Alagebrium in combination with exercise ameliorates age-associated ventricular and vascular stiffness

Abstract

1. Introduction

Age related cardiovascular disease results in part from the formation of advanced glycation end-products (AGEs). These accrue as a result of the irreversible non-enzymatic glycosylation and modification of proteins including collagen, elastin, and other functional proteins (Zieman and Kass, 2004). Cell surface receptors for AGEs, so called “RAGEs,” have been identified and coupled to pathways that initiate cellular inflammation and tissue damage. Interaction of AGE ligands with RAGE enhances receptor expression initiating a positive feedback loop (for review see Ulrich and Cerami, 2001). AGEs increase tensile stiffness and make these molecules less degradable by metalloproteinases leading to tissue accumulation. Although AGE-formation is increased in hyperglycemic conditions, it also occurs under normoglycemic conditions and is time-dependent, suggesting that it could be important in age-related vascular changes. Thus, AGEs may ultimately contribute to the ventricular and vascular stiffness seen in the aging population (Kass et al., 2001). The recent development of agents, for example Alagebrium, that can break carbon– carbon bonds and impair AGE-formation has important implications for treating age related cardiovascular dysfunction (Kass, 2003). Studies have demonstrated enhanced cardiac compliance, and cardiac output in aged dogs (Asif et al., 2000) and improved arterial compliance (Kass et al., 2001) following treatment with Alagebrium.

Reduced physical activity is well recognized to result in cardiovascular deconditioning (McGuire et al., 2001). Sedentary aging is associated with stiffening of the heart and vasculature, while life-long exercise training prevents these developments (Arbab-Zadeh et al., 2004). However, even 1 year of exercise training starting later in life (after the age of 65) does not seem to be effective at improving either arterial (Shibata and Levine, 2012) or cardiac (Fujimoto et al., 2010) stiffness in human studies. This could be due to the fact that in older individuals AGE products are already present in substantial quantities. Thus, it is possible that one mechanism by which exercise prevents stiffening of the heart and blood vessels is by limiting AGE accumulation within cardiovascular structures. The formation of AGE products is not spontaneously reversible and it is plausible that concurrent pharmacological reversal is necessary to allow full expression of the exercise response.

Ventricular–vascular coupling (VVC), or the interaction between ventricular contraction and systemic arterial vessels, is a central determinant of net cardiovascular performance (Kass, 2002). One of the seminal features of the aging vasculature is an increase in arterial stiffness or decreased arterial compliance (CA). Indices of vascular stiffness (increased pulse pressure, increased PWV) are important risk factors for cardiovascular disease, specifically in the elderly (Mitchell et al., 2010). This has led to a search for agents that increase CAand reduce pressure pulsatility such as novel AGE crosslink breakers (Kass et al., 2001). Although multiple pathways are dysregulated in aging blood vessels (Fitch et al., 2001; Wang et al., 2000), including impaired endothelial vasodilator function and NO signaling, increased oxidative stress, upregulation of matrix metalloproteinases (MMP-2) (Wang and Lakatta, 2002; Wang et al., 2003) and monocyte chemotactic factor and its receptor (on vascular smooth muscle cells) (Spinetti et al., 2004), AGE-accumulation may be critically important in limiting the cardiovascular response to exercise in older adults. As a corollary, AGE cross-link breakdown could enhance the response to exercise in cardiovascular structures. We therefore hypothesized that ALT, a novel AGE crosslink breaker, acts synergistically with exercise and improves the parameters of vascular and ventricular stiffness to a greater degree than either intervention alone. In this study, we determined the effect of ALT and exercise alone, or in combination, on cardiovascular parameters including parameters of vascular stiffness and diastolic and systolic functions.

