INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

MAMMALIAN -ADRENERGIC RECEPTORS (-AR) play critical roles in the regulation of blood pressure and smooth muscle tone, cardiac chronotropy and inotropy, and energy metabolism. Virtually all organ systems express at least one of the three known -ARs (₁, ₂, and ₃), which explains the pleiotropic effects that follow nonspecific -AR agonist administration. From a therapeutic standpoint, there are many instances when -AR subtype-selective stimulation or blockade is desired, and therefore, a detailed knowledge of subtype-specific functions is necessary. The traditional approach to the assignment of subtype-specific functions has been to use subtype-selective agents. Although these are indispensable for many in vitro or ex vivo applications, their relative selectivity in vivo can be compromised and may differ significantly from that observed in vitro. As an example, biodistribution of the well-known ₁-AR selective antagonist CGP-20712A in rats differs significantly from the true ₁-AR subtype distribution (48). The ability to selectively block or activate a given receptor subtype is thus dependent on the dose of such agents, as well as their relative selectivity for the "specific" vs. "nonspecific" subtypes in vivo. These variables must be taken into account when attempting to define subtype-specific functions, especially when more than one subtype is present in a given target tissue, or when ligand accessibility to distinct subtypes differs at the target site. The demonstration that most or all -ARs are coexpressed in tissues such as heart (6), adipose (24), and vasculature (4, 32, 42) makes pharmacological isolation of subtype-specific functions of -ARs a significant challenge.

The distribution of -ARs in the heart has been determined both pharmacologically (6) and by quantitation of mRNA levels (46). In mammals, both ₁- and ₂-ARs are expressed in the heart, with ₁-ARs predominating at an ~75:25 ratio in ventricles and an ~60:40 ratio in atria and conduction tissue (6, 40). In human heart preparations studied with subtype-specific agonists and antagonists, both ₁- and ₂-ARs appear to couple to positive inotropic and chronotropic responses (6). In mice, the myocardial ratio of ₁- to ₂-AR and total -AR density are similar to that found in humans. However, our recent studies in ₁-AR gene-targeted mice have demonstrated that cardiac -AR inotropic and chronotropic responsiveness as well as adenylate cyclase stimulation appear to be mediated solely through the ₁-AR (38). It is still not clear whether the differences noted between these studies are because of species-specific coupling behaviors or the inability of subtype-selective agents to effectively discriminate -AR subtype functions in vivo.

Both exercise and stimulation of the baroreflex elicit physiological changes that are dependent on the integration of both sympathetic and parasympathetic nervous system activity. Whereas -ARs are implicated as primary sympathetic nervous system targets in the reflex response to altered hemodynamics, their proportional contribution to the total baroreflex response and the subtype specificity are not clearly defined (3, 13, 18, 41, 49). Short-term adaptive responses to exercise also require an integrated response from multiple neuroeffector systems. Cardiac output increases commensurate with increasing exercise loads to meet peripheral oxygen and energy demands, and the well-known effects of -AR agonists to regulate both heart rate and contractility suggest that these receptors may play an important role in this response. There is controversy, however, on the overall requirement for -ARs during exercise. Whereas some studies have shown that maximal exercise capacity is reduced in the presence of -AR antagonists (1, 9, 10, 19-21, 28, 44, 47, 50), others have failed to demonstrate this effect (5, 25, 29, 36). It is somewhat surprising that functional compensation can occur under -AR blockade, given the fact that -ARs have been invoked in virtually every aspect of the physiological response to exercise. Cardiac chronotropy and inotropy, skeletal muscle vasodilatation, lipid and carbohydrate mobilization, and airway conductance are all enhanced by -AR stimulation during exercise (2, 28, 51).

In the present study, we have complimented what has traditionally been a pharmacological approach to studying receptor function with a genetic approach. Genetic ablation of ₁-AR expression allows for an unambiguous test of ₁-AR function in the mouse and, when combined with the pharmacological approach, can also provide information on other important cardiovascular control mechanisms. Knockout studies can also reveal novel or unexpected functions, such as the prenatal lethality observed in ₁-AR-deficient mice (38), which had only weak support from prior pharmacological studies (23). On the basis of its predominant role in regulating cardiac inotropism and chronotropism, we hypothesized that ₁-AR function would be critical for maintaining both resting and maximally stressed cardiovascular function. However, in mice lacking the ₁-AR, we unexpectedly found that basal cardiovascular indexes were essentially unaltered and, furthermore, that the capacity to respond to stresses such as exercise was normal. The failure to impact exercise performance occurred despite the fact that chronotropic reserve was severely blunted in ₁-AR knockout mice. The use of conscious and unrestrained animals is a critical component of these studies, since many aspects of reflex control and receptor function are lacking in isolated or anesthetized animal preparations. Moreover, the ability to observe conscious mice both at rest and under exercise stress provides additional opportunities to uncover phenotypic alterations relevant to ₁-AR signaling.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Generation of ₁-AR-Deficient Mice

