- Email: [email protected]
Effects of graded exercise on bronchial blood flow and airway dimensions in sheep
ARTICLE IN PRESS
Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189 www.elsevier.com/locate/ypupt
Effects of graded exercise on bronchial blood flow and airway dimensions in sheep R. Bishopa, D. McLeoda, S. McIlveena, R. Blakea, R. Guntherb, J. Davisb, L. Talkenb, D. Cotteec, A. Quaila, G. Parsonsd, S. Whitea, a
Discipline of Human Physiology, University of Newcastle, The Hunter Heart–Lung Research Guild, Hunter Medical Research Institute, Callagham, 2308 NSW Australia b Division of Surgery, University of California, Davis, CA, 95616, USA c Discipline of Surgical Science, University of Newcastle, The Hunter Heart–Lung Research Guild, Hunter Medical Research Institute, Callagham, 2308 NSW, Australia d Divison of Pulmonary and Critical Care Medicine, University of California, Davis, CA 95616, USA Received 10 March 2006; accepted 29 March 2006
Abstract Exercise stimulus–response relationships for airway blood supply and dimensions have not been described in mammalian species. These relationships are vital for postulates concerning integrated reflex factors normally controlling the airways and which may underlie the asthma syndromes of exercise. This study defines airways stimulus–response relationships in exercising sheep. Ewes between 35 and 40 kg were instrumented at left thoracotomy under thiopentone/isoflurane general anaesthesia. Pulsed Doppler ultrasonic transducers were mounted on the bronchial artery, and transit-time plus single-crystal sonomicrometers on the left main bronchus. These recorded simultaneously and continuously bronchial blood flow (Qbr) and conductance (Cbr), bronchial circumference (Circbr) and wall thickness (Thbr). In Protocol 1 (P1), four sheep ran duplicate 5 min protocols on a horizontal treadmill at continuous step-up-and-down speeds of 1 min duration, namely, 0.8, 1.6, 2.2, 1.6 and 0.8 mph (moderate exercise), followed by 10 min recovery. In P2, four sheep ran duplicate 2 min protocols at constant 4 mph (strenuous exercise), and in P3, one sheep ran duplicate protocols each of 3 min at 2.2, 4.4 and 6 mph (severe exercise). Regression analysis and repeated measures ANOVA were used to assess differences between times, runs and exercise intensity. In P1, airway effects were directly related to graded exercise effort sustained over 5 min. Peak effects occurred at 2.2 mph, except for Thbr. Heart rate and Pa rose (to 156% and 111% of resting, respectively), and Qbr and Cbr fell (to 83% and 75%; both Po0.001). Circbr fell to 96% (P ¼ 0.02), and Thbr rose at low speeds early and late, and thinned at the highest speed. In P2 and P3 for all variables the steady-state effects were systematically greater than for P1 (4.4 mph: Cbr to 43%, Circbr to 93%; 6.6 mph: Cbr to 25%, Circbr to 82%). There was no significant recovery hyperaemia, but there was residual post-exercise bronchoconstriction. The exercise stimulus–response relationships from rest to a maximal 6 mph for sheep airway circumference and its bronchial circulation are inverse and functionally constrictor. r 2006 Elsevier Ltd. All rights reserved. Keywords: Exercise; Awake sheep; Sonomicrometry; Aortic pressure; Bronchial blood flow; Bronchial wall thickness; Bronchial circumference; Bronchoconstriction
1. Introduction Exercise stimulus–response relationships for changes in airway dimensions and its circulation have not been described for mammalian species. This is not surprising, Corresponding author. Tel.: +1 61 2 49 633347; fax: +61 2 49 217406.
E-mail address: [email protected] (S. White). 1094-5539/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.pupt.2006.03.003
because techniques are not available for measurements of rapid and sustained changes in airway blood flow and dimensions during controlled exercise work in either man or animal [1]. Implicit in this domain of cardiopulmonary control is that the underlying mechanisms are multiple, integrated, and involve reflex autonomic activity, the effects of which may well differ for the airway, and the bronchial circulation itself [1–3]. For example, it is well
ARTICLE IN PRESS R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
established using conventional spirometry that during exercise in man the airways dilate [3–5]. It is unclear whether at the same time in normal man the airway vascular bed dilates, or constricts. The bronchial bed may constrict in concert with other non-exercising beds during exercise, because it is well endowed with sympathetic nerves [6] and exercise-induced excitation of the sympathetic neural outflow appears generalized to visceral vascular beds [7,8]. Exercise activation of the adrenal medulla and raised levels of circulating adrenaline must also be taken into account. Adrenaline would enhance the constrictor effect of sympathetic nerves on bronchial vessels [6]. Thus, air flow conductance would be mutually enhanced by the reduction in wall thickness by sympathoadrenal constriction of resistance and capacitance vessels, and by the bronchodilator action of b2-adrenergic inhibition on bronchial smooth muscle. An alternative view, however, might be that in healthy man the findings concerning changes in airway dimensions during exercise are conflicting [9]. The reason for this conflict may reside in the provision of average measures of overall pulmonary function by conventional spirometric measures, which are difficult to apply and given to error during exercise. It has been pointed out that a measured rise or fall in expiratory flow using spirometry could be biased by regional heterogeneities in the response of the upper and lower airways, or of the lung parenchyma [10]. A moderate contraction of a lower airway could, therefore, be masked by minimal or opposite changes in the upper airway. The mechanistic factors proposed for exerciseinduced asthma may also be relevant, because the bronchoconstriction of healthy athletes is coupled to an increase in bronchial blood flow [11,12]. Finally, there are static animal models of exercise using spirometric methods [3] which show that different central command sites can either constrict [13] or dilate [14] the airways. The final outcome in terms of a net integrated effect on the bronchial wall and its circulation in freely exercising man and other mammals remains unclear [1]. There is, therefore, recent interest in new techniques which can accurately target specific airways and their blood flow, and examine mechanisms [1,15]. High-resolution computed tomography has shown clearly in man and animals that lower airways of different size, and with different degrees of methacholine-induced tone, vary in their response to controlled lung inflation [10,15]. The technique, however, cannot be used to quantitate and track continuous changes during exercise studies. In our laboratory for this purpose we have designed an instrument for use in animals trained to run on treadmills [1,2]. Our current model is the sheep. The instrument is a composite of ultrasonic instruments which measures simultaneously and accurately high frequency and long-term changes in airway circumference, airway wall thickness and bronchial blood flow. The postulate tested was that in sheep during exercise there are stimulus-dependent, dilator effects on airway dimensions and on its vascular bed.
179
2. Methods Five female sheep (Merino and Suffolk) weighing 36.4–45 kg were used in the study. The sheep were shorn, vaccinated, quarantined, acclimatized and assessed for their ability to run and train on a treadmill prior to entering the experimental programme. The preparation, recovery, conduct of training, and experimental protocols were approved by the Animal Care and Ethics Committees of the University of California, Davis, and of the University of Newcastle. 2.1. Animal preparation Sheep were anaesthetized with intravenous thiopentone, 15 mg kg1, intubated with a cuffed endotracheal tube and ventilated with oxygen and isoflurane 2–3%. At left thoracotomy, the procedure for assessing and mounting a custom-built, lightweight pulsed Doppler flow probe on the bronchial artery was followed as described previously [16]. Airway dimensions (airway circumference and wall thickness) were measured using the Airway Internal Diameter Assessment (AIDA) sonomicrometer [1]. A pair of transittime sonomicrometer (TTS) crystals for measurement of airway hemi-circumference change was attached as previously described [1] to each side of the intraparenchymal left main bronchus within pockets fashioned by blunt dissection. This site is richly innervated with autonomic and sensory-motor nerves [6], and was exposed by reflecting the investing pleura of the upper lung lobe distally using a gauze swab and friction. We refer to this measurement in the text as ‘‘circumference’’. This is because the measurement is the absolute minimal distance around the internal airway wall traversed by ultrasound between the sonomicrometer crystals [1], which in turn is directly proportional to the circumference (and diameter) of the airway. In addition, a single-crystal sonomicrometer (SCS) was placed about 1 cm away for measurement of bronchial wall thickness as described previously [1]. Catheters were positioned in the aortic arch and right atrium via the left superficial cervical artery and vein, respectively [17,18]. The catheters, AIDA wires and pulsed Doppler probe wires were exteriorized on the back and secured in a backpack for later retrieval [17,18]. 2.2. Bronchial flow and conductance Bronchial blood velocity, flow and conductance was measured using a Triton System 6 Model 200 pulsed Doppler flowmeter (Triton Technology, San Diego, CA, USA) as previously described [16]. Bronchial flow conductance, the reciprocal of flow resistance, was calculated from the formula: Conductance ¼ bronchial blood flow velocity divided by aortic pressure [16,18,20]. The use of aortic pressure rather than pressure gradient in the equation is justified by the results of recent detailed studies by Quail et al. [18]. During and after exercise protocols
ARTICLE IN PRESS 180
R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
similar to those used in the current study, the changes in aortic pressure in autonomically intact and blocked sheep reflected quantitatively the changes in pressure gradient for bronchial flow across the three main downstream drainage sites of the bronchial circulation (e.g. right atrium, pulmonary artery, and pulmonary capillary/pulmonary vein/left atrium [18]. The errors involved in the calculation are minor, and the changes in bronchial blood flow conductance are underestimated by some 2–3% if aortic pressure rather than true pressure gradient is used in the calculation. 2.3. Evaluation of AIDA Theoretical aspects of linearity, frequency response and the application of the instrument package have been described [1]. In the current study, the linearity of the SCS and TTS components was confirmed on the bench. For the SCS we used microscopic-micrometry as a comparative standard for the sonomicrometer signal returned from a changing level of air–saline interface. For the TTS we used digital photography and an ImageJs software programme to track the diameter and calculate the distance between crystals placed on each side of an excised bronchus where the internal diameter was changed by an inflatable balloon. The bronchus formed a nearperfect circle and the internal wall was compressed and thin. Regression of output signal on standard did not depart from linear, the slopes of the relationship were significant and did not differ from 1, and the intercepts did not differ from zero. For SCS, 2SE of estimate is 0.8% of the mean wall thickness of 2.6 mm, for TTS, 2 SE of estimate is 1.6% of mean hemicircumference of 12.2 mm, confirming good accuracy. 2.3.1. Analysis of errors using TTS Measurement of internal hemi-circumference (IHC) using TTS crystals attached to the external bronchial wall firstly assumes that the initial signal received takes the shortest path between the crystals around the internal bronchial lining. This was confirmed earlier [1] on the bench, i.e. when it was found that the estimated distance was precisely the distance between the crystals when the bronchus was cut and pinned flat. A second assumption was that the bronchus forms a perfect circle in crosssection, and a third that the bronchial wall is thin. The second assumption is reasonable because the sampled part of left main bronchus at surgery between the TTS transducers appears semicircular. The third assumption may not be valid. At rest in the awake animal, and depending on the amount of resting vagal tone, the wall has a finite thickness, and the signal will traverse a specific geometric pathway (specific hemi-circumference, SHC), which will be longer than the true IHC, depending on the thickness of the bronchial wall. Moreover, the error will change during mechanical and/or reflex activity such as in exercise, depending on whether the bronchus dilates and
