INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

NITRIC OXIDE (NO) and -adrenergic receptor activation are the major determinants of large epicardial coronary artery dilation during exercise. In those vessels, endothelium-derived NO formation during exercise is a flow-dependent process, secondary to increases in shear stress (20). Preventing the rise in coronary blood flow (CBF) (20), endothelial denudation (3), or blockade of NO formation (27) blunts large epicardial coronary dilation. Therefore, NO-dependent dilation of conductance coronary arteries during exercise may be considered as a secondary phenomenon triggered by the dilation of resistance coronary vessels. In addition to NO, -adrenergic receptors directly contribute to conductance vessel dilation during exercise, because their blockade leads to a paradoxical constriction sensitive to -adrenergic receptor blockade (2). In resistance vessels, -adrenergic receptor activation causes dilator responses during exercise and serves as a feedforward mechanism (10, 11). In contrast, NO is not essential for increasing CBF and ensuring a close match between myocardial oxygen supply and demand during exercise (1, 4, 24, 25).

One intriguing observation is the complete failure of the large epicardial coronary artery to dilate during exercise performed after -adrenergic receptor blockade (2). Despite the elevated shear stress caused by the decrease in large epicardial coronary artery diameter (CD) and the rise in CBF, NO production is apparently insufficient to reverse the dominant -adrenergic constriction. NO production may, however, limit the extent of -adrenergic constriction. In this connection, several earlier studies (8, 19, 26) conducted in vitro have reported that the loss of basal NO formation caused by endothelial denudation or pharmacological blockade of the NO synthase (NOS) leads to augmented -adrenergic constriction. In vivo, shear stress-dependent NO formation in the microcirculation has been shown to act as a braking mechanism against immoderate coronary constriction (17, 22).

These data led us to examine the extent to which NO limits -adrenergic constriction of large epicardial coronary arteries during exercise performed after -adrenergic receptor blockade.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Instrumentation. Under general anesthesia with pentobarbital sodium (30 mg/kg iv) and under sterile conditions, nine mongrel dogs (31 ± 1 kg) underwent a left thoracotomy at the fifth intercostal space under artificial ventilation. The pericardium was widely incised parallel to the phrenic nerve. A Tygon catheter (Norton Plastics and Synthetic Division; Akron, OH) was implanted in the thoracic aorta to measure arterial pressure (model 800, Bentley Trantec; Irvine, CA). Mean arterial pressure (MAP) was obtained with an active filter with a time constant of 2 s. Through an apical stab wound, a solid-state pressure transducer (model P6.5, Konigsberg Instruments; Pasadena, CA) was inserted in the left ventricular (LV) cavity to record LV pressure (LVP) and to obtain its first derivative over time (LV dP/dt). A catheter was also implanted in the LV cavity to cross-calibrate the miniature pressure gauge and to eliminate any drift of the instrument through repeated calibrations. A cardiotachometer (model 9857, Sensor Medics; Anaheim, CA) triggered by the LVP pulse was used to monitor heart rate (HR). An ultrasonic Doppler blood flow transducer was placed around the circumflex coronary artery, 1 to 2 cm from the bifurcation of the left main coronary artery. Flow velocity was monitored using a 10-MHz-pulsed Doppler flowmeter (13). Mean flow velocity was obtained with an active filter with a time constant of 2 s. At necropsy, the internal circumference of the vessel under the probe was measured to establish the vessel internal cross-sectional area and to calculate a calibration factor (in ml · min¹ · kHz¹). Distal to the Doppler flow probe, a pair of miniature ultrasonic crystals (1 × 2 mm, 7 MHz) was sutured with 6-0 prolene on opposite sides of the left circumflex coronary artery after adequate alignment. CD was monitored by using an ultrasonic sonomicrometer (model 120.2, Triton Technology; San Diego, CA). Mean CD was obtained with an active filter with a time constant of 2 s. A catheter was implanted into the right atrium for systemic drug delivery. The pericardium was loosely approximated, the chest was closed in layers, and the catheters and wires were exteriorized on the back of the animals. Analgesia was provided postoperatively with 0.3 mg im buprenorphine (Temgesic, Reckitt and Colman Pharmaceuticals; Hull, UK). Prophylactic procaine penicillin G (300,000 U im) and benzathine penicillin G (300,000 U im) were administered for 10 days after the surgery.

