OUP user menu

Autonomic control of cardiovascular performance and whole body O2 delivery and utilization in swine during treadmill exercise

René Stubenitsky, Pieter D Verdouw, Dirk J Duncker
DOI: http://dx.doi.org/10.1016/S0008-6363(98)00102-3 459-474 First published online: 1 August 1998

Abstract

Objective: The present study determined the role of the autonomic nervous system (ANS) in the regulation of systemic and pulmonary circulation and of O2 delivery and utilization in swine at rest and during graded treadmill exercise. Methods: Instrumented swine (n=12) were subjected to treadmill exercise (1–5 km/h) under control conditions and in the presence of single and combined β-adrenergic, α-adrenergic and muscarinic (M) receptor blockade. Results: Exercise produced a four-fold increase in body O2 consumption, due to a doubling of both cardiac output and the arterio–mixed-venous O2 content difference. The latter resulted from an increase in O2 extraction, from 45±1% at rest to 74±1% at 5 km/h, as the O2 carrying capacity [haemoglobin concentration (Hb)] increased by only ∼10%. The increase in cardiac output resulted from a doubling of the heart rate and a small (<10%) increase in stroke volume. The mean aortic pressure (MAP) was unchanged, implying a 50% decrease in systemic vascular resistance (P≤0.05). In contrast, exercise had no significant effect on pulmonary vascular resistance. The sympathetic division of the ANS-controlled O2 delivery via β-adrenoceptors (heart rate and contractility) and Hb concentration via α-adrenoceptor-mediated splenic contraction. In addition, the sympathetic division modulated systemic vascular tone via α- and β-adrenoceptors, but also exerted a vasodilator influence on the pulmonary circulation via β-adrenoceptors. The parasympathetic division controlled O2 delivery in part directly (heart rate) and in part indirectly via inhibition of β-adrenoceptor activity (heart rate and contractility), even during heavy exercise. In addition, the parasympathetic division exerted a direct vasodilator influence on the pulmonary, but not on the systemic, circulation. Conclusions: Thus, in swine, in a manner similar to that in humans, both the sympathetic and parasympathetic division of the ANS contribute to cardiovascular homeostasis during exercise up to levels of high intensity.

Keywords
  • Autonomic nervous system
  • Exercise
  • O2 consumption
  • O2 extraction
  • Pulmonary circulation
  • Systemic circulation
  • Swine

Time for primary review 33 days.

1 Introduction

The large animal most frequently used in exercise studies is the dog, despite differences in their response to exercise compared to that of humans. For example, dogs (i) possess a much higher initial exercise capacity than humans, (ii) are capable of elevating the haemoglobin concentration (Hb) and, hence, the O2-carrying capacity of their blood to a greater extent than humans, and (iii) tend to maintain arterial pH during heavy exercise [1–3]. However, swine are now becoming increasingly more popular in studies of exercise; their cardiovascular response to exercise resembles the response in humans more closely. Thus, both swine and humans have a low initial exercise capacity but can be very well trained and show similar Hb increments and pH decreases in response to exercise [1–4]. Despite the increase in the use of swine in exercise studies, there is a paucity of information on cardiovascular control mechanisms in swine during exercise. Consequently, the role of the autonomic nervous system (ANS) in the regulation of the systemic and pulmonary circulation and of O2 transport and utilization in chronically instrumented swine during treadmill exercise were examined in this study. The importance of sympathetic circulatory control was assessed by studying animals in the absence and presence of either non-selective β-adrenoceptor blockade or α-adrenoceptor blockade. Since presynaptic α2-adrenoceptor blockade can result in enhanced catecholamine release from sympathetic nerve endings [5], we also studied the effect of α-adrenoceptor blockade in the presence of β-adrenoceptor blockade. To determine the contribution of parasympathetic circulatory control, animals were also studied in the absence and presence of non-selective muscarinic (M) receptor blockade. Finally, the contribution of modulation of β-adrenergic activity by the parasympathetic system was evaluated by determining the effects of combined β-adrenoceptor and M-receptor blockade.

2 Methods

Twelve crossbred Landrace×Yorkshire swine (six male and six female; three–four months of age and 24±1 kg at the time of surgery; 30±2 kg at the time of study) were used in this study. All experiments were performed in accordance with the “Guiding Principles in the Care and Use of Laboratory Animals”, as approved by the Council of the American Physiological Society and with prior approval of the Animal Care Committee of the Erasmus University Rotterdam. Adaptation of animals to the laboratory conditions started one week prior to surgery and was continued until one week post-surgery.

2.1 Surgical procedures

Full details of all experimental procedures have been published previously [6–9]. Briefly, swine were sedated with ketamine (30 mg/kg, i.m.), anaesthetized with thiopental (10 mg/kg, i.v.), intubated and mechanically ventilated with a mixture of O2 and nitrous oxide (1:2), to which 0.2–1% (v/v) isoflurane was added. Anaesthesia was maintained with midazolam (2 mg/kg+1 mg/kg/h, i.v.) and fentanyl (10 μg/kg/h, i.v.). Under sterile conditions, the chest was opened via the fourth left intercostal space and an 8 French (Fr) fluid-filled polyvinylchloride (PVC) catheter was inserted into the aortic arch, for the measurement of blood pressure and collection of arterial blood samples, and was secured with a purse string suture. After the pericardium was opened, a precalibrated electromagnetic flow probe (Skalar, Delft, Netherlands) was positioned around the ascending aorta for the measurement of cardiac output. A high fidelity pressure transducer (Konigsberg Instruments, Pasadena, CA, USA) was inserted into the left ventricle (LV) via the apical dimple, for recording of LV pressure and its first derivative (LVdP/dt). An 8 Fr PVC catheter was inserted into the LV for calibration of the Konigsberg transducer signal, and into the left atrium via the left atrial appendage for the measurement of local blood pressure. Similar catheters were also inserted into the pulmonary artery for the measurement of blood pressure, administration of drugs and for collection of mixed venous blood samples. Electrical wires and catheters were tunnelled subcutaneously to the back, the chest was closed and the animals were allowed to recover. All electrical wires and catheters were protected with an elastic vest.

