β₃-Adrenergic receptor antagonist improves exercise performance in pacing-induced heart failure

Abstract

In heart failure (HF), the impaired left ventricular (LV) arterial coupling and diastolic dysfunction present at rest are exacerbated during exercise. We have previously shown that in HF at rest stimulation of β₃-adrenergic receptors by endogenous catecholamine depresses LV contraction and relaxation. β₃-Adrenergic receptors are activated at higher concentrations of catecholamine. Thus exercise may cause increased stimulation of cardiac β₃-adrenergic receptors and contribute to this abnormal response. We assessed the effect of L-748,337 (50 μg/kg iv), a selective β₃-adrenergic receptor antagonist (β₃-ANT), on LV dynamics during exercise in 12 chronically instrumented dogs with pacing-induced HF. Compared with HF at rest, exercise increased LV end-systolic pressure (P_ES), minimum LV pressure (LVP_min), and the time constant of LV relaxation (τ) with an upward shift of early diastolic portion of LV pressure-volume loop. LV contractility decreased and arterial elastance (E_A) increased. LV arterial coupling (E_ES/E_A) (0.40 vs. 0.51) was impaired. Compared with exercise in HF preparation, exercise after β₃-ANT caused similar increases in heart rate and P_ES but significantly decreased τ (34.9 vs. 38.3 ms) and LVP_min with a downward shift of the early diastolic portion of LV pressure-volume loop and further augmented dV/dt_max. Both E_ES and E_ES/E_A (0.68 vs. 0.40) were increased. LV mechanical efficiency improved from 0.39 to 0.53. In conclusion, after HF, β₃-ANT improves LV diastolic filling; increases LV contractility, LV arterial coupling, and mechanical efficiency; and improves exercise performance.

during exercise, the cardiac output increases to meet the elevated oxygen delivery to the muscles. Normally, this is achieved without an abnormal increase in left atrial (LA) pressure (P) (7, 10, 37, 57). Enhanced sympathetic stimulation during exercise increases heart rate and accelerates relaxation, resulting in smaller left ventricular (LV) end-systolic volume (V_ES), lower minimum LV pressure (LVP_min), and increased LA-LV pressure gradient. This produces a marked increase in the peak rate of LV filling (dV/dt_max) without an increase in LAP to abnormal levels. Exercise increases LV contractility (53), both through the adrenergic stimulation and a positive force-frequency relationship (51). These responses result in an increased stroke volume despite reduced duration of ventricular ejection and filling. Moreover, LV contractility, LV-arterial coupling (E_ES/E_A), and LV mechanical efficiency [stroke work (SW)/P-V area (PVA)] are all increased during normal exercise.

In contrast, after the development of heart failure (HF), the slowed relaxation at rest is further impaired during exercise and LVP_min is elevated (11). Increased preload during exercise elevates LV end-diastolic pressure (P_ED) and volume (V_ED), and diastolic P-V relation goes upward throughout the filling phase. Thus mean LAP abnormally increases, which contributes to the development of exertional dyspnea and limited exercise tolerance. Furthermore, contractility during exercise is impaired due to blunted response to β-adrenergic stimulation (40, 60, 64) and an impaired force-frequency relationship (51, 60). Thus the abnormal LV arterial coupling and diastolic dysfunction present at rest are exacerbated during exercise (11, 37).

The β₃-adrenergic receptor (β₃-AR) was initially identified in fat and some vascular tissues and subsequently found in the myocardium. Unlike β₁- and β₂-ARs, β₃-AR stimulation inhibits cardiac contraction and relaxation (14, 21, 22, 31, 46, 47, 58) through its link to inhibitory G proteins (G_i) and involves both nitric oxide (NO)-dependent (14, 22) and -independent effects (2, 14, 21, 22, 46, 67). Stimulation of cardiac β₃-AR has a negative effect on cardiac contraction (i.e., negative inotropy states with the β₃-AR) but has no direct chronotropic effects. In the failing heart, β₃-ARs are upregulated (14, 46) and are relatively resistant to desensitization (58) due to a lack of βARK-1 phosphorylation sites (14, 22, 36, 46, 47), while cardiac β₁-ARs are downregulated and β₁- and β₂-ARs are desensitized and uncoupled from the G_s protein in the failing heart (39). This altered balance of cardiac β₃-, β₁-, and β₂-ARs in HF may have important effects.

