Cholinergic stimulation of the pons depresses respiration in decerebrate cats HIROSHI KIMURA, LESZEK KUBIN, RICHARD 0. DAVIES, AND ALLAN I. PACK Department ofAnimal Biology, School of Veterinary Medicine, and Pulmonary Section, School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania 19104 KIMURA, HIROSHI, LESZEK KUHN, RICHARD 0. DAVIES, AND ALLAN I. PACK. Cholinergic stimulation of the pans depresses respiration in decerebrate cats. J. Appl. Physiol. 69(6):

X280-X289,1990.-The injection of carbachol into the pontine tegmentum of decerebrate cats evokes a postural motor atonia that has many of the characteristics of the atonia of natural rapid-eye-movement (REM) sleep (Morales et al. J. Neurophysiol. 57: 1118-1129, 1987). We have used the carbacholinjected decerebrate cat to study the changes in respiratory neuronal activity that accompany the at,onia. The activities of representative respiratory m&or nerves-phrenic, intercostal, and hypoglossal-and that of a motor branch of C4 were recorded in decerebrate, vagotomized, paralyzed, and artificially ventilated cats. After the microinjection of carbachol, there was a profound suppression of activity in all the nerves and a decrease in respiratory rate. This was a consistent stereotyped response in which the magnitude of the suppression of respiratory-related activity was phrenic (to -65% of control) < inspiratory intercostal (-50%) < hypoglossal (-10%) < expiratory intercostal (-5%). The decrease in respiratory rate (to ~70% of control) was caused by a prolongation of both inspiratory and expiratory durations. Complete reversal of the carbachol effect was elicited by the microinjection of atropine into the same site as the carbachol injection. This allowed us to produce a second episode of atonia by the injection of carbachol into the contralateral pons. Thus we have demonstrated the existence of neural pathways originating in the cholinoceptive cells of the pons that have the potential to powerfully and differentially depress various respiratory motoneuronal pools and to reduce the respiratory rate. These pathways are likely to be activated along with the atonia of REM sleep. atonia; carbachol; hypoglossal nerve; intercostal nerve; phrenic nerve; rapid-eye-movement sleep; control of breathing; pontine tegmentum

ONE

OF THE

CARDINAL

SIGNS

of rapid-eye-movement

(REM) sleep is a profound postural muscle atonia. REM sleep is also accompanied by changes in the activity of various respiratory muscles, changes that differ greatly among the various muscles studied. Rib cage muscles are not uniformly affected; the intercostal muscles become markedly hypotonic throughout REM sleep (5, 20, 25), whereas the interchondral, triangularis sterni and levator costae muscles maintain their inspiratory activity (5,20). Also, the activity of upper airway muscles can be affected differently; pharyngeal and laryngeal abductors become atonic or hypotonic (19,30,3l), but the activity of certain constrictors is maintained (31). Although the diaphragm is spared the intense descending inhibition characteris2280

0161-7567/90

$1.50 Copyright

tically observed in postural muscles during REM sleep (5, 26, 28, 29), recent findings by Orem (24), Sieck et al. (32), Kline et al. (14), and Hendricks et al. (6) indicate that there is indeed some suppression of activity, especially during episodes of phasic REM sleep. A simple construct to explain these variable effects of REM sleep on respiratory muscle activity posits that motoneurons receive converging inputs, excitatory from the respiratory premotor neurons of the medulla and inhibitory from areas of the brain stem reticular formation regulating

posture

or muscle tone, the strength

of

each varying with the importance of the muscle’s functions in respiratory and postural or other mechanisms (5, 24, 28). Thus it may be argued that changes in respiratory motoneuronal firing during REM sleep reflect an instantaneous balance of synaptic excitations and inhibitions (as well as disfacilitations and disinhibitions) that have their origins in central sleep-specific and respiratory-specific processes. The construct has not been tested experimentally. Because of the limitations of experiments on humans and unanesthetized behaving animals, we thought there was a need for an animal model in which the assumptions of the construct could be validated. We adapted the acute animal model recently introduced by Morales et al. (22) for the study of postural mechanisms during REM sleep, the decerebrate cat with microinjections of carbachol into the pons. This model was based on several lines of research. 1) The cardinal signs of REM sleep in intact animals are expressed spontaneously in the decerebrate cat and/or can be induced by the intravenous injection of cholinergic drugs (8, 17, 18, 35). 2) Cholinergic drugs injected locally into the dorsal pontine tegmentum induce a state very similar to naturally occurring REM sleep, including electroencephalographic desynchronization, pontogeniculooccipital waves, rapid eye movements, and postural muscle atonia (1, 34). The postural atonia is the most easily obtainable response, consistently found in various experimental situations, and there is little doubt that it corresponds to the motor atonia observed in naturally occurring REM sleep (22; see 7 for a general review). 3) The inhibition of lumbar motoneurons produced in a decerebrate cat after the injection of carbachol into the pons has the same characteristics as those observed during naturally occurring REM sleep in intact animals (22), including state-specific inhibitory postsynaptic potentials with characteristically large amplitudes and

0 1990 the American

Physiological

Society

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 17, 2019.

RESPIRATORY

SUPPRESSION

rapid rates of rise (21). This model seemed suitable for studying, in an acute experiment, the changes in various central respiratory mechanisms that accompany the carbachol-induced atonia. Because we had no information on the respiratory effects of injecting carbachol into the pons of decerebrate cats, we first determined the motor output pattern to selected representative respiratory muscles. As with many of the studies concerned with the brain stem control of breathing, the cats used in these experiments were also paralyzed, vagotomized, and artificially ventilated. This allowed us to study the effects of carbacholinduced atonia in the absence of changes in sensory feedback arising from pulmonary mechanoreceptors and with blood gases kept relatively constant.

AFTER

PONTINE

CARBACHOL

2281

(CWE MA-821).

