reflex. Splitting the medulla abolished this prolonged response while preserving the ... phasic vagal activity i,e., pulmonary stretch receptor (PSR) activity evo-.
ACTA NEUROBIOL. EXP. 1984, 44: 249-262
RESPONSE OF RESPIRATORY MOTONEURONS TO RELEASE FROM VAGAL INHIBITION IN THE RABBIT K. BuDzIRSKA., K. GEOWICKI and J. R. ROMANIUK Laboratory of
Neurophysiolagy, Medical Research Centre, Polish Academy Science, Dworkowa 3, 00-784 Warsa,w, Poland .
of
, K e y words: control of breathing, vagal reflex
Abstract. Long-lasting effects of vagal input on phrenic (Phr) and external intercostal (EI) motoneuronal output were studied on 16 anesthetized rabbits breathing spontaneously or paralysed and ventilated by a phrenic-nerve-driven pump. Withholding of ventilation by tracheal occlusion or by switching off the servorespirator maintained for seven breaths evoked a progressive increase of Phr and EI from breath to breath. This effect was more evident'in animals ventilated by servorespirator. The higher was the gain of the pump (volume-to-phrenic signal ratio) before the maneuvre, the bigger was the rate of increase of the tidal phrenic amplitude from breath to breath 'at all C02 levels tested. Vagotomy strongly depressed or eliminated this effect. We conclude that with intact vagus nerve the increase of respiratory motoneuronal output was only partialy due to the gradual increase in chemical drive when ventilation was stopped. The character of the response indicates the existence of a long-lasting component of the Breuer-Hering reflex. Splitting the medulla abolished this prolonged response while preserving the inspiratory vagal inhibition indicates that the neuronal pathways crossing the midline of the medulla are important for the effect. INTRODUCTION
Current models of the neural control olf breathing concern the description of the stimulus - response relationship for individual parameters of a respiratory cycle (3, 10, 15). Usually they are also based on
investigations limited to the effects observed during a single respiratory cycle with the assumption that the regulatory mechanism acts in a repetitive fashion during each cycle (7). However, such an assumption automatically excludes from considerations mechanisms with action extended over longer periods of time than one respiratory cyle. One of the experimental procedures commonly used in respiratory investigations is tracheal occlusion for the period of one breath. It is assumed that respiratory activity during this maneouvre is independent of phasic vagal activity i,e., pulmonary stretch receptor (PSR) activity evoked by lung inflation during the preceding non-occluded inspiration. However, several papers have been published (1, 11, 12, 23, 30) which indicate the existence of central mechanisms correlating the respiratory activity during tracheal occlusion with the value of phasic vagal activity exhibited before the procedure. In our work we attempted to determine to what degree the respiratory volume and, indirectly, the value of phasic vagal activity affect the parameters of subsequent respiratory cycles. The effects of tracheal occlusion maintained for seven breaths in spontaneously breathing rabbits, as well as switching off the servorespirator in paralysed animals were examined. Preliminary results have been published previously (28). METHODS