the wall becomes thinner, or constricts, and the wall becomes thicker. The errors were assessed using a symmetrical mathematical model of a bronchial cross-section hemi-circumference consisting of two concentric half-circles, with TTS transducers attached to each end of the outside diameter (OD). The model includes a finite bronchial wall with an IHC, an outside radius (OR), and an inside radius (IR). Wall thickness (WT) was the difference between the OR and IR radii. In this model the SHC equals the distance connecting the transducer attachment point at 9 o’clock with the tangent point of intersection on the IHC, plus the remnant arc of the hemi-circumference to 12 o’clock, multiplied by two: pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi IR 2 2 SHC ¼ 2 OR IR þ arc sin IR . OR The equation suggests that if the OD remained constant and the wall was very thin, then the SHC would equal the IHC. However, should the bronchial wall become progressively thicker with bronchoconstriction, SHC would reduce progressively at a lesser rate than IHC. With near closure of the bronchus, the SHC would equal the OD of the bronchus and IHC would be close to zero. Turning to the present study the approximate average OR (8 mm) observed at surgery for the sheep was entered in the SHC equation together with the IR (5 mm), calculated by subtracting the measured average WT of 2.98 mm (i.e. 3, see Section 3) from the OR. The SHC was calculated as 19.18 mm, in good agreement with the TTSmeasured SHC of 20.571.5 mm. The actual IHC was calculated from 2.p.5/2 ¼ 15.7, and was therefore overestimated at rest by 22%. During maximal exercise SHC fell by 7%, to 17.84 mm, and there was no change in WT. These data permit calculation of the maximal functional changes in bronchial morphometry during exercise in the sheep lung in the current study. First, using an unchanged value for WT of 3 mm and the SHC equation, the value for SHC of 17.84 mm is satisfied in the model by an OR of 7.5 mm, and an IR of 4.5 mm (a 10% reduction). The true IHC can now be calculated using the IR of 4.5 mm, and comes to 14.4 mm. The error overestimate SHC/IHC was 24%. 2.3.2. Limitations of SCS The SCS transducer was applied to the outer wall of the bronchus not knowing whether the sampled luminal tissue–air interface within was anatomically a longitudinal ridge or trough typical of the epithelial surface. Once fixed and focused, however, the epithelial ridge or trough is likely to respond during bronchodilatation and bronchoconstriction in a directionally appropriate and linear way, but not in a quantitatively similar way. Within animals, a given stimulus will produce a consistent WT response, but between animals WT changes may vary considerably in response to a given stimulus, depending on the other structural features of the SCS sample site, e.g. sensing
ARTICLE IN PRESS R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
between, or over cartilage bands, or sensing an epithelial ridge or trough. For this reason no simple relationship between hemi-circumference and WT was expected. The lower airways were suspected of continuous rapid (milliseconds-to-seconds) and longer term (seconds-to-minute) intrinsic, mechanical and reflex changes in morphometry during behaviour at rest, during exercise, and in the postexercise state. It was foreshadowed that such wall motion would be faithfully recorded (Figs. 1–3), and that the records could be analysed either on a beat-by-beat basis, or by using averaging techniques in PowerLab. The AIDA system used in exercising sheep is thus linear and accurate. It is likely the application of the TTS underestimates the degree of bronchoconstriction evoked by exercise. The original data are uncorrected in Section 3, and the significance of the potential errors is presented in Section 4. 2.3.3. Evaluation and comparison of recorded airway variables in open chest anaesthetized sheep, and in awake sheep standing on treadmill A recording of real-time changes in circumference and WT of the bronchus was carried out under isoflurane general anaesthesia in an open-chest sheep (Fig. 1) in the absence of, and during, positive pressure ventilation (tidal volume 7 ml/kg ¼ 250–300 ml; peak inflation pressure 18 cm H2O). The femoral arterial pressure trace lagged the instantaneous records of the TTS and SCS by 0.3 s. The wide arterial pulse-pressure and high heart rate indicates a dilated systemic vascular bed, and the records show aspects of bronchial behaviour hitherto not documented. First, the traces for both the circumference and WT are pulsatile, and respond to the arterial pressure pulse. This is most clearly seen when the ventilator is turned off (Fig. 1A left). The WT tracing appears uncomplicated and is consistently congruent with the arterial pressure pulse. By comparison the circumference trace is less consistent. Second, when the ventilator is turned on (Fig. 1A right, and B), both traces become less congruent as bronchial circumference increases, and WT thins reciprocally, with each inflation. Third, with each positive pressure inflation, the circumference of the left main bronchus increases by 25% of the average deflation value, and the pressure-pulse amplitude may double. Fourth, by contrast, with each inflation, WT thins reciprocally to 92% of the average deflation value, and the arterial pulse amplitude almost halves. Awake sheep show a contrasting record in Fig 2. Inspiration is accomplished by negative intrapleural pressure ventilation at 24 breaths per min, and a right atrial pressure excursion of 5 mmHg about an average of 2.7 mmHg. The sheep was 18-days post-surgery standing on the treadmill prior to, and 7.5 min after, exercise. The sensitivity of the bronchial circumference and thickness measurements to pulse-pressure deformation (recorded in Fig. 1) is still present, but in contrast to the changes observed with positive-pressure ventilation, the amplitude
181
changes in airway dimensions with negative pressure breathing are small. With each inspiration, the circumference of the bronchus increases by 1.0% of the average deflation value, and the pressure-pulse configuration if anything falls slightly. With expiration, the average circumference declines slightly, as expected. Changes in WT with negative pressure ventilation are present but minimal. In trace A the effects of vocalization (bleating) are reproducible, and show that despite a rise in intrathoracic pressure of 20 mmHg, circumference increases by 1.0% at the most. In Trace B, a sigh in the same sheep causes a transient (o2 s) negative intrathoracic pressure excursion which causes circumference to increase over 1 s by 0.5 mm, or 4% above resting, an effect four-fold greater than that of vocalization. This was accompanied by a transient thinning of the bronchial wall. The bronchial blood flow and conductance changes of sighing are also reproduced, as reported in awake dogs and sheep [16,19]. 2.4. Experimental procedures, protocols and statistical analysis During recovery from surgery on 5th–7th day the sheep were placed on the treadmill and trained further to walk unrestrained, apart from loose tethering of their head harness to each side of the treadmill. At this time the postoperative condition of the sheep was assessed and a gentle incremental programme of increasing exercise intensity was implemented under laboratory conditions. If slow surgical recovery was found, i.e. reluctance to walk due to suspected pain or documented raised temperature, the training programme was delayed until the sheep walked willingly at the moderate level of exercise of 2.2 mph. Training and experiments were carried out under laboratory conditions controlled at 2172 1C and 5075% humidity. The treadmill was a modified vibration-free commercial unit programmed for human use with a quiet motor and switching system. The sheep walked on the mat within a separate, clear, Perspex rear-gated frame, set astride the treadmill and supported by the floor, and not touching the treadmill. On the day of the experiment the sheep stood on the treadmill and the AIDA and catheter leads were connected. The aortic and right atrial pressure transducers were referenced to the right atrium of the standing sheep. No formal ventilation measurements were made, however, respiratory rate and an index of changing depth of breathing was recorded by right atrial pressure. The leads were anchored to the sheep’s backpack and to the rim of the walking frame to minimize movement artifact on records. Behaviour, vascular pressures, bronchial artery haemodynamics and bronchial dimensions were recorded continuously and stored on a PowerLab/8SP recording system (ADInstruments, Castle Hill, Australia). At the end of experiments the sheep were terminated and post-mortem scrutiny of the condition of the lungs, other thoracic content, and instrumentation was carried out. The pleura of all sheep under the thoracotomy scar was healed
ARTICLE IN PRESS 182
R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
Fig. 1. Original records from an open chest anaesthetized sheep (40 kg), showing continuous changes in bronchial wall motion associated with arterial pressure pulse and positive-pressure ventilation. Pa ¼ arterial pressure (mmHg); Circbr ¼ bronchial hemi-circumference (mm) recorded by transit-time sonomicrometry; Thbr ¼ bronchial wall thickness recorded by single-crystal sonomicrometry. Arterial pressure-pulse lag 0.3 s. (A) Initial part of record ventilator off; later part positive-pressure ventilator on. (B) During positive-pressure ventilation; Y and X co-ordinates more sensitive, and recording speed greater, than A. Both records show complex interaction effects between pressure-pulse influences and other factors such as myogenic and neural activity regulating circumference and wall thickness, with and without variable stretch effects due to positive-pressure ventilation.
with minimal or no adhesions, and the transducers were firmly fixed by healthy granulation tissue to the bronchus and bronchial artery under the left upper lung lobe.
Three different protocols (P1, P2, P3) were used to examine details of the exercise stimulus-response relationship. The protocols spanned the full range of exercise
ARTICLE IN PRESS R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
183
Fig. 2. Original records of (from top) aortic pressure (Pa, phasic and mean), bronchial blood flow (Qbr, phasic and mean), bronchial wall thickness (Thbr), bronchial wall hemi-circumference (Circbr), right atrial pressure (and respiration; Pra), and heart rate (HR). The record shows the instantaneous effects of vocalization (bleating) and sighing, on changes in bronchial blood flow, circumference and wall thickness of left main bronchus in standing, awake sheep. Panel A: rest before exercise, Panel B: same sheep at rest 7.5 min post-exercise. Further details in methods.
activity, from standing rest to maximum exercise where the heart rate-mean arterial pressure product rose some 2.5 fold, which resembles the effects for maximal exercise in middle aged, untrained men and women. This level of exercise was achieved by a single sheep who achieved 6 mph. None of the other 4 sheep in this study could achieve this running speed for any useful time. In P1, four sheep ran duplicate 5 min protocols on a horizontal treadmill at continuous step-up-and-down speeds of 1 min duration, namely, 0.8, 1.6, 2.2, 1.6 and 0.8 mph (moderate exercise), followed by 10 min recovery. In P2, four sheep ran duplicate 2 min protocols at constant 4.4 mph (strenuous exercise) also followed by 10 min recovery, and in P3, one sheep ran duplicate protocols each of 3 min at 2.2, 4.4 and 6 mph (severe exercise). Regression analysis, repeated measures ANOVA, and 2-way ANOVA were used to analyse differences between times during rest, effects at each level of steady-state exercise, and recovery, and also between replicate runs, i.e. first run versus second [18]. No differences were found, and the data were pooled for analysis. The sample size of the groups was determined by the good reproducibility of the responses, and the significance of mean changes in relation to the hypothesis tested.