The moving averages of the four nerves, together with blood pressure, end-tidal COZ, tracheal pressure, and an event marker were continuously monitored throughout the experiment on a chart recorder (Gould TA2000). The raw nerve recordings, event marker, and selected physiological variables were stored on magnetic tape (Hewlett-Packard 3968A, direct current 2,500-Hz bandwidth). Solutions of carbachol (Sigma Chemical) or atropine SO4 (Sigma Chemical), both dissolved in saline at a concentration of 11 mM, were placed in glass micropipettes having an outer tip diameter of 30-40 pm, Pontamine sky blue (2%) was added to the carbachol solution to mark the injection site. The pipettes were positioned stereotaxically in the dorsal pontine tegmentum at coordinates P3.5, H-4.0, and L2.0. This site was selected because of the large body of evidence that showed that METHODS postural atonia can be provoked most effectively by Animal preparation. The results are reported from carbachol injections within this area of the dorsal pontine experiments on nine adult cats of either sex weighing tegmentum (7; see 33 and 34 for other references). The 2.0-4.0 kg. The cats were preanesthetized with ketamine drugs were injected with short-duration pressure pulses (80 mg im) and diazepam (2 mg im) and were given (NeuroPhore BH-2), and the volumes ejected were didexamethasone (2 mg im). After intubation, they were rectly determined by measuring the movement of the anesthetized with halothane and decerebrated at a pre- meniscus with a pocket microscope and reticule. The collicular level. The recordings were begun -9 h after injected volumes of carbachol ranged from 55 to 420 nl, the injection of ketamine-diazepam and 7 h after the corresponding to 0.11-0.84 pg; those for atropine were decerebration. 80-560 nl, corresponding to 0.59-4.1 pg. A femoral artery and vein were catheterized for blood At the end of the experiment, the animal was perfused pressure recording and drug administration, respectively. with a solution of 10% Formalin in saline, and the brain The Cg branch of the phrenic nerve, the main genioglosstem was removed for subsequent sectioning and histosal branch of the hypoglossal nerve, one intercostal nerve logical verification of the injection sites. Serial transverse (Td-J, and a dorsal motor branch of C4 were isolated, cut sections 50 pm thick were cut on a cryostat. Every fourth distally, and desheathed for whole nerve recording. All section that contained the pontamine blue dye mark was nerves were dissected on the right side. The vagus nerves mounted and counterstained with neutral red. The injecwere cut in the midcervical region. The cat was then tion sites were redrawn and projected onto standard cross placed in a stereotaxic head holder (David Kopf) and sections of the brain stem (3). fixed in a spinal frame with a clamp at T2. For pontine Experimental protocol. Only animals that showed injections, the bony tentorium was removed and the dura strong tonic activity in the C4 nerve branch at the beginwas reflected to expose the anterior cerebellar vermis. ning of the recording were used; this provided a reference The animals were paralyzed with gallamine triethiodide signal that postural muscle atonia was produced by the (initial dose of 10 mg iv, supplemented with a continuous carbachol injection. After a suitable control recording infusion of 5 mg kg-’ . h-l) and artificially ventilated at had been obtained, we injected carbachol into the right a rate of 16-2O/min and a volume adequate to maintain pontine tegmentum, aiming for the stereotaxic coordithe end-expiratory COZ level at 3.5-4.5%. O2 was added nates given above. After a variable latent period, this to the inspired gas mixture (50%), and in some experiinjection produced a fall in the activity of Cq and the ments l-3% COZ was added to raise the end-tidal COZ changes in respiratory motor output that are the subjects above 4% and increase the respiratory drive. of this report. After a steady state was achieved, we Tracheal CO2 was continuously monitored with an removed the carbachol pipette and replaced it with a infrared gas analyzer (Datex). Systemic arterial blood pipette filled with the solution of atropine SO*. This pressure and tracheal pressure were measured with pres- pipette was then positioned in the same site as the sure transducers (Statham P23 Gb), and airflow was carbachol injection, and a volume of atropine one to four measured with a pneumotachograph (Fleisch no. 00) and times that of the preceding carbachol was injected. After differential pressure transducer (Grass PT5). The rectal the injection of atropine had produced a recovery of temperature was maintained at 37.5-38.5”C by a servo- neural activity, we repeated the sequence of injections in controlled heating pad. a symmetrical site on the left side. The dissected nerves were placed on bipolar platinum Data analysis, From the moving average of each nerve, electrodes and immersed in mineral oil pools. The neural we measured the amplitude of the inspiratory and/or expiratory activity and the tonic activity during the activities were amplified using conventional techniques (Grass P511 preamplifiers, 30-Hz to 5-kHz bandwidth). control preinjection period and during the period of The nerve signals were full-wave rectified and processed maximum suppression, or recovery, after carbachol and with third-order Paynter filters with 200-ms time con- atropine injections, respectively. The latencies of the stants to obtain moving averages of the neural activity responses to the microinjections were measured for each l

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 17, 2019.

2282

RESPIRATORY

SUPPRESSION

nerve from the onset of the injection

to the onset of the response, the point at which 50% of the total amplitude change had occurred, and the point where the maximum suppression or recovery occurred during the sustained carbachol or atropine effects, respectively. The phasic, or respiratory-related, activity of each neurogram was measured as the difference between the level of the neurogram at the peak amplitude during the burst and the lowest activity level within the same respiratory cycle. The baseline for determination of tonic levels of activity was established from those segments of the recording when complete atunia occurred after a carbachol injection or when there was no spike activity during a part of the respiratory cycle as evidenced from the raw nerve record. From the phrenic neurogram, we measured the respiratory rate, respiratory period (TT, time from the onset of a phrenic burst to the onset of the next burst), and the durations of inspiration (TI, time from the onset of a phrenic burst to its peak) and expiration (TE, time from the peak of one burst to the onset of the next). For all determinations, at least 20 consecutive respiratory cycles in each condition were measured. Measurements of the various parameters of the responses were normalized for each experiment to the mean control values immediately before each carbachol injection. For comparison of the response magnitudes across the nerves studied, the phrenic nerve activity was used as a reference because this was the only nerve that maintained its phasic activity throughout the various tests. For selected experiments, additional analyses were performed with data acquisition and analysis software (EGAA, RC-Electronics) installed in an IBM PC-AT. This system permitted automatic breath-by-breath detection of the onset, offswitch, and peak of the phrenic moving average signal and determination of the respiratory rate and slope of the phrenic neurogram, as well as the peak inspiratory and expiratory amplitudes of the other neurograms. For this analysis, segments of the moving averages covering one or two adjacent tests were sampled at a rate of 200 Hz and stored on hard disk. Results are expressed as means t SE. Statistical analysis was performed using Student’s t tests for paired comparisons. When multiple comparisons were made, a Bonferroni correction was used. Probabilities were considered significant if they were co.05 divided by the number of comparisons. We present statistical comparisons only for the data from the first carbachol injections; after the second injections, the latencies were significantly longer, and some of the response amplitudes were slightly but significantly decreased. RESULTS