Protocols. Experiments were initiated 3 to 6 wk after surgery in dogs trained to perform a standardized treadmill exercise protocol consisting of three steps: 3 miles/h (0% grade), 4 miles/h (5% grade), and 6 miles/h (10% grade). Each exercise step lasted 3-5 min to ensure that a steady-state hemodynamic response was reached. After continuously monitoring HR, LVP, LV dP/dt, phasic and MAP, phasic and mean CBF, and mean CD in dogs standing quietly on a treadmill, the exercise protocol was initiated. At least 1 h after the first run, the exercise protocol was performed after 10 mg/kg iv N-nitro-L-arginine methyl ester (L-NAME, no. N-5751, Sigma Chemical; St. Louis, MO) given over 10 min to block endothelial NOS (eNOS) activity and prevent NO formation. Adequacy of NO formation blockade with this dose of L-NAME has been demonstrated earlier by others (21) and confirmed by us in a previous study using an acetylcholine chloride challenge (24). To ensure that adequate blockade of NO formation in large epicardial coronary arteries was achieved after L-NAME, CD responses to exercise after L-NAME were compared with those obtained after 35 mg/kg iv N-nitro-L-arginine (L-NNA, no. N-5501, Sigma Chemical), another arginine analogue that blocks eNOS activity.

1

1

Data analysis. Shear stress (in dyn/cm²) was calculated with mean flow velocity (v) and mean internal vessel radius (r) at the site of CD measurements and blood viscosity () by the formula: shear stress = (4v)/r. Vessel internal radius (r) was calculated from mean CD and a wall-to-lumen ratio of 1/11 under baseline conditions. This ratio was obtained from earlier measurements of CD of circumflex coronary arteries of dogs under resting conditions and wall cross-sectional area obtained after death. Vessel wall cross-sectional area should remain constant, i.e., no change in vessel wall cross-sectional area with changes in vessel radius. Blood flow velocity measured with the Doppler flow probe cannot directly reflect flow velocity at the site of CD measurements because the cross-sectional area of the vessel under the Doppler flow probe is fixed. Therefore, mean blood velocity at the site of CD measurements was calculated as the ratio of mean CBF (mean flow velocity × cross-sectional area under the probe) measured upstream with the Doppler flow probe to internal cross-sectional area at the site of CD measurements. Viscosity (in centipoises) was obtained from hematocrit measured before each experiment. A linear relationship between hematocrit and blood viscosity has been previously reported at shear rates >100 s¹ (5). This relationship was used to calculate blood viscosity:  = [0.20 + (0.11 × hematocrit)]/100.  under control conditions was 3.90 ± 0.14 × 10² poises (range 3.14 to 4.44 × 10² poises). Viscosity has been demonstrated to remain stable in vessels with a radius >300 µm (14) and at shear rates exceeding 100 s¹ (6), which is the case in the present study. Data were read directly from the strip charts under baseline conditions while the dogs were standing quietly on the treadmill and near the end of each step of the graded treadmill exercise protocol, when a steady state was reached.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Baseline hemodynamics. Hemodynamic variables measured under control conditions, during exercise, and after the various interventions are reported in Tables 1-3.

Control exercise. Baseline and exercise-induced hemodynamic responses are reported in Table 1. Exercise performed under control conditions led to substantial increases (P < 0.01) in CBF and CD, as reported in Fig. 1 and Table 1. CD was linearly related to shear stress throughout the exercise (Fig. 2). All other hemodynamic variables during exercise were different from baseline (P < 0.01) and are reported in Table 1.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Means ± SE (n = 8) external coronary artery diameter (CD, top) and coronary blood flow (CBF, bottom) at each step of exercise before and after N-nitro-L-arginine methyl ester (L-NAME) or N-nitro-L-arginine (L-NNA) to block nitric oxide (NO) formation. L-NAME and L-NNA decreased baseline CD and impaired exercise-induced CD increases. L-NAME or L-NNA did not alter CBF. NS, not significant. * Different from control, P < 0.01. Differences between L-NAME and L-NNA are directly reported.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Means ± SE (n = 8) external CD at each step of exercise before and after L-NAME or L-NNA plotted against means ± SE shear stress. Comparisons were made between slopes. Blockade of NO formation with L-NAME or L-NNA caused a downward shift in the relationship and a reduction of slope. * Different from control, P < 0.01. Differences in slopes between L-NAME and L-NNA are directly reported.

