The “Heartbreak” of Older Age
- Laboratory of Cardiovascular Science Gerontology Research Center Intramural Research Program National Institute on Aging National Institutes of Health, Baltimore, MD 21224
- Address correspondence to EGL. E-mail lakattae{at}grc.nia.nih.gov; fax (410) 558-8150.
Abstract
It is estimated that by 2035, nearly one in four individuals in the United States will be sixty-five years of age or older. Hypertension, atherosclerosis, and resultant chronic heart failure reach epidemic proportions among older persons, and the clinical manifestations and the prognoses of these worsen with increasing age. The reason is that, in older individuals, specific pathophysiological mechanisms that underlie these diseases become superimposed on heart and vascular substrates that are modified by the process of aging. In other words, cardiovascular aging is “risky.” An understanding of how age, per se, modifies cardiovascular structure and function is critical to the prevention or treatment of cardiovascular diseases in the older person.
An Integrated View of Age-Associated Changes in the Human Heart in the Absence of Clinical Disease
Heart Structure and Function at Rest
A unified interpretation of the cardiac changes that accompany advancing age in otherwise healthy persons suggests that, at least in part, these are adaptive, occurring to some extent in response to arterial changes that occur with aging (Figure 1⇓) (1) . Aging leads to arterial stiffening that results in enhanced pulse-wave velocity and early reflected pulse-waves. These produce a late augmentation in arterial systolic pressure, a reduced or maintained diastolic pressure in the presence of a mild increase in peripheral vascular resistance (PVR), and an increase pulse pressure, accompanied by aortic dilatation and wall thickening. Increases in the thickness of the left ventricle (LV) wall, largely due to an increase in ventricular myocyte size and increased vascular impedance, moderate the increase in LV wall tension. Modest increases in collagen levels also occur with aging.
Arterial and cardiac changes that occur with aging in healthy humans. One interpretation (flow of arrows) of the constellation is that vascular changes lead to cardiac structural and functional alterations that maintain cardiac function. L, left; LV, left ventricular. From (1).
Prolonged contraction of the thickened LV wall maintains a normal ejection time in the presence of the late augmentation of aortic impedance. This preserves the systolic cardiac pumping function at rest. One disadvantage of prolonged contraction is that, at the time of the mitral valve opening, myocardial relaxation is relatively more incomplete in older than in younger individuals and causes the early LV filling rate to be reduced in older individuals. Structural changes and functional heterogeneity occurring within the LV with aging may also contribute to this reduction in peak LV filling rate. However, concomitant adaptations—left atrial enlargement and an enhanced atrial contribution to ventricular filling—compensate for the reduced early filling and prevent a reduction of the end diastolic volume (EDV). Potential age-associated changes in the tissue levels or responses [to growth factors, catecholamines, angiotensin II, endothelin, tumor growth factor β (TGFβ ), or fibroblast growth factor (FGF)] that influence myocardial or vascular cells or their extracellular matrices may also have a role in the schema depicted in Figure 1⇑.
Cardiovascular Reserve
Heart-rate acceleration and impaired augmentation of blood ejection from the LV, accompanied by an acute modest increase in LV EDV, are the most dramatic changes in cardiac reserve capacity that occur with aging in healthy, community-dwelling persons (Table 1⇓). Mechanisms that underlie the age-associated reduction in maximum ejection fraction are multifactorial and include: 1) a reduction in intrinsic myocardial contractility, 2) an increase in vascular afterload, 3) a diminished effectiveness of the autonomic modulation of both LV contractility and arterial afterload, and 4) arterial-ventricular load mismatching. (Ventricular load is the opposition to myocardial contraction and the ejection of blood; afterload is the component of load that pertains to the time following excitation, as opposed to preload, prior to excitation.) Although these age-associated changes in cardiovascular reserve, per se, are insufficient to produce clinical heart failure, they do affect its clinical presentation—that is, the threshold for symptoms and signs, and the severity and prognosis of heart failure secondary to any level of disease burden (e.g., chronic hypertension that causes either systolic or diastolic heart failure).
Exhaustive Upright Exercise Changes In Aerobic Capacity And Cardiac Regulation Between Ages of 20 And 80 Years In Healthy Men and Women
Cardiac Aging In Animal Models
Cardiovascular Structure
Cellular and molecular mechanisms that are implicated in age-associated changes in myocardial structure and function in humans have been studied largely in rodents (Table 2⇓). The altered cardiac structure that evolves with aging in rodents includes an increase in LV mass, due to myocyte enlargement (2) , and proliferation of the matrix where the myocytes reside, which may be linked to alteration of cardiac fibroblast number or function. The number of cardiac myocytes becomes reduced, owing to necrosis predominately, or apoptosis (3) . Stimuli thought to cause cardiac cell enlargement in aging rodents include an age-associated increase in vascular load (due to arterial stiffening) and stretching of cells caused by the death and loss of neighboring myocytes (4) . Stretching of cardiac myocytes and fibroblasts initiates growth-factor-dependent signaling (by angiotensin II or TGFβ ), and, in some cases, apoptosis (5) .