Qualitative effects. In all cats, the injection of carbachol caused, after a variable latency, a suppression of activity in all the nerves studied: phrenic, intercostal, hypoglossal, and C4. Figure IA is a polygraph record showing the time course of the suppression of the moving averages of the nerve activities using a slow chart speed. (The actual time of injection is not shown but preceded this segment of the recording by 70 s.) The decreases in activity had relatively similar time courses in all nerves and were

AFTER

PONTINE

CARBACHOL

seen both in the level of tonic activity (if present) and in the amplitude of the respiratory bursts. The changes in respiratory amplitude were accompanied by simultaneous decreases in the respiratory rate and the rate of rise of the phrenic inspiratory burst. Not shown in Fig. 1 is the fact that the arterial blood pressure also fell in a similar manner, in parallel with or slightly after the motoneuronal activity changes. For all animals (n = 9), the mean pressure was 145/87 mmHg during the control period and 118/69 mmHg after carbachol. After the carbachol injection, the neural activity remained depressed for a prolonged period, with a slow spontaneous return of phrenic and intercostal nerve activity noticeable in some animals after a period of -Z10 min. The spontaneous return of t*he hypoglossal and

Cd activities was appreciably slower, usually beginning after ~6-13 min. (In five animals, however, the activity of at least one of these nerves remained maximally suppressed for the entire period of observation, 18-32 min.) Reversal of the carbachol response, however, could be effected by the microinjection of atropine into the site of the carbachol injection (Fig. 1B). The pattern of recovery of the amplitude of firing, respiratory rate, and blood pressure mirrored that of the carbachol response, except that it had a slower time course (possibly related to the higher

molecular

weight

of atropine

with

conse-

quently slower diffusion or to receptor occupancy). Also seen in Fig. 1 are examples of two other phenomena that were observed in some of the experiments. One is that the injection of atropine sometimes resulted in a level of activity that was higher than the control precarbachol level. This was almost always seen in the phrenic and intercostal nerves, but in Fig. 1 it can also be seen in the C4 nerve branch. The other phenomenon, seen in Fig. 1 during the recovery after atropine, was an oscillation of activity in some animals that could be observed in all the nerves simultaneously. Although they are not shown, similar blood pressure changes followed the neural changes. Such oscillations were often observed in the control state, as well as during the recovery, but never during the maximum suppression of activity after carbachol (see Fig. 9). Once reversal was achieved after the injection of atropine, carbachol was injected into a similar site on the left side. This second carbachol injection was made -30 min after the atropine microinjections and always produced a suppression of activity that was qualitatively similar to the first, although slightly weaker (see below). This suppression of atropine.

was reversed

by a second microinjection

A second example of the effects of carbachol and atropine is given in Fig. 2, which shows, for a different animal and with a faster chart speed, the effects of carbachol and atropine on the unprocessed nerve activities. In this animal, the intercostal nerve fired during expiration as well as inspiration, and the expiratory activity was abolished after the injection of carbachol (Fig. 2, middle). Also clearly seen in Fig. 2 are the changes in respiratory rate, TI, and TE and the decrease in the inspiratory ramp activity. Magnitude of the response. We examined two parameters of the respiratory

neural

activity

that are related

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 17, 2019.

RESPIRATORY

SUPPRESSION

AFTER

PONTINE

2283

CARBACHOL

FIG. 1. Polygraph record showing transition from control state to carbachol-induced atonia (A) and reversal after atropine (B). Traces show moving average of neural activity of phrenic (PHR), intercostal (IC) , hypoglossal (HYPO), and C4 nerves. Carbachol (0.17 pg) was injected 70 s before beginning of record shown in A, and atropine (1.2 pg) was injected 17 min before beginning of record in B. Note similar time courses of changes in all nerves. For further explanation, see text,

PHR

HYPO r\, %A-

-

1 min CONTROL

CARBACHOL

ATROPINE

200 3

PHR IC HYPO c4

0

10 s 2. Effects of carbachol (0.11 pg) and atropine (0.16 pg) injections on unprocessed neural activities. Records were taken during steady-state conditions before and after carbachol injection and after subsequent reversal of carbachol response by atropine. Abbreviations as in Fig. 1. Note bursts of phasic expiratory activity (between phrenic bursts) in intercostal nerve recording during control period and after atropine and its absence during carbachol-induced suppression. There is a profound carbachol-induced prolongation of respiratory cycle. FIG.

to respiratory drive: the peak amplitude of the respiratory burst and the rate of rise of the phrenic neurogram. In all animals, after the injection of carbachol, there was a large decrease in the amplitude of the moving average of the neural activity, and the pattern of this decrease was consistent. In the intercostal and hypoglossal nerves, any tonic activity previously present was abolished. In all respiratory nerves, the peak amplitude of the phasic activity was suppressed. Figure 3 shows the mean suppression of respiratory activity caused by the first injections of carbachol (mean volume 140 t 30 nl). The magnitude of the suppression was: phrenic C inspiratory intercostal < hypoglossal < expiratory intercostal. Also shown is the suppression of C4 activity. All responses are normalized to the amplitude of the neural activity of each nerve during the control period (lOO%, dashed horizontal line). Notably, the subsequent microinjection