Previously, we found that stimulation of β₃-AR by endogenous catecholamine depresses LV contraction and relaxation at rest in conscious dogs with pacing-induced HF (47). In HF, during exercise, markedly elevated levels of catecholamine occur (18), which may cause enhanced activation of cardiac β₃-ARs. Thus the adverse effects of β₃-AR stimulation in HF may be accentuated during exercise. However, the role of β₃-AR during exercise performance in HF is unknown. Whether the beneficial actions of β₃-ANT we observed previously in HF at rest (47) can persist during exercise in HF remains to be determined. Investigations addressing this problem are timely and highly significant, because exercise intolerance is a hallmark of HF, but many existing HF therapies (including aldosterone antagonism, angiotensin-converting-enzyme-inhibitor therapy, and β-blockade) with clear salutary role on cardiac function at rest failed to improve exercise capacity in patients with HF reported by recent clinical trials (16, 28, 32–34, 48).

Accordingly, this study was undertaken to test the hypotheses that after HF there is an enhanced activation of the cardiac β₃-AR system during exercise, which contributes to the abnormal responses of LV contraction, relaxation, and LV-arterial coupling to exercise. We assessed the acute effect of L-748,337, a selective β₃-antagonist, on diastolic filling, LV arterial coupling, and mechanical efficiency at rest and during exercise in a conscious, chronically instrumented dog model with pacing-induced HF (34).

MATERIALS AND METHODS

Instrumentation

This investigation was approved by the Wake Forest School of Medicine Animal Care and Use Committee. Twelve healthy, adult, heartworm-negative mongrel dogs (body wt of 25–35 kg) were instrumented to measure three LV internal dimensions, LVP, and LAP. One myocardial lead (model 4312: Cardiac Pacemakers, Minneapolis, MN) was implanted within the myocardium of the right ventricle (RV), and the lead was attached to unipolar multiprogrammable pacemakers (model 8329: Medtronic, Minneapolis, MN) positioned under the skin in the chest. Hydraulic occluders were placed around the venae cavae by a technique described previously (12, 40, 47, 62).

Data Collection

Studies were performed after full recovery from instrumentation (10 days after original surgery) with the dogs standing and then running on a motorized treadmill (model 1849C; Quinton, Seattle, WA) as previously described (11, 12).

Experimental Protocol

Induction of HF.

Studies were performed after the animals had fully recovered from instrumentation. After completion of the normal baseline studies as previously described (12, 47), rapid RV pacing (at 220–240 beats/min) was initiated using the pacing protocol to induce HF (13, 47). After 4–5 wk of rapid pacing, when the LVP_ED, during the nonpacing period, had increased by >15 mmHg over the prepacing control level, we obtained HF data.

Effect of exercise with HF without and with β₃-ANT.

During the stable HF period, we examined the cardiac response to exercise before and after β₃-ANT. Briefly, before each study, the pacer was turned off, and the dog was allowed to equilibrate for at least 40 min. Then, steady-state measurements were obtained and blood was collected from LA catheter at rest while the dogs stood on a motorized treadmill (model 1849C; Quinton). Variably loaded LV P-V loops were generated by transient occlusion of the venae cavae (VCO) as previously described (11, 47, 62). The first HF exercise was then performed with the dogs running on the treadmill.

