The central respiratory effects of acetylcholine vary with CSF pH

Journal of the Autonomic Nervous System 62 Ž1997. 27–32

The central respiratory effects of acetylcholine vary with CSF pH Melvin D. Burton ) , Douglas C. Johnson, Homayoun Kazemi Department of Medicine, Pulmonary and Critical Care Unit, Massachusetts General Hospital and HarÕard Medical School, Boston, MA 02114, USA Received 26 March 1996; revised 1 August 1996; accepted 27 August 1996

Abstract Hydrogen ion concentration wHqx centrally is a major determinant of ventilation. Its action involves central cholinergic mechanisms. The pointŽs. where increased wHqx induces its changes in the cholinergic system is unclear. If Hq acts presynaptically by increasing endogenous ACh synthesis and release, its effect should be absent when ACh is supplied exogenously. If Hq acts postsynaptically by changing ACh degradation or ACh receptor sensitivity, its effect should persist in the presence of exogenous ACh. We perfused the brain ventricular system in spontaneously breathing anesthetized dogs with progressively higher concentrations of ACh Ž0–52.8 mM. in cerebrospinal fluid ŽCSF. at pH 7.4 and CSF pH 7.1. Increasing concentrations of ACh increased ventilation ) 4-fold in a linear manner in the presence of non-acidic and acidic CSF. With acidic CSF the ACh ventilatory response line was shifted to a higher y-intercept, resulting in a higher ventilation at any wAChx. These findings are consistent with the hypothesis that central acidosis augments ventilation by postsynaptic cholinergic events. Keywords: Control of ventilation; Chemosensitivity; CSF acidosis; Acetylcholine; Acetylcholinesterase

1. Introduction Central Hq chemosensitivity, one of the most vital components of the ventilatory control system, has been extensively described w5,17,26x yet its genesis remains unclear. The cellular elements that give rise to this Hq chemosensitive process have been localized to the ventrolateral medulla ŽVLM. w5,18,21,30x. This area is richly endowed with cellular elements of ACh metabolism and reception w9,14,15x. Observations, using muscarinic blockers, suggest that this Hq chemosensitive process involves facilitated cholinergic transmission w6,10,11,18,25x. It is, however, unclear whether this is a presynaptic or postsynaptic process. Reviews of presynaptic modulation of neurotransmission w27,31x and studies of the kinetic properties of choline acetyltransferase w13,29x have shown that exogenous ACh down regulates the endogenous ACh presynaptic turnover. In this study we provided exogenous ACh to down regulate presynaptic turnover and then assessed H q sensitivity. We propose that the absence of the Hq ventilatory effect with exogenous ACh would favor a

)

Corresponding author. Tel.: q1 617 7262406; fax: q1 617 7262932.

presynaptic cholinergic mechanism, while the presence of the Hq ventilatory effect would favor a postsynaptic cholinergic mechanism. Our findings suggest the latter.

2. Materials and methods Six mongrel dogs Ž9–14 kg body wt.. were purchased from a local kennel where animals are bred for research. The principles approved by the Council of the American Physiological Society regarding animal experimentation were followed in this work. The animals were anesthetized with intravenous pentobarbital sodium Ž30 mgrkg., intubated and attached to an O 2-enriched spirometer ŽCollins, Boston, MA. with an attached CO 2 scrubber and fan for recirculation. A femoral artery, femoral vein and pulmonary artery were catheterized for hemodynamic monitoring and arterial blood sampling. Unassisted ventilation was allowed throughout the course of the experiment. The core temperature was monitored and maintained at 388C. The animals were observed continuously for signs of pain or arousal. A positive blink reflex on eyelid stimulation was considered to be an early sign of awakening and an indication for an additional dose of pentobarbital sodium

0165-1838r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved. PII S 0 1 6 5 - 1 8 3 8 Ž 9 6 . 0 0 1 0 4 - X