3. Results The average resting data for the five sheep were aortic pressure 10073.6 mmHg, heart rate 9172.7 bpm, bronchial blood velocity 13.471.80 cm s1, bronchial blood flow 19.372.60 ml min1, bronchial hemi-circumference 20.571.5 mm, and bronchial thickness 2.9870.137 mm. The continuous effects of moderate exercise at 2.2 mph for 1.5 min are shown in Fig. 3. Aortic pressure rose initially to 133% in the first 6 s, and then fell to a sustained level of approximately 119% of resting in the steady state, rising again transiently as exercise ceased. Aortic pressure then returned to resting levels over the next 2 min. Bronchial flow fell instantly to 60% of resting as exercise commenced, but returned towards and remained below resting levels. As exercise ceased, bronchial flow returned to resting levels. There were also transient and steady-state changes in bronchial circumference and thickness during exercise. At onset there was an immediate fall in circumference (to 86%) and rise in thickness (to 111%). There were oscillations in circumference over 12 s but it remained in the steady state at 86% below resting. During the steady-state, the pulse-excursions of circumference declined, to return in recovery. By contrast, WT after
ARTICLE IN PRESS 184
R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
Fig. 3. Ewe 38 kg.Effects of running on treadmill (between arrows) at moderate exercise of 2.2 mph on aortic pressure (Pa; phasic and mean), heart rate (numbers), bronchial blood flow (Qbr; phasic and mean), bronchial wall thickness (Thbr ; unfiltered signal), and bronchial circumference (Circbr; unfiltered signal).
returning to resting showed a slow continuous decline to 96% during exercise. As exercise ceased the transient fall in circumference and rise in thickness observed at exercise onset was reproduced. In recovery, while bronchial dimensions tended to recover, they remained below resting values for several minutes. The steady-state effects of prolonged graded exercise (P1) in four sheep (0.8, 1.6, 2.2, 1.6 and 0.8 mph) are shown in Fig. 4. There is a direct relationship between exercise intensity and bronchial effects during continuous exercise over 5 min reaching a peak at moderate exercise 2.2 mph. Both heart rate and aortic pressure rose (to 156% and 111%, respectively, Po0.01), and bronchial flow and conductance fell (to 83% and 75%, both Po0.01). Bronchial circumference fell to 96% (P ¼ 0.02), and WT tended to rise at low speeds and return to resting levels at 2.2 mph. During recovery the constriction in airway circumference (Fig. 3) did not reach significance. The effects of strenuous exercise (4.4 mph) in four sheep are shown in Fig. 5, and show greater effects than at 2.2 mph. Heart rate and aortic pressure rose to 182% and 126% (Po0.001); bronchial flow and conductance fell to
73% and 54% (Po0.001). Airway circumference fell to 93% (P ¼ 0.01), and remained significantly depressed throughout recovery (P ¼ 0.05). WT did not change. During exercise the amplitude of the right atrial respiratory excursions increased, there were minimal changes in right atrial pressure, but no panting behaviour occurred postexercise. Finally, the bronchial dimension and vascular effects were variably more intense when one sheep performed severe exercise of 6 mph for 3 min (Fig. 6). In this sheep the regression of peak effects at 2.2, 4.4 and 6 mph showed a linear relationship (Po0.01) for the rise in aortic pressure and fall in bronchial flow, but the bronchial circumference effect tended to plateau at 6 mph. 4. Discussion The hypothesis that the exercise response in sheep consists of dilation of the main lower airways and its circulation is refuted. The new finding is that in exercising sheep the response of both airway circumference and its vascular bed is constrictor.
ARTICLE IN PRESS R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
185
Fig. 4. Mean effects in four sheep of duplicate runs of a continuous exercise protocol of step-up and down speeds, i.e. 0.8, 1.6, 2.2, 1.6, 0.8 mph, and of the recovery pattern over 10 min, on mean aortic pressure (Pa), bronchial blood flow (Qbr), heart rate (HR), bronchial blood flow conductance (Cbr), bronchial wall circumference (Circbr), and bronchial wall thickness (Thbr). The symbol on the left is 2 SEM of the difference between any two time intervals calculated from the error mean square of repeated measures ANOVA. Comparison of the height of the symbol with the difference between any two time intervals gives an approximate estimate of the significance of the change at P ¼ 0.05. The asterisks are actual significance levels of effect assessed by Student–Newman–Keuls multiple comparison post-test. * P ¼ 0.05; ** P ¼ 0.01; *** P ¼ 0.001.
These findings at first sight are surprising, and opposite to those documented for man. In healthy subjects there is a work-dependent increase in upper airway volume due to airway dilatation [4,21]. The response is viewed as a means of minimizing the work of breathing, and the mechanism proposed is vagal withdrawal without a role for sympathetic nerves [1,5]. However, the findings in normal man concerning changes in airway caliber during exercise are conflicting [9]. Traditional pulmonary function tests are open to error particularly during exercise and provide variably average measures of lung function [10]. Therefore, an explanation for the different findings may relate to our target of a specific bronchial segment, rather than the airways in general, and lower airway constrictor effects may be masked by dilator effects in the upper airway. Proof of this must await new techniques, as currently none are available to study specific airway responses in freely ventilating man during exercise. An alternative explanation
is the unknown relationship in resting and exercising sheep between mechanical effects on airways and their interaction with inflation reflexes. The bronchial responses secondary to respiratory behaviours observed in the standing sheep were helpful in this regard. At rest, both eupnoeic breathing, and augmented breaths (sighs), evoke bronchodilatation. These data support the presence and combined action of mechanical events and reflex inhibition by slowly adapting pulmonary stretch receptors of vagal bronchomotor activity [22]. In the post-exercise state, the augmented breath evoked the arterial pressure, heart rate and transient bronchial vasodilator changes reported previously in awake dogs and sheep [16,19]. In addition, there is a rapid increase (over 1 s) in bronchial circumference some four-fold greater than for eupnoeic breathing, and a thinning of the bronchial wall. This is so despite the increase in transmural pressure of 18 mmHg against a background of post-exercise bronchial constriction and
ARTICLE IN PRESS 186
R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
Fig. 5. Mean effects in four sheep of duplicate runs of strenuous exercise at 4.4 mph over 2 min, and of the recovery pattern over 10 min, on mean aortic pressure, bronchovascular conductance, heart rate, and airway wall circumference. Notation as in Fig. 2.
Fig. 6. Mean stimulus–maximal response relationships for a single sheep (40 kg) that ran duplicate trials at 2.2, 4.4 and 6 mph. Variables shown are mean aortic pressure, bronchial blood flow, airway wall internal circumference, and airway wall thickness. Regression data is shown on graph of each variable. Notation as in Fig. 2.