Blockade of NO formation. Baseline and exercise-induced hemodynamic responses after L-NAME and L-NNA are reported in Table 1. Overall hemodynamic responses caused by exercise after L-NAME were characterized by higher LVP (P < 0.05) and MAP (P < 0.01) and lower (P < 0.01) HR and LV dP/dt compared with control. Increases in CBF after L-NAME did not differ from control responses. As expected, baseline CD was reduced (P < 0.01) after L-NAME and exercise-induced dilation was blunted (P < 0.01) after the blockade of NO formation, as reported in Fig. 1 and Table 1. Shear stress increased more (P < 0.01) during exercise after L-NAME. The relationship between CD and shear stress was shifted downward, and the slope was reduced (P < 0.01), as reported in Fig. 2.

Blockade of -adrenergic receptors ± blockade of NO formation. Baseline and exercise-induced hemodynamic responses after -adrenergic receptor blockade with propranolol are reported in Table 2. Exercise performed after -adrenergic receptor blockade caused the expected hemodynamic effects characterized by lower (P < 0.01) LV dP/dt and HR. Exercise after propranolol led to smaller (P < 0.05) CBF levels than in control experiments (Fig. 3 and Table 2). Baseline CD was reduced (P < 0.01) after propranolol- and exercise-induced CD responses were characterized by a paradoxical constriction (P < 0.01), as reported in Fig. 3 and Table 2. Shear stress was higher (P < 0.05) during exercise after propranolol than under control conditions. The slope of the relationship between CD and shear stress after propranolol was substantially smaller (P < 0.01) than control (Fig. 4).

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Means ± SE (n = 9) external CD (top) and CBF (bottom) at each step of exercise before and after propranolol and propranolol + L-NAME. Propranolol reversed CD dilation to constriction. L-NAME after propranolol decreased CD throughout exercise. CBF fell during exercise after propranolol and did not differ from control after the further blockade of NO formation with L-NAME. mph, Miles/hour. * Different from control, P < 0.05;  different from control, P < 0.01. Differences between propranolol and propranolol + L-NAME are directly reported.

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Means ± SE (n = 9) external CD at each step of exercise before and after propranolol or propranolol + L-NAME plotted against means ± SE shear stress. Comparisons were made between slopes. Blockade of -adrenergic receptors with propranolol caused a downward shift in the relationship between CD and shear stress and a reduction of slope compared with control. The additional blockade of NO formation with L-NAME caused a further downward shift in the relationship between CD and shear stress without altering the slope. * Different from control, P < 0.01. Differences in slopes between propranolol and propranolol + L-NAME are directly reported.

Blockade of - and -adrenergic receptors ± blockade of NO formation. Baseline and exercise-induced hemodynamic responses after propranolol + phentolamine are reported in Table 3. Overall hemodynamic responses to exercise after propranolol + phentolamine were characterized by decreases (P < 0.01) in LVP and MAP compared with propranolol alone. CBF responses caused by exercise were maintained after combined - and -adrenergic receptor blockade compared with propranolol alone. Baseline CD was increased (P < 0.01) after propranolol + phentolamine (Fig. 5 and Table 3). Propranolol + phentolamine prevented the exercise-induced fall in CD observed after propranolol alone. Shear stress was lower (P < 0.01) during exercise after propranolol + phentolamine than after propranolol alone. The relationship between CD and shear stress was shifted upward after propranolol + phentolamine, but the slope was not altered (Fig. 6).

View larger version (21K): [in this window] [in a new window]   Fig. 5.   Means ± SE (n = 9) external CD (top) and CBF (bottom) at each step of exercise after propranolol, propranolol + phentolamine, and propranolol + phentolamine + L-NAME. After propranolol + phentolamine, baseline CD increased but CD constriction was prevented compared with propranolol alone. The addition of L-NAME decreased CD throughout exercise compared with propranolol + phentolamine. CBF was not altered by the blockade of -adrenergic receptors. After the additional blockade of NO formation with L-NAME, CBF was higher throughout exercise. * Different from propranolol, P < 0.01. Differences between propranolol + phentolamine and propranolol + phentolamine + L-NAME are directly reported.