2.2 Post-surgical period

On the first two days post-surgery, the animals received daily injections of buprenorphine (0.3 mg i.v.). During the first week after surgery, the animals received intravenous injections of 25 mg/kg amoxicillin and 5 mg/kg gentamycin on a daily basis to prevent infection. Catheters were flushed daily with physiologic saline containing 2000 IU/ml heparin.

2.3 Experimental protocols

Studies were performed two–four weeks after surgery, with the animals exercising on a motor driven treadmill. For each animal, experimental protocols were performed on different days and in random order.

2.3.1 Reproducibility of responses to exercise (n=9)

With swine lying quietly on the treadmill, resting haemodynamic measurements, consisting of LV, left atrial, aortic and pulmonary artery blood pressures and cardiac output were obtained and arterial and mixed venous blood samples were collected. Haemodynamic measurements were repeated and the rectal temperature was measured while the animals were standing on the treadmill. Subsequently, a five-stage exercise protocol was begun (1, 2, 3, 4 and 5 km/h) with each stage lasting 2–3 min. Haemodynamic variables were measured and blood samples were collected during the last 30 s of each exercise stage, when haemodynamics had reached a steady state. At the conclusion of the exercise protocol, the animals were allowed to rest for 90 min. Then, the animals received an intravenous infusion of saline (10 ml), which was infused over 5 min. Five minutes after completion of the infusion, resting measurements were obtained and the five-stage exercise protocol was repeated.

2.3.2 Effects of β-adrenergic receptor blockade (n=12)

Ninety minutes after swine underwent the five stage exercise protocol under control conditions, the animals received intravenous propranolol (0.5 mg/kg dissolved in 10 ml of saline and infused over 5 min). Five minutes after completion of the infusion, resting measurements were obtained and the five-stage exercise protocol was repeated. The dose of propranolol results in >95% inhibition of isoprenaline-induced increases in heart rate and LVdP/dtmax [6], and produces stable haemodynamic alterations over a 30-min period [7]in awake swine.

2.3.3 Effects of α-adrenergic receptor blockade (n=8)

Ninety min after swine underwent the five stage exercise protocol under control conditions, α-adrenergic receptor blockade was produced by intravenous administration of phentolamine (1 mg/kg in 10 ml of saline, infused over 5 min). Five minutes after completion of the infusion, resting measurements were obtained and the exercise protocol was repeated. The dose of phentolamine used results in >95% inhibition of the noradrenaline-induced (0.3 μg/kg, i.a.) increase in carotid vascular resistance [10], and produces stable haemodynamic alterations over a 30-min period [11]in anaesthetized swine.

2.3.4 Effects of α-adrenergic receptor blockade in β-adrenergic receptor blocked swine (n=9)

Swine received intravenous propranolol (0.5 mg/kg in 10 ml of saline, infused over 5 min). Five minutes after completion of the infusion, resting measurements were obtained and the five-stage exercise protocol was performed. After 90 min of rest, propranolol (0.2 mg/kg in 10 ml of saline) was re-administered. Five minutes later, α-adrenergic receptor blockade was produced by infusing phentolamine at a dose of 1 mg/kg (in 10 ml of saline), and 5 min later, the exercise protocol was repeated. We have previously shown that the propranolol dose regimen produces identical responses during two consecutive exercise protocols [9].

2.3.5 Effects of M-receptor blockade and the effects of β-adrenergic receptor blockade in M-receptor-blocked swine (n=12)

Ninety min after swine underwent the five stage exercise protocol under control conditions, M-receptor blockade was produced by the continuous intravenous infusion of atropine (30 μg/kg/min, i.v.). Ten minutes after the start of the infusion, when haemodynamics had reached a new steady state, resting measurements were obtained and the five-stage exercise protocol was repeated. After completion of the exercise protocol, the atropine infusion was stopped and the animals allowed to rest for another 90 min. Then, swine received an intravenous infusion of propranolol (0.5 mg/kg dissolved in 10 ml of saline). After completion of propranolol administration, the infusion of atropine (30 μg/kg/min, iv) was restarted, and 10 min later, the exercise protocol was repeated. The dose of atropine was considered to produce complete vagal blockade, as a doubling of the dose did not produce further increases in heart rate.

2.4 Haemodynamic measurements

Blood pressure in the aorta, left atrium and pulmonary artery were measured using Combitrans® pressure transducers (Braun, Melsungen, Germany), with the reference point at mid-chest level. LV pressure was measured using a Konigsberg micromanometer, which was calibrated using the fluid filled LV catheter; LVdP/dt was obtained via electrical differentiation of the Konigsberg micromanometer LV pressure signal. CO was measured with an electromagnetic flowmeter system (Transflow 601 System, Skalar, Delft, Netherlands).

2.5 Blood gas measurements

Blood specimens were maintained anaerobically in iced heparinized syringes until the conclusion of each exercise trial. Measurements of arterial and mixed venous PO2(mmHg), PCO2 (mmHg) and pH were then performed with a blood gas analyzer (Acid–Base Laboratory Model 505, Radiometer, Copenhagen, Denmark), while arterial and mixed venous O2 saturation (SO2, %) and arterial haemoglobin concentration (Hb, g%) were measured with a haemoximeter (OSM2, Radiometer). From these measurements, arterial and mixed venous blood O2 contents (mmol/l) were computed as (Hbarterial×0.621×O2 saturation)+(0.00131×PO2). Body O2 delivery (DO2) was computed as the product of CO and arterial blood O2 content; body O2 consumption (VO2) was computed as the product of CO and the difference in O2 content between arterial and mixed venous (pulmonary artery) blood. Body O2 extraction was computed as VO2/DO2·100%.