Myocardial Changes With Adult Aging In Rodents
Cell Excitation, Ca2+Regulation, and Contraction
Coordinated changes in the function or expression of proteins that regulate several key steps of cardiac cell excitation-contraction coupling occur in the rodent heart with aging and result in a prolonged action potential (AP), a prolonged cytosolic calcium (Cai ) transient following excitation, and a prolonged contraction (Table 2⇑; Figure 2⇓) (6 –8) . Both the apparent number and the activity of individual cardiac L-type Ca2+ channels increase with age in the rat (9) . The total L-type current inactivates more slowly in myocytes from older versus younger rats (9 , 10) , and this, as well as reductions in outwardly directed K+ currents (10) , might partially account for the prolonged AP of the former. Shorter APs in myocytes from old rat hearts reduce the Cai transient amplitude. This is attributable to a reduction in sarcoplasmic reticulum (SR) Ca2+ load (sequestered Ca2+ ), which may arise from reduced Ca2+ influx via L-type Ca2+ channels (ICaL ) and increased Ca2+ extrusion by the sarcolemmal Na+ -Ca2+ exchanger (Figure 3⇓) (11) .
Differences in various aspects of excitation-contraction coupling mechanisms measured between adult (6–9 months) and senescent (24–26 months) rat hearts. A. Transmembrane action potential, and (B) isometric contraction [from (6) ]. C. Cytosolic calcium transient, measured by a change in the luminescence of aequorin injected into several cells comprising the muscle preparation [from (7) ]. D. Sarcoplasmic reticulum Ca2+ uptake rate [from (8) ].
Schematic representation of the role of age-dependent action potential (AP) prolongation on intracellular Ca2+(Ca2+i) regulation in rat ventricular myocytes. Compared to young rats (4–6 months, left) old rats (23–24 months, middle) display a similar amplitude of the Ca2+i transient triggered by a markedly prolonged AP. Relative to stimulation with their long, native AP, stimulation of old rat ventricular myocytes with a short, “young type” AP (right) increases the amplitude of L-type Ca2+ current (ICa ) but reduces its time integral and diminishes the amplitude and rate of relaxation of the Ca2+i transient. Thus, it appears that ventricular myocytes of old rats utilize AP prolongation to sustain youthful Ca2+i regulation. From (11) .
The prolonged Cai that occurs with aging is explained, in part, by a reduction in the rate of Ca2+ sequestration by the SR (Figure 2D⇑). This is attributable to a reduced transcription of the gene coding for the sarcoplasmic reticulum Ca2+ pump, Serca2 (1) . The abundance of transcripts of the cardiac Na+ -Ca2+ exchanger (NCX-1), which serves as the main trans-sarcolemmal Ca2+ extrusion mechanism, is increased nearly fifty percent in senescent (twenty-four month) compared to the young adult (six month) rat hearts (12) . Age-associated prolonged Ca2+ transient and heart contraction may impair myocardial relaxation during early diastole and may underlie the reduction in early diastolic filling rate that accompanies advancing age.
The expression of cardiac myosin heavy chain (MHC) isoforms changes over time in the senescent rat heart with the β isoform becoming predominant; myosin Ca2+ -ATPase activity declines concomitant with the decline in α MHC content (11) . MHC gene expression can be regulated through binding of the thyroid hormone receptor (THR)—a member of a superfamily of receptors that includes retinoic acid receptors (RARs), vitamin D receptors, and retinoid X receptors (RXRs)—to thyroid response elements in gene 5’ flanking regions. A significant reduction (approximately fifty percent) in the amounts of THRβ and RXRγ mRNA transcripts and proteins occurring between six and twenty-four months has been observed in hearts of senescent rats (13) .
The altered composite cellular phenotype (Table 2⇑), consisting of contractions with reduced velocity and a prolonged time-course, stems from the elevation of cytosolic [Ca2+ ] that follows a prolonged action potential (AP), is adaptive rather than degenerative. This is so because myocardial shortening at reduced velocity is energy efficient. This altered pattern of Ca2+ regulation and myosin protein expression allows the myocardium of older hearts to generate force and active stiffness for a longer time following excitation (13) . The prolonged isovolumic relaxation period in the healthy aging human heart (14) may be partly attributable to prolonged Ca2+ -dependent contractile protein activation, enabling the continued ejection of blood during late systole, a beneficial adaptation with respect to enhanced central arterial stiffness and early reflected pulse-waves (Figure 1⇑) (1) .