The treadmill speed was gradually increased every 1–2 min from 2.5 mph up to the maximum tolerated level (4.5–6 mph). The animals exercised at this level until they could no longer keep up with the treadmill. At submaximal levels of exercise, blood was collected. Both steady-state and VCO data were obtained, and then the treadmill was suddenly stopped. Data were acquired during 12-s periods throughout the exercise protocol. We analyzed the data recorded during submaximal level of exercise to avoid marked fluctuation by respiration. The total exercise time ranged from 4 to 8 min.

After 40 min of recovery, a selective β₃-ANT, L-748,337 (50 μg/kg), was administered intravenously. This dosage of L-748,337 (50 μg/kg iv) has been demonstrated to selectively block β₃-ARs by our group and others (9, 44, 46, 47). We have also verified the adequacy of its effect on β₃-blockade by abolishing BRL-37,344-induced responses.

Ten minutes after L-748,337 treatment, hemodynamic data and blood samples were collected at rest. Then treadmill exercise protocol was again performed (13) and at submaximal levels of exercise, blood was collected. We previously observed that there is no difference in the response to exercise repeated after a 40-min rest period (11, 13). The values of resting controls were also similar before initial exercise and with a 40-min resting period after exercise.

Plasma NO Determination

Plasma NO concentrations were determined by the Hypertension Center Core Laboratory of Wake Health as previously described (1).

Data Processing and Analysis

As previously described, LVV, LV P_ES-V_ES relation and its slope (E_ES), and SW-V_ED relation and its slope (M_SW) were analyzed (12, 13). Relaxation was evaluated by determining the time constant of the isovolumic decrease of LVP (τ). LVP from the time of peak −dP/dt until mitral valve opening was fit to the exponential equation LVP = P_A exp(−t/τ) + P_B, where t is time and P_A, P_B, and τ are constants determined by the data. Although the decrease in isovolumic P is not exactly exponential (65), the time constant, which is derived from the exponential approximation, provides an index of the rate of LV relaxation (24, 42). In addition, τ was also calculated by the Weiss method (monoexponential decay model to zero asymptote). LV-arterial coupling was quantitated as the ratio of E_ES to E_A, determined as P_ES/stroke volume (SV). The LV PVA, which represents the total mechanical energy, was determined as the area under the P_ES-V_ES relation and systolic P-V trajectory above the P_ED-V_ED curve. The efficiency of conversion of mechanical energy to external work of the heart was calculated as SW/PVA (50). Data acquisition and analysis were not blinded to group identity vis-a-vis β₃-ANT treatment due to the study design.

Statistical Analysis

Statistical comparisons were made with repeated-measures ANOVA. If there was a significant overall effect, intergroup differences were assessed by using Fisher's protected least significant difference. Each dog also served as its own control. Significance was established as P < 0.05. Data for steady state are expressed as means ± SD; values for LV P-V relations are expressed as means ± SE.

Drugs

(S)-N-{4-[2-({3-[3-(acetamidomethyl) phenoxy]-2-hydroxypropyl}-amino)ethyl]phenyl} benzene sulfonamide (L-748,337) was a gift from Merck Research Laboratories (Rahway, NJ).

RESULTS

Abnormal Hemodynamic Responses in Pacing-Induced HF: Rest vs. Exercise

At rest, as summarized in Tables 1 and 2, consistent with our past reports (13, 47), chronic RV rapid pacing in a canine model produced progressive LV systolic and diastolic dysfunction. Compared with normal rest, LV P_ED, V_ED, and V_ES significantly increased. LV systolic dysfunction was shown by significant decreases in E_ES and M_sw with reduced LV dp/dt_max. LV diastolic dysfunction was indicated by significantly elevated LVP_min, LVP_ED, and mean LAP, as well as prolonged LV relaxation with increased τ.

Table 1.