PHR

tC,I

HYPO

IC,E

c4

3. Mean levels of activity after carbachol-induced suppression (closed bars) and after atropine reversal (hatched bars) in nerves studied. All activity levels are normalized to their control pre-carbachol level (100%). Bars for different respiratory nerves are shown in rank order according to magnitude of suppression of activity after carbachol injection. Data are means t SE for only 1st injections in 7 experiments for which there were good recordings for 3 respiratory nerves throughout both tests. (In 2 cats there was no expiratory activity in intercostal nerve, and in another 2 cats Cq recording was not suitable for analysis ftir technical reasons.) *P < 0.005, **p C 0.001 vs. control pre-carbachol levels of activity. Note large differences in amount of suppression among respiratory nerves. IC,I, inspiratory intercostal activity; IC,E, expiratory intercostal activity. FIG.

of atropine restored the activity in the respiratory nerves to levels that exceeded the control pre-carbachol levels. Although this was observed in most experiments, the increases were not statistically significant. The consistent pattern of suppression of activity is illustrated in Fig. 4, which shows the data for all the carbachol injections (16 injections in 9 cats). Here the relative depressions of the individual nerves are shown for each injection of carbachol. The injections are arranged in rank order according to the magnitude of the effect on phrenic nerve activity. The relative sensitivities of the different nerves were always (except for test 9 where the hypoglossal response was depressed more than

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 17, 2019.

RESPIRATORY

SIJPPRESSION

AFTER

PONTINE

CARBACHOL

RBACHOL

I 2 3 4 5 6 7 8 9 IO 11-12-13 14 INJECTION

I5 16

NUMBER

FIG. 4. Pattern of carbachol-induced suppression of activities of different respiratory nerves. Magnitude of each nerve activity is given as percentage of control pre-carbachol state. Data from both 1st and 2nd injections are included (16 injections in 9 cats). Injections are arranged in rank order according to magnitude of effect on phrenic nerve activity. In almost all cases relative suppression of different nerves followed the same order (phrenic 5 inspiratory intercostal 5 hypoglossal 5 expiratory intercostal) regardless of absolute amount of suppression observed in an individual test. (In some experiments, activity of a particular nerve could not be adequately measured for technical reasons.) l , Phrenic activity; H, inspiratory intercostal activity; A, hypoglossal activity; q I, expiratory intercostal activity.

the expiratory intercostal) preserved and followed the order given above: phrenic s inspiratory intercostal 5 hypoglossal 5 expiratory intercostal. The variable degree of suppression of activity in different experiments was not related to the dose of carbachol used. After the second microinjection of carbachol (into the contralateral pons, mean volume 200 t 45 nl), a full response that was qualitatively identical to the response evoked by the first injection was always seen. The relative sensitivities of the respiratory nerve activities followed the pattern given above. Compared with the first injection, the magnitude of the suppression was not significantly different for those nerves that usually be= came completely atonic (Cd, hypoglossal, and expiratory intercostal). In contrast, the suppression of the nerves that were least affected by the first injection (phrenic and inspiratory intercostal) was somewhat weaker after the second injection (-20% less affected). For the phrenic nerve, this difference was significant. These differences can, at least in part, be explained by the fact that the second injection site tended to be more medial than the first and not in the optimum site for eliciting atonia (see injection sites in Fig. 8). The second measure of respiratory drive, the rate of rise of the phrenic neurugram, was also suppressed after carbachol. Figure 5A shows for one animal the large decrease in the slope of the rising phase of the neurogram after carbachol. For each condition, control and postcarbachol, three superimposed traces are shown. Figure 5B shows the percent change, compared with control, for all the animals after carbachol and then after atropine. Respiratory timing. After an injection of carbachol, the respiratory frequency decreased with a time course that naralleled that for the change in amnlitude of the nhrenic

B 150f 2 0 2

z 0 0

8 -E 0 J cl3

N.S. lOOF

50.

OCONTROL

CARBACHOL

ATROPINE

FIG. 5. Effects of carbachol on rate of rise of phrenic neurogram. A: individual neurograms from 1 expt, 3 overlapping traces during control

period and during suppression after carbachol injection. B: mean (& SE) change of rate of rise of phrenic neurogram for all 1st injections (72 = 9) vs. control (open bar; 100%). Closed bar, after carbachol; hatched bar, after atropine reversal. Differences were tested in comparison with control conditions.

burst. This decrease in frequency was due to increases in both TI and TE. The la.tter change was greater, so TI/ TT, a measure of the duty cycle, fell. Representative time courses for changes in frequency and TI/TT are shown in Fig. 6A. All timing changes could be reversed by atropine. The average percent change in respiratory rate for all animals after carbachol (first injections only) is shown in Fig. 6B. The average control frequency was 21.7 t 2.8/min; after carbachol it fell to 14.9 t 1.6, and after atropine it returned to 20.9 t 2.9. The change in TT that occurred after carbachol was due to increases in both TI (control 1.3 t 0.3 s, post-carbachol 1.6 t 0.2, post-atropine 1.3 -t 0.2) and TE (control 2.1 t 0.5 s, postcarbachol 3.0 t 0.5, post-atropine 2.1 t 0.4). TI/TT decreased after carbachol from 0.40 & 0.02 to 0.36 t 0.02; after atropine it returned to 0.39 t 0.02. All the carbachol-induced changes in respiratory timing were significant. After a carbachol injection, breathing never became irregular in a manner even remotely similar to that seen in naturally occurring REM sleep. (Consistent with this observation, phasic twitchlike activity was rarely seen in the Cq neural activity.) We never observed periods of tachvnnea.

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 17, 2019.

RESPIRATORY

SUPPRESSION

A 35 r

03. 1

I

I

f

I

t

CARBACHOL 0.56 pg

B

I 30 s/div.

.

NS l

.