28

M.D. Burton et al.r Journal of the Autonomic NerÕous System 62 (1997) 27–32

Ž6 mgrkg body wt. i.v... Access to the cerebral ventricular system was identical to that previously described from this laboratory w6x. Eighteen-gauge needles were placed in both lateral ventricles and the cisterna magna with stereotaxic assistance. Ventriculocisternal perfusion ŽVCP. was performed with simultaneous entry of cerebrospinal fluid ŽCSF. into both lateral ventricles and exit of CSF from the cisterna magna. 2.1. CSF and drug preparation Two different artificial CSF solutions were used. The nonacidic CSF contained the following electrolytes Ž25., Žmmolrl.: Naq Ž140., Cly Ž120., Kq Ž2.6., HCOy 3 Caqq Ž4. and Mgqq Ž2.. The nonacidic CSF was equilibrated with 95% O 2 and 5% CO 2 resulting in pH 7.4 and PCO 2 torr. The acidic CSF contained comparable elecx was decreased 50% and the trolytes, except the wHCOy 3 chloride concentration was increased to 133 mmolrl. The acidic CSF was equilibrated with 95% O 2 and 5% CO 2 resulting in pH 7.1 and PCO 2 40 torr. Acetylcholine ŽSigma Chemical, St. Louis, MO. was dissolved in both nonacidic CSF and acidic CSF at different concentrations as outlined below. The addition of acetylcholine did not alter the pH of the solutions. All artificial CSFs and solutions were prepared on the day of the experiment.

inspiratory duty cycle ŽTirTtot . and inspiratory flow Ž VtrTi . were calculated. 2.4. Analysis Each animal served as its own control. Values at 15 min of VCP with nonacidic CSF, were compared with values at 15 min of VCP with the various concentrations of ACh in nonacidic CSF. The comparison was carried by ANOVA with repeated measures using Fisher test. Values at 15 min of VCP with acidic CSF were compared with values at 15 min of VCP with the various concentrations of ACh in acidic CSF. Again ANOVA was used, with repeated measures using Fisher test. Values at 15 min of VCP with the various concentrations of ACh in nonacidic CSF were compared with values at 15 min of VCP with the corresponding concentration of ACh in acidic CSF; the paired two-tailed t-test was used. Values are presented as means " S.E. P - 0.05 was considered significant. Regression analysis was applied to the Ve and f r values. The 15-min values were used for comparison because a new steady state had been achieved.

3. Results 3.1. Ventilatory findings

2.2. Experimental design Each animal received an acetylcholine challenge by VCP at CSF pH of 7.4 and at CSF pH of 7.1. The order of the challenges, acidic vs. nonacidic, was random. The cerebral ventricular system was perfused for 15 min with either control nonacidic CSF or acidic CSF at a rate of 1 mlrmin. This served as the initial base line. The system was then perfused at consecutive 15-min intervals with incremental concentrations of acetylcholine Ž6.6, 13.2, 26.4, and 52.8 mM.. There was a 20-min washout with control nonacidic CSF between the acidic and nonacidic ACh challenges. The concentrations of ACh in this study were similar to concentrations known to produce ventilatory changes in the anesthetized dog w6x and magnitudes lower than concentrations known to exist in cholinergic synaptic vesicles w32x. 2.3. Measurements At 0, 5, 10 and 15 min of ventriculo-cisternal perfusion, hemodynamics Žheart rate, mean femoral artery pressure, mean pulmonary artery pressure and cardiac output. were recorded and arterial pH and gas tensions were measured at 378C Žmodel 1303, Instrumentation Laboratories, Lexington, MA.. Tidal volume Ž Vt ., respiratory frequency Ž f r ., inspiratory time ŽTi ., expiratory time ŽTe . and total cycle time ŽTtot . were measured using the average of five breaths at the designated time points. Minute ventilation Ž Ve .,