Effects of β₃-blockade on steady-state hemodynamic data at rest and during exercise after HF

	Control		β3-Blockade
	Rest	Exercise	Rest	Exercise
Heart rate, beats/min	133 ± 16	170 ± 17*	136 ± 21	166 ± 17*
Maximum dP/dt, mmHg/s	1,482 ± 280	1,872 ± 404*	1,565 ± 328	1,864 ± 348*
Minimum dP/dt, mmHg/s	−1,453 ± 158	−1,761 ± 293*	−1,477 ± 188	−1,746 ± 242*
Stroke volume, ml	13.4 ± 5.0	14.5 ± 5.3*	14.0 ± 5.4	15.2 ± 5.9*
LV end-diastolic pressure, mmHg	33.2 ± 5.4	41.2 ± 7.1*	31.7 ± 5.2†	37.8 ± 6.5*
LV end-systolic pressure, mmHg	94 ± 7	106 ± 11*	93 ± 7	102 ± 10*‡
Minimum LV pressure, mmHg	16.9 ± 3.7	20.6 ± 4.0*	15.6 ± 2.9†	17.0 ± 3.7*‡
Mean LA pressure, mmHg	25.4 ± 4.9	33.0 ± 5.1*	23.3 ± 3.8†	28.2 ± 4.8*‡
LV end-diastolic volume, ml	52.7 ± 21.4	53.5 ± 21.8	52.1 ± 20.5	53.5 ± 21.8*
LV end-systolic volume, ml	39.4 ± 16.1	38.8 ± 17.0	38.3 ± 16.9†	38.0 ± 15.3‡
Maximum dV/dt, ml/s	172 ± 67	235 ± 86*	191 ± 73	260 ± 97*‡
Stroke work, mmHg·ml	988 ± 452	1,117 ± 412*	1,060 ± 521	1,210 ± 503*‡
Cardiac output, ml/min	1,740 ± 633§	2,383 ± 925*	1,840 ± 744	2,448 ± 918*
Ea, mmHg/ml	8.0 ± 2.6	8.2 ± 2.2	7.6 ± 3.0	7.5 ± 2.8‡
Time constant of relaxation, ms	36.3 ± 5.0	38.3 ± 4.8*	35.2 ± 5.1†	34.9 ± 4.9‡

Values are means ± SD (n =12). HF, heart failure; dP/dt, rate of rise of left ventricular pressure; LV, left ventricular; LA, left atrial; maximum dV/dt, peak rate of mitral flow; E_a, arterial elastance

^*P < 0.05, exercise vs. corresponding control rest.

^†P < 0.05, β₃-blockade rest vs. control rest.

^‡P < 0.05, HF β₃-blockade exercise vs. HF control exercise.

Table 2.

Effects of β₃-blockade on pressure-volume relations at rest and during exercise after HF

	Control		β3-Blockade
	Rest	Exercise	Rest	Exercise
EES, mmHg/ml	3.5±0.7	3.0±0.7*	4.1±0.7†	4.6±0.7*‡
EES/EA	0.51±0.13	0.40±0.09	0.65±0.16†	0.68±0.10‡
MSW, mmHg	61.1±1.0	63.3±6.7	65.5±2.9	75.1±3.2‡
SW/PVA	0.45±0.04	0.39±0.06	0.52±0.03	0.53±0.03‡

Values are means ± SE (n = 4). E_ES, slope of linear end-systolic pressure-volume (P_ES-V_ES) relation; SW, stroke work; PVA, LV pressure-volume area; M_SW, slope of SW-end-diastolic volume (V_ED) relation

^*P < 0.05, exercise vs. control rest.

^†P < 0.05, β₃-blockade rest vs. control rest.

^‡P < 0.05, HF β₃-blockade exercise vs. HF control exercise.

During exercise, as typically displayed in Figs. 1 and 2, such abnormalities at rest were exacerbated. During exercise after HF, heart rate, LVP_ES, τ (38.3 ± 4.8 vs. 36.3 ± 5.0 ms), LVP_ED, LVP_min, and mean LAP (33.0 ± 5.1 vs. 25.4 ± 4.9 mmHg) increased. These changes were accompanied by a consistent rightward and upward shift of the early diastolic portion of the LV P-V loop. During early diastole, at an equivalent LVV, the LVP was significantly higher during exercise than at rest after HF (62). In addition, the impaired LV arterial coupling present at rest were exacerbated during HF exercise. As summarized in Table 2, compared with HF at rest, HF exercise caused a significant decrease in E_ES, while E_A was relatively unchanged, resulting in decreased E_ES/E_A ratio with reduced SW/PVA (0.39 ± 0.06 vs. 0.45 ± 0.04).