Reduced Acute Cellular Response To Stress
The acute reserve function of cardiac myocytes may be recruited through enhanced Ca2+ influx into that cell following excitation. An acute excess of myocardial Ca2+ loading leads to dysregulation of Ca2+ homeostasis, impaired diastolic and systolic function, arrhythmias, and cell death (15). The Ca2+ load in each cell is determined by membrane structure and permeability characteristics, the intensity of stimuli that modulate Ca2+ influx or efflux, and the presence of reactive oxygen species (ROS), which affect both membrane structure and function. Excessive cytosolic Ca2+ loading occurs during physiologic and pharmacologic scenarios that increase Ca2+ influx [for example, rapid beating (tachycardia), neurotransmitters, post-ischemic reperfusion, or oxidative stress] (16 , 17).
Cellular β-Adrenergic Signaling
Deficits in myocardial β -adrenergic receptor (β AR) signaling occur with aging (1) : reduced myocardial contractile response to either β1 AR and β2 AR stimulation is observed (18 –20) . This is due to inability of β AR to augment the Cai to the same extent in cells of senescent hearts as compared to cells from younger adult hearts (18) , and can be attributed to reduced numbers of L-type sarcolemmal Ca2+ channels, leading to a smaller increase in Ca2+ influx in “senescent” hearts (Figure 4⇓) (18 –21) . The decreased postsynaptic responses of aged myocardial cells stimulated with β -adrenergic agonists may arise from multiple changes in the molecular and biochemical steps that couple the β AR to downstream effectors. However, the defect in this signaling pathway in rodents of advanced age appears to be at the point of Gs -mediated coupling of the β -adrenergic receptor to adenylyl cyclase, and to changes in adenylyl cyclase. The defect leads to deficient cAMP production and to impaired activation of protein kinase A, which normally activates key proteins that augment cardiac contractility (19) . The desensitization of β AR signaling in advanced-age rodents does not appear to be mediated by increased β -adrenergic receptor kinase (β ARK) or increased Gi activity (19) . Under conditions of high-intensity β -adrenergic stimulation, isolated hearts, cardiac muscle, and myocytes exhibit signs of Ca2+ overload (see below). The older heart exhibits a reduced threshold for this catecholamine-dependent cardiotoxicity (21 , 22)
The efficacy of β adrenergic receptor stimulation withers with advancing age. A. The effect of norepinephrine on the maximum rate of isometric tension development in isolated trabeculae from hearts of varying age. From (21). B. Velocity of cell shortening, and C. the maximum rate of increases of the Indo-1 fluorescence transient, an index of sarcoplasmic reticulum Ca2+ release into the cytosol, during the electrically stimulated twitch in single cardiac myocytes isolated from the hearts of rats of varying ages and loaded with the fluorescent probe, Indo-1. From (18). D. Norepinephrine increases the L-type sarcolemmal channel current (ICa ). across a range of activating steps to different membrane potentials in single vascular cells potential decline with aging, the norepinephine concentration was 1 x 10-7 M. From (18) . E. Treatment with norepinephrine (NE) increases phosphorylation of troponin I (TNI) in suspensions of heart cells isolated from hearts of rats of varying age. From (20).
Tachycardia
In myocytes from older hearts, enhanced Ca2+ influx, impaired relaxation, and increased diastolic tone occur during tachycardia (Figure 5⇓) (23 –26) . Studies in rodents have demonstrated that impaired relaxation and impaired Ca2+ sequestration that occur with aging can be reduced by increased expression of Serca2 (Figure 5⇓) (23 , 24 , 27 , 28) .
Dependence of LV diastolic pressure (LVDP) in adult hearts (• , n=5), senescent hearts (▪ , n=4), and senescent hearts infected with Ad.SERCA2a ( ▴ , n=6). From (28).
During graded increases of [Ca2+ ] in the fluid perfusing the heart there is a reduced threshold for the occurrence of manifestations of Ca2+ overload in senescent (twenty-four-month old) versus younger adult (six-to-eight-month old) rat hearts (Figure 6⇓) (16) . Young and old hearts exhibit little difference in developed pressure (DP), end diastolic pressure (EDP), or half relaxation time (RT1/2 ) when bathed in 1.5 mM Ca2+ and given gentle electrical stimulation (Figure 6A, B⇓), and these hearts exhibit no aftercontractions—that is, no spontaneous contractions that follow the main contraction in response to the action potential—as exemplified in Figure 6C⇓. Initially, increases in [Cao ] result in a strengthened heart beat due to increased Ca2+ influx through L-type Ca2+ channels and to reduced Ca2+ efflux through Na+ -Ca2+ exchange. But, as [Cao ] is progressively increased, the DP response becomes biphasic: increasing, plateauing, and then decreasing. The maximum increase in DP with increasing [Cao ] is less in twenty-four-month-old than in six-to-eight-month-old hearts (Figure 6A⇓). Additionally, in twenty-four-month-old hearts, the plateau and decrease in DP occur at a lower [Cao ] than in younger hearts. Thus, the [Cao ] response–systolic function curve is significantly shifted leftward, to lower [Ca] in old versus young hearts.