There were prompt ventilatory changes during VCP with ACh in the presence of both nonacidic and acidic CSF. The effects were sustained and reached a plateau within the first 5–10 min. Ventilation promptly returned to baseline during the washout with control nonacidic CSF. Fig. 1 shows that the rise in ventilation was linearly related to the wAChx in the perfused CSF. At CSF pH of 7.4, Ve s 0.17 wAChx q 2.39 Ž r s 0.7, P - 0.001.. At CSF pH of 7.1, Ve s 0.22 wAChx q 4.13 Ž r s 0.8, P - 0.001.. The 95% confidence intervals about the mean Ž x, y . showed no overlap. The slope in the acidic CSF challenge was higher but did not achieve statistical significance. In the presence of acidic CSF, ventilation was higher at any given wAChx. The increase in ventilation was due to increases in both tidal volume and respiratory frequency. There was an early rise and plateau of tidal volume ŽFig. 1.. The tidal volume was higher during the ACh perfusions in the presence of acidic CSF. The respiratory frequency increased as a linear function of CSF wAChx. At CSF pH of 7.4, f r s 0.61 wAChx q 14 Ž r s 0.74, P - 0.001.. At CSF pH of 7.1, f r s 0.62 wAChx q 18 Ž r s 0.77, P - 0.001.. The 95% confidence intervals about the mean Ž x, y . showed overlap. In the presence of acidic CSF this linear function was shifted upwards; however, the corresponding values on the acidic and nonacidic lines were not significantly different. The inspiratory flow response mirrors that of the tidal volume, i.e. an early plateau and a relatively higher response in the presence of acidic CSF ŽFig. 2.. The inspiratory duty cycle

M.D. Burton et al.r Journal of the Autonomic NerÕous System 62 (1997) 27–32

29

Fig. 2. The inspiratory flow Ž Vt r Ti . and duty cycle ŽTi r Ttot . response after 15 min of ventriculocisternal perfusion in anesthetized dogs with various concentrations of acetylcholine in the presence of different CSF pHs. ns6. Values are means "S.E. ), Significant difference Ž P - 0.05. from value at 0 mM of ACh by ANOVA at constant CSF pH. †, Significant difference Ž P - 0.05. by paired t-test between the two different CSFs perfusions at common wAChx. ††, P - 0.1. Fig. 1. The ventilation, tidal volume and breathing frequency response after 15 min of ventriculocisternal perfusion in anesthetized dogs with various concentrations of acetylcholine in the presence of different CSF pHs. ns6. Values are means "S.E. ), Significant difference Ž P - 0.05. from value at 0 mM of ACh by ANOVA at constant CSF pH. †, Significant difference Ž P - 0.05. by paired t-test between the two different CSFs perfusions at common wAChx. ††, P - 0.1.

mimics the respiratory frequency response ŽFig. 2.. Both show a progressive increase with a tendency towards higher values in the presence of acidic CSF.

3.2. Arterial blood findings Normoxia was maintained throughout all experiments. There was a mild respiratory acidosis at baseline which we believe to be due to pentobarbital anesthesia. There was progressive depression of Pa CO 2with increasing wAChx ŽFig. 3.. There was greater depression of Pa CO 2 in the presence of acidic CSF, however, the difference was not significant at all points. There was a progressive rise in arterial pH

Table 1 Cardiovascular response during VCP with acetylcholine at different CSF pHs CSF pH s 7.4

CSF pH s 7.1

ACh ŽmM.

ACh ŽmM.