AFTER

FIG.

breath

6. Effect of carbachol

on respiratory

time

of respiratory

course

of change

timing. A: breath-byrate (top) and TI/TT

(botk~~) after carbachol injection in a single expt. B: mean (2 SE) change of respiratory rate for all 1st injections (n = 9) vs. control (open bar; 100%). Closed bar, after carbachol; hatched bar, after atropine reversal. Differences were tested in comparison with control conditions.

Latency of the effects. The latency to the onset of the effects of carbachol was highly variable, ranging for Cd from 13 s to 9 min, with a mean of 170 t 64 s. In general, larger doses of carbachol had shorter latencies. Because for the least-affected nerves it was difficult to determine the exact time of the onset and maximal magnitude of the response due to a very gradual change in activity or the presence of spontaneous variations (see phrenic and intercostal activities in Fig. l), we chose to evaluate the latency from the time of injection to the time when 50% of the final response occurred. Regardless of which latency measure was used, its pattern for the different nerves consistently indicated that the nerves that were least suppressed, phrenic and inspiratory intercostal, had longer latencies than those most suppressed, hypoglossal

CARBACHOL

2285

and expiratory intercostal. Because of the large scatter in the latency measurements among the different experiments, we normalized them by using the phrenic latency as a reference to reveal the average response latency pattern. The mean relative 50% response latencies after a carbachol injection of the various nerves are shown in Fig. 7A. After the atropine injection, the latencies to 50% recovery were on the average about three times longer, and the recovery of activity tended to follow a reverse order; that is, the hypoglossal and expiratory intercostal were the last to reappear (Fig. 7B). The 50% response latencies after the second carbachol injection followed the same order, although their absolute values were two to three times longer than after the first injection. This difference might again be related to the fact that the second injections tended to be closer to the midline. Only the hypoglossal latency after the first carbachol injection was significantly different from the phrenic latency. Injection site. Although we aimed for one particular stereotaxic site in the dorsal pontine tegmentum, there was some scatter to the exact location of the dye spot, because of the large differences in the size of the cats, and perhaps brain displacements resulting from the decerebration. Figure 8 shows the distribution of the injection sites projected onto standard drawings of a coronal (A) and horizontal (B) section of the pons (2). To simplify the drawings, the sites of both the first and second injections are shown on the same side and are distinguished by different symbols. The most effective sites, those with latencies to the onset of atonia ~5 min,

n E 200 LJJ c!J Z 4x 0

CONTROL CARBACHOL ATROPINE

PONTINE

100

8

0 m

0

A

llil B

0

I-

.L

>

g 200 W Ia -J W

100

> I6 -I w cc

O

PHR

IC,I HYPO IC,E

c4

7. Relative latencies (mean t SE) for activity of each nerve to reach 50% of maximum response after injection of carbachol (A) and atropine (B). Each latency is normalized to that of phrenic nerve (100%). Absolute value of latencies for atropine effect was ~3.0 times longer than for carbachol. *‘la C 0.01 vs. phrenic response latency. Data are from the same set of tests as in Fig. 3. FIG.

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 17, 2019.

2286

O -2

RESPIRATORY

t L

SUPPRESSION

IP4.01

. .

-4

. c

-6

t .

-8

. .

-10

L

L

’

i

’

f

1”

1

’

1

10 8 6 4 2 0 FIG. 8. Locations of carbachol injection sites projected onto standard brain sections from Berman’s atlas (3). A : drawing of coronal section through pons at P4.0; B: drawing of horizontal section at H-3.8. Centers of injection sites are projected onto plane of each drawing. (Two most rostra1 injection sites shown in horizontal drawing were not included on coronal section because they were too far craniad to be accurately projected.) Circles, 1st injections; squares, 2nd injections; closed symbols, phrenic 50% response latency 5 min. Coordinates in millimeters. BC, brachium conjunctivum; BP, brachium pontis; CAE, nucleus coeruleus; LLV, ventral nucleus of lateral lemniscus; MLB, medial longitudinal bundle; P, pyramidal tract; SOM, medial nucleus of superior olive; 5M, motor trigeminal nucleus; 5ST, spinal trigeminal tract; 8N, statoacoustic nerve.

were located in the pontine reticular formation, ventral or ventromedial to the locus coeruleus (cf. 9, 34). Both first and second injections were capable of producing responses with short latencies if they were placed within the effective region. DISCUSSION

We studied the changes in respiratory motor output in a decerebrate cat during pharmacologically induced postural muscle atonia similar to that observed in REM sleep. To provoke the atonia, we injected carbachol into

AFTER

PONTINE

CARBACHOL

the dorsal pontine tegmentum (1, 9, 22,27, 34, 35). After the carbachol injection, we observed a profound suppression of activity in the four motor nerves studied (phrenic, intercostal, hypoglossal, and Cd) and a decrease in respiratory rate. This was a consistent stereotyped response in which the magnitudes of the suppression of the respiratory nerves were always phrenic 5 inspiratory intercostal s hypoglossal 5 expiratory intercostal. Thus, in this reduced preparation (decerebrate, vagotomized, paralyzed, and artificially ventilated), we have unmasked a powerful inhibitory and/or disfacilitatory input from the pons to the motoneuronal pools of various respiratory muscles and to the central mechanisms that control respiratory timing. In a different model of REM sleep atonia, Kawahara and co-workers (10-13) used electrical stimulation to provoke the atonia, with results on respiratory muscle activity similar to ours (see below). Although the carbachol-induced inhibition of lumbar motoneurons appears to be similar to that of natural REM sleep (22), the tonic depressant effects on respiration that we observed were unlike the changes that occur during natural REM sleep in that they were very strong and not accompanied by irregularities in the respiratory pattern (26, 28). Nevertheless, the mechanisms underlying the respiratory and postural effects are likely to be common to both systems, because the effects are produced with similar timing through cholinoceptive neurons located in an area of the pons that has been implicated in the generation of REM sleep and REM sleep atonia (see 33 for a review). Thus our results demonstrate the potential for respiratory inhibition in normal animals, a potential that is suppressed or overridden by excitatory inputs during natural REM sleep, whereas it is unmasked in our preparation. This potentially powerful inhibitory mechanism may intermittently dominate during natural REM sleep in the form of episodic flurries and, under some conditions, could result in sleep-disordered breathing. It is beyond the scope of this paper to discuss the relationship of the sleep state induced by the microinjection of cholinergic agonists into the pons of intact unanesthetized behaving cats to naturally occurring REM sleep. However, there is extensive documentation that many aspects of the sleeplike state evoked by carbachol, certainly the atonia, are similar to REM sleep (see 1, 7, and 33 for reviews). It also appears that, in intact animals, small injections of carbachol into the pons often result in only a partially developed REM sleep-like state; in this state, phasic events are scarce or absent, but the motor atonia is complete (9, 34). The pontine sites from which this incomplete REM sleep-like condition is elicited overlap with those from which a fully expressed state can be evoked, suggesting that variables other than the site of cholinergic stimulation may be involved. Consequently, we think that the tonic respiratory suppression that we observed may represent a subset of the events seen in naturally occurring REM sleep and, as such, can be related to previous studies of respiration during carbachol-induced REM sleep in intact cats (15, 16). In intact cats, during the carbachol-induced REM sleep-like state, breathing is characterized by a large decrease in respiratory frequency (compared with both