The experiments were performed in two series. In series I the experiments were carried out on 12 male rabbits weighing 2.6-3.5 kg. For the initial surgery the animals were premedicated with Fentanyl (0.025 mg/kg - Richter) and Droperidol (1.25 mglkg - Richter). After tracheostomy and application of tubocurarin artificial ventilation was carried out with the use of Universal Laboratory Respirator (Medipan). During the experiments the animals were maintained under halothane anesthesia (0.7Va vol). In series I1 (four rabbits) the animals were anesthetized with a mixture of chloralose (33 mg/l ml) and urethane 400 mg/ 11 ml) in the dose of 2 ml per kg b.w. intravenously. No pharmacological muscle relaxant was administered after tracheostomy. In both series phrenic nerves were prepared, their activity was amplified and integrated (560 Differential Ampl. Int. - Medipan) with a time constant of 100 ms. The end-expiratory C02 concentration and arterial blood pressure in femoral artery were measured (Beckman Medical Gas Analyser LB-2 and Statham P23Db). Body temperature in the range of normotermia (37-38OC) was maintained by an external heating. Gasometric and acid-base equilibrium measurements were carried out on an Automatic pH/Blood gas System (Corning 175). Respiratory volume
(VT) was measured by means of a pneumotachometer (351-Medipan) and Fleisch head No. 00. In series I the artificial ventilation controlled by phrenic nerve activity was applied (18). The procedure consisted in switching-off the respirator for a period of seven breaths (gain 0) (i) retaining conduction in both vagus nerves, (ii) after unilateral and (iii) bilateral vagotomy. In all three states the gain of the respirator servomechanism (i.e., the ratio of V T to the controlling signal) was changed in steps above and below the control value of VT (see 11, 27). In a part of these experiments a mixture'of 5O/a C02 in oxygen was supplied for breathing. It was mixed with air in such a proportion that end-tidal PCOz was the same before and after the change of the respirator servomechanism setting. In series 11, in addition to measurements described above, the EMG of respiratory intercostal muscles from 5 to 9 intercostal space was recorded by means of needle electrodes. The activity was filtred (band pass from 0.1 kHz to 1 kHz) and amplified (560 Differential Amplif. Int. Medipan). In this series of experiments "split-respiratory centre" preparations in non-paralysed rabbits were performed by midline incision of the medulla extending from the level of nucleus facialis about 7 mm rostrally to the obex to 2-3 mrn caudally to the obex (for more details see 17). Before and after the incision it was a routine procedure to occlude the trachea for a period of seven breaths. The integrated neural and muscular activity, end-tidal COz, tidal volume and arterial blood pressure were registered on a 6-channel, recorder (465 Medipan). RESULTS
The use of the phrenic nerve driven respirator allowed for greater variations of vagal input (due to changes in respiratory volume - VT) than possible in spontaneously breathing animals. For this reason, the majority of the results presented here comes from experiments carried out with the use of controlled ventilation. In series I, by analogy to the well known term "fictive locomotion" (locomotion as neural activity in paralysed animals), respiratory activity during "gain 0" will be termed "fictive breathing". During steady state conditions, i.e., 10-15 min after the pump gain setting had been changed, the respirator was switched-off for a period of seven breaths ("gain 0"). Figure 1A shows the record of integrated phrenic nerve activity during such a routine procedure. As can be seen, after switching off the pump, integrated phrenic nerve activity increased from breath to breath up to saturation level. In non paralysed ani-
PHR
h
m
m
m
E.I. EMG
8th
Fig. 1(a) The effect of switching off the pump (gain 0) on integrated phrenic nerve activity during seven fictive breaths. A, before vagotomy, B, after unilateral vagotomy, C , after bilateral vagotomy, BP, blo,cd pressure, CO,, end tidal CO, in ('10, Phr, integrated phrenic nerve activity. (b) The effect of tracheal occlusion on integrated phrenic (PHR) and inspiratory intercostal muscle (EI EMG) activities. EI EMG was recorded from the 8th and 5th intercostal space. On the record of tracheal pressure (PTR)the effects of tracheal occlusion during seven breaths are shown by strong deflections of PTR.