2.6 Data acquisition and analysis

Haemodynamic data were recorded and digitized on-line using an eight channel data-acquisition program ATCODAS (Dataq Instruments, Akron, OH, USA) and stored on a computer for off-line analysis using a program written in MatLab (The Mathworks Inc., MA, USA). A minimum of 15 consecutive beats were used for analysis of the digitized haemodynamic signals. From the individual beats, heart rate (HR), LV peak systolic blood pressure (LVSP), LV pressure at end-diastole (defined as the onset of positive LVdP/dt; LVEDP), mean left atrial- (MLAP), mean aortic- (MAP) and mean pulmonary artery blood (MPAP) pressures, as well as LVdP/dtmax and cardiac output (CO), were determined, while stroke volume (SV=CO/HR), systemic vascular resistance (SVR=MAP/CO) and pulmonary vascular resistance PVR=(MPAP−MLAP)/CO were calculated.

Statistical analysis was performed using three-way (sex, exercise and treatment) analysis of variance for repeated measures. Since no significant differences were observed between males and females in any of the protocols, the data from males and females have been pooled and analyzed together. Subsequently, we performed two-way (exercise and treatment) analysis of variance for repeated measures. When a significant effect of exercise was observed, post-hoc testing was done using Dunnett's test. When a significant effect of treatment was observed, post-hoc testing was done using either a paired t-test or Wilcoxon Signed Rank test, as appropriate. To compare the effects of different treatments that were studied on different days (e.g. the changes produced by phentolamine versus the changes produced by phentolamine in β-adrenoceptor-blocked animals), we performed two-way analysis of variance of the changes produced by the interventions. A P-value of less than or equal to 0.05 was considered to be statistically significant (two-tailed). All data are presented as mean±S.E.M.

2.7 Drugs

Phentolamine (10 mg/ml, Regitine®, Ciba-Geigy, Arnhem, Netherlands) was dissolved in water containing glucose (35 mg/ml) and further diluted in saline to produce a final concentration of 1 mg/kg/10 ml. Propranolol (Sigma-Aldrich, Bornem, Belgium) was dissolved in 30°C saline, to a concentration of 0.5 mg/kg/10 ml. Atropine (Sigma-Aldrich) was dissolved in 30°C saline, to a concentration of 30 μg/kg/ml. Fresh drug solutions were prepared on the day of each experiment.

3 Results

3.1 Exercise responses in male and female swine

Analysis of variance indicated that there were no significant differences between the haemodynamic and blood gas responses to exercise in the six male and six female swine, either during control conditions, or in the presence of adrenergic or muscarinic receptor blockade (not shown). Therefore, data from male and female swine were pooled and analyzed together.

3.2 Reproducibility of responses to exercise

3.2.1 Haemodynamics

During exercise, CO increased from 3.6±0.3 l/min at rest (lying down) to 7.8±0.5 l/min at 5 km/h (P≤0.01), which was principally due to an increase in HR, from 115±4 to 241±3 bpm (P≤0.01), as the SV only increased significantly (∼10%) at lower levels of exercise (Fig. 1). Since preload did not change during exercise (reflected by a maintained LVEDP, not shown) and afterload was elevated [reflected by the increase in LVSP, from 122±5 to 152±5 mmHg (P≤0.01)], the maintained SV was the result of an increase in LV contractility, reflected by the increase in LVdP/dtmax, from 3120±210 to 6500±350 mmHg/s (P≤0.01). MAP decreased slightly when the animals changed position from lying down to standing up, but, during exercise, MAP increased from 91±3 to 98±3 mmHg (P≤0.05). The modest change in MAP in the presence of a doubling of CO implies that the SVR decreased. In contrast, MPAP increased, from 14±1 to 32±2 mmHg (P≤0.01), while the difference between MPAP and MLAP increased virtually in parallel with CO, so that the PVR was not significantly altered. After 90 min of rest, when all haemodynamic variables had returned to baseline resting values, a second period of exercise resulted in nearly identical haemodynamic responses, with the exception of a slightly lower HR and a larger SV during the second exercise period (Fig. 1). Unpublished data from our laboratory indicate that three consecutive exercise trials in four swine did not reveal haemodynamic differences between the three exercise tests (not shown).

Fig. 1

Reproducibility of haemodynamic responses to graded treadmill exercise in chronically instrumented swine. Data points were obtained at rest [lying (0L) and standing (0S)] and during five levels of treadmill exercise (1–5 km/h). LV=left ventricle; PA=pulmonary artery; LA=left atrium. Data are expressed as the mean±S.E.M.; n=9, except for mean LA pressure and pulmonary vascular resistance (n=4); *P≤0.05 vs. rest (lying), †P≤0.05 vs. control at the corresponding time point.

3.2.2 Blood gases and VO2

Exercise resulted in a slight decrease in arterial PCO2, from 43±1 mmHg at rest to 40±1 mmHg at 5 km/h and an increase in pH from 7.44±0.01 to 7.47±0.01 (both P≤0.05) (not shown). Arterial SO2 did not change (96±1% at rest and 96±1% at 5 km/h), but Hb and, hence, arterial O2 content increased by 12±2% (P≤0.05) at the highest level of exercise compared to resting conditions (Fig. 2). Mixed venous PO2 decreased from 43±1 mmHg at rest to 27±1 mmHg at 5 km/h (P≤0.01); similarly, the SO2 and O2 content decreased during exercise, from 53±1% at rest to 25±1% at 5 km/h and from 2.86±0.11 to 1.50±0.12 mmol/l, respectively (both P≤0.01), while PCO2 increased from 54±1 to 59±2 mmHg and pH decreased from 7.39±0.01 to 7.36±0.01 (both P≤0.05). Since both CO and the arterio–venous O2 content difference nearly doubled, whole body VO2 increased four-fold, from 8.2±0.6 to 33.1±2.1 mmol/min (P≤0.01). All variables returned to baseline resting values within 90 min; a second period of exercise resulted in highly reproducible responses.