Fig. 1.

Examples of the effect of β₃-antagonist on left ventricular (LV) diastolic filling during exercise after heart failure (HF). Analog recordings of LV pressure (LVP), left atrial pressure (LAP), and the peak rate of mitral flow (dV/dt_max) at rest, during exercise, and exercise with β₃-antagonist obtained from one conscious chronically instrumented dog after pacing-induced HF. Compared with HF at rest, the maximum mitral peak filling rate (dV/dt_max) increased during exercise. Thus the increase in dV/dt_max during exercise after HF resulted entirely from an increase in LAP. Compared with HF exercise, after β₃-antagonist, exercise caused similar increases in LV end-systolic pressure (P_ES) and heart rate. LAP was lower while τ and LVP_min decreased. There was an increase in the early diastolic mitral valve pressure gradient. dV/dt_max was further augmented with β₃-antagonist during exercise.

Fig. 2.

Examples of the effects of β₃-antagonist on LVP_ES-LV end-systolic volume (V_ES) relations during exercise after HF. LV P-V loops recorded following transient caval occlusions in one conscious dog after HF at rest, during exercise, and during exercise after β₃-antagonist treatment. After HF, compared with rest, exercise caused a decrease in the slope of the LV P_ES-V_ES relation, E_ES. Compared with HF exercise without treatment, HF exercise following treatment with β₃-antagonist produced marked leftward shifts of the LV P_ES-V_ES relation with increase in the slope of E_ES, indicating that β₃-antagonist increased LV contractility after HF during exercise.

β₃-ANT Improves Hemodynamic Responses in Pacing-Induced HF: Rest vs. Exercise

At rest after HF, β₃-ANT significantly improved LV systolic and diastolic function without significant changes of plasma levels of NO (43.0 ± 12.8 vs. 41.8 ± 13.2 μmol/l). As shown in Table 1 and Fig. 1, consistent with our past report (47), compared with HF at rest, β₃-ANT produced no changes in heart rate and LVP_ES but significantly reduced E_A, P_ED, and mean LAP. The dV/dt_max (191 ± 73 vs. 172 ± 67 mmHg/s) was increased due to decreased LVP_min (15.6 ± 2.9 vs. 16.9 ± 3.7 mmHg) and τ. SV and cardiac output were increased. After β₃-ANT, E_ES was also increased. Both the E_ES/E_A ratio and SW/PVA (0.52 ± 0.03 vs. 0.45 ± 0.04) were significantly improved (Table 2 and Fig. 2).

Importantly, during exercise after HF, treatment with β₃-ANT prevented HF exercise-induced adverse effects on LV systolic and diastolic dysfunction (Fig. 2). As shown in Tables 1 and 2, compared with exercise in HF preparations, exercise after β₃-ANT treatment caused similar increases in heart rate, P_ES, and NO (64.6 ± 14.8 vs. 60.2 ± 11.5 μmol/l) but significantly attenuated HF exercise-induced increase in LVP_ED and mean LAP; reversed exercise-induced abnormal increases in LVP_min, τ, and upward shift of the early diastolic portion of LV P-V loop; and further augmented dV/dt_max (Table 1 and Figs. 1 and 3). In addition, during HF exercise with β₃-ANT treatment, there was a reversal of abnormal HF exercise response in ventricular-vascular coupling and cardiac mechanical efficiency. As demonstrated in Fig. 3 and summarized in Table 2, during HF exercise after treatment with β₃-ANT, HF exercise-caused decreases in E_ES and E_ES/E_A were converted to increased E_ES (5.3 vs. 3.7 mmHg/ml) and E_ES/E_A (0.84 vs. 0.55). Thus treatment with β₃-ANT further improved HF exercise SW/PVA from 0.39 to 0.53. The duration of exercise was also significantly increased (7.2 vs. 5.6 min).

