murine myocardium. Moreover, since PKC activation is ineffective in aged hearts, the protective actions of mK_ATP opening do not appear to require downstream PKC activation in older tissue. The influence of mPTP inhibition on cell survival was also preserved in aged hearts. A number of studies confirm protection via mPTP inhibition in young tissue (Halestrap et al., 1997; Liu et al., 2011), and cyclosporin A protects human myocardium in vitro (Schneider et al., 2003) and in vivo (Piot et al., 2008). However, Liu et al. recently reported cyclosporin A is ineffective in aged rat hearts (Liu et al., 2011), and its effects may also be reduced by obesity or diabetes (Huhn et al., 2010). Interestingly, although mPTP inhibition protects aged myocardium, ischemic tolerance remains profoundly depressed in older vs. younger hearts. This could reflect increased activity and altered control of the mPTP (Zhu et al., 2010). Since cyclosporin A limits mPTP opening by reducing cyclophilin D binding (Halestrap et al., 1997), increased cyclophilin D expression together with resistance of the interaction to inhibition with age (Zhu et al., 2013), could render the mPTP resistant to modulation. While preliminary data suggests a 10-fold higher concentration of cyclosporin A does not exert greater effects in aged tissue, this possibility warrants further study.

Preservation of protective responses to p38-MAPK and mitochondrial modulators despite failed protection via IPC, GPCRs and PKC activation, supports dysfunction in protective signal transduction from membrane GPCRs (and PKC) to mitochondrial effectors (Fig. 7). Whether this change involves a profound decline in cellular sensitivity to hormesis stimuli, GPCR agonism and PKC activation vs. complete abrogation of responses is unclear (though data indicates a 2.5-fold elevation in PMA does not induce protection in aged hearts). In either case, the phenotypic outcome is insensitivity to protective stimuli that near maximally enhance stress-resistance in younger tissue. From a clinical perspective, reduced sensitivities to protective interventions might be overcome by strategic combinations of stimuli, whereas unresponsive molecular targets might be by-passed in favor of downstream effectors. From a fundamental perspective, these major changes in the functionality of membrane-to-mitochondrial signaling - an evolutionarily conserved feature of cellular stress-adaptation (Fridolfsson et al., 2012) - could reflect an important mechanistic component of myocardial aging, given evidence the expression and function of cytoprotective signaling paths may govern aging and longevity (Gems and McElwee, 2005; Shore et al., 2012).

4.2. Emergence of a stress-intolerant molecular profile

Whether alterations in signaling and stress-resistance reflect stochastic and/or active (eg. via mTOR or IGF pathways) changes in myocardial protein expression is unclear. We show that additional to repression of caveolar components, aged hearts express increased levels of proteins favoring mitochondrial dysfunction and cell death (Table 3, Fig. 7). These differential changes are consistent with evolving stress-intolerance, and hint at active control (though this requires further study). Increased GRK2 expression may inhibit GPCR function (Jurado-Pueyo et al., 2008) and promote mPTP opening and cell death (Brinks et al., 2010; Chen et al., 2013). Elevated GSK3β levels are predicted to enhance mPTP activity and effects of pro-apoptotic proteins (Juhaszova et al., 2009; Zhu et al., 2013), and are in accord with observations in male and female Fisher 344 rats (Kostyak et al., 2006; Hunter et al., 2007). On the other hand Zhu et al. (2010, 2013) found no age-dependent shift in GSK3β in the same model. Increased cytosolic cytochrome c levels may also sensitize tissue to apoptotic death. The reduced expression of p70s6K may further undermine stress-resistance and signaling, based on involvement in cardioprotection via GPCRs (Gross et al., 2004; Förster et al., 2006) and pre- and post-conditioning (Tsang et al., 2004; Zhu et al., 2006). Since p70s6K phospho-inhibits GSK3β (Sutherland et al., 1993), declining expression will amplify detrimental effects of elevated GSK3β in aged tissue. These changes are consistent with data for middle-aged hearts (Peart et al., 2012), and demonstrate that age induces an expression profile favoring signal dysfunction, mitochondrial injury and cell death. The basis of the differential changes in pro-survival (caveolin-3, p70s6K) vs. pro-injury proteins (GSK3β, GRK2) remains to be established.