2. Methods

2.4. Pressure–volume loops

Closed-chest pressure–volume (P–V) relationships were measured using a combined pressure–conductance Millar catheter (SPR- 838, 2.0 Fr) and the Aria combined pressure–conductance system (Millar). Rats were anesthetized with ketamine/xylazine (120mg/kg, 12 mg/kg; ketamine, xylazine, respectively), and a tracheostomy was performed for mechanical ventilation. The conductance catheter was inserted through the right carotid artery and advanced into the left ventricle (LV). A polyethylene catheter was inserted into the femoral artery for measurement of arterial pressure. In addition, a catheter was inserted into the femoral vein for administration of albumin solution. A 2 Fr Fogarty catheter was advanced through the femoral vein into the IVC. Left ventricular P–V relations were measured through preload reduction by transiently inflating the balloon. Hemodynamic measurements were performed as previously described in detail (Barouch et al., 2002; Varghese et al., 2000). Cardiac preload was determined using left ventricular end-diastolic volume (EDV) and pressure (EDP). Myocardial contractility was determined as the peak rate of rise in LV pressure (+dP/dtmax) divided by instantaneous pressure (dP/dtmax-IP). The load-independent end-systolic elastance (Ees) calculated from the linear end-systolic pressure–volume relationship (ESPVR) fitted in the pressure range of 40 to 110 mm Hg. The effective arterial elastance, Ea, was calculated as the ratio of the left ventricular end-systolic pressure and the stroke volume. Ventricular to arterial coupling was assessed as the Ea/Ees ratio. Active diastolic relaxation was measured by peak −dP/dt (−dP/dtmin), and the time constant of ventricular relaxation (tau). The latter is derived from the mono-exponential equation describing the rate of fall of ventricular pressure (regression of dP/dt vs pressure), and is a well validated measure of the rate of diastolic relaxation (Gilbert and Glantz, 1989). Mean arterial pressure (MAP) was calculated from peak systolic pressure (Pmax) and pressure at dP/dtmax (P@dP/dt<span< a=""> style="font-size: 0.8461em; line-height: 1.6363em; position: relative; vertical-align: baseline; bottom: -0.25em; top: 0.25em;">max) with the equation MAP=P@dP/dt<span< a=""> style="font-size: 0.8461em; line-height: 1.6363em; position: relative; vertical-align: baseline; bottom: -0.25em; top: 0.25em;">max+(Pmax−P@dP/dt<span< a=""> style="font-size: 0.8461em; line-height: 1.6363em; position: relative; vertical-align: baseline; bottom: -0.25em; top: 0.25em;">max)/3, which estimates the center of gravity of the analog triangular pressure waveform in each contraction. The total peripheral resistance index (TPRI) is derived from the ratio of MAP and cardiac output index (COI): TPRI=MAP/COI. Ejection fraction (EF) is derived from stroke volume (SV) and EDV as follows: EF=SV/EDV. The end-diastolic P–V relationship (EDPVR), an assessment of diastolic compliance, was determined from the end-diastolic P–V data obtained utilizing IVC occlusion. A nonlinear fit was used to characterize EDPVR, yielding the curve: EDP=C*e(K*EDV) where C and K are real value constants. End-diastolic pressure increases exponentially with respect to rate constant K as age- and treatment-dependent end-diastolic volume increases with higher loads in consecutive contractile cycles. K values depend on the instantaneous rise in pressure (P) over the increase in volume (V), ΔP/ΔV=1/compliance, and could be used to approximate ventricular stiffness and diastolic function. Characterizing the exponential power of the nonlinear fit for EDPVR, these K values are shown in Fig. 2A.

Fig. 2

Parameters of myocardial contractility including the load-independent measurement end-systolic elastance (Ees) are restored by exercise and Alagebrium: aging is associated with a large decrease in measures of contractility as seen in representative P–V ...

3. Results

Hemodynamic parameters in young (Y) and old (O) rats, and the modulating influence of ALT and Ex, individually and in combination in O rats are outlined in Table 1. A total of two animals died during the study (one old in the ALT treated group and one old in the exercise group).

Table 1

Hemodynamic parameters in young (Y) and old (O) rats and the modulating effects of ALT and EX individually and in combination in O rats.

Heart rate (HR), MAP, COI/CI, SV, EF, ESP, dp/dtmax, −IP, −dp/dt and preload recruitable stroke work (PRSW) were all significantly decreased in O compared to Y animals. EDV, EDP and TPRI were significantly increased in O animals compared to Y. While ALT or Ex each caused partial but statistically significant improvement in age-associated hemodynamic perturbations, their combination resulted in a greater and more consistent improvement of these parameters (Table 1, column 2 vs. 5).