Mouse Instrumentation

1

1

Exercise Protocols and Metabolic Measurements

Physiological Measurements

Measurements of total hemoglobin (Hb) and percent saturation of Hb in mixed venous and arterial blood samples were determined both at rest and under constant treadmill exercise (20 m/min, see Exercise Protocols and Metabolic Measurements). Arterial or mixed venous blood (100 µl) was withdrawn into heparinized 1-ml syringes, and determinations of total Hb (in g/dl) and percent oxygen saturation were performed by hemoximetry (model OSM 3; Radiometer, Copenhagen, Denmark). Total Hb values were corrected for dilution by heparin sodium, which filled the dead space of the syringe.

Statistics

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Cardiovascular Indexes in ₁-AR Knockout Mice at Rest and After Pharmacological Manipulation

Resting values. Mice were instrumented under isoflurane anesthesia and allowed to recover for ~24 h before measurements were taken. There was no difference in the baseline heart rate and blood pressure of conscious, unrestrained wild-type and ₁-AR knockout mice (Fig. 1A), although the high variability in heart rate and blood pressure seen in conscious, unrestrained mice (8) may have obscured the slight trend toward lower values in ₁-AR knockouts.

View larger version (24K): [in this window] [in a new window]   Fig. 1.   Basal and stimulated cardiovascular measurements in wild-type and 1-adrenergic receptor (AR) knockout mice. A: basal heart rate and mean blood pressure, in beats/min and mmHg, respectively, are shown for wild-type mice (WT; solid bars, n = 24) and 1-AR knockouts (1KO; hatched bars, n = 36). B: percent change in heart rate or blood pressure are shown after administration of various agents. Iso, isoproterenol (3 µg/kg), n = 17 for WT, n = 24 for 1KO; SNP, sodium nitroprusside (30 µg/kg), n = 11 for WT, n = 16 for 1KO; Prop, propranolol (3 mg/kg), n = 7 for WT, n = 14 for 1KO; Phen, phenylephrine (100 µg/kg), n = 4 for WT, n = 5 for 1KO. * P < 0.05,  P < 0.005 by unpaired Student's t-test comparing WT and 1KO.

Response to -AR agonists and antagonists. Wild-type mice given isoproterenol responded with a simultaneous drop in mean blood pressure and increase in heart rate. Although ₁-AR knockouts showed a qualitatively similar response, the percent increase in heart rate after isoproterenol was significantly greater in wild types than ₁-AR knockouts, whereas the hypotensive effect was comparable in magnitude (Fig. 1B).

Baroreflex stimulation in ₁-AR knockouts. We tested the hypothesis that the tachycardic effect of isoproterenol in ₁-AR knockouts was because of an indirect, baroreflex-mediated mechanism by administering the direct vasodilator SNP, both alone (Fig. 1B) and in the presence of atropine (see below). The baroreflex response stimulated by SNP leads to a simultaneous increase in sympathetic outflow and a decrease in parasympathetic outflow, which generally results in a rapid restoration of blood pressure and tachycardia. Both ₁-AR knockouts and wild-type mice respond to SNP with an initial drop in blood pressure and consequent increase in heart rate. Neither of these responses was statistically different comparing wild types and ₁-AR knockouts. The magnitude of the tachycardic response in ₁-AR knockouts was similar when comparing isoproterenol to SNP (22.5 ± 4.7 vs. 25.6 ± 3.9% increase in heart rate, respectively; P = NS), suggesting that the baroreflex arc is primarily responsible for the heart rate changes induced by isoproterenol in ₁-AR knockout mice.