mals, tracheal occlusion performed at the end of expiration and maintained for a period of seven breaths led to the same result but more pronounced on integrated EMC of intercostal muscles activities than on Phr, as it is presented i n Fig. 1B. The above results show that the breath to breath increase in tracheal pressure during a prolonged tracheal occlusion is built by diaphgram and intercostal muscles, with a different contribution from them in each breath. In this paper the effect presented in Fig. 1 will be called a prolonged response. A similar reaction has been observed in cats anesthetized with nembutal and spontaneously breathing or ventilated by means of the other type (25) of phrenic nerve driven respirator. In paralysed animals, with increasing gain of the pump the intensity of the prolonged response increased, as shown in Fig. 2. In this Figure the percentage ratio of the amplitude of integrated phrenic nerve activity to the control value measured just before switching off the pump is presented in seven occluded breaths for three different gain settings. With the increase of pump gain the control value of Phr was smaller owing to A PHR toff
Con
c----.vagi intaot
+
vagi cut
No. of breath
Fig. 2. The effect of switching off the pump (gain 0 ) lasting seven "fictive" breaths on integrated phrenic nerve amplitude before and after vagotomy depending the initial gain of ventilation driven by the phrenic nerve activity. The higher the gain of the pump before maneuvre the higher the response to "gain 0" of the pump. Ordinate: the percentage increment of integrated phrenic nerve amplitude (A O/oPhr) in relation to the control value. Abscissa: C , control, off, the moment of pum)p switch off, on, the moment of pump switch on.
volume related vagal inhibition and lower chemical drive. For these reasons Fig. 3 shows a prolonged response in relative units. As can be seen an increased pump gain led, by increasing lung ventilation, to a decrease in PaC02. After bilateral vagotomy the breath by breath increase in phrenic amplitude during switching the pump off was very small or absent during the first 5-7 fictive breaths. Slow increase of integrated phrenic nerve activity persisted after the respirator was switched on again (Fig. 2), indicating a substantial time-lag of the chemical response after bilateral vagotomy. After vagotomy, when vagal inhibition was removed, control phrenic amplitude was higher for comparable PaCOz values 'than before vagotomy. Phr Pa C 4
Q vagi intact
vagi cut
No.of breath
Fig. 3. The effect of "gain 0" of the pump on integrated phrenic nerve amplitude (Phr) a t different levels of PaCO, before (left panel) and after (right panel) vagotomy. On the right side of each panel numbers show PaCOz values. Ordinate: amplitude of integrated phrenic nerve activity in relative units, C, control.
To differentiate the contribution of C02 and vagal input in evoking the prolonged response, switching the pump off was performed in two different sets: a. with constant gain but different C02 levels corresponding respectively to those obtained during ventilation with different volumes (gains) of servorespirator
b. with constant COz but different gains of servorespirator. The response confirmed the assumption that the prolonged response dependent mainly upon vagal afferent activity, since in (a) the only difference in the responses was an increase in phrenic nerve amplitude for higher PaCO2 values - both in control and during- the maneuvre, and in (b) the range and dynamics of prolonged response changed evidently with increasing pump gain in spite of constant COz. Moreover, the parameters of the first fictive breath depended on the pump gain - the larger the pump gain setting (and hence lower phrenic amplitude and time of inspiration (TI)in control), the smaller was the phrenic amplitude and T I of the first "fictive breath"' for the same value of end-tidal PC02. Figure 4 shows the prolonged response for three different pump gains in three states defined by different levels of afferent vagal nerve
One v o g u s c u t
Both v a g i out
B
C
1 2 3 4 5 6 7
No.of breath
Fig. 4. "Prolonged response" to switching .off the pump a t three different gains of the pump set before maneuvre. A, with vagi intact, B, after unilateral vagotomy, C, after bilateral vagotomy. Ordinates. increment of integrated phrenic nerve (Phr) amplitude as the percent of the control value. Letters on the right end of each curve indicate the gain of the pump, h, higher, c, control, 1, lower gains. In the insert PaCO, values for each gain are shown. 2
- Acta Neurobiologiae
activity: A - with both vagi intact, B - after unilateral and C - bilateral vagotomy. The diminution of prolonged response in result of unilateral and then bilateral vagotomy was clearly visible. Incision of the medulla from the level of the obex to 4 mm rostrally to it in rabbits evoked a desynchronization of the respiratory rhythm recorded from left and right phrenic nerves. In such a preparation vagally mediated reflexes persisted, although they were modified to some extent (29). However, the prolonged response to sustained tracheal occlusion disappeared in spite of intact vagi (Fig. 5).