Reduced Ca2+tolerance of senescent hearts. A. Developed pressure (DP) and end-diastolic pressure (EDP) in response to increasing perfusate Ca2+ concentration {[Cao ]} in young and old hearts during stimulation at 2.0 Hz. Responses of both DP and EDP significantly differ with age by repeated-measures analysis of variance. B. Half-relaxation time in young and old hearts in response to an increase in [Cao ] at 2.0-Hz stimulation. Response differs with age. C. Chart recording of left ventricular pressure during and after cessation of electrical stimulation. Note the occurrence after cessation of stimulation in 8 mM [Cao ] of a “hyperrelaxation” and aftercontractions (arrows), followed by a slow decay in diastolic pressure, which eventually reaches the RP level (not shown) D. Aftercontractions occurrence after cessation of electrical stimulation at 2.0 Hz in young and old hearts as a function of [Cao ]. E. Spontaneous ventricular fibrillation (VF) occurrence during pacing at 2.0 Hz as a function of [Cao ]. F. Spontaneous contractile waves generated by spontaneous Ca2+ oscillations in single isolate ventricular myocytes from senescent hearts occur more frequently than in those from young adult hearts as [Cao ] is increased. From (16).
Diastolic function also becomes impaired in older hearts during scenarios that lead to enhanced Ca2+ increase into the cell. RT1/2 decreases in young hearts as [Cao ] increases, but in senescent hearts, it does not change or increases (Figure 6B⇑). EDP in young hearts decreases slightly as [Cao ] increases from 1.5 to 3 mM, and then gradually increases only back to the initial value as [Cao ] is increased further (Figure 6A⇑). However, in twenty-four-month-old hearts, the lesser absolute increase of DP to increasing [Cao ] (Figure 6A⇑) is accompanied by a greater elevation in EDP compared to six-to-eight-month-old hearts.
The occurrence of aftercontractions (Figure 6C⇑, arrows )—driven by spontaneous SR Ca2+ release, another sign of Ca2+ overload (29) —increases as [Cao ] increased (Figure 6D⇑). In twenty-four-month-old hearts, the likelihood of aftercontractions is significantly greater than in younger hearts (Figure 6D⇑). Spontaneous Ca2+ oscillations are manifested as spontaneous Ca2+ or contractile waves (Figure 6F⇑), and the asynchronous force oscillations driven by spontaneous Ca2+ oscillations occur with increasing frequency as [Cao ] rises. Thus, asynchronously occurring cytosolic Ca2+ oscillations become increasingly fused (15) . This summation phenomenon also produces the aftercontraction (Figure 6C⇑) and its precursor, the “hyperrelaxation” of diastolic force that follows the main contribution caused by the action potential (Figure 6C⇑). This results in a resting pressure (RP) that, in the absence of beating, is greater than EDP or the pressure between beats during stimulation. Spontaneous Ca2+ release activates the Na+ -Ca2+ exchanger, leading to spontaneous depolarizations and arrhythmias (15) . Figure 6E⇑ shows that Ca2+ -dependent spontaneous ventricular fibrillation occurs in older but not younger hearts. This trilogy of signs caused by cell Ca2+ overload—reduced systolic performance, measured as developed pressure (DP), increased EDP, and increased likelihood for arrhythmia—is a final common scenario for diverse pathologies that cause heart failure (30) . In fact, the experimental paradigm (Figure 6⇑) leads to acute diastolic and systolic heart failure.
The reduced threshold for Ca2+ overload is a disadvantageous outcome of the aforementioned age-associated adaptation that occurs in the cells of senescent heart (and also in young animals chronically exposed to arterial pressure overload). Causes of reduced Ca2+ tolerance of the older heart include 1) changes in the amounts of proteins that regulate Ca2+ flux, in part, due to altered gene expression (Table 2⇑), 2) to an age-associated alteration in the composition of membranes in which Ca2+ regulatory proteins reside, which includes an increase in the ratio of ω6 :ω3 polyunsaturated fatty acids (PUFAs) in membranes (31) , or 3) to the increased production of ROS following stress.
Changes in Membrane Lipid Composition with Aging
Ca2+ influx and efflux among intracellular organelles and between intracellular Ca2+ storage sites and the cytosol are governed by the function of proteins within the sarcolemma or organelle membranes; the function of these proteins depends upon their conformation and position within these membranes, which, in turn, is determined by the lipid milieu within the membrane. An increase in the content of arachidonic acid (a ω6 PUFA) and a reduction in docosahexaenoic acid (a ω3 PUFA) membranes derived from total heart homogenates or from isolated mitochrondria occur with aging (31 , 32) . The age-associated shift in membrane PUFAs can be reversed or exaggerated by dietary interventions (Table 3⇓).