Role of vagal mechanisms in ₁-AR knockout cardiovascular regulation. We next tested the hypothesis that parasympathetic (i.e., vagal) mechanisms were playing a significant role in the tachycardic response of ₁-AR knockout mice to isoproterenol or SNP by administration of the muscarinic receptor antagonist atropine. The dose of atropine used (1 mg/kg) was determined by its ability to block the hypotensive response to intra-arterially administered carbachol (20 µg/kg). As can be seen in Fig. 2A and summarized in Fig. 2B, atropine significantly elevated heart rates in both wild-type and ₁-AR knockout mice. Heart rate at 5 min after atropine dosing was significantly different between the two genotypes (wild type, 572.5 ± 18.2 beats/min; ₁-AR knockout, 526.8 ± 6.6 beats/min; P < 0.02, n = 11 and 15, respectively). Of note, the qualitative response of ₁-AR knockouts to atropine differs significantly from wild types (Fig. 2A). Wild-type mice showed an initial marked tachycardia, to rates in excess of 600 beats/min, immediately after atropine administration. This brief increase in heart rate partially attenuated over the next 2-3 min, which we suspected was through a withdrawal of sympathetic tone. ₁-AR knockouts also displayed an initial tachycardia in response to atropine; however, the response was sustained and stable and lacked the initial "overshoot" seen in wild-type mice. The normal fluctuations in heart rate observed in conscious mice while grooming, sleeping, or eating (8) were not seen in atropine-treated ₁-AR knockouts, suggesting that chronotropic regulation in these mice depends almost completely on their ability to regulate parasympathetic or vagal tone.

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Effect of muscarinic blockade on cardiovascular indexes in wild-type and 1-AR knockout mice. A: heart rate tracings for wild-type and 1-AR knockout mice after atropine (Atro; 1 mg/kg), SNP (30 µg/kg), and Iso (3 µg/kg) administration. Shown are representative tracings of 2 individual mice, with heart rate shown in beats/min (identical scale for both). B: percent change in heart rate and mean blood pressure after administration of drugs described above. Phenylephrine (Phen) was used at 100 µg/kg. Atro/Iso, Atro/SNP, and Atro/Phen indicate that Iso, SNP, and Phen are given while mouse is under muscarinic blockade with atropine. Definitions are as in Fig. 1. Number of mice used per experiment was as follows: Atro, n = 11 for WT, n = 15 for 1KO; Atro/Iso, n = 9 for WT, n = 6 for 1KO; Atro/SNP, n = 9 for WT, n = 9 for 1KO; Atro/Phen, n = 3 for WT, n = 3 for 1KO. § P < 0.01 by unpaired Student's t-test comparing WT and 1KO. All heart rate increases are significant (P < 0.05) comparing predrug to postdrug rate, except 1KO + Atro/SNP as well as WT and 1KO + Atro/Phen.

Effect of ganglionic blockade with hexamethonium. We tested the effects of hexamethonium, a ganglionic blocking agent, in wild types and ₁-AR knockouts. This agent serves to block autonomic nervous system input at the nicotinic ganglia, which are common to both sympathetic and parasympathetic nervous systems. Hexamethonium (15 mg/kg) was infused slowly over a 10- to 20-min period, and after the infusion, mice were allowed to stabilize for at least 10 min before we administered other agents. There were small but nonsignificant decreases in the heart rates of wild-type and ₁-AR knockout mice after hexamethonium infusion in six wild-type mice and six ₁-AR knockouts (wild type, 528.6 ± 20.5 beats/min pre- vs. 465.8 ± 32.3 beats/min posthexamethonium; ₁-AR knockouts, 463.2 ± 16.9 beats/min pre- vs. 439.2 ± 21.6 beats/min posthexamethonium), whereas mean blood pressure was reduced significantly in both genotypes after hexamethonium administration (wild type, 117.2 ± 6.3 mmHg pre- vs. 98.0 ± 6.2 mmHg posthexamethonium; ₁-AR knockouts, 104.9 ± 6.5 mmHg pre- vs. 91.5 ± 6.4 mmHg posthexamethonium; P < 0.05 for both by paired t-test). These data suggest that when autonomic outflow is blocked, the intrinsic heart rates and blood pressures are comparable between the two genotypes. Under hexamethonium blockade, the baroreflex heart rate response to SNP was blocked by 70% in wild-type mice and 61% in ₁-AR knockouts (data not shown).

