An advanced glycation endproduct cross-link breaker can reverse age-related increases in myocardial stiffness

Abstract

Decreased elasticity of the cardiovascular system is one of the hallmarks of the normal aging process of mammals. A potential explanation for this decreased elasticity is that glucose can react nonenzymatically with long-lived proteins, such as collagen and lens crystallin, and link them together, producing advanced glycation endproducts (AGEs). Previous studies have shown that aminoguanidine, an AGE inhibitor, can prevent glucose cross-linking of proteins and the loss of elasticity associated with aging and diabetes. Recently, an AGE cross-link breaker (ALT-711) has been described, which we have evaluated in aged dogs. After 1 month of administration of ALT-711, a significant reduction (≈40%) in age-related left ventricular stiffness was observed [(57.1 ± 6.8 mmHg⋅m²/ml pretreatment and 33.1 ± 4.6 mmHg⋅m²/ml posttreatment (1 mmHg = 133 Pa)]. This decrease was accompanied by improvement in cardiac function.

One striking characteristic of the normal aging process of humans and other long-lived species is the gradual decrease in the elasticity of the cardiovascular system. These structural and mechanical changes in the heart inexorably lead to increased left ventricle stiffness and, consequently, diastolic dysfunction, a significant contributing factor in heart failure in humans (1). Two major proposed mechanisms for the development of this stiffness are the gradual accumulation of additional myocardial interstitial collagen and glucose-dependent cross-links, termed advanced glycation end products (AGEs), of collagen in situ. Evidence that AGEs contribute importantly to the stiffening process is provided by the observation that hyperglycemia produces a greatly accelerated stiffening of the myocardium of young diabetic animals. Further evidence supporting the AGE hypothesis is the observation that agents that specifically inhibit AGE formation (2, 3) effectively prevent the pathological stiffening process of diabetes and aging. Recently, a new class of thiazolium derivatives has been developed (4) whose members are able to break established AGE cross-links. In vivo efficacy has been confirmed by experiments performed in rats, showing that the increased arterial stiffness associated with diabetes can be reversed by a short treatment with the cross-link breaker phenyl-4,5-dimethylthazolium chloride (ALT-711) (5).

An obvious question is whether AGE breakers can also reverse the cardiovascular stiffness associated with normal aging. In this paper, we have evaluated the effect of 4 weeks of daily treatment with the AGE breaker ALT-711 on the stiffness of the left ventricle in diastole in normal aged dogs. The results show that treatment with the AGE cross-link breaker in these aged subjects is associated with a significant reduction of left ventricle chamber stiffness. As a result, cardiac function significantly improves, as evidenced by increased left ventricular (LV) end-diastolic volume (EDV), stroke volume, and decreased end-diastolic pressure (EDP).

Materials and Methods

Study Design.

Healthy, conditioned, male mongrel dogs (22–32 kg) were entered into the study, which used a protocol approved by the Institutional Animal Care Committee. Subjects were assigned to one of three experimental groups: young animals (group 1; n = 7; 4.8 ± 0.4 years of age); old animals receiving active drug treatment (group 2; n = 8; 10.6 ± 0.7 years); and old animals not receiving active drug treatment (group 3; n = 5; 9.8 ± 0.8 years). The subjects were screened for diabetes (blood glucose and hemoglobin A_1C), anemia (hemoglobin), and kidney dysfunction (serum creatinine) both before (Table 1) and after (data not shown) treatment and were all normal.

Table 1.

Laboratory parameters of studied groups

Group	Glucose, mg/dl	Hb A1C, %	Hematocrit, %	Albumin, g/dl	Creatinine, mg/dl
1, Young	73  ± 5	2.2   ± 0.01	44  ± 1.0	3.6  ± 0.2	0.5  ± 0.06
2, Old treated	74  ± 2	1.75  ± 0.07	47  ± 1.5	3.4  ± 0.1	0.6  ± 0.07
3, Old untreated	85  ± 9	1.79  ± 0.02	46  ± 0.9	3.4  ± 0.2	0.8  ± 0.09

Each subject underwent a baseline hemodynamic study. The aged animals (groups 2 and 3) were studied again 4 weeks later, after group 2 had received the cross-link breaker ALT-711 as a gelatin capsule at a single oral daily dose of 1 mg/kg. A selective cardiac postmortem examination was also carried out for all subjects at the conclusion of study to determine LV mass.