Pufa And Phospholipid Content In Mitochondrial Membranes From Young And Old Rat Hearts
Studies indicate that the composition of PUFAs in membranes affects the ion permeability of those membranes (33) . Membranes enriched in ω3 or ω6 PUFAs are associated, respectively, with a reduced or enhanced susceptibility to Ca2+ overload. The addition of ω3 PUFAs to isolated ventricular myocytes in vitro protects against Ca2+ overload (34) . Mechanisms of the later effect may be related to the effects of ω3 PUFAs on ion-channel properties (35 , 36) . The decline in the content of cardiac-cell-membrane ω3 PUFAs with aging may result in increased vulnerability to Ca2+ overload and other Ca2+ -dependent defects. Indeed, the protective effects of ω3 PUFAs against myocardial Ca2+ overload and fatal arrhythmia following ischemia have been demonstrated both in old animals and in younger animals with diets containing high amounts of saturated fat (32 , 37 –40) (Figure 7⇓).
Influence of age and dietary fat on the duration of ventricular fibrillation (VF) during reperfusion. The influence of dietary fat on myocardial vulnerability to arrhythmia was examined using coronary artery occlusion and reperfusion in the anesthetized rat as a whole animal model of ventricular fibrillation (VF) and sudden cardiac death. Animals were fed a reference (REF) diet alone or supplemented 12% by weight with tuna fish oil (TFO) (rich in ω3 PUFAs), sunflower seed oil (SSO) (rich in ω6 PUFAs), or sheep perirenal fat (SF) (rich in saturated fatty acids). Feeding periods of 6, 12, and 18 months and a total of 108 rats were used. The incidence of ventricular fibrillation in occlusion was reduced from 46% of REF animals to 6% and 21% in TFO and SSO groups, respectively, and increased to 68% in the SF-fed rats. The incidence of ventricular tachycardia was also reduced by TFO and SSO. The duration of arrhythmic episodes was increased by SF and reduced by TFO and SSO. The incidence of fibrillation on reperfusion of acutely ischemic myocardium (15 min. occlusion) was significantly reduced by TFO only (12%, REF = 50%, SSO = 30%, SF = 70%). Severity of arrhythmias increased with age as did the extent of dietary influence. Mortality from fibrillation, which only occurred in rats aged twelve months or older (REF = 13%), was increased by SF (43%) mainly in reperfusion (38%) but did not occur in TFO or SSO. These results indicate the potential benefit of dietary modification to include a higher proportion of polyunsaturated fat, especially fish oil, in reducing risk of sudden cardiac death. From (32).
Increased ROS Generation in the Aged Heart
A third cause for manifestation of a reduced threshold for Ca2+ overload in the old heart is increased intracellular generation of ROS. Mitochondrial production of ROS in the heart, and mitochondrial dysfunction associated with reperfusion, appear to increase with age (Figure 8A⇓) (31 , 41 –45) . Furthermore, several studies indicate that reperfusion that follows ischemia generates ROS and that exogenous application of ROS to cells causes Ca2+ overload, ATP depletion, rigor (i.e., cellular stiffening caused by interlocking of actin and myosin), and occurrence of mitochondrial membrane-permeability transition (MPT) (17) . The MPT involves Ca2+ -dependent opening of the permeability transition pore (PTP), a nonselective, high-conductance channel in the mitochondrial inner membrane that causes the loss of the proton motive force in the respiratory chain and the failure to generation ATP. ROS generated within mitochondria play a role in causing MPT. The release of mitochondrial contents into the cytosol—resulting from the MPT induction—can lead to high [Cai ] and to the generation of additional ROS, whereby necrosis and/or apoptosis are initiated and cell death ensues.