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 17, 2019.

RESPIRATORY

SUPPRESSION

wakefulness and natural REM sleep), with increases in both TI and TE (15). This is similar to what we found in decerebrate animals. The values for tidal volume and mean inspiratory flow (an expression of inspiratory drive) were similarly slightly reduced during carbacholinduced and natural REM sleep in normal cats with intact chemical feedback, but not nearly to the extent that we observed in the decerebrate cat with controlled ventilation. The posterior cricoarytenoid muscle, an upper airway abductor, becomes hypotonic during natural REM sleep and to an even larger degree during carbachol-induced REM sleep (16); again, this is similar to our finding of a large suppression of hypoglossal nerve activity. There may be several reasons why we see such a profound suppression of respiratory nerve activity and respiratory rate that is unlike the respiratory changes seen in natural REM sleep. If, as Orem (23, 24) postulates, there are several mutually opposing excitatory and inhibitory influences impinging on the various components of the respiratory control system during REM sleep, our decerebrate preparation may have removed many of the excitatory inputs or stabilizing circuits. Reticular formation effects could then be expressed unchecked, especially after the injection of carbachol, which seems to trigger the entire pontine atonia-related network (7, 33). In this respect, our preparation may be analogous to the release of exaggerated stretch reflexes seen in the midcollicular decerebrate cat. Particularly important, however, the vagotomy and controlled ventilation removed many of the afferent feedback influences, resulting from lung inflation and changes in blood gas tension, that serve to maintain ventilation during perturbations. In an intact animal, the large changes in ventilation that would accompany decreases in tidal volume and respiratory frequency would cause an increase

A

HYPO c4

-mm

lc--zi 9. Polygraph records (moving averages of nerve activities) showing spontaneous oscillations in activity during control period (A) and transition period after atropine injection (B). Traces as in Fig. 1. Dashed lines at beginning of traces in A indicate levels of zero activity for individual nerves. Carbachol was injected in time between 2 records and resulted in suppression of all activities as described in RESULTS. It also abolished oscillations seen in these records. Atropine (0.6 pg) was injected 3.5 min before beginning of record in B. For further explanation, see text. FIG.

AFTER

PONTINE

CARBACHOL

2287

in Pcoz and provoke a compensatory increase in respiratory drive. Regarding the changes in blood pressure that were parallel to the motor atonia, the smallest decrease in systolic pressure was 10 mmHg and the largest was 45 mmHg. No systematic differences in response characteristics were observed that could be ascribed to differences in the fall in blood pressure. If anything, a fall in blood pressure, by itself, would be expected to cause a rise in the magnitude of the phrenic discharge and an increase in respiratory rate (2). The gallamine used to paralyze the animal has central actions, and these unspecified anticholinergic effects may have had a role. However, respiratory suppression after the injection of carbachol into the pons was observed in intact (16) and decerebrate spontaneously breathing cats (unpublished data). Recently, Kawahara and co-workers (10-13) proposed another decerebrate cat model to study postural atonia and respiratory muscle activity associated with REM sleep. In this model, postural atonia was produced by electrical stimulation of the dorsal pontine tegmentum along the midline in the medial pons (P4-P7), a locus that contains axons from cells in the rostra1 pontine tegmenturn (11). Stimulation caused a profound suppression of hindlimb extensor tone and a parallel suppression of diaphragmatic (lo), hypoglossal (12), or external intercostal (13) muscle or nerve activities. Kawahara et al. (unpublished observations) report that carbachol injected into the pons also causes a parallel suppression of respiration and postural tone. In contrast to the effects on hindlimb extensor activity, the suppression of diaphragmatic activity was not maintained throughout the course of the stimulation; this was probably due, in part, to an increased PCO~ that must have occurred during the stimulation-elicited depression of breathing in these spontaneously breathing cats (cf. 10). The inspiratory activity of the intercostal muscle was suppressed more than the phrenic activity. When rhythmic diaphragmatic activity was reestablished during the period of stimulation, the respiratory frequency was decreased with a prolongation of both TI and ‘rE. Thus these results are somewhat comparable to our findings that phrenic nerve activity was the least affected of all the nerves and the first to recover and that respiratory rate was significantly reduced. The fact that the effects of cholinergic stimulation in our experiments were more sustained than those of electrical stimulation suggests, however, that the latter may simultaneously recruit some mechanisms that oppose the cholinergic effects. Regardless of the exact mechanism that causes the suppression in different nerves (postsynaptic inhibition, disfacilitation, or some combination of the two), we think that the pontine cholinoceptive processes that initiate the suppression of activity are common to all the nerves because of the parallel time course of the suppression of activity seen after carbachol injections and during the subsequent reversal of the response by atropine. Moreover, the same processes may also be involved in the tonic control of the spontaneous activity of respiratory and postural nerves, as well as the cardiovascular system (cf. 33). This is clearly suggested by the records shown in Fig. 9, in which the tonic nerve activities, the ampli-

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 17, 2019.