Fig. 1A illustrates representative P–V loops in a young and an old rat. The P–V loops are substantially shifted to the right in the old compared to the young rat. Moreover, the peak systolic pressure developed and the slope of the ESPVR are significantly attenuated. Fig. 1B illustrates the representative P–V loops at baseline in young and old rats and following ALT and Ex, individually and in combination in O rats. Again the P–V loops in Y and O rats are significantly different. ALT or Ex partially attenuates the influence of age but their combinations render a P–V relationship in O animals that is almost identical to that observed in Y rats.

Fig. 1

Representative traces of families of P–V loops from O and Y rats demonstrating differences in both ESPVR and EDPVR: (A). P–V loops were obtained using a combined conductance/manometric catheter. EDPVR and ESPVR were determined following ...

3.1. Systolic function

As illustrated in Fig. 2, and outlined in Table 1, there was a significant difference in all indices of systolic function in O compared to Y rats. All parameters of systolic function were restored in rats receiving both exercise and ALT such that the parameters were not statistically significantly different from Y rats. For example, dP/dtmax (Fig. 3A) was reduced by 37% in O vs Y rats consistent with depressed contractile function. Also, SV, a load dependent measure of systolic function was decreased in O vs Y rats but was significantly improved in O rats treated with ALT+Ex. In addition, Ees, the slope of the ESPVR (Fig. 3B), effective arterial elastance (Ea) and preload recruitable stroke work (PRSW), were significantly decreased in O rats. Remarkably, O rats who exercised and received ALT demonstrate a substantial restoration of contractility; animals who received either intervention alone also improved, but not to the same degree as the combined intervention (Table 1). Furthermore, there was a significant improvement of Ees and PRSW in rats receiving combined treatment consistent with an additive/synergistic effect of ALT and exercise on contractile function. Exercise, ALT, and combination therapy all improved EF significantly in old animals. However, these improvements remained lower than the normal EF in young rats. EF is a highly load dependent measure. Regardless of the increase in SV, the increase in EDV in O rats was such that a greater restoration of EF was not observed, even though EDV decreased with exercise and ALT toward the level of young animals. With regard to ventricular–vascular coupling, Ea was significantly decreased in O as compared to Y rats and was not significantly different in the O groups treated with Alt or exercise, but was significantly increased in the combined exercise and Alt group such that it was not different from Y. The Ea/Ees ratio, a metric of the efficiency of coupling between the ventricle and vasculature, was not significantly different across all the groups suggesting that efficient coupling is maintained in all groups.

Fig. 3

Combination of exercise and Alagebrium improves parameters of diastolic function in old rats: old rats demonstrate an increase in the slope of the non-linear EDPVR (increased diastolic stiffness). In addition, parameters of active ventricular relaxation ...

3.3. Vascular stiffness

We used PWV as an in vivo measure of vascular stiffness. As illustrated in Fig. 4, PWV was approximately doubled in O rats compared to Y. Both ALT treatment and exercise alone resulted in an equal and significant decrease in PWV in O rats. Remarkably, the combination of exercise and ALT resulted in a further substantial decrease in PWV such that this parameter was not statistically different to that observed in Y rats. This outcome is consistent with a significant impact of combined exercise and ALT on large artery compliance.

Fig. 4

Age is associated with a substantial increase in vascular stiffness as measured by an increase in the pulse wave velocity (PWV). Stiffness is improved with Alagebriumor exercise and by the combination. (*) P<0.01 vs. Y; (#) P<0.01 vs. ...

4. Discussion

We have demonstrated for the first time that the combination of exercise training with the AGE crosslink breaker Alagebrium has a synergistic effect and significantly improves the parameters of both systolic and diastolic functions, and decreases vascular stiffness in a rat model of aging.