Experimental assessment of membrane-permeablity transistion in individual in situ mitochondria by photoexcited ROS production. A. Age-associated differences in the effect of ischemia and reperfusion on ADP-dependent mitochodrial respiration. Mitochondria were isolated and respiratory activities were assessed; glutamate supported State 3 respiration after perfusion (P), ischemia (I), or reperfusion (R) of hearts from 8- and 24-month-old rats. From (45) . B . Confocal linescan imaging of fluorescence in tetramethylrhodamine methyl ester (TMRM) loaded single rat cardiac myocyte, at 2 Hz along a 25 μ m longitudinal segment encompassing approximately twenty mitochondria; time progresses from left to right. TMRM fluorescence serves as the index of mitochondrial membrane potential, ΔΨ . The apparent dark regions between horizontal columns are extra mitochondrial spaces between individual mitochondria (approximately half-sarcomere spacing). The sudden dissipation of TMRM fluorescence (i.e., white-to-black transitions at certain points in time) in individual mitochondria indicates the loss of ΔΨ due to MPT induction. A representative individual mitochondrion is delimited showing its MPT occurrence after approximately 90 sec of photoexcitation. C. Mitochondria from cardiac myocytes of aged (24-month old) rats are more susceptible to ROS-induction of the MPT vs. those from young (3-month old). Cells were examined as in Panel A , and represent the average from 8–10 cells in each group. D. SDS/gel electrophoresis and Western blot analysis of protein from cardiac mitochondria in the experiment in Figure 8A. Only mitochondria from 24-month-old reperfused hearts exhibited protein modified by hydroxynoneal (HNE). Thus, HNE-induced change of mitochondrial enzymes might underlie the exaggerated age-associated decline in mitochondrial state 3 respiration during reperfusion observed in Panel A . From (45) . HNE was detected by HPLC assay in cardiac mitochondria isolated from 6- and 24-month-old rats treated to either ω3 PUFA or ω6 PUFA rich diets for 6 weeks. Hearts (n=8) were subjected to 15-min low-flow global ischemia and reperfused for 30 min prior to isolation of mitochondria. HNE was detected in mitochondria isolated from normoxic controls at a concentration less than 1 pmol/mg mitochondrial protein. E. HNE was detected by HPLC assay in cardiac mitochondria isolated from 6- and 24-month-old rats treated to either ω3 PUFA or ω6 PUFA-rich diets for 6 weeks. Hearts (n=8) were subjected to fifteen minutes of low-flow global ishcemia and reperfused for thirty minutes prior to isolation of mitochondria. HNE was detected in mitochondria isolated from normoxic controls at a concentration less than 1 pmol/mg mitochondrial protein.
Technical limitations had precluded a more precise examination of the local production of ROS in earlier studies and thus could not establish a direct link from ROS to local cellular effects within living, intact cardiac cells. However, newer techniques utilizing chemiluminescence and sequential imaging have revealed that real-time ROS production in liver, during ischemic reperfusion, increases with age (46) . Additionally, more precise methods have been developed that label mitochondrial membranes with fluorescent molecules that generate ROS when excited by visible light laser. The inclusion of a probe that is oxidized by the local production of ROS (e.g., dichlorofluorescein) permits detection of ROS produced by such photoexcitation. Labeling of mitochondria with a membrane-potential-sensitive fluorescent probe permits simultaneous monitoring of ROS generation and mitochondrial membrane potential within cells. Recent studies employing these techniques have discovered a phenomenon referred to as “ROS-induced ROS release” (47) . In other words, the initial or “trigger” ROS generated by the laser excitation of the fluorescent ROS source leads to an amplification of ROS production by the mitochondria themselves. This amplification is associated with induction of the MPT and collapse of the mitochondrial membrane-potential (Figure 8B⇑), which renders the mitochondrial membrane permeable to bulk movement of ions, larger molecules, and water between the cytosol and the mitochondrial matrix. Cells from senescent hearts have a substantially lower threshold for the generation of ROS-induced ROS release and less likelihood of the MPT (Figure 8C⇑).
ROS, such as the hydroxyl radical (·OH), are highly reactive and generally short-lived species. Therefore, they might be expected to cause damage at or near the site of their formation. Membrane PUFAs undergo lipid peroxidation by ROS, producing various aldehydes, alkenals, and hydroxyalkenals, such as malonaldehyde and 4-hydroxy-2-nonenal (HNE) (48) . HNE, potentially the most reactive of these compounds, is formed by superoxide reactions with membrane ω6 PUFAs and reacts with protein sulfhydryl groups to induce altered protein conformation (45 , 48 , 49) . In contrast to many ROS species, HNE is rather long-lived and can therefore diffuse from the site of its origin in membranes to affect potential targets distant from the initial site of ROS production (48) . Thus, if membrane bilayer PUFAs are converted to lipid hydroperoxides, then lipid peroxidation, per se, may be viewed as an “amplifier” for the initial ROS. Furthermore, the reactive aldehydes generated in this process may well act as “toxic second messengers” of the complex chain reactions that follow ROS production (48) .
Because the mitochondrial respiratory chain represents a major subcellular source of ROS during reperfusion of ischemic myocardium (45) , and because ischemic-reperfusion has been associated with decreased rates of NADH-linked, ADP-dependent respiration (50 , 51) , mitochondrial membranes are likely targets of lipid peroxidation and HNE-mediated enzymatic dysfunction. The concentration of HNE is increased in the reperfused post-ischemic myocardium (49) .
HNE modification of mitochondrial proteins occurs exclusively in hearts isolated from senescent rats (Figure 8E⇑), but it has not been established whether ROS-induced peroxidation of mitochondrial membranes leading to the production of HNE (Figure 8D⇑) is related to the amplification of ROS and “mitochondrial catastrophe.” Recent studies show that exposure of intact cardiac mitochondria to HNE leads to decreased NADH-linked respiration (52) partly due to HNE-dependent inactivation of α -ketoglutarate dehydrogenase and pyruvate dehydrogenase. These enzymes contain lipoic acid residues that are prime targets for HNE (52) . Recent evidence indicates age-dependent inactivation of α -ketoglutarate dehydrogenase during reperfusion (53) .