Hemodynamic Evaluation.

Aged dogs were anesthetized with chloralose (75 mg/kg i.v.) after induction with sodium pentothal. A cuffed endotracheal tube was inserted, and ventilation was controlled by a Harvard respiration pump to maintain the arterial pH and PO₂ within physiologic range. Under sterile conditions, Goodale–Lubin catheters (no. 8 French) were introduced into the LV chamber and proximal aorta via the carotid artery. Gould (Cleveland) P23XL transducers with a Honeywell physiological recording system using VR16 and V2203A pressure amplifiers and a fluid-filled catheter system optimally damped for frequency-response recorded LV and arterial pressures. Transducers were placed at the midthoracic level and balanced to provide for equal sensitivity. Pressures obtained from the fluid-filled catheters correlated well on simultaneous comparison to those from a micromanometer catheter system (6). In the intact animal, simultaneous measures of LV pressure and volume were made before and after volume loading was accomplished by infusing 10% dextran-40 (3 ml⋅min⁻¹⋅kg⁻¹ over 3 min; Baxter Health Care, Deerfield, IL). EDP was measured by averaging three to four pressure determinations at the end-expiration phase of the respiratory cycle.

For LV volume determination, two-dimensional echocardiography was performed with the dog lying on its left side (7). By using a 3.5-MHz transducer, recordings were made at a paper speed of 50 mm/sec. Imaging location and time-gain settings were adjusted to yield optimal definition of endocardial borders, which were delineated by bubbling saline into the LV chamber. To minimize the influence of heart rate on these measurements, ventricular dimensions were compared at similar R-R intervals by using the Ultramark 4 system (Advanced Technology Laboratories). The end-diastolic and end-systolic dimensions for three to four consecutive cardiac cycles were measured and averaged, and the ejection fraction and stroke volume were subsequently calculated. Ventricular volume was derived by the length–diameter method (8), with apical views for measurements taken from the inner margins of the endocardial echoes. Endocardial and epicardial borders were traced directly from the video display onto a digitizing tablet. End-diastolic frames were selected for analysis by using the R wave (from lead 2) as a marker for end diastole. Subjects of this study were free of valvular or pericardial disease as assessed by echocardiography.

To characterize the diastolic pressure–volume relationship in the left ventricle (9), the exponential equation P = be^kV was used, where P = pressure in mmHg, V = volume in ml/kg, b = the pressure intercept in mmHg, and k represents the modulus of chamber stiffness in the intact ventricle. Two coordinates of pressure and volume were used. The early diastolic coordinates consisted of the lowest value of diastolic pressure before the mitral valve opened and the end systolic volume. EDP and volume were utilized as the second coordinates. The chamber-stiffness constant k was calculated as the slope of the natural logarithm of pressure to volume: ln(P) = kV + ln(b). Chamber stiffness was derived from the relation dP/dV = kP. The latter was normalized for the EDV, because the slope of dP/dV versus P has been found to be sensitive to variations in cardiac muscle volume (10). With a progressive increase in volumes calculated, myocardial stiffness would be expected to increase as a preload-dependent phenomenon (9, 11). Myocardial stiffness was calculated from E = k⋅stress. This method has been used to describe increased diastolic stiffness during experimental hypertension (7) and during induced collagenolysis associated with reduced chamber stiffness (11).

Statistical Analysis.

All results are expressed as mean ± SEM. Data were analyzed initially as a repeated-measures design by using multivariate ANOVA (12), which indicated that univariate statistics were appropriate. Two-way ANOVA analyses indicated that one or more pairwise comparisons could be significant. These tests were followed, in turn, by 1-degree of freedom pairwise contrasts. The α level of significance was set as P ≤ 0.05.