PHR
L
dm
PHR
R PHR
L PHR
b
-
t
A
Fig. 5. "Prolonged response" of integrated right (R Phr) and left (L Phr) phrenic nerve activities to "gain 0" of the pump in a n intact animal (a) and after splitting the brainstem (b, c). In different animals, in c there is even diminution of inspiratory activity during "gain 0". Arrows indicate the maneuvre.
DISCUSSION
Since the work of Breuer-Hering (4), the effect of tracheal occlusion during respiratory pause on the activity of the phrenic nerve during the next inspiratory burst has been one of the best known respiratory reflexes. With intact vagi the amplitude of integrated phrenic nerve activity is greater during tracheal occlusion than in control conditions and the inspiratory burst lasts longer (9). Tracheal occlusion prolonged over the next few inspirations leads to a breath-by-breath increase in the amplitude of integrated phrenic nerve activity and gradual decrease in the duration of inspiration in relation to the first occluded breath. This respiratory reaction to extended tracheal occlusion had been observed
previously but its implications were not fully appreciated. On the basis of the work of our group as well as of other laboratories a hypothesis has been constructed (1, 28), that the breath-by-breath response to prolonged tracheal occlusion described above is an effect of the existence of central mechanisms which'control phasic vagal activity over longer periods of time than one respiratory cycle and that the vagal effect is in this case more important than chemical changes. The existence of a long-lasting vagally originated mechanism has been confirmed earlier, mainly with the application of electrical stimulation of the vagus nerve (21, 22). Karczewski (21) has found in vagotomized rabbits that electrical stimulation of the vagus nerve triggered by pulmonary stretch receptor activity restored respiratory rhythm to prevagotomy values after a number of breaths. He described this effect as "inertia" of the central system controlling respiratory rhythm parameters. Kahn and Wang (20) have reported -a gradual decrease in the frequency of firing of the pontine neurons during suceesive inflation periods and lack of this effect in medullary neurons when the artificial ventilation was maintained at subthreshold levels. This effect did not occur after vagotomy. The authors suggested that these pontine neurons could form a part of the pontine neural pathway for the Breuer-Hering reflex. progressive Milic-Emili and Pengelly (26) have pointed out thate!!t increase in tidal volume after the addition of an elastic load was attributed to the load-compensating mechanisms. However, little is still known about this mechanism in neurological terms, except that a vagal reflex o a1. (30) have is involved in this phenomenon (26). s a n t 9 ~ m b r o g i et found, in rabbits, t'nat the elimination of vagal afferentation by application of D.C. current for a period of a few breaths resulted in a prolongation of inspiration. However, the maximal reaction was only observed during the first breath and respiratory time gradually decreased in successive breaths. The effect was interpreted in terms of some additional mechanism inhibiting inspiration in the absence of vagal feedback. Callanan and Read ( 6 ) have described, also in rabbits, a gradual increase of airway pressure and tidal volume in successive breaths after tracheal occlusion or addition of elastic load. The effect was referred to chemical changes. However, it was still present after carotid body denervation, but it could be completely eliminated by subsequent vagotomy. Bartoli et al. (1) and particularly Cross ( l l ) , using phrenic nerveactivity controlled respirator (18) and artificial blood gas exchange system - (cardiopulmonary by-pass), have pointed out that the increase
of inspiratory efforts after diminution or elimination of phasic vagal afferentation is purely vagally mediated. On the basis of the literature discussed above and of our results, it can be concluded with certainty that phasic afferent vagal activity affects the successive breaths as a long-lasting central component of the Breuer-Hering reflex. The experimental method used in our work can be compared to the study of the after-effects of electrical stimulation of the vagus nerve. The direction of TI and TE changes in our experiments (TI (Fig. 6) and TE gradually decrease from the first fictive breath) indicates that this is a different phenomenon than the slow increase in TI and TE observed by
L
C
1
2
3
4
5
6
I
7
No. of brmt h Fig. 6. The effect of switching-off the pump during seven "fictive" breaths on integrated phrenic nerve amplitude (Phr) and the inspiratory time (TI). Ordinate: phrenic amplitude in arbitrary units and the inspiratory time TI in seconds.