PUFAs differ with respect to their susceptibility to lipid peroxidation. Indeed, ω6 rather than ω3 PUFA appear to be preferred targets of ROS-induced peroxidation that produces HNE (Figure 8E⇑). Thus, the increase in ω6 :ω3 PUFA that occurs with aging may be a mechanism for increased HNE production following ischemia/reperfusion (I/R), but this effect can be markedly attenuated by an ω3 PUFA-rich diet (Figure 8E⇑). Notably, an ω3 PUFA-rich diet also prevents the age-linked decrement of the mitochondria-specific membrane phospholipid cardiolipin (31) , a crucial cofactor for cytochrome c oxidase and adenine nucleotide translocase (ANT) activity (54) . ANT may participate in the formation of a nonspecific membrane pore through a Ca2+ -mediated, cyclophilin-D–dependent conformational change in ANT (55) . Cyclophilin-D binding to ANT increases following oxidative stress that enhances the Ca2+ -dependent formation of the PTP, which induces MPT. Adenine nucleotides located in the mitochondrial matrix bind to ANT and decrease the sensitivity of the PTP to Ca2+ , but this effect is antagonized by modification of specific thiol groups on ANT—by oxidative stress products such as HNE or by thiol reagents such as carboxyatractyloside—to the extent that MPT induction may be enhanced (55) .
In summary, the amounts and ratios of each class of proteins involved in cell ion-homeostasis, the lipid milieu of membranes in which these proteins reside, and the lower threshold requirement to produce ROS and ROS-derived alkenes change with aging and reduce the threshold for acute Ca2+ -overload that occurs within the older heart. An overview of changes in the aging heart that predispose it to a reduced threshold for abnormal Ca2+ handling during acute stress is shown in Figure 9⇓. With advancing age, various constitutive changes occur in cardiac cells that can make a substantial impact on cardiovascular function, reducing the cellular capacity to tolerate and adapt to stress such as ischemia/reperfusion. Notable among these changes are a relative decrease in the ω3 :ω6 PUFA content ratio in cellular membranes, enhanced cellular production of ROS, and alterations in proteins governing Ca2+ mobilization, especially the decreased expression or function of SERCA2 and Na+ -Ca2+ exchange proteins. These changes lead to disturbances in excitation-contraction coupling (prolonged action potential and Ca2+ transient) and in intracellular Ca2+ compartmentalization (abnormal regulation of systolic Ca2+ and increased diastolic [Ca2+ ]; increased mitochondrial [Ca2+ ] (CaM ) leading, in turn, to spontaneous Ca2+ -oscillations and arrhythmias. These and other changes lead to mitochondrial dysfunction that may impair energy metabolism (with diminishing ATP production) together with excessive ROS production. Overproduction of ROS, and the inability to scavenge excess ROS, leads to damaging lipid-peroxidation and relatively more diffusible, but still potent, reactive intermediates, such as HNE, which affect widespread protein targets and amplify Ca2+ dysregulation and mitochondrial abnormalities. Finally, the abnormal Ca2+ handling and reactive species buildup can induce the MPT with the release of contents that activate a sequence of events leading to cell death (by apoptosis or necrosis). Recent research has demonstrated that these effects of aging are exacerbated by poor nutritional habits but can be ameliorated by diets substituting ω3 PUFAs for ω6 PUFAs (Table 3⇑, Figure 7⇑). Other promising avenues to ameliorate these aging changes include gene therapy to restore the expression of Ca2+ -regulatory proteins and the effective use of antioxidants.
Overview of changes in the aging heart that predispose to a reduced threshold for abnormal Ca2+handling. (See text for details) From (17).