2288

RESPIRATORY

SUPPRESSION

tudes of the respiratory bursts, respiratory rate, and blood pressure (not shown) all exhibited large spontaneous variations with identical time courses during the control (A) and recovery (B) periods. Five of the nine animals displayed pronounced variability of their respiratory, cardiovascular, and somatomotor drive at some stage of the experiment. This tight coupling between the responses of C4 and the respiratory nerves recorded from three very distant levels of the neuraxis is consistent with the assumption that the carbachol acts to trigger a reticular “executive” network that causes the atonia. This is supported by the finding that small highly discrete unilateral injections of carbachol, conjugated to latex spheres to minimize diffusion, cause a full-blown REM sleep-like state (27). The fact that there are differences among the nerves in the latency to reach 50% of the response may appear to be in disagreement with the hypothesis of a single process. We think, however, that the differences in latencies are due to differences in the relative strength of the suppression of activity with respect to the activity threshold in different nerves. Those nerves that are strongly suppressed quickly reach atonia; further decrease of the excitability of these motoneurons cannot be assessed from peripheral nerve recordings. The membrane potential of these motoneurons may be further hyperpolarized to a level below the threshold for firing. Thus, recording from peripheral axons may not give the true time course of the suppression. This reasoning can explain why the nerves that were suppressed most quickly were the slowest to recover their activity after atropine injection; the subthreshold changes in membrane potential took a considerable amount of time before the motoneurons reached their firing threshold. To see the complete time course, one would need to record the membrane potential intracellularly. In summary, we have shown that the injection of carbachol into the dorsal pontine tegmentum of decerebrate, paralyzed, and artificially ventilated cats evokes a profound suppression of respiratory motoneuronal activity and respiratory rate concomitant with the postural muscle atonia. This region of the pons has been shown to be critical for the generation of REM sleep and the atonia associated with REM sleep (see 33 for a review), and the injection of carbachol into this area induces a postural muscle atonia that has physiological characteristics similar to those recorded during naturally occurring REM sleep (22). Although the respiratory changes that we observed are unlike those seen during natural REM sleep in many respects, especially those changes associated with phasic REM sleep, we think that our model is a valid one for the study of the mechanisms underlying the alterations in respiratory neuronal activity associated with REM sleep-like atonia or, perhaps, cataplexy. Further study is needed to determine the relationship of the neural structures and pathways activated during carbachol-induced suppression of respiration and those activated during natural REM sleep. We thank Andrew Kim for the histology secretarial support. This study was supported by National

and Rosemarie Heart,

Lung,

Cohen for and Blood

AFTER

PONTINE

CARBACHOL

Institute Specialized Center of Research Grant HL-42236. H. Kimura was on leave from the Dept. of Chest Medicine, School of Medicine, Chiba University, Japan. L. Kubin was on leave from the Dept. of Neurophysiology, Medical Research Center, Polish Academy of Sciences, Warsaw, Poland. Address for reprint requests: R. 0. Davies, Dept. of Animal Biology, School of Veterinary Medicine, University of Pennsylvania, 3800 Spruce St., Rm. 21lE, Philadelphia, PA 19104-6046, Received 11 December

1989; accepted in final form 11 July 1990.

REFERENCES 1. BAGHDOYAN, H. A., R. LYDIC, C. W. CALLAWAY, AND J. A. HOBSON. The carbachol-induced enhancement of desynchronized sleep signs is dose dependent and antagonized by centrally administered atropine. Neuropsychopharmacology 2: 67-79, 1989. 2. BAINTON, C. R., AND P. A. KIRKWOOD. The effect of carbon dioxide on the tonic and the rhythmic discharges of expiratory bulbospinal neurones. J. Physiol. Land. 296: 291-314, 1979. 3. BERMAN, A. L. The Brainstem of the Cat.- A Cytoarchitectonic Atlas With Stereotaxic Coordinates. Madison: Univ. of Wisconsin Press, 1968. 4. CHASE, M. H., P. J. SOJA, AND F. R. MORALES. Evidence that glycine mediates the postsynaptic potentials that inhibit lumbar motoneurons during the atonia of active sleep, J. Neurosci. 9: 743751,1989. 5. DURON, B., AND D. MARLOT.

6.

7.

8.

9.

10.

11.

12.

13.

Intercostal and diaphragmatic electrical activity during wakefulness and sleep in normal unrestrained adult cats. Sleep 3: 269-280, 1980. HENDRICKS, J. C., L. R. KLINE, R. 0. DAVIES, AND A. I, PACK. The effect of dorsolateral pontine lesions on diaphragmatic activity during REMS. J. Appl. Physiol. 68: 14351442, 1990. Hosso~, J. A., R. LYDIC, AND H, A, BAGHDOYAN, Evolving concepts of sleep cycle generation: from brain centers to neuronal populations. Behau. Brain Sci. 9: 371-448, 1986. JOUVET, M. Recherches sur les structures nerveuses et les mecanismes responsables des differentes phases du sommeil physiologique. Arch. ItaL. BioZ. 100: 125-206, 1962. KATAYAMA, Y., D. S. DEWITT, D. P. BECKER, AND R. L. HAYES. Behavioral evidence for a cholinoceptive pontine inhibitory area: descending control of the spinal motor output and sensory input. Brain Res. 296: 241-262, 1984. KAWAHARA, K., Y. NAKAZONO, S. KUMAGAI, Y. YAMAUCHI, AND Y. MIYAMOTO. Parallel suppression of extensor muscle tone and respiration by stimulation of pontine dorsal tegmentum in decerebrate cats. Bruin Res. 473: M-90, 1988. KAWAHARA, K., Y. NAKAZONO, S. KUMAGAI, Y. YAMAUCHI, AND Y. MIYAMOTO. Neuronal origin of parallel suppression of postural tone and respiration elicited by stimulation of midpontine dorsal tegmentum in the decerebrate cat. Brain Res. 474: 403-406,1988. KAWAHARA, K., Y. NAKAZONO, S. KUMAGAI, Y. YAMAUCHI, AND Y. MIYAMOTO. Inhibitory influences on hypoglossal neural activity by stimulation of midpontine dorsal tegmentum in decerebrate cats. Brain Res. 479: 185-189, 1989. KAWAHARA, K., Y. NAKAZONO, AND Y. MIYAMOTU. Depression of diaphragmatic and external intercostal muscle activities elicited by stimulation of midpontine dorsal tegmentum in decerebrate cats.