Results

Before treatment, the ventricles of older dogs (groups 2 and 3) were significantly less distensible (i.e., stiffer) as compared with the young dogs studied, as illustrated in Fig. 1. Specifically, the ventricles of older dogs exhibited a significantly higher EDP (P < 0.01) and a lower EDV (P < 0.01; Fig. 1A). These differences were accentuated markedly after intravascular volume loading with dextran (Fig. 1A.) In contrast to differences in ventricular characteristics, there were no significant differences between young and old subjects in other cardiovascular parameters measured, including the ejection fraction, heart rate, and systemic pressure, which were all normal (Table 2).

Figure 1.

(A) Under basal conditions, the ventricular dynamics of older (group 2) and younger (group 1) dogs differed in that group 2 exhibited significantly increased EDP and reduced EDV, consistent with increased stiffness. The expansion of intravascular volume by dextran accentuated this difference such that EDV increased in young dogs (group 1) while maintaining EDP below that of older dogs (group 2), which exhibited a smaller increase in volume but a larger increase in EDP. After 4 weeks of treatment with 1 mg/kg ALT-711 orally (group 2*), the EDV–EDP relationship in both basal and volume-expanded states moved toward that of the young animals, consistent with an increase in distensibility. (§ indicates a significant difference when compared with the older animals, whereas † indicates a significant difference between the pre- and posttreatment periods.) (B) Ventricular dynamics at baseline (group 3) and after 4 weeks of sham treatment (group 3‡) of a group of older dogs with identical characteristics to the treated group (group 2). The EDP–EDV relationship remained unchanged at baseline and after volume expansion.

Table 2.

Cardiovascular parameters of studied groups

Group		Heart rate, beats per minute	Systolic pressure, mmHg	Ejection fraction, %	Chamber stiffness, mmHg⋅m2/ml
1	Baseline	108  ±  5.7	153  ±  4.2†	51    ± 5.4	19.8  ±  3.7†
	Dextran vol. expansion	112  ±  7.4	162  ±  8.7	63.1  ± 3.5	35.3  ±  2.7†
2	Baseline	109  ±  3.8	178  ±  4.5	42    ± 5	57.1  ±  6.8
	Dextran vol. expansion	125  ±  3	170  ±  9	50    ± 3	111    ± 12.2
2*	Baseline	112  ± 14	180  ±  9	44    ± 5	33.1  ±  4.6§
	Dextran vol. expansion	132  ±  3.9	183  ± 11.9	42    ± 4	77.3  ± 10.4
3	Baseline	121  ±  4.5	166  ±  3.2	48.2  ± 2.9	54    ±  9.2
	Dextran vol. expansion	130  ±  4	193  ± 11.7	43.1  ± 3.6	104    ± 16.3
3‡	Baseline	124  ±  4.5	166  ±  3.2	48.2  ± 2.9	54    ±  9.2
	Dextran vol. expansion	133  ±  2.5	188  ±  8.6	46.8  ± 2.3	96    ± 14.7

Group 2* indicates after 4 weeks of ALT-711 treatment, and group 3‡ indicates after 4 weeks of sham treatment. 

^†Differs significantly from older groups (P < 0.05). 

^§Differs significantly from baseline of group 2 (P < 0.001). 

Following a 4-week period of daily ALT-711 oral administration, repeat hemodynamic studies demonstrated a significant increase in EDV (P < 0.001) without a change of the EDP in older dogs (group 2* in Fig. 1A). The physiological consequence of this change resulted in a significant increase in the stroke volume index (P < 0.05; Fig. 2). Other cardiovascular parameters measured posttreatment were unchanged (Table 2). Calculated LV chamber stiffness diminished after ALT-711 administration (Table 2) from 57.1 ± 6.8 mmHg⋅m²/ml in the pretreatment period to 33.1 ± 4.6 mmHg⋅m²/ml after treatment (P < 0.001). These effects observed at baseline were amplified after volume expansion such that the EDP–EDV relationship moved closer to that of untreated, young controls (compare groups 2* and 1 in Fig. 1A). Although the stroke volume index also increased further after volume expansion, it did not reach the level of statistical significance (P < 0.1; Fig. 2).

Figure 2.

Stroke volume index at baseline (group 2) significantly improves after treatment with ALT-711 (group 2*). §, P < 0.05 versus pretreatment basal value.