wall thickening, factors which enhance bronchoconstriction rather than bronchodilatation following deep inspiration in asthmatics [23,24]. The findings are common to all sheep studied. These findings together with the satisfactory recovery condition and findings postmortem argue strongly for normal cardio-pulmonary control, and against any
untoward ‘‘retuning’’ of mechanical or reflex mechanisms by surgery and transducer implantation, as an explanation for the differences between exercise-effects on airways in sheep and man. The conclusions are dependent on the satisfactory performance of the high-frequency ultrasonic instruments
ARTICLE IN PRESS R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
used. The linearity and good accuracy of the TTS and SCS [1], was confirmed on the bench. Lung inflation and deflation by both positive and negative pressure breathing caused the bronchial circumference to rise and fall as expected, and WT to respond reciprocally. The amplitude of the excursions also differed considerably between breathing methods (25% above FRC for positive pressure, 1.0% above FRC for negative pressure) as expected, due to the marked differences in inspiratory pressures (positive pressure 18 mmHg; negative pressure about 2–5 mmHg). However, the reduction in bronchial excursion was still apparent when the transmural bronchial pressures were matched during vocalization (+20 mmHg), the sigh (18 mmHg), and were reduced even further during exercise (Fig. 3). Differences in bronchial wall excursions during different experimental states, therefore, may reflect neural constrictor influences on the compliance and other functions of bronchial wall structures. Neural influences are largely blocked by a variety of anaesthetic agents during positive pressure ventilation [25], and are probably enhanced during exercise. An analysis of potential errors in the application of the TTS using mathematical modelling led to several important conclusions. First, the interpretation of the response of the bronchial wall must take into account the simultaneous changes in circumference and WT. In this study the TTS recorded a circumferential shortening but no net change in WT. This means that during maximal exercise there is a true bronchoconstriction and that both external and internal diameters shorten. Moreover, the modelling of data indicates that a false interpretation of bronchoconstriction could occur if circumferential shortening was accompanied by marked wall thinning. However, WT if anything during exercise increases as circumference decreases. This is apparent in Fig. 3, where the continuous record shows transient thickening at the onset and offset of exercise, and longer term, post-exercise residual effects also reflected in the average data. It follows that a false interpretation of exercise-induced bronchoconstriction is unlikely. A point of interest not expressed in the average data is that following the immediate wall thickening coincident with circumferential contraction at the onset of exercise (Fig. 3), there is a slow progressive thinning of the bronchial wall. The effect suggests that secondary to the concomitant bronchovascular constriction there is a rapid fall in microcirculatory blood volume and, based on Starling’s forces, a slower capillary influx of interstitial fluid from the bronchial wall. Second, the modelling analysis quantified potential errors when interpreting the degree of bronchoconstriction, and suggests that the exercise-induced bronchoconstriction is underestimated. At rest, due to a finite WT the error overestimate of internal circumference is 22%. During maximal exercise the average TTS signal gets less and WT does not change, and the error overestimate is similar at 24%. Translated, and taking the errors into account, it is concluded that graded exercise in sheep causes graded
187
bronchoconstriction in the left main bronchus and at maximal exercise the reduction of the airway diameter is approximately 10%, which represents a substantial rise in airflow resistance, if other factors in the Poiseuille relationship remain unchanged. With the onset of exercise the immediate constriction (within seconds), its maintenance, and immediate withdrawal in recovery, suggest that rapidly acting neural controls are responsible. We have not analysed the mechanisms, but the bronchial wall smooth muscle is rich in parasympathetic efferent and sensory-motor innervation, and the bronchovascular bed is similarly endowed but with the addition of sympathetic nerves [6,26,27]. The constrictor effects in the bronchial wall and circulation also appear directly related to the degree of effort. It is therefore possible the degree of autonomic activity is a function of central command excitatory effects involving the mesencephalic and hypothalamic locomotor regions identified by Kaufman and co-workers [13]. The data are notable because during maximal exercise the action of released adrenal medullary hormones might successfully compete with circumferential constriction, through its b2–adrenoceptor bronchodilator action. This would lead to a stimulus-dependent non-linear response for airway circumference. However, this was not the case either during exercise, or in recovery, even at 6 mph. In recovery after strenuous exercise, circumferential constriction remained below resting for 10 min, while bronchial blood flow conductance returned by 2 min to resting levels, suggesting that central de-resetting of the bronchial constrictor mechanism post-exercise is incomplete for some minutes. Mitchell and co-workers demonstrated an uncoupling and recoupling of soft wall tissues from supporting cartilages evoked by electrical field stimulation and acetylcholine [28], but the time-constants of re-coupling in awake animals are unkown. The direct relationship between exercise effort and bronchial constriction was confirmed by a single sheep running at 6 mph, where more intense effects were recorded than when running at 2.2 and 4.4 mph. These findings may not necessarily hold for other mammals. Longer-term exercise beyond 5 min was not studied. Secondary mechanisms could modify the bronchial responses should exercise be prolonged, or carried out under conditions of heat load. Mammals vary in the mechanisms of heat loss and conservation via the airways. We did not measure central temperatures in these studies, but similar protocols showed that pulmonary artery temperatures rose on average from 39.1 to 39.4 1C by the end of exercise [18]. This may not constitute a threshold internal heat load for thermoregulatory inhibition of exercise-induced bronchial wall and vascular constrictor mechanisms, or for induction of thermoregulatory panting behaviour. Awake mammals may also vary in the degree of central command dominance. Upright man has more complex preexercise, exercise, and post-exercise CNS set-points of autonomic excitation than sheep, due to the strong,
ARTICLE IN PRESS 188
R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
postural baroreflex affecting sympatho-adrenal outflows. Circulating adrenaline in exercising upright man may therefore play a more substantial role than in exercising sheep and modulate a prime increase in bronchoconstrictor vagal activity, and reinforce bronchovascular constriction. Awake mammals may differ in the reflex effects of pulmonary stretch afferents when tidal volume increases in exercise. A deep breath at rest [22] in healthy humans and awake sheep causes a bronchodilatation. Such reflexes may be modified by central resetting during and after exercise. Moreover, whether rapid CNS autonomic resetting at the onset of exercise differentially excites and inhibits components of the parasympathetic nerves to the lower airways and heart, as it does the sympathetic nerves to the regional vascular beds and the heart [1,7] has not been examined. These factors together may balance in different mammals the advantages and disadvantages of dilated airways with a larger dead space and smaller resistance to airflow, and constricted airways with a smaller dead space and a raised resistance to airflow. If bronchoconstrictor changes occur in other airways during exercise, the lamina airflow point of the very small airways may shift centrally. It is concluded that in sheep there is a stimulus-dependent constrictor effect on airway dimensions and its vascular bed over a full range of exercise effort of up to 5 min duration. We speculate that central command and its interaction with secondary sensory events evokes an immediate excitation– resetting of parasympathetic pathways to the airway wall, and possibly also to the bronchial circulation. Enhancement of either or both parasympathetic and sympathetic pathways in the bronchial vascular bed cannot be excluded, because recent work shows that both pathways actively determine resting autonomic tone in the bronchial vasculature in awake mammals [16,25,29], and may play a dual role in exercise [2]. It follows that with the onset of exercise in sheep there is a primary selective and sustained (i) inhibition of brainstem vagal neurons to the heart, and (ii) reciprocal excitation of vagal neurons to airway structures. Acknowledgements We wish to acknowledge the help of Triton Technology Inc. for their help and advice on pulsed Doppler flowmetry. The study was supported in part by the Ramaciotti Foundation, and by the Hunter Heart–Lung Research Guild within the Hunter Medical Research Institute, Australia. References [1] White SW, Pitsillides KF, Parsons GH, Hayes SG, Gunther RA, Cottee DB. Coronary–bronchial blood flow and airway dimensions in exercise induced syndromes. Clin Exp Pharmacol Physiol 2001;28: 472–8. [2] McLeod D, Parsons G, Gunther R, McIlveen S, Bishop, White S, et al. Neural factors controlling bronchial blood flow during exercise. FASEB J 2005;19:A175.
[3] Kaufman MP, Forster HV. Reflexes controlling circulatory, ventilatory and airway responses to exercise. In: Rowell L, and Shepherd J, editor. Handbook of Physiology. Exercise: regulation and integration of multiple systems. New York: Oxford University Press; 1996. p. 381-447 [Section12, Chapter 10]. [4] Kagawa J, Kerr HD. Effects of brief graded exercise on specific airway conductance in normal subjects. J Appl Physiol 1970;28: 134–8. [5] Warren JB, Jennings SJ, Clark TJH. Effect of adrenergic and vagal blockade on the normal human airway response to exercise. Clin Sci (London) 1984;66:79–85. [6] Widdicombe JG, Webber SE. Neuroregulation and pharmacology of the tracheobronchial circulation. In: Butler J, editor. The bronchial circulation. New York: Marcel Dekker; 1992. p. 249–89. [7] Rowell LB, O’Leary DS, Kellog DL. Integration of cardiovascular control systems in dynamic exercise. In: Rowell B, Shepherd JT, editor. Handbook of physiology. Exercise: regulation and integration of multiple systems. Bethesda, MD: American Physiology Society; 1996. P.770-838 [Section 12, Chapter 17]. [8] McIlveen SA, Hayes SG, Kaufman MP. Both central command and the exercise pressor reflex reset the carotid sinus baroreflex. Am J Physiol 2001;280:H1454–63. [9] Pichon A, Rouland M, Denjean A, de Bisschop C. Airway tone during exercise in healthy subjects: effects of salbutamol and ipatropium bromide. Int J Sports Med 2005;26:321–6. [10] Brown RH, Mitzner W. Understanding airway pathophysiology with computed tomography. J Appl Physiol 2003;95:854–62. [11] Anderson SD, Daviskas E. Airway drying and exercise induced asthma. In: McFadden ER, editor. Exercise induced asthma: lung biology in health and disease. New York: Marcel Dekker; 1999. p. 77–113. [12] Anderson SD, Holzer K. Exercise-induced asthma: is it the right diagnosis in elite athletes? J Allergy Clin Immunol 2000;106:419–28. [13] Beyaert CA, Hill JM, Lewis BK, Kaufman MP. Effect on airway caliber of stimulation of the hypothalamic locomotor region. J Appl Physiol 1998;84:1388–94. [14] Motekaitis AM, Solomon IC, Kaufman MP. Stimulation of parabrachial nuclei dilates the airways in cats. J Appl Physiol 1994; 76:1712–8. [15] Brown RH, Mitzner W. Functional imaging of airway narrowing. Respir Physiol Neurobiol 2003;137:327–37. [16] McIlveen S, White SW, Parsons G. Autonomic control of bronchial circulation in awake sheep during rest and behaviour. Clin Exp Pharm Physiol 1997;24:940–7. [17] Bishop R, McLeod D, McIlveen S, Blake R, Gunther R, Davis J, et al. Long-term measurement of bronchial vascular resistance in awake sheep. Arch Physiol Biochem 2003;111:313–4. [18] Quail A, Cottee D, McLeod D, Blake R, Bishop R, McIlveen S, et al. Analysis of bronchovascular downstream blood pressure changes in exercising sheep. Arch Physiol Biochem 2003;111:309–13. [19] Porges WL, Hennessy EJ, Quail AW, Cottee DBF, Moore PG, McIlveen SA, et al. Heart-lung interactions: the sigh and autonomic control in the bronchial and coronary circulations. Clin Exp Pharm Physiol 2000;27:1022–7. [20] Quail AW, Cottee DBF, White SW. Limitation of a pulsed Doppler velocimeter for blood flow measurement in small vessels. J Appl Physiol 1993;75:2745–54. [21] Dempsey JA, Forster HV, Ainsworth DM. Regulation of hyperpnea, hyperventilation and respiratory muscle recruitment during exercise. In: Dempsey JA, Pack A, editor. New York: Marcel Dekker; 1994. p. 1065–1134. [22] Widdicombe JG. Overview of neural pathways in allergy and asthma. Pulm Pharmacol Therap 2003;16:23–30. [23] Brown RH, Mitzner W. Delayed distension of contracted airways with lung inflation in vivo. Am J Respir Crit Care Med 2000;162: 2113–6. [24] Brown RH, Mitzner W. Duration of deep inspiration and subsequent airway constriction in vivo. J Asthma 2003;40:119–24.
ARTICLE IN PRESS R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189 [25] Hennessy E, White S, van der Touw T, Quail AW, Porges WL, Glenfield P. Control of resting bronchial hemodynamics in the awake dog. Am J Physiol Heart Circ Physiol 1993;265:H649–H60. [26] Gabella G. Innervation of airway smooth muscle: fine structure. Annu Rev Physiol 1987;49:583–94. [27] Fryer AD. The cholinergic control of the airways. In: Barnes PJ, editor. Autonomic control of the respiratory system. Amsterdam: Harwood Academic Publishers; 1997. p. 59–86.
189
[28] Mitchell HW, Gray PR. Assessment of the dynamic relationship between external diameter and lumen flow in isolated bronchi. Respir Physiol 1999;116:67–76. [29] White S, McIlveen S, Parsons G, Quail A, Cottee D, Gunther R, et al. Neural control of the bronchial circulation. Arch Physiol Biochem 2003;111:303–6.
Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189 www.elsevier.com/locate/ypupt
Effects of graded exercise on bronchial blood flow and airway dimensions in sheep R. Bishopa, D. McLeoda, S. McIlveena, R. Blakea, R. Guntherb, J. Davisb, L. Talkenb, D. Cotteec, A. Quaila, G. Parsonsd, S. Whitea, a
Discipline of Human Physiology, University of Newcastle, The Hunter Heart–Lung Research Guild, Hunter Medical Research Institute, Callagham, 2308 NSW Australia b Division of Surgery, University of California, Davis, CA, 95616, USA c Discipline of Surgical Science, University of Newcastle, The Hunter Heart–Lung Research Guild, Hunter Medical Research Institute, Callagham, 2308 NSW, Australia d Divison of Pulmonary and Critical Care Medicine, University of California, Davis, CA 95616, USA Received 10 March 2006; accepted 29 March 2006
Abstract Exercise stimulus–response relationships for airway blood supply and dimensions have not been described in mammalian species. These relationships are vital for postulates concerning integrated reflex factors normally controlling the airways and which may underlie the asthma syndromes of exercise. This study defines airways stimulus–response relationships in exercising sheep. Ewes between 35 and 40 kg were instrumented at left thoracotomy under thiopentone/isoflurane general anaesthesia. Pulsed Doppler ultrasonic transducers were mounted on the bronchial artery, and transit-time plus single-crystal sonomicrometers on the left main bronchus. These recorded simultaneously and continuously bronchial blood flow (Qbr) and conductance (Cbr), bronchial circumference (Circbr) and wall thickness (Thbr). In Protocol 1 (P1), four sheep ran duplicate 5 min protocols on a horizontal treadmill at continuous step-up-and-down speeds of 1 min duration, namely, 0.8, 1.6, 2.2, 1.6 and 0.8 mph (moderate exercise), followed by 10 min recovery. In P2, four sheep ran duplicate 2 min protocols at constant 4 mph (strenuous exercise), and in P3, one sheep ran duplicate protocols each of 3 min at 2.2, 4.4 and 6 mph (severe exercise). Regression analysis and repeated measures ANOVA were used to assess differences between times, runs and exercise intensity. In P1, airway effects were directly related to graded exercise effort sustained over 5 min. Peak effects occurred at 2.2 mph, except for Thbr. Heart rate and Pa rose (to 156% and 111% of resting, respectively), and Qbr and Cbr fell (to 83% and 75%; both Po0.001). Circbr fell to 96% (P ¼ 0.02), and Thbr rose at low speeds early and late, and thinned at the highest speed. In P2 and P3 for all variables the steady-state effects were systematically greater than for P1 (4.4 mph: Cbr to 43%, Circbr to 93%; 6.6 mph: Cbr to 25%, Circbr to 82%). There was no significant recovery hyperaemia, but there was residual post-exercise bronchoconstriction. The exercise stimulus–response relationships from rest to a maximal 6 mph for sheep airway circumference and its bronchial circulation are inverse and functionally constrictor. r 2006 Elsevier Ltd. All rights reserved. Keywords: Exercise; Awake sheep; Sonomicrometry; Aortic pressure; Bronchial blood flow; Bronchial wall thickness; Bronchial circumference; Bronchoconstriction
1. Introduction Exercise stimulus–response relationships for changes in airway dimensions and its circulation have not been described for mammalian species. This is not surprising, Corresponding author. Tel.: +1 61 2 49 633347; fax: +61 2 49 217406.
E-mail address: [email protected] (S. White). 1094-5539/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.pupt.2006.03.003
because techniques are not available for measurements of rapid and sustained changes in airway blood flow and dimensions during controlled exercise work in either man or animal [1]. Implicit in this domain of cardiopulmonary control is that the underlying mechanisms are multiple, integrated, and involve reflex autonomic activity, the effects of which may well differ for the airway, and the bronchial circulation itself [1–3]. For example, it is well
ARTICLE IN PRESS R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
established using conventional spirometry that during exercise in man the airways dilate [3–5]. It is unclear whether at the same time in normal man the airway vascular bed dilates, or constricts. The bronchial bed may constrict in concert with other non-exercising beds during exercise, because it is well endowed with sympathetic nerves [6] and exercise-induced excitation of the sympathetic neural outflow appears generalized to visceral vascular beds [7,8]. Exercise activation of the adrenal medulla and raised levels of circulating adrenaline must also be taken into account. Adrenaline would enhance the constrictor effect of sympathetic nerves on bronchial vessels [6]. Thus, air flow conductance would be mutually enhanced by the reduction in wall thickness by sympathoadrenal constriction of resistance and capacitance vessels, and by the bronchodilator action of b2-adrenergic inhibition on bronchial smooth muscle. An alternative view, however, might be that in healthy man the findings concerning changes in airway dimensions during exercise are conflicting [9]. The reason for this conflict may reside in the provision of average measures of overall pulmonary function by conventional spirometric measures, which are difficult to apply and given to error during exercise. It has been pointed out that a measured rise or fall in expiratory flow using spirometry could be biased by regional heterogeneities in the response of the upper and lower airways, or of the lung parenchyma [10]. A moderate contraction of a lower airway could, therefore, be masked by minimal or opposite changes in the upper airway. The mechanistic factors proposed for exerciseinduced asthma may also be relevant, because the bronchoconstriction of healthy athletes is coupled to an increase in bronchial blood flow [11,12]. Finally, there are static animal models of exercise using spirometric methods [3] which show that different central command sites can either constrict [13] or dilate [14] the airways. The final outcome in terms of a net integrated effect on the bronchial wall and its circulation in freely exercising man and other mammals remains unclear [1]. There is, therefore, recent interest in new techniques which can accurately target specific airways and their blood flow, and examine mechanisms [1,15]. High-resolution computed tomography has shown clearly in man and animals that lower airways of different size, and with different degrees of methacholine-induced tone, vary in their response to controlled lung inflation [10,15]. The technique, however, cannot be used to quantitate and track continuous changes during exercise studies. In our laboratory for this purpose we have designed an instrument for use in animals trained to run on treadmills [1,2]. Our current model is the sheep. The instrument is a composite of ultrasonic instruments which measures simultaneously and accurately high frequency and long-term changes in airway circumference, airway wall thickness and bronchial blood flow. The postulate tested was that in sheep during exercise there are stimulus-dependent, dilator effects on airway dimensions and on its vascular bed.
179
2. Methods Five female sheep (Merino and Suffolk) weighing 36.4–45 kg were used in the study. The sheep were shorn, vaccinated, quarantined, acclimatized and assessed for their ability to run and train on a treadmill prior to entering the experimental programme. The preparation, recovery, conduct of training, and experimental protocols were approved by the Animal Care and Ethics Committees of the University of California, Davis, and of the University of Newcastle. 2.1. Animal preparation Sheep were anaesthetized with intravenous thiopentone, 15 mg kg1, intubated with a cuffed endotracheal tube and ventilated with oxygen and isoflurane 2–3%. At left thoracotomy, the procedure for assessing and mounting a custom-built, lightweight pulsed Doppler flow probe on the bronchial artery was followed as described previously [16]. Airway dimensions (airway circumference and wall thickness) were measured using the Airway Internal Diameter Assessment (AIDA) sonomicrometer [1]. A pair of transittime sonomicrometer (TTS) crystals for measurement of airway hemi-circumference change was attached as previously described [1] to each side of the intraparenchymal left main bronchus within pockets fashioned by blunt dissection. This site is richly innervated with autonomic and sensory-motor nerves [6], and was exposed by reflecting the investing pleura of the upper lung lobe distally using a gauze swab and friction. We refer to this measurement in the text as ‘‘circumference’’. This is because the measurement is the absolute minimal distance around the internal airway wall traversed by ultrasound between the sonomicrometer crystals [1], which in turn is directly proportional to the circumference (and diameter) of the airway. In addition, a single-crystal sonomicrometer (SCS) was placed about 1 cm away for measurement of bronchial wall thickness as described previously [1]. Catheters were positioned in the aortic arch and right atrium via the left superficial cervical artery and vein, respectively [17,18]. The catheters, AIDA wires and pulsed Doppler probe wires were exteriorized on the back and secured in a backpack for later retrieval [17,18]. 2.2. Bronchial flow and conductance Bronchial blood velocity, flow and conductance was measured using a Triton System 6 Model 200 pulsed Doppler flowmeter (Triton Technology, San Diego, CA, USA) as previously described [16]. Bronchial flow conductance, the reciprocal of flow resistance, was calculated from the formula: Conductance ¼ bronchial blood flow velocity divided by aortic pressure [16,18,20]. The use of aortic pressure rather than pressure gradient in the equation is justified by the results of recent detailed studies by Quail et al. [18]. During and after exercise protocols
ARTICLE IN PRESS 180
R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
similar to those used in the current study, the changes in aortic pressure in autonomically intact and blocked sheep reflected quantitatively the changes in pressure gradient for bronchial flow across the three main downstream drainage sites of the bronchial circulation (e.g. right atrium, pulmonary artery, and pulmonary capillary/pulmonary vein/left atrium [18]. The errors involved in the calculation are minor, and the changes in bronchial blood flow conductance are underestimated by some 2–3% if aortic pressure rather than true pressure gradient is used in the calculation. 2.3. Evaluation of AIDA Theoretical aspects of linearity, frequency response and the application of the instrument package have been described [1]. In the current study, the linearity of the SCS and TTS components was confirmed on the bench. For the SCS we used microscopic-micrometry as a comparative standard for the sonomicrometer signal returned from a changing level of air–saline interface. For the TTS we used digital photography and an ImageJs software programme to track the diameter and calculate the distance between crystals placed on each side of an excised bronchus where the internal diameter was changed by an inflatable balloon. The bronchus formed a nearperfect circle and the internal wall was compressed and thin. Regression of output signal on standard did not depart from linear, the slopes of the relationship were significant and did not differ from 1, and the intercepts did not differ from zero. For SCS, 2SE of estimate is 0.8% of the mean wall thickness of 2.6 mm, for TTS, 2 SE of estimate is 1.6% of mean hemicircumference of 12.2 mm, confirming good accuracy. 2.3.1. Analysis of errors using TTS Measurement of internal hemi-circumference (IHC) using TTS crystals attached to the external bronchial wall firstly assumes that the initial signal received takes the shortest path between the crystals around the internal bronchial lining. This was confirmed earlier [1] on the bench, i.e. when it was found that the estimated distance was precisely the distance between the crystals when the bronchus was cut and pinned flat. A second assumption was that the bronchus forms a perfect circle in crosssection, and a third that the bronchial wall is thin. The second assumption is reasonable because the sampled part of left main bronchus at surgery between the TTS transducers appears semicircular. The third assumption may not be valid. At rest in the awake animal, and depending on the amount of resting vagal tone, the wall has a finite thickness, and the signal will traverse a specific geometric pathway (specific hemi-circumference, SHC), which will be longer than the true IHC, depending on the thickness of the bronchial wall. Moreover, the error will change during mechanical and/or reflex activity such as in exercise, depending on whether the bronchus dilates and