Bruin Res. 491: 180-184, 1989. 14. KLINE, L. R., J. C. HENDRICKS,

Control of activity

of the diaphragm

R. 0. DAVIES, AND A. I. PACK. in rapid-eye-movement sleep.

J. AppE. Physiol. 61: 1293-1300, 1986. 15. LYDIC, R., AND H. A. BAGHDOYAN. Cholinoceptive

pontine reticular mechanisms cause state-dependent respiratory changes in the cat. Neurosci. L&t. 102: 211-216, 1989. 16. LYDIC, R., H. A. BAGHDOYAN, AND C. W. ZWILLICH. State-dependent hypotonia in posterior cricoarytenoid muscles of the larynx caused by cholinoceptive reticular mechanisms. FASEB J. 3: 1625-1631,1989. 17. MAGHERINI, P. C., 0. POMPEIANO,

AND U. THODEN. Cholinergic mechanisms related to REM sleep. I. Rhythmic activity of the vestibulo-oculomotor system induced by an anticholinesterase in the decerebrate cat. Arch. ItuZ. BioZ. 110: 234-259, 1972. 18. MATSUZAKI, M., Y. OKADA, AND S, SHUTO, Cholinergic agents related to para-sleep state in acute brain stem preparations. Bruin Res. 9: 253-267,

1968.

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 17, 2019.

RESPIRATORY

SUPPRESSION

19. MEGIRIAN, II., C. F. L. HINRICHSEN, AND J. H. SHERREY. Respiratory roles of genioglossus, sternothyroid, and sternohyoid muscles during sleep. Eq. Neuro1. 90: 118-128, 1985. 20. MEGIRIAN, D., M. J. POLLARD, AND J. H. SHERREY. The labile respiratory activity of ribcage muscles of the rat during sleep. J. Physiol. Land. 389: 99-110, 1987. 21. MORALES, F. R., P. BOXER, AND M. H. CHASE. Behavioral statespecific inhibitory postsynaptic potentials impinge on cat lumbar motoneurons during active sleep. Exp. NeuroL. 98: 418-435, 1987. 22. MORALES, F. R., J. K. ENGELHARUT, P. J. SOJA, A. E. PEREDA, AND M. H. CHASE. Motoneuron properties during motor inhibition produced by microinjection of carbachol into the pontine reticular formation of the decerebrate cat. J. Neurophysiol. 57: 1118-1129, 1987. 23. OREM, J. Medullary respiratory neuron activity: relationship to tonic and phasic REM sleep. J. Appl. PhysioZ. 48: 54-65, 1980. 24. OREM, J. Neuronal mechanisms of respiration in REM sleep. Sleep 3: 251-267,198O. 25. PARMEGGIANI, P. L., AND L. SABATTINI. Electromyographic aspects of postural, respiratory and thermoregulatory mechanisms in sleeping cats. Electroencephalogr. Clin. Neurophysiol. 33: 1-13, 1972. 26. PHILLIPSON, E. A., AND G. B~WES. Control of breathing during sleep. In: Handbook of Physiology. The Respiratory System. Control of Breathing. Bethesda, MD: Am. Physiol. Sot., 1986, sect. 3, vol. II, pt. 2, chapt. 19, p. 649-689. 27. QUATTROCHI, J. J., A. N. MAMELAK, R. D. MADISON, J. D. MACKLIS, AND J. A. HOBSON. Mapping neuronal inputs to REM sleep

AFTER

PONTINE

CARBACHOL

2289

induction

sites with carbachol-fluorescent microspheres. Science Wash. DC 245: 984-986,1989. 28. REMMERS, J. E. Control of breathing during sleep. In: Regulation of Breathing, edited by T. F. Hornbein. New York: Dekker, 1981, pt. II, p. 1197-1249. (Lung Biol. Health Dis. Ser.) 29. REMMERS, J. E., D. BARTLETT, JR., AND M. D. PUTNAM. Changes in the respiratory cycle associated with sleep. Respir. Physiol. 28: 227-238,1976. 30. SAUERLAND, E. K., AND R. M. HARPER. The human tongue during sleep: electromyographic activity of the genioglossus muscle. Exp. Neural. 51: 160-170, 1976. 31. SHERREY, J. H., M. J. POLLARD, AND D. MEGIRIAN. Respiratory functions of the inferior pharyngeal constrictor and sternohyoid muscles during sleep. Exp. Neural. 92: 267-277, 1986. 32. SIECK, G. C., R. B. TRELEASE, AND R. M. HARPER. Sleep influences on diaphragmatic motor unit discharge. Exp. Neural. 85: 316-335, 1984. 33. SIEGEL, J. M. Brainstem mechanisms generating REM sleep. In: Principles and Practice of Sleep Medicine, edited by M. H. Kryger, T. Roth, and W. C. Dement. Philadelphia, PA: Saunders, 1989, p. 104-120. 34. VANNI-MERCIER, G., K. SAKAI, J. S. LIN, AND M. JOUVET. Mapping of cholinoceptive brainstem structures responsible for the generation of paradoxical sleep in the cat. Arch. Ital. BioZ. 127: 133-164,1989. 35. VILLABLANCA, J. Behavioral and polygraphic study of ‘*sleep” and “wakefulness” in chronic decerebrate cats. Electroencephalogr. C&-z. Neurophysiol. 21: 562-577, 1966.

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 17, 2019.