View larger version (21K): [in this window] [in a new window]   Fig. 6.   Means ± SE (n = 9) external CD at each step of exercise after propranolol, propranolol + phentolamine, and propranolol + phentolamine + L-NAME plotted against means ± SE shear stress. Comparisons were made between slopes. Blockade of -adrenergic receptors with phentolamine caused an upward shift of the relationship between CD and shear stress but failed to alter the slope compared with propranolol alone. The further blockade of NO formation with L-NAME shifted the relationship downward without altering the slope.

Intracoronary adenosine. To examine whether CBF increases during exercise after propranolol were adequate to cause significant flow-dependent dilation, CBF was elevated with intracoronary adenosine (200 or 300 ng · kg¹ · min ¹) in dogs lying quietly on a table after propranolol administration to match exercise-induced CBF responses. In those dogs, CBF increased (P < 0.01) by 29 ± 4 from 45 ± 2 ml/min and CD by 0.25 ± 0.05 from 3.85 ± 0.28 mm during control exercise. Shear stress increased (P < 0.05) by 2.0 ± 0.7 from 6.5 ± 0.7 dyn/cm². Exercise performed after propranolol led to increases (P < 0.01) in CBF by 14 ± 3 from 44 ± 2 ml/min, whereas CD constricted (P < 0.05) by 0.09 ± 0.03 from 3.78 ± 0.28 mm and shear stress increased (P < 0.01) by 3.4 ± 0.8 from 6.9 ± 0.9 dyn/cm². Adenosine given by intracoronary route after -adrenergic receptor blockade in the same animals lying quietly on a table increased (P < 0.01) CBF by 15 ± 3 from 40 ± 3 ml/min, similar to responses during exercise performed after propranolol. Despite adenosine-induced shear stress increases (1.9 ± 0.6 from 7.5 ± 1.1 dyn/cm²) smaller (P < 0.05) than during exercise after propranolol, CD increased substantially (P < 0.01) by 0.09 ± 0.02 from 3.68 ± 0.27 mm (Fig. 7). Adenosine-induced CD increases were abolished by the further administration of L-NAME, in the face of shear stress increases (P < 0.01) by 2.3 ± 0.6 from 9.5 ± 2.4 dyn/cm².

View larger version (15K): [in this window] [in a new window]   Fig. 7.   Means ± SE (n = 6) external CD at baseline and during intracoronary adenosine after propranolol or propranolol + L-NAME plotted against means ± SE shear stress. L-NAME abolished shear stress-induced increases in CD caused by adenosine. * Different from baseline, P < 0.01.

Intracoronary NTG. To examine whether NO-dependent responses were maintained during exercise performed after -adrenergic receptor blockade, intracoronary NTG was administered before exercise while the dogs were lying quietly on a table after propranolol administration and while the dogs ran at 6 miles/h. On the table, NTG increased (P < 0.05) CD by 0.24 ± 0.07 from 3.67 ± 0.31 mm. Changes in CD caused by NTG did not significantly differ during exercise (0.16 ± 0.07 from 3.59 ± 0.31 mm). At peak CD, CBF did not differ from pre-NTG levels.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Our data indicate that NO fails to limit -adrenergic constriction of large epicardial coronary arteries during exercise performed after -adrenergic receptor blockade. The slope of the relationship between CD and shear stress was reduced after propranolol and not further altered by the additional blockade of NO formation during exercise (Fig. 4). For any given increase in shear stress, the fall in CD after propranolol or propranolol + L-NAME was similar. Thus blockade of NO formation after -adrenergic receptor blockade did not lead to greater CD constriction during exercise, as expected if shear stress triggered significant NO production. However, adenosine-induced increases in shear stress, smaller than those caused by exercise after propranolol, were large enough to trigger NO-dependent CD dilation (Fig. 7). In addition, CD dilation to exogenous NO (NTG) was maintained during exercise performed after -adrenergic receptor blockade. Therefore, exercise-induced -adrenergic constriction after -adrenergic receptor blockade was dominant and not limited by shear stress-dependent NO production.