Fig. 2

Reproducibility of systemic O2 transport responses to treadmill exercise. Cardiac output (left panel), haemoglobin (Hb) concentration (upper right panel), arterial O2 saturation (middle right panel) and mixed venous O2 saturation (lower right panel), plotted as a function of whole body O2 consumption (VO2), are shown. Data points were obtained at rest (lying) and during five levels of exercise before (control) and after administration of 10 ml of saline. Data are expressed as the mean±S.E.M., n=9; *P≤0.05 vs. control.

3.3 Effects of β-adrenergic receptor blockade

3.3.1 Haemodynamics

Under resting conditions, propranolol decreased CO by 10±3%, which was bradycardia-mediated (Fig. 3). SV did not change despite a 30±3% decrease in LVdP/dtmax and unaltered LV systolic pressure and was, therefore, most likely due to the increase in LVEDP, from 8±2 to 13±2 mmHg (P≤0.05). Propranolol had no effect on MAP, implying that SVR had increased, while PVR was not affected. Propranolol blunted the exercise-induced increases in CO, HR, LVdP/dtmax and LV systolic pressure compared to control exercise, while SVR was higher at each level of exercise compared to control conditions. Exercise in the presence of propranolol resulted in an increase in SV that was maintained at higher levels of exercise, probably as a result of the propranolol-induced elevation of LVEDP. Despite the lower CO during exercise after propranolol compared to control exercise, the pressure drop across the pulmonary bed was not different from that of control conditions, implying that exercise in the presence of propranolol resulted in a higher PVR compared to control exercise.

Fig. 3

Effects of β-adrenergic receptor blockade on haemodynamic responses to graded treadmill exercise in chronically instrumented swine. Data points were obtained at rest [lying (0L) and standing (0S)] and during five levels of treadmill exercise (1–5 km/h) before (control) and after administration of propranolol (0.5 mg/kg i.v.). LV=left ventricle; PA=pulmonary artery; LA=left atrium. Data are expressed as the mean±S.E.M.; n=12, except for mean LA pressure and pulmonary vascular resistance (n=8); *P≤0.05 vs. rest (lying), †P≤0.05 vs. control at the corresponding time point.

3.3.2 Blood gases and VO2

Propranolol had no effect on arterial PCO2 and pH at rest (44±1 mmHg and 7.46±0.01, respectively) or during exercise (40±1 mmHg and 7.48±0.01 at 5 km/h, respectively). Similarly, propranolol had no effect on mixed venous PCO2 and pH either at rest (54±1 mmHg and 7.38±0.01, respectively) or during exercise (58±1 mmHg and 7.37±0.01 at 5 km/h, respectively) (not shown). Arterial SO2 and Hb were also not altered, but mixed venous SO2 values were lower at rest and particularly during exercise compared to control conditions (Fig. 4), reflecting an increase in O2 extraction, from 43±1 to 48±2% at rest and from 69±2 to 77±2% at 5 km/h (both P≤0.05). The widening of the arterio–venous O2 content difference could only in part compensate for the lower CO, so that body VO2 was 13–17% lower during exercise at 3–5 km/h compared to that found during control exercise (Fig. 4).

Fig. 4

Effects of β-adrenergic receptor blockade on the systemic O2 transport responses to treadmill exercise. Cardiac output (left panel), haemoglobin (Hb) concentration (upper right panel), arterial O2 saturation (middle right panel) and mixed venous O2 saturation (lower right panel), plotted as a function of whole body O2 consumption (VO2), are shown. Data points were obtained at rest (lying) and during five levels of exercise before (control) and after administration of propranolol (0.5 mg/kg i.v.). Data are expressed as the mean±S.E.M., n=7; *P≤0.05 vs. control.

3.4 Effects of α-adrenergic receptor blockade

3.4.1 Haemodynamics

Under resting conditions phentolamine decreased MAP by 11±3%, which was due to a 30±6% decrease in SVR, as CO increased by 17±6% (all P≤0.05); the latter was exclusively mediated by an increase in HR and was accompanied by an increase in LVdP/dtmax (Fig. 5). During exercise, these effects of phentolamine persisted, with the phentolamine-induced increase in CO being even more pronounced at 5 km/h than at rest. Despite increasing CO at rest and during exercise, phentolamine had no effect on MPAP at rest and even caused a decrease at higher levels of exercise compared to exercise under control conditions, indicating that PVR was slightly lower during exercise in the presence of phentolamine, which reached levels of statistical significance at 3 and 5 km/h.

Fig. 5

Effects of α-adrenergic receptor blockade on haemodynamic responses to graded treadmill exercise in chronically instrumented swine. Data points were obtained at rest [lying (0L) and standing (0S)] and during five levels of treadmill exercise (1–5 km/h) before (control) conditions and after administration of phentolamine (1.0 mg/kg i.v.). LV=left ventricle; PA=pulmonary artery; LA=left atrium. Data are expressed as the mean±S.E.M.; n=8, except for mean LA pressure and pulmonary vascular resistance (n=4); *P≤0.05 vs. rest (lying), †P≤0.05 vs. control at the corresponding time point.