Karczewski et al. (24) after elimination of vagal stimulation exciting irritant receptor afferents. This implies that it is the PSR that are responsible for the observed effect as was suggested by D'Angelo (12) and not the irritant receptors as was claimed by Iscoe and Vanner (19). However, there is also the possibility that during "fictive breathing" a collapse of alveoli and resuscitation of irritant receptors occurs. Since 1976 more and more reports describing the effect of afferent vagal activity occurring during a given breath on the parameters of the next breath have been published. Terms "short-term memory" (19, 22), central
summation (24), leaky integration (24, 33), short-term plasticity (23), central adaptation (3) have been introduced. But the majority of these papers concerned respiratory changes evoked by the stimulation of irritant receptors (31) or their afferents (16, 24). Apart from the work of Kahn and Wang (20), which in addition did not strictly deal with the subject of our work, there were no other attempts to elucidate the central mechanism responsible for the prolonged response. In the experiments with the application of the "split-respiratory centre" preparation we have found that the prolonged response disappears after longitudinal transection of the medulla from the level of the obex to 4 mm rostrally to the obex. From anatomical studies it is known that in the intermediate part of NTS (the location of vagal nerve interneurons) there are very few afferent fibres originating from the contralateral side of the IXth and Xth nerve, whereas there is a great number of them in the commissural part of NTS (8). I t has been shown (5) that after splitting the medulla the short-latency phrenic nerve activity inhibition to the single electrical shock is not transmitted to the contralateral side. Maybe the afferent fibres which pass midline at the level of the obex are a part of the reflex pathway of the prolonged response. Different suppositions are also possible: (i) The midsaggital section of the medulla results in the destruction of Nucleus Raphe Magnus (NRM) neurons which are a source of neurotransmitters and neuromodulators (endogenous opioides, serotonin) participating in neural control of breathing (14, 32). The damage of NRM might be responsible for the absence of prolonged response after splitting brain stem. (ii) This effect could be not vagally mediated, but vagally facilitated. In another words, after vagotomy, the increase of respiratory effort to the respiratory loading could be attenuated by the diminution of sensory input to the respiratory controller. This alternative could be supported by the results obtained in "split-brainstem" preparation (see above) and observation made on precollicular decerebrate cat (U. Borecka, S. Kasicki, J. R. Romaniuk - unpublished). In the latter case in one cat, the prolonged response to tracheal occlusion was hardly observed before and after vagotomy. However, electrical stimulation of mesencephalic locomotion area (MLR) which evoked locomotion and respiratory excitation (see 2, 13 for methods and references) restored the prolonged response (Fig. 7). These results show that any input to respiratory controller can facilitate the response to respiratory loading.
Fig. 7. The effect of tralcheal occlusion on integrated diaphragmatic EMG acitivity (L. Diaph. EMG) and transpulmonary pressure (PTP). Horizontal lines indicate the duration of the tracheal occlusim. Vagi are intact. A, tracheal occlusion at endexpiratory PTP; B, control tracheal occlusion at mid-expiratory PTP; C, during stimulation of MLR when locomotion ceased; D, during stimulation of MLR and locomotion (U. Borecka, S. Kasicki, J. R. Romaniuk - unpublished). We wish to thank Professor W. A. Karczewski for his criticism and help during the preparation of the manuscript and to expr&s our gratitude to Mrs. Krystyna Semerau-Siemianowska for her technical assistence. Experiments on cats were carried out in the laboratory of Prof. C. von Euler at Karolinska Institutet in Stockholm with the participation of A. DiMarm, C. van Euler and Y. Yamamoto whom the authors wish to thank for embling the verification of the results of this work in a different experimental set-up. This investigation was supported
by Project 10.4.03.3 of the Polish Academy of Sciences. A preliminary account of relsults presented in this paper was reported a t the XXVIII Congress of Physiological Sciences, Budapest, Hungary, July 13-19, 1980.