Fig. 3.

Group mean data on the differences between HF at rest and HF exercise with and without treatment of β₃-antagonist on LV filling, LV arterial coupling, and stroke work (SW)/P-V area (PVA). Compared with HF at rest, HF exercise caused an increase in dV/dt_max due to significant increases in mean LAP, while τ and LVP_min were increased. In contrast, during HF exercise with treatment of β₃-antagonist, there was a greater increase in dV/dt_max with decreased τ and smaller increase in mean LAP_. During HF exercise, LV arterial coupling (E_ES/E_A) decreased, resulting from a significantly increased E_A, but decreased E_ES, which lead to a reduction in SW/PVA. In contrast, β₃-antagonist reversed the HF exercise-induced abnormal responses and caused significant increases in E_ES, E_ES/E_A, and SW/PVA during HF exercise.

DISCUSSION

The major symptom and cause of disability in HF patients are exercise intolerance. Despite marked progress in the therapeutic approach of patients with HF, exercise intolerance remains to be a hallmark of the disease (16, 33, 39, 56). Current treatment of exercise intolerance is unsatisfactory in HF patients. The β-adrenergic agonist dobutamine and phosphodiesterase inhibitor milrinone were associated with worse survival and clinical outcomes and did not improve quality of life in severe HF patients (39, 52, 61). Aldosterone antagonism, angiotensin-converting-enzyme-inhibitor therapy, and β-blockade treatment improve survival in patients with HF but failed to improve exercise capacity in recent clinical trials (16, 28, 32–34, 48).

In the current investigation, we showed, for the first time, increased activation of cardiac β₃-AR is an important contributor to exercise intolerance in HF and that administration of a β₃-ANT in a canine model of HF enhances the response to exercise with improved LV diastolic filling, increases LV contractility, and decreases arterial elastance with an overall improvement in LV-arterial coupling and mechanic efficiency. This study is important in that it challenges the existing paradigm that general downregulation of adrenergic signaling in congestive HF is protective, unravels a new mechanism of exercise intolerance in HF, and points to a new molecular target for this serious and growing health problem.

How does β₃-ANT administration prevent exercise-induced exacerbation of LV systolic and diastolic dysfunction and restore normal exercise response in HF? Previous observations from serial studies on the mechanism and treatment of exercise intolerance with HF performed by us and other groups have shown that abnormal exercise response in HF is attributed to several factors such as an enhanced sensitivity of LV relaxation to exercise-induced increased systolic load, altered force-frequency relation, chronotropic incompetence, impaired intrinsic contractility, reduced β-adrenergic reserve, abnormal ventricular-arterial interaction (5), and a blunted peripheral arterial vasodilator response to exercise (12, 13, 16, 27, 33, 37–39, 41, 51, 56). The current study indicates that excessive endogenous catecholamine stimulation during exercise in HF (18, 30) alters inotropic response to β-ARs due to enhanced β₃-AR stimulation with higher concentrations of catecholamine (20) with less responsive β₁- and β₂-AR (14, 17, 45). Thus cardiac β-AR desensitization in HF at rest was further exacerbated due to enhanced activation of β₃-AR.

It is noted that β₃-AR stimulation has been reported to produce vasodilatation. However, blocking endogenous β₃-AR with L-748,337 did not produce any evidence of vasodilatation. Compared with HF exercise without β₃-ANT, similar increases in P_ES and heart rate were achieved during HF exercise after β₃-ANT treatment. These effects are also different than that produced by existing HF therapies such as conventional β-blockers and ACE inhibitors. These observations are consistent with the findings of Pelat et al. (55) and suggest that endogenous β₃-AR activation does not have a significant effect on HR or blood pressure in HF. Thus the improvement of exercise performance during HF following treatment with a β₃-ANT should be due to the direct myocardial effects.