0

6.6

13.2

26.4

52.8

0

6.6

13.2

26.4

52.8

HR 188 " 5 199 " 8 199 " 7 199 " 11 190 " 15 185 " 8 192 " 12 190 " 13 195 " 12 196 " 9 MFAP ŽmmHg. 137 " 5 143 " 6 140 " 5 134 " 5 121 " 7 ) 137 " 7 139 " 7 138 " 7 129 " 8 125 " 7 ) MPAP ŽmmHg. 26 " 3 † 25 " 2 † 24 " 2 26 " 2 27 " 3 33 " 3 † 34 " 5 † 29 " 3 31 " 3 31 " 3 CO Žlrmin. 4.3 " 0.5 4.5 " 0.6 4.5 " 0.6 4.9 " 0.7 4.6 " 0.9 4.3 " 0.6 5.3 " 1.0 4.5 " 0.7 4.6 " 0.7 4.5 " 0.6 Cardiovascular parameters after 15 min of ventriculocisternal perfusion with various concentrations of acetylcholine ŽACh. in the presence of different CSF pHs Ž7.4 and 7.1.. Heart rate ŽHR., mean fermoral artery pressure ŽMFAP., mean pulmonary artery pressure ŽMPAP. and cardiac output ŽCO. are given as means "S.E. ) The value is significantly different from value at O mM of ACh by ANOVA. † A significant difference by paired t-test at common wAChx.

30

M.D. Burton et al.r Journal of the Autonomic NerÕous System 62 (1997) 27–32

Fig. 3. The arterial PCO 2 and pH response after 15 min of ventriculocisternal perfusion in anesthetized dogs with various concentrations of acetylcholine in the presence of different CSF pHs. ns6. Values are means "SE. ), Significant difference Ž P - 0.05. from value at 0 mM of ACh by ANOVA at constant CSF pH. †, Significant difference Ž P - 0.05. by paired t-test between the two different CSFs perfusions at common wAChx. ††, P - 0.1.

with increasing wAChx ŽFig. 3.. There was a greater rise in arterial pH in the presence of acidic CSF. 3.3. CardioÕascular findings The cardiovascular parameters were stable ŽTable 1.. There were no changes in the heart rate with increasing wAChx and no differences between the acidic and nonacidic challenges. There was a small decrease in the mean femoral artery pressure during perfusion with the 52.8 mM ACh in both acidic and nonacidic CSF. The mean pulmonary artery pressure was unchanged; however, there was a tendency toward higher pressures in the presence of acidic CSF. The cardiac output was unchanged and there were no differences between the acidic and nonacidic CSF perfusions.

4. Discussion In vertebrates, at a fixed temperature, ventilation is dependent on the wHqx of the brain’s extracellular fluids w5,12,26x. This Hq chemosensitive process of the ventilatory control system has been localized to the microenvironment of the ventrolateral medulla w5,21x. In vitro and in

vivo data have suggested that the Hq chemosensitive ventilatory response involves cholinergic mechanisms w6,10,11,18,20,22,25x. This study addresses how increased Hq may facilitate cholinergic transmission and thus augment ventilation. Presynaptically, Hq may increase endogenous ACh synthesis and release. Postsynaptically, Hq may inactivate synaptic junction acetylcholinesterase ŽAChE. or enhance the sensitivity of postsynaptic ACh receptors. In this work we take advantage of the presynaptic autoinhibitory effect of exogenous ACh on endogenous ACh synthesis and release w13,27,29,31x. By driving ventilation with exogenous ACh and simultaneously assessing Hq sensitivity, we sought to answer whether the Hq-induced ventilatory effect is consistent with a presynaptic or postsynaptic event. We found that the Hq effect on ventilation persists even when ventilation is driven to extreme levels by exogenous ACh ŽFig. 1.. The Hq effect is manifested by an increase in the set point Ž y-intercept. and a mild nonsignificant increase in the gain Žslope. of the ventilatory reponse to ACh challenge ŽFig. 1.. One interpretation of these data is that ACh and CNS acidosis are both ventilatory stimulants which simply produced an additive effect on ventilation. The current data does not exclude this possibility. We think this is not the best interpretation of these data. Cholinergic transmission involvement in central ventilatory Hq chemosensitivity is well established w6,10,11,18,22,25x. Indeed when the acidosis is limited to the CNS as is the case with this acidic CSF perfusion model, the ventilatory response is completely blocked by atropine w6x, suggesting a critical dependence on a cholinergic process. The findings of this study may be best explained by the view that Hq augments ventilation by modifying cholinergic transmission at a postsynaptic site, i.e. inhibition of AChE or enhanced ACh receptor sensitivity. The latter is unlikely because the binding of brainstem muscarinic receptors is insensitive to wHqx between pHs 7 and 10 w1x. Thus the inhibiton of AChE is the best explanation. If this enzyme is indeed part of the central elements which allow ventilatory chemosensitivity, then the functional properties of AChE should match the known functional properties of the central Hq chemosensitive elements. Previous investigations suggest that the central Õentilatory chemosensitiÕe elements must haÕe the following characteristics: Ža. There is strict Hq sensitivity at isothermic conditions w12,26x, with increased ventilation at more acidic pH. Žb. There is relative Hq insensitivity when temperature is the dependent variable, i.e. ventilation remains constant in the face of varying pHs w12x Ž‘alpha-stat hypothesis’.. Žc. There are functionally important imidazole groups w23,24x, which help explain Žb. above. Žd. The elements are located on or near the ventral lateral medulla w30x. Že. Perturbation of the elements mediates a change in ventilation w30x.