REFERENCES
1. BARTOLI, A., CROSS, B. A., GUZ, A., HUSZCZUK, A. and JEFFERIES, R. 1975. The effect of varying tidai volume on the associated phrenic motoneurone output: studies of vegal and chemical feedback. Respir. Physiol. 25: 135-155. 2. BORECKA, U., KASICKI, S. and ROlMANIUK, J. R. 1984. Breuer-Hering reflex in decerebrated walking cat. Proc. S E E R Conf. Chest wall and ventilatory failure, Barcelona: 25. 3. BRADLEY, G. W. 1977. Control of breathing pattern. In J. G. Widdicombe (ed.). Respiratory phyisiology 11. International review af physiology. Vol. 14. University Park Press, p. 185-217. 4. BREUER, J. 1968. Self-stearing of respiration through the nervus vagus. In R. Porter (ed.). Breathing. Ciba Foundation Hering-Breuer Symposium. London, 1970, p. 365-394. 5. BUDZINSKA, K., KIRKWOOD, P., ROMANIUK, J. R. and SBARS, T. A. 1984. Neuronal cross-connections involved in processing the respiratory activity in the medulla of the rabbit. Proc. SEPCR Conf. Chest wall and ventilatory failure, Barcelona: 19. 6. CALLANAN, D. and READ, D. J. C. 1974. The role of arterial chemoreceptors in the breath-by-breath augmentation of inspiratory effort in rabbits during air way occlusion or elastic loading. J . Physid. 241: 33-44. 7. CLARK, F. J. and EULER VON C. 1972. On the regulation of depth and rate of breathing. J. Physiol. 222: 267-295. 8. COTTLE, M. K. 1964. Degeneration studies of primary afferents of IXth and Xth cranial nerves in the cat. J. Comp. Neurol. 122: 329-345. 9. COHEN, M. I. 1975. Phrenic and recurrent laryngeal discharge patterns and the Hering-Breuer reflex. Am. J. Physiol. 228: 1489-1496. 10. COHEN, M. I. and FEUDMAN, J. L. 1977. Mod~elsof respiratory phase switching. Fed. Proc. 36: 2367-2374. 11. CROSS, B. A. 1970. An analysis of pulmonary vagal feedback mechnisms concerned with the control of breathing in the dog. Ph. Thesis, University of London. 12. D'ANGELO, E. 1977. Effects of single breath lutng inflation on the pattern of subsequent breaths. Respir. Physiol. 31: 1-18. 13, DiMARCO, A. F., ROMANIUK, J. R., EULER VON C. and YAMAMOTO, Y. 1983. Emmediate changes ventilation and re(spiratory pattern associated with onset and cessation of locomotion in the cat. J. Physiol. 343: 1-16. 14. ELDRIDGE, F. L. and MILLHORN, D. E. 1981. Central regulation of respiration by endogenous neurotransmitters and neuromodulators. Ann. Rev. Physiol. 43: 121-135. 15. EULER VON C. 1977. Functional organization of the respiratory phase-switching mechanism. Fed. Proc. 36: 2375-21380. 16. GEOWICKI, K. and ROMANIUK, J. R. 1984. Differenthtion of rWpkatmy reflexes by electrical stimulation and cold blockade of the vagus nerve in - the rabbit. Acta Physiol. Pol. (in prea). 17. GROMYSZ, H.and KARCZEWSKI, W. A. 1980. Generation of respiratory pattern
18. 19. 20.
21. 22.
23. 24.
25. 26.
27. 28.