3.4.2 Blood gases and VO2

Phentolamine had no effect on arterial PCO2 and pH at rest (44±2 mmHg and 7.44±0.01, respectively) or during exercise (39±1 mmHg and 7.46±0.01 at 5 km/h, respectively). Similarly, phentolamine had no effect on mixed venous PCO2 and pH either at rest (53±1 mmHg and 7.38±0.01, respectively) or during exercise (55±1 mmHg and 7.37±0.01 at 5 km/h, respectively; not shown). Arterial SO2 and Hb were also not altered, but phentolamine increased mixed venous SO2 at rest and during exercise, compared to control (Fig. 6), reflecting a decrease in O2 extraction, from 44±1 to 39±2% at rest and from 69±2 to 64±2% at 5 km/h (both P≤0.05). In addition, phentolamine abolished the exercise-induced increase in Hb, resulting in a lower arterial O2 content during exercise in the presence of phentolamine (5.30±0.22 mmol/l) compared to control exercise (5.73±0.23 mmol/l, P≤0.05). The resultant narrowing of the arterio–venous O2 content difference was balanced by the increase in CO both at rest and during exercise, resulting in whole body VO2 levels that were not different from those observed during control conditions (Fig. 6).

Fig. 6

Effects of α-adrenergic receptor blockade on the systemic O2 transport responses to treadmill exercise. Cardiac output (left panel), haemoglobin (Hb) concentration (upper right panel), arterial O2 saturation (middle right panel) and mixed venous O2 saturation (lower right panel), plotted as a function of whole body O2 consumption (VO2), are shown. Data points were obtained at rest (lying) and during five levels of exercise before (control) and after administration of phentolamine (1 mg/kg i.v.). Data are expressed as the mean±S.E.M., n=7; *P≤0.05 vs. control.

3.5 Effects of α-adrenergic receptor blockade in β-adrenergic receptor-blocked swine

3.5.1 Haemodynamics

In β-blocked animals, phentolamine caused decreases in MAP, LV systolic pressure, LVEDP and MLAP at rest and during treadmill exercise, while HR and LVdP/dtmax did not change and CO increased only at the highest levels of exercise, compared to exercise in the presence of propranolol alone (Fig. 7). Phentolamine resulted in a lower SVR at rest and during exercise, but did not decrease PVR in the β-blocked animals.

Fig. 7

Effects of α-adrenergic receptor blockade on haemodynamic responses to graded treadmill exercise in chronically instrumented swine in the presence of β-adrenergic receptor blockade. Data points were obtained at rest [lying (0L) and standing (0S)] and during five levels of exercise after administration of propranolol (0.5 mg/kg i.v.) and after propranolol (0.2 mg/kg i.v.) and phentolamine (1 mg/kg i.v.). LV=left ventricle; PA=pulmonary artery; LA=left atrium. Data are expressed as the mean±S.E.M.; n=9, except for mean LA pressure and pulmonary vascular resistance (n=3); *P≤0.05 vs. rest (lying), †P≤0.05 vs. propranolol at the corresponding time point.

3.5.2 Blood gases and VO2

In β-blocked animals, phentolamine had no effect on arterial or mixed venous PCO2 and pH (not shown) or arterial SO2, but increased mixed venous SO2 during exercise, compared to exercise in the presence of propranolol alone (Fig. 8). In addition, phentolamine abolished the exercise-induced increase in Hb, resulting in a lower arterial O2 content during exercise compared to exercise with propranolol alone. The resultant narrowing of the arterio–venous O2 content difference balanced the slight increase in CO, so that, compared to propranolol alone, whole body VO2 was not affected by phentolamine, either at rest or during exercise.

Fig. 8

Effects of α-adrenergic receptor blockade on the systemic O2 transport responses to treadmill exercise in the presence of β-adrenergic receptor blockade. Cardiac output (left panel), haemoglobin (Hb) concentration (upper right panel), arterial O2 saturation (middle right panel) and mixed venous O2 saturation (lower right panel), plotted as a function of whole body O2 consumption (VO2), are shown. Data points were obtained at rest (lying) and during five levels of exercise after administration of propranolol (0.5 mg/kg i.v.) and after administration of propranolol (0.2 mg/kg i.v.) and phentolamine (1 mg/kg i.v.). Data are expressed as the mean±S.E.M., n=7; *P≤0.05 vs. propranolol; ‡P=0.07 vs. propranolol.

3.6 Effects of M-receptor blockade and effects of β-adrenergic receptor blockade in M-receptor-blocked swine

3.6.1 Haemodynamics

Under resting conditions, atropine increased CO, HR, LV systolic pressure and LVdP/dtmax (Fig. 9). The increase in CO (31±6%) was considerably less than the increase in HR (88±7%), as SV decreased by 30±3%; the latter was probably due to the decrease in LVEDP. Atropine had no effect on MAP, implying that SVR had decreased. In contrast, PVR increased, from 2.8±0.5 to 3.4±0.7 mmHg/l/min (P≤0.05). The effects of atropine on all haemodynamic variables gradually waned during exercise, with the exception of HR and LVdP/dtmax, which were significantly elevated at all levels of exercise compared to control. In M-blocked animals, administration of propranolol resulted in decreases in CO, HR, LVdP/dtmax and LV systolic pressure, and increases in LVEDP and SVR at rest and during exercise, but had no effect on MAP. PVR was not affected at rest but was higher during exercise compared to exercise with atropine alone (Fig. 9).

Fig. 9

Effects of M-receptor blockade and the effects of combined β-adrenergic and M-receptor blockade on haemodynamic responses to graded treadmill exercise in chronically instrumented swine. Data points were obtained at rest [lying (0L) and standing (0S)] and during five levels of exercise under control conditions (open circles), during atropine infusion (30 μg/kg/min i.v.; solid circles), and during atropine infusion (30 μg/kg/min i.v.) after administration of propranolol (0.5 mg/kg i.v.; open and solid squares). LV=left ventricle; PA=pulmonary artery; LA=left atrium. Data are expressed as the mean±S.E.M.; n=12, except for mean LA pressure and pulmonary vascular resistance (n=7); *P≤0.05 vs. rest (lying), †P≤0.05 vs. control at the corresponding time point, closed squares indicate P≤0.05 atropine+propranolol vs. atropine at the corresponding time point.