Our findings are comparable with some past reports on the effects of exogenous β₃-AR agonists (31, 46). Exogenous administration of a β₃-agonist induces marked vasodilation (63). Observations of the exogenous administration of β₃-agonists showed that β₃-AR-mediated signaling results in NO generation with elevated levels of NO (8, 14, 21). In rat thoracic aorta, SR59230A, a β₃-ANT, antagonized the β₃-dose-dependent relaxation evoked by SR58611 (23). Such findings raised concern that β₃-ANT might be deleterious in the failing heart by increasing afterload (44). However, current results and our previous study (47) showed that blocking β₃-AR with the current dosage of β₃-ANT does not increase LV afterload. Instead, our studies showed that β₃-ANT decreased E_A both at rest and during exercise. In our study, the plasma levels of NO were not altered by β₃-ANT during exercise, although we could not determine myocardial levels of NO. It has previously been known that activation of PDE2 via a β₃/nitric oxide synthase/cGMP pathway may further reduce inotropy and aggravate systolic function, particularly in the presence of concomitant β₁/β₂-stimulation (43). Recently, we found that in HF, chronic β₃-AR stimulation triggers upregulation of cardiac inducible nitric oxide synthase (iNOS). Oxidant stress from iNOS uncoupling aggravates cardiac dysfunction. These alterations were prevented in β₃-AR knockout (β₃KO) mice and also were reversed with the treatment of a β₃-ANT. Thus it is possible that β₃-ANT may inhibit β₃-AR-iNOS/NO signaling and contribute to its beneficial actions in HF during exercise. However, whether and to what extent administration of β₃-ANT affects the iNOS/NO pathway, thereby contributing to the beneficial action of L-748, 337, are unknown.

Our observation of the improvement of LV mechanical efficiency after acute treatment of β₃-ANT during exercise is consistent with the reports that acute administration of SR59230A, a β₃-ANT, attenuates the imbalance of systemic and myocardial oxygen transport induced by dopamine (25). β₃-ANT by L-748,337 enhances cardiac positive inotropic and lusitropic responses to dobutamine in dogs with pacing-induced HF. Moreover, hypotension or poor ejection would not be anticipated in the use of β₃-ANT, which is also different from the acute use of conventional HF drugs.

The current observation is in agreement with our recent serial studies in HF models showing remarkable salutary effects of the β₃-AR blockade on reversing β-AR desensitization and preventing HF progression. Our preliminary observations (abstract presented at American Heart Association Meeting) revealed that β₃-AR-deficient mice were protected from isoproterenol-induced HF. Recently, we further found that aging-induced desensitization of cardiac β-adrenergic signaling was prevented in β₃-AR knockout (β₃KO) aged mice. β₃KO reverses aging-induced downregulation of LV myocyte sarco(endo)plasmic reticulum Ca²⁺-ATPase and upregulation of iNOS. We also made observations on the restoration of normal cardiomyocyte basal and β-AR subtype modulation of L-type calcium current in HF by β₃KO. Consistent with past reports that these β₃KO animals had normal life spans with no cardiac pathological conditions detectable at any point (67). Our recent investigation is supported by evidence that the efficacy of some β blockers may result from the action of β₃-ARs (3, 19, 26, 39, 59, 66). β₃-AR antagonists improve clinical outcomes and limit HF progression and remodeling in experimental studies (19, 44, 47). Recent clinical studies have demonstrated that chronic treatment (6 mo) with carvedilol (a nonselective β-AR blocker with β₃-AR antagonist) was associated with a positive inotropic effect with significant improvement in load-independent indexes of myocardial contractility beyond what can be attributed to changes in LV chamber size and load after 3 mo (6). However, long-term (6 mo) administration of nebivolol (a β₃-AR agonist/β₁-AR antagonist with associated NO-releasing properties) failed to improve exercise in HF patients (15). Similarly, 3 mo of treatment of nebivolol reduced exercise duration while that of carvedilol did not (P = 0.02 for the difference) in randomized controlled trial (54).