Cardiovascular Response to Exercise

The first phase of exercise experiments, which tested maximal exercise capacity, revealed that wild-type and ₁-AR mice achieved essentially the same work load before exhaustion. On both the graded and constant treadmill protocols, mice ran approximately the same distance, 450-500 m, with no significant differences between wild types and ₁-AR knockouts in either exercise regimen (Fig. 3A). On average, both wild-type and ₁-AR knockout mice terminated their run on the graded protocol during the 27.5 m/min, 20° inclination phase of their run (see MATERIALS AND METHODS). We have determined from other experiments that 27.5 m/min is well below the maximum acute sprinting speed for untrained mice (data not shown), indicating that during graded exercise, stoppage was from exhaustion, not an intrinsic limitation on running speed.

View larger version (32K): [in this window] [in a new window]   View larger version (23K): [in this window] [in a new window]   Fig. 3.   Exercise capacity as well as metabolic and cardiovascular effects of exercise in wild-type and 1-AR knockout mice. A: exercise capacity, measured in total distance run, was determined for wild-type mice and 1-AR knockout mice under graded treadmill (WT, n = 11; 1KO, n = 10) or constant treadmill (WT, n = 9; 1KO, n = 11) exercise protocols (see MATERIALS AND METHODS). B: oxygen consumption or CO2 production (ml · min1 · kg1) is shown for wild-type mice and 1-AR knockouts run under graded treadmill exercise protocol. Wedge is meant to show stepwise increases in work load. WT, n = 10; 1KO, n = 9, except when indicated, next to symbol (as mice stopped running). Data are not shown when 2 mice or less were running. C: cardiovascular effects of graded treadmill exercise on wild-type mice (n = 5) and 1-AR knockout mice (KO, n = 5) mice are shown, with heart rate (HR) in top panel and mean blood pressure (BP) in bottom panel. ¥ P < 0.001 comparing WT vs. 1-KO (from 0 to 20 m/min) by 2-way analysis of variance with repeated measures.

Figure 3B displays the O₂ and CO₂ of wild-type and ₁-AR knockout mice during the graded exercise protocol. There were no differences in O₂ and CO₂ between wild types and ₁-AR knockouts over the range of work loads, although there was a nonsignificant trend toward lower values in ₁-AR knockout mice. Of note, the general shape of both curves is well conserved, suggesting common response pathways during physical exertion. The disparity between the number of wild types and ₁-AR knockouts at the highest work loads reflects the dropout of animals during the highest exercise work loads because of exhaustion. The ratio between CO₂ production and O₂, or respiratory exchange ratio (RER), is one indicator of the transition between aerobic and anaerobic metabolism. In our hands, maximum RER is achieved usually 1-2 min after cessation of exercise. A comparison of basal RER as well as the maximum achievable RER reveals no significant differences between wild-type and ₁-AR knockout mice (basal RER, 0.82 ± 0.02 vs. 0.78 ± 0.02, respectively; maximum RER, 1.04 ± 0.03 vs. 1.05 ± 0.03, respectively; P = NS).

Although lack of the ₁-AR does not appear to affect total exercise performance or the metabolic response to exercise, an examination of the cardiovascular response to exercise reveals striking differences. In the second phase of the exercise protocol where mice were catheterized, blood pressure and heart rate could be directly monitored during all phases of the regimen. Figure 3C displays the chronotropic responses of wild type and ₁-AR knockouts to the graded treadmill exercise protocol. Wild-type mice showed a robust and work load-dependent increase in heart rate during the run, whereas the ₁-AR knockouts displayed little overall tachycardia and almost no work load-dependent increases in heart rate once the run had been initiated. At the maximum work load (20 m/min and 14° inclination), there was an ~200 beats/min difference between wild-type mice and ₁-AR knockouts. Exercise had little influence on mean blood pressure in either genotype, consistent with previous results in wild-type mice of various strains (8).

We also tested the effects of atropine on the chronotropic responses of wild types and ₁-AR knockouts to constant exercise (Fig. 4A). Naive mice of both genotypes were monitored at rest and at 20 m/min, 14°, then given atropine, and the experiment was repeated. Before atropine, wild-type mice showed a robust tachycardic response to the increased treadmill speed, whereas knockout mice showed a modest heart rate increase. After atropine administration, basal heart rates were elevated in both genotypes as expected. Atropine-treated wild-type mice had increased heart rate in response to the increased work load, and ₁-AR knockouts were still able to mount a small but significant increase in heart rate under muscarinic blockade (Fig. 4A). These studies suggest that in both genotypes vagal withdrawal accounts for ~50% of the tachycardic response to exercise, although the absolute increase in heart rate is much smaller for ₁-AR knockouts. Wild-type mice had increased heart rate 232.7 beats/min before atropine vs. 110.7 beats/min after atropine, whereas ₁-AR knockouts had increased heart rate 54.3 beats/min before atropine vs. 28.1 beats/min after atropine.