3.6.2 Blood gases and VO2

Atropine had no effect on arterial PCO2 (45±1 mmHg at rest and 41±2 mmHg at 5 km/h), or pH (7.45±0.01 and 7.48±0.01) (neither shown), or SO2 and Hb (Fig. 10), either at rest or during exercise. Similarly, atropine had no effect on mixed venous PCO2 (51±1 mmHg at rest and 56±1 mmHg at 5 km/h), pH (7.40±0.01 and 7.37±0.01), or SO2 (Fig. 10). Since the arterio–venous O2 content difference was also not affected, body VO2 increased in parallel with CO at rest and during exercise at 1 km/h compared to control conditions, but was not different from control at higher exercise levels. Although atropine had no significant effect on the relation between VO2 and SVR (not shown), the relation between VO2 and mixed venous O2 saturation was shifted slightly upwards (P≤0.05), suggesting that M-activation exerted a small vasoconstrictor influence on the systemic bed (Fig. 10).

Fig. 10

Effects of M-receptor blockade and the effects of combined β-adrenergic and M-receptor blockade on the systemic O2 transport responses to treadmill exercise. Cardiac output (left panel), haemoglobin (Hb) concentration (upper right panel), arterial O2 saturation (middle right panel) and mixed venous O2 saturation (lower right panel), plotted as a function of whole body O2 consumption (VO2), are shown. Data points were obtained at rest (lying) and during five levels of exercise under control conditions, during atropine infusion (30 μg/kg/min i.v.), and during atropine infusion (30 μg/kg/min i.v.) after administration of propranolol (0.5 mg/kg i.v.). Data are expressed as the mean±S.E.M., *P≤0.05 vs. control, †P≤0.05 vs. atropine.

In M-blocked animals, administration of propranolol had negligible effects on arterial and mixed venous PCO2 and pH, or arterial SO2 and O2 content, but mixed venous SO2 was lower at rest and during exercise compared to that found with atropine alone, reflecting an increase in O2 extraction (Fig. 10). The widening of the arterio–venous O2 content difference could only in part compensate for the lower CO, so that body VO2 was reduced at rest and during exercise compared to atropine alone (Fig. 10).

4 Discussion

4.1 Responses to treadmill exercise

Exercise resulted in increases in VO2, from 8 mmol O2/min (∼6 ml O2/min/kg) at rest to 32 mmol O2/min (∼25 ml O2/min/kg) at 5 km/h. HR was 238–251 bpm at 5 km/h, which represents 85–90% of the reported maximum HR in swine (275–285 bpm) [12–14]. Since HR is an excellent indication of VO2 as a percentage of maximum O2 uptake [12], our findings suggest that the maximum O2 uptake for these animals would be approximately 36 mmol/min (∼30 ml/min/kg). Maximum VO2 values in swine (weighing approximately 30 kg) have been reported to be in the range of 25–50 ml/min/kg [2, 12, 14–16]. Ciccone et al. [15]reported that the maximally attainable O2 uptake during exercise in miniswine was approximately 25 ml/min/kg. However, in the same study, stimulation with an electric cattle prod (which was not employed in the present study) during exercise resulted in maximum VO2 levels of approximately 40 ml/min/kg, which suggests that these high values of O2 uptake might have been at least in part the result of exogenously imposed mental stress. Nonetheless, values of 30–50 ml/min/kg are similar to those obtained in normal man [17], but are still much lower than those obtained in dogs [18–20], ponies [21]and horses [22], in which maximum VO2 is in the range of 60–130 ml/min/kg. Although other factors, such as body composition, may contribute to these interspecies differences in maximum VO2, it would appear that the exercise capacity of sedentary swine, like that of sedentary man, is comparatively moderate [15, 23]. Thus, similar values for maximum VO2 have been reported for domestic swine and the leaner miniswine [2, 12, 14–16]. In addition, swine, like humans, can be easily trained and show a dramatic improvement in exercise capacity during training [4, 14, 16, 23], whereas dogs show a more modest increase in exercise capacity during exercise training [23, 24].

Arterial O2 tension and saturation are unaltered during submaximal and maximal exercise in normal humans [17], and in chronically instrumented dogs [2, 25]or horses [26]. In contrast, 7–10 mmHg decreases in arterial O2 tension have been reported in chronically instrumented swine [2, 27], although this is not an ubiquitous finding [16]. An explanation for a decrease in arterial O2 tension in swine is not readily found. For example, the thoracotomy procedure, necessary for instrumentation, is an unlikely explanation, since, in the aforementioned studies, ponies and dogs were also instrumented. That swine have a much lower O2 uptake capacity than dogs or ponies is also an unlikely explanation, as, in humans, a decrease in arterial PO2 is sometimes noted in trained individuals with a markedly elevated O2 uptake capacity, rather than in sedentary individuals [17]. In the present study, we did not observe significant changes in arterial PO2 in going from rest to exercise, although in some of the protocols, a trend towards a decrease was noted (P>0.10). Importantly, the changes in arterial PO2 were never sufficient to produce detectable changes in arterial O2 saturation. The arterial PCO2 decreased slightly during exercise, which was associated with a small increase in arterial pH at higher levels of exercise. Similarly, previous studies in swine also reported a decrease in arterial PCO2, although arterial pH did not change at exercise levels up to 50–70% of maximum O2 uptake [2, 16, 27]; exercise at maximum O2 uptake was associated with a significant decrease in pH [2, 16, 27], which was probably due to anaerobic metabolism, as reflected by increased lactate production [16, 27]. These findings suggest that, in the present study, swine were performing aerobic exercise.