Our observations in old rat hearts are consistent with previous studies on aging (Boluyt et al., 2004; Chang et al., 2004; Raya et al., 1997; Rozenberg et al., 2006; Walker et al., 2006) and those of Pacher et al. (2004, 2002) who demonstrated significant impairments in systolic performance, and in diastolic relaxation and stiffness in ~24 month old Fischer 344 rats. Additionally, we have demonstrated that effective arterial elastance (Ea) is decreased in O rats and is restored toward that of Y with exercise and Alt. Furthermore Ea/Es ratio is modestly decreased in O rats and restored to Y value in the O rats treated with a combination of exercise and Alt. This suggests that the combined intervention restores optimal ventricular–arterial coupling. This finding is somewhat consistent with Chang et al. (2004), who have also demonstrated that aging is associated with a decrease in Ea, that Ea is partially restored with the inhibition of AGE. While we found no statistically significant effect of ALT alone on these parameters, we did however observe similar trends. The differences may be accounted for by the difference in AGE crosslink breaker used (Aminoguanidine vs. ALT) or the dose administered.

It is now well accepted that aging in humans is associated with a significant rise in vascular stiffness which causes a significant increase in left ventricular systolic wall stress (for review see Kass, 2002). Effective arterial elastance (Ea) is coupled to an increase in ventricular stiffness (Ees) maintaining myocardial efficiency at near normal levels. This increase in Ea in aging humans is in contrast to the findings in the rat aging model (Chang et al., 2004; Pacher et al., 2004). This discrepancy is explained on the basis of a markedly reduced stroke volume and therefore blood pressure in the O rats, despite an increase in central vascular stiffness. Efficient coupling is however maintained in this rodent model.

Acute changes in Ees are generally interpreted to represent changes in contractility while chronic changes, specifically increases, are indicative of alterations in chamber geometry and structure of the heart. In contrast, dP/dtmax, a contractility specific index, is not altered with aging in humans who in general demonstrate an increase in Ees. In aging humans diastolic dysfunction is the predominant feature, but systolic dysfunction can also develop, albeit perhaps at a later stage, in the disease process.

One of the interesting observations in this study is that ALT alone significantly improved the parameters of active myocardial relaxation as reflected by a decrease in tau. In addition, there was a significant improvement in the parameters of contractility as reflected by an increase in dp/dtmax. These observations suggest that ALT may have effects that modulate the molecular processes involved in active contraction and relaxation. It has recently been demonstrated that glycated proteins stimulate reactive oxygen species production in cardiac myocytes by an NADPH oxidase-dependent mechanism (Zhang et al., 2006). It is well established that not only does NADPH-dependent ROS production mediate myocardial hypertrophy, but it also leads to depressed myocardial contractility and impaired active relaxation. Thus, ALT may improve these parameters by preventing this process. The improvement in active relaxation with ALT may be particularly important for aged humans since even life-long and extensive exercise does not normalize measures of dynamic relaxation (Prasad et al., 2007).

Another interesting observation relates to the decrease inHR seen in O vs Y rats. This depressed HR is restored in ALT+Ex treated rats. Previous studies support a decrease in intrinsic heart rate (following sympathetic and parasympathetic blockade) in vivo in aging rats (Irigoyen et al., 2000) and humans (Jose and Collison, 1970). This has been confirmed in isolated spontaneously beating rat hearts (Friberg et al., 1985). Another possibility is amore profound depth of anesthesia in old rats compared to young rats, as there is a decrease in minimum alveolar concentration (MAC) of volatile anesthetics in old animals compared to young. Prior studies, however, demonstrated similar depressions of heart rate and blood pressure in old rats despite the use of different anesthetic techniques (Capasso et al., 1990; Pacher et al., 2004). Additionally there is evidence supporting the concept that a significant age-related impairment in baroreflex control of HR and sympathetic nerve activity occurs (Lakatta, 1993). These changes are likely to have significant implications for age-related alterations in cardiovascular regulation. It is unclear from this study why combined Ex and ALT therapy might restore HR toward that of the young. One potential explanation is that exercise might improve baroreflex control of HR in old rats. Alternatively, ALT might decrease central vascular stiffness and in that way alter the sensitivity of the carotid baroreceptor “rheostat,” restoring baroreflex sensitivity.