View larger version (36K): [in this window] [in a new window]   Fig. 4.   Effect of atropine or atropine plus ICI-118,551 on chrontropic response to constant treadmill exercise. A: effect of atropine (1 mg/kg) on chronotropic response to exercise is shown for wild-type mice (n = 3) and 1-AR knockout mice (n = 5) run under constant treadmill protocol. Mice were run at 20 m/min and 14° inclination for 4 min before and after atropine administration. * P < 0.05 by paired Student's t-test comparing rate at 0 vs. 20 m/min within same genotype. B: heart rate increases in response to constant treadmill exercise (20 m/min, 14° inclination, 4 min) were monitored in 1-AR knockout mice (n = 3) with no blockade, muscarinic blockade (atropine, 1 mg/kg), or muscarinic blockade plus 2-AR blockade (atropine plus ICI-118,551, 1 mg/kg). * P < 0.05 by paired Student's t-test comparing no blockade with muscarinic blockade or muscarinic plus 2-AR blockade.

To test the hypothesis that the chronotropic reserve present in ₁-AR knockouts under atropine blockade is because of ₂-AR activation, we administered the ₂-AR selective antagonist ICI-118,551 to atropine-blocked mice and stimulated chronotropy through the constant treadmill exercise protocol. The results, shown in Fig. 4B, demonstrate that ₂-AR blockade has no effect on the chronotropic reserve of atropine-blocked ₁-AR knockouts.

To determine the relative contributions of oxygen extraction and stroke volume in maintaining cardiac output during exercise, we also measured total Hb as well as percent oxygen saturation of Hb in mixed venous and arterial samples, at rest and under constant treadmill exercise. These studies revealed that total Hb content is greater in ₁-AR knockouts compared with wild-type controls (14.92 ± 0.92 vs. 11.44 ± 0.74 g/dl, respectively; P < 0.02, n = 5 for ₁-AR knockouts, n = 8 for wild types). However, oxygen extraction (arterial oxygen content  mixed venous oxygen content) is not significantly different when comparing ₁-AR knockouts to wild-type mice at rest (4.53 ± 1.29 vs. 4.79 ± 0.72 g/dl, respectively; n = 3 for ₁-AR knockouts, n = 5 for wild types) or under constant treadmill exercise (5.42 ± 1.34 vs. 6.73 ± 1.23 g/dl, respectively; n = 3 for ₁-AR knockouts, n = 5 for wild types).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The ₁-AR plays a central role among G protein-coupled receptors in its ability to regulate both resting and stressed cardiac function. The purpose of these experiments was to test the hypothesis that ₁-AR expression would be critical for maintenance of resting heart rate as well as the response to autonomic stimuli and exercise. We have demonstrated that the ₁-AR is an important mediator of the in vivo chronotropic response and have used this model to demonstrate the importance of parasympathetic pathways in murine homeostatic control mechanisms. The use of awake and unrestrained mice is a critical component to the success of these studies, since reflex pathways can be obscured in anesthetized mice and are absent in in vitro preparations.

The finding that resting heart rates are not significantly altered in ₁-AR knockout mice is somewhat surprising given the fact that both nonselective and ₁-selective -AR antagonists cause significant bradycardia in humans and other mammals (16). Despite this apparent lack of difference in heart rate between the two genotypes at rest, experiments using propranolol and atropine suggest that the ₁-AR does play some role in the control of resting heart rate in the mouse. First, propranolol caused a significant bradycardia in wild-type but not in ₁-AR knockout mice, suggesting that ₁-AR signaling does partially regulate resting heart rate in normal mice. Second, the heart rates of ₁-AR knockouts were significantly lower than wild-type mice after atropine administration. Both experiments reveal that a tonic level of ₁-AR signaling is present and contributes to resting heart rate in wild-type but not in ₁-AR knockout mice. Our inability to document statistically significant baseline differences in resting heart rate may result from the high degree of heart rate variability in nonanesthetized animals (8), which can be significantly reduced by either propranolol or atropine (17, 33). With the blockage of parasympathetic outflow with atropine, variability was reduced to the point where differences in heart rate between wild-type mice and ₁-AR knockouts became discernible.