5. Conclusions

This study confirms age-dependent repression of I-R tolerance in human and rodent myocardium, and reveals these changes are associated with impaired hormesis and dysfunctional transduction of cytoprotective signals from membrane GPCRs and PKC to mitochondrial targets (Fig. 7). Impaired membrane micro-domain orchestration of survival-signaling (Fridolfsson et al., 2012; Fridolfsson & Patel, 2013), together with shifts in determinants of signal transduction and mitochondrial preservation may collectively limit resistance and adaptation to damaging stressors (which may, in turn, promote the cardiac aging process). Further studies are warranted to evolve our understanding of mechanisms of cellular- and organ-specific aging, to assess the relevance of age-related myocardial changes reported here, and to rationally evolve strategies to promote resistance to injury in the ‘at-risk’ aged heart.

6. Limitations

A number of study limitations deserve attention. First, while we do not directly compare distinct murine and human tissue models (as our aim was to test whether age commonly impairs stress-responses and protective signaling in different species), it is relevant to highlight that we assessed whole myocardial responses in mice, and isolated atrial muscle strip responses in humans. As discussed previously (Mariani et al., 2000), structural and functional differences may limit the applicability of atrial tissue as a surrogate for ventricular myocardium. However, human ventricular samples are rare to access, with biopsies generally too small to assess functionally and problematic to study as discrete muscle bundles. In contrast, atrial tissue is more readily available and amenable to functional assessment, with the isolated trabeculae model validated and applied in pharmacological and functional analyses (Guo et al., 1983; Keon et al., 1991; Labow et al., 1991), and in the interrogation of ischemic or hypoxic responses (Rosenfeldt et al., 1999; Mariani et al., 2000). Even so, availability of human (particularly younger) tissue precluded the more extensive analysis of protective responses and molecular changes achievable in murine hearts (limiting stimuli assessed in human tissue to a GPCR agonist, and p38-MAPK and mK_ATP channel activation). Sensitivity of the LDH assay also prevented accurate assessment of cell death in this preparation. What is nonetheless remarkable is that despite differing species, tissues and experimental models, data from human and murine analyses remain internally consistent, demonstrating that aging profoundly impairs intrinsic I-R resistance and GPCR-dependent protection, without negating responses to distal p38-MAPK or mK_ATP channel activity.

Additional to examining aspects of cardiac aging, a goal of this work was to contrast the functionality of different molecular determinants of I-R tolerance (thus identifying potential cardioprotective ‘targets’) in aged vs. young tissue. Informed by the changes in membrane-dependent survival signaling and regulatory proteins observed here, future studies are warranted to unravel the basis of these alterations, and rationally refine experimental strategies to bolster I-R tolerance in the aged heart. Since proteins expressed in greater or lower levels in aged hearts are post-translationally modified, additional analysis of this control is also needed. Significant changes in phosphorylation, for example, will either exaggerate or limit effects of altered protein expression. In this regard, increased phosphatase activity detected in aged hearts (Fenton et al., 2005) is likely to exaggerate the functional impacts of reduced p70s6K and increased GSK3β expression documented here.

Highlights.

• Aging repressed ischemic tolerance in human atrial tissue and mouse hearts

• Aging reduced cardioprotection via preconditioning (hormesis) and GPCR/PKC activation

• Protection via modulation of key mitochondrial effectors of protection was preserved

• Survival signal transduction from receptors to mitochondria is dysfunctional, in association with membrane/caveolar changes

• Protein expression in aged mouse heart further favors mitochondrial injury/cell death