the wall becomes thinner, or constricts, and the wall becomes thicker. The errors were assessed using a symmetrical mathematical model of a bronchial cross-section hemi-circumference consisting of two concentric half-circles, with TTS transducers attached to each end of the outside diameter (OD). The model includes a finite bronchial wall with an IHC, an outside radius (OR), and an inside radius (IR). Wall thickness (WT) was the difference between the OR and IR radii. In this model the SHC equals the distance connecting the transducer attachment point at 9 o’clock with the tangent point of intersection on the IHC, plus the remnant arc of the hemi-circumference to 12 o’clock, multiplied by two: pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi IR 2 2 SHC ¼ 2 OR IR þ arc sin IR . OR The equation suggests that if the OD remained constant and the wall was very thin, then the SHC would equal the IHC. However, should the bronchial wall become progressively thicker with bronchoconstriction, SHC would reduce progressively at a lesser rate than IHC. With near closure of the bronchus, the SHC would equal the OD of the bronchus and IHC would be close to zero. Turning to the present study the approximate average OR (8 mm) observed at surgery for the sheep was entered in the SHC equation together with the IR (5 mm), calculated by subtracting the measured average WT of 2.98 mm (i.e. 3, see Section 3) from the OR. The SHC was calculated as 19.18 mm, in good agreement with the TTSmeasured SHC of 20.571.5 mm. The actual IHC was calculated from 2.p.5/2 ¼ 15.7, and was therefore overestimated at rest by 22%. During maximal exercise SHC fell by 7%, to 17.84 mm, and there was no change in WT. These data permit calculation of the maximal functional changes in bronchial morphometry during exercise in the sheep lung in the current study. First, using an unchanged value for WT of 3 mm and the SHC equation, the value for SHC of 17.84 mm is satisfied in the model by an OR of 7.5 mm, and an IR of 4.5 mm (a 10% reduction). The true IHC can now be calculated using the IR of 4.5 mm, and comes to 14.4 mm. The error overestimate SHC/IHC was 24%. 2.3.2. Limitations of SCS The SCS transducer was applied to the outer wall of the bronchus not knowing whether the sampled luminal tissue–air interface within was anatomically a longitudinal ridge or trough typical of the epithelial surface. Once fixed and focused, however, the epithelial ridge or trough is likely to respond during bronchodilatation and bronchoconstriction in a directionally appropriate and linear way, but not in a quantitatively similar way. Within animals, a given stimulus will produce a consistent WT response, but between animals WT changes may vary considerably in response to a given stimulus, depending on the other structural features of the SCS sample site, e.g. sensing
ARTICLE IN PRESS R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
between, or over cartilage bands, or sensing an epithelial ridge or trough. For this reason no simple relationship between hemi-circumference and WT was expected. The lower airways were suspected of continuous rapid (milliseconds-to-seconds) and longer term (seconds-to-minute) intrinsic, mechanical and reflex changes in morphometry during behaviour at rest, during exercise, and in the postexercise state. It was foreshadowed that such wall motion would be faithfully recorded (Figs. 1–3), and that the records could be analysed either on a beat-by-beat basis, or by using averaging techniques in PowerLab. The AIDA system used in exercising sheep is thus linear and accurate. It is likely the application of the TTS underestimates the degree of bronchoconstriction evoked by exercise. The original data are uncorrected in Section 3, and the significance of the potential errors is presented in Section 4. 2.3.3. Evaluation and comparison of recorded airway variables in open chest anaesthetized sheep, and in awake sheep standing on treadmill A recording of real-time changes in circumference and WT of the bronchus was carried out under isoflurane general anaesthesia in an open-chest sheep (Fig. 1) in the absence of, and during, positive pressure ventilation (tidal volume 7 ml/kg ¼ 250–300 ml; peak inflation pressure 18 cm H2O). The femoral arterial pressure trace lagged the instantaneous records of the TTS and SCS by 0.3 s. The wide arterial pulse-pressure and high heart rate indicates a dilated systemic vascular bed, and the records show aspects of bronchial behaviour hitherto not documented. First, the traces for both the circumference and WT are pulsatile, and respond to the arterial pressure pulse. This is most clearly seen when the ventilator is turned off (Fig. 1A left). The WT tracing appears uncomplicated and is consistently congruent with the arterial pressure pulse. By comparison the circumference trace is less consistent. Second, when the ventilator is turned on (Fig. 1A right, and B), both traces become less congruent as bronchial circumference increases, and WT thins reciprocally, with each inflation. Third, with each positive pressure inflation, the circumference of the left main bronchus increases by 25% of the average deflation value, and the pressure-pulse amplitude may double. Fourth, by contrast, with each inflation, WT thins reciprocally to 92% of the average deflation value, and the arterial pulse amplitude almost halves. Awake sheep show a contrasting record in Fig 2. Inspiration is accomplished by negative intrapleural pressure ventilation at 24 breaths per min, and a right atrial pressure excursion of 5 mmHg about an average of 2.7 mmHg. The sheep was 18-days post-surgery standing on the treadmill prior to, and 7.5 min after, exercise. The sensitivity of the bronchial circumference and thickness measurements to pulse-pressure deformation (recorded in Fig. 1) is still present, but in contrast to the changes observed with positive-pressure ventilation, the amplitude
181
changes in airway dimensions with negative pressure breathing are small. With each inspiration, the circumference of the bronchus increases by 1.0% of the average deflation value, and the pressure-pulse configuration if anything falls slightly. With expiration, the average circumference declines slightly, as expected. Changes in WT with negative pressure ventilation are present but minimal. In trace A the effects of vocalization (bleating) are reproducible, and show that despite a rise in intrathoracic pressure of 20 mmHg, circumference increases by 1.0% at the most. In Trace B, a sigh in the same sheep causes a transient (o2 s) negative intrathoracic pressure excursion which causes circumference to increase over 1 s by 0.5 mm, or 4% above resting, an effect four-fold greater than that of vocalization. This was accompanied by a transient thinning of the bronchial wall. The bronchial blood flow and conductance changes of sighing are also reproduced, as reported in awake dogs and sheep [16,19]. 2.4. Experimental procedures, protocols and statistical analysis During recovery from surgery on 5th–7th day the sheep were placed on the treadmill and trained further to walk unrestrained, apart from loose tethering of their head harness to each side of the treadmill. At this time the postoperative condition of the sheep was assessed and a gentle incremental programme of increasing exercise intensity was implemented under laboratory conditions. If slow surgical recovery was found, i.e. reluctance to walk due to suspected pain or documented raised temperature, the training programme was delayed until the sheep walked willingly at the moderate level of exercise of 2.2 mph. Training and experiments were carried out under laboratory conditions controlled at 2172 1C and 5075% humidity. The treadmill was a modified vibration-free commercial unit programmed for human use with a quiet motor and switching system. The sheep walked on the mat within a separate, clear, Perspex rear-gated frame, set astride the treadmill and supported by the floor, and not touching the treadmill. On the day of the experiment the sheep stood on the treadmill and the AIDA and catheter leads were connected. The aortic and right atrial pressure transducers were referenced to the right atrium of the standing sheep. No formal ventilation measurements were made, however, respiratory rate and an index of changing depth of breathing was recorded by right atrial pressure. The leads were anchored to the sheep’s backpack and to the rim of the walking frame to minimize movement artifact on records. Behaviour, vascular pressures, bronchial artery haemodynamics and bronchial dimensions were recorded continuously and stored on a PowerLab/8SP recording system (ADInstruments, Castle Hill, Australia). At the end of experiments the sheep were terminated and post-mortem scrutiny of the condition of the lungs, other thoracic content, and instrumentation was carried out. The pleura of all sheep under the thoracotomy scar was healed
ARTICLE IN PRESS 182
R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
Fig. 1. Original records from an open chest anaesthetized sheep (40 kg), showing continuous changes in bronchial wall motion associated with arterial pressure pulse and positive-pressure ventilation. Pa ¼ arterial pressure (mmHg); Circbr ¼ bronchial hemi-circumference (mm) recorded by transit-time sonomicrometry; Thbr ¼ bronchial wall thickness recorded by single-crystal sonomicrometry. Arterial pressure-pulse lag 0.3 s. (A) Initial part of record ventilator off; later part positive-pressure ventilator on. (B) During positive-pressure ventilation; Y and X co-ordinates more sensitive, and recording speed greater, than A. Both records show complex interaction effects between pressure-pulse influences and other factors such as myogenic and neural activity regulating circumference and wall thickness, with and without variable stretch effects due to positive-pressure ventilation.
with minimal or no adhesions, and the transducers were firmly fixed by healthy granulation tissue to the bronchus and bronchial artery under the left upper lung lobe.
Three different protocols (P1, P2, P3) were used to examine details of the exercise stimulus-response relationship. The protocols spanned the full range of exercise
ARTICLE IN PRESS R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
183
Fig. 2. Original records of (from top) aortic pressure (Pa, phasic and mean), bronchial blood flow (Qbr, phasic and mean), bronchial wall thickness (Thbr), bronchial wall hemi-circumference (Circbr), right atrial pressure (and respiration; Pra), and heart rate (HR). The record shows the instantaneous effects of vocalization (bleating) and sighing, on changes in bronchial blood flow, circumference and wall thickness of left main bronchus in standing, awake sheep. Panel A: rest before exercise, Panel B: same sheep at rest 7.5 min post-exercise. Further details in methods.
activity, from standing rest to maximum exercise where the heart rate-mean arterial pressure product rose some 2.5 fold, which resembles the effects for maximal exercise in middle aged, untrained men and women. This level of exercise was achieved by a single sheep who achieved 6 mph. None of the other 4 sheep in this study could achieve this running speed for any useful time. In P1, four sheep ran duplicate 5 min protocols on a horizontal treadmill at continuous step-up-and-down speeds of 1 min duration, namely, 0.8, 1.6, 2.2, 1.6 and 0.8 mph (moderate exercise), followed by 10 min recovery. In P2, four sheep ran duplicate 2 min protocols at constant 4.4 mph (strenuous exercise) also followed by 10 min recovery, and in P3, one sheep ran duplicate protocols each of 3 min at 2.2, 4.4 and 6 mph (severe exercise). Regression analysis, repeated measures ANOVA, and 2-way ANOVA were used to analyse differences between times during rest, effects at each level of steady-state exercise, and recovery, and also between replicate runs, i.e. first run versus second [18]. No differences were found, and the data were pooled for analysis. The sample size of the groups was determined by the good reproducibility of the responses, and the significance of mean changes in relation to the hypothesis tested.