Whereas we have previously demonstrated that wild-type and ₁-AR knockout mice did not statistically differ with respect to basal heart rate and blood pressure values (38), we report these values here as well since the surgical anesthetic agent was altered. Of interest, this change allowed us to observe a chronotropic responsiveness to isoproterenol that was lacking in our earlier studies (38). We attribute this to the variable effect of different anesthetic agents on cardiac reflexes, even after a 12- to 24-h recovery period. In the present study, chronotropic responses could be observed in ₁-AR knockouts to a variety of stimuli, including isoproterenol, SNP, and exercise. Of note, the basal heart rate value we previously obtained after surgery with methoxyflurane (542.6 ± 28.9 beats/min) is comparable to the peak heart rate value obtained after isoproterenol administration in mice instrumented under isoflurane studied here (547.3 ± 10.0 beats/min), suggesting that tachycardic reserve was limited or absent in mice studied within 24 h of methoxyflurane anesthesia.

The importance of vagal or parasympathetic mechanisms in the control of murine heart rate is clearly shown by these studies. Whereas ₁-AR knockouts display a qualitatively normal baroreflex response, the chronotropic component following a hypotensive stimulus can be almost totally blocked by pretreatment with atropine. Wild-type mice under the same conditions, while exhibiting an attenuated chronotropic response, retain significant chronotropic reserve in comparison with ₁-AR knockouts. The difference in chronotropic response noted between wild types and ₁-AR knockouts occurs despite the fact that the magnitude of the hypotensive stimulus is equivalent between the two genotypes. This suggests that vascular responsiveness and peripheral ₂-AR function is intact in ₁-AR knockouts. Given the extremely high heart rates in comparison with other species, a common assumption has been that sympathetic drive predominates in maintaining resting heart rate in the mouse (17). However, the demonstration that a significant portion of chronotropic reserve, whether stimulated by drugs or exercise, is blocked by atropine reveals the importance of vagal mechanisms. The relative contributions of parasympathetic vs. sympathetic control in the mouse are reminiscent of that found in larger mammals and humans (26), with the exception that the set point for intrinsic rate is considerably higher in the mouse. The phenomenon of high set points with balanced antagonism between parasympathetic and sympathetic input is also reminiscent of that seen in neonates of many species. The range for heart rate in normal mice can extend to a low of ~300 beats/min, and under significant stress such as exercise can exceed ~800 beats/min. On the basis of our observations, the upper end of this heart rate spectrum, above ~550 beats/min, is mediated by sympathetic stimulation through the ₁-AR. In comparison, in the human, the transition between vagal and sympathetic control occurs at a heart rate of ~100 beats/min (37, 39). In the mouse, vagal outflow can also increase under some conditions, such as the baroreflex response to phenylephrine, and serves to drive heart rate downward to the ~300 beats/min range in both wild-type mice and ₁-AR knockouts. Maximum achievable heart rates, whether stimulated by drugs or exercise, were unaffected by atropine within a given genotype. This suggests that under conditions of maximal stress, vagal input is small or nonsignificant.

Perhaps the most striking phenotype of the ₁-AR knockout mouse is the fact that, despite the large chronotropic deficit seen during exercise, total exercise capacity is not different from wild types. Treadmill exercise is one of the most effective means to increase heart rate, and, coupled with our previous findings which demonstrated a lack of inotropic response in ₁-AR knockouts, we expected that this disparity between wild types and ₁-AR knockouts would produce a functional deficit. Surprisingly, neither exercise capacity nor the metabolic response to exercise is significantly altered in these mice. The possible mechanisms by which the ₁-AR knockout mouse can compensate for this chronotropic deficit are revealed by an examination of the Fick equation, which yields the following relationship: O₂ = heart rate × stroke volume × (A-V O₂ difference), where A-V O₂ difference represents the difference in oxygen content between arterial and venous blood. At all work loads, O₂ is essentially equivalent between wild types and ₁-AR knockouts, whereas the heart rate response is markedly attenuated in ₁-AR knockout mice. To achieve the same O₂ as wild types, ₁-AR knockouts must compensate by either larger increases in stroke volume or larger increases in oxygen extraction than wild types. On the basis of our examination of arterial and venous oxygen content, our data do not support the hypothesis that ₁-AR knockouts compensate through increased oxygen extraction, but rather that compensation must be occurring through increases in stroke volume. In support of the idea that stroke volume changes underlie the compensatory changes in ₁-AR knockouts, their lower heart rates would allow for increases in diastolic filling time (15). This is further supported by human studies where subjects under -AR blockade maintained equivalent cardiac outputs by increasing stroke volume via the Frank-Starling mechanism (5), whereas chronotropic and inotropic responses remained largely suppressed. Pharmacological approaches to study the impact of -AR blockade on exercise have either shown deficits (1, 9, 10, 19, 20, 28, 47, 50) or no effect (5, 25, 29, 36) on exercise performance. Our attempt to test the effect of genetic ablation of the ₁-AR has shown that acute exercise performance is not affected by the congenital absence of ₁-ARs and also suggests that the chronotropic component of the exercise response may be less critical than previously thought.