The exercise-induced four-fold increase in VO2 was the result of a 60–65% increase in O2 extraction, in combination with a 130–150% increase in O2 delivery, the latter being met by a 120–140% increase in CO and a 10–15% increase in Hb. In exercising dogs [28–30], horses [26, 31, 32]and sheep [33], O2 delivery is facilitated by an increase in the O2-carrying capacity of arterial blood due to an increase in Hb (by 20–50%). The latter is a result of splenic contraction, which expresses erythrocyte-rich blood into the general circulation [26, 30, 31]. In swine and humans, an increase in Hb also occurs but is usually smaller (<15%) [2, 16, 27, 34, 35]and may, in part, result from extravasation of plasma during exercise resulting from increased capillary plasma filtration, thereby leading to haemoconcentration [36]. Under resting conditions, mixed venous SO2 is considerably higher in ponies (74% [26]) and dogs (64% [29]and 72% [37]) than in swine (55%, present study). Consequently, ponies and dogs can nearly triple their fractional O2 extraction during heavy exercise compared to resting conditions [26, 29], whereas it can at most double in swine.

The exercise-induced increase in CO was principally the result of an increase in HR, as SV increased only at the lower three levels of exercise. Previous studies in swine have documented either no change [1, 38]or a small (<15%) increase in SV [2, 12]during exercise. SV also increases slightly in ponies [21, 39]or dogs [28, 40], but it can increase by up to 100% in man during exercise in the upright position [41]. However, when humans exercise in the supine position, which is more comparable to the body position of quadrupeds, SV increases are much less (∼25%) [41]. This is the result of an increase in SV to approximately 80% of the maximum SV when changing from the upright to the supine position, so that only a slightly further increase is possible during subsequent supine exercise [41]. In the present study, SV was maintained despite an increase in afterload (LVSP), which must have been due to the increase in contractility (LVdP/dtmax), as preload (LVEDP) was not altered. The increase in CO did not result in appreciable increases in MAP, indicating that SVR decreased commensurately.

During exercise, MPAP increased in excess of the increase in MLAP. The increase in pressure difference across the pulmonary bed tended to be smaller than the increase in CO, so that PVR tended to decrease (P>0.05). In man, PVR decreases by more than 50% during exercise, but in quadrupeds such as sheep and horses, there is only a 25% decrease in PVR [42, 43]. The present study is the first to report the responses of the pulmonary vascular bed in swine to treadmill exercise. Our findings indicate that, in swine, PVR decreases even less during exercise than in other quadrupeds.

4.2 Sympathetic control of cardiovascular function during exercise

4.2.1 β-Adrenoceptor blockade

β-adrenoceptors are present in the heart where they modulate HR and contractility, and are located on blood vessels, where they mediate vasodilation. In awake resting swine, β-adrenoceptor blockade resulted in lower HR and LVdP/dtmax and higher SVR under resting conditions, confirming previous observations from our laboratory that significant β-adrenergic activity is present in swine in the resting state [9]. This contrasts with studies in awake resting dogs [44, 45]and resting (supine) humans [46, 47], in which β-adrenergic activity can be minimal. The effect of β-adrenoceptor blockade on HR, contractility and CO became progressively greater during exercise. Consequently, the lower CO was no longer compensated for by an increase in O2 extraction, so that whole body VO2 was 13–17% lower compared to that found when animals were undergoing control exercise. Our observations correlate well with data obtained in humans, which indicate that there is impaired exercise capacity during β-adrenoceptor blockade [48]. In the present study, we failed to find evidence for anaerobic metabolism during the higher levels of exercise, however, it is possible that lactate production was only modest, due to the relatively short duration of each exercise stage (2–3 min) and was, therefore, buffered by the blood. It is likely that longer periods of exercise at the highest level of exercise would have resulted in decrements of arterial pH.

Propranolol produced an increase in SVR that could have been caused, at least in part, by unopposed α-adrenergic vasoconstriction. However, comparison of the SVR in the phentolamine-treated swine (22.7±1.0 mmHg/l/min under resting conditions, Fig. 5) and the phentolamine plus propranolol-treated animals (26.2±0.9 mmHg/l/min, Fig. 7), representing a propranolol-induced increase in SVR of 3.5 mmHg/l/min, to the increase in SVR produced by propranolol alone (from 28.0±1.7 to 31.2±2.0 mmHg/l/min, Fig. 3), which was almost the same [3.2 mmHg/l/min, P=nonsignificant (NS)], suggests that the increase in SVR was primarily due to blockade of β-adrenoceptors per se. Interestingly, propranolol increased PVR during exercise, demonstrating for the first time that, in swine, exercise is associated with an increased β-adrenergic vasodilator influence in the pulmonary circulation. The observation of β-adrenergic vasodilation in the pulmonary bed is in agreement with findings in several other species, including the cat [49], dog [50]and sheep [51].

4.2.2 α-Adrenoceptor blockade

Blockade of α-adrenergic receptors influences cardiovascular performance and O2 transport through several mechanisms. First, blockade of prejunctional α2-adrenergic receptors interrupts the negative feedback control of catecholamine release (norepinephrine) [5, 52]. The resultant increase in norepinephrine levels augments cardiac β-adrenergic stimulation, thereby increasing HR and contractility. Second, α-adrenergic blockade can decrease vascular tone by interrupting post-junctional α1- and α2-receptor-mediated vasoconstriction of resistance vessels of inactive skeletal muscle and visceral organs [35, 53]. Conversely, α2-adrenergic receptors on coronary vascular endothelium stimulate the release of nitric oxide, which opposes the direct vasoconstrictor effect [54]. Third, α-adrenergic blockade decreases Hb concentrations by interrupting post-junctional α-mediated splenic contraction [33, 55].