3. Results The average resting data for the five sheep were aortic pressure 10073.6 mmHg, heart rate 9172.7 bpm, bronchial blood velocity 13.471.80 cm s1, bronchial blood flow 19.372.60 ml min1, bronchial hemi-circumference 20.571.5 mm, and bronchial thickness 2.9870.137 mm. The continuous effects of moderate exercise at 2.2 mph for 1.5 min are shown in Fig. 3. Aortic pressure rose initially to 133% in the first 6 s, and then fell to a sustained level of approximately 119% of resting in the steady state, rising again transiently as exercise ceased. Aortic pressure then returned to resting levels over the next 2 min. Bronchial flow fell instantly to 60% of resting as exercise commenced, but returned towards and remained below resting levels. As exercise ceased, bronchial flow returned to resting levels. There were also transient and steady-state changes in bronchial circumference and thickness during exercise. At onset there was an immediate fall in circumference (to 86%) and rise in thickness (to 111%). There were oscillations in circumference over 12 s but it remained in the steady state at 86% below resting. During the steady-state, the pulse-excursions of circumference declined, to return in recovery. By contrast, WT after
ARTICLE IN PRESS 184
R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
Fig. 3. Ewe 38 kg.Effects of running on treadmill (between arrows) at moderate exercise of 2.2 mph on aortic pressure (Pa; phasic and mean), heart rate (numbers), bronchial blood flow (Qbr; phasic and mean), bronchial wall thickness (Thbr ; unfiltered signal), and bronchial circumference (Circbr; unfiltered signal).
returning to resting showed a slow continuous decline to 96% during exercise. As exercise ceased the transient fall in circumference and rise in thickness observed at exercise onset was reproduced. In recovery, while bronchial dimensions tended to recover, they remained below resting values for several minutes. The steady-state effects of prolonged graded exercise (P1) in four sheep (0.8, 1.6, 2.2, 1.6 and 0.8 mph) are shown in Fig. 4. There is a direct relationship between exercise intensity and bronchial effects during continuous exercise over 5 min reaching a peak at moderate exercise 2.2 mph. Both heart rate and aortic pressure rose (to 156% and 111%, respectively, Po0.01), and bronchial flow and conductance fell (to 83% and 75%, both Po0.01). Bronchial circumference fell to 96% (P ¼ 0.02), and WT tended to rise at low speeds and return to resting levels at 2.2 mph. During recovery the constriction in airway circumference (Fig. 3) did not reach significance. The effects of strenuous exercise (4.4 mph) in four sheep are shown in Fig. 5, and show greater effects than at 2.2 mph. Heart rate and aortic pressure rose to 182% and 126% (Po0.001); bronchial flow and conductance fell to
73% and 54% (Po0.001). Airway circumference fell to 93% (P ¼ 0.01), and remained significantly depressed throughout recovery (P ¼ 0.05). WT did not change. During exercise the amplitude of the right atrial respiratory excursions increased, there were minimal changes in right atrial pressure, but no panting behaviour occurred postexercise. Finally, the bronchial dimension and vascular effects were variably more intense when one sheep performed severe exercise of 6 mph for 3 min (Fig. 6). In this sheep the regression of peak effects at 2.2, 4.4 and 6 mph showed a linear relationship (Po0.01) for the rise in aortic pressure and fall in bronchial flow, but the bronchial circumference effect tended to plateau at 6 mph. 4. Discussion The hypothesis that the exercise response in sheep consists of dilation of the main lower airways and its circulation is refuted. The new finding is that in exercising sheep the response of both airway circumference and its vascular bed is constrictor.
ARTICLE IN PRESS R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
185
Fig. 4. Mean effects in four sheep of duplicate runs of a continuous exercise protocol of step-up and down speeds, i.e. 0.8, 1.6, 2.2, 1.6, 0.8 mph, and of the recovery pattern over 10 min, on mean aortic pressure (Pa), bronchial blood flow (Qbr), heart rate (HR), bronchial blood flow conductance (Cbr), bronchial wall circumference (Circbr), and bronchial wall thickness (Thbr). The symbol on the left is 2 SEM of the difference between any two time intervals calculated from the error mean square of repeated measures ANOVA. Comparison of the height of the symbol with the difference between any two time intervals gives an approximate estimate of the significance of the change at P ¼ 0.05. The asterisks are actual significance levels of effect assessed by Student–Newman–Keuls multiple comparison post-test. * P ¼ 0.05; ** P ¼ 0.01; *** P ¼ 0.001.
These findings at first sight are surprising, and opposite to those documented for man. In healthy subjects there is a work-dependent increase in upper airway volume due to airway dilatation [4,21]. The response is viewed as a means of minimizing the work of breathing, and the mechanism proposed is vagal withdrawal without a role for sympathetic nerves [1,5]. However, the findings in normal man concerning changes in airway caliber during exercise are conflicting [9]. Traditional pulmonary function tests are open to error particularly during exercise and provide variably average measures of lung function [10]. Therefore, an explanation for the different findings may relate to our target of a specific bronchial segment, rather than the airways in general, and lower airway constrictor effects may be masked by dilator effects in the upper airway. Proof of this must await new techniques, as currently none are available to study specific airway responses in freely ventilating man during exercise. An alternative explanation
is the unknown relationship in resting and exercising sheep between mechanical effects on airways and their interaction with inflation reflexes. The bronchial responses secondary to respiratory behaviours observed in the standing sheep were helpful in this regard. At rest, both eupnoeic breathing, and augmented breaths (sighs), evoke bronchodilatation. These data support the presence and combined action of mechanical events and reflex inhibition by slowly adapting pulmonary stretch receptors of vagal bronchomotor activity [22]. In the post-exercise state, the augmented breath evoked the arterial pressure, heart rate and transient bronchial vasodilator changes reported previously in awake dogs and sheep [16,19]. In addition, there is a rapid increase (over 1 s) in bronchial circumference some four-fold greater than for eupnoeic breathing, and a thinning of the bronchial wall. This is so despite the increase in transmural pressure of 18 mmHg against a background of post-exercise bronchial constriction and
ARTICLE IN PRESS 186
R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
Fig. 5. Mean effects in four sheep of duplicate runs of strenuous exercise at 4.4 mph over 2 min, and of the recovery pattern over 10 min, on mean aortic pressure, bronchovascular conductance, heart rate, and airway wall circumference. Notation as in Fig. 2.
Fig. 6. Mean stimulus–maximal response relationships for a single sheep (40 kg) that ran duplicate trials at 2.2, 4.4 and 6 mph. Variables shown are mean aortic pressure, bronchial blood flow, airway wall internal circumference, and airway wall thickness. Regression data is shown on graph of each variable. Notation as in Fig. 2.