Despite the central role that we have attributed to ₁-AR function in the control of heart rate, several lines of evidence indirectly suggest that other chronotropic regulators may play a minor role in both resting and stimulated heart rate. For example, there is a small but significant increase in heart rate after isoproterenol administration or exercise in atropine-blocked ₁-AR knockouts (atropine blockade is complete based on the failure of phenylephrine to produce bradycardia). Also, full autonomic blockade with hexamethonium results in lower heart rates than that achieved by parasympathetic blockade with atropine, suggesting that sympathetic nervous system-derived neuroeffectors may be modulating heart rate. On the basis of the failure of propranolol to produce bradycardia in ₁-AR knockouts at rest, and the inability of the ₂-AR-selective antagonist ICI-118,551 to block chronotropic responses to exercise, we assume that the ₂-AR is not fulfilling this role. Our previous study also supports the lack of ₂-AR involvement in direct chronotropic responses (38). Whereas overexpression of the ₂-AR in mice (30) results in hyperfunctioning hearts (showing increases in rate, inotropy, and adenylate cylase activation), it appears that the endogenous ₂-ARs show little if any functional coupling in the heart. These differences may be reconciled by the differences in methodology between the studies: supraphysiological expression of the ₂-AR by a nonhomologous promoter can clearly give rise to constitutive signaling; however, one cannot extrapolate from this transgenic model to the function of endogenous ₂-ARs expressed at physiological levels. Of the several receptor subtypes whose activity may be affected by ganglionic blockade but not by -blockade, both ₁-ARs (34) and the receptors for neuropeptide Y (11) have shown direct chronotropic coupling in isolated atria. Other potential chronotropic modulators that could be indirectly affected by ganglionic blockade include angiotensin II (27), serotonin (31), and endothelin (35). Further studies will be necessary to clarify the role of these mediators in controlling resting mouse heart rates and whether they can functionally compensate for ₁-AR signaling defects in mice lacking the ₁-AR.

Together with our previous characterization of the ₁-AR knockout, these studies suggest that, at least postnatally, loss of the ₁-AR has little functional impact on either resting or stressed cardiovascular performance. These results are striking given the central role that has been attributed to ₁-AR function in the heart. The loss of ₁-AR density and function in many forms of heart failure has been hypothesized to be a primary pathophysiological mechanism underlying the functional decline in patients suffering from this disease (6). Our studies show that whereas decreased ₁-AR function may lead to some functional deficits, downregulation of this receptor cannot be the primary cause of pump failure in a structurally normal heart. Such alterations in -AR density and/or function may be an "epiphenomenon," playing a minor role in the etiology of heart failure. It is clear, however, that inappropriate stimulation of the -AR system in chronic heart failure is detrimental (43), whereas -AR blocker therapy has beneficial effects (12, 14, 22). Once born, ₁-AR knockout mice appear normal in all respects in comparison with their wild-type counterparts and clearly show the proper response to stresses such as exercise. Cardiac histology (38) and echocardiographic studies (7) also fail to uncover any defects in ₁-AR knockout mice. This genetic model of ₁-AR ablation will allow future studies to address the importance of ₁-AR signaling to the development and progression of heart failure.

In summary, our studies indicate that within the murine sympathetic nervous system, ₁-ARs are probably the largest single contributor to cardiac chronotropic responses. In addition to the utility of the ₁-AR knockout in determining a direct role for ₁-ARs in cardiovascular function, this knockout model is also valuable for its ability to confirm the role of other homeostatic control mechanisms that are important for mammalian cardiovascular regulation.