29.
30.
31.
32.
33.
in the rabbit-brainstem transection revisited. Acta Neurobiol. Exp. 40: 985-992. HUSZCZUK, A. 1970. A respiratory pump cantrollled by phrenic nerve activity. J. Physiol. 210: 183-184. ISCOE, S. and VANNER, S. 1979. Respiratory periodicity following stin~ulation of vagal affwents. Can. J. Physiol. Pharmacol. 58: 823-829. KAHN, N. and WANG, S. C. 1%7. Mectrophysiological basis for pontine apneustic center and its role i n integration of the Hering-Breuer reflex. J. Neurophysiol. 30: 301-318. KARCZEWSKI, W. A. 1963. A model of proprioceptive information from the lungs. Proc. V Int. Conf. Med. Electronics, Liege. KARCZEWSKI, W. A. BUDZINSKA, K., GROMYSZ, H., H E R C Z Y ~ K I R. . and ROMANIUK, J. R. 1976. Some responses of the respiratory complex to stimulation of its vagal and mesencephalic inputs. In B. Duron (ed.). Respiratory centres and afferent system, INSERM, 59, p. 107-115. KARCZEWSKI, W. A. and ROMANIUK, J. R. 1980. Neural control of breathing a n d central nervous system plxticity. Acta Physiol. Pol. S u ~ p l .31: 1, 10. KARCZEWSKI, W. A., NASEONSKA, E. and ROMANIUK, J. R. 1980. Respiratory responses to stimulation of afferent vagal fibres in rabbits. Aota Neurobiol. Exp. 40: 543-562. KNOX, C. K. 1973. Characteristics of inflation and deflation reflexes during expiration in the rat. J. Neurophysiol. 36: 284-295. MILIC-EMILI, J. and PENGELLY, L. D. 1970. Ventilatory effects of mechanical loading. In E. J. M. Cambell, E. Agostoni, J. New,som Davis, LloydLuka (ed.). The reslpiratory muscles. Mechanics and neural control. London, p. 271-290. ROMANIUK, J. R., RYBA, M. and GROTEK, A. 1976. The effects of C 0 2 on the companents of breathing pattern. Acta Physiol. Pol. 27: 215-233. ROMANIUK, J. R., BUDZINSKA, K. and GEOWICKI, K. 1980. Activity of phrenic motoneurons output during release from vagal inhibition. Proc. Int. Union Physiol. Sci. XXXIII Int. Congr., Budapest, 667, 2948. ROMANIUK, J. R. and BUDZINSKA, K. 1982. The control of respiratory outputs i n "split respiratory centre" preparation. The brain in health and disease. I World Congr. IBRO. Lausanne, Neuroscience, Suppl. 7, p. 180. SANT'AMBROGIO, G., C ~ P O R E S I ,E., SELLICK, H. and MORTOLA, J. 1972. Respiratory changes induced by the im'mediate block of nervous conduction in the vagus nerves. Q. J. Exp. Physiol. 57: 207-212. SELLICK, H. and WIDDICOMBE, J. G. 1970. V a m l deflation and inflation reflexes mediated by lung irtritant receptors. Q. J. Exp. Physiol. 55: 1531.63. SESSLE, B. J., BALL, G. J. ar$ LUCIER, G. E. 1981. Suwresive influences from periaqueductal gray and nucleus raphe magnus on respiration and related activities and on solitary tract neurons and effect of naloxone. Brain Res. 216: 145-161. YOUNES, M., BAKER, J. P. POLACHECK, J. and REMMERS, J. E. 1978. Tery In mination of inspiration thro,ugh graded inhibition of i n ~ p i r a ~ t o ractivity. R. S. Fitzgerald, H. Gautier, S. Lahiri (ed.). The regulation of respiration during sleep and anaesthesia. Adv. Exp. Med. Biol. Plenum Press, New York, 99, 383.
Accepted 9 July 1984