wall thickening, factors which enhance bronchoconstriction rather than bronchodilatation following deep inspiration in asthmatics [23,24]. The findings are common to all sheep studied. These findings together with the satisfactory recovery condition and findings postmortem argue strongly for normal cardio-pulmonary control, and against any
untoward ‘‘retuning’’ of mechanical or reflex mechanisms by surgery and transducer implantation, as an explanation for the differences between exercise-effects on airways in sheep and man. The conclusions are dependent on the satisfactory performance of the high-frequency ultrasonic instruments
ARTICLE IN PRESS R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
used. The linearity and good accuracy of the TTS and SCS [1], was confirmed on the bench. Lung inflation and deflation by both positive and negative pressure breathing caused the bronchial circumference to rise and fall as expected, and WT to respond reciprocally. The amplitude of the excursions also differed considerably between breathing methods (25% above FRC for positive pressure, 1.0% above FRC for negative pressure) as expected, due to the marked differences in inspiratory pressures (positive pressure 18 mmHg; negative pressure about 2–5 mmHg). However, the reduction in bronchial excursion was still apparent when the transmural bronchial pressures were matched during vocalization (+20 mmHg), the sigh (18 mmHg), and were reduced even further during exercise (Fig. 3). Differences in bronchial wall excursions during different experimental states, therefore, may reflect neural constrictor influences on the compliance and other functions of bronchial wall structures. Neural influences are largely blocked by a variety of anaesthetic agents during positive pressure ventilation [25], and are probably enhanced during exercise. An analysis of potential errors in the application of the TTS using mathematical modelling led to several important conclusions. First, the interpretation of the response of the bronchial wall must take into account the simultaneous changes in circumference and WT. In this study the TTS recorded a circumferential shortening but no net change in WT. This means that during maximal exercise there is a true bronchoconstriction and that both external and internal diameters shorten. Moreover, the modelling of data indicates that a false interpretation of bronchoconstriction could occur if circumferential shortening was accompanied by marked wall thinning. However, WT if anything during exercise increases as circumference decreases. This is apparent in Fig. 3, where the continuous record shows transient thickening at the onset and offset of exercise, and longer term, post-exercise residual effects also reflected in the average data. It follows that a false interpretation of exercise-induced bronchoconstriction is unlikely. A point of interest not expressed in the average data is that following the immediate wall thickening coincident with circumferential contraction at the onset of exercise (Fig. 3), there is a slow progressive thinning of the bronchial wall. The effect suggests that secondary to the concomitant bronchovascular constriction there is a rapid fall in microcirculatory blood volume and, based on Starling’s forces, a slower capillary influx of interstitial fluid from the bronchial wall. Second, the modelling analysis quantified potential errors when interpreting the degree of bronchoconstriction, and suggests that the exercise-induced bronchoconstriction is underestimated. At rest, due to a finite WT the error overestimate of internal circumference is 22%. During maximal exercise the average TTS signal gets less and WT does not change, and the error overestimate is similar at 24%. Translated, and taking the errors into account, it is concluded that graded exercise in sheep causes graded
187
bronchoconstriction in the left main bronchus and at maximal exercise the reduction of the airway diameter is approximately 10%, which represents a substantial rise in airflow resistance, if other factors in the Poiseuille relationship remain unchanged. With the onset of exercise the immediate constriction (within seconds), its maintenance, and immediate withdrawal in recovery, suggest that rapidly acting neural controls are responsible. We have not analysed the mechanisms, but the bronchial wall smooth muscle is rich in parasympathetic efferent and sensory-motor innervation, and the bronchovascular bed is similarly endowed but with the addition of sympathetic nerves [6,26,27]. The constrictor effects in the bronchial wall and circulation also appear directly related to the degree of effort. It is therefore possible the degree of autonomic activity is a function of central command excitatory effects involving the mesencephalic and hypothalamic locomotor regions identified by Kaufman and co-workers [13]. The data are notable because during maximal exercise the action of released adrenal medullary hormones might successfully compete with circumferential constriction, through its b2–adrenoceptor bronchodilator action. This would lead to a stimulus-dependent non-linear response for airway circumference. However, this was not the case either during exercise, or in recovery, even at 6 mph. In recovery after strenuous exercise, circumferential constriction remained below resting for 10 min, while bronchial blood flow conductance returned by 2 min to resting levels, suggesting that central de-resetting of the bronchial constrictor mechanism post-exercise is incomplete for some minutes. Mitchell and co-workers demonstrated an uncoupling and recoupling of soft wall tissues from supporting cartilages evoked by electrical field stimulation and acetylcholine [28], but the time-constants of re-coupling in awake animals are unkown. The direct relationship between exercise effort and bronchial constriction was confirmed by a single sheep running at 6 mph, where more intense effects were recorded than when running at 2.2 and 4.4 mph. These findings may not necessarily hold for other mammals. Longer-term exercise beyond 5 min was not studied. Secondary mechanisms could modify the bronchial responses should exercise be prolonged, or carried out under conditions of heat load. Mammals vary in the mechanisms of heat loss and conservation via the airways. We did not measure central temperatures in these studies, but similar protocols showed that pulmonary artery temperatures rose on average from 39.1 to 39.4 1C by the end of exercise [18]. This may not constitute a threshold internal heat load for thermoregulatory inhibition of exercise-induced bronchial wall and vascular constrictor mechanisms, or for induction of thermoregulatory panting behaviour. Awake mammals may also vary in the degree of central command dominance. Upright man has more complex preexercise, exercise, and post-exercise CNS set-points of autonomic excitation than sheep, due to the strong,
ARTICLE IN PRESS 188
R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189
postural baroreflex affecting sympatho-adrenal outflows. Circulating adrenaline in exercising upright man may therefore play a more substantial role than in exercising sheep and modulate a prime increase in bronchoconstrictor vagal activity, and reinforce bronchovascular constriction. Awake mammals may differ in the reflex effects of pulmonary stretch afferents when tidal volume increases in exercise. A deep breath at rest [22] in healthy humans and awake sheep causes a bronchodilatation. Such reflexes may be modified by central resetting during and after exercise. Moreover, whether rapid CNS autonomic resetting at the onset of exercise differentially excites and inhibits components of the parasympathetic nerves to the lower airways and heart, as it does the sympathetic nerves to the regional vascular beds and the heart [1,7] has not been examined. These factors together may balance in different mammals the advantages and disadvantages of dilated airways with a larger dead space and smaller resistance to airflow, and constricted airways with a smaller dead space and a raised resistance to airflow. If bronchoconstrictor changes occur in other airways during exercise, the lamina airflow point of the very small airways may shift centrally. It is concluded that in sheep there is a stimulus-dependent constrictor effect on airway dimensions and its vascular bed over a full range of exercise effort of up to 5 min duration. We speculate that central command and its interaction with secondary sensory events evokes an immediate excitation– resetting of parasympathetic pathways to the airway wall, and possibly also to the bronchial circulation. Enhancement of either or both parasympathetic and sympathetic pathways in the bronchial vascular bed cannot be excluded, because recent work shows that both pathways actively determine resting autonomic tone in the bronchial vasculature in awake mammals [16,25,29], and may play a dual role in exercise [2]. It follows that with the onset of exercise in sheep there is a primary selective and sustained (i) inhibition of brainstem vagal neurons to the heart, and (ii) reciprocal excitation of vagal neurons to airway structures. Acknowledgements We wish to acknowledge the help of Triton Technology Inc. for their help and advice on pulsed Doppler flowmetry. The study was supported in part by the Ramaciotti Foundation, and by the Hunter Heart–Lung Research Guild within the Hunter Medical Research Institute, Australia. References [1] White SW, Pitsillides KF, Parsons GH, Hayes SG, Gunther RA, Cottee DB. Coronary–bronchial blood flow and airway dimensions in exercise induced syndromes. Clin Exp Pharmacol Physiol 2001;28: 472–8. [2] McLeod D, Parsons G, Gunther R, McIlveen S, Bishop, White S, et al. Neural factors controlling bronchial blood flow during exercise. FASEB J 2005;19:A175.
[3] Kaufman MP, Forster HV. Reflexes controlling circulatory, ventilatory and airway responses to exercise. In: Rowell L, and Shepherd J, editor. Handbook of Physiology. Exercise: regulation and integration of multiple systems. New York: Oxford University Press; 1996. p. 381-447 [Section12, Chapter 10]. [4] Kagawa J, Kerr HD. Effects of brief graded exercise on specific airway conductance in normal subjects. J Appl Physiol 1970;28: 134–8. [5] Warren JB, Jennings SJ, Clark TJH. Effect of adrenergic and vagal blockade on the normal human airway response to exercise. Clin Sci (London) 1984;66:79–85. [6] Widdicombe JG, Webber SE. Neuroregulation and pharmacology of the tracheobronchial circulation. In: Butler J, editor. The bronchial circulation. New York: Marcel Dekker; 1992. p. 249–89. [7] Rowell LB, O’Leary DS, Kellog DL. Integration of cardiovascular control systems in dynamic exercise. In: Rowell B, Shepherd JT, editor. Handbook of physiology. Exercise: regulation and integration of multiple systems. Bethesda, MD: American Physiology Society; 1996. P.770-838 [Section 12, Chapter 17]. [8] McIlveen SA, Hayes SG, Kaufman MP. Both central command and the exercise pressor reflex reset the carotid sinus baroreflex. Am J Physiol 2001;280:H1454–63. [9] Pichon A, Rouland M, Denjean A, de Bisschop C. Airway tone during exercise in healthy subjects: effects of salbutamol and ipatropium bromide. Int J Sports Med 2005;26:321–6. [10] Brown RH, Mitzner W. Understanding airway pathophysiology with computed tomography. J Appl Physiol 2003;95:854–62. [11] Anderson SD, Daviskas E. Airway drying and exercise induced asthma. In: McFadden ER, editor. Exercise induced asthma: lung biology in health and disease. New York: Marcel Dekker; 1999. p. 77–113. [12] Anderson SD, Holzer K. Exercise-induced asthma: is it the right diagnosis in elite athletes? J Allergy Clin Immunol 2000;106:419–28. [13] Beyaert CA, Hill JM, Lewis BK, Kaufman MP. Effect on airway caliber of stimulation of the hypothalamic locomotor region. J Appl Physiol 1998;84:1388–94. [14] Motekaitis AM, Solomon IC, Kaufman MP. Stimulation of parabrachial nuclei dilates the airways in cats. J Appl Physiol 1994; 76:1712–8. [15] Brown RH, Mitzner W. Functional imaging of airway narrowing. Respir Physiol Neurobiol 2003;137:327–37. [16] McIlveen S, White SW, Parsons G. Autonomic control of bronchial circulation in awake sheep during rest and behaviour. Clin Exp Pharm Physiol 1997;24:940–7. [17] Bishop R, McLeod D, McIlveen S, Blake R, Gunther R, Davis J, et al. Long-term measurement of bronchial vascular resistance in awake sheep. Arch Physiol Biochem 2003;111:313–4. [18] Quail A, Cottee D, McLeod D, Blake R, Bishop R, McIlveen S, et al. Analysis of bronchovascular downstream blood pressure changes in exercising sheep. Arch Physiol Biochem 2003;111:309–13. [19] Porges WL, Hennessy EJ, Quail AW, Cottee DBF, Moore PG, McIlveen SA, et al. Heart-lung interactions: the sigh and autonomic control in the bronchial and coronary circulations. Clin Exp Pharm Physiol 2000;27:1022–7. [20] Quail AW, Cottee DBF, White SW. Limitation of a pulsed Doppler velocimeter for blood flow measurement in small vessels. J Appl Physiol 1993;75:2745–54. [21] Dempsey JA, Forster HV, Ainsworth DM. Regulation of hyperpnea, hyperventilation and respiratory muscle recruitment during exercise. In: Dempsey JA, Pack A, editor. New York: Marcel Dekker; 1994. p. 1065–1134. [22] Widdicombe JG. Overview of neural pathways in allergy and asthma. Pulm Pharmacol Therap 2003;16:23–30. [23] Brown RH, Mitzner W. Delayed distension of contracted airways with lung inflation in vivo. Am J Respir Crit Care Med 2000;162: 2113–6. [24] Brown RH, Mitzner W. Duration of deep inspiration and subsequent airway constriction in vivo. J Asthma 2003;40:119–24.
ARTICLE IN PRESS R. Bishop et al. / Pulmonary Pharmacology & Therapeutics 20 (2007) 178–189 [25] Hennessy E, White S, van der Touw T, Quail AW, Porges WL, Glenfield P. Control of resting bronchial hemodynamics in the awake dog. Am J Physiol Heart Circ Physiol 1993;265:H649–H60. [26] Gabella G. Innervation of airway smooth muscle: fine structure. Annu Rev Physiol 1987;49:583–94. [27] Fryer AD. The cholinergic control of the airways. In: Barnes PJ, editor. Autonomic control of the respiratory system. Amsterdam: Harwood Academic Publishers; 1997. p. 59–86.
189
[28] Mitchell HW, Gray PR. Assessment of the dynamic relationship between external diameter and lumen flow in isolated bronchi. Respir Physiol 1999;116:67–76. [29] White S, McIlveen S, Parsons G, Quail A, Cottee D, Gunther R, et al. Neural control of the bronchial circulation. Arch Physiol Biochem 2003;111:303–6.