M.D. Burton et al.r Journal of the Autonomic NerÕous System 62 (1997) 27–32

Acetylcholinesterase is known to haÕe the following characteristics: Ža. The hydrolytic activity shows strict Hq sensitivity at isothermic conditions w3x, with less hydrolytic activity at more acidic pH. Žb. The hydrolytic action, at least in the poikilotherm, appears to be temperature adaptative, i.e. AChE activity remains constant in the face of varying environmental temperatures w4x. Žc. AChE is known to have imidazoles in its active site w16x, and indeed the imidazole groups are important in conferring the hydrolytic action w28x. Žd. AChE is present in the ventral lateral medulla w14x. Že. Perturbation of ventral medullary AChE with inhibitors increases ventilation w10x. We realize that the striking similarities between AChE and the ventilatory chemosensitive elements do not provide proof that they are one and the same. More work is needed in this area. However, it is intriguing to speculate that medullary AChE’s strict Hq sensitivity at isothermic conditions may account for the ventilatory findings in this study. The cardiovascular changes in our study were unimpressive. This is in agreement with the cardiovascular response to VCP with ACh in our previous study w6x. Others, using different preparations have found more prominent cardiovascular responses with cholinergic agents w2,10x. Species differences or technique differences may account for the varying outcome. The depth of penetration may be a factor. In our study, the brainstem surface is bathed with ACh by VCP. Studies that have shown the most impressive cardiovascular changes have employed microinjections into the subsurface of the rostral medulla w2x. In any event, the relative absence of cardiovascular changes in the face of dramatic ventilatory changes was surprising. This finding is in striking contrast to that of neuroactive amino acids w8x and catecholamines w7x which show both ventilatory and cardiovascular changes when given by VCP. These findings suggest that ACh given by VCP in the dog may have a relative specific action on ventilatory related pathways.

31

concentrations of ACh w13,29x. The time course of transmitter turnover at cholinergic synapses is rapid, ranging from seconds for synthesis and vesicular uptake to microseconds for release w19x. The 15-min perfusion period used in this study was adequate time to allow stabilization of perturbations occurring at such synapses. The VCP model exposes the entire brain stem to the CSF perfusate. We, therefore, cannot specify which brain stem siteŽs. are responsible for the changes we observed. It is likely that the observed respiratory effects are due to interactions at both ventral lateral medullary ŽVLM. sites and non-VLM sites. Indeed, recent data have shown that respiratory acid chemosensitive neurons are present in both VLM and non-VLM sites. The animals were deeply anesthetized with pentobarbital. This blunted their chemosensitivity as evidenced by the resting hypercapnia and small ventilatory change with acidic CSF alone. We did not control arterial PCO 2 and pH. During the ACh challenges, a severe respiratory alkalosis developed. This hypocapnia and alkalemia undoubtedly dampened the magnitude of the ventilatory changes observed. If the arterial acid-base state had been held constant, an even greater separation of the ACh ventilatory response lines in the presence of non-acidic and acidic CSF would be expected. Finally, this work addresses only the role of cholinergic transmission. The ventilatory control system is complex and is influenced by a multitude of neurotransmitter systems. It is quite possible that part of the Hq and cholinergic effects may interact with these other systems. Cholinergic mechanisms are important in the central ventilatory chemosensitive response. This work has shown that the Hq chemosensitive effect persists even when ventilation is driven by exogenous ACh. This finding implies that the cholinergic mechanism of the Hq chemosensitivity may involve postsynaptic events, e.g. changes in acetylcholinesterase activity or ACh receptor sensitivity. Since the known functional properties of acetylcholinesterase are virtually identical to the known functional properties of the ventilatory Hq chemosensitive elements, we suggest that CSF wHqx may augment ventilation by modifying acetylcholinesterase activity.