In the present study, phentolamine increased HR and contractility at rest (which were probably reflex-mediated in response to the hypotension) but also during exercise. The latter cannot be explained by activation of the baroreceptor reflex, since, at higher levels of exercise and hence sympathetic activity, systemic hypotension produced by vasodilators like adenosine or dipyridamole do not produce further increases in HR during exercise in swine [14, 16], or ponies [26]. The phentolamine-induced increases in HR and contractility were β-adrenoceptor mediated, as they were abolished by propranolol. Phentolamine decreased SVR during exercise but also at rest, indicating that, even under resting conditions, significant systemic α-adrenergic tone is present in resting swine, a finding that is consistent with observations in resting dogs [56, 57]and man [46]. The phentolamine-induced decrease in SVR may have been, at least in part, mediated by increased β-adrenergic vasodilation (possibly secondary to increased catecholamine release), as pretreatment with propranolol blunted the phentolamine-induced decrease at rest by approximately 50% (P≤0.05). Phentolamine decreased PVR during exercise at 3 and 5 km/h. The slight vasodilation produced by phentolamine was blocked by propranolol, indicating that it was probably the result of increased β-adrenoceptor-mediated vasodilation. The lack of effect of α-adrenoceptor blockade on PVR in the presence of propranolol may be due to simultaneous blockade by phentolamine of vascular smooth muscle α1-receptors and endothelial α2-adrenoceptors that mediate pulmonary vasoconstriction and vasodilation, respectively [58]. Future studies using selective α-receptors are needed to determine the importance of the α-receptor subtype in the control of pulmonary vasomotion in swine during exercise.

In this study, we observed that the exercise-induced increase in Hb was abolished when swine were pretreated with the nonselective α-adrenergic receptor blocker, phentolamine, which is in agreement with earlier observations in exercising sheep [33]. If the increase in Hb had been due to extravasation of plasma, phentolamine would have been expected to further increase Hb, as arteriolar vasodilation will lead to increased intracapillary filtration pressures [36]. Sato et al. [55]demonstrated in dogs that catecholamine-induced splenic contraction is elicited via α-adrenergic receptor stimulation. Thus, the increase in Hb during exercise in swine is most likely due to α-adrenoceptor-mediated splenic contraction.

4.3 Parasympathetic control of cardiovascular function during exercise

M-receptors are located in the heart, particularly in the atria, where they modulate HR directly and indirectly by influencing the release of catecholamines from the sympathetic nerve endings. Muscarine receptors are also present in blood vessels, where they can produce vasoconstriction via direct vascular smooth muscle contraction or vasodilation via endothelium-dependent mechanisms. Atropine nearly doubled HR in awake resting swine, an effect that was blunted in animals in which the β-adrenoceptors were blocked, indicating that the parasympathetic nervous system controls HR both indirectly, via β-adrenoceptor activity, and directly. These findings also indicate that, in resting swine, the parasympathetic part of the ANS dominates over the sympathetic part, as it does in other large mammalian species, such as the dog [53], horse [39]and man [59], but not in rodent species [60, 61]. M-receptor blockade also markedly increased LVdP/dtmax, but this effect was abolished when β-adrenoceptors were blocked. The observation that combined M-and β-adrenoceptor blockade produced similar decreases in LVdP/dtmax compared to β-adrenoceptor blockade alone suggests that the parasympathetic system did not directly modulate contractility. It is more likely that modulation occurred via alterations in HR (“Treppe” effect [62]) or via inhibition of β-adrenoceptor activity. Interestingly, atropine still resulted in significant increases in HR, even at 5 km/h when HR was 238 bpm (>85% of maximum HR), indicating that vagal withdrawal was not complete, even during heavy exercise. In humans and dogs, the parasympathetic division also exerts an influence on the heart, up to near maximum levels of exercise [63–65]. HR, after β-adrenoceptor blockade alone (176±5 bpm), was not different from that after combined β-adrenoceptor and M-receptor blockade (169±4 bpm), indicating that vagal control during exercise occurred via inhibition of β-adrenergic activity.

An unexpected observation was that atropine increased VO2 under resting conditions by as much as 45%. In a previous study in anaesthetized dogs, an increase (16%) in whole body VO2 after administration of atropine was also found [66]. Although a direct effect on VO2 is thus possible in the present study, the animals appeared to be more alert during infusion of atropine (although they were still lying down), which may have contributed to the increase in VO2. The atropine-induced decrease in SVR was probably caused by the higher VO2 levels, although a small vasodilator influence may have been exerted via an increase in β-adrenergic activity. In contrast, the increase in PVR that followed blockade of M-receptors was not mediated via increased β-adrenoceptor vasodilation. The finding of a direct vasodilator effect of M-receptor activation is consistent with the presence of vasodilator M-receptors on pulmonary vascular endothelium [58].

5 Conclusions

Autonomic control of global LV function, systemic and pulmonary circulation and whole body O2 balance in awake swine at rest and during treadmill exercise are described in this study. The sympathetic division controlled O2 delivery via β-adrenoceptors (HR and contractility) and Hb concentration via α-adrenoceptor-mediated splenic contraction. Also, besides modulating systemic vascular tone via α- and β-adrenoceptors, the sympathetic division exerted a vasodilator influence on the pulmonary circulation via β-adrenoceptors. The parasympathetic division controlled O2 delivery in part directly (HR) and in part indirectly, via inhibition of β-adrenoceptor activity (HR and contractility). In addition, the parasympathetic division exerted a direct vasodilator influence on the pulmonary, but not on the systemic, circulation. Thus, in a manner similar to that in man, in swine, both the sympathetic and parasympathetic division of the ANS contribute to cardiovascular homeostasis at rest and during exercise, even at high intensity.

Acknowledgements

The authors gratefully acknowledge Mr Rob van Bremen for technical assistance. The research of Dr. Duncker has been made possible by a Research Fellowship of the Royal Netherlands Academy of Arts and Sciences.

References

View Abstract