There is increasing evidence that exercise attenuates age-associated diastolic dysfunction in animals (Brenner et al., 2001; Starnes et al., 2003) and humans. For example, sedentary aging results in a large increase in LV diastolic stiffness compared to young but similarly sedentary subjects (Arbab-Zadeh et al., 2004). In contrast, master athletes had normal P–V relations compared to young controls, and similarly normal stress–strain curves demonstrating conclusively that life-long exercise (more than 2 decades in this study) preserves ventricular compliance (Arbab-Zadeh et al., 2004). While the effect of a lifetime of exercise on maintaining ventricular compliance is well established, until recently the effect of exercise training starting later in life remains unknown. Dr. Levine and coworkers have recently demonstrated that 1 year of endurance training had little effect on ventricular compliance in previously sedentary seniors (Fujimoto et al., 2010). On the other hand, there was a significant improvement in ventricular–vascular coupling. Interestingly, this is consistent with our observations in the exercised but not ALT treated rats. EDP and EDPVR were only modestly altered by exercise in contrast to the significant effect of exercise on vascular stiffness.

The large increase in vascular stiffness observed and its secondary effects on the heart (VVC) in the old rats are consistent with what has been previously described by our group and others (Atkinson, 2008; Kim et al., 2009; Santhanam et al., 2010; Soucy et al., 2006). Furthermore, the significant improvement in vascular stiffness, as measured by PWV, in the animals who received either exercise or ALT, and the improvement in the group that received both interventions, are consistent with the proposed biologic effect of each intervention and our original hypothesis: a physiologically meaningful interaction between the two interventions. The increase in vascular stiffness with age results from a number of interacting/underlying mechanisms involving all the components of the vascular wall including the endothelium, structural proteins, vascular smooth muscle and intercellular matrix. The accumulation of AGE in the matrix contributes to vascular stiffness changes with aging. In animal studies, Alagebrium improved vascular and ventricular properties (Asif et al., 2000; Candido et al., 2003; Liu et al., 2003; Vaitkevicius et al., 2001) and in one of the first human trials, ALT decreased pulsatile load as measured by pulse pressure and increased vascular compliance (Kass et al., 2001). Additional studies demonstrated improvement in left ventricular diastolic filling, quality of life, decreased left ventricular mass, left ventricular ejection and reduced aortic stiffness (Little et al., 2005; Liu et al., 2003), while another trial failed to demonstrate significant improvements in cardiac function, or systolic blood pressure with ALT administration without exercise (Hartog et al., 2011). In a large double-blind multicenter trial, ALT was compared to placebo for the treatment of hypertension in patients with and without hypertension (SAPPHIRE and SILVER studies) (Bakris et al., 2004). The primary outcome (drop in systolic blood pressure by 5 mm Hg) was not achieved in this study, however ALT decreased systolic blood pressure by 2–3 mm Hg more than the placebo group. These results are unfortunately difficult to interpret as all groups (both ALT and placebo) had a significant drop in blood pressure of 6–10 mm Hg during the first 2 weeks of the study (Bakris et al., 2004). Currently, there is no clear evidence that ALT alone has lesser or more profound effects in humans than in rodents. Furthermore, there are no human studies that have investigated the combined effect of ALT and exercise on vascular stiffness. It should also be noted that at least in theory, AGEs could de-stabilize the fibrous cap of an atherosclerotic cap pre-disposing to myocardial infarction. However in the more than 1000 subjects studied in clinical trials of Alagebrium to date, there has been no signal of increased myocardial infarction or death. Longer trials of this agent involving a full year of therapy in otherwise healthy seniors are currently ongoing (clinicatrials.gov NCT01014572) and will be necessary to ensure safety of this agent.

Although others have demonstrated a significant effect of ALT in spontaneously hypertensive aged rats (Susic et al., 2004), this study represents the first demonstration of the synergistic effects of ALT in combination with exercise in a purely aging model. Exercise has also been shown to improve vascular characteristics. For example, Tanaka et al. have demonstrated that the predicted age-related increase in central arterial stiffness did not develop in physically active women (Tanaka et al., 1998, 2000). Our data demonstrates a significant improvement in vascular stiffness in old exercised rats compared to age-matched sedentary controls. This observation is consistent with the idea that exercise can improve vascular properties.