4.1. Limitations of study We used a whole body approach and based upon the observations made some reasonable assumptions about cellular events. Thus the assumed cellular events remain to be proven. This will require a cellular approach using a reduced preparation. This study did not document inhibition of endogenous ACh synthesis or ACh release during the exogenous ACh challenge. We believe that endogenous ACh synthesis and release were decreased during the exogenous ACh perfusions. This is based upon in vitro work showing Ža. that the release of ACh is modulated presynaptically by inhibitory autoreceptors w27,31x, and Žb. that choline acetyltransferase is inhibited by millimolar

Acknowledgements We thank Elizabeth Cargill and Mahnaz Nouri for superb technical assistance. This work was supported by National Heart, Lung, and Blood Institute ŽNHLBI. Grant HL-29493 and by an American Lung Association Research Award. References w1x Anthony, B.L. and Aronstam, R.S., Effect of pH on muscarinic acetylcholine receptors from rat brainstem, J. Neurochemistry, 46 Ž1986. 556–561.

32

M.D. Burton et al.r Journal of the Autonomic NerÕous System 62 (1997) 27–32

w2x Arneric, S.P., Giuliano, R., Ernsberger, P., Underwood, M.D. and Reis, D.J., Synthesis, release and receptor binding of acetylcholine in the C1 area of the rostral ventrolateral medulla: contributions in regulating arterial pressure, Brain Res., 511 Ž1990. 98–112. w3x Augustinsson, K.B., Cholinesterases: a study in comparative enzymology, Acta Physiol. Scand., ŽSuppl. 52. 15 Ž1948. 1–182. w4x Baslow, M.H. and Nigrelli, R.F., The effect of thermal acclimation on brain cholinesterase activity of the killfish, Fundulus heteroclitus, Zoologica, 49 Ž1964. 41–51. w5x Bruce, E.N. and Cherniack, N.S., Central chemoreceptors, J. Appl. Physiol., 62 Ž1987. 389–402. w6x Burton, M.D., Johnson, D.C. and Kazemi, H., CSF acidosis augments ventilation through cholinergic mechanisms, J. Appl. Physiol., 66 Ž1989. 2565–2572. w7x Burton, M.D., Johnson, D.C. and Kazemi, H., Adrenergic and cholinergic interaction in central ventilatory control, J. Appl. Physiol., 68 Ž1990. 2092–2099. w8x Chiang, C., Pappagianopoulos, P., Hoop, B. and Kazemi, H., Central cardiorespiratory effects of glutamate in dogs, J. Appl. Physiol., 60 Ž1986. 2056–2062. w9x Cortes, R., Probost, A. and Palacios, J.M., Quantitative light microscopic autoradiographic localization of cholinergic muscarinic receptors in the human brain: brainstem, Neuroscience, 12 Ž1984. 1003– 1026. w10x Dev, N.B. and Loeschcke, H.H., Cholinergic mechanism involved in respiratory chemosensitivity of the medulla oblongata in the cat, Pfluegers Arch., 379 Ž1979. 29–36. w11x Fukuda, Y. and Loeschcke, H.H., Cholinergic mechanism involved in neuronal excitation by Hq in rats in vitro, Pfluegers Arch., 379 Ž1979. 125–135. w12x Hitzig, B.M., Temperature induced changes in turtle CSF pH and central control of ventilation, Resp. Physiol., 49 Ž1982. 205–222. w13x Kaita, A.A. and Goldberg, A.M., Control of acetylcholine synthesis. The inhibition of choline acetyltransferase by acetylcholine, J. Neurochem., 16 Ž1969. 1185–1191. w14x Kasa, P., The cholinergic systems in brain and spinal cord, Prog. Neurobiol., 26 Ž1986. 211–272. w15x Kobayashi, R.M., Brownstein, M., Sanvedra, J.M. and Palkovits, M., Choline acetyltransferase content in discrete regions of the rat brainstem, J. Neurochem., 24 Ž1975. 637–640. w16x Krupka, R.M., Chemical structure and function of the active center of acetylcholinesterase, Biochemistry, 5 Ž1966. 1988–1997. w17x Leusen, I., Chemosensitivity of the respiratory center, Am. J. Physiol., 176 Ž1954. 39–51.