To assess the reproducibility of the LV hemodynamic measurements over time, a group of older dogs (group 3) underwent a repeat hemodynamic study after a 4-week interval. No significant differences were observed in any of the parameters measured between the two studies in the basal state or after dextran infusion (compare groups 3 and 3‡ in Fig. 1B and Table 2). Postmortem determination of LV weight revealed no differences among groups 1, 2, and 3 at 4.4 ± 0.3, 4.6 ± 0.4, and 4.3 ± 0.5 g/kg of body weight, respectively.

Discussion

This study in the aged dog with normal systolic function examined the ability of selective AGE cross-link cleavage by ALT-711 to modify diastolic stiffness of the left ventricle. Before treatment, basal diastolic stiffness of older dogs was increased compared with young subjects, as evidence by increased EDP with decreased EDV (Fig. 1). Distending the ventricle by expansion of the intravascular volume with dextran accentuated these differences, as expected, because the pressure needed to expand further a stretched myocardium increases in direct proportion to diastolic ventricular volume (9).

Daily ALT-711 administration for 1 month was associated with a significant increase of EDV. Because the filling pressures observed pre- and posttreatment were not different, this result is consistent with decreased stiffness: A given pressure will distend a less stiff myocardium more than a stiff one. An analogous increase in EDV has been observed previously during pharmacologic inhibition of collagen cross-links by another agent in a rodent model (11). The significant rise of stroke volume observed in the basal state and a further increase during dextran infusion (Fig. 2) directly results from an enhanced filling volume after cross-link cleavage in these aged subjects.

In the absence of systolic dysfunction, a reduction of early diastolic filling has been a relatively common finding in asymptomatic aged humans without coronary artery disease. If hypertension or hypertrophy (13, 14) is absent, this minimal diastolic dysfunction does not appear to be accompanied invariably by impaired isovolumic relaxation. Consequently, the diastolic filling alteration presumably is related to the passive properties of the left ventricle as reflected in the relationship between EDV and EDP. In this context, the upward shift of this relation to the left after dextran treatment indicates that increased pressure was required to distend the ventricle during diastole in the untreated animals.

The influence of afterload (as expressed by aortic pressure) may also affect LV function. In this study, however, blood pressure at baseline and after volume expansion was similar in group 2 both before and after ALT-711. Further, the adjustment of aortic pressure by the subtraction of LV EDP (13) indicates that the degree of hypertension in the anesthetized state was relatively mild. This finding is also reflected in the absence of LV hypertrophy, as determined by LV chamber mass measurements.

Increased afterload in aged humans in the basal state has been associated with a normal or increased LV EDV (15), so that afterload increments per se would not contribute to the reduced EDV in this canine model. However, the stroke volume rise after treatment may result from a reduced aortic as well as ventricular stiffness. In this regard, in a prior study during early treatment of diabetics with aminoguanidine, reduction of aortic stiffness did not result in a greater stroke volume than in untreated diabetic rats (16), suggesting that stroke volume increments are not related invariably to reduction of aortic stiffness.

Before treatment, the aged subjects had LV EDVs in the resting state that were nearly 50% less than those found in young animals. The latter had a higher stroke volume index, 27.5 ± 2.0 ml, as compared with 13.0 ± 2.7 ml in the aged animals. The rise of stroke volume and cardiac output with an unchanged ejection fraction after treatment is consistent with increased filling of the ventricle, a mechanism for increasing stroke volume in the aged. A previous study in diabetic rats with reduced cardiac output showed a rise of ≈30% in the basal state after ALT-711 treatment (5).

These data confirm a significant role of the accumulation of AGE cross-links in promoting the decreased cardiovascular compliance of aging. In addition, it is clear that substantial reversal of LV stiffness can be obtained after a brief treatment with ALT-711. Such reversal is associated with improved cardiac function. Whether additional courses of ALT-711 might produce even more recovery of function and whether the rate of reaccumulation of AGEs differs from that of normal aging remains to be determined. To date, two successful strategies targeting AGEs have been demonstrated, one inhibiting formation and the other breaking established AGE linkages. Such therapeutic approaches may affect appreciably the normal aging process, as well as the accelerated aging of diabetes.

Abbreviations

AGE

advanced glycation endproduct

LV

left ventricular

EDV

end-diastolic volume

EDP

end-diastolic pressure