What is the underlying mechanism contributing to the exercise and ALT-mediated reduction in vascular stiffness? Vascular stiffness is determined by dynamic (the vascular smooth muscle tone) and structural properties of the vascular wall. NO may regulate vascular stiffness both by acutely altering the tension developed by vascular smooth muscle cells, and by modifying the structure of blood vessels via its effects on vascular matrix protein characteristics (Wilkinson et al., 2004). There is emerging evidence that exercise can improve vascular properties by NO-dependent mechanisms. For example, in rats exercise increases the expression of NOS in the aorta and skeletal muscle arterioles (Spier et al., 2004; Tanabe et al., 2003). In humans, exercise increases NOS phosphorylation and thereby NO signaling (Hambrecht et al., 2003). In addition to promoting vascular remodeling, exercise causes activation of mechano-transduction mechanisms in both endothelial and vascular smooth muscle cells, which result in: 1) stimulation of signaling pathways that up-regulate genes which promote anti-atherogenesis, eNOS expression and phosphorylation, and NO bioavailability, 2) up-regulation of anti-oxidant mechanisms (e.g. superoxide dismutase (SOD)), and 3) potentiation of anti-inflammatory mechanisms (Fukai et al., 2000; Gielen et al., 2005; Hambrecht et al., 2003; Kojda and Hambrecht, 2005). This may further explain the findings by Guo et al. that ALT contributes to oxidative stress reduction by up-regulating activities of SOD and glutathione peroxidase (Guo et al., 2009).

There are a number of potential limitations of this study. We have been unable to measure AGE or RAGE receptors in cardiac tissue using immunohistochemical staining and biochemical assays. This would allow us to confirm that the mechanism by which ALT and Ex are working depends on its crosslink breaking activity. Additionally, we do not have measures of LV mass pre- and post-treatment (or any other echocardiographic data) in order to determine whether ALT cross-link breaking activity alters ventricular structure or wall thickness, and exercise capacity and catecholamine levels were also not measured. Some past studies have questioned the validity of measuring absolute ventricular volume via conductance catheters (Boltwood et al., 1989) (for a detailed review, see Burkhoff, 1990). These studies found that the offset volume is not constant but rather varies with loading conditions and stroke volume appears to be inconsistent if measured by ventriculography vs. a conductance catheter. However, the studies lack adequate validations utilizing sonomicrometry and ventriculography to measure a broad range of left ventricular volumes. At this point conductance catheters still provide accurate measurements during the cardiac cycle against which new methods are compared. Furthermore, conductance catheters offer the advantage of obtaining a continuous signal without the need to invade the chest cavity (Burkhoff, 1990; Kottam et al., 2011). An additional limitation is that PWV was not adjusted for either heart rate or blood pressure, both of which can theoretically change the measured value for PWV without affecting the underlying arterial wall properties. Currently there is no clear consensus in the literature regarding the association of heart rate and pulse wave velocity. While some studies demonstrate an association between PWV and heart rate (Amar et al., 2001; Sa Cunha et al., 1997), others fail to show this effect (Blacher et al., 1999; Nurnberger et al., 2003; Papaioannou et al., 2008). Despite these limitations, our physiologic measurements confirm the salutatory effects of the intervention on both systolic and diastolic functions in the aging rat heart. Lastly there is ongoing debate on the role of exercise intensity on improving the cardiovascular status (Cameron and Dart, 1994; Hoydal et al., 2007; Kemi et al., 2005). Our model of continuous moderate exercise resembles the regime mostly practiced and recommended for adults to improve cardiovascular function (Cameron and Dart, 1994). Recent animal studies however indicate that high intensity aerobic interval training might be more efficient at improving the cardiovascular status (Hoydal et al., 2007; Kemi et al., 2005).

In summary, exercise and ALT in combination improve diastolic stiffness, systolic and diastolic dysfunctions, LV contractility and large artery compliance in a rodent model of aging. This result may have important therapeutic implications for the aging population if it can be confirmed along with its safety profile over a longer period of time in seniors. Further studies extending these results to humans would seem warranted.