w18x Loeschcke, H.H., Central chemosensitivity and the reaction theory, J. Physiol. 332 Ž1982. 1–24. w19x McIlwain, H. and Bachelard, H.S., Chemical factor in neurotransmission; acetylcholine. In: Biochemistry and the Central Nervous System, Churchill Livingstone Publishing, Edinburgh, London, Melbourne and New York, 1985, pp. 413–440. w20x Metz, B., Hypercapnia and acetylcholine release from the cortex and medulla, J. Physiol. Lond., 186 Ž1966. 321–322. w21x Mitchell, R.A., Loeschcke, H.H., Massion, W.H. and Severinghaus, J.W., Respiratory responses mediated through superficial chemosensitive areas on the medulla, J. Appl. Physiol., 18 Ž1963. 523–533. w22x Monteau, R., Morin, D. and Hilaire, G., Acetylcholine and central chemosensitivity: in vitro study in the newborn rat, Respir. Physiol., 81 Ž1990. 241–254. w23x Nattie, E.E., Diethyl pyrocarbonateŽan imidazole binding substance. inhibits rostral VLM CO 2 sensitivity, J. Appl. Physiol., 61 Ž1986. 843–850. w24x Nattie, E.E., Diethyl pyrocarbonate inhibits rostral ventrolateral medullary Hq sensitivity, J. Appl. Physiol., 64 Ž1988. 1600–1609. w25x Nattie, E.E., Wood, J., Mega, A. and Goritski, W., Rostral ventrolateral medulla muscarinic receptor involvement in central ventilatory chemosensitivity, J. Appl. Physiol., 66 Ž1989. 1462–1470. w26x Pappenheimer, J.R., Fencl, V., Heisey, S.R. and Held, D., Role of cerebral fluids in control of respiration as studied in unanesthetized goats, Am. J. Physiol., 208 Ž1965. 436–450. w27x Raiteri, M., Marchi, M. and Paudice, P., Presynaptic Muscarinic receptors in the central nervous system, Annals NY Acad. Sci., 604 Ž1990. 113–129. w28x Roskoski, Jr., R., Choline acetyltransferase and acetylcholinesterase: evidence for essential histidine residues, Biochemistry, 13 Ž1974. 5141–5144. w29x Ryan, R. and McClure, W.O., Ž1980. Physical and kinetic properties of choline acetyltransferase from rat and bovine brain, J. Neurochem., 34 Ž1980. 395–403. w30x Schlafke, M.E., See, W.R. and Loeschcke, H.H., Ventilatory response to alterations of Hq concentration in small areas of the ventral medullary surface, Resp. Physiol., 10 Ž1970. 198–212. w31x Vizi, E.S., Presynaptic modulation of neurochemical transmission, Prog. Neurobiol., 12 Ž1979. 181–290. w32x Whittaker, V.P., Essman, W.B. and Dowe, G.H., The isolation of pure cholinergic synaptic vesicles from the electric organs of eleasmobranch fish of the family Torpidinae, Biochemical J., 128 Ž1972. 833–846.