Study Limitations

Several methodological issues should be considered in the interpretation of our data: 1) Observations were obtained from a pacing-induced HF canine model. Although rapid pacing produces an animal model of HF that closely mimics that of clinical congestive cardiomyopathy, this experimental model demonstrated biventricular chamber dilatation with increased LV and RV filling pressures and striking abnormalities in systolic and diastolic function similar to that found in patients with dilated congestive cardiomyopathy (7, 12, 35). We cannot ascertain that these results are applicable to HF from other causes. 2) We did not examine the effects of β₃-ANT on LV end-diastolic P-V relation during exercise in this investigation. Further studies are needed to focus on this point in HF. 3) Our study did not define the mechanisms of the action of β₃-ANT during exercise of HF in addition to its ready known blocking β₃-AR activation-induced enhanced inhibition on LV contraction and relaxation. It is possible that its beneficial actions are attributable to the alteration on β₃-AR stimulation activated subtypes of NOS/NO pathways. We also did not measure oxidative stress. Indeed, it is emerging that cardioactive amines are also able to bypass the receptors, enter the myocytes, and trigger the activity of monoamine oxidases, leading to increased oxidative stress due to the monoamine oxidase-driven generation of H₂O₂. This has been shown both in the case of ischemia-reperfusion (4) and in transverse aortic constriction-induced HF (29). Clearly, more insight will be gained from ongoing work designed to assess the influence of β₃-ANT on the interaction between β₃-AR stimulation and the subtypes of NOS/NO pathways. It should be clarified in a future study whether β₃-ANT may blunt an increased oxidative stress during exercise in HF preparations.

Our novel findings indicate that antagonizing detrimental effects of endogenous β₃-adrenergic stimulation may help preserve β₁- and β₂-AR function, lead to resensitizing the activity of the β-adrenergic system, and provide inotropic support and thus improve exercise performance in HF. However, our study only elucidated the positive systolic and diastolic effects of acute β₃-ANT administration after HF at rest and during exercise. Since acute gain of β₃-ANT may not correlate with long-term benefits, whether these beneficial actions of acute β₃-ANT administration during exercise in HF will persist after chronic long-term application is unclear. Thus the precise therapeutic role of chronic β₃-ANT for exercise tolerance of HF remains to be determined.

In conclusion, we found that in dogs with pacing-induced HF, β₃-AR activation contributes to exercise intolerance in HF. β₃-ANT improves LV diastolic filling; increases LV contractility, LV arterial coupling, and mechanical efficiency; and improves response to Ex. The current study supports the hypothesis that upregulation of cardiac β₃-ARs is the critical determinant of the dysfunctional β-AR regulation that occurs in HF at rest and during exercise and provides unique insight to the functional implications of β₃-AR blockade in exercise intolerance of HF.

GRANTS

This study was supported, in part, by National Institutes of Health Grants AA-12335 and HL-074318) and American Heart Association Grant 0530079N (to C. P. Cheng) and National American Heart Association Scientist Development Grant 0530079N (to H.-J. Cheng).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: S.M., H.J.C., and C.P.C. conception and design of research; S.M., H.J.C., A.M., H.H., Q.-H.H., and C.P.C. performed experiments; S.M., H.J.C., A.M., H.H., Q.-H.H., and C.P.C. analyzed data; S.M., H.J.C., A.M., H.H., Q.-H.H., and C.P.C. interpreted results of experiments; S.M. and C.P.C. prepared figures; S.M. drafted manuscript; S.M., H.J.C., A.M., H.H., Q.-H.H., W.C.L., and C.P.C. approved final version of manuscript; W.C.L. and C.P.C. edited and revised manuscript.