Exercise-induced CD dilation has been reported to be blunted by impairing CBF increases (20), by endothelial denudation (3), or by the systemic administration of an arginine analogue to block the NOS (27). Thus dilation of large epicardial coronary arteries during exercise involves flow-dependent NO production as a primary mechanism. Our data are consistent with these conclusions. The slope of the relationship between CD and shear stress observed during control exercise was dramatically reduced by the blockade of NO formation with L-NAME (Fig. 2). The residual CD dilator response after L-NAME or L-NNA was blunted by propranolol, consistent with the involvement of -adrenergic receptor activation during control exercise.

The present data agree with an earlier report (2) showing that -adrenergic receptor blockade leads to a paradoxical CD constriction during exercise, reversed by -adrenergic receptor blockade. The rise in shear stress during exercise after -adrenergic receptor blockade would be expected to trigger NO formation thereby limiting CD constriction. In that situation, CD constriction should be magnified by the blockade of NO formation. Our analyses revealed a downward shift of the relationships between CD and shear stress after propranolol + L-NAME compared with propranolol alone (Fig. 4). However, the slopes of these relationships were not altered. Therefore, the blockade of NO formation after propranolol had no further effects beyond those on baseline CD. Presumably, the rise in shear stress during exercise after -adrenergic receptor blockade could not elicit a large enough NO production to limit CD constriction. Given that -adrenergic receptor blockade alone leads to constrictor responses, but not the blockade of NO formation, -adrenergic receptor activation most certainly plays a greater role than shear stress-dependent NO formation in antagonizing -adrenergic CD constriction.

Subthreshold increases in shear stress may conceivably explain the limited contribution of NO during exercise performed after -adrenergic receptor blockade. To directly address that issue, we have used intracoronary adenosine infusions to match peak CBF increases observed during exercise after propranolol alone. Previously, we had shown that CD dilation caused by intracoronary adenosine involved a flow-dependent process sensitive to the blockade of NO formation and suppressed by arterial constriction to prevent CBF increases (18). In the present study, adenosine caused smaller increases in shear stress than those observed during exercise after -adrenergic receptor blockade. Despite these limited shear stress increases, adenosine triggered substantial NO-dependent CD dilation, sensitive to L-NAME (Fig. 7). Consequently, the failure of CD to dilate during exercise after -adrenergic receptor blockade cannot be explained by an inadequate rise in shear stress.

Our data differ from those of Canty and Schwartz (7) where flow-dependent CD dilation caused by adenosine infused distally to the site of CD measurements was resistant to the blockade of NO formation. In our hands, L-NAME blunted CD dilation to adenosine consistent with an earlier study (18) where adenosine-induced CD dilation was also prevented by impairing CBF increases. Therefore, adenosine-induced dilation was flow dependent and not a receptor-operated process. The reason for the apparent discrepancy between the effects of blockade of NO formation on adenosine-induced CD dilation may be related to the disproportionate CBF increases used to trigger flow-dependent dilations, i.e., 38% in the present study and ~330% in the study by Canty and Schwartz (7). NO may not be the sole mechanism accounting for CD dilation caused by sustained flow increases, as suggested by Canty and Schwartz (7). An alternate pathway becomes apparent particularly when a sustained elevation of CBF by several folds over baseline levels is created.

Conceivably, NO-dependent effects on vascular smooth muscle cells were impaired during exercise after -adrenergic receptor blockade thereby causing -adrenergic influences to be dominant. To examine that possibility, we have used NTG as a surrogate for NO production to determine whether large epicardial coronary arteries remain responsive to NO during exercise performed after propranolol. NTG caused substantial dilator responses of large coronary arteries while the dogs exercised at 6 miles/h after propranolol administration. Compared with responses obtained in resting animals (after propranolol), changes in CD caused by NTG given during exercise did not statistically differ. Thus the failure of large epicardial coronary arteries to dilate during exercise after propranolol cannot be accounted for by a lack of vascular responsiveness to NO. An alternate mechanism has to be invoked to explain our observations.