Reduced Chronic Adaptative Capacity of the Older Heart
Many of the multiple changes in cardiac structure, excitation, myofilament activation, contraction mechanisms, Ca2+ dysregulation, deficient β AR signaling, and altered gene expression of proteins involved in excitation-contraction coupling that occur with aging (Table 2⇑) also occur in the hypertrophied myocardium of younger animals with experimentally induced chronic hypertension (1 , 56) and in failing animal or human hearts, in which they have been construed as an adaptive response to a chronic increase in LV loading. The expression of atrial natriuretic (57) and opioid (58) peptides—molecules that are usually produced in response to chronic stress—is increased in the senescent rodent heart (Figure 10⇓) (57 , 59) . This suggests that the old rodent heart is “stressed” on a chronic basis. When additional chronic mechanical stresses that evoke substantial myocardial hypertrophy, such as pressure or volume overload, are imposed upon the older heart, the response in many instances is reduced. There is evidence that transcriptional events associated with hypertrophic stressors become altered with advancing age: the nuclear binding activity of the transcription factor NF-κ B is increased and that of another transcription factor, Sp1, is diminished (60) . Because these transcription factors each influence the expression of several genes, they may contribute to the pattern of gene expression observed in the hearts of senescent rodents as well as dictate the limits of adaptive responses to the imposition of additional chronic stress. The acute induction of both immediate-early and late genes that are expressed during the hypertrophic response is blunted in hearts of aged rats after aortic constriction (61 , 62) . Similarly, the acute induction of heat shock protein genes in response to either ischemia or heat shock is reduced in hearts of senescent rats (63 , 64) . A similar loss of adaptive capacity is observed in younger rats that have utilized a part of their reserve capacity prior to a growth factor challenge (65) .
Impaired Ischemic Preconditioning In The Human Elderly Heart
The most potent form of cardiac protection known against prolonged periods of ischemia, such as during coronary artery occlusion of an acute myocardial infarction (MI), results from the activation of endogenous mechanisms triggered by brief episodes of transient ischemia closely preceding the infarct (66) . This phenomenon is known as ischemic preconditioning (PC) and occurs in most species, including humans (67) . Patients with preinfarction angina, which is arguably the clinical manifestation of “spontaneous” PC, experience a better clinical outcome after MI than those without preceding angina. During the past decade, multiple studies in acute MI patients undergoing coronary reperfusion with thrombolytic agents were able to show that patients who experienced bouts of preinfarction angina manifest significant clinical benefits, such as improved LV function and reduced myocardial infarct size, episodes of congestive heart failure and cardiogenic shock, arrhythmias, and in-hospital mortality, compared to those without angina (68 –74) . However, the aging process appears to reduce or eliminate the protection afforded the adult heart by PC as demonstrated in experimental animal studies (75 , 76) as well as in several clinical studies (77 , 78) . It is possible that this age-associated reduction in the protective efficacy of PC by preinfarction angina could explain at least part of the higher mortality in the elderly—as compared to younger adult patients—who have coronary heart disease (79) . Intriguingly, in the rat model (80) as well as in elderly patients (79) , physical activity and exercise training restores the cardiac protective effect of PC and preinfarction angina otherwise lost during aging.
CONCLUSION
Our understanding of the age-associated alterations in vascular and cardiac structure and function has advanced at both the cellular and molecular levels, providing valuable clues that may assist in the development of effective therapies to prevent, delay, or attenuate the cardiovascular changes that accompany aging. Many of these age-associated changes are being increasingly recognized as risk factors for cardiovascular diseases. Future efforts to modulate these risks imparted by aging will require the close collaboration of researchers, including molecular cardiologists, cardiovascular physiologists, pharmacologists, and clinically based translational research.
Acknowledgments
The authors would like to thank Christina R. Link for her secretarial support.
- © American Society for Pharmacology and Experimental Theraputics 2002
References
Edward G. Lakatta, MD, (Left) is the Director of the Laboratory of Cardiovascular Science, National Institute on Aging, National Institutes of Health. He also holds adjunct appointments as Professor, Department of Physiology, University of Maryland School of Medicine, and Professor, Cardiology Division, Johns Hopkins School of Medicine. The overall goals of his research program are: 1) to study basic mechanisms in cardiac excitation-contraction coupling and how these are modulated by surface receptor signaling pathways in cardiac muscle; 2) to determine mechanisms of normal and abnormal function of vascular smooth muscle and endothelial cells; 3) to identify age-associated changes that occur within the cardiovascular system and to determine the mechanisms for these changes; 4) to study myocardial structure and function and to determine how age interacts with chronic disease states to alter function; and 5) to establish the potentials and limitations of new therapeutic approaches such as gene transfer techniques. He is an Inaugural Fellow of the Council on Basic Cardiovascular Sciences of the American Heart Association, and the recipient of the Eli Lilly Award in Medical Science.
Steven J. Sollott, MD, (Right) is a Senior Investigator and the Head of the Cardioprotection Unit, Laboratory of Cardiovascular Science, National Institute on Aging, National Institutes of Health. He also holds an adjunct appointment as Assistant Professor, Cardiology Division, Johns Hopkins University School of Medicine. His research focuses on the structure and function of cardiovascular cells along three principal lines: 1) mechanisms of cardiac contractility; 2) nature and control of mitochondrial instability during oxidant stress; and 3) cellular changes after vascular injury. One of his more noteworthy findings relates to the observation that paclitaxel (Taxol), a drug used to treat cancer, could markedly attenuate vascular restenosis after angioplasty. Human clinical trials currently in progress in Europe, Asia, and the United States indicate that paclitaxel may prevent human restenosis with minimal toxicity.