The apparent lack of contribution of NO to CD responses during exercise after propranolol may have resulted from augmented -adrenergic influences interfering with flow-dependent NO production. In support of this possibility, a recent human study reported that baroreceptor unloading caused by lower body negative pressure resulted in a major reduction of flow-mediated dilation of the brachial artery (15). More importantly, -adrenergic receptor blockade reversed the impairment of flow-mediated arterial dilation. It is tempting to speculate that conditions leading to impaired flow-mediated dilation such as mental stress (9) may involve a similar process triggered by the elevated peripheral sympathetic drive and -adrenergic constriction.

If -adrenergic receptor activation directly caused a reduction of NO formation after -adrenergic receptor blockade, the slope of the relationship between CD and shear stress should become more positive after propranolol + phentolamine compared with propranolol alone. Conversely, the further blockade of NO formation after propranolol + phentolamine would be expected to reverse this effect. Our data reveal no change in slope of the relationship between CD and shear stress after either propranolol + phentolamine or propranolol + phentolamine + L-NAME compared with propranolol alone (Fig. 6). Thus -adrenergic receptor activation during exercise after propranolol cannot be causally related to the loss of NO activity. A synergistic effect between -adrenergic receptor activation and eNOS activity during exercise most likely accounts for the present observations. However, this inhibitory effect of -adrenergic receptor blockade on flow-dependent dilation may be selectively displayed during exercise because propranolol does not interfere with flow-dependent CD dilation caused by reactive hyperemia (16).

At the level of the microcirculation, convincing evidence has to be provided to support the notion that shear stress-dependent NO formation acts as a braking mechanism against immoderate constrictor responses (22, 23). Endothelin- and vasopressin-induced constriction of small arteries and arterioles were augmented in vivo by the blockade of NO formation (22). In contrast, blockade of NO formation failed to augment endothelin-induced contractions of isolated microvessels occurring against basal NO release in vitro. This led to the conclusion that shear stress was the trigger for NO release that limited constriction in vivo. In the same connection, ₁- and ₂-adrenergic agonists caused greater constriction of arterioles after the blockade of NO formation in vivo (17). Our data concerning -adrenergic constrictor responses of large epicardial coronary arteries during exercise differed. The efficacy of shear stress-induced NO production in antagonizing constrictor responses is apparently less in large coronary arteries than in smaller arterial vessels of the microcirculation where the ability to regulate shear stress can however differ markedly according to the size of the arterial microvessels (23).

In microvessels, -adrenergic dilation during exercise serves as a feedforward mechanism to ensure a better match between myocardial oxygen supply and demand (10, 11). Gorman et al. (11) using a mathematical model to quantitate interstitial norepinephrine levels, estimated that norepinephrine contributed to ~25% of CBF responses to exercise through feedforward -adrenergic dilation. The present data extend this conclusion to large epicardial coronary arteries where -adrenergic receptor activation was essential for causing CD dilation during exercise. Therefore, -adrenergic effects and NO production acted synergistically to dilate large epicardial coronary arteries during exercise. Exercise-induced responses of the microcirculation differ. Whereas -adrenergic dilation directly contributes to CBF increases, NO is not essential for ensuring the match between myocardial supply and demand (1, 4, 24, 25). Therefore, a link between NO production and -adrenergic receptor activation is not apparent at the level of resistance coronary vessels.

In the present study, L-NAME given after propranolol restored CBF to levels close to those achieved under control conditions (Fig. 3). A simple explanation for this observation is the disproportionate rise in LVP in the face of a small reflex bradycardia after L-NAME that would be expected to augment cardiac metabolic demand and CBF. Myocardial oxygen consumption should also be increased at any given level of cardiac work during exercise after L-NAME because of the suppression of the oxygen sparing effect of NO described earlier (4). A differential alteration by L-NAME of the relationship between cardiac work and myocardial oxygen consumption under control conditions and after propranolol could explain the disparate effects of blockade of NO formation on CBF in those conditions.

In conclusion, large epicardial coronary artery dilation during exercise is the result of the concerted action of shear stress-dependent NO formation and -adrenergic receptor activation. Both mechanisms are required to cause CD dilation during exercise. Suppression of -adrenergic receptor activation paradoxically leads to impaired shear stress-dependent NO formation and allowed -adrenergic constrictor effects to become dominant.